HEAT3 User

Disclaimer
CRANES SOFTWARE, INC. MAKES NO WARRANTY OR

REPRESENTATIONS IN CONNECTION WITH NISA/HEAT, NISA/EMAG,
NISA/ROTOR, NISA/CIVIL, NISA II, DISPLAY III/IV, NISA-COMPOSITE,
ENDURE, NISAOPT, FEAP, NISA/3D-FLUID, DYMES, OR ANY OTHER
SOFTWARE PROGRAM OF CRANES SOFTWARE, INC. (HEREINAFTER
REFERRED TO AS NISA/DISPLAY SOFTWARE), EXPRESS OR IMPLIED,
INCLUDING ANY IMPLIED WARRANTY OF MERCHANTABILITY OR
FITNESS FOR A PARTICULAR PURPOSE. CRANES SOFTWARE, INC.
MAKES NO WARRANTY AND ASSUMES NO LIABILITY FOR THE NISA/
DISPLAY SOFTWARE. CRANES SOFTWARE, INC. ASSUMES NO
RESPONSIBILITY FOR THE USE OF THE PROGRAMS OR FOR THE
ACCURACY OR VALIDITY OF ANY RESULTS OBTAINED FROM THE
NISA/DISPLAY SOFTWARE. CRANES SOFTWARE, INC. SHALL NOT BE
LIABLE FOR LOSS OF PROFIT, LOSS OF BUSINESS, OR OTHER
FINANCIAL LOSS WHICH MAY BE CAUSED DIRECTLY OR INDIRECTLY
BY THE NISA/DISPLAY SOFTWARE, WHEN USED FOR ANY PURPOSE OR
USE, OR DUE TO ANY DEFECT OR DEFICIENCY THEREIN.
Any questions relating to the use or interpretation of the SOFTWARE or their

operation should be directed to:
Cranes Software, Inc.

1607 E. Big Beaver Road, Suite 250
Troy, MI, 48083, USA
Tel: (248) 689-0077
Fax: (248) 689-7479
NOTICE FOR U.S. GOVERNMENT USERS ONLY: RESTRICTED RIGHTS

LEGEND.
Use, duplication or disclosure by the Government is subject to restrictions as set

forth in subparagraph (c)(1)(ii) of the Rights in Technical Data and Computer
Software clause at DFARS 252.227-7013.
NISA, DISPLAY III/IV, ENDURE, NISAOPT, FEAP, NISA/3D-FLUID, NISA/
EMAG, NISA/HEAT, NISA/ROTOR, NISA/CIVIL and DYMES are registered
trademarks of Cranes Software, Inc.
Copyright 2005-2007 by the Cranes Software, Inc. 1607 E. Big Beaver Road,
Suite 250, Troy, Michigan 48083, U.S.A.
All rights reserved. No part of this manual may be reproduced, stored in a retrieval
system, or transmitted in any way without the prior authorization of Cranes
Software, Inc.
1-iii
Table of Contents
Preface - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - x
Version 15.0 New Features - - - - - - - - - - - - - - - - - - - xii
1 Introduction - - - - - - - - - - - - - - - - - - - - - - - - - - 1.1
1.1 General Description - - - - - - - - - - - - - - - - - - - - 1.1
1.1.1 NISA/HEAT Capabilities - - - - - - - - - - - - - - - 1.1
1.2 NISA Suite of FEA Software - - - - - - - - - - - - - - 1.3
1.2.1 NISA Softwares - - - - - - - - - - - - - - - - - - - - - 1.3
1.2.2 NISA Documentations - - - - - - - - - - - - - - - - 1.8
2 Theoretical Overview - - - - - - - - - - - - - - - - - - - - 2.1
2.1 Introduction - - - - - - - - - - - - - - - - - - - - - - - - - - 2.1
2.2 Governing Equations - - - - - - - - - - - - - - - - - - - - 2.1
2.3 Boundary Conditions - - - - - - - - - - - - - - - - - - - 2.6
2.4 Finite Element Formulation - - - - - - - - - - - - - - - 2.8
2.4.1 Galerkin Method - - - - - - - - - - - - - - - - - - - - - 2.8
2.4.2 Element Shape Function - - - - - - - - - - - - - - - - 2.9
2.4.3 Finite Element Formulation of Energy
Equation - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - 2.15
2.5 Surface Radiation Heat Transfer - - - - - - - - - - - 2.17
2.5.1 Introduction - - - - - - - - - - - - - - - - - - - - - - - 2.17
2.5.2 Governing Equations - - - - - - - - - - - - - - - - - 2.17
2.5.3 Computation of View Factors - - - - - - - - - - - 2.20
2.5.4 Solution Procedure - - - - - - - - - - - - - - - - - - 2.21
2.5.5 Using Artificial Convection Boundary
Condition to Simulate Surface Radiation - - - - - - - - 2.24
2.5.6 Special Cases - - - - - - - - - - - - - - - - - - - - - - 2.25
iv
2.6 Phase Change Effects - - - - - - - - - - - - - - - - - - - 2.30
2.7 Convection Link - - - - - - - - - - - - - - - - - - - - - - 2.32
2.8 Radiation Link - - - - - - - - - - - - - - - - - - - - - - - 2.32
2.9 Beam Element - - - - - - - - - - - - - - - - - - - - - - - 2.34
2.10 Heat Transfer in Composites - - - - - - - - - - - - - 2.35
2.11 References - - - - - - - - - - - - - - - - - - - - - - - - - 2.40
3 NISA/HEAT Solution Technique - - - - - - - - - - - - 3.1
3.1 Introduction - - - - - - - - - - - - - - - - - - - - - - - - - - 3.1
3.2 Frontal Solver - - - - - - - - - - - - - - - - - - - - - - - - - 3.1
3.3 Iterative Solver - - - - - - - - - - - - - - - - - - - - - - - - 3.7
3.4 Solution Errors - - - - - - - - - - - - - - - - - - - - - - - 3.12
3.5 NISA/HEAT Files - - - - - - - - - - - - - - - - - - - - - 3.15
4 Element Library - - - - - - - - - - - - - - - - - - - - - - - - 4.1
4.1 Introduction - - - - - - - - - - - - - - - - - - - - - - - - - - 4.1
4.1.1 Element Types - - - - - - - - - - - - - - - - - - - - - - 4.1
4.1.2 Element Input - - - - - - - - - - - - - - - - - - - - - - 4.5
4.2 2-D Element (NKTP = 2) - - - - - - - - - - - - - - - - - 4.7
4.3 Axisymmetric Element (NKTP = 3) - - - - - - - - - - 4.9
4.4 3-D Element (NKTP = 4) - - - - - - - - - - - - - - - - 4.13
4.5 3-D Thin Shell Element for Heat Transfer
(NKTP=20) - - - - - - - - - - - - - - - - - - - - - - - - - - - - 4.17
4.6 3-D Laminated Composite Shell for heat transfer
(NKTP = 32) - - - - - - - - - - - - - - - - - - - - - - - - - - - 4.21
4.7 3-D Bar Element for Heat Transfer (NKTP = 12) 4.25
4.8 3-D Convection Link Element (NKTP = 49) - - - 4.27
4.9 3-D Radiation Link Element (NKTP = 50) - - - - 4.29
5 Input Setup - - - - - - - - - - - - - - - - - - - - - - - - - - - 5.1
5.1 Introduction - - - - - - - - - - - - - - - - - - - - - - - - - - 5.1
5.2 Free Format Input Features - - - - - - - - - - - - - - - - 5.5
6 Executive Commands - - - - - - - - - - - - - - - - - - - - 6.1
v
6.1 General Description - - - - - - - - - - - - - - - - - - - - 6.1
6.2 Executive Commands - - - - - - - - - - - - - - - - - - - 6.4
7 Model Data - - - - - - - - - - - - - - - - - - - - - - - - - - - 7.1
7.1 Introduction - - - - - - - - - - - - - - - - - - - - - - - - - - 7.1
7.2 Title - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - 7.4
7.2.1 *TITLE Data Group - Problem Title - - - - - - - - 7.4
7.3 Element Data - - - - - - - - - - - - - - - - - - - - - - - - - 7.5
7.3.1 *ELTYPE Data Group - Element Type Selection
Table - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - 7.5
7.3.2 *ELEMENTS Data Group - Element Definition 7.6
7.3.3 *E1 Data Group - Alternate Form for Element
Definition - - - - - - - - - - - - - - - - - - - - - - - - - - - - - 7.13
7.3.4 *RCTABLE Data Group - Real Constant
Table - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - 7.16
7.3.5 *LAMANGLE Data Group - Rotation Angles
for Laminated Composite Elements - - - - - - - - - - - 7.18
7.3.6 *LAMSEQ data group - Composite lamination
sequence data - - - - - - - - - - - - - - - - - - - - - - - - - - 7.20
7.4 Nodal Data - - - - - - - - - - - - - - - - - - - - - - - - - 7.29
7.4.1 *LCSYSTEM Data Group - Definition of
Local Coordinate Systems - - - - - - - - - - - - - - - - - - 7.29
7.4.2 *NODES Data Group - Nodal Coordinates - - 7.33
7.5 Material Data - - - - - - - - - - - - - - - - - - - - - - - - 7.40
7.5.1 *MATFLUID Data Group - Material Property
Data for Fluid - - - - - - - - - - - - - - - - - - - - - - - - - - 7.40
7.5.2 *PRINTCNTL data group - Selective printout
control - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - 7.43
7.6 Miscellaneous Data - - - - - - - - - - - - - - - - - - - - 7.45
7.6.1 *SETS Data Group - Definition of Sets - - - - - 7.45
7.6.2 *TEMPFN Data Group - Temperature
Dependency Curves - - - - - - - - - - - - - - - - - - - - - - 7.48
7.6.3 *TIMEAMP Data Group - Definition of
Time Amplitude Curves - - - - - - - - - - - - - - - - - - - 7.49
7.6.4 *CPVAR Data Group - Coupled Variable Data 7.53
vi
8 Analysis Data - - - - - - - - - - - - - - - - - - - - - - - - - - 8.1
8.1 Analysis Control Data - - - - - - - - - - - - - - - - - - - 8.3
8.1.1 *FLCNTL Data Group - Fluid Load Case
Control - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - 8.3
8.2 Boundary Conditions - - - - - - - - - - - - - - - - - - - - 8.6
8.2.1 *BCDVAR Data Group - Variable Nodal
Boundary Condition - - - - - - - - - - - - - - - - - - - - - - 8.6
8.2.2 *BCGVAR Data Group - Variable Elemental
Boundary Condition - - - - - - - - - - - - - - - - - - - - - - 8.8
8.2.3 *ICDS Data Group - Specified Dirichlet Type
Initial Conditions - - - - - - - - - - - - - - - - - - - - - - - -8.16
8.2.4 *ICGRAD Data Group - Neumann Type
Initial Conditions - - - - - - - - - - - - - - - - - - - - - - - -8.17
8.2.5 *PCHANGE - - - - - - - - - - - - - - - - - - - - - - -8.19
8.2.6 *STEPSIZE - - - - - - - - - - - - - - - - - - - - - - - -8.19
8.3 Output Control Data - - - - - - - - - - - - - - - - - - - 8.22
8.3.1 *PRINTCNTL data group - Selective
printout control - - - - - - - - - - - - - - - - - - - - - - - - - -8.22
8.4 Data Terminator - - - - - - - - - - - - - - - - - - - - - - 8.24
8.4.1 *ENDDATA Data Group - Input Data
Terminator - - - - - - - - - - - - - - - - - - - - - - - - - - - - -8.24
9 Hints for Using NISA/HEAT - - - - - - - - - - - - - - - 9.1
9.1 Introduction - - - - - - - - - - - - - - - - - - - - - - - - - - 9.1
9.2 Running NISA/HEAT - - - - - - - - - - - - - - - - - - - 9.1
9.2.1 Monitoring Convergence - - - - - - - - - - - - - - - 9.3
9.2.2 Managing Disk Space - - - - - - - - - - - - - - - - - 9.3
9.2.3 Killing a Running NISA/HEAT Job - - - - - - - 9.4
9.3 Modeling Hints - - - - - - - - - - - - - - - - - - - - - - - - 9.5
9.4 Analysis Hints - - - - - - - - - - - - - - - - - - - - - - - - 9.9
A Appendix A: Verification Problem - - - - - - - - - A.1
B NISA/HEAT to NISA/HEAT3 Translator - - - - B.1
C FLUTL: NISA/HEATs Utility Package- - - - - - C.1
C.1 About this Appendix - - - - - - - - - - - - - - - - - - - C.1
vii
C.2 Introduction - - - - - - - - - - - - - - - - - - - - - - - - - C.1
C.3 Basic Features of FLUTL - - - - - - - - - - - - - - - - C.1
C.4 Executing FLUTL - - - - - - - - - - - - - - - - - - - - - C.4
C.5 Menus and Pertinent Notes - - - - - - - - - - - - - - - C.5
C.6 Some Examples of FLUTL Usage - - - - - - - - - C.15
C.6.1 Example #1 - - - - - - - - - - - - - - - - - - - - - - - C.15
C.6.2 Example #2 - - - - - - - - - - - - - - - - - - - - - - - C.15
C.6.3 Printouts for Example #1 and Example #2 - - C.16
D Solving Thermal Stress Problems Using
NISA/HEAT and NISA II- - - - - - - - - - - - - - - - - - - D.1
E System of Units - - - - - - - - - - - - - - - - - - - - - - - - E.1
F Keyword Index - - - - - - - - - - - - - - - - - - - - - - - - F.i
viii
ix
Preface
You have with you the latest version of NISA - A standard in Finite Element
Analysis. NISA is one of the few, commercially available, proven and robust
Finite Element Analysis software that has enjoyed a long-standing presence in
the arena of engineering analysis and design. Today it is the result of more than
three decades of innovation and dedication of highly skilled scientists, technol-
ogy architects and software engineers. As a result, generations of scientists,
engineers and researchers have come to depend on NISA to solve their most
complex engineering problems.
NISA which has a heritage of more than 30 years, changed hands from EMRC,
to Cranes Software, Inc. in July of 2005. Cranes Software, Inc is a wholly
owned subsidiary of Cranes Software International Limited - a global software
products and solutions provider. With this change comes an induction of fresh
talent and resources which is poised to take NISA to a new level in the world of
FEA. NISA Version 15.0 is the achievement of a great development team
which has worked rigorously for the past year in accordance to the best in breed
software development life cycle management practices. As manuals are very
important to us, a lot of thought has gone into the design and content of the
manuals. These have been totally revamped as per the new features and look of
the product. New additions to the various modules have been consolidated and
presented in an integrated manner.
The NISA Shell is now more appropriately called Application Launcher. The
new application launcher makes selecting modules, input/ output files, analysis
type and various CAD/FEA translators a simple task. Significant updates and
error correction have been made to individual modules of the NISA suite of
programs.
xi
Version 15.0 New Features
NISA II: Improved iterative and sparse matrix equation solvers; end release for
pipe and elbow element; General spring element (NKTP=38) upgraded to a
general spring and damper element; facility to input user element stiffness,
mass, and damping matrices; rigid link forces output for linear transient
dynamic analysis; ability to post process larger problems involving multiple
load cases.
ADVANCED DYNAMICS: Multiple Support Excitation, for shock spectrum

analysis to handle non-uniform support excitations. This feature also includes
seven modal summation rules namely ABS, SRSS, CQC, Grouping Method,
Ten percent Method, Double Sum Method, and with & without Missing Mass;
Centroidal Stresses, Stress Resultants at Element Centroid and Base Shear
computation for Shock Spectrum Analysis.
DYSPAN: Spectrum compatible Power Spectral Density generation.
ENDURE: Improved functionality of features in Shell; expansion of crack

propagation configurations; Automatic identification of EDI path; Automatic
mesh generation for plate, pipe, and elbow with different types of cracks; and
fatigue initiation theory based on the MANSON approach.
DISPLAY III: Pre & post processing support new features of NISA modules.
In addition to this, general features are: Viewing the input function such as time
amplitude, spectra etc. as a graph; Realistic plot for 3-D General Beam, 3-D
Straight pipe, 3-D Elbow elements (NKTP=48); Automatic selection of Master
and Slave node for Rigid links; Post processing of Non-linear Beam element
(NKTP 39) stresses; Post processing of 3-D beam results such as - filtering of
the results and report generation, reporting maximum stresses and ASME ratios
at critical points across section; Option to plot the XY points in the graph with-
xii
out the connecting lines; Area under the curve with respect to X-Axis; History
plot for nonlinear results for external results; Reading/ Writing of Multiple
External results for linear and nonlinear static analyses; Viewing results for
complete model using Symmetry for external results; Variation of stress triaxi-
ality along a line of nodes; Crack mouth opening displacement graph; Crack
opening area calculation; facility to integrate fluid quantities such as pressure,
temperature on surface.
DISPLAY IV: Pre & post processing support new features of NISA modules.
Key additional features in DISPLAY IV are: Enhancement of Dialog boxes,
accelerator keys. XP style file open dialog; enhancement of Entity-Status view
in the workspace to facilitate the deletion of a single entity or the entire group
using mouse right click or by using Delete key on the keyboard; a wizard that
helps navigate analysis data.
Up-gradation of translators such as SAT2NEU translator to support ACIS R16

geometry kernel; Solid Works translator S/W 2007; geometric import from
IGES 5.3 version.
EMAG: 3D-Magnetodynamic analysis (harmonics) and 3D Transient Mag-

netic analysis capabilities using magnetic vector potential and electric scalar
potential (with and without massive conductors).
NISA/HEAT3 and 3DFLUID: Heat Flux computation and Printing of Heat

Balance Sheet, Sparse Matrix Equation Solver has been implemented; printing
of Local Reynolds and Peclet numbers.
NISA/CIVIL: Revised code of practice conforming to ACI 318 -2005, BS

8110-1997, BS 5950-1:2000 and LRFD 2002 for concrete and steel designs,
Module to design pipes conforming to ASME-NB, NC & ND codes is now part
of NISA/CIVIL. More emphasis is given to produce good structural design
drawings of RCC slabs, Beams, Columns, Footings, & structural steel draw-
ings, Inclusion of concrete and steel quantities in structural drawings, custom-
ized design report generation in ASCII, MS-Word and MS-Excel formats,
stress resultants, contours for shell elements, reinforcement contouring for shell
elements, combined isolated footings for expansion joint columns, improved
realistic plots, standard animation feature for all Stress Resultants and Eigen
Modes, animation of Color Contours.
xiii
Chapter
1
Introduction
1.1 General Description

NISA/HEAT (Numerically Integrated elements for System Analysis for HEAT
transfer) is a general purpose finite element program to analyze a wide spectrum
of problems encountered in heat transfer. NISA/HEAT has a graphical user
interface called DISPLAY III which can be used to set up a heat transfer problem
and also post-process the results. An overview of NISA/HEAT capabilities is
given below. Highlights of the main features of the pre- and post-processor
DISPLAY III are also included.
1.1.1 NISA/HEAT Capabilities

NISA/HEAT solves energy equation for temperature with appropriate initial and
boundary condition. The energy equation is described in detail in Chapter 2. The
following capabilities are currently supported:
Steady and transient analysis

2D, 3D and axisymmetric domains
Phase change
Orthotropic thermal conductivity
Surface radiation with view factor computation
Volumetric/nodal heat generation
Boundary Conditions
1-1
General Description
a
- Specified temperature ( T )
- Heat flux normal to a surface ( q nc )
- Heat flux along global axes qcx , qcy , qzc

c
- Convection h , Ta

b a
- Radiation , T s

h: heat transfer coefficient
T : ambient temperature
: emmissivity
Ts : temperature of surroundings
a: variation with time is allowed
b: variation with temperature is allowed
c: variation with both time and temperature is allowed
Temperature dependent material properties
- Polynomial
- Tabular
Post-processing with DISPLAY III
- Temperature contour plots
1-2
Introduction
- Temperature vs. time plots
- Temperature vs. distance plots

Post-processing with FLUTL
- Heat flux calculation at boundary faces
- Temperature vs. time data for a node

Output available in text format
- Nodal temperature
- View factors
- Radiation heat flux

Choice of direct (frontal) and iterative (GMRES and conjugate gradient) solv-
ers
1.2 NISA Suite of FEA Software
1.2.1 NISA Softwares

NISA provides an integrated and comprehensive family of general and special
purpose programs for Computer Aided Engineering (CAE). The NISA suite of
FEA software covers a wide spectrum of engineering applications, e.g., linear and
nonlinear structural and heat transfer analysis, structural and shape optimization,
fatigue and fracture analysis, fluid flow analysis and printed circuit board stress
and heat transfer analysis. All analysis programs are directly interfaced with the
DISPLAY program for pre- and post-processing. A brief description of the NISA
suite of FEA software is given below. Complete description of these programs is
available in their pertinent users manuals listed in Section 1.2.2.
NISA
This is the parent program of the NISA family. It is a general purpose finite
element analysis program to which all other programs in the NISA family are
1-3
NISA Suite of FEA Software
interfaced. NISA capabilities and input specification are given in this manual.
Other NISA related documentations are listed in Section 1.2.2.
DISPLAY
A 3-D interactive color graphics finite element pre- and post-processing program.
Highlights of the main capabilities are given in Display III/IV Users Manual.
COMSYN
A free interface component mode synthesis program to compute natural
frequencies and mode shapes of a structure from NISA eigenvalue analyses of its
components. As a special feature, this program can also compute natural
frequencies and mode shapes of cyclic symmetric structures given the results of
NISA eigenvalue analysis of a single sector.
STROPT
A general purpose finite element structural optimization program employing the
state-of-the-art optimization techniques including design sensitivity computation.
The optimization process involves appropriately re-sizing the cross-sectional
parameters (design variables) while simultaneously satisfying prescribed limits
(constraints) on the structural response to minimize material volume, mass or
weight. Constraints include displacements, stresses and natural frequencies in
addition to limits on the design variables.
The design variables may be fixed (constant valued) or free (variable valued), and
they may be grouped and linked in many possible ways. This facility is especially
useful for uniformity of design in certain regions of the structure, for creating a
symmetrical design under unsymmetric loading conditions, and for satisfying
certain design specifications. Optimization may be performed for multiple loading
conditions.
SHAPE
A finite element program for shape optimum design of structures. SHAPE adopts a
novel approach for the shape optimization of 2-D and 3-D continuum structures.
The novelty lies in that the design variation is not limited to the original
boundaries. Instead, new boundaries may be created by removal of material from
1-4
Introduction
inside the structure. There is no need to model the boundaries by parametric

curves. In addition, symmetry conditions may be fully exploited in modeling, and
major design changes without the need for mesh refinement can be accomplished.
Shapes can be derived for new applications by starting from a very general model.
Various displacement and stress constraints may be specified for optimization

under multiple loading conditions. Furthermore, parts of the original design may be
frozen by the user in order to satisfy specific manufacturing requirements.
SHAPE uses NISA as the analysis module and accepts the same NISA input file.
Specific optimization information, e.g., constraints, number of iterations, frozen
regions, are read from a separate file. For each new design, a new NISA input file
and an updated optimization input file are generated. In addition, SHAPE also
provides a NISA input file with smoothed boundaries. Furthermore, a separate file
for design history is output, to enable tracing of the changes in design. Each
improvement in the design may be plotted and postprocessed in DISPLAY.
SECOPT
A program for optimizing the area of beam cross-section under a set of constraints
on section properties as well as other constraints imposed by code requirements.
Additional constraints may be imposed to control local buckling of webs or flanges
and to satisfy other practical considerations. The section properties of the optimum
section may be written to a file in NISA input format.
ENDURE
A program for analyzing the fatigue performance of engineering structures.

Durability studies including crack initiation as well as crack propagation can be
performed. Both stress-life and strain-life approaches are available for crack
initiation studies. Damage tolerance analysis can be carried out conveniently with
the availability of built-in crack configurations and crack growth models. Crack
growth models include: Paris, Foreman, Collipriest, Walker and Elber. Fracture
parameters include stress intensity factor and J-integral.
Single and multi-channel loading may be used. Both deterministic and probabilistic
description of loading are available. The rainflow method of cycle-counting is used
for the general case of sequential load history. Stress input may be specified
1-5
manually during preliminary design or directly interfaced from NISA static or

dynamic analysis.
Fatigue life contours may be plotted using the postprocessing module of the
DISPLAY program.
FEAP
A program for the analysis of Printed Circuit Boards (PCBs) and electronic
systems. FEAP is completely interfaced with NISA for various types of analysis
including stress and thermal analysis, heat transfer with convective heat flow and
random vibration analysis.
The finite element model of the PCB is created interactively using a large built-in
library of commercial components such as: Dual Inline Packages (DIPs), hybrid
packages, flat packs, Leadless Ceramic Clip Carriers (LCCCs), etc. Direct
interface with several CAD geometry data bases is also available.
The Integrated Circuits (ICs) are modeled according to the method by which they
are mounted on the PCBs, either the traditional poke-through attachment or the
more recent surface-mounted technology. Once all components on a PCB are
defined, a finite element mesh is automatically generated for each IC and the entire
board. Appropriate elements, including isotropic or composite elements, are
chosen from the NISA element library.
Other capabilities include: Analysis of several PCBs or an entire Avionics box,

special sub-modeling and automated mesh refinement for an area of interest,
detailed solid model of end pins in addition to prediction of reliability and fatigue
life of any PCB component.
NISA-3D/FLUID
A finite element program for analyzing a wide range of 2-D, axisymmetric and 3-D
fluid flow problems. Analysis capabilities include steady state and transient,
laminar and turbulent, incompressible and compressible and Newtonian and Non-
Newtonian fluid flow. In addition conductive and convective heat transfer
problems may be analyzed.
1-6
Introduction
NISA-3D/FLUID interfaces with DISPLAY for graphical representation of results,

e.g., color contour plots of pressure, velocity, temperature, strain rates, in addition
to stream lines.
DYMES
A general purpose program for mechanical system dynamics to analyze
constrained, multi-body and spatial mechanical systems, in which body elements
are connected through mechanical joints such as spherical, revolute and
translational joints and force elements such as nonlinear springs, dampers and
actuators.
Major areas of applications of the code include machine mechanisms, robotics,

vehicle dynamics, controlled systems and space craft dynamics.
NISA/P-ADAPT
A program for linear, static, structural analysis based on the adaptive p-version of
finite element method. The order of the displacement shape functions can be varied
from one to eight. The model may be constructed by using different shapes of
linear or quadratic elements. The adaptive program may be controlled either
manually or automatically. Certain types of geometric curves/surfaces are
represented exactly in the analysis, thereby increasing the accuracy of the analysis.
Zero and/or non-zero boundary constraints can be imposed on the nodes, edges
and/or faces of a model.
NISA/CIVIL
A finite element program that offers CAD based solutions to a wide variety of
problems encountered in the Analysis and Design of Reinforced Concrete and Steel
Structures. Backed by powerful NISA II Analysis and DISPLAY III/IV - the
graphical Pre and Post processor of NISA family of programs, NISA/CIVIL
provides excellent tools for modeling, associating design information and carry out
design process in Limit state and working stress methodologies of design. Design
results are processed to produce structural engineering drawings in AutoCAD
environment. Equipped with an extremely user friendly GUI and graphic displays,
NISA/CIVIL, presents an elegant platform for analysis and design of different
types of structures encountered in practice.
1-7
NISA/EMAG
A general purpose finite element program used to analyze a wide spectrum of

problems in electrical engineering. In particular NISA/EMAG can be used to
perform 2D and 3D electric and magnetic field analysis in electromagnetic devices.
NISA/HEAT
A finite element program for analyzing a wide range of 2D, axisymmetric and 3D
thermal problems. Analysis capabilities include steady and transient analysis, 2D,
3D and axisymmetric domains, phase change, orthotropic thermal conductivity,
surface radiation with view factor computation, volumetric/nodal heat generation.
In addition conductive and convective heat transfer problems may be analyzed. In
addition heat transfer in laminated composite shells can be modeled.
NISA-3D/Heat interfaces with DISPLAY for graphical representation of results,

e.g., color contour plots temperature.
NISA/AERO
A program for aeroelastic analysis. In the present version of NISA/AERO,

divergence, flutter and aeroelastic response solution techniques (for both lifting
surfaces and panels) at subsonic and supersonic speeds are available.
1.2.2 NISA Documentations

The NISA family of programs are documented in the following manuals:
User and verification manuals
1. NISA User's Manual

2. NISA Verification Manual
3. DISPLAY III/IV User's Manual
4. NISA/HEAT User's Manual
5. NISA/3D-FLUID (Incompressible) User's Manual
6. NISA/3D-FLUID (Compressible) User's Manual
1-8
Introduction
7. NISA/EMAG User's Manual

8. NISA/OPT (Sectional Optimization) User's Manual
9. NISA/OPT (Shape Optimization) User's Manual
10. NISA/OPT (Structural Optimization) User's Manual
11. FEAP User's Manual
12. ENDURE User's Manual
13. ENDURE Examples & Verification Manual
14. NISA/ROTOR User's Manual
15. NISA/DYSPAN User's Manual
16. DYMES User's Manual
17. NISA/AERO User's Manual
18. NISA/CIVIL User's Manual
19. NISA/CIVIL Examples & Verification Manual
Training manuals
1. DISPLAY III Training Manual

2. DISPLAY IV Training Manual
3. NISA/HEAT Training Manual
4. NISA/3D-FLUID Training Manual
5. NISA/EMAG Training Manual
6. FEAP Training Manual
7. NISA/ROTOR Training Manual
8. ENDURE Training Manual
1-9
1-10
Chapter
2
Theoretical Overview
2.1 Introduction
In this chapter, a brief review of the theoretical aspects of NISA/HEAT is given.
Section 2.2 presents the basic equations for heat transfer, namely the energy
equation. Initial and boundary conditions for this equation is discussed in
Section 2.3. Finite element formulation of the energy equation is given in
Section 2.4. Theoretical aspects for surface radiation and phase change are given
in Section 2.5 and Section 2.6, respectively.
2.2 Governing Equations

Consider a three-dimensional body; Figure 2.2.1, of volume V and bounded by a
surface area S. We assume that the material obeys Fourier's law of heat
conduction;
T
q = [ K ] ------ (2.2-1)
x
or in component form:
qx k xx k xy k xz T x

q y = k yx k yy k yz T y (2.2-2)

qz k zx k zy k zz T z
2-1
Governing Equations
where q i is the rate of heat flow conducted per unit area in the direction i, [K] is the
thermal conductivity tensor for the material (Eqn. 2.2-1 assumes anisotropic
material behavior), and T x is the temperature gradient vector in Cartesian
coordinates.
Figure 2.2.1 Three dimensional conduction in a body
2-2
In the material principal directions, Eqn. 2.2-2 reduces to the diagonal form
q1 k 11 0 0 T x 1

q 2 = 0 k 22 0 T x 2 (2.2-3)

q3 0 0 k 33 T x 3
or simply,
T
q n = k n ------ (2.2-4)
n
The thermal conductivity matrix in Eqn. 2.2-2 may be obtained from that in Eqn.
2.2-3 by simple second order tensor transformation as
T
K x, y, z = T K 1, 2, 3 T (2.2-5)
where T is the second order transformation tensor between the two sets of axes
x,y,z and 1,2,3. It is noted that the principle of irreversible thermodynamics shows
that the tensor K has to be symmetric. If the material is homogeneous and isotropic,
then:
k xy = k yz = k zx = 0 , and
(2.2-6)
k xx = k yy = k zz = k
The heat conduction equation or the law of conservation of energy is obtained by

considering the heat flow equilibrium in the interior of the body and it gives:
T T T T
k x ------ + k y ------ + k z ------ + q = c ------
b
(2.2-7)

x x
y y
z z t
In deriving the above equation, it is assumed that the principal material directions
are along the global xyz axes. kx, ky, kz are the same as kxx, kyy, kzz, respectively, qb is
the rate of heat generated per unit volume, t is the time, is the mass density of the
material and c is the specific heat (where in the case of conduction in solids, no
2-3
Governing Equations
distinction is made between the specific heat at constant pressure, c p , and the
specific heat at constant volume, c v ).
Eqn. 2.2-7 is the general equation for heat conduction in solids and it may take
one of the following special cases:
- If the material is isotropic with constant conductivity, k, the equation

reduces to
b 1 T
2T + q k = --- ------ (2.2-8)
t
where = k c is the thermal diffusivity
- If the material is isotropic with constant conductivity, k, and steady state

conditions prevail:
b
2T + q k = 0 (2.2-9)
which is Poissons equation.
- If the material is isotropic with constant conductivity, k, and no heat gen-

b
eration q :
1 T
2T = --- ------ (2.2-10)
t
which is the diffusion equation
- If the material is isotropic, with constant conductivity, k, no heat genera-

b
tion q , and steady state conditions exist:
2T = 0 (2.2-11)
which is Laplaces equation.
In the above equations,
(2.2-12)
= e + e + e
z x y y z z
2-4
where ex, ey, and ez are the unit vectors in x,y and z directions, respectively, and
2 T 2 2
-
2
- + -------
= = -------2- + ------- 2
(2.2-13)
2
x y z
2-5
Boundary Conditions
2.3 Boundary Conditions

Eqn. 2.2-7 may be generally subjected to one or more of the following boundary
conditions (B. C.);
(a) Prescribed temperature or Dirichlet B.C.:
T ( x, t ) = T s ( x, t ) on S 1, S 1 CS (2.3-1)
(b) Prescribed heat flow or flux (Cauchys or Neumann B.C.):
T s
k n ------ =q , S 2 CS (2.3-2)
n s = s 2
(c) No heat flow or adiabatic B.C.:
T
------ =0 , S 3 CS (2.3-3)
n s = s
3
which is the natural boundary condition and is a special case of the Neu-
mann of Cauchy B.C.
(d) Convective heat exchange (Convection B.C.):
When part of the surface of the body, say S c where S c CS , is in con-

tact with a fluid media with a fluid temperature, T f , the rate of heat
exchange between the body and the fluid is given by Fouriers law in the
form
k T
------ = h ( T s T f ) (2.3-4)
n
where T s is the body surface temperature, T f is the fluid temperature, n
is the normal to the surface S c and h is the convective heat transfer coef-
ficient. The convective heat transfer coefficient, h, may be a function of
temperature or time and the fluid temperature, T f , may be a function of
time.
2-6
(e) Radiation boundary condition:
Assuming a body of surface temperature T s , is enclosed within a space

with wall temperature, T e (only part of the body surface, say S r , may be
considered exposed to radiation from another body or surface with wall
temperature T e ). We only consider gray bodies or surfaces for which the
emissive power is independent of the radiation wavelength. Under the
above assumptions, the radiant energy emitted from the surface of the
r
body per unit time and area is given by q :
r 4
q = T s (2.3-5)
where is the emissivity of the surface, is the Stefan-Boltzman con-
stant and Ts is the absolute temperature of the surface. On the other hand,
the radiant energy absorbed by the surface due to the wall temperature
a
T e is given by q :
a 44
q = T e (2.3-6)
where is the absorbtivity of the surface. Owing to Kirchoffs law, = and the
heat flux out of the surface is given by q:
4 4 T (2.3-7)
q = [ T s T e ] = k n ------
n
Note that may be function of temperature, surface finish and view factor between
the two bodies.
It is convienent to put equation (2.3.7) into the following form which is similar to
the convective boundary condition form:
q = h ( Ts Te ) (2.3-8)
where
2 2
h = ( T s + T e ) ( T s + T e ) (2.3-9)
2-7
Finite Element Formulation
is an equivalent radiation coefficient which is strongly dependent on temperature.
2.4 Finite Element Formulation
2.4.1 Galerkin Method

The objective of the finite element method is to approximate the differential
equations by a system of algebraic equations. This is done by discretizing the flow
domain into a number of simply shaped regions called elements. Within each
element, the dependent variables are approximated by simple polynomial
functions. The coefficient of these polynomials are obtained from nodal values of
the dependent variable.
Mathematically, the temperature in an element can be written as [Chung, 1978]:
T
T ( x, t ) = T ( t ) (2.4-1)
where the unknown T is column vector of nodal points and is the column vectors
of the shape functions. Substituting these into the governing equation yields an
equation of the form
f ( , T ) = R (2.4-2)
where R is the residual resulting from the use of the approximations of Eqn. 2.4-1.
The Galerkin form of the method of weighted residuals seeks to reduce the error
(residual) R to zero. This is done by achieving orthogonality between the residual
and weighting functions of the element which is expressed as
E ( f w ) d E ( R w ) d = 0 (2.4-3)
where E is the element domain.
2-8
2.4.2 Element Shape Function

NISA/HEAT contains a library of element types which can be used to solve a wide
range of problems. These element types can be divided into two main categories,
namely two-dimensional (or axisymmetric) elements and three dimensional
elements.
For two-dimensional or axisymmetric problems, the following elements may be

used:
(a) 4 noded isoparametric quadrilateral
(b) 8 noded isoparametric quadrilateral
(c) 3 noded isoparametric triangle
(d) 6 noded isoparametric triangle
For three-dimensional problems, the following elements may be used:
(a) 8 noded isoparametric hexa- (b) 10 noded isoparametic tetra-

hedral hedron
(c) 20 noded isoparametric (d) 4 noded quadrilateral shell
hexahedral
(e) 6 noded isoparametric (f) 8 noded quadrilateral shell
wedge
(g) 5 noded isoparametric (h) 3 noded triangular shell
wedge
(i) 4 noded isoparametric tetra- (j) 6 noded triangular shell
hedron
For each of the element types mentioned above, the shape functions used are given
in terms of the natural coordinates of the element, i.e., and . For a 4 noded
isoparametric quadrilateral, the shape function is defined as
2-9

1 4(1 )(1 )
1 4(1 + )(1 )
= (2.4-4)
1 4(1 + )(1 + )
1 4(1 )(1 + )

An 8 noded isoparametric quadrilateral is given by
1 4(1 ) ( 1 ) ( 1 )

2
1 2(1 )(1 )
1 4(1 + )(1 )( 1)

1 2(1 + )(1 )
2
= (2.4-5)
1 4(1 + )(1 + )( + 1 )
2
1 2(1 )(1 + )
1 4(1 )(1 + )( + 1 )

2
1 2(1 )(1 )
The 3 noded isoparametric triangular element has shape function defined by
L1

= L2 (2.4-6)

L3
where L1, L2, and L3 are the natural coordinates related by L1 + L2 + L3 = 1.
For 6 noded isoparametric triangular element, the shape function is
2-10
L 1 ( 2L 1 1 )

4L 1 L 2

L 2 ( 2L 2 1 )
= (2.4-7)
4L 2 L 3

L 3 ( 2L 3 1 )

4L 3 L 1
An 8 noded isoparametric hexahedral element is defined by

1 8(1 )( 1 )( 1 )
1 8(1 + )(1 )(1 )

1 8(1 + )(1 + )(1 )
1 8(1 )(1 + )(1 )

= (2.4-8)
1 8(1 )(1 )(1 + )

1 8(1 + )(1 )(1 + )

1 8(1 + )(1 + )(1 + )
1 8(1 )(1 + )(1 + )

The 20 noded hexahedral element has a shape function given by
1

= 2 (2.4-9)

3
where
2-11
1 8(1 ) ( 1 ) ( 1 ) ( 2.0 )

2
1 4(1 )(1 )(1 )
1 8 ( 1 + ) ( 1 ) ( 1 ) ( 2.0 + )

2
1 4(1 + )(1 )(1 )
1 =
1 8 ( 1 + ) ( 1 + ) ( 1 ) ( 2.0 + + )
2
1 4(1 )(1 + )(1 )
1 8 ( 1 ) ( 1 + ) ( 1 ) ( 2.0 + )

2
1 4(1 )(1 )(1 )
2

1 4(1 )(1 )(1 )

1 4( 1 + )( 1 )( 1 )
2
2 =
1 4(1 + )(1 + )(1 )
2

1 4(1 )(1 + )(1 )
2

1 8(1 ) ( 1 ) ( 1 + ) ( 2.0 + )

2
1 4(1 )(1 )(1 + )
1 8 ( 1 + ) ( 1 ) ( 1 + ) ( 2.0 + + )

2
1 4(1 + )(1 )(1 + )
3 =
1 8 ( 1 + ) ( 1 + ) ( 1 + ) ( 2.0 + + + )
2
1 4(1 )(1 + )(1 + )
1 8 ( 1 ) ( 1 + ) ( 1 + ) ( 2.0 + + )

2
1 4(1 )(1 )(1 + )
A 6 noded wedge element uses linear shape function defined by
2-12
1 2 L1 ( 1 Z)
1 2 L2 ( 1 Z)
1 2 L3 ( 1 Z)
= (2.4-10)
1 2 L1 ( 1 + Z)
1 2 L2 ( 1 + Z)
1 2 L3 ( 1 + Z)
where (L1, L2, L3, Z) are the natural coordinates for the element and L1 + L2 + L3 = 1.
Also Z lies between -1 and 1.
A 15 noded wedge element has shape functions given by
2-13
2
1 2 L 1 ( 2L 1 1 ) ( 1 Z ) 1 2 L1 ( 1 Z )
2L 1 L 2 ( 1 Z )

2
1 2 L 2 ( 2L 2 1 ) ( 1 Z ) 1 2 L 2 ( 1 Z )

2L 2 L 3 ( 1 Z )
2

1 2 L 3 ( 2L 3 1 ) ( 1 Z ) 1 2 L 3 ( 1 Z )

2L 3 L 1 ( 1 Z )

2
L1 ( 1 Z )

= L2 ( 1 Z )
2
(2.4-11)

L3 ( 1 Z )
2

2
1 2 L 1 ( 2L 1 1 ) ( 1 + Z ) 1 2 L 1 ( 1 Z )

2L 1 L 2 ( 1 + Z )

1 2 L 2 ( 2L 2 1 ) ( 1 + Z ) 1 2 L 2 ( 1 Z )
2

2L 2 L 3 ( 1 + Z )

2
1 2 L 3 ( 2L 3 1 ) ( 1 + Z ) 1 2 L 3 ( 1 Z )

2L 3 L 1 ( 1 + Z )
A 4 noded isoparametric tetrahedron element is defined by
L1

L2
= (2.4-12)
L3

L4
where L1, L2, L3, and L4 are the natural coordinates related by
L1 + L2 + L3 + L4 = 1 .
2-14
A 10 noded isoparametric tetrahedron element is defined by
L 1 ( 2L 1 1 )

4L 1 L 2

L 2 ( 2L 2 1 )

4L 2 L 3

L 3 ( 2L 3 1 )
= (2.4-13)
4L 3 L 1

4L 1 L 4

4L 2 L 4

4L 3 L 4

L 4 ( 2L 4 1 )
For shell elements shape functions are same as those for corresponding 2D
elements.
2.4.3 Finite Element Formulation of Energy Equation

The energy equation can be written as
T- T T T
Cp ----- = ----- kx------ + k y ------ + k z------ + q (2.4-14)
t x x y y z z
If implicit method is used, one can write the above equation as
n 1
TT - T T T
Cp --------------------- = ----- kx------ + k y ------ + k z------ + q
t x x y y z z
2-15
Using Galerkins method weighted residual, the above equation becomes
n1
T ----- T- T k T
C p ------ -----
- k k ------
t x x x y y y z z z
------ d = T
C p ------------- + q d
t

T T T T n1
C p ------- -----
- k ---------- k ----------
t x x x y y y z z z
k ---------- Tj d = C p ------------
T -
t
+ q d

C p T T T
---------- T + kx ------ -
- --------
+ k y ------- --------- + k z ------- --------- T j d
t x x y y z z
(2.4-15)
C
= T
kx ------ n x + ky T ------ z d + ----------p- T T jn 1 d + q d
------ y + k T
x y z z t
The boundary integral term appearing on the right hand side is evaluated based on
the type of boundary condition (i.e., heat flux, convection, radiation, surface
radiation) applied to the boundary surfaces.
Note that in deriving the above equation, vector identity

W A = ( WA ) A W has been used. In addition Green-Gauss theorem
[Kreyzig, 1972] has been used to convert volume integral WA into surface
integral appearing on the right hand side. The above can be written as
[A] {T} = {B} (2.4-16)
In the case of steady state problems, the terms containing t are set to zero.
2-16
2.5 Surface Radiation Heat Transfer
2.5.1 Introduction
Heat transfer by surface radiation is usually neglected or approximations are made
to include its effect on real life problems. Computed temperature distributions may,
therefore, be different than as would be expected. With the availability of more
(and increasing) computing resources, it is becoming possible to account for the
heat transfer via the surface radiation. Among a long list of applications involving
surface radiation, a few important ones relate to cooling of automobile engines, air
conditioning of passenger compartments, heat transfer in furnaces, etc.
Motivated by these applications, NISA/HEAT now allows its users to account for
heat transfer in computational domains by surface radiation in addition to
conduction and convection.
It must be pointed out here that the radiative heat transfer phenomenon is different
from its other two counterparts, i.e., conduction and convection. In the latter two
modes, the heat transfer rate is a function of local temperatures and/or their
gradients. However, in surface radiation heat transfer, all surfaces are connected
with each other in radiation heat transfer sense and heat transfer rate is proportional
to the difference in the fourth power of temperatures.
2.5.2 Governing Equations

The analysis of radiation heat transfer is presently based on the following
assumptions:
(i) the surfaces considered are gray
(ii) the surfaces considered are diffuse
(iii) the surfaces considered are isothermal
The first two assumptions imply a consideration of radiative energy independent of

wavelength and direction of its transmission. In our future versions, some of these
restrictions will be relaxed. The third assumption, however, can be controlled by
2-17
Surface Radiation Heat Transfer
the discretization process (as will be explained later) at an increased computational

cost.
Based on the above assumptions, conservation of radiative energy in an enclosure

leads to the following equation [Siegel and Howell]:
N N
ij 1 j
-----
( ij Fij )Tj
4
---------- F ij q r = (2.5-1)
j j
j
j=1 j=1
For convenience, Equation (2.5.1) may be represented in matrix form as
[R] {qr} = [S] {T} (2.5-2)
where
1 j
R ij = -----ij ------------- F ij (2.5-3)
j
j
and
3
S ij = ( ij F ij )T j (2.5-4)
where ij is the Kronecker Delta, j refers to the surface emissivity of the jth
surface, Fij is a view factor between an ith and jth surface, q rj refers to radiative
flux leaving (+ve) the jth surface, is the Stefan-Boltzman constant, Tj is the
absolute temperature of the jth surface and N is the total number of surfaces with
radiative heat exchange. It is the radiative flux, qr, that couples the radiative heat
2-18
balance with the overall energy conservation expressed by Eqn. . This will be
explained in Section 2.5.4.
Figure 2.5.1 Definition of terms used in view factor calculations
2-19
2.5.3 Computation of View Factors

In order to compute qr s (in Eqn. 2.5-1), view factors Fij, between all radiating
surfaces must be available. In this section, the physical meaning of viewfactor and
its calculation will be discussed.
In the above defined Eqn. 2.5-1, Fij, the view factor between the surfaces i and j
must be computed. For a better understanding, i,j in the above equation can be
replaced with 1 and 2. Thus, view factor, F1-2, between two arbitrary surfaces (see
Figure 2.5.1), 1 and 2 is defined as a fraction of diffuse radiant energy leaving
surface 1 that arrives at surface 2. Mathematically,
1
F 1 2 = ------
A1 Cos 1 Cos 2
-------------------------------- dA 1 dA 2
2
(2.5-5)
A1 A2 r 12
where A1 and A2 are the areas of surfaces 1 and 2, respectively, r12 is the distance
between the two elemental areas dA1 and dA2, 1 is the angle between the position
dependent normal vector n and the line connecting dA1 and dA2. Angle 2 is
defined in a similar way. It must be noted that in defining the above expression for
the view factor, it is assumed that the two surfaces are gray, diffuse and isothermal.
Another important relation connecting the view factors F1-2 and F2-1 is as follows:
A1 F1-2 = A2 F2-1 (2.5-6)
The above relationship, known as reciprocal relationship, can be used to evaluate

F2-1 if F1-2 and the areas A1 and A2 are known. It must be noted that Cos 1 and
Cos 2 must be positive in order for the surface dA1 and dA2 to see each other. If
either of the cosines has a negative value, the corresponding view factor,
F dA dA should be set to zero. Such cases, in which the inactive side of the
1 2
radiating face acts as an obstructor, will be termed as self-obstruction cases. Also,
view factor F1-2 should be set to zero, if a third surface obstructs the view between
surfaces 1 and 2.
2-20
In order to calculate view factors internally, the user must specify the radiation
surfaces in terms of the finite element faces of a discretized domain. The user must
also specify which of the two sides is a radiatively active side. These pieces of
information can be supplied very easily via the already existing *BCGVAR card in
the NISA file of NISA/HEAT. Each radiating face is taken as one radiation surface.
View factors between the radiating surfaces are automatically generated by NISA/
HEAT taking into account self-obstruction and obstructions due to a third surface.
However, in the absence of any third surface obstruction, setting IOBST=0 in the
Executive command,SRAD = ON, IOBST, NAXI, NDIV, , will result in a
saving of substantial computer time.
As can be assessed from the preceding discussion, computing view factors can
result in usage of excessive computer time. To economize this computation,
different techniques are used depending on whether the geometry being analyzed is
2D, 3D or axisymmetric. For example, double area integration method [Siegel and
Howell] is employed in comparison with contour integration method [Siegel and
Howell] when a 3D geometry, with radiation surfaces, is being analyzed. No
special directives are required when computing view factors for axisymmetric
geometries. NISA/HEAT internally generates a complete 3D model (with the axis
of symmetry as the X-axis [NISA/HEAT]) to calculate the required view factors.
Furthermore, for 2D problems, a completely different approach, called Hottels
crossed-string method [Siegel and Howell], is employed for its computational
efficiency and accuracy. Reference [Siegel and Howell] provides more details for
evaluating view factors for interested readers.
2.5.4 Solution Procedure

In order to account for surface radiative exchange, discretized Eqn. 2.5-1 must be
solved for the radiative flux, q r . Then its effect must be reflected in the overall
energy balance at the affected surfaces.This is achieved by evaluating the first
integral on the right hand side of finite element formulation Eqn. .
q d (2.5-7)

where
2-21
q = qa + qc + qr
In the above equality qa, qc and qr refer to the applied heat flux, effect due to
convection boundary condition (*BCGVAR card) and effect due to radiation on
this surface, S respectively. Radiative effects can be specified via *BCGVAR card
with labels TG, and/or RAD. In the latter case, gray-body radiative effects will be
considered and q r is evaluated by solving Eqn. 2.5-2. It must be pointed out here
that a temperature field (see the right hand side of Eqn. 2.5-2) must be available, if
q r s, due to gray-body radiation effects are to be included. On the other hand, q r s
affect the distribution of temperature in the domain. This leads to a coupling
between the energy and radiative energy transfer systems. An effective solution
strategy must be established to solve this system of equations.
In NISA/HEAT, a sequential solution procedure is adapted to solve for temperature

distribution, Ts and then evaluation of q r s using discretized Eqn. ) and Eqn. 2.5-
1, respectively. This procedure is undertaken to reduce the wave front size (hence
solution time), which would otherwise be large for simultaneous solution of T and
q r . Both steady and transient solutions can be attempted using this approach. The
transient solution for q r is representative of temperature distribution existing at
that instant of time on the radiative surfaces and the domain in general. Different
solution approaches are undertaken based on the presence of forced or natural
convective effects. For example, when the heat transfer analysis has predominantly
forced convective effects, a converged velocity field is used to obtain a temperature
field in the domain. The temperature distribution is employed to evaluate q r s
using Eqn. 2.5-1. This in turn may alter the temperature distribution since the
forcing function, integral shown in Eqn. 2.4-14, may have changed. This process
of obtaining q r s and temperature distribution is repeated until these distributions
do not change.
It has been observed that q r (and hence T) solution may not converge or may do so
slowly. An under relaxation of q r leads to its stabilization. This is achieved as
follows:
2-22
i+1
q ri + 1 = q r + ( 1 )q ri (2.5-8)
* *
where the superscript i refers to an iteration no., and is a relaxation factor

defined by the user via an executive card SRAD = ON, IOBST, NAXI, NDIV, ,
and q r refers to a relaxed value of q r . Note that the relaxation parameter, ,
*
depends on radiation-conduction parameter, NR1, which is defined as
4
T h L
NR1 = -------------------------
-
k ( Th Tc )
where is the Stefan-Boltzman constant, Th and Tc are upper and lower absolute
temperatures, k is the material conductivity and L is characteristic length. If
radiation-conduction parameter is much larger than one, low values of relaxation
parameters must be used in order to get convergence. For problems involving
convection boundary conditions there is an important dimensionless grouping,
NR2, which is defined as
4 4
( T h Tc )
NR2 = ------------------------------
-
h ( Th Tc )
small value of NR2 helps convergence. The relaxation factor, should be picked
on the basis of the minimum of these dimensionless parameters (whichever ones
are relevant for a particular problem).
Together with checking for convergence of the velocity and temperature fields,
convergence is also checked for q r . L2 Norm (See Section 3.4) is evaluated and a
check is made against a user-defined tolerance (in the *FLCNTL card). The
solution procedure is continued until L2 Norm falls below the specified tolerance.
2-23
2.5.5 Using Artificial Convection Boundary Condition to Simulate Surface

Radiation
The sequential solution procedure for temperature and radiative flux is effective in
reducing the wavefront size, which would otherwise be large for simultaneous
solution of temperature and radiative flux. However, this solution strategy has
some limitations. Consider a problem with surface radiation and heat flux
boundary condition (see Figure 2.5.5.1).
Figure 2.5.5.1 Problem with heat flux and surface radiation

In such cases, temperature cannot be solved if temperature and radiative flux are
solved in a segregated manner. NISA/HEAT makes use of artificial convection
boundary condition. The boundary radiation surfaces are replaced by convection
boundary condition (h = guessed value, . T = guessed value). With the presence of
convection boundary condition, it is possible to solve for temperature. Our goal is
to seek a value of convection coefficient and effective surrounding temperature
such that it simulates surface radiation exactly. The temperature distribution from
the solution of energy equation is used to compute radiative heat flux. This value of
radiative heat flux is used to compute a new value for the convection coefficient
and effective surrounding temperature. It is through the artificial convection
2-24
boundary condition, the effect of surface radiation is brought into the energy
equation. Contribution of radiative heat flux is not directly added. Note that with
this technique, temperature and radiative flux are still solved in a segregated
manner.
Users can invoke this radiation stabalization technique by setting IRADHX to 1 in

the executive card UFIJ ( Section 6.2). This is the default method.
2.5.6 Special Cases

There are a few special cases which require a slight modification to the above
methodology for including the gray-surface radiative effects in the heat transfer
analysis. These are as follows:
(i) Plane(s) of symmetry in the domain
(ii) exchange of radiative flux through a window in a domain
(iii) exchange of radiative flux between a domain and surroundings
Whenever a plane of symmetry is encountered in the domain, local temperature

gradient normal to this plane must be set to zero, in the absence of surface
radiation. On the other hand, a local energy balance on this surface, between the
surface radiative effects and other modes of heat transfer must be established to
satisfy the balance of energy at this surface. Mathematically,
q = 0 = qr = qc
where q c is the local heat transfer flux (via conduction) and q r is the radiative flux
evaluated using Eqn. 2.5-1. This is automatically satisfied on the surfaces, where
no flux or temperature is prescribed (as is the case for a plane of symmetry).
Use of symmetry plane must be made very carefully. Figure a and Figure b show
two conduction problems with external boundaries as radiation surfaces. It is
appropriate to use symmetry plane for the problem shown in Figure a. However,
use of symmetry plane is not appropriate for the problem shown in Figure b . This
2-25
is due to the fact that the use of symmetry plane will lead to, effectively, neglecting
radiation exchange between certain surfaces, for example, between surfaces A and
B. Therefore, if a problem such as the one shown in Figure b is to be solved, it is
recommended to model the complete geometry.
In some cases, radiation surfaces do not form a complete enclosure. In such cases,
radiant energy is received from the surroundings beyond the opening. In these
cases, it is still possible to complete the enclosure by considering a pseudo surface
at the opening. The pseudo surface is black with a temperature that yields the
equivalence of the black body emissive power and the radiant energy from the
surroundings. The total radiation heat leaving the pseudo surface is given by
j Aj Fjk T j
4 4
Q k = A k T s
j =1
j k, other openings
j Ak Fkj T j
4 4
= A k T s (2.5-9)
N
Q
j Aj Fkj T j
4 4
q k = -----k- = T s
Ak
j =1
j k , other openings
This value of heat flux is applied as a boundary condition at the pseudo surface
(opening), surface k, for solving radiation exchange Eqn. 2.5-1. In the above
equation, Ts is the temperature of the surroundings beyond the opening. The user
just specifies the temperature, Ts. Heat flux, q k , is calculated internally in the
program. However, heat flux values at radiation outlets are not used as boundary
conditions for the energy Eqn. 2.2-7.
In certain situations where external surfaces exchange radiation energy with

surroundings, it may not be possible to complete an enclosure. Such a case is
shown in Figure 2.5.6.2. All external boundaries of the domain are radiating.
These surfaces can exchange radiation energy among each other and/or with the
2-26
surroundings. In order to account for this energy exchange, the term

N
Fij Ts
4
1 must be included on the right hand side of Eqn. 2.5-9. Thus
j=1
this equation is modified as
N
1 j N N
-----ij ------------
j
- Fij q rj =
( ij
4
F ij )T j
1
Fij Ts4

(2.5-10)
j=1
j
j=1 j=1
Note that the last term will have contribution only for external surfaces. For an
internal surface,
Fij = 1 , implying that the last term in the above equation becomes zero. If Fij
j=1
= 0, for all j, Eqn. 2.5-9 reduces to radiation boundary condition
4 4
q rj = j ( T j Ts ) (2.5-11)
as desired.
2-27
Figure 2.5.6.1 Figures illustrating the use of symmetry plane
2-28
Figure 2.5.6.2 Figure showing external surfaces exchanging radiative energy

among each other and/or with surroundings
2-29
Phase Change Effects
2.6 Phase Change Effects

The phenomenon of phase change is encountered in many engineering
applications. For example, continuous casting of metals and alloys, welding
operations, permafrost in cold climates, operations of condensors and evaporators
involve phase changes from one state of matter to another. Numerical modeling of
this phenomenon is desirable for better process control and overall design
parameter required for manufacturing the relevant components. The phase change
effects are associated with energy balances in the systems. For example, when a
liquid is cooled and begins to solidify, latent heat is generated. This amount of heat
must be acocunted for in the overall conservation of energy. Another issue that
must be addressed is the range of temperatures over which the phase change is
accomplished. Pure water freezes isothermally at 0C while the liquid steel may
solidify over a range of temperatures between 1400C to 1700C. At the boundaries
or surfaces where the phase change is taking place evolution (during solidification)
or absorption (during melting) of energy occurs. Thus the following conditions
must be satisfied at such boundaries [Tamma and Namburu, 1990].
T = Tf (2.6-1)
dV
QA = L ------- (2.6-2)
dt
where Tf is the phase change temperature, Q is the energy that must be balanced in
converting the volume V with interface area A evolving or absorbing the latent
heat L.
There are several ways to handle the above mentioned conditions (Eqn. 2.6-1) and
(Eqn. 2.6-2) Reference [Tamma and Namburu], a review paper, evaluates some of
the techniques employed to satisfy Eqn. 2.6-1 and Eqn. 2.6-2. NISA/HEAT
employs the following methods for handling the phase change effects:
(i) ENTHALPY METHOD: In this method, the enthalpy is evaluated in

the domain (at the nodes) based on the temperature distribution. The
enthalpies and temperatures at the nodes are used to calculate the spe-
cific heat values at the Gauss points. Thus obtained specific heat is
employed in the energy equation for calculation of temperature
distribution. Mathematically,
2-30
dH dt
C = ----------------
dT dt
n n1
H(T ) H(T )
or, C = ------------------------------------------- (2.6-3)
n n1
T T
where H refers to the enthalpy, T refers to the temperature, C is the spe-

cific heat, t refers to time, and the superscript n denotes the time step
number, It must be noted that the enthalpy evaluation accounts for both
the sensible and the latent heat energies. The conditions given by (Eqn.
2.6-1) and (Eqn. 2.6-2) are handled in an indirect manner.
(ii) SOURCE MEHTOD: This method attempts to satisfy the equations

(Eqn. 2.6-1) and (Eqn. 2.6-2) rather directly. The heat generated (or
absorbed) during the phase change is evaluated and included in the
energy equations to obtain a temperature distribution until the phase
change is completed. In NISA/HEAT run with phase change effects,
each node is associated with a certain volume of the material. When
this node (i) is undergoing phase change, it releases (or absorbs) energy
given by
C ( T i T f )V i
Q i = ----------------------------------
- (2.6-4)
t
where C is the specific heat, is the mass density, Tf is the phase

change temperature, V is the volume associated with node i, t is the
time step size. Note that there is no summation associated with the index
i. Qi time t is summed over time steps until it is equal to or becomes
greater than
L Vi (2.6-5)
At that instant Eqn. 2.6-1 and (Eqn. 2.6-2) will be satisfied and the
phase change at node i is said to be completed. Note that for nodes in
liquid or solid regions, Qi will be 0.
2-31
Convection Link
2.7 Convection Link

NISA/HEAT supports convection link element which allows heat to be convected
from one node in the model to another. This is achieved by the following element
stiffness matrix corresponding to convection link.
hA 1 1
1 1
Note that the element consists of two nodes and the temperature need not be known
at either of nodes. The heat transfer coefficient and area for heat convection must
be specified and is given under *RCTABLE card in NISA/HEAT.
2.8 Radiation Link

In addition to surface radiation, NISA/HEAT also allows radiation link element
which can be used to model heat transfer via radiation between two nodes.
Radiation heat transfer between two nodes i and j can be written as (Siegel and
Howell, 1992)
4 4
Q r = A i i G ij ( T i T j )
where is the Stefan-Boltzman constant, j is the emissivity of node i, Ai is the

surface area for radiation, Ti and Tj are the node temperatures, Gij is the fraction of
energy emitted by node i that is absorbed by node j after reaching the
absorbing surface by all possible paths. Note that Gii is generally not zero. Also
note that
A i i G ij = A j j G ji
The effect of radiation heat transfer is modeled using the following element
stiffness matrix.
2-32
2 2
A i i G ij ( T i + T j ) ( T i + T j ) 1 1
1 1
Note that the element stiffness matrix is linearized using the most recent available
node temperatures dependent and an iterative procedure is necessary to get the
final solution. This data needed to evaluate the element stiffness matrix (i.e.,
Stefan-Boltzman constant, emissivity, gray-body radiation factor Gij) is specified
under *RCTABLE card in NISA/HEAT.
The Gij values depend upon view factor matrix, Fij and emissivity values.
Calculation of matrix Gij is computationally very expensive. Each row of this
matrix obtained by solving NxN matrix, the coefficients of which depend on
emissivity values and view factors. Thus a different system of NxN equations must
be solved for each node. The user is referred ro Siegel and Howell (1992) for
further details.
2-33
Beam Element
2.9 Beam Element

Beam elements can be used to model one-dimensional coinduction through rods,
wires that connect two nodes of a 3-dimensional domain. This is implemented in
NISA/HEAT III by dropping terms containing ky and kz from the element stiffness
matrix (see Eqn. ). Heat flux, convection, radiation boundary conditions are
supported. Volumetric and nodal heat surfaces are also supported. All boundary
conditions or heat sources/sinks may be time and/or temperature dependent.
2-34
2.10 Heat Transfer in Composites

Introduction:
The properties of composite materials have led to their widespread applications in
several industries like aerospace, automotive, ship building, nuclear and other high
tech industries. Apart from their strength bearing properties and applications,
composite materials have been employed in high temperature applications. NISA
15.0 can be used for structural analysis of composites. 3D-FLUID version 15.0 and
HEAT III can be used for heat transfer analysis. In this section more details
pertaining to the heat transfer capabilities in laminated composites are described.
Theory:
In this first release, the first order lamination theory for composites is adapted to
understand the heat transfer analysis. Based on this, a linear temperature
distribution is assumed through the thickness of the laminated composites.
However, in-plane variation of the first order or higher orders can be used with a
choice of higher order elements using NKTP=32 and appropriate NORDR value.
In the laminated composites, the layers are stacked up three-dimensionally as
shown in Figure 2.10.1. Each layer can have principal directions for the thermal
conductivity and must possess a finite thickness. The local thermal conductivity
principal directions must be specified with respect to the global coordinate system.
Based on the first order lamination theory, the temperature field in three
dimensions can be written as [Noor, Argyris]:
T (x,y,z) = T0 (x,y) + z T1 (x,y) (2.10-1)
where
1
T 0 = --- ( T t + T b ) (2.10-2)
2
and
1
T 1 = --- ( T t T b ) (2.10-3)
h
where Tt and Tb are the temperatures at the top and bottom surfaces of a laminate
2-35
Heat Transfer in Composites
and h represents the plate (or shell) thickness.
Figure 2.10.1 Representation of Multilayered Composites
2-36
Figure 2.10.2 Fiber Orientation
2-37
Heat Transfer in Composites
Figure 2.10.2 shows the thermal conductivity directions (fiber or sometimes

called orientation) of a layer with respect to the local and global coordinate axes.
The user must provide Kxx, Kyy, Kzz and angle. 3D-FLUID and HEAT III will
compute the relevant transformations and convert the thermal conductivities in the
global axes and use these values for conduction heat transfer through the laminates.
The energy Eqn. 2.2-7 is solved for obtaining the temperature distribution in the
laminated plates (or shells). In this equation K, the thermal conductivity can be
written as a tensor Kij where i, j both vary from 1 through 3, ie,
K xx K xy K xz
K yx K yy K yz (2.10-4)
K zx K zy K zz
where subscripts x, y, z represent global coordinate directions.
Eqn. 2.2-7 is discretized using the Galerkin Procedure and after some algebraic
manipulations, one obtains algebraic equations with two degrees of freedom at
each node. (This is the main difference between the thin shell formulation with one
degree of freedom per node with NKTP=20 implemented in earlier versions of 3D-
FLUID and HEAT III).
The two degrees of freedom here are the average of the top and bottom temperature
gradient through the thickness and the temperature.
The boundary conditions of convection, radiation, heat flux and heat generation
can be aplied and have been described under *BCGVAR and *BCDVAR cards.
Whenever (average laminate) temperatures are known, these must be applied using
*BCDVAR card. The second degree of freedom will automatically be set to 0 at
those nodes when the boundary conditions are applied using DISPLAY III.
Another important point to remember is the order in which laminates must be
specified. These are further elaborated in Model Data under groups *LAMSEQ and
*LAMANGLE.
2-38
Nonlinear Heat Transfer Analysis in Composites:

In many instances, thermal conductivity of the laminate depends on its
temperature. 3D-FLUID and HEAT-3 allow the users to vary the thermal
conductivity of each layer based on temperature. This temperature dependence of
thermal conductivity renders the analysis to be nonlinear.
For users convience, the thermal conductivitys variation with temperature can be
described by
(a) a polynomial
(b) a table
Both of the above capabilities are supported in 3D-FLUID and HEAT 3. The user
must input either the coefficient of the polynomial (up to 3 rd order, i.e., 4
coefficients) or provide a table specifying the amplification values of thermal
conductivity as described in *MATFLUID card.
Please note that the material id of a layer identifies its property.
2-39
References
2.11 References
Kreyszig, E., Advanced Engineering Mathematics, John Wiley, 1972.
Rohsenow, W.M., and Harnett, J.P., Handbook of Heat Transfer, McGraw Hill,
1973.
Seigel, R. and Howell, J.R.; Thermal Radiation Heat Transfer, McGraw-Hill

Book Company, New York, 1981.
Tamma, K.K. and Namburu, R.R., Recent Advances, Trends and New
Perspectives via Enthalpy Based Finite Element Formulations for Applications to
Solidification Problems, International Journal for Numerical Methods in
Engineering, Vol. 30, pp. 803-820, 1990.
2-40
Chapter
3
NISA/HEAT Solution Technique
3.1 Introduction
NISA/HEAT supports both direct and iterative solution methods. The direct solver
used by NISA/HEAT in frontal solution technique. Two iterative solvers are
available: (1) conjugate gradient method and (2) Generalized minimal residual
(GMRES). Each of these methods are described in this chapter. A description of
different types of error reporting is also included.
3.2 Frontal Solver

NISA/HEAT uses the wavefront technique for the solution of the overall finite
element equilibrium equations, which are in the form of simultaneous linearized
algebraic equations. For most practical problems, the computer time required for
the solution of the system of equations represents a substantial portion of the total
computer time of the run. In the wavefront (frontal) method, the solution time is
proportional to the square of the wavefront size. Therefore, it is important to
minimize the wavefront size. For similar reasons, it is important to minimize the
band-width in banded solution methods. The frontal technique uses the Gauss
elimination method for the solution of the simultaneous linear equations. It
alternates between assembly and decomposition of the coefficient matrix, with the
assembly proceeding in the order in which the elements are defined. As a result,
the entire (global) coefficient matrix is never totally in core. This is unlike the
band-width techniques in which the element contributions to the overall equations
are fully assembled before factorization.
3-1
Frontal Solver
In the frontal technique, the ordering of the elements determines the wavefront
size. The node numbering sequence does not have any effect on the wavefront size.
Again, this is unlike the banded solvers in which the node numbering is what
affects the bandwidth. The instantaneous wavefront size is defined as the number
of active equations that are in core when a typical element is assembled to the
coefficient matrix in the front. When an equation is fully summed, meaning that the
corresponding unknown has made its last appearance and will not appear again in
any subsequent element, it is eliminated from the current set of equations, thus
making room for a new unknown to occupy the same place. As an element is
brought into the front, unknowns making their first appearance occupy vacant or
new positions in the coefficient matrix, and unknowns which appeared earlier are
assembled to the existing coefficients. The assembled coefficient matrix, therefore,
keeps expanding and contracting as elements are assembled and unknowns making
their last appearance are eliminated.
Three quantities are defined to reflect the varying size of the frontwidth. The
maximum wavefront (Wmax) is the maximum size that the instantaneous wavefront
can attain:
Wmax = max(Wi), i = 1, 2,..., M (3.2-1)
where M is the number of elements, and Wi is the instantaneous wavefront given

by
W i = W i 1 + F i L i 1 , i = 1, 2,..., M (3.2-2)
where Fi is the number of variables making their first appearances as element (i) is
assembled to the front, and Li-1 is the number of variables that make their last
appearances when element number (i-1) was assembled to the front. Clearly, WM +
1 must be zero. The average wavefront size (Wav) is defined as
M
1-
W av = ----
M Wi (3.2-3)
i=1
3-2
The average wavefront size represents the average length of an eliminated

equation. The eliminated equations constitute the decomposed coefficient matrix
and are written to NISA files 30 and 31. Thus, a rough estimate of the number of
words in the decomposed matrix is M * Wav. The solution time is proportional to
the root mean square wavefront (Wrms) defined by
12
M
1
Wi
W rms = ---- 2 (3.2-4)
-
M
i= 1
The largest problem that can be solved using NISA/HEAT depends on the memory
available at a particular installation. On most main frames and mini computers, the
problem size is limited to a maximum wavefront of 2000. The maximum wavefront
should not be confused with the total number of degrees of freedom in the problem.
The wavefront limit is sufficiently large for most practical problems. It is not
unusual to have a problem with 30,000 DOF and a maximum wavefront of 500.
The three measures of the wavefront size are printed in the output file. In addition,
the wavefront history (the instantaneous frontwidth as the front advances through
the elements) may also be printed, see the executive command EXECUTION in
Section 6.2.
As mentioned earlier, the wavefront size depends on the element sequence. A nodal
degree of freedom becomes active as the element in which it appears first is
assembled into the front, and it becomes eligible for elimination when it makes its
last appearance. In order to reduce the wavefront size, the elements should be
ordered such that the first and last appearances of a typical node, as implied by the
element connectivity, are as close as possible. This is evident from Eqn. 3.2-2,
where the objective is to make the quantity (Fi - Li-1) minimum.
In other words, the element should be sequenced such that the wavefront advances
through the model continuously from one end to the other in the direction which
has the largest number of nodal points.
It is important to note that the element sequence is defined in ascending order of

the element identification numbers and not by the sequence in which they are
entered in the data deck.
3-3
Frontal Solver
Figure 3.2.1 Effect of element numbering sequence on the wavefront

for a simple mesh (one DOF per node is assumed)
As an example of the effect of element numbering on the wavefront, consider the
mesh shown in Figure 3.2.1, where two different element numbering schemes are
used. It is evident from the Figure that the minimum wavefront results when the
front sweeps across the model in the direction of the largest number of nodes.
Wavefront Optimizer
One may easily optimize the wavefront size for a simple mesh such as the
rectangular mesh shown in Figure 3.2.1. But it becomes exceedingly difficult and
3-4
more time consuming to minimize the wavefront for a more complicated mesh. In
addition, the elements are generally numbered in an order which is convenient to
the user for easier identification of the regions in the model. In many cases, the
user element sequence, defined by the ascending order of the element ID numbers,
does not yield a cost-effective wavefront size. As mentioned earlier, the solution
time is proportional to the square of the frontwidth.
NISA/HEAT has a built-in wavefront optimizer which does not require additional
input from the user, and can be controlled using a single command (see the
executive command RESEQUENCE in Section 6.2). The internal element
sequence is invisible to the user. All communications between the user and the
program, as far as the input specification and the output request are concerned, are
in terms of the users element identification numbers.
The wavefront optimizer is a direct element resequencing algorithm based on

topology, and level structure, and it favors the minimum front growth criterion (the
quantity Fi - Li-1 in Eqn. 3.2-2). Several tie-breakers are used when the criterion is
satisfied by several candidates.
The specified boundary conditions are not considered in the resequencing process,
since these conditions are load case dependent, and can only reduce the frontwidth.
The calculated frontwidth is hence an upper bound. The optimizer is invoked only
once prior to the solution of all iterations.
The wavefront optimizer is cost-effective. The CPU time spent in the resequencing
process is only a small fraction of the total CPU time. The latter is, of course,
reduced due to reduction of the CPU time in the solution phase. For medium size
problems, say less than 5000 DOF, the wavefront optimization takes about 5 to
10% of the CPU time taken in a single pass through solution phase, and about 1 to
10% for larger size problems, depending on the computer installation.
The algorithm performs five independent trials, each with a different starting
element, for the minimization of the frontwidth. The starting elements are chosen
such that they are the least connected to the model or the most distant from each
other (in the topological sense). A typical trial may be terminated before
completion if it becomes apparent that the previous trial has better wavefront
statistics. The final internal element sequence is chosen from the completed trial
3-5
Frontal Solver
with the least mean square frontwidth (this is the default option, see executive
command RESEQUENCE).
It should be noted that the wavefront statistics reported in the wavefront

optimization module are upper bounds but are close estimates for the actual
wavefront parameters, for the following reasons:
1. The optimizer works with nodes on the front rather than nodal degrees of
freedom on the front. During the solution, some nodal degrees of free-
dom may be eliminated before all the elements connected to the node are
assembled.
2. The specified boundary conditions, which reduce the number of active
DOF, consequently the frontwidth, are not considered in the
resequencing process.
3-6
3.3 Iterative Solver

The nonlinear equations described in the previous section can be solved either by a
direct solver (e.g. Gaussian Elimination) or by an iterative procedure. The
advantage of a direct solver is that a solution to the system of equation is always
obtained after a finite number of operations. However, for large problems, the
computer resources (CPU time and disk space) can be prohibitively large. Iterative
methods are very attractive for solving those kinds of problems. Iterative methods
start with an initial guess and build up the final solutions with each iteration. They
have the advantage of preserving the sparsity of the coefficient matrix, which is
very important for finite element methods where the matrices of large number of
zero entries. Use of iterative solvers for large problems typically result in reduces
storage and computational time. The convergence pattern of the iterative solver
improves dramatically if the initial guess value is close to the actual solution.
Therefore, the number of iterations required for convergence reduces in successive
time iterations because the previous time iteration solution can be used as the guess
value.
Three different iterative solution procedures are currently available in NISA/

HEAT.
(a) Conjugate Gradient Method (CG)
(b) Generalized Minimal Residual Method (GMRES)
(c) Pre-conditioned Conjugate Gradient Method (PCG)
The basic system to be solved is of the form: Ax = b
where A is the coefficient matrix, x is the solution of the system of length N and b
is the right hand side.
Conjugate Gradient Method
In this type of solution procedure, the coefficient matrix (A) has to be symmetric
and positive definite, which is true for pure heat transfer problems. The basic
algorithm is as follows [2]:
3-7
Iterative Solver
1. Start with an initial guess x0

2. Find the residual, r0 = b - Ax0
3. Set q0 = r0
4. For i = 0, 1, 2,..... do following operations till convergence is achieved
(i.e. r0 is smaller than a pre-determined tolerance limit).
(i) qi = (ri, ri) / (qi, Aqi)
(ii) xi+1 = xi + aiqi
(iii) ri+1 = ri - aiAqi
(iv) bi = (ri+1, ri+1) / (ri, ri)
(v) qi+1 = ri+1 + biqi
(where (a,b) indicates dot product of two vectors a & b).
CG has modest storage requirements and converges to the solution in at

most N steps.
Generalized Minimal Residual Method (GMRES)
GMRES is applicable to unsymmetric coefficient matrices as well. This method is

based on Arnoldis process [1] for constructing an orthogonal basis of Krylov
subspaces. GMRES does not break down even for problems with indefinite
symmetric coefficient matrices. The disadvantage as compared to CG method is
that the storage requirements increase a lot with the number of Arnolds steps and it
becomes increasingly expensive as the number of steps increases. Therefore, the
number of orthogonal vectors are limited to a predetermined number, k, and the
algorithm is restarted when this number is reached. The solution procedure is as
follows:
1. Start with an initial guess x0

2. Find the residual r0 = b - Ax0
3-8
(1)
3. Set = r0 r0 2
4. For i = 1, 2,......, k do the following till the residual norm W(2) = ||b - A xi||2
is less than or equal to the tolerance limit or i = k (k is preset integer - no.
of orthogonal vectors to be formed before restart).
(i)
(i) W = A
(ii) For m = 1,.....,i
(m)
hm,i = ( w, )
(m)
W = W h m, i
end
(iii) hi+1,i = || W ||2
(i + 1)
(iv) = W/hi+1,i
(v) apply J1,....., Ji-1 on (h1,i,......, Hi+1,i)
3-9
Iterative Solver
(vi) Construct Ji acting on the ith & (i+1)st

Component of h ,i such that (i+1)st
Component of Ji hi is 0
(vii) s:= Jis
5. Compute y as the solution of Hy = s in which the upper ixi triangular

part of H has hi,j as its elements, and s represents the first i components
of s.
(0) (1) (1) (2) (2) (i) (i)
6. x = x + y + y + + y
7. s(i+1) = b Ax 2
8. if x is accurate enough approximation then quit
else x(0) = x
9. Restart the algorithm (go to step 2).
GMRES is guaranteed to converge to the exact arithmetic for any N x N matrix A.
3-10
References:
1. W.E. Arnold, The Principle of Minimized Iterations in the solution of the
Matrix Eigenvalue Problem, Quart. Appl. Math., V. 9 (1951), pp. 17-29.
2. R. Chandra, Conjugate Gradient Methods for Partial Differential Equations,
Ph.D. Thesis, Department of Computer Science, Yale University, Connecticut,
1978.
3-11
Solution Errors
3.4 Solution Errors

NISA/HEAT discretizes the partial differential equations, relevant to the analysis
being performed, using the Galerkin approach. A set of linear algebraic equations
is obtained with this discretization procedure.
These equations are represented as

Ax = b (3.4-1)
where A is the coefficient matrix and x and b are the vectors of unknown variable
and known right-hand side, respectively. It must be pointed out here that x refers to
the solution at each iteration in the case of steady state analysis. For transient
analysis, the solution is obtained at every inner iteration in each time step. In
problems dealing with nonlinear heat transfer, the matrix A is a function of the
solution from the previous iteration(s). The difference in solutions obtained at
subsequent iterations can be termed as Error at each node. For transient analysis,
it is possible to have one inner iteration per time step. In such cases, the above-
mentioned difference in solutions at subsequent iterations refer to that between two
consecutive time steps.
There are two methods of reporting the change in nodal solutions at two
consecutive iterations. These are termed as maximum difference and L2 Norm
errors and are reported in the non-dimensionalized form of temperatures.
Temperature is non-dimensionalized as either of the two methods can be selected
via an executive command ERROR (Chapter 6).
Maximum Difference
In this method of reporting the errors, a maximum value of the difference in
solutions at nodes between two consecutive iterations is reported. This can be
written as follows:
k k k 1
D error = Max [ T i T i ] , i=1,...., N (3.4-2)
where Derror is the error in the variable T, the superscript k represents the iteration
number, the subscript i indicates the node number and N is the total number of
nodes in the model.
3-12
When reporting this error in the output file (and log file, if any), also given are the
corresponding variable name, iteration number and the node number where this
maximum difference is found. A typical message for the variable temperature is as
follows:
TEMPERATURE ITERATION NO. = 3 MAX. ERROR = 0.52301E - 02 AT

NODE = 1051
The maximum error reported is a dimensionless temperature difference.

Temperature is non-dimensionalized as
T T ref1
T = ------------------------------
- , where Tref1 and Tref2 are the minimum and maximum
T ref2 T ref1
allowable temperatures in the domain, and inputted through *FLCNTL Card (see
Chapter 8 for details).
L2 Norm
Another popular and a more strict measure of error is obtained by considering L2

norm of vectors. Bigness or Smallness of a vector or matrix can be indicated in
terms of a norm. Therefore for comparing two sets of solution vectors, a norm of
the difference between them can be defined. One such norm denoted as an L2 norm
or Euclidian norm, can be used for error estimation. NISA/HEAT employs the
following expression as an indicator of error.
N
k 2
( xi
k+1
xi )
i=1
L 2 Norm = -------------------------------------------------
-
N
k+1 2
( xi )
i=1
where xi is any nodal quantity at the in node, k is a step no. or an iteration no., and
N is the total no. of nodes.
3-13
Solution Errors
Monitoring Convergence
Error at each iteration or time-step is also written to file 56. A plot of error vs.
iteration can be obtained by typing the following command:
monitor [- options...]
where options include:
- time seconds time interval after which graph is to be updated.

- file filename filename should be the name of file 56.
- scale type scale type can either be log or linear for y-axis, x-axis is
always linear. Default scale type is linear.
- by color background color, Default is white.
Note that we must save file 56. File 56 can be saved by including unit 56 in the
executive command:
SAVE = 26, 27, 54, 56.

Note:
The above capability is available only on UNIX and WINDOWS NT workstations.
3-14
3.5 NISA/HEAT Files

NISA/HEAT III uses several secondary storage files for reading and writing data as
well as storing analysis information on a temporary or permanent basis. A brief
description of the files used in NISA/HEAT III is given in Table .
Table 3.5-1 NISA files used in fluid flow analysis

FILE
TYPE CONTENTS
CODE+
Formatted, sequential Users input data
Formatted, sequential Printout
Formatted, sequential Log file, brief messages for the
analysis progress
11 Binary, sequential Scratch
16 Binary, sequential Optimal wavefront route
17 Binary, sequential Optimal wavefront route
18 Binary, sequential Element linear stiffness matrices
26 Binary, direct access Basic model and analysis data
27 Binary, direct access Analysis results (post data)
30 Binary, direct access Decomposed stiffness matrix/Global
mass or geometric stiffness
matrix
3-15
NISA/HEAT Files
Table 3.5-1 NISA files used in fluid flow analysis

31 Binary, direct access Decomposed stiffness matrix/Global
mass or geometric stiffness
matrix
44-53 Binary, sequential Scratch
54 Binary, sequential Used for restart runs
55 Ascii, sequential Used to generate *NDTEMPER data
group for thermal stress analysis
using NISA II
56 Ascii, sequential Used to report iteration/time-step
error for iterative graphical display
+
file codes designated by dash () are system dependent
3-16
Chapter
4
Element Library
4.1 Introduction
4.1.1 Element Types

A comprehensive library of finite elements is available in NISA/HEAT. Each
element in the library is identified by two variables, NKTP and NORDR. The
variable NKTP specifies the element type whereas the variable NORDR specifies
the element shape and number of nodes. For example, NKTP = 2 and NORDR = 1
designates a two-dimensional, 4-noded quadrilateral element, whereas NKTP = 2
and NORDR = 2 designates a two-dimensional 8-noded quadrilateral element.
Table lists all heat transfer elements available in the library. Table also shows the
heat transfer elements along with their features.
The available NORDR values for each element type are given in the detailed
description of each element. NORDR controls both the element shape functions
(linear through cubic) and the geometry of the element (quadrilateral or triangular,
with straight or curved sides). Considering all available combinations of NKTP
and NORDR, more than twenty-five different element types are available in the
NISA/HEAT element library.
4-1
Introduction
Table 4.1-1 Summary of heat transfer elements

Element Type DOF per
Element Description No. of Nodes
No. (NKTP) Node
2 2-D Element 3 to 9 1
3 Axisymmetric Element 3 to 9 1
4 3-D Element 6 to 27 1
20 3-D Shell Elements 3 to 8 1
32 3-D Laminated 3 to 8 1
Composite Shell
Elements
49 and 50 3-D convection and 2 1
radiation links
12 Beam Element 2 1
4-2
Element Library
Table 4.1-2 Heat Transfer element classification

NKTP and
SHAPE FEATURES
NORDR
2-D and NKTP = 2, 3 - Heat transfer problems
axisymmet NORDR = 1,2,
ric 10,11, 30 - Steady state and tran-
elements sient
- Radiation and convec-

tion boundary condi-
tions
3-D NKTP = 4 - Heat flux and heat gen-

elements NORDR = 1,2, eration rate
10,11, 20, 21, 30
- Surface radiation
- Phase Change
3-D shell NKTP = 20

elements NORDR = 1,2,
10, 11
4-3
Introduction
Table 4.1-2 Heat Transfer element classification

3-D NKTP = 20
Lamited NORDR = 1,2,
composite 10, 11
shell
elements
3-D NKTP = 49, 50 - Steady and transient

convection NORDR = 1 3D heat transfer prob-
and lems
radiation
links
3-D bar NKTP = 12 - Heat Transfer prob-

elements NORDR = 1 lems
- Steady state and tran-

sient problems
All the elements used in NISA/HEAT are isoparametric which means that the same
shape functions are used to represent both the element geometry and the
temperature variation within an element. Elements with mid-side nodes can have
curved sides. For example, NORDR = 2 elements use three nodes to define each
side. Since a parabolic curve can exactly fit three points, the sides of these elements
can exactly match a parabolically curved boundary. For shell elements, coordinates
of a generic point, with M nodal points are given by
M M
tk k Vni
k k
xi = k x i + ( 2)
k=1 k=1
4-4
Element Library
where
xi Cartesian coordinates (x1, x2, x3) of a generic point on the shell

element
k Cartesian coordinate of a nodal point k in the shell middle surface
xi
Vn i Component of the unit vector Vn in direction i and at nodal point k
k Shape function at nodal point k, k = k ( , n )
tk Thickness of the shell in the direction and at nodal point k
( , , ) natural coordinates
4.1.2 Element Input

Element input definition includes specifying element type, element connectivities
and material properties. The common data groups that specify most of the above
input are listed in Table , which also shows the requirement of each data group for
specific element types (NKTP). Other data groups that specify boundary conditions
are mentioned explicitly in the description of each element. These data groups
(specifying boundary conditions) are detailed in Chapter 8.
4-5
Introduction
Table 4.1-3 Data groups for element definition

Sec-
Group ID Name tion Purpose Requirement
No.
Element Selection and
Definition
*ELTYPE 7.3.1 Select element types used in the Always
model
*ELEMENTS 7.3.2 Define element connectivities Always
and identification numbers
(pointers) for element types and
material properties
Element Material
Properties
*MATFLUID 7.5.1 Define element material For all
properties, e.g., density, specific elements
heat, conductivity, etc.
4-6
Element Library
4.2 2-D Element (NKTP = 2)

Description
A unit thickness is assumed for this element. The theoretical basis of the element is
discussed in Chapter 2. An element reference guide is given in Table with the
available boundary conditions given in Table .
The element can be shaped as a 4-, 8-, or 9-noded quadrilateral, or as a 3- or 6-

noded triangle depending on the selected NORDR value. The element
configuration, node locations and face numbering convention are shown in Figure
4.2.1. A general description of the element input data is given in Section 4.1.2.
Table gives the output pertinent to this element.
Figure 4.2.1 Element configuration for available NORDR values, NKTP = 2

(circled numbers indicate face numbers)
4-7
2-D Element (NKTP = 2)
Table 4.2.1 Element reference guide (NKTP = 2)

Element Type NKTP = 2 (2-D element)
Analysis Types HEAT
Degrees of Freedom 1 per node: T,
NORDR - Quadrilateral: 4, 8 or 9 nodes (NORDR = 1,
(Shape / No. of nodes) 2, 30)
- Triangle: 3 or 6 nodes (NORDR = 10, 11)

Material Properties
Isotropic:
Transient Heat Transfer 3 properties: DEN, COND, SPEC
Steady State Heat Transfer 1 property: COND
Anisotropic:
Transient 4 properties: DEN, KXX, KYY, SPEC
Steady State 2 properties: KXX, KYY
Nodal Output - TEMPERATURE (T)
Table 4.2.2 Available loading (NKTP = 2)

Nodal Boundary - Specified temperature (T)
Condition
- Nodal heat source
Element Boundary - Heat flux (qx, qy) along global axes or (qn)
Condition along outward normal to the element face
- Radiation or convection on any face of the

element
Heat Generation - Specified heat generation in an element
4-8
Element Library

Surface Radiation - Specify element faces as radiation surfaces to
be included in surface radiation analysis
Note:
1. The element connectivity must be given in the order shown in Figure 4.2.1, in
which the node numbering sequence starts at a corner node and proceeds along
the perimeter of the element in a counterclockwise direction.
4.3 Axisymmetric Element (NKTP = 3)

Description
This element is based on the assumption that its geometry, material properties and
loadings are rotationally symmetric about an axis. The axis of symmetry coincides
with the Z-axis of the cylindrical system (Z-R- ) and R-axis denotes the radial
direction. The element must lie in the +ve Z and +ve R plane. The theoretical basis
of this element is given in Chapter 2. An element reference guide is given in Table
with the available loading given in Table .
The element can be shaped as a 4-, 8-, or 9-noded quadrilateral, or as a 3- or 6-

noded triangle depending on the selected NORDR value. The element
configuration, node locations and face numbering convention are shown in Figure
4.3.1. A description of the input and output pertinent to theis element is given in
Table .
4-9
Axisymmetric Element (NKTP = 3)
Figure 4.3.1 Element configuration for available NORDR

values, NKTP = 3
4-10
Element Library

Element Type NKTP = 3, (Axisymmetric element)
Analysis Types HEAT
NORDR - Quadrilateral: 4, 8 or 9 nodes (NORDR = 1,
(Shape / No. of nodes) 2, 30)

Material Properties
Isotropic:
Anisotropic:
Transient 4 properties: DEN, KZZ, KRR, SPEC
Steady State 2 properties: KZZ, KRR

Condition
- Nodal heat source
Element Boundary - Heat flux (qz, qr) along global axes or (qn)

element
4-11
Axisymmetric Element (NKTP = 3)

Notes:
2. Once a reference is made to NKTP = 3, the first coordinate entry in the
description of nodes is interpreted as that in the direction of the axis of
symmetry (i.e., Z-axis of the Z-R- system) while the second entry refers to
the radial direction (i.e., R-axis). If the modelling plane is represented by X-Y
plane (as in the case of NISA pre-processor DISPLAY III), this can be easily
achieved by making Z-R of the axisymmetric system coincident to the X-Y
system of the modelling plane. Also, the model must be contained in the
positive Z - positive R quadrant (or if using DISPLAY III, then in the positive
X - positive Y of the DISPLAY III modelling plane).
3. Axis of symmetry must be X or Y-axis. Default is X-axis. Axis of symmetry is
specified in the executive card DIME (Section 6.2).
4-12
Element Library
4.4 3-D Element (NKTP = 4)

Description
The theoretical basis of the element is discussed in Chapter 2. An element
reference guide is given in Table with the available loading given in Table .
The element can be shaped as an 8-, 20- or 27- noded hexahedron (brick) element,
or as a 6-, 15- or 18-noded wedge element depending on the selected NORDR
value. The element configuration, node locations and face numbering convention
are shown in Figure 4.4.1. A general description of the input and output data
pertinent to this elemet is given in table Table .
4-13
Figure 4.4.1 Element configurations and face numbering convention

for available NORDR values, NKTP = 4
4-14
Element Library

Element Type NKTP = 4, (3D-Element)
Analysis Types HEAT
NORDR - Hexahedron (brick): 8, 20 or 27 nodes
(Shape / No. of nodes) (NORDR = 1, 2, 30)
- Wedge: 6, 15 or 18 nodes (NORDR = 10,

11, 12)
- Tetrahedron: 4 or 10 nodes (NORDR =

20,21)
Material Properties
Isotropic:
Transient Heat Transfer 3 properties: DENS, COND, SPEC
Anisotropic:
Transient 5 properties: DENS, KXX, KYY, KZZ,
SPEC
Steady State 3 properties: KXX, KYY, KZZ
4-15

Condition
- Nodal heat source
Element Boundary - Heat flux (qx, qy,qz) along global axes or
Condition (qn) along outward normal to the element
face

element
Notes:
the perimeter of a face on the element. This face will be considered face-1 of
the element. Following the right hand rule, the midside nodes of the edges are
then defined with the first edge midside node above the first node of face-1.
Face-2 of the element follows similarly. For wedge elements, face-1 must be a
triangular face.
2. The linear quadrilateral element (NORDR = 1) uses trilinear shape functions
to represent the element geometry.
4-16
Element Library
4.5 3-D Thin Shell Element for Heat Transfer (NKTP=20)

Description
This is a curved thin shell element oriented arbitrarily in space and is suitable for
modeling heat transfer in thin shell structures. The element has one degree of
freedom per node (T). The theoretical basis of the element is discussed in Chapter 2
and Section 4.1.1. An element reference guide is given in Table with the available
boundary conditions given in Table .
The element can be shaped as a 4-, or 8-noded quadrilateral, or as a 3- or 6-noded

triangle depending on the selected NORDR value. The element configuration, node
locations and face numbering convention are shown in Figure 4.2.1. Table gives
the output pertinent to this element.
4-17
3-D Thin Shell Element for Heat Transfer (NKTP=20)
Figure 4.5.1 Element configuration for available NORDR values,

NKTP = 2 (circled numbers indicate face numbers)
4-18
Element Library

Element Type NKTP = 20 (3-D thin shell)
Analysis Types HEAT
NORDR - Quadrilateral: 4, or 8 nodes (NORDR = 1,
(Shape / No. of nodes) 2)

Material Properties
Isotropic:
Anisotropic:
SPEC
4-19
3-D Thin Shell Element for Heat Transfer (NKTP=20)

Nodal Boundary - Specified temperature (T) or heat source at
Condition the node (for heat transfer)
(top or bottom only)
- Radiation or convection on top, bottom or

side faces of the element
Surface Radiation - Specify element faces as radiation surfaces
to be included in surface radiation analysis
(only top and bottom)
Notes:
2. The element is suitable for modeling curved thin shell structures. Heat flux,
convection and radiation, if any, may be applied to the top and bottom faces.
Top and bottom are determined as follows:
(a) Choose a view point so that the element connectivities are defiend coun-
terclockwise.
(b) The top face of the element is now closer to the observer. The bottom
and top faces are numbered 1 and 2, respectively. The inplane faces of
quadrilateral elements are numbered 3 to 6 (3 to 5 for triangles), as
shown in Figure 4.5.1.
4-20
Element Library
4.6 3-D Laminated Composite Shell for heat transfer (NKTP = 32)
This is a curved thin 3D Shell element oriented arbitrarily in space and is suitable
for modeling heat transfer in moderately thick to thin laminated composite shells.
The element consists of a number of layers of perfectly bonded orthogonal
materials. The element has two degrees of freedom per node (T and T n ). The
theoretical basis of the element is discussed in section 2.10. An element reference
guide is briefed in Table 4.6.1 with the available loading listed in Table 4.6.2.
Figure 4.6-1 3-D laminated composite shell (NKTP = 32), (a) element configuration,
(b) layer setup, (c) face numbering convention for top and bottom surfaces
The element can be shaped as a 4 to 12 node quadrilateral, or a 3 or 6 node triangle

depending on the selected NORDR value. The element configuration, layer setup,
4-21
3-D Laminated Composite Shell for heat transfer (NKTP = 32)
node locations and face numbering convention are shown in Table 4.6-1. The
lamination sequence is between the top and bottom faces of the element with the
layer setup starting from the top face. Table 4.6-1 gives the output pertinent to this
element.

Element Type NKTP = 32 (3-D laminated composite shell )
Analysis Types HEAT
Degrees of Freedom 2 per node: T and T n .
NORDR - Quadrilateral: 4, or 8 nodes (NORDR = 1,

(Shape / No. of nodes) 2)
Material Properties
Isotropic:
Anisotropic:
SPEC
4-22
Element Library

Nodal Boundary - Specified temperature (T) or heat source at
Condition the node (for heat transfer)
(top or bottom only)
- Radiation or convection on top, bottom or

side faces of the element
Surface Radiation - Specify element faces as radiation surfaces
to be included in surface radiation analysis
(only top and bottom)
Notes:
1. The element may be oriented anywhere in space. The element connectivity

must be given in the order shown in Figure 4.6-1, in which the node numbering
sequence starts at a corner node and proceeds along the perimeter of the
element.
2. The bottom and top faces of the element are numbered as faces 1 and 2,
respectively (Figure 4.6-1). Top and bottom faces are determined as follows:
(a) Choose a viewpoint so that the element is defined counterclockwise
(b) The top face is now closest to you
For pressure loading, the pressure (force/area) is integrated over the area of the
loaded face. For curved shells, this area will be somewhat different than the
area of the shell midsurface.
3. Transverse temperature gradients are included in the element formulation.
4-23
3-D Laminated Composite Shell for heat transfer (NKTP = 32)
4. For most NISA elements, the variable MATID in the *ELEMENTS data group
refers to a material property table defined in *MATERIAL data group.
However, this is inadequate for composite shells, since each layer may be of a
different material. Therefore, for this element, MATID is a pointer to an entry
in the *LAMSEQ (lamination sequence) data group. The *LAMSEQ data
group in turn contains pointers to the layer thicknesses, rotation angles, and
materials for all layers (*RCTABLE, *LAMANGLE and *MATERIAL data
groups, respectively). In this way, we retain the ability to model composite
shells which have tapered thickness or variable rotation angles. There is no
limitation on the number of layers or the leap sequence. Note that the number
of real constant tables, lamination angle tables, or material property tables may
not necessarily be equal to the number of layers, since more than one layer
may be of the same material type, thickness type, or lamination angle
(orientation) type.
5. There is no need to enter a non-zero value for the variable IDRC (real constant
ID number) in the *ELEMENTS data group for this element. Any value
entered for IDRC will be reset to zero.
6. The element consists of a number of layers of perfectly bonded orthotropic
materials. For each different material, for transient or steady state heat transfer
analysis DENS, COND, SPEC or DENS, COND respectively are to be
specified. Several options are available for defination of material principal
axes, see *LAMSEQ for details.
7. The following special plots are available for this element. These plots may be
obtained using the postprocessing module of the DISPLAY program:
(a) Contour plots of temperature on the top and bottom faces of the elements.
4-24
Element Library
4.7 3-D Bar Element for Heat Transfer (NKTP = 12)

Description
This element is a 2-node bar element in three dimensions for heat transfer analysis.
The element may be oriented any where in space. The element has the temperature
(T) as the only degree of freedom at each node. The theoretical basis of the element
is discussed in Section 2.9. An element reference guide is given in Table with the
available loading given in Table .
The element configuration and face number convention are shown in Figure 4.7.1.
Table gives the output pertinent to this element.
Figure 4.7.1 Element configuration and face number convention

(NKTP = 12, NORDR = 1)
4-25
3-D Bar Element for Heat Transfer (NKTP = 12)

Element Type NKTP = 12, 3-D Bar element for heat transfer
Analysis Types Steady state and Transient heat transfer
Degrees of Freedom 1 per node: T (temperature)
NORDR - Line, 2 nodes (NORDR = 1)
(Element shape, nodes)
Real Constants 4; cross sectional areas at node 1 and 2 and
perimeters at node 1 and 2, respectively (A1,3 A2,
P1, P2).
Material Properties 3 constants: KXX, DENS, C, see note 1

Nodal Loading - Specified non-zero nodal temperature values,
see *BCDVAR
- Specified nodal heat generation rate, see

*BCDVAR (NHS)
Element Boundary Heat flux (qx, qy) along global axes or (qn) along
Condition outward normal to the element face (top or bot-
tom only)
Radiation or convection on top, bottom or side
faces of the element
Heat Generation Specified heat generation in an element
Notes:
1. For steady state heat transfer analysis, the values of DENS and C are not
required.
4-26
Element Library
2. The element has three faces. Area and perimeter are specified at the two end
nodes (input in *RCTABLE data group).
4.8 3-D Convection Link Element (NKTP = 49)

Description
This element is a 2-node convection link element in three dimensions and it may be
used for heat convection between two nodes. The element can be used in three-
dimensional applications. The element has the temperature (T) as the only degree
of freedom at each node and can be oriented anywhere in space. The theoretical
basis of the element is discussed in Section 2.7. An element reference guide is
given in Table with the available loading given in Table .
The element configuration is shown in Figure 4.8.1.
Figure 4.8.1 Element configuration for convection link

4-27
3-D Convection Link Element (NKTP = 49)

Element Type NKTP = 49, 3-D convection link element for
heat transfer
NORDR - 2 nodes (NORDR = 1), see note 2
Real Constants 2; convection surface area and film coefficient
(A, h) respectively, see note 1
Material Properties none
Element Output - Nodal point heat flow at nodes with speci-
fied temperature

Nodal Loading - Specified non-zero nodal temperature values at
any one of the element nodes, see *BCDVAR
Notes:
1. The convection surface area, A, in *RCTABLE data group, for this element is
the surface area over which heat is convected. In an axisymmetric analysis, the
convection area must be the total circumferential area.
2. Nodal points may or may not be coincident.
4-28
Element Library
4.9 3-D Radiation Link Element (NKTP = 50)

Description
This element is a 2-node radiation link element in three dimensions and may be
used for heat radiation between two nodes. The element can be used in two-
dimensional (including axisymmetric) or in three-dimensional applications. The
element has the temperature (T) as the only degree of freedom at each node and can
be oriented anywhere in space. An element reference guide is given in Table with
the available loading given in Table
The element configuration is shown in Figure 4.9.1.
Figure 4.9.1 Element configuration for radiation link

4-29
3-D Radiation Link Element (NKTP = 50)

Element Type NKTP = 150, 3-D radiation link element for heat
transfer
NORDR - 2 nodes (NORDR = 1), see note 3
Real Constants 4; surface area for radiation (A), surface emmis-
sivity (), shape factor (F), and Stefen-Boltz-
man constant ( ), see notes 1 and 2
Material Properties none
Element Output - Nodal point heat flow output at nodes with
specified temperature
Transient Capabilities - Variable, user defined, or equal time step
increments with various time integration
schemes

Nodal Loading - Specified non-zero nodal temperature values at any
one of the element nodes, see *SPTEMP
Notes:
1. If the emmissivity is temperature dependent, then the curve identification
number is given in *RCTABLE data group and the corresponding curve is
input in *TEMPFN data group.
4-30
Element Library
2. The input quantity, A in *RCTABLE data group, for this element is the surface
area over which heat is radiating. In an axisymmetric analysis, the radiation
area must be the total circumferential area.
3. Nodal points may or may not be coincident.
4-31
3-D Radiation Link Element (NKTP = 50)
4-32
Chapter
5
Input Setup
5.1 Introduction
The input data setup for a typical NISA/HEAT analysis run, is described in this
chapter.
This chapter also explains the rules of the free format input. A detailed description
of the executive commands, which constitute the first data block in a typical
NISA/HEAT deck, and their applicable analysis types is given in Chapter 6.
Detailed description of the other two data blocks, namely, the model data block
and the analysis data block are given in Chapter 7 and Chapter 8, respectively. A
list of all available cards in NISA/HEAT is given in Table .
The input data for a typical NISA/HEAT analysis type consists of three data
blocks and a data terminator as shown in Figure 5.1.1.
The three data blocks and the data terminator constituting a NISA/HEAT input
deck should appear in the sequence shown in Figure 5.1.1, namely:
- Executive commands data block
- Model data block

- Analysis data block
- Data terminator
A brief description of each data block and its general format are as follows:
5-1
Introduction
1. Executive Commands Data Block:

This is the first data block in a typical NISA input deck. It consists of command
cards which specify general control parameters for the execution of the program,
such as: type of analysis, restart option, post-processing files, etc. The syntax for
a typical executive command is:
Command name = option or data
where the left hand side (Command name) is a character string, and the right
hand side (option or data) is an applicable option(s) or data for the specific com-
mand. Detailed description of these commands is given in Chapter 6.
2. Model Data Block:

The model data block describes the model characteristics of the domain, e.g.,
coordinates, connectivities, material properties, etc. It consists of a *TITLE
card(s), followed by distinct data groups that are arranged arbitrarily to form the
block. The *TITLE data group should be the first group in the model data in
order to enable printing the title of the problem on every page of the printout.
Each data group consists of a group identification card which has to be the first
card in the group, followed by the pertinent free format data of the group. For
example, all nodal point coordinates are defined in one data group which is pre-
ceded by its group identification card, *NODES. Detailed descriptions of all the
groups of the model data block are given in Chapter 7.
5-2
Input Setup
Figure 5.1.1 NISA/HEAT Input deck setup for a typical analysis
5-3
Introduction
3. Analysis Data Block:

The analysis data block describes data pertinent to boundary conditions. Print
controls and specific output features are also described in the analysis data.
Specified zero and non-zero boundary conditions are considered as applied
boundary conditions for a particular degree of freedom and they are defined in the
analysis data block.
The analysis data block consists, in general, of a delimiter data group followed by
distinct data groups that can be arranged in arbitrary form in the block. The
delimiter data group is *FLCNTL. Each data group, including the delimiter group,
consists of a group identification card that has to be the first card in the group,
followed by the pertinent free format data of the group.
Detailed descriptions of all the groups of the analysis data block are given in
Chapter 8.
4. Data Terminator:
The input data terminator signals the end of the data deck. The *ENDDATA group
identification card represents the data deck terminator which must be the last card
in the input data deck.
5-4
Input Setup
5.2 Free Format Input Features

1. Overview and Comment Cards:
NISA input is designed in a convenient and user friendly free format with various
automatic input generation features. As discussed above, the input data is
composed of main blocks, each of which consists of distinct data groups. Data
groups are identified by a unique group identification card which must be the first
card in the group, followed by the pertinent free format data of the group. The
following sub-sections describe the general rules and guidelines for the use of the
free format input.
Comment card(s) may be inserted anywhere in a NISA deck. Any card starting
with two asterisks (**) is a comment card, e.g.:
** boundary conditions
Comment cards are for the users reference only and they are ignored by the
program.
2. Group Identification Cards:

A group identification card must precede the free format data of a typical data
group. The general format of the group identification card is as follows:
*group name, parameter = data
where the group name is a character string constituting a valid group name. The
parameter is a pertinent group feature and the data is a value assigned to the
parameter.
Group identification names may be abbreviated to a minimum number of unique

characters. If names are not abbreviated, however, they must be spelled out
correctly to any number of characters specified between the minimum abbreviation
form and the full form. For example, the *ELEMENT group identification name
may be abbreviated to *ELEM, *ELEME, or *ELEMEN, but *ELEMZ or
*ELEMENV are not acceptable. Embedded blanks within group identification
names are not allowed.
5-5
Free Format Input Features
3. Data Cards:
The data for each group follows the appropriate group identification card. The
precise content of each data group is detailed in Chapter 7 and Chapter 8. In what
follows, we provide some general rules and/or restrictions applicable to all data
cards:
- Leading, trailing, and embedded blanks are allowed anywhere in a card. A

blank card is totally ignored and does not imply null entries.
- A comma (,) must be entered between two successive entries. A repeated

comma (,,) implies a null entry. A comma after the last entry in a card is
optional.
- A slash (/) serves as a horizontal repetition character, e.g., 2.5// means 2.5, 2.5,
2.5. Slashes need not be separated by commas.
- Special tab character ($) is marked on many data cards. Inserting the tab char-
acter on a data card has the effect of supplying null entries for all variables up
to the variable following the tab character. For example, in the data for the
*NODES data group (see Chapter 7 ), there is a tab character ($) between the
4th and 5th entry on the card. Thus the following entries are equivalent:
1 $ 1.0, 1.0, 0.0 and

1, 0, 0, 0, 1.0, 1.0, 0.0
A comma before or after a tab character is optional.
5-6
Input Setup
4. Input File Switching (Input Re-direction):

The input data can be re-directed at any point in the data deck, from the primary
input file to a secondary file by using the *READ, filename command. The
subsequent data will be read from the secondary file until the end of file is reached,
then the input will be automatically re-directed to the primary file. Only one level
of file switching is allowed. That is, the file re-direction command must be entered
in the primary file only. An example is shown below:
Primary File (FILE1) Secondary File (FILE2)

*TITLE *ELEMENT
Example for file switching 1, 1, 1, 0
1, 2, 3, 4
*ELTYPE
1, 2, 1
*READ, FILE2
...
...
...
5. Input Data Echo (*ECHO)

The echo of the input data may be suppressed or printed (part or all) by using the
*ECHO = ON or *ECHO = OFF commands. The *ECHO command may appear
anywhere in the data deck. However, it will be in effect for the data group(s)
succeeding it. This command may be used as many times as needed to switch the
echo on or off.
5-7
Table 5.2-1 Alphabetical list of all valid command and input data groups
Section
Group ID(1) Description
No.
ANALYSIS 6.2 Specify analysis type
ANISO Indication of presence of direction-dependent
6.2
properties
*BCDVAR 8.2.1 Variable Dirichlet type boundary conditions
*BCGVAR 8.2.1 Variable Neumann type boundary conditions
BFACE 6.2 Generate a list of boundary faces
CNVBC Lump matrix due to convection boundary
6.2
condition
*CPVAR 7.6.4 Coupled variable data
DIMENSION 6.2 Specify dimension of problem
DTAUTO 6.2 Automatic time step size control
*E1 7.3.3 Alternate element definition
ELEMENTECHO 6.2 Echo element input
*ELEMENTS 7.3.2 Element definition
*ELTYPE 7.3.1 Element type specification
*ENDDATA 8.4.1 Input data terminator
ERROR 6.2 Specify convergence criteria
EXECUTION 6.2 Select execution/checking run
FILENAME 6.2 Specify NISA file prefix
*FLCNTL 8.1.1 Fluid load case control
*ICDS 8.2.3 Dirichlet type initial condition
*ICGRAD 8.2.5 Neumann type initial condition
INITIALC 6.2 Specify a uniform initial condition
ITERATION 6.2 Iteration strategy
*LCSYSTEM 7.4.1 Local coordinate system definiton
MASS_LUMPING 6.2 Lump mass matrix
*MATFLUID 7.5.1 Material property data
MEMORY 6.2 Initial captured memory
5-8
Input Setup
Table 5.2-1 Alphabetical list of all valid command and input data groups
*NODES 7.4.1 Nodal coordinates
NODEECHO 6.2 Echo node input
PCHANGE 6.2 Turn on phase change computation
*PCHANGE 8.2.6 Specification of phase change parameter
*PRINTCNTL 8.3.1 Selective printout control
(1) Acceptable minimum abbreviations are in bold face
Table 5.2-2 Alphabetical list of all valid command and input data groups (cont.)
Section
Group ID(1) Description
No.
RESEQUENCE 6.2 Element resequencing
RESTART 6.2 Restart options
SAVEFILE 6.2 Save specific NISA file
*SETS 7.6.1 Specification of a set of numbers
SRAD 6.2 Analysis with surface radiation
SOLUTION 6.2 Solution method
STDS 6.2 Steady state analysis
STEP 6.2 Time step size control
TCUTOFF 6.2 Temperature cutoff
TEMP_OFFSET 6.2 Temperature offset
*TEMPFN 7.6.2 Temperature dependency curves
*TIMEAMP 7.6.3 Time amplitude curves
*TITLE 7.2.1 Problem title
UFIJ 6.2 View factor and/or iteration control
WARNING 6.2 Set warning flag
(1)
Acceptable minimum abbreviations are in bold face
5-9
5-10
Chapter
6
Executive Commands
6.1 General Description

As described above, the executive commands constitute the first data block in a
typical NISA/3D-FLUID input deck, and they define general control parameters
for program execution. The syntax for a typical executive command is:
Command name = option or data
The command name is a character string that may be abbreviated to a minimum of

the first four characters.
The option or data is a character string or a data value that will be assigned to the
specific command. Section 6.2 gives detailed description of executive commands
that are applicable to specific analysis types. The listing of the commands within
each section is arranged alphabetically and an overall alphabetical listing is given
in Table . The following general notes apply to all executive commands:
- Embedded blanks in executive commands are ignored;
- Command names may be abbreviated. Minimum abbreviations are

shown in bold face letters.
6-1
General Description
- Various command options are shown in square brackets. Default values

of options are shown in braces.
Table 6.1 Alphabetical list of executive commands

Executive Command Name(1) Description
ANALYSIS Specify analysis type
ANISO Indication of presence of direction
dependent properties
BFACE Generatre a list of boundary faces
CNVBC Lump matrix due to convection boundary
condition
DIMENSION Specify dimension of problem
DTAUTO Automatic time step size control
Executive Command Name(1) Description
ELEMENTECHO Echo element input
ERROR Specify convergence criteria
EXECUTION Select execution/checking run
FILENAME Specify NISA file prefix
INITIALC Specify a uniform initial condition
ITERATION Iteration strategy
MASS_LUMPING Lump mass matrix
MEMORY Initial captured memory
NODEECHO Echo node input
PATH Directory for temporary files
PAT1,..., PAT4 Directory for temporary files
PCHANGE Turn on phase change computation
RESEQUENCE Element resequencing
RESTART Restart options
SAVEFILE Save specific NISA binary data file
SOLUTION Solution procedure
6-2
Executive Commands
Table 6.1 Alphabetical list of executive commands (cont.)

SRAD Analysis with surface radiation
STDS Steady state analysis
STEP Time step size control
TCUTOFF Temperature cutoff
TEMP_OFFSET Temperature offset
TRANS Transient solution termination control
UFIJ View factor and/or radiation control
WARNING Set warning flag
(1)
Acceptable minimum abbreviations are in bold face.
6-3
Executive Commands
6.2 Executive Commands

ANALYSIS - Specify analysis type
ANALYSIS = [HEAT]
where
HEAT : indicates heat transfer analysis
ANISO - Indication of presence of direction-dependent properties
ANISO = ON
{ OFF }
where,
ON : Thermal conductivity is taken to be direction dependent

OFF : Thermal conductivity is direction-independent
BFACE - Generate a list of boundary faces
BFACE = ON
{ OFF }
ON : The program generates a list of boundary element faces and writes it

to file bound.dat.
CNVBC - Lump matrix due to convection boundary condition
6-4
Executive Commands
CNVBC = LUMP
{ OFF }
LUMP : Lumping of contribution due to convection boundary condition

to the stiffness matrix will be invoked.
This feature is useful when there is a large heat gain or loss due to convection
boundary condition. This option is useful in avoiding any unphysical over-shoot or
under-shoot in the temperature solution.
DTAUTO - Automatic time step size control
DTAUTO = ON, TEND, DTMIN, DTMAX, DTINC, DTDEC

{ OFF, 1.3, 0.7 }
where
ON/OFF : Turn automatic time stepping on or off

TEND : Time upto which analysis is to be run
DTMIN : Minimum time step size
DTMAX : Maximum time step size
DTINC : Factor for increasing time step size (DTINC >1)
DTDEC : Factor for decreasing time step size (0 < DTDEC < 1)
Notes:
1. There are no defaults for TEND, DTMIN and DTMAX. The user must supply
these values.
2. Currently this capability is available for ANAL = HEAT only.
6-5
Executive Commands
DIMENSION - Dimension of fluid or heat transfer problem
DIMENSION = {2D}
AX, AXIS
3D
where
2D : Two-dimensional problem
AX : Axisymmetric problem
AXIS : Axis of symmetry. It must be either x or y-. The default is x, i.e.,
x-axis.
3D : Three-dimensional problem
ELEMENT_ECHO - Echo element input
ELEMENT_ECHO = { ON }
OFF
where
ON : Element input echo is on

OFF : Element input echo is off
ERROR - Specify convergence criteria
ERROR = { MAXDIF }
L2NORM
6-6
Executive Commands
where
MAXDIF : Maximum error in a dimensionless form will be used to check

for convergence against user specified tolerance
L2NORM : L2 Norm will be used to check for convergence against user
specified tolerance
For more information, the user may refer to section 3.4.
EXECUTION - Select execution / checking run
EXECUTION = {GO}
CHECK
CGO
where
GO : Indicates execution run only, no wavefront check

CHECK : Indicates wavefront check only, no execution to be performed
CGO : Indicates wavefront check followed by execution run
FILE_NAME - Specify NISA file prefix
FILE_NAME = fname
{ no default }
where
fna : is an alphanumeric character string that will be used as a file name

me prefix followed by the NISA file number to be saved. Up to
64 characters may be used (system dependent).
6-7
Executive Commands
This command must be used in conjunction with the SAVE FILE command which
specifies the NISA file numbers to be saved (e.g., 26, 27,...).
INITIAL_CONDITION - Specify a uniform initial value for a variable
INITIAL_CONDITION = VALUE ,T
{ T lower , T }
where
VALU : Initialization value of the variable given in LABEL.

E
T : Indicates the name of the variable (temperature) being initial-
ized. The user should also refer to the *ICDS data group in
Analysis Data.
If temperature has been specified on some particular nodes in the *ICDS data
group, then the initialization values given here will be superceded by those in the
*ICDS card. If this card is not present, then temperature is initialized to TLower as
specified in the *FLCNTL card (ref. Analysis data). Further, *BCDVAR will take
precedence over initial condition.
ITERATION - Specify iteration strategy
ITERATION = METHOD ,PARAMETER ( S )

{ FIXED ,1.0 }
where
METHOD : Name iteration type adopted in each time step.
= FIXED Fixed point iteration
= MNR Modified Newton-Raphson iteration
(currently unavailable)
6-8
Executive Commands
PARAMETER(S) : Parameter(s) necessary to qualify the iteration strategy

The nonlinear finite element system of equations at the
nth time step for the ith iteration can be written as:
n
n i
K u i u in = f u

where, u is a typical variable (say velocity), and
n n n1
ui = u i 1 + ( 1 ) u
while a value of = 1 retains the full non-linearity,

< 1 would usually expedite convergence ( = 0.5 is
a popular choice). It should be noted, however, that in a
true transient problem only a value of = 1
should be used.
6-9
Executive Commands
MASS_LUMPING - Lump mass matrix
MASS_LUMPING = C1, C2,

{ none }
where
TE : Lump mass matrix for energy equation
Notes:
1. Usually lumping of mass matrix for the energy equation can help in avoiding
unphysical overshoot or undershoot in temperature solution.
2. This option is useful for transient problems only.
MEMORY - Initial captured memory
MEMORY = { 1000000 }
N
where
N : is an integer specifying the number of double precision words of

memory
NISA/HEAT starts off by capturing N (default = 1000000) words of memory. If the

program needs more memory to run the problem, it captures more memory at run
time. However, in some cases the user may be prompted to increase the amount of
captured memory by putting a higher value for N.
6-10
Executive Commands
NODE ECHO -Echo node input
NODEECHO = { ON }
OFF
where
ON : Node input echo is on

OFF : Node input echo is off
Note:
Users are advised to use the more general *ECHO data group (Section 5.2) since
the NODE ECHO command will not be supported in future versions.
PATH - Directory for temporary files
PATH = Pname
{ no default }
where
Pnam : is a character string that will be used as a directory for NISA

e temporary files. Up to 64 characters may be used (system
dependent). The directory name should include the last character just
before the file name. For example, a directory /user/nisa/tmp in
UNIX operating system should be given as /user/nisa/tmp/. The last
slash is required. The user must also have write permission
for the directory. Path names such as./tmp/ will be taken relative to
the current working directory.
6-11
Executive Commands
Note:
NISA temporary file will be opened in users current working directory if this
command is not used.
PAT1
PAT2
PAT3
PAT4 - Directory for NISA files
PAT1 = Pname, n1, n2,

PAT2 { no default }
PAT3
PAT4
where
Pname : is a character string that will be used as a directory for NISA

files n1, n2, etc. Up to 64 characters may be used (system
dependent). The directory name should include the character
just before the file name. For example, a directory /user/nisa/
tmp in UNIX operating system should be given as /user/nisa/
tmp/. The last slash is required. The user must also have write
permission for the directory. Path names such as /tmp/ will be
taken relative to the current working directory.
n1, n2,...:
NISA file numbers which should be open in pname directory. These four
commands allow the user to open files in four different directories other than the
current working directory during a NISA run. Saved files, however, will be opened
in the current working directory.
6-12
Executive Commands
PCHANGE - Phase Change (solidification)
PCHANGE = { OFF }
ON, METHOD
where
OFF : Indicates that phase change does not occur

ON : Indicates that phase change may occur
METHOD : Indicates the method chosen
If 1 is chosen for METHOD, it implies that enthalpy formulation for the phase
change is desired while METHOD = 2 invokes source method formulation. The
default value of 1 is chosen in case MEHTOD is not indicated.
RESEQUENCE - Element resequencing
RESEQUENCE = { ON }, MAX, LIST

OFF
where
ON : Element resequence for wavefront minimization is on.

MAX : Select the trial that gives the least maximum wavefront. The default
option is to select the trial that gives the least root-mean-square
(RMS) wavefront. The MAX may be used if the available memory is
the determining factor. In most cases, the trial that gives the least
RMS wavefront also gives the least maximum wavefront.
LIST : Output a table of correspondence between the element IDs and the
internal numbering sequence. The default option is not to provide
this table.
6-13
Executive Commands
OFF : Do not resequence the elements. The element resequencing is

transparent to the user. The element IDs are not changed. All
communications between the user and the program as far as input
and output are concerned are in terms of the users element
identification numbers.
The element resequencing is transparent to the user. The element IDs are not
changed. All communications between the user and the program as far as input and
output are concerned are in terms of the users element identification numbers.
The command RESE = ON (default option) resequences the elements and selects
the trial with the least RMS wavefront, but does not provide a table of
correspondence for element numbering. The options may appear in any order.
RESTART - Restart analysis for Incompressible Flow
RESTART = { 0 }
n, ISTEP
where
n : Is the restart option number

=0 no restart, execution proceeds from scratch
=1 uses the same model data and same boundary conditions of the
previous analysis, and continues solving the problem with the
results of ISTEP time step of previous run.
=2 uses the same model data of previous analysis but new
boundary conditions, solving the problem with the zero initial
guess. ISTEP has no relevance in this case.
=3 uses the same model data of the previous analysis but new
boundary conditions, solving the problem with the results of
ISTEP time step of the previous run as the initial guess.
for n > 0 restart, files 26, 27 should be saved in the previous analysis.
6-14
Executive Commands
Whenever ISTEP is applicable and is not supplied, the last step is used. Every time
a RESTART is made, the post-processing files, i.e., files 26 and 27, are
overwritten. This means that if the results prior to any RESTART are needed later
on, the files 26 and 27 should be copied with a different set of file names before
starting the RESTART run. Currently ISTEP is fixed at the last time step. Option 2
is currently unavailable.
In case of a RESTART > 0, EXECUTIVE cards, material data (*MATFLUID), and

fluid control card (*FLCNTL) are always read.
SAVEFILE - Save Specific NISA files
SAVEFILE = n1, n2,

{ no default }
where
n1, : NISA file numbers to be saved

n2,...
This command must be used in conjunction with the FILE NAME command which
specifies a character prefix (fname) for files to be saved. Refer to Chapter 3 for
NISA file numbers.
Note:
For certain computer systems, the NISA files to be saved are actually specified by
the Job Control Language (JCL) and this command is not needed. Refer to
pertinent system manual for details.
SOLU - Solution Procedure
{ FRONT }
SOLU = ITERATIVE, MAXIT, TOLER, METHOD, IOCORE
SPARSE
6-15
Executive Commands
where
FRONT : Direct, frontal solver used for the solution of the system of
linearized algebraic equations.
ITERATIVE : Iterative solver used for the solution of the system of
linearized algebraic equations
MAXIT : Maximum number of iteration for the iterative solver
TOLER : Required level of convergence or tolerance for the iterative
solver
METHOD : Iterative solution method
1 : Conjugate gradient (CG)
2 : Generalized minimization of residual (GMRES)
3 : Pre-conditioned Conjugate Gradient Method (PCG)
4 : Pre-conditioned Conjugate Gradient Method (PCKSS)
IOCORE : The out-of-core scheme activated instead of the in-core
scheme for the iteration process to speed up the
performance of the solver
= 0 in-core scheme for iteration is used.
= 1 out-of-core scheme for iteration is used.
Unless specified by the user, the iterative solver uses the following default values;
MATRIX = 400, TOLER = 10-6 and METHOD = 2.
SPARSE : Direct, sparse solver used for the solution of the system of
linearized algebraic equations.
SRAD - Analysis with surface radiation
SRAD = ON, IOBST, NAXI, NDIV, , IRVIEW

{ OFF }
6-16
Executive Commands
where
ON : Surface radiation effects will be included

OFF : Surface radiation effects will not be included
IOBST : =1, Third surface obstruction and self-obstruction checks
will be made for view factor calculation.
=0, Third surface obstruction checks will be skipped.
However,
self-obstruction checks will still be made
NAXI : Number of divisions along the circumference. This value is
used only for axisymmetric problems and is meaningless
for 2D and 3D problems.
NDIV : Number of sub-divisions of a surface for view-factor
calculation
: Relaxation parameter for radiative heat flux 0 < ( ) < 1
IRVIEW : This option is used for restarts only.
0 : do not recompute view factors
1 : re-compute view factors
Notes:
1. Whenever possible, the user should use IOBST = 0. This leads to substantial
savings of CPU time for view factors calculation.
2. If NAXI is specified to be less than 8, NISA/HEAT will set this value equal to
8. If NAXI is specified to be a very large value (say 9999), it will be reset to the
maximum allowable value based on available memory.
3. For 2D-problems having third surface obstructions, the user can instruct
NISA/3D-FLUID to subdivide a radiation surface into NDIV number of
divisions for view factor computation. This leads to more accurate values of
view-factors. However, CPU time will also be higher. Maximum allowable
value for NDIV is equal to 10. There is no provision to subdivide a surface for
3D-problems.
4. Relaxation parameter for radiative flux should lie between 0 and 1
0 < ( ) < 1 . A small value of can slow down the rate of convergence and a
6-17
Executive Commands
high value of may cause solution to diverge. Values between 0.1 and 0.3 are
recommended.
5. Radiation view factors are stored in file 54 (binary), only if obstruction checks
are on. Therefore, view factors do not have to be recomputed at restarts. How-
ever, the user can force NISA/3D-FLUID to recompute the view factors at
restarts using the executive command SRAD=ON, IOBST, NAXI, NDIV,
RELAXQ, IRVIEW by setting IRVIEW to 1 will trigger recomputation of
view factors on restarts. Users may want to recompute view factors if they
wish to
(i) change NDIV for 2D problem at restarts
(ii) change NAXI for axisymmetric problems at restarts
(iii) switch on or off obstruction checks at restarts
Note that view factors will not be stored in file 54 if obstructions checks are not
performed. This is done to save the disk space required to store the view factors. If
view factors are not stored in file 54, they will automatically get recomputed at
restarts.
6. Unless specified by the user, the following default values will be used; IOBST
= 1, NAXI = 32, NDIV = 1, =0.2 and IRVIEW = 0.
STDS - Steady State Analysis
STDS = ON, , ,
{ OFF }
6-18
Executive Commands
where
ON : Steady state analysis option will be used. Maximum number of

iterations is to be specified in the *FLCNTL card (see Section
8.1.1 ).
OFF : Transient analysis option will be used. Number of time steps
and the maximum number of iterations per time step are to be
specified in the *FLCNTL card.
, , : = not used
= not used
= relaxation factor for temperature
Notes:
1. The nonlinear finite element system of equations at nth iteration can be written
as
n 1
n 1 T n = f T
K T

n1
where T is a typical variable (say temperature), T is the under-relaxed value
of T after n -1 iteration. Solution after n iterations is taken to be
n n n1
T = T + ( 1 )T
when Tn is obtained after solving the above given system of equations, is called
the relaxation parameter.
2. Relaxation factors , and are set to 1.0 by default. For nonlinear energy
equation, a value of less than 1.0 may be used for temperature relaxation
parameter, .
6-19
Executive Commands
STEP - Time step size control
The users can subdivide the time step given in *FLCNTL card into still smaller
time steps using the executive command.
STEP = N
{1}
where, N is the number of subdivisions of the time step with default value of N
being 1.
In DISPLAY, the user will be able to look at the results at time steps given in the
*FLCNTL or *STEPSIZE cards. However, NISA/HEAT will use smaller time
steps size (DT/N) for time accurate solution.
TCUTOFF - Upper and lower limits of temperature
TCUTOFF = ON
{ OFF }
If TCUTOFF = ON, REFT1 and REFT2 will be the upper and lower cutoff
temperatures. Otherwise, these are just used to non-dimensionalize the energy
equation.
TEMP_OFFSET - Temperature offset
TEMP_OFFSET = TVAL
{ 0, 0 }
TVAL : Temperature offset value
This card can be used by users who wish to input data in degree Celsius in a
radiation problem. In such a case TVAL should be set to 273.15. The final results
will be output in degrees Celsius.
6-20
Executive Commands
TRANS - Transient solution termination control
TRANS = NSTMIN, EPSUVW, EPSTMP

{ 1, 1.0e 06, 1.0e 06 }
where
NSTMIN : Minimum number of time steps for which solution must be run.
EPSUVW : Velocity tolerance for checking steady state
EPSTMP : Temperature tolerance for checking steady state.
Notes:
1. This card is used for transient analysis only.
2. Tolerance giving in *FLCNTL card is used to check convergence within the
time step. Tolerances given in this card are used to check if steady state has
been reached. If change in solution between two successive time steps is
smaller than these tolerances, solutions will be terminated.
6-21
Executive Commands
UFIJ - View factor and/or surface radiation control
User defined view factors can also be read by NISA/HEAT. In addition, view
factors computed by NISA/3D-FLUID can be printed out in the output file. This is
done using an executive command.
Executive card UFIJ can be used to:
1. Read user defined view factors.

2. Print view factors computed by NISA/HEAT.
3. Solve surface radiation problems involving only heat flux, heat generation (i.e.
no specified temperature boundary condition.
UFIJ = IUSRVF, IPRVF, IRADHX, HRAD

{ 0, 0, 1, 1.0 }
where
IUSRVF : 0 = Do not read user defined view factors. Compute view factors
internally.
This is a default.
1 = Read user defined view factors, do not compute view factors
internally.
IPRVF : 0 = Do not print view factor matrix in the output file. This is
default.
1 = Print the view factor matrix in the output file.
IRADH : 0 = Do not use artificial convection boundary condition method
X for solving surface radiation problems with heat flux.
element heat generation etc. type of b.c.
1 = Use artificial convection boundary condition method.
HRAD : Typical value of radiative heat transfer coefficient in the domain
2 2
hr = ( T + Tr ) ( T + T r )
6-22
Executive Commands
Notes:
1. If user defined view-factors are to be read, they should be given the file vfac-
nisa.dat. Format for this file is given under Analysis Hints (Section 9.4 ).
2. Solution is independent of the value of HRAD. However, use of this option
can help stabalize the solution.
WARNING - Set warning flag
WARNING = { GO }
STOP
where
GO : directs the program to continue after a warning(s) is produced during

input processing
STOP : program will stop if a warning is produced during input processing
6-23
Executive Commands
6-24
Chapter
7
Model Data
7.1 Introduction
The NISA data deck consists of distinct data blocks as has been mentioned in
Chapter 5. The data blocks must be ordered in the sequence shown in Figure5.1.1.
The model data block represents the majority of the input data and is detailed in
this chapter. It describes the physical characteristics of the finite element model,
but without any boundary condition. The latter is described in the next chapter.
The description of the model primarily consists of:
1. Element Data
(a) Element types selection: A table of the element types used in the model.
(b) Element definition: Specification of element connectivities (node list),

and identification numbers (or pointers) for element types and material
properties.
2. Nodal Data
(a) Definition of local coordinate systems
(b) Nodal coordinates
3. Material Data
This includes specification of material properties.
7-1
Introduction
4. Miscellaneous Data
This includes the definition of sets (groups) of nodes or elements, definition of
temperature and time amplitude curves, vectors, etc., which may be referenced
in the analysis data.
The above input items of the model data are arranged in distinct data groups. Each
data group consists of a group ID card followed by its free format data, which may
consist of one or more card sets. Each card set may consist of one or more cards.
The free format rules are given in Section 5.2. Each group has a descriptive
identification name indicating the function of the data group. For example, the
element definition is given in the *ELEMENTS data group, the nodal coordinates
in *NODES data group.
The detailed description of the data groups is presented in this chapter in the
sequence shown in Table 7.1.1. This sequence is adopted according to the function
of each data group as shown above. For easy reference, an alphabetical list of all
data groups available in NISA/HEAT is given in Table . It should be noted that the
data groups may be arranged in the model data quite arbitrarily, with the following
exceptions:
1. The *ELTYPE data group must precede the *ELEMENTS data group.
2. It is recommended that the *TITLE data group be the first data group in
the model data set. This will allow the title of the problem being ana-
lyzed to be printed on every page of the output file.
It should also be noted that not all of the data groups given in this chapter need to
be present in the model data for the heat transfer analysis.
7-2
Model Data
Table 7.1.1 List of model data groups
Section No. and Group ID(1) Description

7.2 Title
7.2.1 *TITLE Problem title
7.3.1 Element Data
7.3.1 *ELTYPE Element type selection
7.3.2 *ELEMENTS Element definition
7.3.3 *E1 Alternate element definition
7.3.4 *RCTABLE Real constant
7.3.5*LAMANGLE Rotation angles for laminated
composite elements
7.3.6*LAMSEQ Composite lamination sequence data
7.4 Nodal Data
7.4.1 *LCSYSTEM Local coordinate system
7.4.2 *NODES Nodal coordinates
7.5 Material Data
7.5.1 *MATFLUID Material property data
7.6 Miscellaneous Data
7.6.1 *SETS Definition of a set of numbers (e.g.,
for *PRINTCNTL)
7.6.2 *TEMPFN Definition of temperature amplitude
curves
7.6.3 *TIMEAMP Definition of time-amplitude curves
7.6.4 *CPVAR Coupled variable data
(1) Acceptable minimum abbreviations are in bold face
7-3
Title
7.2 Title
7.2.1 *TITLE Data Group - Problem Title

This data group is always required and is used to print a title for the analysis. It is
recommended that this data group be the first data group in the model data set so
that the title of the problem may be printed on every page of the output file.
Group ID card: *TITLE
Problem title card set:
Entry No: 1
Variable: Problem title
Max char: 80
entry variable description

1 -- alphanumeric title (up to 80 characters). This title will appear
on the output file.
Note:
1. Up to 6 cards (80 characters per card) may be used for the title. The content of
all cards will be printed when this data group is read in. The content of the first
card will be printed as a heading on all subsequent pages of the output file.
7-4
Model Data
7.3 Element Data
7.3.1 *ELTYPE Data Group - Element Type Selection Table

This data group is always required, and must precede the *ELEMENTS data
group. This group selects all types of elements used in the model, for subsequent
reference in the *ELEMENTS data group.
Group ID card: *ELTYPE
Element type selection card set:
Entry No: 1 2 3
Variable: NSRL NKTP NORDR
Max char: 2 3 2

1 NSRL serial number to be used for element identification index (in
ascending order starting with 1, maximum limit of 50).
NISA elements are identified by two variables, NKTP and
NORDR. NKTP indicates the element type number.
NORDR selects the element shape functions (e.g., linear,
parabolic), and hence, the number and location of its nodes.
Element definitions in the *ELEMENTS data group refer to
this NSRL value as a shorthand notation for the NKTP and
NORDR of the element.
2 NKTP element type number - see NISA element library
3 NORDR element order - see NISA element library
Note:
1. The same element NKTP and NORDR values may be specified more than
once, with a different NSRL in this data group. The user may wish to do this so
7-5
Element Data
that, for example, changes to a model are easily recognized. It may also be
useful in plotting, since an option is available in the DISPLAY program to sort
elements by the NSRL values before plotting.
7.3.2 *ELEMENTS Data Group - Element Definition

This data group is always required, provided that the alternate form of the element
definition, the *E1 data group, has not been used. (Note that both the
*ELEMENTS and the *E1 data groups cannot coexist in the same data deck).
Group ID card: *ELEMENTS
Elements may be defined using any combination of three methods: individual

element definition, first level generation and second level generation. The element
generation schemes are exemplified later in the text. Each element is defined by
two sets of cards: the first set consists of one card which defines element
parameters, and also generation parameters for first level generation; the second set
defines the element node list and may consist of more than one card if the element
has more than ten nodes.
For individual element definition, the first card set (one card) will include only
element parameters. For first level element generation, the first card set will
include both element and generation parameters. For second level element
generation, the first card set will include only generation parameters and will be
followed by the appropriate sets of cards defining the base element(s) either
individually or via first level generation. Examples for various methods of element
definitions are given below.
7-6
Model Data
Card set 1: one card
Entry No: 1 2 3 4 5 6 7
Variable: NELID MATID NSRL IDRC KISO NSET NODINC
Max char: 6 6 2 6 1 6 6
Entry No: 8 9
Variable: NELSET NELINC
Max char: 6 6

Entries 1-5 are not needed for second level generation, -
enter zeros or start the card with a tab ($) -
1 NELID element ID number
2 MATID material ID number. This number refers to a material with an
ID number of MATID, defined in the *MATFLUID data
group.
7-7
Element Data

3 NSRL serial number of element identification index (see
*ELTYPE data group).
4 IDRC not used
5 KISO not used
---- Remaining entries on this card are required only for ----
---- first or second level element generation ----
---- tab ($)----
6 NSET the absolute value of NSET indicates the number of
sets to be generated from each element in the base set,
including the base set. A negative value for NSET
activates
second level generation.
7 NODINC node number increment
8 NELSET number of elements in the base set, see notes 2, 3
9 NELINC element number increment
Card set 2: more than one card if the element has more than ten nodes
Entry No: 1 2 3 4-9 10

Variable: NODE1 NODE2 NODE3 ... NODE10
Max char: 6 6 6 ... 6

1 NODE1 first node of the element connectivity list
2 NODE2 second node of the element connectivity list
3 NODE3 third node of the element connectivity list
7-8
Model Data
4 NODE4 fourth node of the element connectivity list (if required)

5 NODE5 fifth node of the element connectivity list (if required)
6 NODE6 sixth node of the element connectivity list (if required)
7 NODE7 seventh node of the element connectivity list (if required)
...
... ... as required...
...
10 NODE10 tenth node (if required), see note 1
Notes:
1. If the element has more than 10 nodes (e.g., the 20-noded solid element), con-
tinue on additional cards in card set 2 to complete the element node list.
2. For first level element generation: (see example below)
- If there is only one element in the base set (NELSET = 1), then only two
sets of cards are required to generate NSET number of elements.
- If there is more than one element in the base set (NELSET > 1), then the
first base element along with the generation parameters will be defined
by the first two card sets. Definition for additional elements in the base
set should follow immediately. The total number of elements generated
will be equal to NELSET*NSET.
3. For second level element generation, card set 1 is used to define the generation
parameters, and should include only entries 6-9, i.e., start the card with a tab
($). Immediately after the definition of the second level generation parameters,
the NELSET number of base elements are defined on subsequent cards using
individual definition or first level generation. See examples below.
Example 7.3.1: Individual element generation
The following cards define one 15-noded wedge element whose element ID is 10,
material ID is 1, identification index is 1, and node list is 1 to 15. Comment cards
are included for explanation.
7-9
Element Data
**
**Card set 1: defines element parameters, no generation parameters needed
**
10, 1, 1, 0, 0
**
**Card set 2: element node list; note that two cards are required since the element has more
than 10
**nodes
1, 2, 3, 4, 5, 6, 7, 8, 9, 10,
11, 12, 13, 14, 15
**
Example 7.3.2: First level element generation
The 15 element mesh shown in Figure7.3.1 is generated using first level

generation. The base set has 3 elements, which are defined individually. Five sets
including the base set are required.
**
**Card set 1: defines element parameters, for the first element in the base set, along with first
level
**generation parameters.
**Five sets are generated including the base set which contains 3 elements
**
1, 1, 1, 1, 0, 5, 1, 3, 1
**
**Card set 2: defines node list for the first base element
**
1, 2, 8, 7
**
**Second element in the base set. Two sets of cards are needed. No generation parameters
are specified.
**
6, 1, 1, 0
7-10
Model Data
7, 8, 14, 13
**
**Third element in the base set.
**
11, 1, 1, 0
13, 14, 20, 19
**
Figure 7.3.1 First level element generation

Example 7.3.3: Second level element generation
The 15 element mesh shown in Figure7.3.2 is generated using second level

generation, 3 sets including the base set are required. The base set has 5 elements,
which are defined using first level generation.
7-11
Element Data
**
**Card set 1: defines second level generation parameters only 3 sets are to be generated from
a base set
**which contains 5 elements
**
$-3, 6, 5, 5
**
**Card set 1: defines the base element parameters and generation parameters for first level
generation
**
1, 1, 1, 0, 0, 5, 1, 1, 1
**
**Card set 2: defines node list for the base element of the first level generation
**
1, 2, 8, 7
**
Figure 7.3.2 Second level element generation
7-12
Model Data
7.3.3 *E1 Data Group - Alternate Form for Element Definition

This data group is required if the standard form for element definition, the
*ELEMENTS data group, is not being used. (Note that both the *E1 and
*ELEMENTS data groups cannot coexist in the same data deck).
Users are advised to use the standard form of element definition, the *ELEMENTS
data group, since this data group (*E1) will not be supported in future versions.
Group ID card: *E1
Elements may be defined using any combination of three methods: individual

definition, first level generation and second level generation. Each element is
defined by one set of cards which consists of one card except if the element has
more than eight nodes. Element definition includes specifying the element node
list, element parameters and generation parameters, if any.
For first level element generation, the first card set will define the first base
element node list and element parameters as well as the generation parameters. The
remaining elements in the base set, if any, should follow immediately with no
generation parameters. For second level generation, the first card set will define
only generation parameters. All elements in the base set should follow
immediately.
Card set 1: one card (except if the element has more than 8 nodes)
$
Entry No: 1 2 3 4-8 9 10
Variable: NELID NODE1 NODE2 ... NODE8 MATID
Max char: 6 6 6 ... 6 4
7-13
Element Data
Entry No: 11 12 13 14 15 16
Variable: NSRL IDRC NSET NODINC NELSET KISO
Max char: 2 4 4 6 3 1

----- Entries 1-12 are not needed for second level
-----
generation,
----- enter zeros or start card with two tabs ($$) -----
1 NELID element ID number
2 NODE1 first node of the element connectivity list
3 NODE2 second node of the element connectivity list
4 NODE3 third node of the element connectivity list (if
required)
5 NODE4 fourth node of the element connectivity list (if
required)
6 NODE5 fifth node of the element connectivity list (if
required)
7 NODE6 sixth node of the element connectivity list (if
required)
8 NODE7 seventh node of the element connectivity list (if
required)
9 NODE8 eighth node of the element connectivity list (if
required), see note 1.
--- tab ($) ---
7-14
Model Data
10 MATID material ID number. This number refers to a

material with an ID number of MATID, defined in
the *MATFLUID data group.
11 NSRL serial number of element identification index (see
*ELTYPE data group).
12 IDRC not used
--- tab ($) ---
----- Entries 13-14 are required only for first -----
----- or second level element generation ----
13 NSET the absolute value of NSET indicates the number of
sets to be generated from the base set, including the
base set. A negative value of NSET activates
second level generation.
14 NODINC node number increment
15 NELSET number of elements in the base set, see notes 2, 3.
16 KISO not used
Notes:
1. If the element has more than 8 nodes, (e.g., the 20-noded solid element), com-
plete the node list on additional cards (8 nodes per card).
2. For first level element generation:
- If there is only one element in the base set (NELSET = 1), then one set of
cards (one card except if the element has more than 8 nodes) will be
required to generate the number of elements equal to NSET.
- If there is more than one element in the base set (NELSET > 1), then only
the first element definition will contain the generation parameters, i.e.,
entries 13-15. Definition for additional elements in the base set, if any,
should follow immediately. The total number of elements generated will
be equal to NELSET * NSET.
7-15
Element Data
3. For second level element generation, card set 1 is used to define the generation
parameters and should include only entries 13-15, i.e. start the card with two
tabs ($$). Immediately after the definition of the second level generation
parameters, start the definition of NELSET base elements using individual or
first level element generation.
7.3.4 *RCTABLE Data Group - Real Constant Table

This data group is required if any element type used in the model requires real
constants (e.g., shell thickness, cross-sectional properties, spring constants, etc.).
Refer to the element library for the required real constants corresponding to the
selected elements in *ELTYPE data group.
Group ID card: *RCTABLE
Card set 1: Real constant index card
Entry No: 1 2 3 4
Variable: IDRC NUMRC IFORM IDCRV
Max char: 6 2 1 6
1 IDRC real constant table identification number. It is the number
referred to in the *ELEMENTS data groups to define the
element real constants.
2 NUMRC number of real constants in this table to be input in the next
card(s).
3 IFORM key for number of entries per card in card set 2
= 0 (or blank)-8 entries per card
> 0-4 entries per card
7-16
Model Data

4 IDCRV temperature dependency curve identification number for
heat transfer analysis only. The temperature dependency
curves are defined in *TEMPFN data group. It is used to
identify the temperature dependency curve for (a) film
coefficient in convection link element (NKTP = 49) or (b)
emmissivity in radiation link element (NKTP = 50). Enter
zero if film coefficient or emmissivity in these elements is
not temperature dependent.
Card set 2: Real constant table card(s), entered according to IFORM
IFORM = 0
Entry No: 1 2 3 4 5 6 7 8
Variable: RC1 RC2 RC3 RC4 RC5 RC6 RC7 RC8
Max char:
OR
IFORM >0
Entry No: 1 2 3 4
Variable: RC1 RC2 RC3 RC4
Max char: 20 20 20 20

1 RC1 first real constant
2 RC2 second real constant
...
... ... as required (number of entries per card must be consistent
.
with IFORM)
...
7-17
Element Data
Notes:
1. Use as many cards as needed in card set 2 to enter NUMRC number of real
constants in this table. Data should be entered according to the IFORM param-
eter (8 entries per card if IFORM = 0 OR 4 entries per card if IFORM > 0).
2. If there is more than one real constant table in this data group repeat the same
card set sequence (card set 1 and 2) to define the additional tables.Group ID
card:
IFORM = 0 OR 4 entries per card if IFORM > 0).
7.3.5 *LAMANGLE Data Group - Rotation Angles for Laminated Composite Ele-
ments
This data group is required if the model includes any composite element (NKTP =
32). Two sets of cards are required for a typical rotation angles table.
Group ID card: *LAMANGLE
Card set 1: Rotation angle index card
Entry No: 1 2
variable: IDRT NUMRT
Max char:

1 IDRT rotation angle table ID (up to 4 digits). This is the
number referred to in *LAMSEQ data group to define
the rotation angles at each node of an element.
2 NUMRT number of rotation angles in this table to be input in
the next card(s).
7-18
Model Data
Card set 2: Rotation angles table card
Entry No: 1 2 3 4 5 6 7 8
Variable: RA1 RA2 RA3 RA4 RA5 RA6 RA7 RA8
Max char: 10 10 10 10 10 10 10 10

1 RA1 rotation angle (in degrees) at node 1
2 RA2 rotation angle (in degrees) at node 2
...
... ... as required...
...
8 RA8 rotation angle (in degrees) at node 8 (if necessary)
Notes:
1. If the element has more than eight nodes, use additional cards as required.
2. Rotation angles are measured with respect to the coordinate system defined
with variable IDRA in *LAMSEQ data group.
3. If there is more than one rotation angles table in this group repeat the same
card set sequence (card set 1 and 2) to define additional tables.
7-19
Element Data
7.3.6 *LAMSEQ data group - Composite lamination sequence data

This group is required if the model includes composite elements (NKTP = 32,).
It should be noted that the input for this data group does not include a table
identification number. However, each data set entered in this group will be given an
identification number by the program, in ascending order starting with one. These
identification numbers must match with the entry MATID in the *ELEMENT data
group.
Group ID card: *LAMSEQ
Lamination sequence card set:
Entry No: 1 2 3 4 5-19 20

variable: NLAY IDRA 11 12 ... 118
Max char: 4 4 4 4 ... 4

1 NLAY number of layers
2 IDRA coordinate system ID for defining rotation angles, see
examples below:
= 0 - right-hand-rule rotation about the normal to the
shell surface, measured from the line connecting
the first two nodes of the element (or from the tan-
gent, at node 1, to the line connecting nodes 1 and
2, if the element has curved sides).
= 1 - right-hand-rule rotation about the global Z-axis,
measured from the global X-axis.
= 2 - right-hand-rule rotation about the global X-axis,
measured from the global Y-axis.
= 3 - right-hand-rule rotation about the global Y-axis,
measured from the global Z-axis.
7-20
Model Data

shell surface, measured from the projection of the
global X-axis onto the shell surface.
global Y-axis onto the shell surface.
global Z-axis onto the shell surface.
= -i - the number i is the local coordinate system ID
(IDSYS) defined in *LCSYSTEM data group,
such that the rotation angle is a right-hand-rule
rotation about the third axis of system i (local z),
measured from the first axis of system i (local x).
Note that the system i is flaged with a negative
sign here.
7-21
Element Data
begin thickness definitions for all layers
thickness table ID number for first (top) layer (up to 4

digits). This number refers to a table with an ID number
of I1, defined in the *RCTABLE data group.
thickness .... table ID number for second layer, if any.
thickness table ID number for last layer.
end thickness definitions
begin rotation angle definitions for all layers
rotation angle table ID number for first (top) layer (up to

4 digits). This number refers to a table with the same ID
number, defined in the *LAMANGLE data group.
rotation angle table ID for second layer, if any.
rotation angle table ID number for last layer.
end rotation angle definitions
begin material property definitions for all layers
material ID number for first (top) layer (up to 4 digits).

This number refers to a material with the same ID
number, defined in the *MATERIAL data group.
material ID number for second layer, if any.
7-22
Model Data
material ID number for last layer.
end material property definitions
Notes:
1. Each card can have a maximum of 20 integer entries. Use additional cards if
required.
2. The layer numbering sequence always starts at the top surface of an element.
Example 7.3.6 (a)
Consider a flat plate lying in the global XY plane and modeled using 8 node
laminated composite shell elements (NKTP = 32, NORDR = 2), as shown in
Figure7.3.3. The plate has 4 layers. The lamination angles, thicknesses and
material specification are shown in the figure. The lamination angles (45/0/90/-45)
are about the global Z-axis and are measured from the global X-axis (i.e., IDRA =
1). The lamination data for this problem are as follows:
Figure 7.3.3 Composite flat plate
7-23
Element Data
**
*LAMSEQ
**NLAY IDRA -thickness IDS- -angles IDS- -material IDS-

**
4, 1, 1,2,2,1, 1,2,3,4, 1,2,2,1
**
*RCTABLE
**thickness table ID no.1
**
1, 8
0.0055, ///////
**thickness table ID no.2
**
2, 8
0.0075, ///////
**
layer 1, = 45o, t = 0.0055", graphite-epoxy
7-24
Model Data
layer 2, = 0o, t = 0.0075", fiberglass
layer 3, = 90o, t = 0.0075", fiberglass
Model Data (*LAMSEQ) NISA Users Manual
Version 90.0 6.2-20 (04/15/90)
*LAMANGLE
**lamination angle table ID no.1
**
1, 8
45.0, ///////
**
2, 8
0.0, ///////
**lamination angle table ID no. 3
**
3, 8
90.0, ///////
**
7-25
Element Data
4, 8
-45.0, ///////
**
*MATERIAL
**material ID no.1, graphite-epoxy
**
EX, 1, 0, 25.0E6
**other material properties for graphite-epoxy
**
**material ID no. 2, fiberglass
**
EX, 2, 9.0E6
**other material properties for fiberglass
**
Example 7.3.6 (b)
To understand the need for more general ways to define fiber angles, consider the
complication that arises in the problem of the previous example when the flat plate
is changed to a cylindrical shell, as shown in Figure7.3.4. The rotation angle of the
top layer is still 45o. However, it is no longer a rotation of 45o about the global Z-
axis measured from the global X-axis. It is a rotation of 45o about the normal to the
shell surface measured from the projection of the global X-axis onto the shell
surface (i.e., the projection of the global X-axis on the plane tangent to the shell
7-26
Model Data
surface at each node). Thus, IDRA may be set to 4 in the previous example to
define the lamination angles for this problem.
Figure 7.3.4 Composite cylindrical shell

It should be noted that IDRA = 0 is similar to IDRA = 4, 5 and 6, in as much as it
describes a rotation about the normal to the shell; however, the reference line from
which angles are measured is associated with each element instead of being
defined globally. If the element has straight sides then the angle is defined as in
Figure7.3.5(b), i.e., the reference line connects the first two nodes of the element. If
the element has curved sides (Figure7.3.5(a)) then the reference line is the tangent,
at node 1, to the line connecting nodes 1 and 2.
These examples of using IDRA = 0 in Figure7.3.5(a) and Figure7.3.5(a) are for flat
plates; the generalization to shells consists in realizing that the rotation angle is
7-27
Element Data
measured from the projection of the reference line defined above onto the tangent
plane to the shell surface at each node.
Figure 7.3.5 Rotation angles for IDRA = 0,
7-28
Model Data
7.4 Nodal Data
7.4.1 *LCSYSTEM Data Group - Definition of Local Coordinate Systems

Local coordinate systems may be specified in this group for the purpose of:
Defining a local coordinate system which may be referenced in the *NODES data
group to specify nodal coordinates and boundary conditions in local coordinate
systems.
Group ID card: *LCSYSTEM
The free format data for a typical local coordinate system consists of one card set
which may contain one or two cards depending on the method of definition.
$
Entry No: 1 2 3 4-12
IDSYS ITYPE METHOD Variable, depends on the value
Variable:
of METHOD
Max char: 6 1 1 max. 10 characters each
Local system card set:

1 IDSYS coordinate system ID, must be 3. Systems 0, 1 and 2
are reserved for global cartesian, cylindrical and spherical
systems, respectively.
2 ITYPE type of local coordinate system
= 0 - local system is cartesian
= 1 - local system is cylindrical
7-29
Nodal Data
3 METHOD coordinate system definition method, see Figure7.4.1

= 0 - defined by 3 node numbers
= 1 - defined by the coordinates of 3 points
= 2 - defined by the local system origin and 3 successive
rotation angles
= 3 - defined by the local system origin and two direction
cosine vectors
--- tab ($) ---
4 - 12 - - definition parameters (see below for explanation). Entries
4 to 9 are entered on this card. Entries 10 to 12 are required
only if METHOD = 1 or 3 and they are entered on a
continuation card which must begin with a tab character
($).
For METHOD = 0 (requires entries 4 - 6)

4 NODE1 a node number given in the *NODES group and defines
the origin of the local system.
5 NODE2 a node number given in the *NODES group and defines a
point on the +x axis of the local system.
6 NODE3 a node number given in the *NODES group and defines a
point on the xy plane of the local system.
For METHOD = 1 (requires entries 4-9 on the first card, entries 10-12 on the
second card)

4-6 XYZ1 global cartesian coordinates (X,Y,Z) of the origin of the
local system.
7-9 XYZ2 global cartesian coordinates (X,Y,Z) of a point on the +x
axis of the local system.
7-30
Model Data
10-12 XYZ3 global cartesian coordinates (X,Y,Z) of a point on the xy

plane of the local coordinate system. Entries 10-12 are
entered on second card which must begin with a tab ($).
For METHOD = 2 (requires entries 4 - 9)

local system.
7-9 THETAX, three successive rotation angles (in degrees) about the
THETAY, global X, updated Y, and final Z axes, respectively, that
THEATZ define the orientation of the local system, see Figure.
6.3.1 (c).
For METHOD = 3 (requires entries 4-9 on the first card, entries 10-12 on the
second card)

local system.
7-9 DIRCOS1 direction cosines of a vector lying on the +x axis of the
local system (3 entries).
10-12 DIRCOS2 direction cosines of a vector lying on the xy plane of the
local system, other than the local x-axis itself (3 entries).
Entries 10-12 are input on second card which must begin
with a tab ($).
7-31
Nodal Data
Notes:
1. Refer to Section 7.4.2 for an example of defining a local coordinate system.
2. If the local coordinate system is cylindrical (ITYPE = 1 on first card) and is
referenced in the *NODES data group to define nodal coordinates, then the
corresponding nodal coordinates must be cylindrical (R, ,Z). The same holds
for a local spherical coordinate system.
3. A local cylindrical (or spherical) coordinate system may be referenced in the
*NODES data group to define both coordinates and local velocity coordinate
systems at more than one node. The local velocity coordinate system directions
will be dependent on the coordinates of each node.
4. The user is not allowed to specify only a particular component of velocity in
local coordinate system. All velocity components (2 for 2-D and 3 for 3-D)
must be specified whenever a local coordinate system is used to specify bound-
ary conditions at nodes.
Figure 7.4.1 Definition of local coordinate systems, (X,Y,Z) = global coordinate

system, (x,y,z) = local coordinate system, 0' = shifted origin: (a) using three nodes,
(b) using coordinates of three points, (c) using shifted origin and three successive
rotations, (d) using shifted origin and two vector {v1} and{v2}.
7-32
Model Data
7.4.2 *NODES Data Group - Nodal Coordinates

This data group is always required. It defines the nodal coordinates and may also
be used to specify local displacement coordinate systems at nodes.
Group ID card: *NODES
Nodal coordinates card set:
ndividual definition, first level generation and second level generation. Second
level generation is activated whenever a base set, defined using individual nodes
definition and/or first level generation, is preceded by the second level generation
card as shown below. Examples for first and second level node generations are also
given below
$ $
Entry No: 1 2 3 4 5 6 7 8
Variable: NODE IDCSYS INDEX NSET X Y Z IDDSYS
Max 1
6 6 6 2 12 12 6
char: 2
7-33
Nodal Data

1 NODE - for individual definition and first level generation, this
is the node number (positive integer up to 6 digits).
- for second level generation, a negative value of
NODE indicates that this card is a second level
generation card. The absolute value of NODE is the
extrapolation increment for the node numbers defined
in the base set. The nodes in the base set are defined
immediately following this card.
2 IDCSYS geometry coordinate system ID (in which the coordinates
of the current node is defined)
= 0 - global cartesian
= 1 - global cylindrical
3- local coordinate system, as defined in *LCSYSTEM
data group
3 INDEX - not used for individual node definition, enter zero
- for first level generation, this is the node number

increment. If a card with INDEX > 0 follows an
individually defined node, then first level generation
will create nodes in increments of INDEX between the
two NODEs and assign coordinates to these nodes,
interpolating the coordinates data for the two NODEs.
- for second level generation, this is the number of

nodes in the base set (maximum of 100). Nodes in the
base set are defined immediately following this
card.
7-34
Model Data

4 NSET - not used for individual definition and first level
generation, enter zero.
- for second level generation, this is the number of sets
to be generated from the base set, including the
base set (maximum of 99).
--- tab ($) ---
5-7 X,Y,Z nodal coordinates referred to the selected coordinate
system (IDCSYS). (For second level generation, these
entries are the extrapolation increments for the
coordinates).
- global cartesian coordinates (X,Y,Z) if IDCSYS = 0
- global cylindrical coordinates (R, , ) if IDCSYS =

1
- consistent with the local coordinate system type if

IDCSYS 3
8 IDDSYS local displacement coordinate system ID.

= 0 - same as global cartesian coordinate system.
= 1 - provided that the node is defined in the global
cylindrical coordinate system (i.e, IDCSYS = 1 also), a
local displacement coordinate system dependent on the
location of the node point will be generated internally.
= 2 - provided that the node is defined in the global
spherical coordinate system (i.e., IDCSYS = 2 also), a
local displacement coordinate system dependent on the
location of the node point will be generated internally.
3 - local coordinate system as defined in the
*LCSYSTEM data group. The local displacement
coordinate system will be dependent on the location of the
node point if the referenced system is cylindrical or
spherical.
7-35
Nodal Data
Notes:
1. For second level generation, the nodes in the base set may be defined individu-
ally or by using first level generation. They may not be defined using first level
generation.
2. All nodes defined using first or second level generation must have the same
local displacement coordinate system ID (IDDSYS).
Example 7.4.1: First level node generation
The six nodes shown in Figure7.4.2 are generated using first level generation.
Nodes 1 and 6 are defined in the global cartesian coordinate system. The
coordinates for the generated nodes (2, 3, 4 and 5) are linearly interpolated between
nodes 1 and 6 in the same coordinate system.
**
**Individual definition for node number 1
1, 0, 0, 0, 0.0, 0.0, 0.0
**
**Definition of node number 6. First level generation is activated between nodes 1 and 6
with node
**increment of 1
6, 0, 1, 0, 5.0, 0.0, 0.0
**
Figure 7.4.2 First level node generation
7-36
Model Data
Example 7.4.2: Second level node generation
The 18 nodes in Figure7.4.3 are generated using second level generation. The base
set is generated via first level generation (individual definition may also be used).
Three sets including the base set are generated.
**
**Define the second level generation parameters as well as the extrapolation increments for
nodal
**coordinates
-10, 0, 6, 3, 0.0, 1.0, 0.0
**
**Define nodes of the base set (nodes 1 to 6). First level generation is used
1, 0, 0, 0, 0.0, 0.0, 0.0
6, 0, 1, 0, 5.0, 0.0, 0.0
**
Figure 7.4.3 Second level node generation in the master set
7-37
Nodal Data
Example 7.4.3: Use of local coordinate systems
Consider the geometry shown in Figure7.4.4 where it is required to define the

coordinates of nodes 1 through 17. It is convenient to define nodes 1-6 in the global
cartesian coordinate system, nodes 7-10 in a local cartesian coordinate system
(IDCSYS = 3) and nodes 13-17 in a local cylindrical coordinate system (IDCSYS
= 4), as shown in the Figure. In addition, it is required that nodes 13-17 have a local
displacement coordinate system, consistent with the local cylindrical system in
which the coordinates are defined. The input data is as follows:
Figure 7.4.4 Geometry and node locations for example 7.4.3

**
*LCSYTEM
**IDSYS ITYPE METHOD origin--- ---angles---
3, 0, 2, 3.0, 1.0, 0.0, 0.0, 0.0, 30.0
4, 1, 2, 1.0, 1.0, 0.0, 0.0, 0.0, 45.0
**
*NODES
**NODE IDCSYS INDEX NSET X Y Z IDDSYS
**nodes 1-6 defined in global cartesian, using first level generation
1, 0, 0, 0, 0.0, 0.0, 0.0, 0
6, 0, 1, 0, 5.0, 0.0, 0.0, 0
7-38
Model Data
** nodes 7-10 defined in local system no.3, using 1st and 2nd level generation
-3, 3, 3, 2, 0.0, 2.0, 0.0, 0
7 3, 0, 0, 0.0, 0.0, 0.0, 0
9 3, 1, 0, 1.0, 0.0, 0.0, 0
**nodes 13-17 defined in local system no.4, using first level generation,
**a local displacement coordinate system, IDDSYS = 4 is specified
13, 4, 0, 0, 0.5, 0.0 0.0, 4
17 4, 1, 0, 0.5, 180.0, 0.0, 4
**
7-39
Material Data
7.5 Material Data
7.5.1 *MATFLUID Data Group - Material Property Data for Fluid

This group is always required
Group ID card: *MATFLUID
Material properties card set
$
Entry No: 1 2 3 4 5 6
Variable: LABEL MATID KTEMP COEF0 COEF1 COEF2
Max
4 6 1 12 12 12
char:
Entry No: 7 8
Variable: COEF3 COEF4
Max
12 12
char:
7-40
Model Data

1 LABEL label for the material property. Allowable labels are:
(a) used for general material property
DENS Density ( )
COND Conductivity (k).
SPEC Specific heat (Cv)
(b) used for direction dependent thermal conductivity
Kxx Thermal conductivity in the direction of x axis
Kyy Thermal conductivity in the direction of y axis
Kzz Thermal conductivity in the direction of z
axis
2 MATID material identification number (must be an integer
between 1 and 999999). This is the material ID referenced
on the element definition group (*ELEMENTS).
3 KTEMP temperature dependence key
= 0 material property is constant

= 1 material property is defined by a polynomial
(maximum order = 4) (also see notes)
>1 property is defined with the help of a base value
(COEF0) and a temperature function described by a
curve with the ID as KTEMP (see *TEMPFN in Sec-
tion 7.6.2).
4 COEF0 constant (base) value of the property.
5 COEF1 coefficient of linear term in the polynomial
expression for variation of property with
temperature.
6 COEF2 coefficient of quadratic term.
7 COEF3 coefficient of cubic term.
8 COEF4 coefficient of quartic term.
7-41
Material Data
Notes:
1. NISA is entirely independent of any system of units. The only requirement is
that the units used must be self consistent.
2. If P = P(T) is any property given as a function of temperature T, then P is
found as follows:
4
i
P ( T ) = 0 +
i T
i=1
where i are the coefficients COEF0 through COEF4.
3. If any of the five coefficients (i.e., COEF0 through COEF4) is not specified,
then it is taken as zero. Thus, with a KTEMP value of 1, if only COEF0 is
specified, it amounts to declaring the material property to be constant.
4. If KTEMP = 0, then even if COEF 1 through COEF4 are specified, these val-
ues are ignored (only COEF0 is taken as constant property)
5. For anisotropic thermal conductivity (i.e. if executive card ANISO = ON,
TEMP is used), Kxx must be specified. If Kyy or Kzz (or both) is not specified,
the value of Kxx is picked up for the unspecified directional property. If COND
is specified for solids, Kxx, Kyy and Kzz is set equal to COND.
6. Anisotropic conductivity can be specified only for solids. For fluids if COND
along with Kxx, Kyy, and Kzz is specified, only COND is taken as conductivity
value. While, if Kxx is specified, COND is set equal to Kxx. Only Kyy or Kzz
(or Kyy and Kzz) can not be specified for fluid conductivity.
7. For heat transfer problems (ANALYSIS = HEAT), conductivity (COND or

Kxx) must be specified for each material type.
8. For transient heat transfer problems (ANALYSIS = HEAT and STDS = OFF or
no STDS card), density (DENS) and specific heat (SPEC) must be specified.
7-42
Model Data
7.5.2 *PRINTCNTL data group - Selective printout control

This card may be used to control whether temperature and fluxes are to be printed
to the output file. The printout of a typical output quantity may be provided in its
entirety, totally suppressed or provided for a subset(s) of nodes or elements. A
default option applies for any output quantity not specified in this group.
Group ID card: *PRINTCNTL
Printout control card set
Entry No: 1 2 3 4 5 6 7 8 9 10
Variable: Label 11 12 13 14 15 16 17 18 19
Max char: 12 6 6 6 6 6 6 6 6 6

1 LABEL Allowable labels are:
Default
option
TEMPERATURE nodal temperatures 0 (all)
HFLT total heat flux on element
boundary face 1(none)
(excluding concentrated nodal
heat fluxes)
HFLC heat flux due to convection on
element 1(none)
boundary face
7-43
Material Data
HFLR heat flux due to radiation on

element boun- dry face 1(none)
NDFLUXES nodal heat flow at nodes with
specified 1(none)
Temperatures.
HEATFLUXES Heat balance sheet
1(none)
(Setting the option to -1 directs the program not to print the output designated by
label in
the output file).
----- tab ($) -----
2 I1 output option or set identification number.
< 0 output designated by LABEL will be sup-
pressed.
= 0 output designated by LABEL will be pro-
vided in its entirety (e.g. element stresses for all
elements).
> 0 output designated by LABEL will be pro-
vided for members of the set with an ID of I1,
which has been defined in the *SETS data
group.
3-6 I2 to I9 additional set ID numbers additional set ID
numbers to complete the definition of the subset
for which the output is requested, valid if I1 > 0
only.
Notes:
1. If more than 9 sets are required to define the subset for a typical output quan-
tity,continue on additional cards. Additional cards must start with a tab ($)
character. Zero set ID numbers are ignored.
7-44
Model Data
7.6 Miscellaneous Data
7.6.1 *SETS Data Group - Definition of Sets

Sets of integer numbers (nodes IDs) may be defined in this data group for
subsequent reference in the *PRINTCNTL data group, for output requests.
Group ID card: *SETS
Sets definition card set
$ $
Entry
1 2 3 4 5 6 7
No:
Variable: IDSET LABEL I1 J1 K1 I2 J2
Max
6 1 6 6 6 6 6
char:
Entry
8 9 10 11
No:
Variable: K2 I3 J3 K3
Max
6 6 6 6
char:

1 IDSET set identification number, integer > 0
7-45
Miscellaneous Data
2 LABEL card identification label (to interpret the remaining

entries on this card). Allowable labels are:
S = entries on this card are single numbers
R = entries on this card define range of numbers
E = entries on this card are single numbers to be
excluded from any ranges of numbers defined in
this set
3 - 11 I1,J1,K1, - I1, J1, K1, ..., K3 are single numbers defined in
..., K3 the set, when LABEL is set to S.
- (I1,J1,K1), (I2,J2,K2), (I3,J3,K3) define three

ranges of numbers beginning with Ii, ending with
Ji and in increment of Ki, for i=1,2,3, respectively,
when LABEL is set to R.
- I1,J1,K1,....,K3 are single numbers to be excluded

from any ranges of numbers defined in this set,
when LABEL is set to E.
Notes:
1. LABEL must not be blank for the first card used to define the set.
2. If LABEL is left blank for any subsequent card (defining the same set), the
LABEL of the previous card is used.
3. Define as many sets as desired. Each set must start with a unique non-zero
IDSET.
4. If any set requires more than one card, continue with additional cards. Addi-
tional cards must start with a tab character $. If any additional card starts with
two consecutive tab characters $$, then the LABEL of the previous card is
used.
5. Once a set is defined, a non-zero value of IDSET on any subsequent card will
begin the definition of a new set, if any.
7-46
Model Data
Example 7.6.1:
Set identification number 101 defines the set of numbers:

1, 4, 8, 20 to 30, 50 to 60, 200 to 220, 1000 to 1030 (in increments of 10)
Set identification number 201 defines the set of numbers:
100 to 150, 152, 153, 155, 157, 163, 171, 182, 183, 191, 195, and excludes the
numbers 101, 109, 140
Input is as follows:
**
*SETS
** IDSET LABEL I1 J1 K1 I2 J2 K2 I3 J3 K3
101, S, 1, 4, 8
$, R, 20, 30, 1, 50, 60, 1 200, 220, 1
$, $, 1000, 1030, 10
**
102, R, 100, 150, 1
$, S, 152, 153, 155, 157, 163, 171, 182, 183, 191
$, $, 195,
$, E, 101, 109, 140,
7-47
Miscellaneous Data
7.6.2 *TEMPFN Data Group - Temperature Dependency Curves

This data group may be used to define temperature dependency curves. These
curves may be referenced in other data groups (e.g., *MATFLUID, etc.) to indicate
the variation of a quantity of interest with respect to temperature. For example, this
includes temperature dependent material properties. Thus, a typical curve is used to
compute the value of a quantity of interest at a given temperature. Two sets of cards
are required for a typical curve. Repeat the same card set sequence for additional
curves, if any.
Group ID card: *TEMPFN
Entry No: 1 2
Variable: IDCURV NPOINT
Max char: 6 2

1 IDCURV temperature curve identification number, integer > 1.
2 NPOINT number of data points used to define the curve.
Card set 2: may require more than one card.
Entry No: 1 2 3 4 5 6 7 8
Variable: T1 P1 T2 P2 T3 P3 T4 P4
Max char: 10 10 10 10 10 10 10 10

1-8 (Ti, Pi) temperature and scaling factor values (entered in pairs) for
a typical point on the curve.
7-48
Model Data
Notes:
1. Use as many cards as necessary to define the NPOINT number of data points
(4 points per card). The data points must be entered in ascending order of tem-
perature.
2. A typical curve should cover the entire range of possible temperature values
encountered in the analysis, i.e. REFT1 T REFT2, where REFT1and
REFT2 are the reference temperatures specified in the *FLCNTL card (section
8.1.1). For temperature values outside the range of *TEMPFN table, a multi-
plier corresponding to the nearest temperature value will be used.
3. Linear interpolation is used to compute the scaling factor for a given
temperature. The exact value of a quantity of interest (e.g., conductivity) at a
given temperature is equal to the scaling factor times the base value given in
the corresponding data group (e.g., *MATFLUID).
4. The temperature axis corresponds to dimensional temperature.
5. Density cannot be specified as a function of temperature.
7.6.3 *TIMEAMP Data Group - Definition of Time Amplitude Curves

This data group may be used to define a time-amplitude curve. This curve may be
referenced in some analysis data groups such as, *BCDVAR, *BCGVAR, etc. The
curve will specify the value of the pertinent quantity at each step or time instant.
See example below.
Group ID card: *TIMEAMP
Two sets of cards are required for each time amplitude curve to be defined. The
first set consists of one card which defines the curve ID and the number of points
on the curve. The second set defines time and amplitude values and may consist of
more than one card if more than four points are specified on the curve.
7-49
Miscellaneous Data
Entry No: 1 2
Variable: IDCURV NPOINT
Max char: 6 2

1 IDCURV curve identification number, integer > 0. This is
the number that will be referenced by other
analysis data groups.
2 NPOINT number of data points used to define the curve.
Card set 2: more than one card if more than four points are specified
Entry No: 1 2 3 4 5 6 7 8
Variable: T1 A1 T2 A2 T3 A3 T4 A4
Max char: 10 10 10 10 10 10 10 10

1-8 (Ti, Pi) pairs of time and amplitude values of a typical
point on the curve. Use as many cards as required
to define the NPOINT data points, four points
per card. Time values must be entered in
ascending order.
Notes:
1. A typical curve should cover all possible range of time values encountered in
the analysis. For time values outside the specified range, an amplitude corre-
7-50
Model Data
sponding to the nearest time values will be used. Amplitudes within the speci-
fied time range will be linearly interpolated.
2. The total value of a quantity will be equal to the instantaneous amplitude value
times the base value of the quantity specified on its corresponding data group.
See example 1 for for details.
3. Default time amplitude curves are internally defined by two points as follows:
(0.0,0.0),(t,1.0), where t = max (1.0, max. time specified by user).
4. The time axis corresponds to dimensional time.

Example 7.6.2:
Time amplitude curve definition
The following cards define the time amplitude curve shown in Figure7.6.1. The
curve identification number is 111 and 7 data points are specified. Analysis data
groups may reference the curve identification number to specify changes in a
specific quantity. Details of the program interpretation for incremental and total
values of the pertinent quantity are shown after the card input.
**
**Card set 1, defines curve ID and number of points.
111, 7
**Card set 2, defines the time-amplitude pairs. Two cards are required since 7 points are
specified
**(4 points per card).
0.0, 0.0, 2.0, 0.0, 5.0, 1.0, 6.0, 1.0
7.0, 0.0, 8.0, 0.0, 10.0, 0.5
**
To explain the interpretation of the program for the above curve, we assume the
following case:
- An analysis group, e.g. *BCDVAR, refers to the above curve ID.
- The base value of the variable, specified on *BCDVAR, is 15 units.
7-51
Miscellaneous Data
- The analysis uses 11 steps or time increments.
Then, the values of the variable interpreted by the program at each step will be as
follows (see note 1 above):
Step
1 2 3 4 5 6 7 8 9 10 11
no.
Value 0.0 0.5 5.0 10.0 15.0 15.0 0.0 0.0 3.75 7.5 15.0
Figure 7.6.1 Time-amplitude curve for example 7.6.2
7-52
Model Data
7.6.4 *CPVAR Data Group - Coupled Variable Data

This data group may be used to couple temperature at a node with that at another
node.
Group ID card: *CPVAR
Coupled variable card set:
Coupled variables sets may be defined using two methods: individual definition
and first level generation as shown below
$
Entry No: 1 2 3 4 5 6
Variable: LABEL NSET NODINC NODSET NODE1 NODE2
Max char: 10 10 10 10 10 10
Entry No: 7-13 14

Variable: ... NODE10
Max char: 10 10

1 LABEL label representing variable. Allowable labels are:
T - temperature
2 NSET - not used in individual definition, enter zero
- for first level generation, this is the number of sets of

coupled variables to be generated from the base set,
including the base set.
7-53
Miscellaneous Data
3 NODINC - not used in individual definition, enter zero
- for first level generation, this is the node number

increment. Additional sets of coupled nodes will be
generated by adding NODINC to the node numbers in
the base set.
4 NODSET number of coupled nodes in this set (or in the base set for
first level generation). This variable may be entered as
zero, see note 1.
- tab ( $ ) -
5 NODE1 first node in this set (the referenced variable at the first
node is the independent DOF)

6 NODE2 second nod
...
... ... as required ...
...
14 NODE10 tenth node in this set
Notes:
1. Up to 10 coupled nodes may be entered on one card. The specification of the
number of coupled nodes (NODSET, entry 4) is optional since the number of
coupled nodes is counted internally. Additional cards needed to define
additional coupled nodes must start with a tab character ($). For example, if
7-54
Model Data
the U velocity is the same at nodes 1 through 11, the input may be entered as
follows:
U $ 1, 2, 3, 4, 5, 6, 7, 8, 9, 10
$ 11
Or alternative as:
U $ 1, 2, 3, 4
$ 5, 6, 7, 8
$ 9, 10, 11
2. The variable of the first node (NODE1) is the independent (master) degree of
freedom; the others are dependent degrees of freedom.
3. A dependent degree of freedom must not be referenced again (as independent
or dependent) in any subsequent coupled variable definition.
4. Dependent degrees of freedom which appear in this data group can not be con-
strained in the *BCDVAR data group.
7-55
Miscellaneous Data
7-56
Chapter
8
Analysis Data
The analysis data for fluid flow is presented in this chapter. The analysis data
generally defines the boundary conditions on the model and selects the output
options. It must follow the model data block in a typical NISA/HEAT data deck as
shown in Figure 5.1.1.
The analysis data block consists of distinct data groups. Each data group consists
of a group identification card followed by its free format data, which may consist
of one or more card sets, each card set may consist of one or more cards. The free
format rules are given in Section 5.2. Each group has a descriptive identification
name indicating the function of the data group. For example, the specified
Dirichlet boundary condition is given in the *BCDVAR data group, and the
Neumann boundary condition is given in the *BCGVAR data group.
A detailed description of the analysis data groups is presented in this chapter in

the sequence shown in Table 8.0.1. This sequence is adopted according to the
function of each data group. For easy reference an alphabetical list of all data
groups available in NISA/HEAT is given in Table . The input data setup for the
above analysis types is discussed in Table 5.1.
The completion of whole analysis data block must be followed by the data deck
terminator, the *ENDDATA group ID card.
8-1
Table 8.0.1 List of analysis data groups
Section No. and Group ID (1) Description

8.1 Analysis Control Data
8.1.1 *FLCNTL Load case control
8.2.1 *BCDVAR Variable Dirichlet boundary
condition
8.2.2 *BCGVAR Variable Neumann boundary
condition
8.2.3 *ICDS Initial condition at nodes
8.2.4 *ICGRAD Neumann type Initial condition
8.2.5 *PCHANGE Specification of phase change
parameters
8.2.6 *STEPSIZE Variable time step size
8.3 Output Control Data
8.3.1 *PRINTCNTL Selective printout control
8.4 Data Deck Terminator
8.4.1 *ENDDATA Input data terminator
(1) Acceptable
minimum abbreviations are in bold face
8-2
Analysis Data
8.1 Analysis Control Data
8.1.1 *FLCNTL Data Group - Fluid Load Case Control

This data group is always required in heat transfer analyses. It defines some
computation keys needed in the analysis, and it must be the first data group in the
analysis set.
Group ID card: *FLCNTL
Heat transfer control card set: one card (if not provided, default values are used)
Entry No 1 2 3 4 5 6 7
Variable: NLSTP ITMAX NTPN DT EPS PENLT REFU
Max char: 4 4 4 8 8 12 8
Entry No 8 9 10
Variable: REFL REFT1 REFT2
Max char: 8 8 8

1 Total number of time steps used in this load case (maxi-
NLSTP
mum is 50)
2 ITMAX Maximum inner iterations for each time step
3 NTPN Printout control key to indicate that at every NTPN time
step, results will be printed in the hard copy output file.
Results at the end of each time step still will be saved in
File 27 for post-processing.
4 DT Value of time step
5 EPS Tolerance for each iteration
6 PENLT not used
8-3
Analysis Control Data
7 REFU not used

8 REFL Reference length
9 REFT1 Reference temperature (lower bound)
10 REFT2 Reference temperature (upper bound)
Notes:
1. Default values for these variables are:
NLSTP = 1, ITMAX = 1, NTPN = 1, DT = 1.0, EPS = 0.001, RFFL = 1.0,
REFT1 = 0, REFT2 = 1.0
2. For steady state problems, steady state analysis option may be used (ref.
EXECUTIVE command STDS). Total number of time steps and the parameter
NTPN should be set to 1. Maximum number of iterations to be used in
obtaining the steady state solution is specified through ITMAX. If the
convergence is not achieved in the given maximum number of iterations,
RESTART(s) may be necessary (ref. EXECUTIVE command RESTART).
Value of time step is meaningless.
3. Steady state problems may also be solved using a pseudo-transient approach.
One may set NLSTP to its maximum value (i.e., 9800) and once the solution
converges, the program will automatically stop.
4. The value of DT is fixed for all time steps.
5. The value of EPS is compared with the solution errors (ref. Section 3.4
between two time steps to check for convergence. Once the solution error goes
below the prescribed EPS, program stops automatically.
6. In case ITMAX > 1, EPS is used to check for solution error between two itera-
tions. If the error is below EPS, the computation is forwarded to the next time
step.
7. REFL is the representative length scale for the problems being analyzed. This
is used to non-dimensionalize the governing equations.
8. If TCUTOFF = ON, REFT1 and REFT2 are lower and upper cutoff tempera-
tures. The program uses them to non-dimensionalize the governing equa-
tions. If the problem is such that exact values of these limits are unknown prior
8-4
Analysis Data
to solving the problem, a conservative estimate should be used (i.e., a wider

range could be specified).
9. In problems having heat flux b.c. and/or internal heat generation (with TCUT-
OFF = ON), the user may encounter a temperature error of 1.0. In such cases,
temperature in the domain is trying to go above the upper cut off temperature
REFT2 specified by the user in the *FLCNTL card. In such cases, REFT2
should be set to a higher value.
10. If TCUTTOFF = ON (Section 6.2) the lower cut off temperature, REFT1,
should also be chosen carefully. It should be the minimum of the prescribed
temperature boundary condition, ambient temperature (in convective
boundary condition) and the temperature of surroundings (in radiation bound-
ary condition). Note that minimum temperature in the domain may still be
lower if some heat sink is present in the domain.
11. In the absence of *ICDS card, REFT1 is taken as the initial temperature in the
domain for unsteady state problems. For steady state problems, REFT1 will be
taken as the initial guess.
12. If NLSTP is set to zero then the program will stop just before getting into the
solution phase. Some valuable information, such as approximate estimates of
total memory and disk space needed, can be obtained at this stage. This may be
particularly useful for large problems.
13. If RESTART feature is used for transient problems, results can be post-pro-
cessed for all the time steps. Results of the previous run will not be over-writ-
ten.
8-5
Boundary Conditions
8.2.1 *BCDVAR Data Group - Variable Nodal Boundary Condition

This data group may be used to define variable nodal boundary conditions at nodes.
Group ID card: *BCDVAR
Specified time dependent Dirichlet boundary card set:
Entry No: 1 2 3 4 5 6
Variable: NODE LABEL P0 LASTND NODINC IDCRV1
Max char: 6 4 12 6 6 6
$
Entry No: 7 8 9 10-11 12
Variable: IDCRV2 LABEL LABEL ----- LABEL
Max char: 6 4 4 -- 4

1 NODE node number
2 LABEL label for degree of freedom (see note). Allowable labels
are:
T temperature
NHS Nodal heat generation
3 P0 specified real value (base value)
4 LASTND last node of a range of nodes with the same specified
value
5 NODINC increment for the range of nodes
6 IDCRV1 > 0 refers to the identification number of any valid time-
amplitude curve defined in the *TIMEAMP data group.
8-6
Analysis Data
7 IDCRV2 is only used when analyzing fluid flow problems with

free surface(s).
tab ($)
8 - 12 LABEL the same specified value can be assigned to the additional
degrees of freedom by specifying the proper labels
Nodal Heat Source:
(i) A concentrated nodal heat source can be applied through this card, in
order to simulate internal heat generation at a node.
(ii) For a 3D model, NHS has the same unit as heat generation rate, e.g.
Watt. However, for a 2D model, it is implied that the NHS is applied
per unit width of the fluid dimension, i.e. for a 2D geometry the unit of
NHS is Watt/m.
(iii) If LABEL = NHS, IDCRV2 is the identification number of any valid

temperature function curve defined in the *TEMPFN data group.
Miscellaneous:
(i) In case of duplication, the last one is considered.
8-7
Boundary Conditions
8.2.2 *BCGVAR Data Group - Variable Elemental Boundary Condition

This data group may be used to specify heat flux, convective and radiative
boundary conditions on the faces (or edges) of elements. For unsteady state
problems, these boundary conditions may be time dependent. The *BCGVAR data
group can also be used to specify surfaces for surface radiation heat transfer.
Element heat generation can also be specified through this card.
Group ID card: *BCGVAR
Card set 1: one card; always required.
Entry No: 1 2 3 4 5 6
Variable: LABEL NELID LASTEL INC IDFACE NCARDS
Max
4 6 6 6 1 2
char:
Entry No: 7 8 9 10 11
IDCRV
Variable: LFN IDCRV2 P0 IDCRV3
1
Max
3 6 6 13
char:

1 LABEL Label indicating the variable. Allowable labels are:
TG heat flux (LFN = -1, 0)
8-8
Analysis Data
data for convection b.c., radiation b.c. and

heat generation (LFN = 1, 2, or 3)
RAD data for radiation surfaces, symmetry
planes and radiation outlets
(LFN = 0 or 1)
2 NELID element number
3 LASTEL last element of a range of elements with the same gradient
boundary condition.Enter 0 or 1 if this card defines a
boundary condition on a single element.
4 INC increment for the range of elements. Enter 0 or 1 if loading
just a single element.
5 IDFACE face number on which the gradient acts (see NISA element
library for face numbering convention).
6 NCARDS Number of cards in card set 2 required to complete the
boundary condition data for the element face. If NCARDS
= 0, only one value is required to complete the boundary
condition data. Parameter P0 has the desired value and the
second card is not required. If LABEL = RAD, NCARDS
is also used to identify radiation surfaces (NCARDS = 1),
symmetry plane (NCARDS = -1) and radiation outlet
(NCARDS = -2). If LABEL = RAD and NCARDS = -1,
P0 is meaningless. If LABEL = RAD and NCARDS = -2,
P0 has the value of temperature of surroundings.
7 LFN this entry has the following interpretation depending on
the integer value and the corresponding label.
Label LFN Description
TG -1 heat flux normal to the element face
0 heat flux along coordinate axes
1 convective heat transfer boundary
condition at the
element face
8-9
Boundary Conditions
2 radiative boundary condition at the

element face
3 heat generation in the element
(IDFACE is ignored)
RAD 0 inward element face is the radiation
surface
1 outward element face is the radiation
surface
8 IDCRV1 > 1 refers to the identification number of any valid time-
amplitude curve defined in the *TIMEAMP data group.
9 IDCRV2 > 1 refers to the identification number of any valid tem-
perature-function curve defined in the *TEMPFN data
group
tab ($)
10 P0 uniform gradient base value in all directions (used if
NCARDS = 0). P0 value is ignored if NCARDS is not
equal to zero.
11 IDCRV3 refers to the identification number of any valid time-
amplitude curve defined in *TIMEAMP data group. Used
for convection boundary condition only.
Entry No: 1 2 3
Variable: P1 P2 P3
Max char: 13 13 13
8-10
Analysis Data
Card set 2: required only if NCARDS > 0

1-3 P1 - P3 gradient values. See notes 1 - 11.
Notes:
Heat Flux Boundary Condition:
(i) Heat flux can be specified normal to the element face (LABEL = TG,
LFN = -1) or along the global axes (LABEL = TG, LFN = 0). If LFN =
0, P1-P3 have values q x , q y and q z . If LFN = -1, P0 has a value of q n .
Heat flux, q n , is taken to be positive along the outward normal to the
element face. Therefore, if heat is to be added to the domain, q n should
be given a negative value.
(ii) Specified heat flux normal to an element face or along the global axes
become time and/or temperature dependent whenever a IDCRV1 > 0
and/or IDCRV2 > 1 is entered, respectively. The value of the heat flux
on an element face or along the global axes at any instant is the heat
flux value specified in this group times the scale factor given on the
temperature function curve times the scale factor given on the time
amplitude curve.
Convective Boundary Condition:
(i) For convective heat transfer boundary condition (LABEL = TG, LFN =
1), two values are required. P1 is the convective heat transfer coeffi-
cient ( h c ) and P2 is the ambient temperature (Tc). These values are
used to input heat flux given by q = hc (T Tc).
(ii) The convective heat transfer coefficient ( h c ) becomes temperature

dependent whenever a IDCRV2>1 is entered. The value of the convec-
tive heat transfer coefficient at any instant is the convective heat trans-
fer coefficient value specified in this group times the scale factor given
8-11
Boundary Conditions
on the temperature function curve. IDCRV3 is the time amplitude curve

id for time dependent heat transfer coefficient.
(iii) The ambient temperature ( T c ) becomes time dependent whenever a

IDCRV1>0 is entered. The value of the ambient temperature at any
instant is the ambient temperature value specified in this group times
the scale factor given on the time amplitude curve.
(iv) Heat transfer coefficient becomes time dependent whenever a IDCRV3

> 0 is entered. The value of heat transfer coefficient at any instant is the
heat transfer coefficient specified in this group times the scaling factor
given on the time amplitude curve.
Radiation Boundary Condition:
(i) For radiative heat transfer boundary conditions (LABEL=TG, LFN =

2), three values are required. P1 is Stefan-Boltzman constant ( ) , P2 is
the temperature of surroundings ( T s ), and P3 is the emissivity ( ) of
the radiation surface. These values are used to input heat flux given by
4 4
q = ( T T s )
(ii) Temperature must be specified in absolute units.
(iii) The emissivity ( ) becomes temperature dependent whenever a

IDCRV2 > 1 is entered. The value of the emissivity value specified in
this group times the scale factor given on the temperature function
curve.
(iv) The temperature of surroundings ( T s ) becomes time dependent when-

ever a IDCRV1 > 0 is entered. The value of the surrounding tempera-
ture at any instant is the surrounding temperature value specified in this
group times the scale factor given on the time amplitude curve.
(v) If due to some factor on temperature function curve, the emissivity

value becomes greater than 1, it is truncated to 1 as emissivity for a sur-
face must always be less than or equal to 1.
8-12
Analysis Data
(vi) The Stefan-Boltzman constant is an universal constant, and must have

same constant value for all the elements.
Heat Generation:
(i) For heat generation (LABEL = TG, LFN = 3), P0 is heat generation rate
(heat/volume) and IDFACE is ignored.
(ii) Specified internal heat generation in an element becomes time and/or

temperature dependent whenever a IDCRV1>0 and/or IDCRV2>1 is
entered, respectively. The value of the heat generation in an element at
any instant is the heat generation value specified in this group times the
scale factor given on the temperature function curve times the scale fac-
tor given on the time amplitude curve.
Surface Radiation:
(i) For radiation surfaces (LABEL = RAD, NCARDS = 1), P1 is Stefan-

Boltzman constant ( ), P2 is temperature of the surroundings ( T s ),
and P3 is emissivity ( ) of the radiation surface. Heat flux at the sur-
face is calculated through radiation energy balance. Temperature of sur-
roundings must be specified in absolute units. Units such as C of F
may be used with TEMP_OFFSET executive command.
(ii) For radiation symmetry plane (LABEL = RAD, NCARDS = -1), LFN
is set internally in the HEAT code. Value of P0 is meaningless. Radia-
tive heat flux is internally set to zero.
(iii) For radiation outlets (LABEL = RAD, NCARDS = -2), LFN is set to 0
by default in DISPLAY III. P0 is the temperature of surroundings.
Radiative heat flux at the surface is calculated internally in the pro-
gram.
(iv) If both radiation boundary condition and surface radiation cards are
present on an element face, radiation boundary condition will be
retained and the element face will be deleted from the list of radiation
surfaces.
8-13
Boundary Conditions
(v) If a surface does not interact with other surfaces in surface radiation
sense, the user is advised to make use of radiation boundary condition
(LABEL=TG, LFN=2). This will lead to reduction in disk space
requirement as well as cpu time required to solve the problem.
(vi) If view factors with obstructions are to be evaluated, then all radiation
surfaces are taken as potential obstructors. Therefore, all possible
obstructors must also be specified as radiation surfaces.
(vii) If radiating surfaces form a complete enclosure, temperature of sur-

roundings is meaningless. For enclosure, the sum of view factors for
each surface should be equal to 1. However, due to the numerical
errors/inaccuracies, the sum may come out to be less than 1 (. 0.98 or
so). This can make solution dependent on the ambient temperature
given in surface definition. One can let IDCRV3 > 1 for a surface to
force the sum of view factors to be 1.
(viii) A radiation outlet cannot be specified as a symmetry plane.
(ix) At restarts, the number of surfaces and surfaces themselves should be

same as those in the previous run. However, for a restart of 3, the user
can change the value of Stefan-Boltzman constant, surface emissivity
and surrounding temperature.
(x) Surface radiation may be switched off or on at any restart.
(xi) Points to be kept in mind when solving 3D and axisymmetric problems.
(a) Accuracy of view factor computations can be enhanced by com-

pleting the enclosure.
(b) If radiation is the primary mode of heat transfer, it is advised to

use finer mesh in order to get accurate values of view-factors.
(c) If obstruction are on, use of finer mesh can enhance the accuracy
of view factor computation.
8-14
Analysis Data
(xii) It is possible to analyze surface radiation effects between disjoined

structures.
(xiii) Radiation is not allowed into elements belonging to solid materials.

Fluids are assumed to be transparent and non-participating.
(xiv) Radiation heat flux, area of radiation surfaces, radiation heat loss/gain
is written to the output file.
(xv) The emissivity ( ) becomes temperature dependent whenever a

IDCRV2 > 1 is entered. The value of the emissivity value specified in
this group times the scale factor given on the temperature function
curve.
(xvi) The temperature of surroundings ( T s ) becomes time dependent when-

ever a IDCRV1>0 is entered. The value of the surrounding temperature
at any instant is the surrounding temperature value specified in this
group times the scale factor given on the time amplitude curve.
(xvii) If due to some factor on temperature function curve, the emissivity

value becomes greater than 1, it is truncated to 1 as emissivity for a sur-
face must always be less than or equal to 1.
Miscellaneous:
(i) In the absence of a valid IDCRV1 and/or IDCRV2, this data group
degenerates to *BCGRAD.
(ii) In case of duplication, the last one is considered.
8-15
Boundary Conditions
8.2.3 *ICDS Data Group - Specified Dirichlet Type Initial Conditions

This data group is used to define specified Dirichlet type initial conditions at
nodes.
Group ID card: *ICDS
Specified Dirichlet boundary card set:
$
Entry No: 1 2 3 4 5 6
Variable: NODE LABEL P0 LASTND NODINC LABEL
Max char: 6 4 12 6 6 4
Entry No: 7 8-9 10

Variable: LABEL ----- LABEL
Max char: 4 -- 4

1 NODE node number
LABEL label for degree of freedom (see note).
2
Allowable labels are:
T temperature
3 P0 specified real value
LASTND last node of a range of nodes with the same
4
specified value
5 NODINC increment for the range of nodes
--- tab ($) ---
8-16
Analysis Data
6 - 10 LABEL the same specified value can be assigned to

the additional degrees of freedom by specify-
ing the proper labels
Note:
1. In case of duplication, the last one is considered.
8.2.4 *ICGRAD Data Group - Neumann Type Initial Conditions

This data group may be used to define gradient type initial conditions on faces (or
edges) of elements, radiation on faces and heat generation in an element.
Group ID card: *ICGRAD
Card set 1: one card; always required.

$
Entry No: 1 2 3 4 5 6 7 8
Variable: LABEL NELID LASTEL INC IDFACE NCARDS LFN P0
Max char: 3 6 6 6 1 1 3 10

1 LABEL Label indicating the variable. Allowable labels are:
TG - gradient of temperature
2 NELID element number
3 LASTEL last element of a range of elements with the same gradient
boundary condition. Enter 0 or 1 if this card defines boundary
condition on a single element.
4 INC increment for the range of elements. Enter 0 or 1 if loading just
a single element
5 IDFACE face number on which the gradient acts (see NISA element
library for face numbering convention).
8-17
Boundary Conditions
6 NCARDS number of cards in card set 2 required to complete the gradient

data for this face. If NCARDS = 0, gradient is assumed to be
uniform with a value of P0 (entry number 8 on this card), and
the second card set is not required in this case.
7 LFN this entry has the following interpretation depending on the inte-
ger value.
= -1 normal gradient to the element face
= 0 gradient along coordinate
--- tab ($) ---
8 P0 uniform gradient value in all directions (used if NCARDS = 0).
U0 value is ignored if NCARDS > 0.
Card set 2: required only if NCARDS > 0
Entry No: 1 2 3 4 5 6 7 8 9
Variable: P1 P2 P3 P4 P5 P6 P7 P8 P9
Max char: 8 8 8 8 8 8 8 8 8

1-9 P1-P9 gradient values at all directions. See notes 1 - 3.
Notes:
1. For temperature gradient (TG) in 3-D problem we need three values. If LFN =
0, P1 - P3 have values TX, TY, TZ. If LFN = -1, P1 - P3 have values Tn, 0, 0.
In 2-D problem, if LFN = 0, P1-P2 have values TX, TY, and if LFN = -1, P1
- P2 have values Tn, 0.
2. In case of duplication, the last one is considered.
8-18
Analysis Data
8.2.5 *PCHANGE
This data group should be used when the phase change effects are to be considered
in the heat transfer analysis with/without fluid flow. The details of this card follow:
Group ID card: *PCHANGE
Entry No.: 1 2 3 4
Variable: MATID LHEAT TSOLDS TLQDS
Max Char.: 6 13 13 13
1 MATID Material ID for the material undergoing phase change

2 LHEAT Latent heat associated with this material
3 TSOLDS Solidus temperature of this material
4 TLQDS Liquidus temperature of this material
Notes:
1. All valid material IDs specified in *MATFLUID data group can be given.
2. The MATID provided under this group will be considered to be undergoing
phase change.
3. The units for this data must be consistent with the rest of the data.
4. LHEAT should be specified per unit mass.
8.2.6 *STEPSIZE
This group is used for transient problems to specify variable time step size.
Group ID card: *STEPSIZE
8-19
Boundary Conditions
Entry No.: 1 2
Variable: NSTEP DT
Max. Char.: 6 13

1 NSTEP number of steps to be run with time step size DT
2 DT time step size
Notes:
1. More than one entry may be given to get variable time step size.
2. Results are saved in 26 and 27 files for the time steps given in this card for a
minimum of 50 steps per run.
3. The program will run for the number of steps specified in the *FLCNTL card.
4. For a step number that is more than the total number of steps specified in the
*STEP card, the last given time step size will be used.
5. For a new run (Restart = 0), the user may specify more number of steps in
*STEP card than the number of steps in the *FLCNTL card. The extra infor-
mation given in *STEP card will be used for restart runs. The users may add
information to *STEP card at restart. For restarts, step number is taken to be
the sum of the step number for the current run and the total number of steps in
the previous runs.
6. Also see executive card STEP.
Example:
*STEPSIZE
10, 0.1
5, 0.2
8-20
Analysis Data
5, 0.5
10, 1.0
20, 2.0
100, 5.0
In the above example, the first 10 steps will be run using a time step size of 0.1.
The next five steps will use a step size of 0.2 and the next 5 will use a step size of
0.5 and so on.0
8-21
Output Control Data
8.3 Output Control Data
8.3.1 *PRINTCNTL data group - Selective printout control

This card may be used to control whether temperature and fluxes are to be printed
to the output file. The printout of a typical output quantity may be provided in its
entirety, totally suppressed or provided for a subset(s) of nodes or elements. A
default option applies for any output quantity not specified in this group..
Group ID card: *PRINTCNTL
Printout control card set:
Entry No: 1 2 3 4 5 6 7 8 9 10
Variable: LABEL I1 I2 I3 I4 I5 I6 I7 I8 I9
Max. Char: 12 6 6 6 6 6 6 6 6 6

1 LABEL Allowable labels are:
Default Option
TEMPER-
- nodal temperatures 0 (all)
ATURE
- total heat flux on element bound-
HFLT ary face (excluding concentrated -1 (none)
nodal heat fluxes)
- heat flux due to convection on ele-
HFLC -1 (none)
ment boundary face
NDFLUX nodal heat flow at nodes with speci-
-1 (none)
ES fied Temperature
HEAT-
- Heat balance sheet -1 (none)
FLUXES
8-22
Analysis Data
(Setting the option to -1 directs the program not to

print the output designated by label in the output
file).
----- tab ($) -----
output option or set identification
2 I1
number
-- output designated by LABEL will
<0
be suppressed
=0 be provided in its entirety (e.g., ele-
ment stresses for all elements)
be provided for members of the set
>0
with an ID of I1, which has been
defined in the *SETS data group.
additional set ID numbers to com-
plete the definition of the subset for
3-6 12 to 19
which the output is requested, valid
if I1 > 0 only.
Note:
1. If more than 9 sets are required to define the subset for a typical output quan-
tity, continue on additional cards. Additional cards must start with a tab ($)
character. Zero set ID numbers are ignored.
8-23
Data Terminator
8.4 Data Terminator
8.4.1 *ENDDATA Data Group - Input Data Terminator

Group ID card: *ENDDATA
The group ID card must be the last card in the data deck.
8-24
Chapter
9
Hints for Using NISA/HEAT
9.1 Introduction
This chapter gives useful hints for running NISA/HEAT. Both modeling and
analysis hints are discussed in this chapter. Modeling hints give information that a
HEAT user should know about DISPLAY III. The analysis hints cover the NISA/
HEAT code itself. In addition, it is also shown how users can run jobs in the
background, save screen messages, monitor convergence etc.
9.2 Running NISA/HEAT

On UNIX based machines, the users can run NISA/HEAT by typing
heat3
The code prompts the user to type in input nisa file name and the output file name.
When the job is done, the user gets back the UNIX prompt. Users can run their
jobs in the background. For example, to run a nisa file named test.nis, user should
type in the following commands in a file, say nisa.go.
heat3>>run.log<<!
test.nis
test.out
y
!
9-1
Running NISA/HEAT
Save the above file and give the following command.
chmod +x nisa.go
In order to run the job in background, type
nisa.go &
All the screen messages get directed to the file run.log. One can use an editor to
view this file, use page command or use tail command. The tail command is
particularly very useful. Type
tail -f run.log
This command will update the contents of the file run.log on screen. One can use
CTRL-C to kill the above tail command. This, however, will not kill the fluids run.
One can also submit a job to be run at a later time using at command. For example,
if one wants to submit batch file nisa.go at 22:00 hrs, type the following at the
UNIX prompt.
at 22:00
nisa.go
CTRL-D
The job will start at 10 p.m. the same day. Use
at -l
to see different jobs that are queued to run at a later time. Use the following
command to remove a job from the queue.
at -r jobid
where jobid corresponds to job one wants to remove from queue and the id can be
obtained by using at -l command.
9-2
9.2.1 Monitoring Convergence

NISA/3D-FLUID prints out the maximum iteration error on screen as well as in the
output file. Users can get a graphical plot of convergence with iterations using a
utility monitor (see Section 3.4). If a job is running in the background, users can
use the egrep command to grep iteration errors from the log file. Users are
advised to create the following alias in the .cshrc file.
alias mklog egrep ERROR|L2NORM|TIME STEP = |DISK run.log >

err
Type
mklog
to grep the errors in run.log file and direct them to a file named err. The user can
edit this file to view the iteration errors.
9.2.2 Managing Disk Space

To run large 3D problems a lot of disk space is required. An estimate of disk space
requirement can be obtained by running a check run. To run a check run, add the
following executive command in the nisa file.
EXECUTION = CHECK
The user may not have all the required disk space available on one disk. However,
the user may have two or more disks available to him and total free space on those
disks may be more than the required disk space to run the problem. Under such
situations, the users are encouraged to make use of executive commands PATH or
PAT1/PAT2/PAT3/PAT4 (see Section 6.2) in their nisa file. These executive
commands can be used to direct temporary files to different directories. Users
should note that files 30 and 31 take up most of the disk space.
9-3
Running NISA/HEAT
9.2.3 Killing a Running NISA/HEAT Job

If a job is running interactively, typing CTRL-C kills the job. If the job is running
in the background, use ps command to find its process id. To kill the job, type
% kill -9 pid
where pid is the process id shown by the ps command.
Note that if a NISA/HEAT job is killed using kill command or using CTRL-C, one
may not be able to post-process the results. If one wishes to post-process results
and at the same time kill the job, one should create a file stopnisa.dat in the
working directory. If the file is present, NISA/HEAT will exit normally save all the
results for post-processing and restarts.
9-4
9.3 Modeling Hints

The results one obtains using the finite element method are approximate as the
continuous physical system is represented by a discrete mathematical model. The
physical system has infinite number of degrees of freedom while the finite element
model has finite number of degrees of freedom. The degree of approximation in the
calculated results depends on how closely the physical system has been modeled.
The modeling process involves many variables: element types and shapes, number
of elements and mesh grading, material properties, applied boundary conditions
etc. Each of these factors may cause some misrepresentation of the actual physical
system. As the degree of nonlinearity in the governing equation (for a given
process) increases, the accuracy of the solution becomes more critically dependent
on the quality of the model.
In the following some guidelines for solving a heat transfer problem using finite
element are given. These are very general in nature and should not be interpreted as
strict rules. However, they should help the user in making a good model the very
first time.
Usually quadrilateral elements would provide better results as compared to an

equivalent number of triangular elements of same interpolating order. How-
ever, there may be geometries where it may be impossible (or at least greatly
difficult) to avoid triangles or wedges. For general triangular and circular
regions (where one is likely to generate triangles), some alternate element pat-
terns using only quadrilaterals are given in Figure 9.3.2.
Element connectivities must be consistent. Correct the connectivities if some
are inconsistent. These operations should be done before any analysis/
boundary condition data are included in the model.
Care must be taken in merging the nodes and a satisfactory boundary line plot
should be checked for. There should be no undesirable cracks in the model.
Also for reasons of compatible interpolations across the inter-element
boundaries, mid-side node of one element must not be connected to a corner
node of another element.
Boundary conditions should be given in sets for easy identification by the user
while inspecting the NISA input file.
Oftentimes idealization of a physical process by a finite element model may
give rise to conflicting boundary conditions at some nodes. It should be noted
9-5
Modeling Hints
that only one set of boundary conditions can be applied at one time to a given
node. Hence suitable choices must be made. Section 9.3.2 shows such an
example.
In NISA/HEAT if similar boundary conditions are applied more than once
then only the last one will be retained.
For axisymmetric models the X-axis of the modelling plane is to be treated as
the axis of symmetry and the model should be contained in the first quadrant
of the X-Y plane. (also refer to the description of NKTP = 3 in Chapter 4).
The users may also use Y-axis as the axis of symmetry using an executive
command; DIME = AX, Y.
In a typical heat transfer problem, appropriate boundary conditions are to be
prescribed on all the boundaries, either using the Dirichlet type (i.e. nodal
temperature etc.) or Neumann type (i.e. heat transfer etc.) of boundary condi-
tions. If no boundary condition is specified, normal gradient of temperature is
assumed zero; this is implicit in the formulation (ref. Chapter 2).
9-6
Figure 9.3.1 Discretization of Triangular and circular domains using only Quadri-
laterals
9-7
Modeling Hints
Figure 9.3.2 Example of Non-unique Corner Boundary Condition
9-8
9.4 Analysis Hints

Phase Change Problems
1. Phase change problems are run as transient problems.

2. Since phase change problems are non-linear, relaxation factor should be used
for temperature. This can be input through the executive card:
STDS = OFF, , ,
where , , has the same meaning as described before (Section 6.2).
3. Small time steps should be used for phase change problem. The executive card
STEP may be used for controlling the time STEP size (Section 6.2).
4. For time accurate solution, enough iterations should be given within each time
step for the temperature (and if applicable, velocity) solution to converge.
Reference Values in *FLCNTL Card
1. In NISA/HEAT3 the analysis is primarily controlled by the executive cards

(ref. Chap. 6) and the Heat Control card (*FLCNTL, ref. Chapter 8). The vari-
ables in these cards have been explained in their respective chapters.
2. Typical reference time Tref should be chosen based on the analysis type. For
forced convection problem, the reference time should be based on
2
C L ref
- where = density, C = specific heat, Lref = reference
T ref = --------------------
k
length, and k = conductivity. The step size DT is non-dimensionalized by the
program with respect to Tref. As such DT should be a fraction of Tref (pro-
vided, of course, Tref indeed represents the most relevant time scale of the
problem). In case of problems with multiple time scales (for example a free
convection problem where convection time scale may be much smaller than
the conduction time scale), the suitable time scale may be limited by the reso-
lution needed for the smaller one.
3. In problems having heat flux b.c. and/or internal heat generation (with TCUT-
OFF = ON), the user may encounter a temperature error of 1.0. In such cases,
temperature in the domain is trying to go above the upper cut off temperature
9-9
Analysis Hints
REFT2 specified by the user in the *FLCNTL card. In such cases, REFT2
should be set to a higher value.
4. If in the executive deck, TCUTOFF = ON, the lower cut off temperature,
REFT1, should also be chosen carefully. It should be the minimum of the pre-
scribed temperature boundary condition, ambient temperature (in convective
boundary condition) and the temperature of surroundings (in radiation bound-
ary condition). Note that minimum temperature in the domain may still be
lower if some heat sink is present in the domain.
Steady State Problems
1. For problems involving non-linearity in energy equations (say temperature

dependent thermal conductivity, or radiation boundary condition), a low relax-
ation factor for temperature ( ) should be used.
Surface Radiation
1. The user is advised to read the surface radiation theory (Section 2.5). Surface
radiation analysis can be controlled through executive cards SRAD and
UFIJ. These cards are described in Section 6.2. The radiation surfaces are spec-
ified in *BCGVAR card (LABEL = RAD). The user is strongly urged to read
the notes given in Section 8.2.2 for surface radiation.
2. NISA/HEAT can internally compute the view factor matrix for the radiation
surfaces defined in the *BCGVAR card. The user can, however, also supply
externally computed view factor matrix to be read NISA/HEAT from an exter-
nal file by setting IUSRVF = 1 in the executive card (Section 6.2).
UFIJ = IUSRVF, IPRVF
Numbering radiation surfaces (for user defined view factors):
Surfaces are assigned numbers 1 through N in the order they appear in the nisa
file. Relevant data such as Stefan-Boltzman constant, surface emissivity and
T are given in the nisa file along with radiation surfaces.
Reading of user view factors:

If IUSRVF 1, NISA/HEAT will read user defined view factors given in vfac-
nisa.vfc instead of computing view factors on its own. The file vfacnisa.vfc
must be present in the current working directory of the user. The following
9-10
FORTRAN subroutine can be used to create view factor input file vfac-
nisa.vfc for NISA/HEAT.
SUBROUTINE VFNISA (NSIZE, NSURF, FIJ)
IMPLICIT NONE
INTEGER*4 NSIZE, NSURF, I, J
REAL*8 FIJ
DIMENSION FIJ (NSIZE, NSIZE)
OPEN (UNIT = 21, FILE = vfacnisa.dat, STATUS = UNKNOWN)
C NSIZE : SIZE OF THE VIEW FACTOR MATRIX
C NSURF : NO. OF SURFACES
C FIJ : VIEW FACTOR MATRIX (KNOWN TO THE USER)
DO 1000 I = 1, NSURF
WRITE (21,2000) (FIJ (I, J), J = 1, NSURF)
2000 FORMAT (5 (2X, E13.6))
1000 CONTINUE
CLOSE 21
RETURN
END
Any other format than the one given in the above subroutine will cause reading
error in NISA/HEAT.
Special Note:
All of the user defined surfaces must be valid surfaces. See Section 8.2.2 for
the definition of invalid surfaces. If any invalid surface is found in the nisa file,
the program will stop. the list of the invalid surfaces can be found in the output
file. The user must remove the invalid surfaces from the list of radiation sur-
faces and supply new view factor matrix in a data file vfacnisa.dat.
Transient Problems
1. For transient problems, if the time step errors oscillate a lot or diverge,
the time step should be reduced. In addition, number of iterations within
each time step may have to be increased.
9-11
Analysis Hints
2. In NISA/HEAT steady state problems can also be solved via (pseudo)

unsteady approach. The solution is progressed just like a transient prob-
lem and when the steady state solution is reached, temperature does not
change with time. Thus for a steady state solution any suitable time step
could be used along with one or more iterations per time step. If the
steady state solution is not achieved within the user specified number
time steps (maximum value of NLSTP in *FLCNTL card), one may use
the RESTART capability until the solution converges.
3. For unsteady problems, initial conditions are very critical (everything
else remaining same, different initial conditions will produce different
solutions).
4. The user can subdivide the time step given in *FLCNTL card into still
smaller time steps using the executive command STEP = N where N is
the number of subdivisions of the time step. In DISPLAY, the user will be
able to look at the results at time steps given in the *FLCNTL card. How-
ever, NISA/HEAT will use smaller step size (DT/N) for time accurate
solution.
9-12
Appendix
A
Appendix A: Verification Problem
Verification Problem No.A.1
Title:
One dimensional heat conduction in a bar
Element Type:
3-D eight node solid element (NKTP = 4, NORDR = 1)
Analysis Highlights:
Linear steady state heat transfer analysis.
Problem:
A 1.0 m long bar with a square cross section (0.2 m x 0.2 m) is subjected to a flux
of 1000 W/m2 at one end. The other end of the bar is fixed at 0C. All other
surfaces are insulated.
The temperature distribution at steady state is calculated.
A-1
Properties:
Material:
KXX = KYY = KZZ = 200 W/mC (Thermal conductivity).
Finite Element Model:
The bar is modeled using five equal size 8 node solid elements (Figure A.1.1). The
flux load is specified as a surface flux on the element face at X = 1.0 m. The nodes
at X = 0 are fixed at 0C temperature.
Results and Comparison:
The nodal temperatures along the bar axis are listed in Table A.1.1. The
temperature varies linearly from 5C (at X = 1m) to 0C (at X = 0) along the length
of the bar. Using one dimensional steady state heat conduction equation [1], the
theoretical solution is T = 5X. Results obtained by NISA compare exactly.
Reference:
1. J. P. Holman, Heat Transfer, McGraw-Hill Book Company, Fifth edition
(1981).
Table A.1.1 Comparison of temperature distribution along the bar axis
X meters 0.2 0.4 0.6 0.8 1.0

Theory [1] (C) 1.0 2.0 3.0 4.0 5.0
NISA (C) 1.0 2.0 3.0 4.0 5.0
A-2
Figure A.1.1 One dimensional heat conduction in a bar.
A-3
Input Data for Verification Problem No. A.1
**EXECUTIVE data deck for HEAT

ANALYSIS = HEAT
DIMENSION = 3D
ELEMENT_ECHO = OFF
FILE = heat01
SAVE=26,27,54
NODEECHO = OFF
STDS = ON,1.0,1.0,1.0
SOLU=ITERATIVE,200,0.1E-03,1
*TITLE
ONE DIMENSIONAL HEAT CONDUCTION IN A BAR.
LINEAR STEADY STATE HEAT TRANSFER ANALYSIS.
*ELTYPE
1, 4, 1
*NODES
1,,,, 0.00000E+00, 0.00000E+00, 0.00000E+00, 0
2,,,, 2.00000E-01, 0.00000E+00, 0.00000E+00, 0
3,,,, 4.00000E-01, 0.00000E+00, 0.00000E+00, 0
4,,,, 6.00000E-01, 0.00000E+00, 0.00000E+00, 0
5,,,, 8.00000E-01, 0.00000E+00, 0.00000E+00, 0
.. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. ..
.. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. ..
33,,,, 4.00000E-01, 0.00000E+00,-2.00000E-01, 0
34,,,, 6.00000E-01, 0.00000E+00,-2.00000E-01, 0
35,,,, 8.00000E-01, 0.00000E+00,-2.00000E-01, 0
36,,,, 1.00000E+00, 0.00000E+00,-2.00000E-01, 0
*ELEMENTS
1, 1, 1, 0, 0
1, 2, 32, 31, 11, 12, 22, 21,
2, 1, 1, 0, 0
2, 3, 33, 32, 12, 13, 23, 22,
3, 1, 1, 0, 0
3, 4, 34, 33, 13, 14, 24, 23,
A-4
4, 1, 1, 0, 0
4, 5, 35, 34, 14, 15, 25, 24,
5, 1, 1, 0, 0
5, 6, 36, 35, 15, 16, 26, 25,
*MATFLUID
DENS, 1,0, 1.00000E+00,
COND, 1,0, 2.00000E+02,
SPEC, 1,0, 1.00000E+00,
*FLCNTL, ID = 1
** last two entries are lower and upper
** cut-off temperatures
1,1,1,1.0,0.001,0.1E+09,1.0,1.0,0.0,10.0
*BCDVAR
** BCDVAR SET = 1
** specified temperature boundary condition
1,T , 0.00000E+00,,, 0, 0
11,T , 0.00000E+00,,, 0, 0
21,T , 0.00000E+00,,, 0, 0
31,T , 0.00000E+00,,, 0, 0
*BCGVAR
** BCGVAR SET = 1
** heat flux normal to element face
** negative value indicates heat being added to the
** system
TG , 5,,,4, 0, -1, 0, 0, -0.10000E+04
*ENDDATA
A-5
Title:
Heat conduction in a square plate with prescribed temperature
Element Type:
2-D four node planar element (NKTP = 2, NORDR = 1)
Linear steady state heat transfer analysis with prescribed temperature conditions.
Problem:
A square plate (1.0 in x 1.0 in) with unit thickness has one side at 100F
temperature. Other three sides are at 0F temperature. All four sides and the two
faces of the plate are insulated.
The temperature distribution in the plate at steady state condition is calculated.
Properties:
Material:
KXX = KYY = 1.0 BTU/sec.in.0F (Thermal conductivity)

The plate is modelled using 64 (8x8) four node plane elements (Figure A.2.1). All
nodes on the edges (X = 0, 1.0 in. and Y = 1.0 in.) are specified at 0F. Nodes on
the edge Y = 0 are fixed at 100F (except the extreme ones at X = 0, and X = 1.0
in).
A-6

The isotherms for the calculated temperature distribution are shown in Figure
A.2.1. The theoretical solution for this problem is given by the following Eqn. [1].
( 2n + 1 )x ( b y ) ( 2n + 1 )
sin --------------------------- sinh -----------------------------------------
4T b 1 a a
T ( x, y ) = ---------
-------------------- ----------------------------------------------------------------------------------------
( 2n + 1 ) ( 2n + 1 )b
sinh ---------------------------
n=0 a
where Tb (= 100F) is the temperature at Y = 0.0 boundary and 'a' (= b = 1.0 inch)
is the size of the square plate. Isothermal contour lines are shown in Figure A.2.2
and the temperature for the nodes at X = 0.5 inch are compared in Figure A.2.3.
Excellent agreement between NISA and theoretical results is evident.
Reference:
1. H. S. Carslaw, J. C. Jaeger, "Conduction of Heat in Solids," Second Edi-
tion, Oxford University Press (1959).
A-7
Figure A.2.1 Heat conduction in a square plate with prescribed

temperature boundary conditions
A-8
Figure A.2.2 Isotherms for square plate with prescribed temperature

boundary conditions
A-9
Figure A.2.3 Comparison of Temperature Distribution at X = 0.5 inch
A-10

ANALYSIS = HEAT
ELEMENT-ECHO = OFF
FILE = heat02
SAVE=26,27,54
NODEECHO = OFF
RESEQUENCE = OFF
STDS = ON, 1.0, 1.0, 1.0
SOLU=ITERATIVE,200,0.1 E-03,l
*TITLE
HEAT CONDUCTION IN A SQUARE PLATE WITH PRESCRIBED TEMPERA-
TURE
*ELTYPE
1, 2, 1
*NODES
1 ,,,, 0.00000E+00, 0.00000E+00, 0.00000E+00,0
2 ,,,, 1.25000E-0 1, 0.00000E+00, 0.00000E+00,0
3 ,,,, 2.50000E-0 1, 0.00000E+00, 0.00000E+00,0
4 ,,,, 3.750008-0 1, 0.00000E+00, 0.00000E+00,0
*ELEMENTS
1, 1, 1, 0, 0
1, 2, 1 1, 1 0
2, 1, 1, 0, 0
2, 3, 12, 11,
3, 1, 1, 0, 0
3, 4, 13, 12
.. .. .. .. .. .. .. ..
.. .. .. .. .. .. .. ..
63, 1, 1, 0, 0
70, 71, 80, 79,
A-11
64, 1, 1, 0, 0
71, 72, 81, 80,
*MATFLUID
** material properties
DENS, 1,0, 1.00000E+00,
COND, 1,0, 1.00000E+00,
SPEC, 1,0, 1.00000E+00,
*FLCNTL, ID = 1
** heat control parameters
1,1,1,1.0,0.001,0.1E+09,1.0,1.0,0.0,100.0
*BCDVAR
** BCDVAR SET = 1
l,T , 0.00000E+00,,,0, 0
9,T , 0.00000E+00,,,0, 0
10,T , 0.00000E+00,,,0 ,0
18,T , 0.00000E+00,,,0, 0
.. .. .. .. .. .. .. .. .. ..
.. .. .. .. .. .. .. .. .. ..
79,T , 0.00000E+00,,,0, 0
80,T , 0.00000E+00,,,0, 0
81,T , 0.00000E+00,,,0, 0
** BCDVAR SET = 2
2,T , 1.00000E+02,,, 0, 0
3,T , 1.00000E+02,,, 0, 0
4,T , 1.00000E+02,,, 0, 0
5,T , 1.00000E+02,,, 0, 0
6,T , 1.00000E+02,,, 0, 0
7,T , 1.00000E+02,,, 0, 0
8,T , 1.00000E+02,,, 0, 0
*ENDDATA
A-12
Verification Problem No.A.3 cribed temperature
Title:
Heat conduction in a long hollow cylinder with prescribed temperature
Element Type:
Axisymmetric four node solid elements (NKTP = 3, NORDR = 1)
Linear steady state heat transfer analysis with prescribed temperatures.
Problem:
A long hollow cylinder is subjected to an inside temperature of 100F and an
outside surface temperature of OF. The cylinder is insulated to prevent axial heat
flow. The inner and outer radii are 1 and 2 inches, respectively, see Figure A.3.1.
The steady state temperature distribution along the radial direction is calculated
and compared with theoretical solution.
Properties:
Material:
KXX = 100.0 BTU/sec.in.oF (Thermal conductivity)

Since the temperature distribution is not a function of the axial direction, an
arbitrary length of 0.5 inch is modeled by 10 four node axisymmetric elements
(NKTP = 3), see Figure A.3.1. The two nodes at the inner radius (X = 1.0 in) are
specified at 100F and the two nodes at outer radius (X=2.0 in) are specified at 0F.
A-13
Results and Comparisons:

The theoretical temperature distribution along an arbitrary radius of the cylinder is
given as [Ref. 1];
Ti T0 r
T ( r ) = T i ----------------- ln ---
ro r i
ln ----
ri
where Ti (= 100F) and To (= 0F) are temperatures at the inner and the outer radii,
respectively.
The calculated temperature distribution along the radius is compared with the
theoretical values in Table A.3.1. An excellent agreement with the theoretical
values is achieved.
Reference:
1. J. P. Holman, "Heat Transfer," McGraw-Hill Book Company,Fifth edition
(1 963).
Table A.3.1 Comparison of nodal temperature along the

radius of the cylinder
Radial dist (in) 1.0 1.2 1.5 1.8 2.0
o
Theory [1] ( F) 100.000 73.700 41.504 15.200 0.0000
NISA o 100.000 73.703 41.511 15.203 0.0000

( F)
A-14
Figure A.3.1 Long Hollow Cylinder with Prescribed

Temperature Boundary Conditions
A-15

ANALYSIS = HEAT
DIMENSION = AX,Y
ELEMENT_ECHO = OFF
FILE = heat03
SAVE=26,27,54
NODEECHO = OFF
STDS = ON, 1.0, 1.0, 1.0
*TITLE
HEAT CONDUCTION IN A LONG HOLLOW CYLINDER WITH PRE-
SCRIBED TEMPERATURE
*ELTYPE
1, 3, 1
*NODES
1,,,, 1.00000E+00, 0.00000E+00, 0.00000E+00,0
2,,,, 1.00000E+00, 5.00000E-0 1, 0.00000E+00,0
3,,,, 1.10000E+00, 0.00000E+00, 0.00000E+00,0
4,,,, 1.10000E+00, 5.00000E-0 1, 0.00000E+00,0
... .. .. .. .. .. .. .. .. .. .. ..
.. .. .. .. .. .. .. .. .. .. .. .. .. ..
20,,,,1.90000E+00,5.00000E-01, 0.00000E+00,0
21,,,,2.00000E+00,0.00000E+00, 0.00000E+00,0
22,,,,2.00000E+00,5.00000E-01, 0.00000E+00,0
*ELEMENTS
1, 1, 1, 0, 0
1, 3, 4, 2,
2, 1, 1, 0, 0
3, 5, 6, 4,
.. .. .. ..
.. .. .. ..
A-16
9, 1, 1, 0, 0
17, 19, 20, 18,
10, 1, 1, 0, 0
19, 21, 22, 20,
*MATFLUID
DENS, 1,0, 1.00000E+00,
COND, 1,0, 1.00000E+02,
SPEC, 1,0, 1.00000E+00,
*FLCNTL, ID = 1
1,1,1,1.0,0.001,0.1E+09,1.0,1.0,0.0,100.0
*BCDVAR
** BCDVAR SET = 1
**specified temperature boundary condition
l,T , 1.00000E+02,,, 0, 0
2,T , 1.00000E+02,,, 0, 0
** BCDVAR SET = 2
**specified temperature boundary condition
21,T , 0.00000E+00,,, 0, 0
22,T , 0.00000E+00,,, 0, 0
*ENDDATA
A-17
Title:
Column with varying convective boundary condition on its faces
Element Type:
2-D four node planar elements (NKTP = 2, NORDR = 1)
Linear steady state heat transfer analysis with convection boundary condition.
Problem:
A column with a rectangular cross section (14 in x 36 in) is built in a wall of a

building with part of the column exposed from both sides of the wall, see Figure
A.4.1. The column surfaces facing interior of the building is exposed to 100F. The
outside environment is at OF ambient temperature. The variation of the heat
transfer film coefficient on the faces of the column is shown in Figure A.4.1.
Steady state temperature distribution in the column cross section is calculated.
Properties:
Material:
KXX = 1.0 BTU/hr.ft.OF = 0.0833333 BTU/hr.in.OF (Thermal conductivity)
Only one half the column cross section is modelled. The area in which the
convective coefficient is not constant has more elements. Totally 96 four node
elements are used in the mesh (Figure A.4.2). The ambient temperature and the
convective film coefficients are specified at the appropriate element faces (edges).
A-18
The center line temperature over the width of the column (from X = 0 to X = 36 in)
is plotted in Figure A.4.3. The results are compared with the results of a Fourier
analysis solution by Peavy [I] where good agreement is shown.
Reference:
1. B. A. Peavy, "Steady-State Heat Conduction on Exposed Exterior of

Rectangular Cross Section", J Res. Nat. Bureau of Standards, 69 (1965)
145- 15 1.
Figure A.4.1 Problem Geometry and Boundary Conditions
A-19
Figure A.4.2 Finite Element Mesh with Node and Element Numbers
A-20
Figure A.4.3 Temperature Distribution Comparison for

Column Cross Section
A-21

ANALYSIS = HEAT
ELEMENT-ECHO = OFF
FILE = heat04
SAVE=26,27,54
NODEECHO = OFF
STDS = ON, 1 .0, 1 .0, 1.0
*TITLE
COLUMN WITH VARYING CONVECTIVE BOUNDARY CONDITIONS
ON FACES
*ELTYPE
1,2, 1
*NODES
1 ,,,, 0.00000E+00, 0.00000E+00, 0.00000E+00,0
2 ,,,, 2.00000E+00, 0.00000E+00, 0.00000E+00,0
3 ,,,, 4.00000E+00, 0.00000E+00, 0.00000E+00,0
4 ,,,, 6.00000E+00, 0.00000E+00, 0.00000E+00,0
.. .. .. .. .. .. .. .. .. .. .. .. .. .. ..
.. .. .. .. .. .. .. .. .. .. .. .. .. .. ..
123,,,, 3.20000E+01,7.00000E+00, 0.00000E+00,0

124,,,, 3.40000E+01,7.00000E+00, 0.00000E+00,0
125,,,, 3.60000E+01,7.00000E+00, 0.00000E+00,0
ELEMENTS
1, 1, 1, 0, 0
1, 2, 27, 26,
2, 1, 1, 0, 0
26, 27, 52, 51,
.. .. .. .. ..
.. .. .. .. ..
A-22
95, 1, 1, 0, 0
74, 75, 100, 99,
96, 1, 1, 0, 0
99, 100, 125, 124,
*MATFLUID
DENS, 1,0, 1.00000E+00,
COND, l,O, 8.333338-02,
SPEC, 1,0, 1.00000E+00,
*FLCNTL, ID- 1
1,1,1,1.0,0.001,0.1E+09,1.0,1.0,0.0,100.0
*BCGVAR
** convection boundary conditions
TG, 1, 4, 1, 4,1,1,0,0
3.47220E-03 1.00000E+02 0.00000E+00
TG, 1, 21, 4, 1,1,1,0,0
3.47220E-03 1.00000E+02 0.00000E+00
TG, 93, 93, 1, 1,1,1,0,0
4.16667E-02 0.00000E+00 0.00000E+00
TG, 93, 96, 1, 2,1,1,0,0
4.16667E-02 0.00000E+00 0.00000E+00
*ENDDATA
A-23
Verification Problem No.A.5 Heat conduction analysis in an infinitely

long slab.
Title:
Heat conduction analysis in an infinitely long slab
Element Type:
Transient heat transfer analysis with specified initial temperature.
Problem:
An infinitely long slab with a 10 in x 1.0 in cross section is at an initial temperature

of 100F. The two ends of the slab are maintained at 0F.
Transient heat conduction analysis is performed to obtain the spatial temperature

distribution within the structure.
Properties:
Material:
KXX = 1.0 BTU/sec.in.F (Thermal conductivity)

C = 1.0 BTU/lb .F (Specific heat)
DENS = 1.0 lb/in3 (Density)
Since the geometry is infinite in Z-direction, any cross-section of the structure will
have the same temperature distribution.
A-24
Twenty 4-noded plane heat transfer elements (NKTP = 2, NORDR = 1) are used to
model one such cross-section, see Figure A.5.1. The initial temperature of the
structure is specified as 100F and the end nodes (1, 31, 21 and 51) have a
prescribed temperature value of 0F.
Solution Procedure:
Transient heat transfer analysis is performed using Crank-Nicholson time

integration scheme. Eighty equal time step sizes of t = 0.25 sec. are used.

Spatial temperature distribution results of the analysis are presented in Figure
A.5.2 for time = 4, 8, 12, 16 and 20 seconds. These results are compared with the
theoretical solution [1] and are in excellent agreement.
Reference:
1. H. S. Carslaw, and J. C. Jaeger, "Conduction of Heat in Solids," Second

edition, Oxford University Press (1959).
A-25
Figure A.5.1 Slab Cross Selection Model with Node and Element Numbers
Figure A.5.2 Temperature Distribution in the Slab Cross section

at various time instants
A-26

ANALYSIS = HEAT
ELEMENT-ECHO = OFF
FILE = heat05
SAVE=26,27,54
NODEECHO = OFF
** set initial temperature to 100.0
INITIAL-CONDITION = 100.0,T
** subdivide time step specified in *FLCNTL card in 4 steps
STEP = 4
** select iterative solver
*TITLE
HEAT CONDUCTION ANALYSIS IN AN INFINITELY LONG SLAB WITH
SPECIFIED TEMPERATURES AT BOTH ENDS
TRANSIENT HEAT CONDUCTION PROBLEM
*ELTYPE
1, 2, 1
*NODES
1 ,,,, 0.00000E+00, 0.00000E+00, 0.00000E+00,0
2 ,,,, 5.00000E-0 1, 0.00000E+00, 0.00000E+00,0
3 ,,,, 1.00000E+00, 0.00000E+00, 0.00000E+00,0
4 ,,,, 1.50000E+00, 0.00000E+00, 0.00000E+00,0
.. .. .. .. .. .. .. .. .. .. .. ..
.. .. .. .. .. .. .. .. .. .. .. ..
49 ,,,, 9.00000E+00, 1.00000E+00, 0.00000E+00,0
50 ,,,, 9.50000E+00, 1.00000E+00, 0.00000E+00,0
51 ,,,, 1.00000E+00, 0.00000E+00, 0.00000E+00,0
*ELEMENTS
1, 1, 1, 0, 0
1, 2, 32, 31,
2, 1, 1, 0, 0
2, 3, 33, 32,
.. .. .. .. ..
A-27
.. .. .. .. ..
19, 1, 1, 0, 0
19, 20, 50, 49,
20, 1, 1, 0, 0
20, 21, 51, 50,
*MATFLUID
DENS, 1,0, 1.00000E+00,
COND, 1,0, 1.00000E+00,
SPEC, 1,0, 1.00000E+00,
*FLCNTL, ID= 1
20,1,20,1.0,0.001,0.1E+09,1.0,1.0,0.0,100.0
*BCDVAR
** BCDVAR SET = 1
l,T , 0.00000E+00,,, 0,0
21,T , 0.00000E+00,,, 0,0
31,T , 0.00000E+00,,, 0,0
51,T , 0.00000E+00,,, 0,0
*ENDDATA
A-28
Verification Problem No.A.6 Transient heat conduction through a wall
Title:
Transient heat conduction through a wall
Element Type:
Nonlinear transient heat transfer analysis with temperature dependent material
properties and time dependent specified temperatures
Problem:
An infinitely long slab with 20 cm x 1.0 cm cross section is at an initial
temperature of 100C. The left and right end faces have time dependent prescribed
temperatures. The thermal conductivity is a polynomial function of temperature
whereas the specific heat and the density are constant.
Transient heat transfer analysis is performed to obtain the spatial temperature

distribution in the structure.
Properties:
Material:
KXX = Ko+aT = 2.0 + (0.01)T W/cm C, (Thermal conductivity)

C = 1.0 J/KgC, (Specific heat)
DENS = 8.0 Kg/cm 3, (Density)
A-29
Since the geometry is infinite in Z-direction, any cross-section of the structure will
have the same temperature distribution.
Twenty 4-node plane heat transfer elements (NKTP = 1, NORDR = 1) are used to
model one such cross-section (Figure A.6.1). The end nodes (1, 2, 41, and 42) have
a prescribed temperature values which are time dependent (Figure A.6.1). The
initial temperature of the structure is specified to be 100C.
Solution Procedure:
As the material properties are temperature dependent, nonlinear analysis is

performed using Newton-Raphson iterative technique. Transient analysis using
Crank-Nicholson time integration scheme with fifteen equal time steps (1.0 sec)
each is performed. Within each time step a maximum of 25 iteration are allowed.
The effective conductivity matrix is reformulated every tenth iteration.

Spatial temperature distributions at 10 seconds and 13 seconds for X = 0 to 10 cm
is shown in Table A.6.1. These results are compared with results by Orivuori [I]
showing excellent agreement.
Reference:
1. S. Orivuori, "Efficient Method for Solution of Nonlinear Heat Conduc-

tion Problems," Int. J. Num. Meth. Engng, 14(1979) 1461 -1476.
A-30
Table A.6.1 Comparison of temperature distribution (C)
X Time=10 seconds Time=13 seconds

(cm) NISA Reference [1] NISA Reference [1]
0.0 200.000 200.000 100.000 100.000
1.0 176.161 176.165 129.407 128.525
2.0 153.205 153.213 140.732 139.970
3.0 133.459 133.467 137.721 136.951
4.0 118.596 118.600 124.820 124.719
5.0 108.985 108.985 114.146 114.140
6.0 103.718 103.715 107.183 107.182
7.0 101.294 101.292 103.243 103.241
8.0 100.370 100.368 101.292 101.290
9.0 100.081 100.083 100.449 100.448
10.0 100.014 100.014 100.134 100.134
A-31
Figure A.6.1 Finite Element Mesh and Boundary Conditions for

Wall Cross Section
A-32

ANALYSIS = HEAT
ELEMENT-ECHO = OFF
FILE = heat06
SAVE=26,27,54
NODEECHO = OFF
STEP = 4
*TITLE
TRANSIENT HEAT CONDUCTION THROUGH A WALL WITH NONLINEARITY
IN CONDUCTIVITY
BOUNDARY TEMPERATURES OF THE LEFT AND RIGHT END SURFACES ARE
TIME DEPENDENT
*ELTYPE
1, 2, 1
*NODES
1 ,,,, 0.00000E+00, 0.00000E+00, 0.00000E+00,0
2 ,,,, 0.00000E+00, 1.00000E+00, 0.00000E+00,0
3 ,,,, 1.00000E+00, 0.00000E+00, 0.00000E+00,0
.. .. .. .. .. .. .. .. .. .. .. ..
.. .. .. .. .. .. .. .. .. .. .. ..
40 ,,,, 1.90000E+00, 1.00000E+00, 0.00000E+00,0
41 ,,,, 2.00000E+00, 0.00000E+00, 0.00000E+00,0
42 ,,,, 2.00000E+00, 1.00000E+00, 0.00000E+00,0
ELEMENTS
1, 1, 1, 0, 0
1, 3, 4, 2,
2, 1, 1, 0, 0
3, 5, 6, 4,
.. .. .. .. ..
.. .. .. .. ..
19, 1, 1, 0, 0
A-33
37, 39, 40, 38,

20, 1, 1, 0, 0
39, 41, 42, 40,
*MATFLUID
DENS, 1,0, 8.00000E+00,
COND, 1 , 1 , 2.00000E+00, 1.00000E-02, 0.00000E+00, 0.00000E+00,
0.00000E+00,
SPEC, 1,0, 1.00000E+00,
*TIMEAMP
1 ,6
0.0,200.0,5.0,200.0,10.0,200.0,10.01,100.0
12.0,100.0,25.0,100.0
2,5
0.0,1.0,5.0,1.0,9.9999,1.0,10.0001,1.0
100.0,1.0
**FLCNTL, ID= 1
15,10,15,1.0,1.Oe-20,0.1E+09,1.0,1.0,0.0,200.0
**BCDVAR
** BCDVAR SET = 1
l,T , 1.00000E+00,,, 1, 0
2,T , 1.00000E+00,,, 1, 0
** BCDVAR SET = 2
41,T , 1.00000E+02,,, 2, 0
42,T , 1.00000E+02,,, 2, 0
*ICDS
** ICDS SET = 1
l,T , 1.00000E+02
2,T , 1.00000E+02
3,T , 1.00000E+02
4,T , 1.00000E+02
.. .. .. .. .. .. .. .. .. .. .. ..
.. .. .. .. .. .. .. .. .. .. .. ..
40,T , 1.00000E+02
41,T , 1.00000E+02
42,T , 1.00000E+02
*ENDDATA
A-34
Verification Problem No.A.7 Transient heat conduction in a stainless

steel cylinder
Title:
Transient heat conduction in a stainless steel cylinder
Element Type:
Axisymmetric eight node solid elements (NKTP = 3, NORDR = 2)
Transient heat transfer analysis with prescribed initial temperature.
Problem:
A cylinder initially at 600K is quenched in an oil bath maintained at 300K. The

cylinder is assumed to have a film coefficient of 500 w/m2.0K. The length of the
cylinder is 60 mm and the diameter is 80 mm (Figure A.7.1).
Transient heat conduction analysis is performed for 3 minutes. The temperature is

obtained (a) at the center of the cylinder, (b) at the center of the circular face, (c) at
the midplane of the side and (d) at the edge of the cylinder.
Properties:
Material
KXX = 17.4 W/m.OK (Thermal conductivity)
C = 526 J/Kg.OK (Specific heat)
DENS = 7900 Kg/m3 (Density)
The axisymmetric model of the cylinder is shown in Figure A.7.2, and is composed
of twenty four (4 x 6) eight node axisymmetric solid elements, (NKTP = 103 and
A-35
NORDR = 2).
The initial temperature of the structure is specified as 600K and the ambient
temperature is maintained at 300K. Convection boundary conditions are
prescribed on the faces X = 40 mm, Y = 0, and Y = 60 mm.
Solution Procedure:
Transient analysis is performed using Crank-Nicholson time integration scheme.

Equal time step sizes of 5 seconds are used until a time of 3 minutes (180 seconds)
is reached.

The temperatures: (a) at the center of the cylinder, (b) at the center of the circular
face, (c) at the midplane of the side and (d) at the edge of the cylinder, are
compared in Table A.7.1 with the theoretical solution[l] at time equal to 3 minutes.
The parameters used in calculation of the theoretical solution are obtained by using
graphical charts. In accordance, NISA results are in very good agreement with the
theoretical solution.
Reference:
1. F. P. Incyopera, and D.P. Dewitt, "Fundamentals of Heat Transfer," John-Wiley

& Sons (1981).
Table A.7.1 Comparison of nodal temperatures at time = 3 minutes

Location NISA Ref. [l] Percent of error
Center of the cylinder, 402.7 3860K 4.3%
(node 55)
Center of the circular 370.5 3570K 3.7%
face, (node 1 or node 109)
Midplane of the side, 362.4 3510K 3.2%
(node 63)
A-36
Edge of the cylinder, 342.8 3350K 2.3%

(node 9 or node 1 17)
Figure A.7.1 Geometry and model of the cylinder
A-37
Figure A.7.2 Finite Element Mesh with Element and Node Numbers
A-38

ANALYSIS = HEAT
DIMENSION = AX,Y
ELEMENT_ECHO = OFF
FILE = heat07
SAVE=26,27,54
NODEECHO = OFF
INITIAL_CONDITION = 600.0,T
**SOLU=ITERATIVE,200,0.E1- 03,l
*TITLE
TRANSIENT HEAT CONDUCTION IN A STAINLESS STEEL CYLIN-
DER
*ELTYPE
1, 3, 2
*NODES
1 ,,,, 0.00000E+00, 0.00000E+00, 0.00000E+00, 0
2 ,,,, 5.00000E-03, 0.00000E+00, 0.00000E+00, 0
3 ,,,, 1.00000E-02, 0.00000E+00, 0.00000E+00, 0
4 ,,,, 1.50000E-02, 0.00000E+00, 0.00000E+00, 0
.. .. .. .. .. .. .. .. .. ... .. .. .. ..
.. .. .. .. .. .. .. .. .. ... .. .. .. ..
115 ,,,, 3.00000E-02, 6.00000E+02, 0.00000E+00, 0
115 ,,,, 3.50000E-02, 6.00000E+02, 0.00000E+00, 0
115 ,,,, 4.00000E-02, 6.00000E+02, 0.00000E+00, 0
*ELEMENTS
1, 1, 1, 0, 0
1, 2, 3, 12, 21, 20, 19, 10
2, 1, 1, 0, 0
3, 4, 5, 14, 23, 22, 21, 12
.. .. .. .. .. ..
.. .. .. .. .. ..
23, 1, 1, 0, 0
A-39
95, 96, 97, 106, 115, 114, 113, 104,

24, 1, 1, 0, 0
97, 98, 99, 108, 117, 116, 115, 106,
*MATFLUID
DENS, l,0, 7.90000E+03,
COND, 1,0, 1.74000E+01,
SPEC, 1,0, 5.26000E+02,
*FLCNTL, ID = 1
36,1,36,1.0,0.001,0.1E+09,1.0,1.0,300.0,600.0
*BCGVAR
** BCGVAR SET = 1
TG , 1,,,1, 1, 0, 0,
0.50000E+03, 0.30000E+03, 0.00000E+00
TG , 2,,, 1,1,1, 0, 0,
0.50000E+03, 0.30000E+03, 0.00000E+00
TG , 3,,, 1,1, 1, 0, 0,
0.50000E+03, 0.30000E+03, 0.00000E+00
TG , 4,,, 1,1,1,0, 0,
0.50000E+03, 0.30000E+03, 0.00000E+00
** BCGVAR SET = 2
TG , 4,,,,2,1, 1, 0, 0,
0.50000E+03, 0.30000E+03, 0.00000E+00
TG , 8,,,,2,1, 1,0, 0,
0.50000E+03, 0.30000E+03, 0.00000E+00
TG , 12,,,2,1, 1, 0, 0,
0.50000E+03, 0.30000E+03, 0.00000E+00
TG , 16,,,2,1, 1, 0, 0,
0.50000E+03, 0.30000E+03, 0.00000E+00
TG , 20,,,2, 1, 1, 0, 0,
0.50000E+03, 0.30000E+03, 0.00000E+00
TG , 24,,,2, 1, 1, 0, 0,
0.50000E+03, 0.30000E+03, 0.00000E+00
** BCGVAR SET = 3
TG , 21,,,3, 1, 1, 0, 0,
0.50000E+03, 0.30000E+03, 0.00000E+00
TG , 22,,,3, 1, 1, 0, 0,
0.50000E+03, 0.30000E+03, 0.00000E+00
A-40
TG , 24,,,3, 1, 1, 0, 0,
0.50000E+03, 0.30000E+03, 0.00000E+00
TG , 24,,,3, 1, 1, 0, 0,
0.50000E+03, 0.30000E+03, 0.00000E+00
*ENDDATA
A-41
Title:
Heat conduction in a convectively cooled orthotropic bar
Element Type:
2-D eight node planar elements (NKTP = 2, NORDR = 2)
Transient heat transfer analysis in an orthotropic material
Problem:
An infinitely long bar of rectangular cross section, (4 in x 2 in) is made of

orthotropic material. The orthotropic axes coincide with the global axes. The bar is
initially at 500F and it is plunged into a bath of liquid which is maintained at
100F. The convective heat transfer (film) coefficient on the bar surface is 240
BTU/hr.ft2.0F.
Temperature at different points of the cross section is calculated at time equal to 3

seconds.
Properties:
Material:
KXX = 20 BTU/hr.ft.0F (Thermal conductivity in X-direction)
KYY = 3.6036 BTU/hr.ft.0F (Thermal conductivity in Y-direction)
C = 0.00909 BTU/lbm.0F (Specific heat)
DENS = 400 lbm/ft3 (Density)
A-42
Since the bar has infinite length in Z-direction, any cross section will have the same
temperature distribution. Also the geometry and the boundary conditions of the
cross section are symmetric. Hence only one quarter of the bar cross section is
modelled using 36 (6 x 6) 8-node planar elements, see Figure A.8.1. The initial
temperature is specified as 500F. Convective boundary conditions are imposed on
the element faces (edges) which are in contact with the liquid, i.e. at X = 0.1667 ft.
(2 in) and at Y = 0.0833 ft. (1 in).
Solution Procedure:
Thirty equal time steps of 2.77778 x 10-5 hr. (0.1 sec.) are used with Crank-
Nicholson time integration scheme.
Temperature at time = 8.333 x 10-4hr. (3 sec.) for the four nodes (1, 13, 157 and
169) are tabulated in Table A.8.1. These temperatures are compared with the
analytical solution [1] and results are in very good agreement.
Reference:
1. P. J. Schneider, "Conduction Heat Transfer," Addision-Wesley Publishing

Company (1956).
Table A.8.1 Comparison of nodal temperature (0F)

Source Node 1 Node 13 Node 157 Node169
Ref: [1] 460 279 202 151
NISA 462 285 200 151
A-43
Figure A.8.1 Quarter Model of the Bar Cross Section with

Element and Node Numbers
A-44

ANALYSIS = HEAT
ELEMENT_ECHO = OFF
FILE = heat08
ANISO = ON,TEMP
SAVE = 26,27,54
NODEECHO = OFF
INITIAL_CONDITION = 500.0,T
SOLU = ITERATIVE,200,0.1 E-03,l
*TITLE
HEAT CONDUCTION IN A CONVECTIVELY COOLED ORTHOTROPIC BAR
*ELTYPE
1, 2, 2
*NODES
1 ,,,, 0.00000E+00, 0.00000E+00, 0.00000E+00,0
2 ,,,, 1.388898-02, 0.00000E+00, 0.00000E+00,0
3 ,,,, 2.777788-02, 0.00000E+00, 0.00000E+00,0
.. .. .. .. .. .. .. .. .. .. .. .. .. ..
.. .. .. .. .. .. .. .. .. .. .. .. .. ..
167,,,,1.38889E-01, 8.33328E-02, 0.00000E+00,0
168,,,,1.52778E-01, 8.33328E-02, 0.00000E+00,0
169,,,,1.36667E-01, 8.33328E-02, 0.00000E+00,0
*ELEMENTS
1, 1, 1, 0, 1
1, 2, 3, 16, 29, 28, 27, 14,
2, 1, 1, 0, 1
3, 4, 5, 18, 31, 30, 29, 16,
.. .. .. .. .. ..
.. .. .. .. .. ..
35, 1, 1, 0, 1
139, 140, 141, 154, 167, 166, 165, 152,
36, 1, 1, 0, 1
A-45
141, 142, 143, 156, 169, 168, 167, 154,

*MATFLUID
DENS, 1,0, 4.00000E+02,
KXX , 1,0, 2.00000E+0l,
KYY , l,0, 3.60360E+00,
SPEC, 1,0, 9.00900E-03,
*FLCNTL, ID- 1
50,1,50,.2778E-4,0.001,0.1 E+09,1.0,1.0,100.0,500.0
*BCGVAR
** BCGVAR SET = 1
TG , 6,,,2, 1, 1, 0, 0,
0.24000E+03, 0.10000E+03, 0.00000E+00
TG , 12,,,2, 1, 0, 0,
0.24000E+03, 0.10000E+03, 0.00000E+00
TG , 18,,,2 , 1, 1, 0, 0,
0.24000E+03, 0.10000E+03, 0.00000E+00
TG , 24,,,2, 1, 1, 0, 0,
0.24000E+03, 0.10000E+03, 0.00000E+00
TG , 30,,,2, 1, 1, 0, 0,
0.24000E+03, 0.10000E+03, 0.00000E+00
TG , 36,,,2,1, 1, 0, 0,
0.24000E+03, 0.10000E+03, 0.00000E+00
** BCGVAR SET = 2
TG , 31,,,3, 1, 1, 0, 0,
0.24000E+03, 0.10000E+03, 0.00000E+00
TG , 32,,,3, 1, 1, 0,0,
0.24000E+03, 0.10000E+03, 0.00000E+00
TG , 33,,,3, 1, 1, 0, 0,
0.24000E+03, 0.10000E+03, 0.00000E+00
TG , 34,,, 3, 1, 1, 0, 0,
0.24000E+03, 0.10000E+03, 0.00000E+00
TG , 35,,, 3, 1, 1, 0, 0,
0.24000E+03, 0.10000E+03, 0.00000E+00
TG , 36,,, 3, 1, 1, 0, 0,
0.24000E+03, 0.10000E+03, 0.00000E+00
*ENDDATA
A-46
Verification Problem No.A.9 eld problem)
Title:
Electric potential distribution inside a void (Analogous field problem)
Element Type
2-D-Four node planar elements (NKTP=2, NORDR=l)
Linear steady state potential flow analysis with prescribed potentials
Problem:
A rectangular void of dimensions (2m x lm) in a grounded conductor is biased to a
sinusoidal voltage of 10 sin y as shown in Figure A.9.1. The potential
distribution inside the void is computed.
Properties:
Material = Isotropic
Kxx = (Dielectric permittivity of air) = 8.854E-12 (F/m)
Due to the symmetry of the problem, the upper half of the void is considered to be
the analysis domain. The calculation domain is modeled with 2D-linear quadratic
elements (NKTP=102, NORDR = l). Nodes on the line x = 2m are fixed at voltages
given by 10 sin y and the nodes on lines x = 0 m, y = 0.5 and y = 1.0 m are kept
at 0 volts because of the conductor being grounded. Specifying voltages here is
analogous to specifying temperatures in a heat transfer analysis run.
A-47

The values obtained by NISA for the potentials are in excellent agreement with the
values obtained from the analytical solution given in reference 1 (See Table A.9.1
and Figure A.9.2).
References:
1. M. Javid and P. M. Brown, "Field analysis and Electromagnetics," McGraw-

Hill Book Company (1963).
Table A.9.1 Comparison of results from NISA with analytical results
Location (Analytical) (NISA)
Node 3 (0.2, 0.5) 0.0250419 0.023988

Node 18 (1.7, 0.5) 3.896535 3.866077
Node 24 (0.2, 0.6) 0.0238162 0.022814
Node 36 (1.4, 0.6) 1.443830 1.42 1357
Node 46 (0.3, 0.7) 0.032885 0.03 15422
Node 61 (1.8, 0.7) 4.3159715 4.29344
Node 66 (0.2, 0.8) 0.014719 0.01410031
Node 82 (1.8, 0.8) 3.135735 3.119373
Node 88 (0.3, 0.9) 0.012561 0.012048
Node 103 (1.8, 0.9) 1.64855 1.6399518
A-48
Figure A.9.1 Finite element model and dimensions
A-49
Figure A.9.2 Potential distribution in a void
A-50

ANALYSIS = HEAT
ELEMENT-ECHO = OFF
FILE = heat09
SAVE=26,27,54
NODEECHO = OFF
STDS = ON, 1 .0, 1 .0, 1.0
*TITLE
POTENTAL DISTRIBUTION IN A BIASED SLOT
*ELTYPE
1, 2, 1
*NODES
1 ,,,, 0.00000E+00, 5.00000E-0 1, 0.00000E+00, 0
2 ,,,, 1.00000E-0 1, 5.00000E-0 1, 0.00000E+00, 0
3 ,,,, 2.00000E-0 1, 5.00000E-0 1, 0.00000E+00, 0
.. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. ..
.. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. ..
124,,,, 1.80000E+00, 1.00000E+00, 0.00000E+00, 0
125,,,, 1.90000E+00, 1.00000E+00, 0.00000E+00, 0
126,,,, 2.00000E+00, 1.00000E+00, 0.00000E+00, 0
*ELEMENTS
1, 1, 1, 0, 0
1, 2, 23, 22,
2, 1, 1, 0, 0
2, 3, 24, 23,
.. .. .. .. ..
.. .. .. .. ..
100, 1, 1, 0, 0
104, 105, 126, 125,
*MATFLUID
DENS, 1,0, 1.00000E+00,
A-51
KXX , 1,0, 1.00000E+00,

*FLCNTL, ID = 1
1,1,1,1.0,0.001,0.1E+09,1.0,1.0,0.0,10.0
*BCDVAR
** BCDVAR SET = 1
l,T , 0.00000E+00,,, 0,0
22,T , 0.00000E+00,,, 0,0
43,T , 0.00000E+00,,, 0,0
.. .. .. .. .. .. .. .. ..
.. .. .. .. .. .. .. .. ..
124,T , 0.00000E+00,,, 0,0
125,T , 0.00000E+00,,, 0,0
126,T , 0.00000E+00,,, 0,0
** BCDVAR SET = 2
21,T , 1.00000E+01,,, 0,0
42,T , 9.5 1056E+00,,, 0,0
63,T , 8.09017E+00,,, 0,0
84,T , 5.87785E+00,,, 0,0
105,T , 3.09017E+00,,,0,0
*ENDDATA
Verification Problem No.A.10 e (Analogous Field Problem)
Title:
Magnetostatic Potential Distribution inside a Coaxial Cable (Analogous Field
Problem)
Element Type:
2-D Linear quadrilateral elements (NKTP=2, NORDR= 1)
Linear steady state potential flow analysis with prescribed flux conditions.
A-52
Problem:
A coaxial cable carrying 10A, has a hollow cylindrical region with an inner radius
of 1 m and an outer radius of 2 m, between the conductors. Magnetic scalar
potential distribution is computed in this hollow region between the inner and the
outer conductors (Figure A.11.1)
Properties:
Material: Isotropic
KXX : (permeability of air, o ) = 12.566371E-7 H/m

Due to the symmetry of the problem a quarter of the model is considered in global
cylindrical coordinates. Along the line = 0 the magnetic scalar potential is set
to an arbitrary value of 9.2737E3 and the magnetic flux density along =

--- is
2
specified according to Eqn. A.11, [see results and comparison]. The analysis 2
domain is modeled by using 2D linear quadratic elements (NKTP = 2, NORDR =
1) (see Figure A.11.1). The magnetic flux density is given as a distributed flux
along the line =
---
2

In Table A.11.1 the magnetic scalar potentials obtained by NISA are compared
with the analytical values [1,2]. The analytical expression for the magnetic scalar
potential is obtained using the flux density inside the cable which is given by [1, 2]:
o I
B = ----------
- (A.11)
2R
Where o is the permeability of air, I is the current through the cable, and R is the
radial location of any point in the calculation domain. The results obtained by
A-53
NISA, given in the Table A.11.1, are in excellent agreement with those obtained
using the analytical expressions (Figure. 10.12.2 ).
Table A.11.1 Comparison of Magnetic Scalar Potential

with the Analytical Solution [1,2]]
Location in (Analytical) in (NISA)
Node 15 ( = 22.50) 9274.325 9274.369
Node 24 ( = 33.750) 9274.637 9274.554
Node 32 ( = 450) 9274.95 9274.738
Node 40 ( = 56.250) 9275.262 9275.923
Node 45 ( = 67.50) 9275.575 9275.108
Node 50 ( = 78.750) 9275.887 9275.299
References:
1. J. D. Jackson, Classical electrodynamics, John Wiley and sons, Inc., (1967)
2. M. A. Plonus, Applied electromagnetics, McGraw-Hill Book Company
(1978
A-54
Figure A.11.1 )Finite element model and dimensions
A-55
Figure A.11.2 Magnetostatic potential distribution

inside a coaxial cable
A-56
Input Data for Verification Problem No.A.10

ANALYSIS = HEAT
ELEMENT-ECHO = OFF
FILE = heat10
SAVE=26,27,54
NODEECHO = OFF
STDS = ON, 1 .0, 1 .0, 1.0
**SOLU=ITERATIVE,200,0.E1- 03,l
*TITLE
MAGNETOSTATIC POTENTIAL DISTRIBUTION INSIDE A COAXIAL CABLE
*ELTYPE
1, 2, 1
*NODES
1, 1 ,,, 1.00000E+00, 0.00000E+00, 0.00000E+00,0
2, 1 ,,, 1.16667E+00, 0.00000E+00, 0.00000E+00,0
3, 1 ,,, 1.33333E+00,0.00000E+00, 0.00000E+00,0
.. .. .. .. .. .. .. .. .. ... .. .. .. ..
.. .. .. .. .. .. .. .. .. ... .. .. .. ..
61, 1,,, 1.66667E+00, 9.00000E+01, 0.00000E+00,0
62, 1,,, 1.83333E+00, 9.00000E+01, 0.00000E+00,0
63 1,,, 1.83333E+00, 9.00000E+01, 0.00000E+00,0
*ELEMENTS
1, 1, 1, 0, 0
1, 2, 9, 8,
2, 1, 1, 0, 0
2, 3, 10, 9,
.. .. .. .. ..
.. .. .. .. ..
47, 1, 1, 0, 0
54, 55, 62, 61,
48, 1, 1, 0, 0
A-57
55, 56, 63, 62,

*MATFLUID
DENS, 1,0, 1.00000E+00,
KXX , l,0, 1.257008-06,
*FLCNTL, ID- 1
1,1,1,1.0,0.001,0.1E+09,1.0,1.0,9274.0,10000.0
*BCDVAR
** BCDVAR SET = 1
l,T , 9.273708+03,,, 0,0
2,T , 9.273708+03,,, 0,0
3,T , 9.273708+03,,, 0,0
4,T , 9.273708+03,,, 0,0
5,T , 9.27370E+03,,, 0,0
6,T , 9.27370E+03,,, 0,0
7,T , 9.27370E+03,,, 0,0
*BCGVAR
TG, 43, 43, 1, 3,0,-1,0,0 -1.42850E-06
TG, 44, 44, 1, 3,0,-1,0,0 -8.03500E-07
TG, 45, 45, 1, 3,0,-1,0,0 -7.08000E-07
TG, 46, 46, 1, 3,0,-1,0,0 -6.33000E-07
TG, 47, 47, 1, 3,0,-1,0,0 -5.72500E-07
TG, 48, 48, 1, 3,0,-1,0,0 -7.72500E-07
*ENDDATA
A-58
Verification Problem No.A.12 ransfer
Title:
Combined surface radiation and conduction heat transfer
Element Type:
2-D four noded linear quadrilateral element (NKTP=2, NORDR=1)
Surface radiation analysis with view factor computation.
Problem:
A simple example of conduction heat transfer with surface radiation effects is
solved to illustrate surface radiation heat transfer capability of NISA/HEAT. Figure
A.11.1 shows schematic of the problem.Surfaces AB and CD are held at
temperatures Th (= 500K) and TC (= 300K), respectively. Surface BC exchanges
radiation energy with itself and with the surroundings that are at a temperature T
(= 10K). Emissivity of surface BC is 1.0. Surface AD is adiabatic and is not a
radiating surface. Temperature distribution in the domain shown in Figure A.12.1
is calculated.
Properties:
Material: Isotropic
KXX = 50 W/m2 - K (Thermal Conductivity)

Figure A.12.2 shows the finite element mesh used for the problem. There are a total
of 64 elements and 85 nodes in the model. Surface BC is divided into 16 radiation
surfaces which correspond to the element faces along surface BC.
A-59
Solution Procedure:
View factors between the radiating surfaces are computed first. Radiative heat flux
which takes into account radiation energy exchange among the surfaces and with
the surroundings is calculated using Equation (2.6.39). Equation (2.6.39) is nothing
but conservation of radiative energy. Note that a temperature field must be
available, if radiative flux is to be evaluated. On the other hand, radiative flux
affects temperature distribution. This leads to coupling between solution of
temperature and radiative fluxes. NISA/HEAT solves them in a sequential manner
in an iterative loop. Solution is continued until norm for radiative flux falls below a
user specified tolerance for a tolerance of 0.001, 22 iterations were required for
convergence.
Figure A.12.3 shows the temperature distribution for the problem described above.
For the sake of comparison, the same problem was run using radiation boundary
condition (*RADBC car) instead of surface radiation (*RADSURFACE card).
Note that the use of radiation boundary conditions in this case will amount to
neglecting surface radiation exchange among the surfaces. Only the radiation
exchange with the surroundings is taken into account. Temperature distribution for
this case is shown in Figure A.12.4.
Note that the two temperature distributions are different from each other. One can
conclude that one must use surface radiation instead of radiative boundary
condition in situations where surface to surface radiation exchange is significant.
A-60
Figure A.12.1 Schematic diagram
Figure A.12.2 Finite element mesh
A-61
Figure A.12.3 Temperature distribution (Surface Radiation)
Figure A.12.4 Temperature distribution (Radiation

Boundary Condition)
A-62

ANALYSIS = HEAT
ELEMENT_ECHO = OFF
FILE = ht11a
SAVE=26,27,54
NODEECHO = OFF
** steady state analysis
STDS = ON,1.0,1.0,1.0
** iterative solver
** surface radiation
SRAD = ON,1,32,1,0.2,0
*TITLE
surface radiation
*ELTYPE
1, 2, 1
*RCTABLE
1, 4,1, 0
1.0000000E+00, 1.0000000E+00, 1.0000000E+00, 1.0000000E+00,
*NODES
1,,,, 1.50000E+00, 0.00000E+00, 0.00000E+00, 0
2,,,, 1.87500E+00, 0.00000E+00, 0.00000E+00, 0
3,,,, 2.25000E+00, 0.00000E+00, 0.00000E+00, 0
.. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. ..
.. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. ..
98,,,,-2.25000E+00, 4.66351E-13, 0.00000E+00, 0
99,,,,-2.62500E+00, 5.44076E-13, 0.00000E+00, 0
100,,,,-3.00000E+00, 6.21802E-13, 0.00000E+00, 0
*ELEMENTS
1, 1, 1, 1, 0
1, 6, 7, 2,
2, 1, 1, 1, 0
2, 7, 8, 3,
.. .. .. .. ..
A-63
.. .. .. .. ..
63, 1, 1, 1, 0
93, 98, 99, 94,
64, 1, 1, 1, 0
94, 99, 100, 95,
*MATFLUID
DENS, 1,0, 1.00000E+00,
KXX , 1,0, 5.00000E+01,
SPEC, 1,0, 1.00000E+00,
*FLCNTL, ID= 1
1,50,1,1.0,0.001,0.1E+09,1.0,1.0,10.0,500.0
*BCDVAR
** BCDVAR SET = 1
1,T , 3.00000E+02,,, 0, 0
2,T , 3.00000E+02,,, 0, 0
3,T , 3.00000E+02,,, 0, 0
4,T , 3.00000E+02,,, 0, 0
5,T , 3.00000E+02,,, 0, 0
** BCDVAR SET = 2
96,T , 5.00000E+02,,, 0, 0
97,T , 5.00000E+02,,, 0, 0
98,T , 5.00000E+02,,, 0, 0
99,T , 5.00000E+02,,, 0, 0
100,T , 5.00000E+02,,, 0, 0
*BCGVAR
** BCGVAR SET = 1
** surface radiation
RAD , 1,,,1, 1, 1, 0, 0,
0.56690E-07, 0.10000E+02, 0.10000E+01
RAD , 5,,,1, 1, 1, 0, 0,
0.56690E-07, 0.10000E+02, 0.10000E+01
.. .. .. .. .. .. .. .. .. .. .. .. .. ..
.. .. .. .. .. .. .. .. .. .. .. .. .. ..
RAD , 57,,,1, 1, 1, 0, 0,
0.56690E-07, 0.10000E+02, 0.10000E+01
RAD , 61,,,1, 1, 1, 0, 0,
A-64
0.56690E-07, 0.10000E+02, 0.10000E+01

*ENDDATA
ANALYSIS = HEAT
ELEMENT_ECHO = OFF
FILE = ht11b
SAVE=26,27,54
NODEECHO = OFF
STDS = ON,1.0,1.0,1.0
** iterative solver
*TITLE
radiation boundary condition
*ELTYPE
1, 2, 1
*RCTABLE
1, 4,1, 0
1.0000000E+00, 1.0000000E+00, 1.0000000E+00, 1.0000000E+00,
*NODES
1,,,, 1.50000E+00, 0.00000E+00, 0.00000E+00, 0
2,,,, 1.87500E+00, 0.00000E+00, 0.00000E+00, 0
3,,,, 2.25000E+00, 0.00000E+00, 0.00000E+00, 0
.. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. ..
.. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. ..
98,,,,-2.25000E+00, 4.66351E-13, 0.00000E+00, 0
99,,,,-2.62500E+00, 5.44076E-13, 0.00000E+00, 0
100,,,,-3.00000E+00, 6.21802E-13, 0.00000E+00, 0
*ELEMENTS
1, 1, 1, 1, 0
1, 6, 7, 2,
2, 1, 1, 1, 0
2, 7, 8, 3,
.. .. .. .. ..
.. .. .. .. ..
A-65
63, 1, 1, 1, 0
93, 98, 99, 94,
64, 1, 1, 1, 0
94, 99, 100, 95,
*MATFLUID
DENS, 1,0, 1.00000E+00,
KXX , 1,0, 5.00000E+01,
SPEC, 1,0, 1.00000E+00,
*FLCNTL, ID= 1
1,50,1,1.0,0.001,0.1E+09,1.0,1.0,10.0,500.0
*BCDVAR
** BCDVAR SET = 1
1,T , 3.00000E+02,,, 0, 0
2,T , 3.00000E+02,,, 0, 0
3,T , 3.00000E+02,,, 0, 0
4,T , 3.00000E+02,,, 0, 0
5,T , 3.00000E+02,,, 0, 0
** BCDVAR SET = 2
96,T , 5.00000E+02,,, 0, 0
97,T , 5.00000E+02,,, 0, 0
98,T , 5.00000E+02,,, 0, 0
99,T , 5.00000E+02,,, 0, 0
100,T , 5.00000E+02,,, 0, 0
*BCGVAR
** BCGVAR SET = 1
TG , 1,,,1, 1, 2, 0, 0,
0.56690E-07, 0.10000E+02, 0.10000E+01
TG , 5,,,1, 1, 2, 0, 0,
0.56690E-07, 0.10000E+02, 0.10000E+01
.. .. .. .. .. .. .. .. .. .. .. .. .. ..
.. .. .. .. .. .. .. .. .. .. .. .. .. ..
TG , 57,,,1, 1, 2, 0, 0,
0.56690E-07, 0.10000E+02, 0.10000E+01
TG , 61,,,1, 1, 2, 0, 0,
0.56690E-07, 0.10000E+02, 0.10000E+01
*ENDDATA
A-66
Verification Problem No.A.13 nd specified heat flux boundary condi-

tion at the ends.
Title:
Heat conduction in a 3-D slender bar with convective and specified heat flux
boundary condition at the ends.
Element Type:
3-D four node tetrahedral elements (NKTP = 4, NORDR = 20) and 2nd order ten
node tetrahedral elements (NKTP = 4, NORDR = 21).
Tetrahedral elements
Problem:
A slender bar of square cross-section (1m x 1m) is subjected to a convective
boundary condition at one end and specified heat flux boundary condition at the
other end. The sides of the slender bar is insulated. The specified heat flux value is
100 W/m2, while the convective heat transfer coefficient is 10W/m2-k and the
ambient temperature is 30C.
Properties:
KXX = 1.0 W/m-C

3-D tetrahedral elements are used. Two different patterns of tetrahedron are used
for 1st order elements (NKTP = 4, NORDR = 20). Also 2nd order 10 noded
tetrahedron is used (NKTP = 4, NORDR = 21). Finite element meshes used in this
problem are shown in Figure A.13.2a. To apply the convective boundary condition
and specified heat flux at the ends, conventional H/*/FA/* can not be used
conveniently for tetrahedral elements as at a hyperpatch face [say the top end of the
bar] all the elements do not have the same face number (unlike hexahedron for
A-67
example). Therefore, the boundary conditions can be conveniently applied using

one of the following two options.
(i) DMAP ON ELEMENTS
Select hyperpatch no. and face no. and apply the boundary condition
[element face is immaterial]
(ii) Geometric Associativity Hyperpatch
Select Geometric Associativity under More Options menu from ADD

on elements and then choose hyperpatch face number. Here also the
element face information is immaterial.

The results obtained using three different types of tetrahedral elements are shown
in Figure A.13.3 to Figure A.13.4. Since the four sides are insulated, the
temperature varies only in z-direction. The corresponding analytical solution for
one-dimensional heat transfer problem is given by
z 1
T = q --- + --- + T
k h
The variation of temperature in z-direction for all 3 types of tetrahedral elements as

well as the analytical result is shown in Figure A.12.6, showing an excellent
agreement.
A-68
Figure A.13.1 Schematic diagram
A-69
Figure A.13.2a Finite element mesh (NORDR=1, Pattern=1)
A-70
Figure A.13.2b Finite element mesh (NORDR=1, Pattern=2)
A-71
Figure A.13.2c Finite element mesh (NORDR=1, Pattern=1)
A-72
Figure A.13.3 Temperature distribution
A-73
Figure A.13.4 Temperature distribution along the z-direction
A-74
Input data for Verification Problem No. A.13

ANALYSIS = HEAT
DIMENSION = 3D
ELEMENT_ECHO = OFF
FILE = ht12a
SAVE=26,27,54
NODEECHO = OFF
STDS = ON,1.0,1.0,1.0
** iterative solver
*TITLE
Tetrahedral Element (Pattern No. = 1, Order = 1).
Convective h.t. coefficient at one end, Dflux at other end.
*ELTYPE
1, 4, 20
*NODES
1,,,, 0.00000E+00, 0.00000E+00, 0.00000E+00, 0
2,,,, 2.50000E-01, 0.00000E+00, 0.00000E+00, 0
3,,,, 5.00000E-01, 0.00000E+00, 0.00000E+00, 0
4,,,, 7.50000E-01, 0.00000E+00, 0.00000E+00, 0
.. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. ..
.. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. ..
273,,,, 5.00000E-01, 1.00000E+00, 1.00000E+01, 0
274,,,, 7.50000E-01, 1.00000E+00, 1.00000E+01, 0
275,,,, 1.00000E+00, 1.00000E+00, 1.00000E+01, 0
*ELEMENTS
1, 1, 1, 1, 0
1, 2, 7, 27,
2, 1, 1, 1, 0
1, 7, 6, 31,
.. .. .. .. ..
.. .. .. .. ..
A-75
799, 1, 1, 1, 0
249, 269, 245, 275,
800, 1, 1, 1, 0
269, 270, 245, 275,
*MATFLUID
DENS, 1,0, 1.00000E+00,
KXX , 1,0, 1.00000E+00,
SPEC, 1,0, 1.00000E+00,
*FLCNTL, ID= 1
1,1,1,1.0,0.001,0.1E+09,1.0,1.0,0.0,1500.0
*BCGVAR
** BCGVAR SET-ID = 1
TG, 723, 723, 1, 3,0,-1,0,0 -1.00000E+02
TG, 725, 725, 1, 2,0,-1,0,0 -1.00000E+02
TG, 728, 728, 1, 3,0,-1,0,0 -1.00000E+02
.. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. ..
.. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. ..
TG, 795, 795, 1, 3,0,-1,0,0 -1.00000E+02
TG, 798, 798, 1, 3,0,-1,0,0 -1.00000E+02
TG, 800, 800, 1, 2,0,-1,0,0 -1.00000E+02
TG, 1, 1, 1, 1,1,1,0,0
1.00000E+01 3.00000E+01 0.00000E+00
TG, 2, 2, 1, 1,1,1,0,0
1.00000E+01 3.00000E+01 0.00000E+00
.. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. ..
.. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. ..
TG, 76, 76, 1, 1,1,1,0,0
1.00000E+01 3.00000E+01 0.00000E+00
TG, 77, 77, 1, 1,1,1,0,0
1.00000E+01 3.00000E+01 0.00000E+00
*ENDDATA
ANALYSIS = HEAT
DIMENSION = 3D
ELEMENT_ECHO = OFF
A-76
FILE = ht12b
SAVE=26,27,54
NODEECHO = OFF
STDS = ON,1.0,1.0,1.0
**SOLU=ITERATIVE,600,0.1E-03,1
*TITLE
Convective Boundary Condition at One End, Dflux at the Other.
*ELTYPE
1, 4, 20
*NODES
1,,,, 0.00000E+00, 0.00000E+00, 0.00000E+00, 0
2,,,, 2.50000E-01, 0.00000E+00, 0.00000E+00, 0
3,,,, 5.00000E-01, 0.00000E+00, 0.00000E+00, 0
.. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. ..
.. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. ..
273,,,, 5.00000E-01, 1.00000E+00, 1.00000E+01, 0
274,,,, 7.50000E-01, 1.00000E+00, 1.00000E+01, 0
275,,,, 1.00000E+00, 1.00000E+00, 1.00000E+01, 0
*ELEMENTS
1, 1, 1, 1, 0
6, 1, 2, 26,
2, 1, 1, 1, 0
6, 2, 7, 32,
.. .. .. .. ..
.. .. .. .. ..
799, 1, 1, 1, 0
244, 250, 274, 270,
800, 1, 1, 1, 0
250, 274, 270, 275,
*MATFLUID
DENS, 1,0, 1.00000E+00,
KXX , 1,0, 1.00000E+00,
A-77
SPEC, 1,0, 1.00000E+00,

*FLCNTL, ID= 1
1,1,1,1.0,0.001,0.1E+09,1.0,1.0,0.0,1500.0
*BCGVAR
TG, 723, 723, 1, 3,0,-1,0,0 -1.00000E+02
TG, 725, 725, 1, 3,0,-1,0,0 -1.00000E+02
TG, 728, 728, 1, 3,0,-1,0,0 -1.00000E+02
.. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. ..
.. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. ..
TG, 795, 795, 1, 2,0,-1,0,0 -1.00000E+02
TG, 798, 798, 1, 3,0,-1,0,0 -1.00000E+02
TG, 800, 800, 1, 3,0,-1,0,0 -1.00000E+02
TG, 1, 1, 1, 1,1,1,0,0
1.00000E+01 3.00000E+01 0.00000E+00
TG, 2, 2, 1, 1,1,1,0,0
1.00000E+01 3.00000E+01 0.00000E+00
.. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. ..
.. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. ..
TG, 76, 76, 1, 1,1,1,0,0
1.00000E+01 3.00000E+01 0.00000E+00
TG, 77, 77, 1, 1,1,1,0,0
1.00000E+01 3.00000E+01 0.00000E+00
*ENDDATA
ANALYSIS = HEAT
DIMENSION = 3D
ELEMENT_ECHO = OFF
FILE = ht12c
SAVE=26,27,54
NODEECHO = OFF
STDS = ON,1.0,1.0,1.0
**SOLU=ITERATIVE,600,0.1E-03,1
*TITLE
Convective Boundary Condition at One End, Dflux at the Other
*ELTYPE
A-78
1, 4, 21
*NODES
1,,,, 0.00000E+00, 0.00000E+00, 0.00000E+00, 0
2,,,, 2.50000E-01, 0.00000E+00, 0.00000E+00, 0
3,,,, 5.00000E-01, 0.00000E+00, 0.00000E+00, 0
.. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. ..
.. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. ..
1539,,,, 3.75000E-01, 1.00000E+00, 9.50000E+00, 0
1540,,,, 6.25000E-01, 1.00000E+00, 9.50000E+00, 0
1541,,,, 8.75000E-01, 1.00000E+00, 9.50000E+00, 0
*ELEMENTS
1, 1, 1, 1, 0
1, 276, 2, 500, 7, 966, 1342, 717, 1143, 27,
2, 1, 1, 1, 0
134
1, 966, 7, 280, 6, 496, 1142, 721, 31,
6,
.. .. .. .. .. .. .. .. .. ..
.. .. .. .. .. .. .. .. .. ..
799, 1, 1, 1, 0
112 114
249, 1340, 269, 1537, 245, 1541, 1341, 275,
5, 1,
800, 1, 1, 1, 0
153
269, 491, 270, 960, 245, 1141, 715, 1341, 275,
7,
*MATFLUID
DENS, 1,0, 1.00000E+00,
KXX , 1,0, 1.00000E+00,
SPEC, 1,0, 1.00000E+00,
*FLCNTL, ID= 1
1,1,1,1.0,0.001,0.1E+09,1.0,1.0,0.0,1500.0
*BCGVAR
TG, 723, 723, 1, 3,0,-1,0,0 -1.00000E+02
A-79
TG, 725, 725, 1, 2,0,-1,0,0 -1.00000E+02

TG, 728, 728, 1, 3,0,-1,0,0 -1.00000E+02
.. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. ..
.. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. ..
TG, 795, 795, 1, 3,0,-1,0,0 -1.00000E+02
TG, 798, 798, 1, 3,0,-1,0,0 -1.00000E+02
TG, 800, 800, 1, 2,0,-1,0,0 -1.00000E+02
TG, 1, 1, 1, 1,1,1,0,0
1.00000E+01 3.00000E+01 0.00000E+00
TG, 2, 2, 1, 1,1,1,0,0
1.00000E+01 3.00000E+01 0.00000E+00
.. .. .. .. .. .. .. .. .. .. .. .. .. .. ..
.. .. .. .. .. .. .. .. .. .. .. .. .. .. ..
TG, 76, 76, 1, 1,1,1,0,0
1.00000E+01 3.00000E+01 0.00000E+00
TG, 77, 77, 1, 1,1,1,0,0
1.00000E+01 3.00000E+01 0.00000E+00
*ENDDATA
A-80
Verification Problem No.A.14 Solidification of molten metal in a cylin-

der
Title:
Solidification of molten metal in a cylinder
Element Type:
Axisymmetric quadrilateral element (NKTP = 3, NORDR = 1)
Phase change capability using temperature dependent material property.
Problem:
The Solidification of molten alloy in a cylinder of diameter D and height H1
followed by an another material of diameter D and height H2 is considered (Figure
A.14.1). The bottom molten metal is kept at an initial temperature of TI whose
material property is temperature dependent. The top metal is kept at an initial
temperature of T2. The top surface of the top metal is exposed to atmosphere.
Properties:
D = 2.0 H1 = 5.0
H2 = 5.0 1 = 2.7
2 = 6.9 K1(Cond.) = 0.449
K2(Cond. = 0.154 Cp1 = 0.215

)
Cp2 = 0.128 T1 = 680oC
Tc = 20oC T = 20oC
h = 0.05 Latent Heat = 93oC
A-81
Solidus = 570oC Liquidus Temp. = 649oC

Temp.
The axisymmetric model of the cylinder having two materials one above the other
is shown in Figure A.14.2. The Finite Element Model consist of twenty (20 x 1)
elements in each material. All the nodes in material 1 is kept at initial temperature
of T1 (6800 C) and all the nodes in material 2 is kept at initial temperature of T2
(20" C). Convective boundary conditions is applied on top surface of element no
40 which is exposed to atmosphere as shown in Figure A.14.1. Temperature
dependent Specific Heat are applied for material 1 by using KTEMP card and
Temperature Function card. Latent Heat, Solidus Temperature and Liquidus
Temperature values are applied using Phase Change card.
Solution Procedure:
Transient analysis is performed using Crank-Nicholson time integration scheme.
Equal time step sizes of 3 seconds are used until a time of 2 minutes 30 seconds (1
50 seconds) is reached.
Results:
Temperature contours shows that after 150 seconds the hot molten metal reaches to
maximum temperature of 41 1.70 C. Graphs are also plotted at selected node nos. to
show how temperature at different location is varying with time.
A-82
Figure A.14.1 Geometry and applied boundary conditions
A-83
Figure A.14.2 Finite element mesh
A-84
A-85
Figure A.14.3 Temperature distribution
A-86
Figure A.14.4 Temperature history plot
A-87
A-88
Appendix
B
NISA/HEAT to NISA/HEAT3 Translator
Appendix B
NISA/HEAT nisa files can be converted to NISA/HEAT3 format using heat2to3

(h2h3 on workstations) conversion program. Given below is the list of cards
that can be converted to HEAT3 format and their equivalent cards in HEAT3.
Table B.1 Conversion table for HEAT2 to HEAT3
NISA/HEAT NISA/HEAT3
BLANK This card is not required in HEAT3
EXECUTION EXECUTION
FILE FILE
NODE NODE
PATH PATH
PAT1 - PAT4 PAT1 - PAT4
RESEQUENCE RESEQUENCE
RESTART RESTART
SAVE SAVE
WARNING WARNING
INITIAL INITIAL
MASS MASS
ORTH ANISO
B-1
SIGMA Used in *BCGVAR

STEP Used in *FLCNTL
SRAD SRAD
UFIJ UFIJ
*TITLE *TITLE
*ELTYPE *ELTYPE
NISA/HEAT NISA/HEAT3
*RCTABLE *RCTABLE
*LCSYS *LCSYS
*NODES *NODES
*MATHEAT *MATFLUID
*TEMPFN *TEMPFN
*PCHANG1 *PCHANG, PCHANG
*CPTEMP *CPVAR
*TIMEAMP *TIMEAMP
*HEATCNTL *FLCNTL
*INITEMP *ICDS
*DFLUX *BCGVAR
*CONVBC *BCGVAR
*RADSURF *BCGVAR
*TIMEINT *FLCNTL
*SPTEMP *BCDVAR
*RADBC *BCGVAR
*TEMPHISTORY Use FLUTL package
*STEPSIZE *STEPSIZE
*CFLUX *BCDVAR (NHS)
B-2
*NDHEAT *BCDVAR
*PRINT *PRINT
*TEMPOUT *FLCNTL
*ELHEAT *BCGVAR
Notes:
1. File 39 of NISA/HEAT is file 55 in HEAT3.

2. Time dependent emmissivity is currently not supported in HEAT3.
3. The user must confirm the converted values for *NDHEAT card.
B-3
B-4
Appendix
C
FLUTL: NISA/HEATs Utility Package
Appendix C
C.1 About this Appendix

In this section various features of the NISA/HEATs Utility Package, called
FLUTL are described. Since the same utility package is shared with NISA/3D-
FLUID, some reference to the fluid flow related variable is present.
C.2 Introduction
Ordinarily when a heat transfer problem is solved, temperature field is obtained. It
is usually calculated at all the computation points. This means that there is often
more data than the analyst needs as he may be interested in only a specific portion
of the computational domain. Also, instead of seeking values of temperature at
discrete points in the domain, the analyst may want them in an overall (integrated)
sense or one may be integrated in heat flux computation or heat transfer
coefficient calculation within (or across) a portion of the domain. The main
purpose of FLUTL is to provide such facilities to the analyst.
C.3 Basic Features of FLUTL

FLUTL is an interactive program designed to accompany the NISA/HEAT
package. It runs in alpha-numeric mode. Input data for FLUTL comes from the
binary result files (26 and 27) produced by NISA/HEAT, and from the answers
supplied by the user during the session. As FLUTL runs, all the answers provided
by the user are stored in a session file. The session file name is fixed by FLUTL
and has the generic form flutlnn.ses, where nn is a number between01 and 99.
When FLUTL is invoked, it finds the last session file name in the current
C-1
Basic Features of FLUTL
directory and creates a file with the next higher (in alphanumeric order) file name.
The highest filename FLUTL allowed to create is flutl99.ses. The entire session,
complete with questions and remarks made by FLUTL (i.e. practically everything
that the user sees on the screen), can also be logged simultaneously in a log file
(with a user-supplied name). Please note that the session files and the log files are
in ASCII format, and may therefore occupy substantial disk space; it is advised that
these files be purged from time to time.
In broad categories, the main capabilities of FLUTL can be divided under the
following headings:
(i) To produce time history of temperature.
(ii) To make a temperature profile along a line in space at a given time step.
(iii) Integrate temperature over a defined surface.
Options (i) and (ii) will produce X-Y type of plot data that could be appended to
the NISA/HEAT binary result files (26 and 27), or dumped into a user supplied
ASCII file, or both. The plots may then be viewed using the NISA post-processor
DISPLAY III POST (ref. command GNI and GPH for plotting of binary data and
ASCII data, respectively).
Please note that if a job is RESTARTed, like most other old results, the appended
results will also be lost(ref. EXECUTIVE command RESTART).
Option (iii) will usually produce ASCII formatted output, which will be dumped in
the hard copy output file (the name of this file could be supplied by the user).
Default file names for the log file, the ASCII X-Y type plot file and the hard copy
output file are flutl.log, flutl.xyz, and flutl.out, respectively. Files with default
names are always opened as new files; this means that if a previous copy of any
of these files exists in the current directory, it is lost. Checks for duplication in user
supplied file names are performed.
In addition to the above functions, this package is equipped with handy capabilities
for user convenience. For example, the user may check all the nodal boundary
conditions using FLUTL.
C-2
FLUTL is largely menu driven and most questions require no explanation. Often
the answers are either YES or NO with the default demarcated within double
quotes (for example (Y/N) means that the default is NO). Answers are to be
given on-screen in most cases; in a few instances, FLUTL can be redirected to
read from an external file containing, for example, a list of nodes or elements. The
external file should contain data laid out in the same fashion as they would appear
if entered from the screen, and should preferably end with a data terminating line
containing only *END. It should be noted here that whenever more than one
number is expected, the numbers should be separated by either a comma or a blank,
and each number should not occupy more than a 15 character field.
FLUTL execution is sequential and the user may choose to climb up or down the
menu ladders by the appropriate menu option number. If either !Q or !E is typed
as a response to a FLUTL prompt, the program execution terminates.
Data (e.g., values of temperature) that has been retrieved could be modified as
follows:
DATA = DATA * FACTOR + OFFSET
Factors and offsets will be asked for by FLUTL at appropriate times. This feature is
helpful in scaling or in non-dimensionalizing the data.
FLUTL can also be run in session file mode. In this mode, the user puts the
responses (to the questions which would be asked by FLUTL) in an ASCII file and
supplies its name to FLUTL instead of typing the responses on the screen. The user
should look at any flutlnn.ses files to get a better idea of the format.
Care must be taken to ensure that the answers given in this session file are correct;
if FLUTL encounters a wrong answer, it may continue to read the session file, but
could interpret the rest of the answers wrongly. A typical example of such an error
is as follows: in the first pass of FLUTL, one or more of the files could be newly
opened. In the next pass, FLUTL views these files as existing files and may ask if
the user wishes to overwrite or append to them. These additional answers must be
inserted into the sequence of answers given during the first pass of FLUTL for the
rest of the answers to be interpreted correctly. Please note that an empty carriage
return is indicated by CR in the session file. Lines starting with ** are comment
C-3
Executing FLUTL
lines. While reading a session or an external data file, comment lines and blank
lines are ignored.
C.4 Executing FLUTL

FLUTL can be executed by typing FLUTL preceded by the necessary path name, if
applicable. The following questions appear immediately after the welcome
message:
1. Would you like to read in a session file (Y/N) >

If Yes
1.1) Enter name of session file or <CR> to quit >
2. Would you like to log your session (Y/N) >

If Yes
2.1) Default Log File Name is: flutl.log. Enter a log file name or <CR>
for default >
3. Do you wish to append results to File 26 and File 27 (Y/N)>

4. Enter name of file 26 or <CR> to quit >
5. ...Default File 27 is: Default Name
Enter file 27 name or <CR> to quit >
6. ...Default Hard Copy Output File is: flutl.out
Enter name of hard copy output file or <CR> to use default >
7a)...Default ASCII Plot File is: flutl.xyz
Enter a file name for writing ASCII X-Y plot data or <CR> for default>
7b) Enter file name for writing ASCII X-Y plot data or <CR> for ignore >
Maximum number of trial answers allowed for questions 4 and 5 are 3. The default
file27 name, whenever possible, is constructed from the file26 name. Depending
on whether the answer to question 3 is no or yes, either question 7a or 7b is
asked. If any of the hard copy output file or the ASCII X-Y plot file already exists,
C-4
then FLUTL asks the user what has to be done with the existing files (i.e.,
overwrite or append).
Once question 7 is answered, FLUTL brings up the Main Menu.
C.5 Menus and Pertinent Notes

As mentioned earlier, FLUTL is menu driven. Starting at any level, the user climbs
up or down by choosing the appropriate option number. At times, when an option is
selected, some questions must be answered by the user.
FLUTL begins with Main Menu which appears as:
Menu.1
=======================================================
===========
MAIN MENU
=======================================================
===========
(0) = Quit execution of FLUTL
(1) = Change File26 and File27
(2) = Change Hard Copy Output File
(3) = Change X-Y Plot File
(4) = Process NISA/3D-FLUID Results
(5) = Check Nodal B.C. or Wall Function Nodes
Enter a Number or <CR> for Default (def = 4) >
Option 0 will terminate execution of FLUTL. Options 1 through 3 allow the user to
change some of the filenames during execution of FLUTL.
Option 5 lets the user verify the nodal boundary conditions and the wall function
nodes visually. In this case, the user should make only a CHECK run (ref.
EXECUTIVE command EXEC = CHECK) and supply the resulting file26
name to FLUTL. If the user selects to check nodal boundary conditions for a
C-5
Menus and Pertinent Notes
particular variable (say, Temperature T), FLUTL will ask the user to supply an
initialization value. FLUTL will first initialize the selected variable (in this case T)
with this initialization value, and then superimpose the applied nodal boundary
conditions of T on the initialized values. These nodal values will be written out as
results of a fictitious HEAT run (Load Case = 1, Time Step = 1) in the
corresponding file27. The user may then plot contours of T using DISPLAY III
POST. As all the free nodes would have the user specified initialization value, there
would be spots of this value among the legitimate boundary condition values. In
order to distinguish these spots, the initialization value should be any number (the
order of the number being comparable to the applied boundary condition) that has
not been used in specifying any of the boundary conditions for the chosen variable.
In Figure C.1, an example of one such missing temperature boundary condition is

shown. The original intent was to apply a temperature boundary condition of 10.0
on all the nodes on the exterior. An initialization value of 1.0 was used. Note the
spot on the side ABCD which indicates that a node was missed.
An important note that concerns the viewing of boundary conditions in this way is
that to clearly view the contours the user may have to judiciously zoom into
various corners of the model, and set the range (ref. DISPLAY POST command
RNG) to current view etc., to clearly view the contours, since in the case of small
elements and bad graphic resolution, the diagnostic spots may be missed.
If Option 4 of Menu.1 is selected, the following menu appears:
Menu.2
========================================================
==========
MENU FOR POST PROCESSING OF 3D-FLUID RESULTS
========================================================
==========
(0) = Go Back to Main Menu
(1) = Produce Time History Plots
(2) = Create Profiles of Fluid Qtys
(3) = Miscellaneous Capabilities
C-6
Enter a number >
Option 1 produces time history data for any of the fluid variables. By choosing
option 2, spatial variation of a fluid variable at a given time step can be obtained.
The menus for obtaining time histories and profiles are give in the following menus
(e.g., Menu.3 and Menu.4).
C-7
Menu.3
========================================================
==========
TIME HISTORY PLOTS CAN BE OBTAINED FOR ANY OF THE FOL-
LOWING:
========================================================
==========
(0) = Previous Menu
(1) = U Velocity [U]
(2) = V Velocity [V]
(3) = W Velocity [W]
(4) = Kinetic Energy [K]
(5) = Dissipation Rate [D
(6) = Eddy Viscosity [EDD]
(7) = Stream Lines [SL]
(8) = Temperature [T]
(9) = Pressure [P]
(10) = X- Normal Stresses [Sxx]
(11) = Y- Normal Stresses [Syy]
(12) = Z- Normal Stresses [Szz]
(13) = XY- Shear Stresses [Sxy]
(14) = XZ- Shear Stresses [Sxz]
(15) = YZ- Shear Stresses [Syz]
Enter a number >
C-8
Menu.4
========================================================
==========
PROFILES CAN BE MADE FOR ANY OF THE FOLLOWING:
CURRENT TIME STEP = 3
========================================================
==========
(0) = Previous Menu
(9) = Pressure [P]
Enter a number >
Miscellaneous capabilities (Option 3 of Menu.2) of FLUTL are currently limited to

integration of a temperature over a user defined surface. If this option is selected,
the following menu appears.
Menu.5
========================================================
==========
C-9
MENU FOR MISCELLANEOUS CAPABILITIES

========================================================
==========
(0) = Go Back to Previous Menu
(1) = Integrate a Fluid Quantity
Enter a number >
The fluid quantities that could be integrated are given in the following menu:
C-10
Menu.6
========================================================
==========
ANY OF THE FOLLOWING QUANTITY CAN BE INTEGRATED OVER A
SURFACE CURRENT TIME STEP = 3
========================================================
==========
(0) = Previous Menu
(1) = U Velocity [U] (2) = V Velocity [V]
(3) = W Velocity [W] (4) = Volume Flow Rate [Q]
(5) = Temperature [T] (6) = Pressure [P]
(7) = Stream Function [S] (8) = Kinetic Energy [K]
(9) = Dissipation [D] (10 = Eddy Viscosity [EDD]
(1 = X-Normal Stress [SXX] ) = Y-Normal Stress [SYY]
1) = Z-Normal Stress [SZZ] (12 = XY-Shear Stress [SXY]
(1 = XZ-Shear Stress [SXZ] ) = YZ-Shear Stress [SYZ]
3) (14
(1 )
5) (16
)
(1 = Surface Stresses (Normal and Shear)
7) = Film Heat Transfer Coefficient [h]
(1
8)
Enter a number >
It should be pointed out here that the TOTAL AREA reported in the hard copy
output is a scalar sum of all the elemental areas while the various components of
area are vector sums. These components could be viewed as projections of the
chosen surface onto planes represented by X, Y or Z = constant, respectively. The
surface normals are always computed to be positive when pointing outwards. Also
the words face and area in the output mean edge and length in a 2D case.
C-11
In the following, some menus are presented that indicate choices currently
available in FLUTL to define node numbers for time history or profiles, to define
planes and lines, etc.
Menu.7
========================================================
==========
...SELECTING NODE NUMBER FOR TIME HISTORY PLOTS:
========================================================
==========
(0) = Back to Previous Menu
(1) = By Giving Node Number
(2) = By Giving Approx. X, Y and Z coords.
Enter a Number or <CR> for Default (def. = 1) >
Menu.8
========================================================
==========
...COLLECTING USER DEFINED NODES:
========================================================
==========
How Do You Want to Pick the Nodes ?
(0) = Quit
(1) = By Specifying the Nodes ON-SCREEN
(2) = By Specifying a Plane
(3) = By Specifying a Line
(4) = By Reading in a File with Node Numbers
Enter a number >
C-12
Menu.9
========================================================
==========
OPTIONS FOR DEFINING A PLANE:
========================================================
==========
How Would You Like to Define the Plane?
(0) = Quit Defining the Plane
(1) = By Specifying a Constant X, Y or Z Value
(2) = By Specifying 3 Nodes
(3) = By Specifying 3 Points
(4) = By Specifying 1 Node and 1 Normal
(5) = By Specifying 1 Point and 1 Normal
Enter a number >
C-13
Menu.10
========================================================
==========
OPTIONS FOR DEFINING A LINE:
========================================================
==========
How Would You Like to Define the Line?
(0) = Quit Defining Line
(1) = By Intersection of Two Planes
(2) = By Specifying Two Nodes
(3) = By Specifying Two Points
(4) = By One Node and One Direction Vector
(5) = By One Point and One Direction Vector
Enter a number >
When nodes are picked up using a plane or a line (in case of creating profiles), all
the nodes that are within the user specified tolerance are picked. Sorting options are
available in FLUTL; the user is advised to use them since the sequence of node
numbers may not be the desired ordering of the nodes. The user also has an option
of restricting FLUTL to a subdomain of the original computational domain by
providing minimum and maximum extents of the subdomain.
C-14
C.6 Some Examples of FLUTL Usage

In this section examples of some actual sessions with FLUTL are presented. The
output files generated by FLUTL in these sessions are also given.
C.6.1 Example #1
In the first example, the result files for flow over a cylinder (using 2D elements) are
taken. This is the same as Problem No. 8 in the NISA/3D-FLUID Demo Problems.
The following tasks are to be performed:
(i) make a time history plot of U-velocity at point (x,y) = (5,1)
(ii) make a U-velocity profile at X = 15.0 for Y between 0.0 and 4.0
The names of the binary results files are CYL26.DAT and CYL27.DAT. The time
history plot data are put in the file CYL1.XYZ and the profile data are put in the
file CYL2.XYZ. The log of this particular session is given in the file CYL.LOG.
Printouts of these files are given in Section C.6.3.
C.6.2 Example #2
In this example result files for a Couette flow problem is considered. The problem
is solved using 3D elements. This is the first problem in the NISA/3D-FLUID
Demo Problems. The following tasks are to be performed:
(i) make a U-velocity profiles along the line defined by the intersection of the
planes X = 0 and Z = 0.
(ii) find total flow rate through the entry plane.
(iii) find total flow rate through the exit plane.
The names of the binary result files are COUTETTE26.DAT and

COUETTE27.DAT. The profile data are put in the file COUETTE.XYZ. In order to
find the volume rate flow through the entry and exit planes, the element numbers
along with the appropriate face numbers that make up these plane had to be known
(this knowledge can be obtained from the finite element model using the NISA pre-
C-15
Some Examples of FLUTL Usage
processor DISPLAY PRE). For the case of the entry plane, this data is entered
interactively, while for the case of the exit plane, the data are compiled in a file
named ELEM.DATA and FLUTL is made to read this file. These two approaches
are used just to exemplify the two different modes of data entry to FLUTL. The
output detailing the evaluation of the flow rates is in the file COUETTE.OUT. The
log of this particular session is saved in the file COUETTE.LOG. Printouts of these
files are given in Section C.6.3.
C.6.3 Printouts for Example #1 and Example #2
Begin file: CYL.LOG

Do you wish to append results to File26 and File27 (Y/N) >!CR
Enter Name of File26 or <CR> to quit >
CYL26.DAT
..IN SUBROUTINE (INFO01:1)
..MEMORY REQUESTED (IN INT*4 WORDS) = 2161
..POST PROCESSING NISA/3D-FLUID RESULTS .................
..JOB TITLE: FLOW OVER A CYLINDER SHOWING SEPERATION
(RE=20)
..DATE RUN MADE: 4-19-1989
..RUN MODE: GO
..PROBLEM DIMENSION: 2D
..TOTAL NUMEBR OF NODES = 270
..LARGEST NODE I.D. = 270
..TOTAL NUMEBR OF ELEMENTS = 230
..LARGEST ELEMENT I.D. = 230
..MINIMUM X-COORDINATE = .00000E+00
..MAXIMUM X-COORDINATE = .15000E+02
..MINIMUM Y-COORDINATE = .00000E+00
..MAXIMUM Y-COORDINATE = .50000E+01
..MINIMUM Z-COORDINATE = .00000E+00
..MAXIMUM Z-COORDINATE = .00000E+00
C-16
..TOTAL NUMBER OF LOAD CASES = 1

.. Default File27 is: CYL27.DAT
Enter File27 name or <CR> for default >!CR
.. Default Hard Copy Output File is: futl.out
Enter name of hard copy output file or <CR> to use default >!CR
..Default ASCII X-Y Plot File is: futl.xyz
Enter file name for writing ASCII X-Y plot data or <CR> for default
CYL1.XYZ
..PLEASE WAIT..
..SEARCHING FOR END-OF-FILE FOR FILE26..
..PLEASE WAIT..
C-17
========================================================
==========
MAIN MENU
========================================================
==========
(0) = Quit Execution of FLUTL
Enter a Number or <CR> for Default (def = 4) >4
========================================================
==========
========================================================
==========
Enter a Number >1
..Total Number of Load Cases = 1

..Default Load Case is = 1
Enter Load Case Number or <CR> for Default >!CR
C-18

LOWINGS:
(0) = Previous Menu
(9) = Pressure [P]
Enter a number >1
..IN SUBROUTINE (THIST0:1)

..SELECTING NODE NUMBER FOR TIME HISTORY..

(0) = Back to Previous Menu
(1) = By Giving Node Number
(2) = By Giving Approx. X, Y and Z coords.
Enter a Number or <CR> for Default (def.= 1) >2
C-19

Enter Approx. X, Y and Z Coordinates >
5,1,0
..NODE PICKED = 33
..X-Y-Z COORDINATES: .50000E+01 .11000E+01 .00000E+00
..DISTANCE FROM THE GIVEN POINT = .10000E+00
..TOTAL NUMBER OF TIME STEPS = 14
..Default Starting Step No. = 1
..Default End Step No. = 14
..Default Step Increment = 1
Enter Start, End Time Step# and Increment or <CR> for Default >!CR
..X-AXIS: TIME STEP#; Y-AXIS: FLUID QTY.
..PUTTING 8-CHARACTER PLOT CODE..
Enter Plot Code (max. 8-char) >U @N33
Do you wish to label your plot (Y/N) >!CR
..SETTING FACTORS AND OFFSETS FOR DATA MODIFICATIONS..
..*NOTE: DATA = DATA*FACTOR + OFFSET
..CURRENT VALUES OF THE FACTORS AND OFFSETS ARE:
..X-FACTOR = .10000E+01
..Y-FACTOR = .10000E+01
..X-OFFSET = .00000E+00
..Y-OFFSET = .00000E+00
Do you wish to change the above values (Y/N) >!CR
..RESULTS ARE READY TO BE WRITTEN OUT..
C-20
Would You Like to Write Out the Results (Y/N) ? >!CR

..> RESULT PROCESSING COMPLETED <
- Hit <CR> to Continue
!CR

LOWINGS:
(0) = Previous Menu
(9) = Pressure [P]
(10 = X- Normal Stresses [Sxx]
) = Y- Normal Stresses [Syy]
(11 = Z- Normal Stresses [Szz]
) = XY- Shear Stresses [Sxy]
(12 = XZ- Shear Stresses [Sxz]
) = YZ- Shear Stresses [Syz]
(13
)
(14
)
(15
)
Enter a Number >0
C-21
========================================================
==========
========================================================
==========
Enter a Number >0
C-22
========================================================
==========
MAIN MENU
========================================================
==========
..Default ASCII X-Y Plot File is: CYL1.XYZ

Enter file name for writing ASCII X-Y plot data or <CR> for default >
CYL2.XYZ
========================================================
==========
MAIN MENU
========================================================
==========
Enter a Number >4
C-23
========================================================
==========
Enter a Number >2

Enter Load Case Number or <CR> for Default >!CR
..DEFAULT TIME STEP = 14
Enter Time Step # or <CR> to Use Default >!CR

C-24
(0) = Previous Menu

(9) = Pressure [P]
(13
)
(14
)
(15
)
Enter a Number >1
..IN SUBROUTINE (PRFIL0:1)

..RETRIEVING FLUID RESULTS..
..COLLECTING USER DEFINED NODES..
How Do You Want to Pick the Nodes?
(0) = Quit
C-25
Enter a Number >2

Do You Want to Give Min. and Max. XYZ Cutoffs (Y/N)? >Y
Enter Xmin, Xmax, Ymin, Ymax, Zmin, Zmax >
14,15,0,4
(0) = Quit Defining a Plane
Enter a Number >1
..SPECIFYING A PLANE AT A CONST. X OR Y OR Z..
..FOR EXAMPLE, TO SPECIFY A PLANE AT X=2.2 ENTER X2.2"..
Enter X, Y or Z and the Constant >X15
..CURRENT TOLERANCE = .10000E-02
Enter New Tolerance or <CR> for Curr. Value >!CR
..TOTAL NUMBER OF NODES SELECTED = 9
Do You Want to Sort the Node Numbers (Y/N)? > !CR
Available Sorting Choices Are:
(0) = Quit Sorting
(1) = Sorted by X-coordinates
(2) = Sorted by Y-coordinates
(3) = Sorted by Z-coordinates
(4) = Sorted by Node Numbers
Enter a Number >2
C-26
Enter Sorting Order (1=Ascending, -1=Descending)
or <CR> for Default (def = 1) >!CR

What Would You Like to Plot on X-axis?
(1) = X-coordinates
(2) = Y-coordinates
(3) = Z-coordinates
(4) = Node Numbers
(5) = Calculated Fluid Qty
Enter a Number >5
What Would You Like to Plot on Y-axis?
(1) = X-coordinates
(2) = Y-coordinates
(3) = Z-coordinates
(4) = Node Numbers
Enter a Number >2
..X-AXIS: FLUID QTY ; Y-AXIS: Y-COORD ..
Enter Plot Code (max. 8-char) >U @X15
Do you wish to label your plot (Y/"N") >!CR
..X-FACTOR = .10000E+01
..Y-FACTOR = .10000E+01
..X-OFFSET = .00000E+00
..Y-OFFSET = .00000E+00
C-27

Would You Like to Write Out the Results (Y/N)? >!CR
..> RESULT PROCESSING COMPLETED <
------ Hit <CR> to Continue
- !CR

(0) = Previous Menu
(9) = Pressure [P]
(13
)
(14
)
(15
)
C-28
Enter a Number >0
========================================================
==========
========================================================
==========
Enter a Number >0
========================================================
==========
MAIN MENU
========================================================
==========
NORMAL COMPLETION OF FLUTL
End of file: CYL.LOG
C-29
Begin file: CYL1.XYZ

.10000E+01 .12980E+01;U @N33
.20000E+01 .12952E+01
.30000E+01 .12953E+01
.40000E+01 .12952E+01
.50000E+01 .12943E+01
.60000E+01 .12927E+01
.70000E+01 .12905E+01
.80000E+01 .12881E+01
.90000E+01 .12860E+01
.10000E+02 .12841E+01
.11000E+02 .12827E+01
.12000E+02 .12815E+01
.13000E+02 .12807E+01
.14000E+02 .12800E+01
-99999 -99999
End file: CYL1.XYZ
Begin file: CYL2.XYZ

.77252E+00 .00000E+00;U @X15
.78195E+00 .26516E+00
.81310E+00 .53033E+00
.85353E+00 .79549E+00
.90526E+00 .10607E+01
.99930E+00 .15859E+01
.10762E+01 .22425E+01
.11041E+01 .30303E+01
.10714E+01 .39495E+01
-99999 -99999
End file: CYL2.XYZ
Begin file: COUETTE.LOG
C-30
Do you wish to append results to File26 and File27 (Y/N) >N
Enter Name of File26 or <CR>to quit >

COUETTE26.DAT
..IN SUBROUTINE (INFO01:1)
..MEMORY REQUESTED (IN INT*4 WORDS)= 513
..POST PROCESSING NISA/3D-FLUID RESULTS .................
..JOB TITLE: 3D COUETTE FLOW
..DATE RUN MADE: 4-19-1989
..RUN MODE: GO
..PROBLEM DIMENSION: 3D
..TOTAL NUMEBR OF NODES = 64
..LARGEST NODE I.D. = 64
..TOTAL NUMEBR OF ELEMENTS = 27
..LARGEST ELEMENT I.D. = 27
..TOTAL NUMBER OF LOAD CASES = 1
.. Default File27 is: COUETTE27.DAT
Enter File27 name or <CR>for default >!CR
.. Default Hard Copy Output File is: futl.out
Enter name of hard copy output file or <CR>to use default >
COUETTE.OUT
..Default ASCII X-Y Plot File is: futl.xyz
Enter file name for writing ASCII X-Y plot data or <CR>for default >
COUETTE.XYZ
C-31
..PLEASE WAIT..
..PLEASE WAIT..
========================================================
==========
MAIN MENU
========================================================
==========
========================================================
==========
========================================================
==========
C-32
Enter a Number >2

Enter Load Case Number or <CR>for Default >!CR
Enter Time Step # or <CR>to Use Default >!CR
C-33

(0) = Previous Menu
(9) = Pressure [P]
(13
)
(14
)
(15
)
Enter a Number >1
..IN SUBROUTINE (PRFIL0:1)

..RETRIEVING FLUID RESULTS..
..COLLECTING USER DEFINED NODES..
How Do You Want to Pick the Nodes?
C-34
(0) = Quit
Enter a Number >3
Do You Want to Give Min. and Max. XYZ Cutoffs (Y/N)? >!CR
How Do You Want to Define the Line?
(0) = Quit Defining Line
(1) = By Intersection of Two Planes
(2) = By Specifying Two Nodes
(3) = By Specifying Two Points
(4) = By One Node and One Direction Vector
(5) = By One Point and One Direction Vector
Enter a Number >1
..CURRENT TOLERANCE = .10000E-02
Enter New Tolerance or <CR>for Curr. Value >!CR
..DEFINING A LINE BY TWO PLANES..
..DEFINING THE FIRST PLANE..
Enter a Number >1
..FOR EXAMPLE, TO SPECIFY A PLANE AT X=2.2 ENTER X2.2..
C-35
Enter X, Y or Z and the Constant >X0

..DEFINING THE SECOND PLANE..
Enter a Number >1
..FOR EXAMPLE, TO SPECIFY A PLANE AT X=2.2 ENTER X2.2..
Enter X, Y or Z and the Constant >Z0
..TOTAL NUMBER OF NODES SELECTED = 4
Do You Want to Sort the Node Numbers (Y/N)? >!CR
Available Sorting Choices Are:
(0) = Quit Sorting
(1) = Sorted by X-coordinates
(2) = Sorted by Y-coordinates
(3) = Sorted by Z-coordinates
(4) = Sorted by Node Numbers
Enter a Number >2
Enter Sorting Order (1=Ascending, -1=Descending) or <CR> for Default (def =
1) >1
What Would You Like to Plot on X-axis?
C-36
(1) = X-coordinates
(2) = Y-coordinates
(3) = Z-coordinates
(4) = Node Numbers
Enter a Number >5
What Would You Like to Plot on Y-axis?
(1) = X-coordinates
(2) = Y-coordinates
(3) = Z-coordinates
(4) = Node Numbers
Enter a Number >2
..X-AXIS: FLUID QTY ; Y-AXIS: Y-COORD ..
Enter Plot Code (max. 8-char) >U @X0,Z0
Do you wish to label your plot (Y/N) >!CR
C-37

..X-FACTOR = .10000E+01
..Y-FACTOR = .10000E+01
..X-OFFSET = .00000E+00
..Y-OFFSET = .00000E+00
Would You Like to Write Out the Results (Y/N)? >!CR
..>>RESULT PROCESSING COMPLETED <<
- Hit <CR>to Continue
!CR
C-38

(0) = Previous Menu
(9) = Pressure [P]
(13
)
(14
)
(15
)
Enter a Number >0
C-39
========================================================
==========
========================================================
==========
Enter a Number >3

Enter Load Case Number or <CR>for Default >!CR
========================================================
==========
MENU FOR MISCELLANEOUS CAPABILITIES
========================================================
==========
(0) = Go Back to Previous Menu
(1) = Integrate a Fluid Quantity
Enter a Number >1
C-40
Enter Time Step # or <CR>to Use Default >!CR
ANY OF THE FOLLOWING QUANTITIES MAY BE OVER A SURFACE

P
(0) = Previous Menu

3) (14
(1 )
5) (16
)
(1
8)
Enter a Number >4
..IN SUBROUTINE (INTGR0:1)

..COLLECTING USER DEFINED ELEMENT FACES..
How Do You Want to Pick the Elements?
C-41
(0) = Quit
(1) = By Specifying the Elements ON-SCREEN
(2) = By Reading in a File with Element Numbers
Enter a Number >1
..TOTAL NO. OF ELEMENTS = 27

..SMALLEST ELEMENT NUMBER = 1
..LARGEST ELEMENT NUMBER = 27
Enter Element Numbers (Start,End,Inc,Face# or <CR>to Stop) >
1,7,3,6
10,16,3,6
19,25,3,6
!CR
..NUMBER OF ELEMENTS SELECTED SO FAR = 9
..CHECKING VALIDITY OF INPUT ELEMENTS..
..TOTAL NUMBER OF ELEMENTS SELECTED = 9
C-42
..SETTING FACTOR AND OFFSET FOR DATA MODIFICATIONS..

..CURRENT VALUES OF THE FACTOR AND OFFSET ARE:
..FACTOR = .10000E+01
..OFFSET = .00000E+00
Enter a Case Title (Max 80 Char) >
Volume Flow Rate at the Entry Plane for the Couette Flow Problem
..RESULTS OF INTEGRATION PERFORMED..
INTEGRATION OPTION = Volume Rate of Out-Flow
CASE TITLE: Volume Flow Rate at the Entry Plane for the Couette Flow Prob-
lem
RESULT INTEGRATED = -.44170E+00
OVER THE TOTAL AREA = .10000E+00
COMP# 1 OF AREA = -.10000E+00
COMP# 2 OF AREA = .00000E+00
COMP# 3 OF AREA = .00000E+00
EXTENTS OF THE CHOSEN ELEMENT FACES ARE:
MINIMUM X-COORD. = .00000E+00
MINIMUM Y-COORD. = .00000E+00
MINIMUM Z-COORD. = .00000E+00
MAXIMUM X-COORD. = .00000E+00
MAXIMUM Y-COORD. = .10000E+01
MAXIMUM Z-COORD. = .10000E+01
W- Hit <CR>to Continue
-!CR
C-43

ANY OF THE FOLLOWING QUANTITIES MAY BE OVER A SURFACE
XV

(0) = Previous Menu
3) (14
(1 )
5) (16
)
(1
8)
Enter a Number >4
C-44
..IN SUBROUTINE (INTGR0:1)

..COLLECTING USER DEFINED ELEMENT FACES..
How Do You Want to Pick the Elements?
(0) = Quit
(1) = By Specifying the Elements ON-SCREEN
(2) = By Reading in a File with Element Numbers
Enter a Number >2
Enter File Name to be Read or <CR>to Quit>
ELEM.DATA
**
** This data file has the element and corresponding face numbers
** at the outflow plane for the Couette flow problem
**
3,9,3,4
12,,,4
15,,,4
18,,,4
21,27,3,4
*END
..NUMBER OF ELEMENTS SELECTED SO FAR = 9
..CHECKING VALIDITY OF INPUT ELEMENTS..
..TOTAL NUMBER OF ELEMENTS SELECTED = 9
..SETTING FACTOR AND OFFSET FOR DATA MODIFICATIONS..
C-45
..CURRENT VALUES OF THE FACTOR AND OFFSET ARE:

..FACTOR = .10000E+01
..OFFSET = .00000E+00
Enter a Case Title (Max 80 Char) >
Volume Rate Flow through the Exit Plane for the Couette Flow problem
CASE TITLE: Volume Rate Flow through the Exit Plane for the Couette Flow
problem
RESULT INTEGRATED = .44170E+00

COMP# 1 OF AREA = .10000E+01
COMP# 2 OF AREA = .00000E+00
COMP# 3 OF AREA = -.16653E-15
- Hit <CR>to Continue
!Q
** PROGRAM STOPPED DUE TO USER REQUEST **
C-46
End of file: COUTTE.LOG
Begin file: COUTTE.XYZ

.00000E+00 .00000E+00;U @X0,Z0
.24598E+00 .33333E+00
.57910E+00 .66667E+00
.10000E+01 .10000E+01
-99999 -99999
End of file: COUTTE.XYZ
Begin file: COUTTE.OUT

POST PROCESSING NISA/3D-FLUID RESULTS .................
JOB TITLE: 3D COUETTE FLOW
DATE RUN MADE: 4-19-1989
RUN MODE: GO
PROBLEM DIMENSION: 3D
TOTAL NUMEBR OF NODES = 64
LARGEST NODE I.D. = 64
TOTAL NUMEBR OF ELEMENTS = 27
LARGEST ELEMENT I.D. = 27
MINIMUM X-COORDINATE = .00000E+00
MAXIMUM X-COORDINATE = .20000E+01
MINIMUM Y-COORDINATE = .00000E+00
MAXIMUM Y-COORDINATE = .10000E+01
MINIMUM Z-COORDINATE = .00000E+00
MAXIMUM Z-COORDINATE = .10000E+01
TOTAL NUMBER OF LOAD CASES = 1
C-47
** CASE TITLE: Volume Flow Rate at the Entry Plane for the Couette Flow
Problem **
Factor = .10000E+01, Offset = .00000E+00
.. Details of the Elements and Results Used for Integration ..
Fig C.1 Missing nodal boundary conditions checks made

with the help of FLUTL
C-48
CASE TITLE: Volume Flow Rate at the Entry Plane for the Couette Flow Prob-
lem
RESULT INTEGRATED = -.44170E+00
COMP# 1 OF AREA = -.10000E+01
COMP# 2 OF AREA = .00000E+00
COMP# 3 OF AREA = .00000E+00
** CASE TITLE: Volume Rate Flow through the Exit Plane for the Couette
Flow problem **
Factor = .10000E+01, Offset = .00000E+00
.. Details of the Elements and Results Used for Integration ..

C-49
CASE TITLE: Volume Rate Flow through the Exit Plane for the Couette Flow
problem
RESULT INTEGRATED = .44170E+00
COMP# 1 OF AREA = .10000E+01
COMP# 2 OF AREA = .00000E+00
COMP# 3 OF AREA = -.16653E-15
End of file: COUTTE.OUT
Begin file: ELEM.DATA

**
** This data file has the element and corresponding face numbers
** at the outflow plane for the Couette flow problem
**
3,9,3,4
12,,,4
15,,,4
18,,,4
21,27,3,4
*END
End of file: ELEM.DATA
C-50
Appendix
D
Solving Thermal Stress Problems Using NISA/
Appendix D
HEAT and NISA II
Many of the engineering problems require coupled analysis of heat transfer and
statics. Many of the engineering problems require coupled analysis of heat
transfer and statics. Consider the design of an exhaust manifold of an automobile.
Exhaust gases are emitted through the exhaust manifold. The hot gases flowing
through the exhaust manifold heat up the manifold. This causes the structure to
deform and induces unwarranted stresses. Therefore, one should be able to
perform thermal stress analysis using temperature distribution obtained by a
thermal analysis package.
A method to solve thermal stress problems using NISA/HEAT and NISA II has
been outlined. The temperature distribution predicted by the NISA/HEAT package
is available for thermal stresses calculation in the solid using NISA II.
NISA/HEAT is a general purpose heat transfer analysis package. Arbitrary 2D, 3D

and axisymmetric domains can be analyzed. On the other hand NISA II can be
used to analyze stresses and deformations in a structure due to various loadings
such as pressure, force, moment, linear and angular acceleration, displacement
and temperature. Both NISA II and NISA/HEAT are finite element based
packages and have a common pre- and post-processor DISPLAY III. Both the
packages use the same format to define the geometry of the model. An interface
between the two packages has been developed so that data generated by one can
be easily transferred to the other.
D-1
Solving Thermal Stress Problems Using NISA/HEAT and NISA II
Modeling and Analysis Procedure
The user creates a model in DISPLAY III to run heat transfer analysis
(ANAL=HEAT). All the relevant boundary conditions are applied to the model.
For any NISA/HEAT analysis, files 26 and 27 must be saved which are used for
post processing of final results. For thermal stress analysis, the user must also save
file 55. This can be done via the executive card SAVE = 26,27,55. File 55
contains the temperature distribution in the solid domain, that serves as the
temperature boundary condition for the static analysis (*NDTEMPER card).
The finite element mesh created for heat transfer analysis can also be used for
doing static analysis. The static analysis model is generated in the following
manner: first all the boundary condition applied for heat transfer analysis are
removed using the command HBC,DEL,BC_DATA or by the following sequence
of operations in menu mode
HOT B.C. Delete All
All the relevant boundary conditions for the static analysis, other than temperature,
are applied to this model. Since a reference temperature is needed at which thermal
stress is zero, this information must be inputted for the static model.
Since the statics model considers multiple boundary conditions from heat transfer
model as different load cases, the analysis procedure is somewhat different for
steady state and transient analysis. For steady state analysis, there is only one load
case, and the boundary condition file 55 created by the fluids module can be
included in the static nisa file by using *READ, FILE_NAME.DAT command. The
statics code will read the boundary conditions from the file FILE_NAME.DAT. On
the other hand, for transient analysis, there are multiple load cases, and if the
thermal stress analysis is performed after certain number of time steps, the data
files created by the fluid flow analysis are read in the following fashion:
*NDTEMPER, TCRV = n, FN = FILE_NAME, Step = m, where TCRV = n is
only required for non-linear static analysis, while step = m indicates the time step
identification number on the file, FILE_NAME that would be accessed by the
static analysis for the present load case.
The user first creates two models for performing heat and static analysis. In the
statics model he/she includes the appropriate *READ card, and the Heat-Static
D-2
Interaction module, heatst (for example for interaction of Incompressible NISA/

HEAT with NISA-II) can be invoked, which in turn calls NISA/HEAT to perform
the fluid flow analysis and create the necessary temperature boundary conditions
for the static analysis. After the heat transfer analysis is done and the necessary
data files are created, NISA II is called which then performs the static analysis,
using the static NISA file created beforehand (with the *READ cards or
FN=FILE_NAME.DAT to include appropriate boundary conditions). At the end of
this analysis, one can get the stresses and displacement in the structure due to
pressure and/or thermal loading as a result of fluid flow. For this class of problem,
the flow chart is given in Figure D.1. An alternative approach to transfer pressure
using DISPLAY is given in Figure D.2.
Fig D.1 Fluid-structure interaction between fluid flow and solid structure
D-3
Fig D.2 Fluid-sructure interaction between fluid flow and solid shell
D-4
Appendix
E
System of Units
Appendix E
NISA/HEAT is entirely independent of the units of the physical quantities

specified in the input data. The only requirement is that the units used must be self
consistent. To aid the user in preparing the input, consistent units for some
common engineering and physical quantities used in various analysis types are
given in Table E.1 and Table E.2. Conversion factors for various units are
provided in Table E.3 through Table E.9. Some commonly used dimensionless
groups in fluid and heat transfer analyses are shown in Table E.9 .
In these tables, three systems of units are cited, the International System (SI), the
metric absolute system (CGS) and the British Gravitational System (BGS).
The SI system is based on the fundamental dimensions of length (L), mass (M)
and time (t). Length is measured in meters (m), mass in kilograms (kg) and time in
seconds (sec). The program expects, however, the mass and the mass density to be
specified in force units (i.e., F.t2/L and F.t2/L4, respectively). Therefore, for the
users convenience, Table E.1 lists force as a fundamental unit, i.e., mass is
derived from force through Newtons law F = Ma, where a is the gravitational
acceleration.
The CGS system (Centimeter, Gram and Seconds) is based on the fundamental
dimensions of length (L), mass (M) and time (t). Length is measured in
centimeters (cm), mass in grams (gm) and time in seconds (sec). As in the SI
units, Table E.1 lists force as a fundamental unit and mass and mass density are
specified in force units (i.e., F.t2/L and F.t2/L4, respectively).
E-1
System of Units
The BGS system is based on the fundamental dimensions of length (L), force (F)
and time (t). Length is measured in feet (ft), force in pounds (lbf) and time in
seconds (sec).
It should be noted that in Table E.2, the dimension of heat energy is given the
symbol (E) for convenience. The units may be in Joules (J) or British Thermal
Units (Btu) which are equivalent to force times length (F.L), see Table E.7 for
conversion factors. Also, time is kept explicit in the units given in Table E.2 (e.g.,
J/sec or Btu/sec). Note however, that Watt (W) is defined as the power that
produces energy at the rate of one Joule per second, i.e., W = J/sec.
To use Tables Table E.3 through Table E.7, the units to be converted should be
located in the left-hand column whereas the required units should be located in the
top row of the table. The conversion factors common to both units should be
multiplied by the quantity to be converted. As an example, in Table E.3 one meter
(1m) is equal to 39.37008 inches.
The conversion factors have been reported to seven significant digits. However,
this does not mean that the factors are known to this accuracy. Factors given with
fewer than seven digits are exact.
To derive conversion factors for quantities not listed in Table E.3 through Table
E.9, it is always convenient to use unity factors. The following two examples
explain the use of such factors:
Example 1:
4
To change the Stefan-Boltzmann constant from W/m2.K4 to Btu/hr.ft2. R , use the
following
factors:
E-2
System of Units
8 W 8 J hr ft 2 R4 Btu
= 5.67033 10 ------------------- = 5.67033 10 -------------------------------- ------------------------------- -------------------------------
2 4 2 4 Btu 2 4
m K sec m K hr ft R
2 4
8 J hr ft R Btu
= 5.67033 10 --------- ------- ------- --------- -------------------------------
Btu sec 2 4 2 4
m K hr ft R
8 4 3 2 5 4 Btu see Table E.3,
= 5.67033 10 ( 9.479735 10 ) ( 3.6 10 ) ( 0.3048 ) ( --- ) -------------------------------
9 2 4 E.7 and E.9
hr ft R
8 2 4
= 0.171257 10 Btu hr ft R
Note:
When changing from one system of units to another, all the relevant
nondimensional parameters (Biot number, Numlet number, etc., as applicable),
must be held constant.
E-3
System of Units
Table E.1 Units for commonly used quantities++

UNITS
QUANTITY DIMENSIONS
SI METRIC (CGS) IMPERIAL (BGS)
Fundamental
Quantities
Force F N dyne lbf
Length L m cm ft
Time t sec sec sec
Temparature T K K R
Derived Quanti-
ties
Angle rad rad rad
Coefficient of
thermal expan- 1/T 1/K 1/K 1/R
sion
Mass density 2 4 3 2 4 3 2 4 3 2 4 3
F t L M LN sec m kg m dyne sec cm gm m lb
cm f sec ft slug ft
++
applicable notes are given following Table E.9
E-4
System of Units
Table E.2 Units for commonly used quantities++

UNITS
DIMEN-
QUANTITY METRIC IMPERIAL
SIONS(1) SI
(CGS)(1) (BGS)(2)
Convective heat J/sec.cm2.K
E/t.L2.T J/sec.m2.K Btu/sec.ft2 R
transfer coefficient
Density(2) M/L3 kg/m3 gmm/cm3
lbm/ft3
Heat energy E(J or F.L) JNm J 10
7
dyne.cm Btu
Heat flux
E/t.L2 J/sec.m2 J/sec.cm2 Btu/sec.ft2
Heat generation
E/t.L3 J/sec.m3 J/sec.cm3 Btu/sec.ft3
rate
E/M J/kg J/gmm Btu/lbm
Latent heat(2)
Specific heat(2) J/kg.K

E/M.T
J/
J/gmm.K Btu/lbm R
Stefen-Boltzmann E/t.L2.T4 J/sec.cm2.K
sec.m2.K4 Btu/sec.ft2. R
4
constant(3)
Temparature(1) T
K K R
Thermal
conductivity
E/t.L.T J/sec.m.K J/sec.cm.K Btu/sec.ft. R
++
applicable notes are given following Table E.9
E-5
System of Units
Table E.3 Conversion factors for LENGTH
Centi-
Meter Millimeter Inch Foot
LENGTH meter
m mm in ft
cm
Meter
1.0 1.0 x 10+2 1.0 x 10+3 3.937008 x 10+2 3.280840
m =
Centimeter
1.0 x 10-2 1.0 1.0 x 10+1 3.937008 x 10-1 3.280840 x 10-2
cm =
Millimeter
1.0 x 10-3 1.0 x 10-1 1.0 3.937008 x 10-2 3.280840 x 10-3
mm =
Inch
2.54 x 10-2 2.54 2.54 x 10+1 1.0 8.333333 x 10-2
in =
Foot 3.048 x
3.048 x 10-1 3.048 x 10-2 1.2 x 10+1 1.0
ft = 10+1
Table E.4 Conversion factors for FORCE
Newton Dyne Kilogramf Poundf Poundal

FORCE
N dyne kgf lbf Pdl
Newton
1.0 1.0 x 10+5 1.019716 x 10-1 2.248089 x 10-1 7.233015
N =
Dyne 7.233015 x 10-
1.0 x 10-5 1.0 1.019716 x 10-6 2.248089 x 10-6
dyne = 5
Kilogramf 7.093165 x
9.80665 9.80665 x 10+5 1.0 2.204623
kgf = 101
Poundf 3.217405 x
4.448221 4.44822 x 10+5 4.535924 x 10-1 1.0
lbf = 10+1
Poundal 1.382549 x 10-

1.382549 x 104 1.409808 x 10-2 3.108095 x 10-2 1.0
Pdl = 1
E-6
System of Units
Table E.5 Conversion factors for MASS
dyne.sec2/
N.sec2/m lbf.sec2/ft
MASS kg f.sec2/m cm pdl.sec2/in lbf.sec2/in
(kg) (gmm) (slug)
N.sec2/m 1.019716 x 10-

=
1.0 1.0 x 10+3 6.852178 x 10-2 2.204623 5.710148 x 10-3
(kg) 1
kg f.sec2/ 9.80665 x
9.80665 1.0 6.719691 x 10-1 2.161997 x 10+1 5.599743 x 10-2
m = 10+3
dyne.sec2/ 1.019716 x 10-

= 1.0 x 10-3 1.0 6.852178 x 10-5 2.204623 x 10-3 5.710148 x 10-6
cm (gmm) 4
lbf.sec2/ft 1.459389 x 1.459389 x

1.488152 1.0 3.217403 x 10+1 8.333333 x 10-2
(slug) = 10+1 10+4
pdl.sec2/ 4.535924 x 10- 4.625354 x 10- 4.535924 x

3.108096 x 10-2 1.0 2.590081 x 10-3
in = 1 2 10+2
lbf.sec2/ 1.751268 x 1.785796 x 1.751268 x

1.2 x 10+1 3.860883 x 10+2 1.0
in = 10+2 10+1 10+5
Table E.6 Conversion factors for MASS DENSITY
N-sec2/ dyne-sec2/
kg f-sec2/m4 lbf-sec2/ft4 pdl-sec2/ft4
DENSITY m4 cm4 lbf-sec2/in4
(slug/ft3) (lbm/ft3)
(kg/m3) (gmm/cm3)
N-sec2/m4 (kg/ 9.357253 x 10-

=
1.0 1.0 x 10-3 1.019716 x 10-1 1.940321 x 10-3 6.242795 x 10-2
m3) 6
dyne-sec2/cm4 6.242795 x 9.357253 x 10-

1.0 x 10+3 1.0 1.019716 x 10+2 1.940321
(gmm/cm3) = 10+1 3
kg f-sec2/m4 9.80665 9.80665 x 10-3 1.0 1.902804 x 10-2 6.122091 x 10-1 9.176331 x 10-
= 5
lbf-sec2/ft4 5.153789 5.153786 x 3.217405 x 4.822531 x 10-

5.255401 x 10+1 1.0
(slug/ft3) = x 10+2 10-1 10+1 3
pdl-sec2/ft4 1.601847 1.601847 x 1.498889 x 10-

1.633429 3.108095 x 10-2 1.0
(lbm/ft3) = x 10+1 10-2 4
lbf-sec2/in4 1.068697 1.068690 x 2.073600 x 6.671661 x

1.089761 x 10+4 1.0
= x 10+5 10+2 10+2 10+3
E-7
System of Units
Table E.7 Conversion factors for ENERGY

dyne.cm IT++
N.m(J)++ ft.lbf
ENERGY
(erg)
ft.pdl
( cal ) Btu
N.m(J)+ 7.375621 x 2.373036 x 2.388459 x 9.478172 x

1.0 1.0 x 10+7
= 10-1 10+1 10-1 10-4
dyne.cm 7.375621 x 2.373036 x 2.388459 x 9.478172 x

(erg) 1.0 x 10-7 1.0
= 10-8 10-6 10-8 10-11
ft.lbf 1.355818 x 3.217405 x 3.238315 x 1.285067 x

1.355818 1.0
= 10+7 10+1 10-1 10-3
ft.pdl 4.214011 x 4.214011 x 3.108095 x 1.006499 x 3.994110 x 10-

1.0
= 10-2 10+5 10-2 10-2 5
(cal)IT+++ 9.935427 x 3.968321 x

4.1868 4.1868 x 10+7 3.088025 1.0
= 10+1 10-3
Btu 1.055056 x 1.055056 x 7.781693 x 2.503687 x 2.519958 x

1.0
= 10+3 10+10 10+2 10+4 10+2
++
1J = 0.999835 int J
+++
(cal)IT indicates international steam table calorie, (cal)IT = 1.000669 (cal), where (cal) indicates thermochemical
calorie.
E-8
System of Units
Table E.8 Conversion equations for TEMPERATURE++

TEMPER-
ATURE T ( C ) T(K) T ( F ) T ( R )
5--- 5---
T ( C ) = T ( C ) T(K) - 273.16
9
( T ( F ) 32 ) ( T ( R ) 491.68 )
9
T(K) 5--- 5
T ( C ) + 273.16 T(K) ( T ( F ) + 459.68 ) --- ( T ( R ) )
= 9 9
9--- 9---
T ( F ) =
( T ( C ) + 32 )
5
( T ( K ) 255.38 ) T ( F ) T ( C ) 459.68
5
9--- 9---
T ( R ) =
( T ( C ) + 491.68 ) T(K) ( T ( F ) + 459.68 ) T ( R )
5 5
++
Note that conversions of the actual temperature values are governed by the
equations given in the table, but the relations between the intervals of the
temperature scales are given by:
1 degree Celsius = 1 degree Kelvin
= (9/5) degree Fahrenheit
= (9/5) degree Rankine
For example, to change a temperature of 100 C to F, we use the following
equation from the table:
T ( F ) = (9/5) T ( C ) + 32
= ( 9 5 ) ( 100 ) + 32 = 212 F
6
However, to change a quantity of 1.0 10 ( 1 C ) to( 1 F ) we use
6 1 6 1 F 6 F 1
1.0 10 ------ = 1.0 10 ------ ( ----- ) = 1.0 10 ( ------ ) -----
C C F C F
6 F 1
- ----- = 5--- 10 6 ( 1 F )
= 1.0 10 -------------------
( 9 5 ) F F 9
E-9
System of Units
Table E.9 Dimensionless groups
Group* Symbol Name

2 Fo Fourier number
t r 0
4 Radiation Conduction
T h L ( K ( T h T c ) ) Nrc
Parameter
c Specific Heat
d Characteristic length (usually transverse)
h Heat transfer coefficient
k Thermal conductivity
L Characteristic length
r0 Characteristic length
t Time
Thermal Diffusivity k c
Volumetric coefficient of thermal expansion
Mean force path
Viscosity
Density
Surface Tension
Notes:
1. Units of temperature may be interchanged, i.e., degree Kelvin ( K ) with

degree Celsius ( C ) and degree Rankine ( R ) with degree Fahrenheit ( F ) ,
provided that other quantities involving temperature are compatible and no
radiation boundary conditions are specified.
2. For convenience, the units of specific heat and other mass related quantities
are specified per unit mass. In this case the units of density should be mass per
volume (M/L3). If weight density is used instead, i.e., units of F/L3, the units of
the specific heat and other mass related quantities should be per unit weight.
3. If radiation boundary conditions are specified, the Stefan-Boltzmann constant
as well as all quantities involving degree temperature in their units (e.g., heat
E-10
System of Units
transfer coefficient, conductivity, specific heat), have to be specified in terms

of absolute degrees, i.e., degree Kelvin ( K ) or degree Rankine ( R ) .
E-11
System of Units
E-12
Keyword Index
Keyword Index
Appendix F
A
Anisotropic properties
input cards - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - 6.2-4, 7.5.1-40
Axisymmetric problems - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - A.3-13
C
Conduction heat transfer
example problems - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - A.1-1
conduction heat transfer
Example problems - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - A.12-59
Convection heat transfer coefficient
computation using FLUTL - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - C.3-1
Convergence - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - 3.4-2-14
convergence - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - 3.3-7
Covection boundary condition
example problems - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - A.4-18
input cards - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - 8.2.2-8
D
Disk space management - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - 9.2.2-3
E
Element library - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -1
Error indicators/monitoring - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - 3.4-12
Executive Commands - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -1
F
Finite element formulation - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - 2.4-8
F-i
Keyword Index
Fluid utility program, FLUTL - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -1

Fluid-structure interaction
procedure - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -D.1-3
G
Galerkin formulation - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - 2.4.1-8
Governing equations - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - 2.5.2-17
H
Heat flux - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - 8.2.2-8
example problems - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -A.1-1
Heat generation
input cards - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - 8.2.2-8
I
Iterative solver
input cards - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - 2-16
iterative solver
example problems - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -A.12.4-63
N
Nonlinear problems - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - A.6-29
P
Phase change
example problems - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - A.14-81
theory . - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - 2.1-1
R
Radiation boundary condition
input cards - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - 10-5
radiation boundary condition
Running NISA/HEAT - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - 9.1-1
S
Shape function - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -2.4-1-8
Specified temperature
F-ii
Keyword Index
input cards - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - 8.2.2-8

Steady state problems
input cards - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -2-4
useful hints - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - 9.1-1
Surface radiation
input cards - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - 2-16
theory - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - 2.5-17
useful hints - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - 1-10
T
Temperature dependent boundary condition
input cards - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - 8.2.2-8
Temperature dependent material properties
example problem - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - A.6-29
Transient analysis
Transient problems
Theory - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -2.10-35
useful hints - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - 2-11
W
Wavefront - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - 3.2.1-4
Wavefront Optimizer - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - 3.2.1-4
Wavefront solver - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - 3.2-1
F-iii

HEAT3 User

Transféré par

Informations du document

Titre original

Copyright

Formats disponibles

Partager ce document

Partager ou intégrer le document

Options de partage

Avez-vous trouvé ce document utile ?

Ce contenu est-il inapproprié ?

Droits d'auteur :

Formats disponibles

HEAT3 User

Transféré par

Droits d'auteur :

Formats disponibles

Disclaimer

CRANES SOFTWARE, INC. MAKES NO WARRANTY OR

Any questions relating to the use or interpretation of the SOFTWARE or their

Cranes Software, Inc.

NOTICE FOR U.S. GOVERNMENT USERS ONLY: RESTRICTED RIGHTS

Use, duplication or disclosure by the Government is subject to restrictions as set

ADVANCED DYNAMICS: Multiple Support Excitation, for shock spectrum

DYSPAN: Spectrum compatible Power Spectral Density generation.

ENDURE: Improved functionality of features in Shell; expansion of crack

Up-gradation of translators such as SAT2NEU translator to support ACIS R16

EMAG: 3D-Magnetodynamic analysis (harmonics) and 3D Transient Mag-

NISA/HEAT3 and 3DFLUID: Heat Flux computation and Printing of Heat

NISA/CIVIL: Revised code of practice conforming to ACI 318 -2005, BS

1.1 General Description

1.1.1 NISA/HEAT Capabilities

Steady and transient analysis

- Heat flux normal to a surface ( q nc )

- Heat flux along global axes qcx , qcy , qzc

h: heat transfer coefficient

a: variation with time is allowed

b: variation with temperature is allowed

c: variation with both time and temperature is allowed

Temperature dependent material properties

- Temperature contour plots

- Temperature vs. time plots

- Temperature vs. distance plots

- Heat flux calculation at boundary faces

- Temperature vs. time data for a node

- Radiation heat flux

1.2 NISA Suite of FEA Software

1.2.1 NISA Softwares

inside the structure. There is no need to model the boundaries by parametric

Various displacement and stress constraints may be specified for optimization

A program for analyzing the fatigue performance of engineering structures.

manually during preliminary design or directly interfaced from NISA static or

Other capabilities include: Analysis of several PCBs or an entire Avionics box,

NISA-3D/FLUID interfaces with DISPLAY for graphical representation of results,

Major areas of applications of the code include machine mechanisms, robotics,

A general purpose finite element program used to analyze a wide spectrum of

NISA-3D/Heat interfaces with DISPLAY for graphical representation of results,

A program for aeroelastic analysis. In the present version of NISA/AERO,

1.2.2 NISA Documentations

User and verification manuals

1. NISA User's Manual

7. NISA/EMAG User's Manual

1. DISPLAY III Training Manual

2.2 Governing Equations

Figure 2.2.1 Three dimensional conduction in a body

The heat conduction equation or the law of conservation of energy is obtained by

- If the material is isotropic with constant conductivity, k, the equation

- If the material is isotropic with constant conductivity, k, and steady state

- If the material is isotropic with constant conductivity, k, and no heat gen-

- If the material is isotropic, with constant conductivity, k, no heat genera-

In the above equations,

2.3 Boundary Conditions

(a) Prescribed temperature or Dirichlet B.C.:

(b) Prescribed heat flow or flux (Cauchys or Neumann B.C.):

(c) No heat flow or adiabatic B.C.:

(d) Convective heat exchange (Convection B.C.):

When part of the surface of the body, say S c where S c CS , is in con-

(e) Radiation boundary condition:

Assuming a body of surface temperature T s , is enclosed within a space

is an equivalent radiation coefficient which is strongly dependent on temperature.