Académique Documents
Professionnel Documents
Culture Documents
Author:– Date:–
Athol J Carr 25 September 2004
RUAUMOKO
The program is designed to run in an interactive mode or in a batch mode with an input data file that may be
prepared using a word-processor, a spreadsheet or a text editor. If a word-processor or spreadsheet is used
to prepare a data file then the data MUST be saved as a plain ASCII text file. In the default form the program
will prompt the user for the input data. The input data supplied may be logged so that it can be used for later
analyses and prepared data files may be used for all or part of the input data. For further details on the use of
the command processor used in the program type $HELP at any prompt. Complete all responses by pressing
the RETURN/ENTER key.
To run the program call the program by the method appropriate to your operating system. On a personal
computer just type RUAUMOKO2D assuming that the file RUAUMOKO2D.EXE and FQWIN.HLP are in your
current directory or path.
In Microsoft Windows operating systems another option is to create a shortcut on the desktop and for this
purpose a suitable icon for RUAUMOKO, Ruaumoko2D.ico, is supplied with the program.
The program prompts for responses to a series of questions. Default responses, where appropriate, are
enclosed in square brackets, [ ]. File names must match the conventions of your operating system but file
names, with paths where necessary, must not exceed 60 characters in length and must not contain blanks.
The first question asks for the name of the output file. The default is the computer console or terminal screen.
The next two questions are for names of the files containing the earthquake accelerograms or the excitation
histories. The default value assumes that the excitations are appended to the rest of the input data. If there
are less than three inputs records just hit the RETURN key for the subsequent requests. If the file names are
included in the excitation part of the data file then just hit the RETURN key at each prompt.
The next question asks for the name of the DYNAPLOT post-processor file. This file will only be opened if
requested by a greater than zero value of the variable KPA later in the data. This file is automatically given
the extension .RES or .RAS if required, or if the name shown, which uses the filename of the output file, is
acceptable, just hit the RETURN key.
The next question asks for the selection of on-screen graphics. Any answer not of the form Y, YE or YES
(upper or lower case) will suppress the graphics. If the answer was YES and the screen supports colour
graphics you will later be asked if you want COLOUR graphics or not. Note; if on-screen graphics is enabled
the screen cursor is invisible at the end of all future prompts.
The next question asks if this is a data check run only. If so type Y, YE or YES. In the case of the data check
run the analysis is terminated at the completion of the modal analysis and just before the start of the time-
history analysis. The total storage requirements for the time-history analysis is checked.
The program then displays information on logging or adding data files
The next prompt is for the title name of your analysis which is the first line of the data described in the
following sections of this manual. If the user already has a data file ready then type $ADD filename, where
filename is the name and path if necessary of the data file, followed by RETURN. If the user is going to key
the data in as prompted, it is suggested that the all of the data be logged in a file for later use and to save re-
supplying this data in the event of a data error by giving the command $LOG filename, where filename is the
name of the file where data is to be logged, followed by RETURN. Note that all responses prior to this one will
not be logged even if logging had been enabled at an earlier prompt.
At the completion of reading and checking of all of the input data for the analysis, the program, if it is running
in an interactive mode and the on-screen graphics is enabled, the user will be prompted as to whether colour
graphics is wanted.
The on-screen graphics draws a picture of the structure with all members numbered and the analysis pauses
until the RETURN key is pressed. If you want a hard copy use the pull-down menu at the top left of the
window to print the picture or to save it as a The program will next plot the accelerograms or excitation
histories scaled to the magnitudes used in the analysis. These are cleared by pressing the RETURN key.
2
After the modal analysis is complete the user is asked if the mode shapes are to be displayed in an animated
display. Respond by typing the mode number followed by RETURN. Five and a quarter cycles of vibration
are displayed unless a different number of cycles is specified. The default response END terminates the
plotting of mode shapes and starts the time-history analysis.
During the time-history, on-screen pictures of the deformed shape of the structure showing the locations of
inelastic behaviour are presented but require no user action because the pictures are replaced automatically
as soon as a new picture is ready. Excessive use of the screen capture during the time-history may cause the
system to crash.
The program RUAUMOKO-2D may be run in a batch mode as well as the interactive mode described above.
In the batch mode the graphics is disabled and the test data option is also disabled.
The START/w is only required for Windows95 or Windows98 operating systems to prevent the system from
attempting to multi-task following command lines. For WindowsNT, Windows2000 or WindowsXP this item
should be removed.
If the RUAUMOKO-2D executable file RUAUMOKO2D.EXE is not in your path then the path should be part of
the command.
The outputfile is the name (including the path) of the output file. This file must NOT exist. If the file does
exist the analysis will terminate immediately. The post-processor file, if requested in the input data, will have
the same file name (and path) but with the extension .RES or .RAS .
The inputfile is the name (including path) of the input data file (in the form required in the following part of this
manual) and the file MUST exist. If the file cannot be found, because the name or path is incorrect then the
analysis will terminate.
The files quake1file and quake2file are the names (including paths) of the excitation files. These file names
are optional, if they are missing or the word INPUT (in upper case) is supplied instead than they are assumed
to be part of the inputfile above. If files names are supplied then the files MUST exist or the analysis will be
terminated.
If there are less than two arguments following the command RUAUMOKO2D then the programs will run in the
normal interactive mode.
As many such command lines may be arranged in a .bat file, the best example was a visitor to the department
who set up 475 runs in a batch file and went away for days skiing. The only real limitation is the amount of
disk required for the output files, especially if the post-processing file is also output.
Enviroment Settings.
RUAUMOKO-2D makes great use of a local random access scratch file during th analysis. At each time-step
and in each iteration a read and a write operation is carried out for each member in the structure. The default
location for this file RUAUMOKO.SCR is in the users current working directory. When the analysis is
completed this file is deleted by the program, however, if the program is terminated prematurely by the user
this file may be left in the directory. This will not cause any difficulties as it will over-written the next time the
program runs. However, if the program is being run over a network and the users working directory is on a
server, then a very large penalty will be paid for the network traffic and the program may appear to run at
about 40% of its normal speed. The tmp and temp environment variable should be set to a local hard disk on
the computer where the analysis will actually take place. On all of the Windows workstations in the School of
Engineering, University of Canterbury, computer laboratories these variables are set to C:\TEMP which is a
3
local scratch directory on each workstation. If the programs are being run on a single personal computer there
is no necessity to set the environment variables.
On the latest personal computer version of RUAUMOKO-2D most of the scratch file is emulated in memory
which has increased the execution speed of the program by factors of up to 8. However, should the number
of sectors required for the scratch file exceed the memory allocated the remainder of the disk sectors will be
written to the hard disk file mentioned above. In all cases, even if it is not used, the file RUAUMOKO.SCR will
be created at the start of the program execution and deleted at the end of the analysis. (This disk in memory
approach has been successfully used in VMS and unix workstations since 1989).
Do not attempt to carry out simultaneous executions of the program unless they are run from different
directories and the environment variable are not set as the scratch file would be accessed by the different
analyses with great confusion as the result.
Note: In the following user guide, each line of required data is indicated by a box containing the data items.
Below each box is a description of the data items. The data items on each line may be separated by commas
or blank spaces. The format for the items are indicated by the letter at the end of each descriptive line with A
indicating a character string, I indicating an integer value and F indicating a floating point number. A floating
point number may or may not have a decimal point and may also take a scientific or exponent form such as
1.5E6 which could also be expressed as 1500000.0. Character strings will be upper-cased unless enclosed in
double or single quotes and will terminate at the first blank space unless the string is enclosed in quotes.
4
OVERVIEW OF THE DATA FOR "RUAUMOKO-2D"
The analysis data for the structure is described by the following sequence of input lines or card images. Each
data set consists of the following items of data.
12 EQUAKE - one line - see section 24. This is only required if dynamic excitation is applied.
This is followed by a line with the parameters for the accelerogram or the loading history.
This then may be followed by the accelerogram or loading history though in most cases the
accelerogram is read from a separate file.
13 EQUAKE - one line - see section 24. This is only required if there are two components of the
accelerogram or loading history.
This is followed by a line with the parameters for the accelerogram or the loading history.
This then may be followed by the accelerogram or loading history though in most cases the
accelerogram is read from a separate file.
14 EQUAKE - one line - see section 25. This data is only required for a Cyclic Adaptive Pushover Analysis.
This is followed by a line with the parameters for the application of the displacement history.
This then may be followed by the displacement history though in many cases the history may be read
from a separate file.
5
15 SPECTRA - one line - see section 26. This is only required if the analysis is for a Response Spectra
Analysis.
This is followed by a line with the parameters for the Acceleration Response Spectra.
This then may be followed by the data for the response spectra though the spectra could be read from a
separate file.
6
THE DATA FILE FOR "RUAUMOKO-2D"
The analysis data for the structure is described by the following sequence of input lines or card images.
IPANAL IFMT IPLAS IPCONM ICTYPE IPVERT INLGEO IPNF IZERO ORTHO IMODE
7
Notes: If IPANAL is greater than 0, a static analysis is carried out if any static loads are specified and then
the modal analysis is performed unless suppressed, or the order of the operations is reversed, by the
variable IPNF.
For IPANAL = 4 or 5. Degrees of Freedom flagged as for travelling wave input are treated as fixed for
the static and modal analyses and are then released at the start of the relative displacement time
history integration.
For IPANAL = 9. Degrees of Freedom flagged as input nodal displacement degrees of freedom are
fixed for the static and modal analyses and are then released at the start of the total displacement
time history integration.
IPANAL = 6 is for an Adaptive Push-Over Analysis where the input load pattern is adapted as the
structure deforms and the analysis will terminate when the limit displacement is reached, the factored
elastic displacement is reached or the structure stiffness reduces to 0.1% of the initial stiffness.
IPANAL = 7 is for a Cyclic Adaptive Push-Over Analysis where the input load pattern is adapted as
the structure deforms and when the change points in the prescribed displacement history is reached
the loading again reverts to the initial loading pattern but in the opposite direction and the adaptive
loading again continues until the next reversal point in the displacement history is reach. The analysis
will terminate when the displacement history is completed or the structure stiffness reduces to 0.1% of
the initial stiffness.
IPANAL = 8 is for the case for simulating a laboratory excitation where a displacement is applied to
the structure at a particular degree of freedom or a set of degrees of freedom. In this case the
degrees of freedom where the displacements are to be applied are treated as free for the static and
modal analyses, i.e. gravity and free-vibration cases and then the displacement histories are applied
to these degrees of freedom. The analysis is a Relative Displacement formulation.
IPANAL = 9 is for the case for the analysis of a structure where the ground is to be subjected to a set
of specified ground displacement histories. This would be appropriate if the different motions are to
be applied at different parts of the foundation or at different depths in the foundation. For the static
and modal analyses these degrees of freedom need to be fixed to prevent rigid body displacements of
the structure. The analysis is a Total Displacement formulation
IPANAL = 10 is for the case for the analysis of a structure where the ground motion is based on an
acceleration response spectra. The data for a normal dynamic analysis is read but then the analysis
is reset to linear elastic, the damping model is reset to Rayleigh damping, the mass matrix is reset to
diagonal and the analysis is set to small displacement theory. There may be separate spectra for the
each earthquake direction or the same spectra may be use for all earthquake component. The
analysis is a Relative Displacement formulation.
ORTHO is a flag that enables the user to show the orthogonality of the mode shapes computed with
respect to the mass matrix. Ideally the diagonal terms of the orthogonality matrix should be 1.0 and
the off-diagonal terms should be 0.0.. This check is normally only used if, for some reason, there is a
suspicion that the modes computed are not orthogonal with respect to the mass or stiffness matrices.
IMODE is used to select the eigenvalue algorithm used to compute the natural frequencies and mode
shapes of free-vibration. The default is the Householder and QR algorithm that has been used since
the mid-1960s. This has always been reliable but in 2001 3 instances were reported where non-
orthogonal modes were computed The Jacobi algorithm is the most stable algorithm available but is
slow, and if any mode shapes are required they all have to be found, recommended for small
problems only. The Sub-space iteration method is suitable for very large problems but must be used
with care if parts of the structure have very small masses with small stiffnesses to match as the
iteration process may then be unpredictable. The iteration uses a generalised Jacobi eigenvalue
routine.
8
Rayleigh or Proportional Damping Model ICTYPE=0 or 1
9
3 Frame Control Parameters
Note: If both MODE1 and MODE2 above are zero then the coefficients C1 and C2 are to be taken as the
constants " and $ of the Rayleigh damping model.
If there is only one dynamic degree of freedom then if MODE1=1 and MODE2=0 then mass
proportional damping is modelled.
If there is only one dynamic degree of freedom then if MODE1=0 and MODE2=1 then stiffness
proportional damping is modelled.
If ICTYPE = 4, User specified Modal damping, then MODE1 is the number of modes for which
damping is specified, 1 # MODE1 # 10. Note that MODE2, C1 and C2 should be zero.
KP KPA KPLOT JOUT DSTORT DFACT XMAX YMAX NLEVEL NUP IRESID KDUMP
Note: The variables JOUT and DSTORT are no longer used in Ruaumoko but are retained to maintain
backwards compatibility of the data files.
If KP is zero for a response spectra analysis (IPANAL=10) then only the combined modal response
will be output whereas if KP is greater than zero each modal response will also be output.
10
5 Iteration Control and Wave Velocities
Notes: 1. If IPVERT is negative then travelling wave models are not permitted.
2. If the ground wave velocities WAVEX and/or WAVEY are input as zero they are treated as
infinite.
3. When the Central Difference marching method is being used for the time-history integration
then the actual time-step DT/MAXIT must be less than half the shortest natural period of the
structure if numerical instability is to be avoided in the time-history integration.
4 The Displacement limits are absolute values in that only positive numbers are used.
5 The default value for FTEST used when either one or both of MAXIT and/or MAXCIT are
specified is 0.00001. Note that the squares of the vector norms are used in the test so that
0.001 implies a 3% tolerance and 0.0001 implies a 1% tolerance.
6 If the variable MAXCIT is applicable the maximum number should lie in the range of 1 to 3.
As a rule, experience has shown that very seldom is more than one cycle of iteration
required.
7 As a general observation not more than about 5 Newton Raphson iterations should be
needed in any time-step
8 If Dispersion is applied to the travelling wave, then the amount of dispersion applied to the
accelerogram applied to nodes after the first node to receive the input depends on the wave
velocity, the distance between the nodes (i.e. the time delay) and the dispersion parameter D
and the characteristic earthquake frequency OMEGA. The dispersion is in the form of a
white noise which is designed not to significantly alter the response spectra of the input
ground motion. The first node to receive the input motion has no dispersion applied to its
input motion.
9. The dispersion amplitude may be scaled by the factor F.
11
5a User Specified Modal Damping Parameters (Only if ICTYPE = 4)
Notes:
1. Number of Modes n specified must equal MODE1 of line 2, Principal analysis options.
2. Damping for modes less than the first mode above is taken as DAMP1 and damping for modes
greater than the last above is taken as DAMPn. This model may be considered as a variation of the
trilinear damping model.
3. For modes with numbers in-between the modes specified, the Percentage of Critical damping is
interpolated with respect to the modal frequencies of the modes at the beginning and end of the gap
in the mode numbers.
Output control during the time-history integration is under the control of the parameters KP and KPA on the
Section 4 above. If the parameter KP is non-zero, = J, the results are written to the output file, or device, every
J time steps starting at step 0. Likewise, if KPA is non-zero, = I, then results go to the post-processing
DYNAPLOT disk file every I time steps starting at step zero.
The control of which nodes or members have their results output is determined by the parameter IOUT
associated with each of the nodes or members in the following data sections.
IOUT = 0; if required, the results will be sent to both the output file and written to the DYNAPLOT file.
= 1; if required, will only be written to the DYNAPLOT file
= 2; if required, will only be sent to the output file.
= 3; will neither be printed or written to the DYNAPLOT file or to the output file.
Default values of IOUT are set on the lines with the words NODES or MEMBERS in Sections 7 or 8 below and
these values will apply for all nodes or members unless over-ridden by new values on the data for that node or
member. Interpolated nodes or members take the value for the first node or member in the sequence.
At every time step the following information may be written to the DYNAPLOT post-processing file.
1. The time, ground accelerations and the inertia, damping and elastic energies, the external work done
by the applied loads and the plastic work.
2. For every requested node, the nodal displacements, velocities, accelerations, damping forces, inertia
forces and applied forces.
3. For every requested member, the member forces, the member deformations, the member action
stiffness flags as a percentage of the elastic stiffness. (axial, flexure at end 1, flexure at end 2) or
(longitudinal displacement, transverse displacement and moment). For the four-hinge beam members
the moments, curvatures and stiffness flags for the internal hinges are also output. For Frame and
Wall members the shear forces at each end are also output.
12
7 Nodal Point Input
NODES IOUT
One input line required per nodal point, except that the presence of sequential nodal points with equally
spaced nodal point coordinates need only be implied by the presence of the first and last nodal points of the
sequence. Boundary conditions will be taken (except for the last nodal point in the sequence) as being
identical to those of first nodal point in the sequence. Coupling nodes will also be interpolated between the
first and last nodal points in the sequence.
Note: There is no restriction of the ordering of master and slave nodes in the program and daisy chaining of
slaving has also been used. However, care must be exercised to ensure that circular chaining does
not occur, i.e. that the node is not effectively slaved back to itself. The unconstrained equations are
numbered and then the program works out the equation numbers of the slaved nodes. Once the
member data has been input the equations are automatically renumbered to maximise the execution
speed in the subsequent equation solving. This means that if one later wishes to apply foundation
compliance to what was a fixed-base structure then the original node numbering does not have to be
changed, the new numbers for the foundation nodes are just appended to the existing node
sequence. Similarly, adjoining frames in a pounding study can be numbered sequentially rather than
numbering across floors of both structures to minimise band-width. Dummy nodes may be left in the
data set as long as they are fully fixed.
13
7a. Inter-storey Drift Input. (Only if NLEVEL on data line 5 is greater than 2)
DRIFT
This is followed by a line or lines with a list of a list of NLEVEL nodal point numbers starting with a node point
at the ground level and then with a node point at each storey progressing upwards from the bottom of the
structure. If the nodes cannot be listed on one line then take as many lines as is necessary. The program will
continue reading data lines until all NLEVEL numbers have been read.
N1 N2 N3 . . . Ntop
14
Four Node Member
15
8 Member Topology or Geometry
ELEMENTS IOUT
One input line is required per member in ascending order, except that sequential members with a nodal point
numbering that uniformly increments may be implied by the provision of the first and last member in the
sequence. The member type for all interpolated members will be taken as that for the first member in the
sequence.
N Member number I
MT Member type number, refers to the member properties that follows I
NODE1 Nodal point number at end 1 of the member I
NODE2 Nodal point number at end 2 of the member I
NODE3 Inner nodal point at end 1 of the member I
(if NODE3 is zero or blank it is reset = NODE1)
NODE4 Inner nodal point at end 2 of the member I
(if NODE4 is zero or blank it is reset = NODE2)
IOUT Output control flag (see section 6 above) I
Notes: The 4-node members are connected to NODE1 and NODE2 only with no stiffness contributions to
nodes NODE3 and NODE4. These latter two nodes NODE3 and NODE4 are purely for a geometric
definition (they are assumed to be connected to nodes NODE1 and NODE2 respectively by rigid links
with the flexible section of the member lying between the inner nodes). The boundary conditions for
the inner nodes NODE3 and NODE4 must be fully fixed.
In a large displacement analysis the coordinates of nodes NODE3 and NODE4 are updated from the
incremental displacements and rotations of nodes NODE1 and NODE2.
All four nodal points are required and must be distinct for the quadrilateral finite elements and the
masonry panel elements. The nodal points must be given in an anti-clockwise sequence around the
element.
16
9 Member Property Tables
PROPS
One set of input for every different cross-section type (this may be due to the different initial load conditions
etc.). These must be arranged in ascending section number order and each must be one of the following
section types;
N MTYPE LABEL
N Section Number I
MTYPE = FRAME => Beam or Beam-Column member A
= SPRING => Translational and Rotational Spring member
= WALL => Taylor Structural-Wall member
= DAMPER => Viscous Damping member
= TENDON => Active Tendon or Constant Force member
= CONTACT => Contact or Impact member
= QUAD => Quadrilateral Finite Element.
= PANEL => Masonry In-fill Panel
= GROUND => Foundation Element
= MULTISPRING => Multi-Spring Hinge Element.
LABEL Maximum 30 character label for section. If you wish to retain case or if the A
label contains blank characters enclose the label in single or double quotes.
The MTYPE may be supplied with just enough letters to be distinct , in this case just the first letter is sufficient.
The above line for each section shall be followed by the remaining data for the appropriate member types.
See the following sections.
17
11 FRAME type member properties
Frame Member
Order of all of the components required for Frame Members are:
18
Giberson One Component Beam Model
19
11a. Basic section properties
Note:
1. If ITYPE above is 6 or 7 then IHYST may not be 1, 5, 18 or 19, IPIN must be 0 and the bi-linear factor
RF in section 11c must have a value greater than or equal to 0.01.
2. If the member is linearly elastic then there is no effective distinction between members ITYPE 1 to 6.
3. To represent truss members either IPIN = 3 or I (in section 11b) is set equal to zero. The latter option is
preferred as then all beam hysteresis rules can be used for the truss behaviour. If the first option is used
then only the elasto-plastic and bi-linear rules may be used.
4. The value of IDAMG is used as a weight when computing the weighted damage index for the whole
structure. Only integer values are accepted. If all damage index flags are 1 then the weighted damage
index for the structure is the average damage index.
20
11b. Elastic section properties
Note: The weight/unit length is used to compute the member mass and does not contribute to the static loads
on the member or the structure.
The end-block lengths END1 and END2 are within the beam between nodes NODE3 and NODE4.
11c. Member Bi-linear Factors and Hinge Properties (only if IHYST above not equal 0)
RA RF H1 H2 H3 H4 H5 H6 H7
Notes:
1. For One-Component and Four-Hinge Beams and all Beam-Columns the plastic hinge lengths are reset
to 1.0 if 0.0 is specified making the plastic curvature the same as the plastic hinge rotation.
2. The plastic hinge lengths are not used for the Two-Component Beam members.
3. The plastic hinge lengths for Variable Flexibility Beams should be between 0.1 and 0.5 of the member's
clear span.
4. The Fractions of length to the interior plastic hinges of the Four-Hinge Beam member should be between
0.1 and 0.4, the default is 0.3.
5. The Central Stiffness ratio is the fraction of the end stiffnesses that is used by the central region between
hinges 3 and 4. The minimum is 0.1. All stiffness quantities, AE, EI and GA are adjusted by the same
scale factor.
21
11d. Material Specific Damping (only if ICTYPE = 5 in section 2)
This option allows different sections to have their own values of ALPHA and BETA for Rayleigh Damping. This
is useful if one is modelling a pounding study between two adjoining buildings where the buildings have different
damping characteristics or where different parts of the system have different damping characteristics, i.e. in soil-
structure interaction studies. Damping matrix C is given by
22
11e. Member Initial Fixed End Forces (only if ICOND = 1)
Note: The prestress force which is Tension positive affects only the internal member axial force used in the
determination of yield states and has no effect on the static analysis of the structure.
The Axial force applied to the member is positive if it acts in the direction NODE3 to NODE4 and one half
of this force is applied to each of the nodes at the ends of the member.
Positive member moments cause tensile stresses on the bottom fibres and positive shear forces act in the local
y direction on the positive local x face of the beam section. Tensile axial forces are positive. As an example if
a downward load is applied to the above member M1, M2 and V1 are negative while V2 is positive.
23
11f. BEAM Yield Conditions
PYT PYC MY1+ MY1- MY2+ MY2- MY3+ MY3- MY4+ MY4-
Note: Yield moments for hinges 3 and 4 are only required for the Four-hinge BEAM members. If both yield
moments at a hinge are zero then the hinge remains elastic. If both PYC and PYT are zero then axial
yield is suppressed. If both positive and negative yield moments at a hinge are zero then yield at that
hinge is suppressed.
There is no interaction between axial forces and the yield moments.
24
11g. Concrete BEAM-COLUMN Yield Surface at End 1 of member
Note: If M1B = M2B = 0.0 then a straight line is used between the points (PB, MB) and (0.0, MO) and not the
cubic curve.
The interaction is symmetric about the M = 0.0 axis.
25
11h. Steel BEAM-COLUMN Yield Surface at End 1 of member
26
11i. Quadratic BEAM-COLUMN Yield Surface at End 1 of member
27
11j. In-elastic Shear Parameters (Only if IGA on line 11a is 1)
One line for in-elastic shear for flexure about the local z-z axis.
Vy Vcr Vcc Alfa R Duct1 Duct2 Vres Phi1 Phi2 Pres Ido
Notes: The flexural ductility used for the strength degradation is based on the flexural ductilities at the hinges
at the ends of the member.
The action taken on the reaching of the shear yield strength follows the rules of Satyarno in his study on
the behaviour of shear-flexure interaction in framed structures.
The in-elastic shear follows the SINA hysteresis rule. Please see the Appendices.
Beta = Alfa and the rule assumes symmetry in both force and displacement axes.
If Vy, the shear area or the shear modulus are less than or equal to 0.0 then in-elastic shear is disabled
SINA Hysteresis
28
11k. In-elastic Flange Shear-Link Parameters (Only if IGA on line 11a is 2)
One line for in-elastic shear for flexure about the local z-z axis.
Flex Bear V1 V2 V3
Notes: The joint flexibilities specified at the ends of the member are complemented by the shear link flexibility
computed for the shear link. The shear force at each end of the member is used to adjust the joint
flexibility at that end of the member.
29
11l. Strength Degradation Parameters (only if ILOS is > 0) (See Appendix A)
For those rules which require sets of data such as SINA or Muto etc., then there are 2 sets of such data:
1: Flexure at End 1
2: Flexure at End 2
MUT MUC MU1+ MU1- MU2+ MU2- MU3+ MU3- MU4+ MU4- BETA1 BETA2
Notes: Ductility data for hinges 3 and 4 is to be provided only for the Four-hinge Beam members.
All ultimate ductilities must be greater than 1.0.
The ultimate ductilities have no effect on the dynamic analysis but are only used at the end of the
analysis to compute the damage indices.
See Appendix B for information on which hysteresis rules can allow damage indices to be computed.
30
12 SPRING type member properties
Spring Member
31
12a. Basic Section properties
ITYPE IHYST ILOS IDAMG KX KY GJ WGT RF RT PSX PSY PSZ THETA ITRUSS IOP
Notes:
1. The weight/unit length is used to compute the member mass and does not contribute to the static loads
on the member or the structure.
2. If the length of the member is zero then the local X and Y axes coincide with the global axes.
3. The ITYPE=2 model may be used to model the horizontal x and y behaviour of say a bridge pier where
the bridge is modelled in plan and the pier acts in flexure with the vertical axis in the z direction
perpendicular to the model of the structure. The behaviour of the pier represents the yield force
interaction in the x and y directions where the yield forces would be the yield moments about the y-y and
x-x axes at the base of the pier multiplied by the height of the pier.
4. The pre-load forces are internal to the member and have no contribution to the forces acting on the
structure. They affect only the initial member deformation before any structure deformations are
imposed.
32
5. This member may be used to represent truss members where the local X-direction represents the truss
action. The stiffness KX should equal the AE/L of the truss member.
6. If the Crisafulli hysteresis rule 33 is to be used to model in-fill panels then the AREAs asked for by the
rule, (see IHYST=33, Appendix B), are used to set up the stiffness of the spring at each time step.
However, the spring member does not use the length of the strut in forming the stiffness and so the
AREAs should be the strut cross-sectional areas divided by the strut length. Also, the stiffness KX is
used for the initial stiffness of the strut until the stiffness is checked to find its status in the hysteresis rule
behaviour, this is in the static analysis, where static loads are applied or at time = 0.0 if not. This means
that an initial estimate of KX should be made. In general the Crisafulli model does not carry significant
tensile forces which means that normally two strut members are required acting along opposite diagonals
and each would only act in the longitudinal, or X, direction.
7. IDAMG, m is used as a weight when adding member damage indices to obtain structure damage index,
see notes for Frame members , section 11a.
8. If ITYPE= 5 then the transverse yield force is a function of the axial force. The transverse hysteresis is
elasto-plastic representing the transverse friction. The other actions in the axial direction and rotation
are linearly elastic. There must be axial and transverse stiffness but the rotational stiffness may be zero.
This option allows different sections to have their own values of ALPHA and BETA for Rayleigh Damping. This
is useful if one is modelling a pounding study between two adjoining buildings where the buildings have different
damping characteristics or where different parts of the system have different damping characteristics, ie.e. in soil-
structure interaction studies. Damping matrix C is given by
33
12c. Yield Surface ( ITYPE = 1 ) (Not if IHYST = 0, 18, 19, 33 or 37)
Note: There is no interaction between torsion and spring yield and there is no interaction between the spring
yield in the rotation, x or y directions.
FX1 FY1 FX2 FY2 MZ+ MZ- RX1 RY1 RX2 RY2
Note: There is an elliptic interaction curve between x and y yield force in each quadrant of the diagram below.
The force-displacement curve in this case is tri-linear, it is modelled by two elasto-plastic members and
one elastic member in parallel. The yield forces are symmetric about the origin. There is no interaction
between yield in rotation and the x and y directions. The rotational component follows a bi-linear
hysteresis rule.
34
12e. Yield Surface ( ITYPE = 3 ) Flexure-Shear interaction. (Not if IHYST = 0, 18, 19, 33 or 37)
FX+ FX- FY+ FY- MZ+ MZ- FCRY FCCY ALFA HINGE DUCT1 DUCT2 RES DUCT3 I
Notes:
1. The SINA hysteresis rule governs the Shear (or Y action). The cracking forces in the Y direction are
symmetric about the origin. The strength loss in the Y direction is based on the absolute values of the
rotational component ductility.
2. The Shear Stiffness KY = G AS / L where G is the shear modulus, AS is the Shear Area and L is the
plastic hinge length. The Flexural Stiffness GJ = E I / L where E is the elastic modulus and I is the
second moment of area of the equivalent beam section. To get curvatures from the rotations divide the
rotations by the plastic hinge length L.
3. If I above is non-zero then shear strength degradation will also occur with shear ductility. A line with the
degradation data as for Appendix A of this manual must immediately follow this line. The normal strength
degradation controlled by the variable ILOS will apply to the x and rotational actions in the spring
member. The case of I above is 2 is for columns where retrofit is assumed to have been applied and
the sequence of shear failure is to be shown. When I is 3 is the case where retrofit is assume not to occur
and shear failure in a column implies failure of the structure.
4. These last two cases of I = 2 or 3 are really applicable to columns, 2 informs the user that shear yield has
occurred but implies a retrofit has been applied so one can see where the next yield will occur. Option
3 covers the case for columns without retrofit where failure of a column implies failure of the structure.
35
12f. Yield Surface ( ITYPE = 4 ) Axial Force - Yield Moment interaction (Not if IHYST = 0, 18, 19, 33 or 37)
36
12g. Yield Surface ( ITYPE = 5 ) Axial Force - Transverse Force interaction (Not if IHYST = 0, 18, 19, 33 or 37)
F0 MU IOP
Note: The axial force is the local x component action. The friction force acts in the local y direction.
12i. Stiffness Degradation Parameters (only if IHYST > 0) (See Appendix B.)
For those rules which require sets of data, such as SINA or Muto, then there are 2 sets of such data.
ITYPE=1
1: Longitudinal or X direction action. (The same parameters are used for rotational action)
2: Transverse or Y direction action.
ITYPE=3
1: Longitudinal or X direction action.
2: Rotation or Theta-Z action.
For the Widodo hysteresis (IHYST = 19) a third set of data is used for the rotational action.
37
13 STRUCTURAL-WALL type properties
Wall Member
Notes:
1. The weight/unit length is used to compute the member mass and does not contribute to the static loads
on the member or the structure.
2. Wall members usually require smaller time-steps than most other structures, i.e. 0.001 to 0.0001 seconds
compared with 0.01 seconds and Newton-Raphson iteration, say 3 cycles, should be provided. Changes
in the flexural stiffness also lead to large changes in the axial stiffness and the movement of the neutral
axis implies coupling between the axial and flexural actions as well.
3. If the member is in-elastic and Newton-Raphson iteration has not been enabled the variable MAXIT is
automatically reset to a maximum of 3 cycles of iteration and FTEST to 0.10. It has been found that a
smaller value of FTEST would be more appropriate.
4. The original Taylor element had a mid-length node with an hierarchical axial degree of freedom so that
38
the longitudinal strains due to the axial displacements had a quadratic variation with length as do the
longitudinal strains due to the beam-like cubic flexural displacement with length. It has been found that
the mid-length degree of freedom was never excited in the analyses and this seventh deformation degree
of freedom has been deleted.
5. At each section along the length the centre-line axial strain is computed, the curvature computed as a
function of the end rotations and assuming that the cross-section remains plane the longitudinal strain
in each segment of the section is computed. For each segment in the cross-section the concrete and
steel stresses and tangent modulii are computed and then the moment and axial force on the section are
calculated and the new effective cross-sectional area, second moment of area and the location of the
neutral axis is computed.
6. The section properties are integrated along the length using a Lobatto integration rule to get the member
stiffnesses. Three point Lobatto integration is identical to a three point Simpson’s rule.
7. Taylor sometimes used up to three elements to represent the wall over a single storey height.
8. IMAX < 1 is accounted for by intially cracking the section from both extreme segments inwards so that
the effective I is less than or equal to IMAX*IGROSS. AMAX is no longer used but as the flexural stiffness
reduces with the pre-cracking of the section the effective cross-sectional area also reduces.
9. NSECT = 2 is for an Giberson type beam-column model with an elastic central region with the plastic
hinge zones modelled by a Taylor-like segmental section. This allows for the neutral axis to move within
the hinge zone.
39
13b. Concrete and Steel Properties
Notes:
1. The concrete is assumed to carry no tensile stress once the material has cracked and the concrete is
assumed to carry no tensile stress once the compressive strain has exceeded EPSB.
2. IHYST = 0, The uncracked section properties are used for all time-steps.
3. IHYST = 1, The concrete elastic modulus is assumed to be a constant EMODC until the stress SIGA is
reached. The peak strain EPSA = SIGA/EMODC. The steel assumes a bi-linear stress-strain hysteresis.
This follows the original Taylor Element-A model
4. IHYST = 2, The concrete stress-strain law follows a quadratic rule until the stress SIGA is reached. The
peak strain EPSA = 2.0*(SIGA/EMODC). The steel hysteresis rule has an elasto-plastic first yield
excursion but subsequent cycles follow the Al-Bermani Bounding Surface rule with "=$=0.3 to model
the Bauschinger effect. This default value of ALFA may be overridden by specifying a different value
in the range 0.0 to 1.0 in the data above where ALFA = 1.0 implies a bi-linear hysteresis. The steel bi-
linear factor SBILIN, used to define the in-elastic bounding surface, takes a minimum of 0.0001 for the
Bounding Surface algorithm to work.
40
‘Kent and Park’ Concrete Stress-Strain Relationship
41
13c Material Specific Damping (only if ICTYPE = 5 in section 2)
This option allows different sections to have their own values of ALPHA and BETA for Rayleigh Damping. This
is useful if one is modelling a pounding study between two adjoining buildings where the buildings have different
damping characteristics or where different parts of the system have different damping characteristics, ie.e. in soil-
structure interaction studies. Damping matrix C is given by
42
13d. Input for Standard Wall Section (ICOND = 0)
Note: All distances D1, D2, D3 and D4 are measured from the left side of the wall when the section is viewed
down the member in the -X direction
The structural axis of the member is assumed to be at D/2
43
13e. Input for Numerical Wall Section (ICOND = 1)
One line is required for each of the sub-areas in the section. NIP lines required.
I Sub-area number I
X(I) Location of centre of sub-area I F
AS(I) Area of steel in sub-area I F
AC(I) Area of concrete in sub-area I F
44
14 DAMPER (dash-pot) type member properties
Dashpot Member
Notes:
1. C1, C2 and C3 are the dash-pot coefficients (force / unit velocity) for the different actions in the member.
These act in the positive direction only if ITYPE = 2.
2. These members are only included in the time-history analysis if IPANAL = 2 or more when their
contributions are added to the damping matrix of the structure.
4. Sign Conventions are the same as those for the SPRING member types, see section 12.
5. If the length of a DAMPER type member is zero then the local X and Y axes coincide with the global axes
unless a non-zero value has been specified for THETA above.
6. If the member deformation is within the range of the gap then the dashpot has no resisting force.
45
7. For Non-linear dash-pots all components use the same ALFA. The dash-pots show no hysteresis in their
Force-Velocity relationship, it is a non-linear elastic force-velocity relationship. The dash-pot force for
component i is related to the dash-pot velocity by the following equation
8. If ITYPE = 2 the dashpot has different properties in the positive and negative directions and the rules
for the data in the negative direction are the same as those for the positive direction components.
9. If IPLAS = 0, Elastic time-history analyses only, (see section 2, Principal analysis options), then the initial
gaps are reset to zero and ITYPE is reset to 0 after all the input data has been read.
46
15 TENDON type member properties
Tendon Member
Notes:
1. These members are only included in the time-history analysis IPANAL > 1 when they are considered in
the dynamic equilibrium equations.
2. Sign Convention: This is the same as for the SPRING members. Tensile forces and member elongations
are positive.
3. If the member is of zero length then the TENDON local member axis is assumed to coincide with the
global x axis unless a non-zero has been specified for THETA above. THETA is only used for zero length
members.
47
Or Active Tendon Force member
Notes:
1. These members are only included in the time-history analysis IPANAL > 1 when they are considered in
the dynamic equilibrium equations.
2. Sign Convention: Tensile forces and member elongations are positive.
3. If the member is of zero length then the local member axis is assumed to coincide with the global x axis
unless a non-zero value has been specified for THETA above. THETA is only used for zero length
members.
4. Care must be taken when setting out the magnitude and sign of the multiplier as an inept choice may
lead to very large displacements and loss of precision and failure of the analysis. The deformation and
force in the member is tension positive. The tendon member in tension pulls on both joints with equal and
opposite forces. It is worth noting that if the displacement of the structure has an equivalent circular
frequency T (radians/sec) then the magnitude of the velocities are of the order of T times the magnitude
of the displacements and the magnitude of the accelerations is T2 times the magnitude of the
displacements.
48
Or Semi-Active Tendon, Resettable Actuator, member
ITYPE = -1 I
STIFF Equivalent stiffness (Force per unit displacement) F
FPP Saturation Force (Same value used in tension and compression) F
DELAY Time Delay on Response (seconds). (Maximum of 92 time-steps) F
THETA Angle between Global X and Local x axes (Degrees) F
Notes:
1. These members are only included in the time-history analysis IPANAL > 1 when they are considered in
the dynamic equilibrium equations.
2. Sign Convention: Tensile forces and member elongations are positive.
3. If the member is of zero length then the local member axis is assumed to coincide with the global x axis
unless a non-zero value has been specified for THETA above. Theta is only used for zero length
members.
4. The force displacement relationship follows the Resettable Actuator Hysteresis rule IHYST=43, please
see Appendix B. The equivalent yield values are Fy+ = +FPP and Fy- = -FPP where FPP is the
saturation force.
5. The time-delay is internally converted to an integer number of time-steps
49
Or Variable Force member
ITYPE = -2 I
STIF Equivalent stiffness coefficient (Force per unit displacement) F
ALFA Power factor for element displacement (ALFA > 0.0) F
DAMP Equivalent damping coefficient (Force per unit velocity) F
BETA Power factor for element velocity (BETA > 0.0) F
Notes:
1. These members are only included in the time-history analysis IPANAL > 1 when they are considered in
the dynamic equilibrium equations.
2. Sign Convention: Tensile forces and member elongations are positive.
3. The force in the member is given by the following relationship. The displacement is the elongation of the
member and the velocity is the rate of change of member length.
4. This member is very useful where a damper or spring member is required where the power factors ALFA
and BETA are less than 1.0. In these cases the equivalent tangent stiffness or damping coefficients tend
to infinity when the displacement or the velocity tend to zero leading to numerical difficulties in the
analyses. In these limiting cases the forces are usually small and large stiffness or damping coefficients
are not a problem as the tendon member has neither stiffness or damping coefficients
50
16 CONTACT type member properties
Contact Member
16a. Basic section properties
Notes:
1. If the length of the member is zero then the local X and Y axes coincide with the global axes unless a
non-zero value has been specified for THETA above.
2. These members stiffness and damping matrices are only included in the time-history part of the analysis
when IPANAL = 2 or 3.
3. The Coefficient of restitution r (supplied as CX above) is related to the Percentage of Critical damping
8 by the following formula
4. The Effective End masses are only required if ITYPE is 1 or 2 and are used to generate the equivalent
fraction of critical damping in the member. The masses are usually taken as the mass in the frame at
the floor at that end of the member.
5. The transverse friction force is proportional to the transverse displacement since the onset of contact or
the reversal of transverse displacement and is limited to MU times the contact force.
6. The friction forces only exist if there is a non-zero contact force.
51
7. The Fractions of Critical Damping (for ITYPE = 1, or 2 using 8 in note 3) are related to the Damping
Coefficient c in the axial deformation mode of the member by the following expression where k is the
current member tangent stiffness and the masses m1 and m2 are the masses at the ends of the member
(When ITYPE is 0 the damping coefficient c is CX)
16c. Hysteresis data . See section 14 for the data required. This will complete the data for this member type.
52
17 QUADRILATERAL type member properties
53
17a. Basic section properties
Notes:
1. The weight/(unit volume) is used to compute the member mass and does not contribute to the static
loads on the member or the structure.
2. The element is a Hybrid Stress Type 2 quadrilateral finite element. All four nodes must be distinct (i.e.
no triangles). The stress distribution is assumed to have a cubic variation in both the x and y directions
and the element edge displacements are assumed to vary linearly along the element edge for the
tangential displacement and to have a cubic variation along the edge for the normal displacement. The
element actually assumes a linear variation of thickness between nodes but in the application within the
program a constant thickness is assumed. Three point Guass quadrature is used in each direction to
integrate the flexibility matrix and three point Guass quadrature is used for the integrals along the
element edges. Any reduction of the order of integration gives inaccurate integrals and a higher order
gives no change in the integrals.
3. The element is assumed, at present, to be linearly elastic.
This option allows different sections to have their own values of ALPHA and BETA for Rayleigh Damping. This
is useful if one is modelling a pounding study between two adjoining buildings where the buildings have different
damping characteristics or where different parts of the system have different damping characteristics, i.e. in soil-
structure interaction studies. Damping matrix C is given by
54
18 MASONRY PANEL ELEMENTS
These elements are designed to represent the behaviour of an unreinforced masonry panel infill in a reinforced
concrete frame. The model assumes that the infill is such that there are no gaps between the frame and the
panel. As generally constructed in South America the frame is often placed after the panel is erected (Crisafulli,
1997).
The panel is modelled with two parallel struts in each diagonal to represent the thrust carried across the panel
in diagonal compression. A fifth strut, which acts between the two opposite diagonal carriers on the diagonal
carrying compression, transfers the shear force between the top and bottom of the panel.
In general, the panel elements are rectangular but other geometries are permitted .
Four or Three lines of data required. An extra line will be required if the Material Specific Damping model is being
used.
1. Geometric Data
2. General Properties
3. Strut Data (2 lines or 1 line)
4. Material Specific Damping Data (if ICTYPE=5 only)
55
18a. Geometric Data
56
18b. Properties 18a
Notes:
The input data required for the application of the panel element includes the parameters described in for defining
the cyclic axial behaviour of masonry and the properties of the equivalent strut. In addition, the following variables
need to be defined in relation to the shear behaviour of the masonry panel:
Vertical separation between struts, hz: values of hz between z/3 and z/2 seems to lead to adequate results,
where z is the contact length between the panel and the frame.
Horizontal and vertical offset, xoi and yoi: these parameters define the horizontal and vertical distance,
respectively, measured from the external nodes to the internal nodes. This is intended to represent the reduction
of the dimensions of the panel due to the depth of the frame members.
Bond shear strength, and coefficient of friction, :: these parameters are usually obtained from direct shear
tests or following design specifications. It is recommended, however, the use of the reduced values resulting from
the modified shear theory proposed by Crisafulli. This theory takes into account the complex stress state in the
panel due to the composite nature of masonry.
Maximum shear stress, : this is the maximum shear stress permissible in the masonry panel, whose value
can be selected from the shear failure envelope of masonry. This is to avoid a large shear strength due to high
axial forces in the struts.
Shear stress factor, "s: this parameter defines the ratio of the maximum shear stress to the average stress in
the masonry panel. It normally varies from 1.40 to 1.65.
Shear stiffness factor, (s: this factor represents the fraction of the total stiffness assigned to the shear spring,
usually ranging from 0.5 to 0.75 but should not exceed 0.85 to 0.90. The remaining part, , is assigned to
the struts. The total stiffness of the panel element is controlled by the strut area and the elastic modulus.
Weight per unit volume: this is used to calculate the mass of the panel and is used only to form the structure
mass matrix. It does not contribute to loads on the structure.
57
18c. Either Strut Hysteresis (see also Appendix B IHYST=33)
or Elastic Properties
EMO AREA1
58
Notes:
FC The compressive strength f'm2 is the main parameter controlling the resistance of the strut. It must be
noted that FC does not represent the standard compressive stress of masonry but should be adopted
taking into account the inclination of the compressive principal stresses and the mode of failure expected
in the masonry panel. See Crisafulli 1997.
FT Tensile strength f't represents the tensile strength of the masonry or the bond strength of the panel-frame
interface, whichever is smallest. The consideration of the tensile strength has been introduced in the
model in order to gain generality. However, results obtained from different examples indicate that the
tensile strength, which is generally much smaller than the compressive strength, has no significant
influence on the overall response. Therefore, in the absence of more detailed information, the tensile
strength can be assumed to be zero.
UC The strain at maximum stress g'm usually varies between -0.002 and -0.005 and its main effect on the
overall response of the infilled frame is the modification of the secant stiffness of the ascending branch
of the stress-strain curve.
UUL The ultimate strain gu is used to control the descending branch of the stress-strain relationship. When
a large value is adopted for, example gu = 20 g'm , a smooth decrease of the compressive stress is
obtained.
UCL The closing strain gcl defines the limit strain at which the cracks partially close and compressive stresses
can be developed. Values of the closing strain ranging between 0 and 0.003 lead to results which agree
adequately with experimental data. If a large negative value is adopted, for example gcl = gu , this effect
is not considered in the analysis.
EMO The elastic modulus Emo represents the initial slope of the stress-strain curve and its value can exhibit
a large variation. Various expressions have been proposed for the evaluation of the elastic modulus of
masonry. It is worth noting, however, that these expressions usually define the secant modulus at a
stress level between 1/3 and 2/3 of the maximum compressive stress. In order to obtain an adequate
ascending branch of the strength envelope it is assumed that Emo $ 2 f'm2 / g'm .
GUN The unloading stiffness factor (un controls the slope of the unloading branch. It is assumed to be greater
than or equal to 1.0 and usually ranges from 1.5 to 2.5.
ARE The reloading strain factor "re defines the point where the reloading curves reach the strength envelope.
The calibration of the hysteretic model for the axial behaviour of masonry showed that good results are
obtained using values ranging between 0.2 and 0.4. However, higher values, for example 1.5, are
required to model adequately the cyclic response of the infilled frames. This is because other sources
of nonlinear behaviour, such as sliding shear, need to be indirectly considered in the response of the
masonry struts.
Four parameters are required to represent the cross-sectional area of the masonry strut. These are the initial area
Ams1 = AREA1 and final area Ams2 = AREA2 and the axial displacements at which the cross-sectional area
changes, )R1 = R1 and )R2 = R2. In a simplified model, it can be assumed that AREA1 and AREA2 are the
same using a low value of the strut area to avoid an excessive increase in the axial strength. In a more refined
analysis, a higher value of the initial area can be adopted, whereas the final area can be reduced by about 10%
to 30%. The displacement R1 and R2 can be estimated as g'm dm/5 and g'm dm (where dm is the length of the
masonry strut) respectively, at least until more precise information becomes available. Several empirical
expressions, which are described in section 6.2.1.3 of the reference, have been proposed for the evaluation of
the equivalent width of the masonry strut, whose value normally ranges from 0.1 to 0.25 of the diagonal length
of the infill panel.
IENV The descending branch of the stress-strain curve is usually modelled with a parabola instead of the curve
associated with Sargin's equation in order to obtain a better control of the response of the strut after the
maximum stress has been reached.
59
18e. Material Specific Damping (only if ICTYPE = 5 on section 2)
This option allows different sections to have their own values of ALPHA and BETA for Rayleigh Damping. This
is useful if one is modelling a pounding study between two adjoining buildings where the buildings have different
damping characteristics or where different parts of the system have different damping characteristics, ie.e. in soil-
structure interaction studies. Damping matrix C is given by
60
19 GROUND type member properties
These members can be regarded as a form of beam member and may be used to represent the flexibility of the
soil continuum beneath foundation beams and footings. The elements have a cubic variation of normal thrust
and a linear variation of axial shear along their lengths so their edge displacements are fully compatible with
beam members. Three possible representations of the soil are included:
Vlazov Foundation
General Two-Parameter Foundation (Pasternak model)
Vogt Foundation
The normal force-deflection relationship of both the Vlazov and the General Two-Parameter Foundation members
are governed by the following equation (Selvadurai,1979):
In the Vlazov soil model, the parameters k and G are derived from the elastic properties of the soil and specified
constraints to the variation of normal displacement of the soil with depth.
In the General Two-Parameter soil model, k and G are input directly by the user.
The Vogt foundation model was derived using Boussinesq’s formulas and has been used in the past for modelling
arch dams (Holand, 1968; Clough,1973). In two-dimensional analyses, the normal force-deflection relationship
for the Vogt model is similar to that of the Vlazov and General Two-Parameter soil models, but does not contain
a shear term. This model is analogous to a distributed Winkler spring.
The models can be regarded as ‘Pseudo-Dynamic’ foundation compliance models. They do not automatically
account for the frequency dependence of the dynamic stiffness of the soil. Such effects can be included by
appropriately scaling the input parameters to account for the fundamental period of the combined soil-structure
system.
The effects of radiation damping are incorporated into the Vlazov model by way of dashpots, placed in parallel
with the soil springs. Radiation damping is highly dependent on frequency and should only be included if the
fundamental period of the combined soil-structure system is below the fundamental period of the site.
Appropriate scale factors for radiation damping and the elastic stiffness of the soil, as a function of the
fundamental frequency of the soil-structure system, can be obtained from Wolf (Wolf, 1994)
61
Foundation members should be placed beneath foundation beams and footings and additional members should
be provided to model the soil surface adjacent to the structure being modelled. These members should extend
either side of the structure at least two to three times the width of the structure.
References:
Clough, Ray W et al,. ADAP – a computer program for static and dynamic analysis of arch dams. University of
California, Berkeley, California, 1973
Holand, Ivar and Aldstedt, Erik, Arch Dam Analysis by Finite Element Analysis, Institute for Structural Mechanics,
University of Trondheim, Norway, January 1968
Wolf, John P., Foundation Vibration Analysis Using Simple Physical Models. Prentice Hall , Englewood Cliffs,
NJ, 1994.
62
19a. General Data for Foundation Member
ITYPE ISHEAR IHYSTA IHYSTK IHYSTG ILOSA ILOSK ILOSG ISTF IRAD
Note:
IHYSTA, IHYSTK and IHYSTG can take on values of 0,1,2,3,4,8,10,11,12,13,14,20,21,23,26 and 28 see
Appendix B.
Notes:
A linear variation of displacement with depth is only appropriate for shallow soil layers
The model is formulated assuming that Plane Strain conditions apply.
63
19c. Data for Vlazov Soil Model. Only if ITYPE = 2
(Hyberbolic Variation of Displacement with Depth – deep soils)
19e. Data for Pasternak General Two-Parameter Soil Model. Only if ITYPE = 4.
AS KS GS WEFF
64
19g. Data for Non-linear Foundation Behaviour (only if IHYSTA, IHYSTK or IHYSTG > 0)
Note: PYTA, PYCA, PYT, PYC and TY are in units of force per unit area
Note: AMASS, KMASS and GMASS can take values between 0.0 and 1.0
Notes:
ADAMP, KDAMP and GDAMP can take values between 0.0 and 1.0
Radiation damping is highly frequency dependent and scale factors should be evaluated according to
fundamental frequency of the soil structure system.
65
19j. Longitudinal Strength Degradation Parameters (only if ILOSA > 0), see Appendix A.
19k. Longitudinal Stiffness Degradation Parameters (only if IHYSTA > 0), see Appendix B.
19l. Normal Strength Degradation Parameters (only if ILOSK > 0) see Appendix A.
19m. Normal Stiffness Degradation Parameters (only if IHYSTK > 0), see Appendix B.
19n. Shear Strength Degradation Parameters (only if ILOSG > 0), see Appendix A.
19o. Shear Stiffness Degradation Parameters (only if IHYSTG > 0), see Appendix B.
This option allows different sections to have their own values of ALPHA and BETA for Rayleigh Damping. This
is useful if one is modelling a pounding study between two adjoining buildings where the buildings have different
damping characteristics or where different parts of the system have different damping characteristics, i.e. in soil-
structure interaction studies. Damping matrix C is given by
66
20 MULTI-SPRING type member properties
Spring Typ 1
Spring Typ 2
Spring Typ 3
h1 h Spring Typ 4
2
h3
Node 2
Spring Typ 5
h4
h5 Node 3
Node 4 h
Node 1
Multi-Spring Element
If the Bi-Linear with Slackness hysteresis rule is used, then this element will become unstable if the whole
element goes into tension. This is because the element has no stiffness if all the multi-springs have positive
(tensile) displacements. To prevent this instability a compressive force should be imposed across the member
either by loads applied to the structure or by other prestressed members acting across the member.
The individual springs making up the multi-spring member have a sign convention where tensile deformations
and forces are positive.
67
20a. Control Parameters.
Note: The THETA term is only used for zero length members. In all other cases the local axial or longitudinal
axis is defined as being along the member length.
The IHYST in the notes above refers to the Appendices A and B covering the hysteresis models
NS ITYPE SA SV YT YC YP YN RA RV
Note: If IHYST > 0 and YP > 0.0 or YN < 0.0 then a Bi-linear Hysteresis is used for the Shear Spring.
20c. Longitudinal Strength Degradation Parameters (only if ILOS > 0), see Appendix A.
20d. Shear Strength Degradation Parameters (only if KLOS > 0) see Appendix A.
68
20e. Material Specific Damping (only if ICTYPE = 5 in section 2)
This option allows different sections to have their own values of ALPHA and BETA for Rayleigh Damping. This
is useful if one is modelling a pounding study between two adjoining buildings where the buildings have different
damping characteristics or where different parts of the system have different damping characteristics, i.e. in soil-
structure interaction studies. Damping matrix C is given by
69
21 Lumped Weights at the nodes
WEIGHTS INTERP
One input line is required for each nodal point. However, if any nodal points are omitted and INTERP above is
non-zero, the omitted nodes will be assumed to have the same weights as those of the nodal point on the
previous input line. The information for the last nodal point should be supplied. The nodal point numbers must
be provided in an ascending order. It must be noted that the nodal weights and that the self-weight computed
from the member weights are not added to the static load analysis.
N WX WY WM
Note: If only a static analysis is to be carried out, then the nodal weights may be omitted. Similarly the weight
may be provided by self-weight of the members (see the section tables in part 10 of this manual). Note
that these weights are converted to masses internally within the program.
70
22 External (static) nodal loads
LOADS
One input line is required for each nodal point. The information for the last nodal point should be supplied even
if it is not loaded. The nodal point numbers must be provided in an ascending order. It must be noted that the
nodal forces are independent of the nodal weights and that a self weight will not be computed from the nodal
masses and added to the static analysis.
N FX FY FM
71
23 Dynamic load factors (Only if IPVERT is less than 0).
One set of data if IPVERT = -1, two sets if IPVERT = -2, etc.
SHAPE ALPHA
One input line is required for each loaded nodal point. The last nodal point should be provided even if no loads
are applied to it or a line with node number 0 will terminate the input.
The load components will be multiplied by the loading, or excitation, history (see section 23). The first component
is multiplied by the first excitation history and the second component by the second excitation history.
This enables a maximum of three component loading, a constant static component and two time-varying
components
N FX FY FM
If there is more than a single loading pattern repeat all of the above SHAPE data for the second loading pattern
72
24. Earthquake Accelerograms, Displacement Time-Histories or Dynamic Load Time-Histories.
(Only for IPANAL = 2, 3, 4, 5, 8 or 9)
The following set of input lines is required for each earthquake or dynamic loading component.
There is 1 component if IPVERT = -1, 0 or 1
There are n components if IPVERT = -n or n
For Earthquake Accelerograms the components act in the directions specified in the Earthquake Transformation
matrix supplied in section 2b of this manual.
For Dynamic loading patterns each component multiplies each of the load vectors described under the SHAPE
data sets.
For Input Displacement Histories (IPANAL = 8 or 9) the excitations act at the degrees of freedom specified in
the Nodal Point Boundary Condition data. The displacement must be scaled, via the divisor ASCALE to have
the displacement units chosen for the structure.
EQUAKE FILE
73
24b. Control parameters for Accelerogram, Displacement Time-History or Loading Time-History
74
The rest of the input for the accelerogram (sections 24c and 24d) is in the Accelerogram File, the Displacement
History File or the Force History File specified by the user in the initial file name prompts at the beginning of the
program execution or on the line with the word EQUAKE. The default response is that the accelerograms, or
excitation is appended to the above data file.
24c. Accelerogram, Displacement History or Load-History flag. This line is the start of the accelerogram etc. for
all but the PEER or SAC format records.
The PEER format records are preceded by 4 lines of header information and this START line is not required.
The SAC format records are preceded by 2 lines of header information and this START line is not required
One input line with the word STAR, START or DATA: (the colon is mandatory) starting in column 1 and the word
must be in upper case. This line may be preceded by as many header lines as desired.
START
The remainder of the input is the Acceleration, Displacement History or Load-history record itself. The record is
in the form of a series of lines each of which starts with a Line Sequence Number (which must be in an
ascending order) followed by either (i) a group of 4 or 1 successive time-acceleration points (BERG or FREE
or EXCEL Formats), or (ii) a sequence of 10, 8, 5 or 6 uniformly time spaced acceleration values at DELTAT time
intervals apart, the (CALTECH, NCEER or CSMIP Formats).
Note that the NCEER, CSMIP, EXCEL or PEER records do not have a sequence number. The analysis
acceleration record, or forcing function, will begin at the first time on or implied by the beginning of the
accelerogram line ISTART and there must be sufficient lines remaining to span the analysis time-history length
TR.
If the input is a force history IPVERT < 0 then the so-called accelerations are the actual force multiplier values
which will multiply the input load shape patterns (see section 24 above).
If the input are Displacement Histories (IPANAL = 8) these histories are applied to the degrees of freedom
specified in the Nodal Point Boundary Conditions.
The record must be on one of the following formats depending on the value of IBERG on section 25b. The
FORTRAN format is provided in parentheses for each case.
The I3 format implies 3 characters for the number which is right justified, 6X implies 6 blank characters. The F8.4
implies 8 characters for the number and if omitted the decimal point is located in front of the 4th to last character
in the 8 character set which is assume to be right justified, if the decimal point is provided the number may be
located anywhere in the 8 character space. A number preceding the format, i.e. 10F8.4 means that there are
10 numbers each having an 8 character field. Each format starts on a new line.
75
(1) BERG FORMAT (I3,4(F8.4,F9.6)) (Default filename extension is .eqb)
ISEQ T1 G1 T2 G2 T3 G3 T4 G4
If the line sequence number is greater than 999 it is not read or checked by the program.
(2) CALTECH FORMAT (I4,6X,10F6.0) or more precisely (I4,6X,10I6)(Default filename extension is .eqc)
ISEQ G1 G2 G3 G4 G5 G6 G7 G8 G9 G10
If the line sequence number is greater than 9999 it is not read or checked by the program.
G1 G2 G3 G4 G5 G6 G7 G8 G9 G10
ISEQ T1 G1
The three items may be placed anywhere on the line and separated by at least one blank column or a comma.
The lines must be in consecutive order with ISEQ starting at 1 and increasing line by line. This format is
particularly useful where the excitation record has been generated on a spreadsheet. It is, however, not as
compact a format as the NCEER format.
76
(5) CSMIP FORMAT (8F10.3) (Default filename extension is .eqs)
G1 G2 G3 G4 G5 G6 G7 G8
T1 G1
The two items may be placed anywhere on the line and separated by at least one blank column or a comma.
The times MUST be in ascending order but do not have to be at equal intervals. This format is particularly useful
where the excitation record has been generated on a spreadsheet. It is, however, not as compact a format as
the NCEER format.
G1 G2 G3 G4 G5
Note DELTAT is usually 0.004 seconds. The records are usually in the units of acceleration of gravity.
G1 G2 G3 G4 G5 G6
Note DELTAT is usually 0.020 seconds. The records are usually in the units of acceleration of gravity.
77
25. Cyclic Adaptive Push-Over Displacement Time-History. (Only if IPANAL = 7)
The following set of input lines is required for the required displacement history. This displacement history acts
as a control on the displacement at the node and direction specified on line 25b.
EQUAKE FILE
78
The rest of the input for the accelerogram (sections 25c and 25d) is in the Displacement-History File the name
of which is specified by the user in the initial file name prompts at the beginning of the program execution. The
default response is that the accelerograms, or excitation is appended to the above data file.
One input line with the word STAR, START or DATA: (the colon is mandatory) starting in column 1 and the word
must be in upper case. This line may be preceded by as many header lines as desired.
START
The remainder of the input is the Displacement History record itself. The record is in the form of a series of lines
each of which contains a sequence of displacements at which the Adaptive Load Sequence reverses. Each of
these values is read under the FORTRAN Free Format so that the numbers are separated by one or more blanks
or a comma. All NSTEP points MUST be supplied. The initial 0.0 is implied and should not be supplied. The
sequence of displacement values should have alternating signs. Use as many lines as are necessary to span
all NSTEP points.
79
26. Acceleration Response Spectra. (Only if IPANAL =10)
The following set of input lines is required for the acceleration response spectra. These spectral accelerations
are assumed to apply to the base of the structure and will apply to all earthquake excitation directions as
specified by the parameter IPVERT in line 2 of the data set.
SPECTRA FILE
Notes: The Acceleration Response Spectra are supplied in the units of Acceleration of Gravity (g). The
response spectra are then multiplied by the gravitational constant GRAV supplied in line 3 of the data
set.
The spectra are supplied in the following section with one line required for each natural period of free-
vibration. The first line must be for the zero natural period of free-vibration. The total number of lines
is NSTEP.
In the response spectra calculations if the natural period of a mode is greater than the value of the
NSTEPth natural period the spectral acceleration is assumed to remain constant for all greater natural
periods.
80
The rest of the input for the response spectra (sections 26c and 26d) is in the Response Spectra File the name
of which is specified by the user in the first excitation file name prompt at the beginning of the program execution.
The default response is that the spectra is appended to the data file following lines 26a and 26b above.
One input line with the word STAR, START or DATA: (the colon is mandatory) starting in column 1 and the word
must be in upper case. This line may be preceded by as many header lines as desired.
START
The remainder of the input is the Response Spectra.. The record is in the form of a series of lines each of which
contains the natural period of free-vibration and the necessary spectral values. Each of these values is read
under the FORTRAN Free Format so that the numbers are separated by one or more blanks or a comma. All
NSTEP lines MUST be supplied. The number of spectral accelerations required on each line depends on the
values of IPVERT in line 2 of the data set and on the value of ISAME in line 26b above. If IPVERT = 0 or 1, or
ISAME = 0 then there must be only one spectral acceleration on each line. If IPVERT > 1 and ISAME = 1then
there must be IPVERT spectral accelerations on each line. Use as many lines as are necessary to span all
NSTEP points.
End
81
This page is intentionally left blank.
82
Department of Civil Engineering COMPUTER PROGRAM LIBRARY
Author:– Date:–
Athol J Carr September 25, 2004
DYNAPLOT
Purpose
This program is designed to plot the time history results or the hysteresis loops computer by the non-linear
dynamic analysis programs RUAUMOKO-2D and RUAUMOKO-3D. The output is in the form of time-history
plots of nodal displacements or forces or member actions or deformations or combinations of the above
quantities. Alternatively hysteresis loops may be plotted for members or nodes.
Method of Analysis
The program generates each graph, one by one, requesting the control information for each graph then
scanning the Post-processor file generated by RUAUMOKO-2D or RUAUMOKO-3D to select the required
function to plot. The program can only plot the results from one non-linear analysis at a time.
All data for DYNAPLOT is prompted for by the program using the TinyCLIP command processor. For help on
the input data type $HELP at any prompt. To get help on the use of the DYNAPLOT programme type HELP
or ? At any prompt.
The number of nodes and members with data stores in the .RES file is recorded at the beginning of the file
together with the time-step between stored items are recorded as the first item in the .RES file. The structure
geometry and mode shapes are also saved at the beginning of the file.
Running the post-processor program DYNAPLOT.
1
To run the program call the program by the method appropriate to your operating system. On a personal
computer just type DYNAPLOT assuming that the files DYNAPLOT.EXE and FQWIN.HLP are in your current
directory or path.
In Microsoft Windows operating systems another option is to create a shortcut on the desktop and for this
purpose a suitable icon for DYNAPLOT, Dynaplot.ico, is supplied with the program.
The program prompts for responses to a series of questions. Default responses, where appropriate, are
enclosed in square brackets, [ ]. File names must match the conventions of your operating system but file
names, with paths where necessary, must not exceed 60 characters in length and must not contain blanks.
The first question asks for the name of the output file. The default is the computer console or terminal screen.
This is normally the response for most users, however if an output file is created then the envelope values of
each graph line drawn is output to that file.
The second question asks for the name of the DYNAPLOT post-processor file, do not supply the extension
part of the name. This file has been given the extension .RES or .RAS by RUAUMOKO-2D or RUAUMOKO-
3D and DYNAPLOT will attempt to open a file with the extension .RES but if that does not exist will attempt to
open the file with the extension .RAS. If a file with the given name and path with either extension cannot be
found the program will again prompt for the file name, and path. Once the file has been found DYNAPLOT
will scan the file to determine whether it was created by RUAUMOKO-2D or RUAUMOKO-3D. The program
DYNAPLOT will also scan the file to find the duration of the data stored in the file and if this is less that the
duration TIME supplied in the RUAUMOKO data the user is informed of the duration that is available for
plotting. The program will then ask if colour graphics is wanted. The program then prompt following the
normal options in DYNAPLOT. Please see the manual pages for DYNAPLOT.
In Microsoft Windows operating systems to get hard copies of the graphs use the pull down ‘file’ menu and
select the Print or Save options to send the graph to the printer or to save the plot as a bitmap file (.BMP).
On unix systems using GKS graphics select the Hard-copy option from the Choice window
2
Output from RUAUMOKO-2D post-processor file
The data that can be plotted consists of the data stored in the RUAUMOKO-2D post-processor file for which
the default name is RUAUMOKO.RES or RUAUMOKO.RAS. The content of the file is controlled by the
parameter IOUT in the program RUAUMOKO. At every time-step stored in the post processor file the
following data is stored.
b. For each node the following items are stored: (controlled by IOUT in RUAUMOKO-2D)
(18 items)
1 = X and Y displacements, Z rotation
2 = X and Y velocities, Z rotational velocity
3 = X and Y accelerations, Z angular acceleration
4 = X and Y applied forces, Z applied moment.
5 = X and Y damping forces, Z damping moment
6 = X and Y inertia forces, Z inertia moment
c. For each member the following items are stored: (controlled by IOUT in RUAUMOKO-2D)
3
For Wall members (13 items)
1 = Axial force and
2 = Bending Moments at End 1 and End 2
3 = Axial strain
4 = Curvatures at End 1 and End 2
5 = % axial stiffness
6 = % flexural stiffness at End 1 and End 2
7 = Shear Forces at End 1 and End 2
8 = Neutral axis eccentricities at End 1 and End 2
4
Output from RUAUMOKO-3D post-processor file
The data that can be plotted consists of the data stored in the RUAUMOKO-3D post-processor file for which
the default name is RUAUMOKO.RES or RUAUMOKO.RAS. The content of the file is controlled by the
parameter IOUT in the program RUAUMOKO-3D. At every time-step stored in the post processor file the
following data is stored.
b. For each node the following items are stored: (controlled by IOUT in RUAUMOKO-3D)
(36 items)
1 = X, Y and Z displacements, X, Y and Z rotations
2 = X, Y and Z velocities, X, Y and Z rotational velocities
3 = X, Y and Z accelerations, X, Y and Z angular accelerations
4 = X, Y and Z applied forces, X, Y and Z applied moments.
5 = X, Y and Z damping forces, X, Y and Z damping moments
6 = X, Y and Z inertia forces, X, Y and Z inertia moments
c. For each member the following items are stored: (controlled by IOUT in RUAUMOKO-3D)
Note: Groups 8 and 9 are for the In-elastic Beam-Shear option only
Groups 10 and 11 are for the In-elastic Shear-Link model only.
5
For Damper members (18 items)
1 = x, y and z forces, x, y and z moments
2 = x, y and z velocities, x, y and z rotational velocities
3 = x, y and z % stiffnesses, x, y and z % rotational stiffnesses
For Hybrid Quadrilateral finite elements, Forces and Moments per unit length.(24 items)
1 = N-xx, N-yy, N-xy, M-xx, M-yy, Mxy at node I
2 = N-xx, N-yy, N-xy, M-xx, M-yy, Mxy at node J
3 = N-xx, N-yy, N-xy, M-xx, M-yy, Mxy at node K
4 = N-xx, N-yy, N-xy, M-xx, M-yy, Mxy at node L
6
Control data for DYNAPLOT
c. The name of the output file, the default being the output screen. If you wish the results to be
output to a file then give the name (with path if necessary).
d. The name of the Post-processor file output by RUAUMOKO-2D or RUAUMOKO-3D, the
default name of which is RUAUMOKO. The file must have the extension .RES or .RAS.
e. If you have a colour monitor you will be prompted to ask if you wish to use colour. If you are
also going to hard-copy the plots and you have only a mono-chromatic printer then answer N
or NO as all lines would appear as solid black lines. In monochromatic systems different line
styles are used for each line.
f. What sort of graph do you want? Minimum of the first character required.)
The time ranges used in the plots must not exceed the time range stored in the post-processor file. If the
range is exceeded then a message like “Ran out of RUAUMOKO results” will appear on the screen.
Graph titles which are to contain lower case letters or consist of more than one word must be enclosed in
single or double quotes and in any case must not contain more than 32 characters, including blanks.
Note: Plots for which the horizontal axis is time will have the text “Time (Seconds)” appended to the title
supplied by the user for the horizontal axis.
7
Time-History Plots
The example below plots two lines of nodal data with a time interval of 1.0 with the first and last grid lines
being 1 and 10 and IYN and IYP being -1 and 1 respectively.
8
Hysteresis Plots
The example below plots the hysteresis curve for moment versus curvature. Note that although the member
is bi-linear the results file has only stored data at every 5 time-steps and hence some loops appear to have
had the corners cut off
9
Combination Time-History Plots
This option enables combinations of node and/or member results to be combined into one plot line. Other
aspects are identical to that of the Time-History plots. The maximum number of contributions to any one plot
line is 50.
In the example below the Base Shear is obtained by summing the shears in the two ground floor columns.
10
Base Shear Combination Hysteresis Plots
This option enables combinations of node and/or member results to be combined into one plot line and to be
plotted in the form of a Hysteresis plot by plotting the combined results against a second set of combined
results. An example would be the plot of base-shear (a combined input) versus the interstorey drift (a second
combined input). In the example below the Base Shear is plotted against the top floor displacement. The
maximum number of contributions to each axis value is 50.
11
Work-Energy History Plots
This option enables the Time-history of work or energies to be plotted. An option is to plot rate-of-change of
work or “power” to be plotted.
The example below plots four lines, the bottom line is the kinetic energy, the second line is the kinetic energy
and damping work done, the third line is the sum of the kinetic energy, the damping work done and the strain
energy. The fourth line is the External Applied Work. The vertical grid spacing is 1.0 seconds and the first and
last grids being 0 and 10 respectively. The vertical spacing was the default 0.0 and IYN and IYP are 0 and 1
respectively.
If one selects the Default option then the following energy plot is produced without the necessity of specifying
all the data items above.
12
Snap-shots of Deflected Shapes
This option enables the displaced shape of the structure to be plotted at various time-steps.
The example below plots two lines, the solid line is the deflected shape at 2.0 seconds, the dashed line is the
deflected shape at 2.5 seconds. The data was stored for nodes 1, 3, 5, 7 and 9.
13
Movie-like Presentation of Deflected Shapes and Plasticity
The option presents the movie-like displays of deflected shapes and plasticity. This show the same deformed
mesh pictures that are able to be displayed in RUAUMOKO as it carries out the time-history analysis.
However, in DYNAPLOT they may be displayed picture by picture so that hard copies of the plots may be
produced.
If continuous effects are selected then no further action is required until the sequence finishes except that on
some systems a key-press will stop the sequence and a second key-press will then terminate the plot. For
stepped effects press the Enter key to proceed to the next picture. In the later case the sequence must
proceed to the end of the sequence to terminate. The user is then asked if a further sequence is to be shown.
This feature is only available if all nodes and all members have had their results saved to the .RES post-
processor file.
The option presents the movie-like displays of the modes shapes of free vibration. This is similar to the
displays in RUAUMOKO except that here if the number of cycles is S (upper or lower case) rather than the
number of cycles then the mode shape is displayed only in its maximum plotted position so that a hard copy
may be produced.
This feature is only available if all nodes and all members have had their results saved to the .RES post-
processor file.
This produces the plot of the mesh as is shown in Ruaumoko. However, in this case you are asked for the
range of members, or elements that are to be shown in the plot. The default range is for all elements. For
three-dimensional structures, i.e. the data has come from Ruaumoko3D, you are also asked if you wish to
change that plot view axes.
This feature is only available if all nodes and all members have had their results saved to the .RES post-
processor file.
This produces the plot of thenodal points or joints in the mesh. As in the case for plotting the element mesh,
case you are asked for the range of nodes or joints that are to be shown in the plot. The default range is for
all nodes.
This feature is only available if all nodes and all members have had their results saved to the .RES post-
processor file.
This option enables the user to redraw the last graph plotted. This is normally used where the Enter key has
been pressed before the user obtained a hard-copy. This enables the graph to be redrawn without going
through the selection of the data items again.
14
Keeping the Last Graph
This enables to user to save the graph data into an ASCII file so that it may be retrieved into the program
COMBPLOT or into a spreadsheet etc. for combining with other graphical information. For Hysteresis data
each output line saves the coordinates of the points on the plot. For Time-history plots the first item on the
line for each point is the time and this is followed by the values of each line at that time. The last two items
saved are the x and y titles for the graph. There are several choices for the output format, one fitting an 80
column line another a 132 column line as well as formats suitable for spreadsheet input such as for Microsoft
Excel.
There is an option for a single line Time-history graph and that is to save the file as a Caltech Format
accelerogram. This is particularly useful if the graph is for the total acceleration for node in the structure and if
one wishes to analyze a substructure located at that point in the structure or if one wants a Response Spectra
for the motion at that point. The resulting file may be input to RUAUMOKO or SPECTRA as a Caltech format
accelerogram.
If one selects the OPEN option from the main menu, DYNAPLOT restarts and prompts for the name of a new
post-processor (.RES or .RAS) file.
15
This page is intentionally left blank.
16
Department of Civil Engineering COMPUTER PROGRAM LIBRARY
Author:– Date:–
Athol J Carr September 25, 2004
APPENDICES
for programs
1
APPENDIX A
Notes:
1. ILOS, the parameter that controls the strength degradation (see Properties tables)
ILOS = 0, No Strength Degradation.
ILOS = 1, Strength loss in each direction is a function of the ductility in that direction.
ILOS = 2, Strength loss in each direction is a function of the number of inelastic cycles.
ILOS = 3, Strength loss in each direction is a function of the maximum ductility.
ILOS = 4, As for ILOS = 1 above but strength loss is also proportional to the number of inelastic
cycles.
ILOS = 5, As for ILOS = 4 above but strength loss is also proportional to the number of inelastic cycles
and the strength due to ductility for ductilities greater than DUCT2 remains at the level of
RDUCT until the ductility reaches DUCT3 when the strength suddenly reduces to 1% of the
original strength.
ILOS = 6, As for ILOS = 3 above but strength loss is also proportional to the number of inelastic
cycles.
ILOS = 7 As for ILOS = 6 above but strength loss is also proportional to the number of inelastic cycles
and the strength due to ductility for ductilities greater than DUCT2 remains at the level of
RDUCT until the ductility reaches DUCT3 when the strength suddenly reduces to 1% of the
original strength.
2. If Strength Loss is based on cycle number rather than the ductility then DUCT1 is the cycle number that
the strength starts to reduce and DUCT2 is the cycle number at which the strength reaches the residual
value. It must be noted that the cycle number is computed as the number of times the hysteresis rule
leaves the post-yield back-bone or skeleton curve divided by 2 and this might be greater than the
number of cycles of hysteresis particularly if there has been a one sided ratchet-like behaviour of the
hysteresis. The minimum value permitted for RDUCT is 0.01. If the strength was to reduce to 0.0,
Ruaumoko would then take the member behaviour as elastic which would not be the intention of the
user.
3. If a number is provided for the variable DUCT3 above then the strength decreases linearly from RDUCT
times the initial strength at DUCT2 to 0.01 of the initial strength at ductility (cycle number) DUCT3. If
this number is omitted then the strength remains constant after DUCT2 is reached.
4. See Appendix B for information on which Hysteresis rules are able to accept strength degradation.
5. If ILOS is greater than 0 then as the strength is reduced the stiffness is reduced to match. This means
that the yield displacement, rotation or curvature remains constant as the strength decreases making
the definition of member ductility consistent. If the hysteresis loop being used has other strength
parameters such as a cracking force or moment then these are also reduced proportionally to the yield
strength. If this is not done then some of the hysteresis loops may be impossible to follow where the
yield strength would become less than the cracking strength.
6. If ILOS is supplied as a negative number, i.e. -5, the strength degradation rule would follow that for
ILOS=5 but the stiffness would not be reduced and other hysteresis rule actions would also not be
reduced. This means that the definition of ductility would be difficult to follow as the yield displacement.
rotation or curvature, which is the denominator in the expression for ductility, would decrease as the
strength decreases. Care would also be necessary insetting the levels of strength degradation to
ensure that the hysteresis loop does not become impossible to follow. See the note below.
2
Strength Reduction Variation
In earlier versions of Ruaumoko when the strength degraded the stiffness remained using its input values.
This causes problems with the definition of ductility in that as the yield force reduces and the stiffness
remains constant the yield displacement reduces and therefore for a given member deformation the apparent
ductility increases. This has shown up in that the residual strength, or the 1% strength, is reached at much
smaller displacements, or curvatures, than the user had expected. The program has now been modified such
that as the yield forces, or moments, degrade the stiffness also degrades. This means that the yield
displacements remain constant and the definitions of ductilities remain more consistant. As some hysteresis
rules have other force, or moment quantities such as cracking forces, or intercept forces (see Appendix B),
which can also cause difficulties when the yield strength degrades, such that the yield strength may reduce to
a smaller level than say the cracking moment leading to confusion within the hysteresis rule, such force
quantities are now also degraded as the yield strength degrades. This is more likely to be realistic than the
earlier operation of the strength degradation in that for most member sections the yield point is defined by the
extreme fibre yield strain and given the section properties the yield strain, or curvature, is more likely to
remain constant than is the yield force or moment. There are still some difficulties when there are different
degradations in each direction, ILOS =1, 4 or 5 as the stiffness will vary depending whether the member
displacement is positive or negative.
3
APPENDIX B
0 = Elastic (default)
1 = Elasto-plastic
2 = Bi-linear
3 = Ramberg-Osgood
4 = Modified Takeda Degrading Stiffness
5 = Bi-linear with Slackness
6 = Kivell Degrading Stiffness
7 = Origin Centered Bi-linear Hysteresis
8 = SINA Degrading Stiffness
9 = Stewart Degrading Stiffness with Slackness
10 = Bi-linear Degrading Stiffness
11 = Clough Degrading Stiffness
12 = Q-HYST Degrading Stiffness
13 = Muto Tri-linear Degrading Stiffness
14 = Fukada Tri-linear Degrading Stiffness
15 = Bi-linear Elastic
16 = Non-linear Elastic (Un-Reinforced Masonry)
17 = Degrading Elastic
18 = Ring-Spring Isolator or Damper
19 = Hertzian Contact Non-linear Spring
20 = Mehran Keshavarzian's Degrading Stiffness
21 = Widodo Foundation Compliance
22 = Li Xinrong Reinforced Concrete Column Degrading Stiffness
23 = Bouc Degrading Stiffness
24 = Remennikov Out-of-plane Buckling Steel Brace
25 = Takeda with Slip Degrading Stiffness
26 = Al-Bermani Bounding Surface Hysteresis
27 = Peak Oriented Hysteresis
28 = Matsushima Strength Decay model
29 = Kato Shear model
30 = Elastomeric Damper Spring
31 = Composite Section, modified SINA model
32 = Different Stiffness in Positive and Negative directions. Modified Bi-linear rule
33 = Masonry Strut Hysteresis
34 = Hyperbolic Hysteresis
35 = Degrading Bi-linear with Gap Hysteresis
36 = Bi-linear with Differing Positive and Negative Stiffness Hysteresis
37 = Non-linear Elastic Power Rule Hysteresis
38 = Revised Origin Centred Hysteresis
39 = Dodd-Restrepo Steel Hysteresis
40 = Bounded Ramberg-Osgood Hysteresis
41 = Modified (Payne) Ramberg-Osgood Hysteresis
42 = HERA-SHJ Steel Hinge unit.
43 = Resetting Origin Loop
44 = Pampanin Reinforced Concrete Hinge hysteresis
45 = Degrading Stiffness Ramberg-Osgood Hysteresis
46 = Dean Saunders Reinforced Concrete Column
47 = Multi-linear Elastic
48 = Isotropic/Kinematic Strain Hardening Bi-Linear
49 = Isotropic/Kinematic Strain-Hardening Ramberg-Osgood
50 = Flag-shaped Bi-linear Hysteresis
4
Use of Hysteresis Rules for Frame members in RUAUMOKO-2Dand RUAUMOKO-3D
43 Resetting Origin No No No No No No
5
Use of Hysteresis Rules for Frame members in RUAUMOKO-2Dand RUAUMOKO-3D
48 Isotropic Strain Hard. Bi-linear Yes Yes Yes Yes Yes Yes
49 Isotropic Strain Hard. Ramberg Yes Yes Yes Yes Yes Yes
6
Use of Hysteresis Rules for Spring members in RUAUMOKO-2D
42 HERA-SHJ Yes
7
Use of Hysteresis Rules for Spring members in RUAUMOKO-2D
8
Use of Hysteresis Rules for Spring members in RUAUMOKO-3D
42 HERA-SHJ No
9
Use of Hysteresis Rules for Spring members in RUAUMOKO-3D
10
Use of Hysteresis Rules for Foundation members in RUAUMOKO-2D and RUAUMOKO-3D
42 HERA-SHJ No
43 Resetting Origin No
11
Use of Hysteresis Rules for Foundation members in RUAUMOKO-2D and RUAUMOKO-3D
12
Use of Strength Degradation and Damage Indices in RUAUMOKO-2D and RUAUMOKO-3D
0 Elastic No No
1 Elasto-Plastic Yes Yes
2 Bi-linear Yes Yes
3 Ramberg-Osgood Yes Yes
4 Takeda Yes Yes
5 Bi-linear - Slackness Yes Yes
6 Kivell Yes Yes
7 Origin-Centered Yes Yes
8 SINA Yes Yes
9 Stewart Yes Yes
10 Degrading Bi-linear Yes Yes
11 Clough Yes Yes
12 Q-HYST Yes Yes
13 Muto Yes Yes
14 Fukada Yes Yes
15 Bi-linear Elastic No No
16 Non-Linear Elastic No No
17 Degrading Elastic No No
18 Ring-Spring No No
19 Hertzian Contact No No
20 Keshavarzian Yes Yes
21 Widodo Foundation No Yes
22 Li-Xinrong Column No Yes
23 Bouc No* Yes
24 Remennikov No Yes
25 Takeda with slip Yes Yes
26 Al-Bermani Bound-Surface Yes Yes
27 Peak Oriented Yes Yes
28 Matsushima Degrading No* Yes
29 Kato Degrading Shear No Yes
30 Elastomeric Spring No No
31 Composite Section Yes Yes
32 Different +/- Stiffness Yes Yes
33 Masonry Strut Hysteresis No Yes
34 Hyperbolic Hysteresis Yes Yes
35 Degrading Bi-linear Hysteresis No No
36 Bi-linear Differing +/- Stiffness Yes Yes
37 Non-linear Elastic Power No No
38 Revised Origin Centred Yes Yes
39 Dodd-Restrepo Steel No Yes
42 HERA-SHJ Yes No
43 Resetting Origin No No
13
Use of Strength Degradation and Damage Indices in RUAUMOKO-2D and RUAUMOKO-3D
44 Pampanin No No
45 Degrading Ramberg-Osgood Yes Yes
46 Dean Saunders Concrete Column Yes Yes
47 Multi-linear Elastic Yes No
48 Isotropic Strain Hard. Bi-linear No Yes
49 Isotropic Strain Hard. Ramberg No Yes
50 Flag-shaped Bi-linear Yes Yes
14
Notes on notation for hysteresis rules:
In all of the diagrams associated with the hysteresis rules the following notation is used.
If a hysteresis rule is selected and the rule is not allowed for that member then an error message is
printed in the output for the section properties and the analysis will be terminated.
Yes* implies that the hysteresis rule is now allowed for column members. However, the effects on the
small-cycle hysteresis loops of the yield moments varying with the changes in the axial force in the
member have not been studied.
If both Strength Degradation and Damage Indices are selected then the effects of Strength Degradation
on the computed Damage Indices is uncertain and a warning is printed after reading the member
properties.
If Strength Reduction or Damage Indices are not allowed for the specified Hysteresis Rule and they are
specified in the input data, the data is read and then the control parameters ILOS and/or IDAMG are
reset to zero. Warnings of these re-settings are printed in the output.
No*. The Bouc and Matsushima Hysteresis rules have their own strength degradation capability.
Damage indices for the masonry strut hysteresis only outputs the hysteretic work done.
For the Ruaumoko (2D) version Spring member ITYPE=2 is tri-linear hysteresis only. If ITYPE=3 then
the SINA hysteresis is used for the transverse (local y) direction.
For the Ruaumoko (2D) version the Frame member ITYPE=7, the four hinge beam has the allowable
hysteresis table that follow the same rules as that for the variable flexibility beam.
15
DEGRADING STIFFNESS parameters
16
IHYST = 1 Elasto-Plastic Hysteresis. - No further data required.
Note: This rule is not available for Variable Flexibility Beam Members.
Elasto-Plastic Hysteresis
17
IHYST = 2 Bi-Linear Inelastic. - No further data required.
Bi-Linear Hysteresis
18
IHYST = 3 RAMBERG-OSGOOD Hysteresis [Kaldjian 1967] - No further data required.
Note: The bi-linear factor in the section data is used as the Ramberg-Osgood factor r and must be
greater than or equal to 1.0
Ramberg-Osgood Hysteresis
19
It must be noted that the Ramberg-Osgood loop works well when large cycle loops are exercised but an off-
set of the forces can occur in some small cycles as is shown in the diagram above. In 1984 the Ramberg-
Osgood hysteresis loop in Ruaumoko was modified to bound the forces within an envelope obtained by the
loops from the maximum and minimum displacements. In the year 2000 the Ramberg-Osgood loop reverted
to its original definition and the bounded version was moved to IHYST = 40.
20
IHYST = 4 Modified TAKEDA Hysteresis [Otani 1974].
Note: Increasing ALFA decreases the unloading stiffness and increasing BETA increases the reloading
stiffness. The power factor NF is usually taken as 1.0
21
IHYST = 5 Bi-linear with Slackness Hysteresis.
The bi-linear with slackness model can be used to represent diagonal braced systems where yield in in one
direction may stretch the members leading to slackness in the bracing system. The model allows for either
yield in compression, in say a cross-braced system, or for a simple elastic buckling in compression which
would be more appropriate in a single brace member.
Notes:
1 Concrete and Steel Beam-column sections require GAP+ = - GAP-.
2 If no value is prescribed for RCOMP, i.e. there are less than four items on the line or RCOMP is the
word DEFAULT or D then the bi-linear factor in compression is the same as that for tension.
3 For the SPRING members the hysteresis data is the same for all actions. If different properties are
desired in the different actions then separate members should be used for the different actions. The
default bi-linear factor is that for the force components.
4 If the strain rate constant C is non zero then the positive and negative yield forces are multiplied by the
factor
where is the current strain rate and is the quasi-static strain rate EPS0.
22
HYST = 6 KIVELL Degrading Stiffness [Kivell 1981].
The pinching model of Kivell was designed to represent the behaviour of the nails in steel nail-plates
connecting timber members together at the joints. The assumed cubic unloading-reloading curve is
represented by three straight lines.
ALFA
23
IHYST = 7 Origin-Centered Bi-linear Hysteresis - No further data is required,
On unloading and on subsequent reloading the path is on a line passing through the origin.
Origin-Centred Hysteresis
24
IHYST = 8 SINA Degrading Tri-linear Hysteresis. [Saiidi 1979].
Notes:
1. The i refers to the different actions on the member, see the member data descriptions for the number of
actions and which action they refer to.
2. Concrete and Steel Beam-column sections require symmetry in moments and thus
FCR(i)- = - FCR(i)+ etc. and that ALFA = BETA
25
IHYST = 9 Wayne Stewart Degrading Stiffness Hysteresis. [Stewart 1984].
This very general rule was initially developed by Wayne Stewart for the representation of timber framed
structural walls sheathed in plywood nailed to the framework. The model allows for initial slackness as well as
subsequent degradation of the stiffness as the nails enlarged the holes and withdrew themselves from the
framework.
Notes: Concrete and Steel Beam-column sections require GAP+ = - GAP- and that all other
components maintain symmetry about the zero force or moment axis
This rule is not available for the Variable Flexibility or 4-Hinge Beam members.
Modified loop;
Member section yield values are taken as Fu+ and Fu- and the Fu on this line is taken as Fy+ and Fy-.
This modification allows the use of strength degradation. In the original model strength degradation
affects only the cracking moments Fy and not Fu. Also the PTRI (read as part of this data line) and r
(read as part of the member section data) are interchanged.
26
Operation of Wayne Stewart Hysteresis rule
Note: Vos = Fi etc.
27
Wayne Stewart used the following hysteresis values in his plywood wall examples [Stewart 1984].
Example: The diagram below shows the use of the modified Wayne-Stewart hysteresis loop to model a pre-
1970 reinforced concrete column hinge where plain round longitudinal reinforcement bars are used
[Liu,2001]. The two loops compare the observed experimental loop with that computed using the
program HYSTERES using the following parameters
28
IHYST = 10 Degrading Bi-linear Hysteresis. [Otani 1981].
This is similar to the Bi-linear rule except that the stiffness degrades with increasing inelastic deformation.
ALFA
29
IHYST = 11 CLOUGH Degrading Hysteresis. [Otani 1981] - No further data is required.
This rule was the first degrading stiffness rules to represent reinforced concrete members. The rule is the
same as the modified TAKEDA rule with the parameters ALFA and BETA both equal to 0.0.
30
IHYST = 12 Q-HYST Degrading Stiffness Hysteresis. [Saiidi 1979].
This rule is the same as the Modified Takeda rule with the parameter BETA set to 0.0 and unloading as per
Emori and Schnobrich.
ALFA
Note: This rule is not available for the Variable Flexibility and 4-Hinge Beam Members.
31
IHYST = 13 MUTO Degrading Tri-linear Hysteresis. [Muto 1973].
After cracking the model is an Origin-Centered rule. After yield is reached the model become a Bi-linear
hysteresis with the equivalent elastic stiffness equal to the secant stiffness to the yield point.
Notes:
1. The i refers to the different actions on the member, see the member data descriptions for the number of
actions and which action they refer to.
2. Concrete and Steel Beam-column sections require symmetry in moments and therefore
FCR(i)- = - FCR(i)+ etc.
32
IHYST = 14 FUKADA Degrading Tri-linear Hysteresis. [Fukada 1969].
Notes:
1. The i refers to the different actions on the member, see the member data descriptions for the number of
actions and which action they refer to.
2. Concrete and Steel Beam-column sections require symmetry in moments and thus
FCR(i)- = - FCR(i)+ etc.
33
IHYST = 15 Bi-linear Elastic Hysteresis. - No further data is required.
This rule is similar to the Bi-linear hysteresis except that the rule unloads elastically down the same path which
means that no hysteretic energy is dissipated..
34
IHYST = 16 Non-linear Elastic Hysteresis - No further data is required.
This non-linear elastic model represents the non-linear behaviour of face-loaded masonry wall units.
No hysteretic energy is dissipated.
35
IHYST = 17 Degrading Elastic Rule. - No further data is required.
The degradation of the elastic stiffness is proportional to the amount of equivalent ductility. This equivalent
ductility is equal to the displacement divided by the nominal yield deformation.
36
IHYST = 18 Ring-Spring Hysteresis. [Hill 1994]
This device can be used as a seismic energy dissipation device. The default model operates in the compressive
force - compressive displacement quadrant of the force-displacement plot.
Ring-Spring
Note:
1. This rule is normally only available for the Spring Members. In this case do not supply yield data as the
yield point is defined by DXINIT, see below.
2. The rule may be used for the flexural components of the Giberson one-component beam option of the
FRAME members when it would normally be expected to be used in the bi-directional mode. It may also
be used for the axial component of the Giberson beam members provided the beam has no flexural
stiffness i.e. EI is zero, representing a truss-like action. In both of these cases the appropriate yield
moments or yield forces must be provided with dummy non-zero values (the actual yield values are
computed internally by the hysteresis rule but non-zero yield forces or moments are required in order that
the member is treated as non-linear).
3. When initial pre-stress (FRAME members) or pre-load (SPRING members) forces are applied to the Uni-
directional Ring-spring (KTYPE = 0) they must be compressive (i.e. negative) forces.
Ring-Spring Hysteresis
37
IHYST = 19 Hertzian Contact Spring Hysteresis Rule. [Davis 1992]
The Hertzian contact spring is useful for modelling the contact between impacting structures. It really models
contact between spheres but this seems to be used in wider applications. It is only available for the SPRING
members and the CONTACT members.
38
IHYST = 20 MEHRAN KESHAVARZIAN Degrading and Pinching Hysteresis. [Keshavarzian 1984]
ALFA
Note: This rule is not available for the Variable Flexibility and 4-Hinge Beam Members.
39
IHYST = 21 WIDODO Foundation Compliance Model.[Widodo 1995]
These non-linear elastic rules are designed to model foundation compliance springs including the modelling of
a wall footing that can suffer partial or tip uplift. This is only appropriate to SPRING members, see section 12..
Notes:
1. The i refers to the different actions on the member, see the member data descriptions for the number
of actions and which action they refer to.
2. This rule is not available for the Variable Flexibility Beam members.
40
IHYST = 22 Li XINRONG Reinforced Concrete Column Hysteresis. [Li Xinrong 1994]
This rule is only available for the Reinforced Concrete column option for the FRAME members, see section 11.
The degrading rule modifies the stiffness of the member to allow for the effects due to variation of the axial force
acting in the column.
41
IHYST = 23 BOUC Hysteresis Rule. [Wen 1976]
This very general parametric hysteresis rule gives a smooth transition of the change of stiffness as the
deformation of the member changes. It has been used to represent lead-rubber bridge bearings or energy
dissipators [Bessasson 1992] and has been used for the analysis of inelastic buildings subjected to random
vibration [Baber 1981].
A1 A2 A3 A4 A5 N D3 D4 D5 MODE INIT
Constantinou
where and are the displacement velocity, is the elastic stiffness, is the bi-linear factor and is the yield
displacement. is the yield force and is the change in displacement. Q is the force in the spring and F is
the current stiffness factor. The tangent stiffness is . and are the Bouc hysteresis control
parameters, is the power factor.
42
The Bouc rule is controller by the parameter , which in RUAUMOKO is initially 0.0 and is integrated step-by-step
as in the above equations. The rule is such that at the static analysis which means that the initial
stiffness is the bi-linear stiffness and the force in the member is proportional to the bi-linear stiffness and the
displacement. To over come some of these effects an option is to force the rule to be bi-linear until reversal after
the first yield. A result is that there is a marked reduction in member force at the change of rule. further work is
being done to fully understand the implications of the use of this hysteresis.
43
IHYST = 24 REMENNIKOV Steel Brace Member Hysteresis.
Notes:
1. This hysteresis is available only for the Giberson One-component beam and the Steel beam-column
options of the FRAME member type. (see Section 11a)
2. The member only permits this hysteresis in the axial component. The member is assumed to be bi-linear
in flexure (provided the yield moments are non-zero in sections 11e or 11g).
3. The beam or beam-column cross-section area and the axial yield forces in section 11e or the yield
interaction forces and moments in section 11g must be supplied.
4. It is recommended that Iteration on Residuals, say MAXIT = 3 and FTEST = 0.001, be used with this rule
(see section 5).
5. It is recommended that a small time step be used so that the transitions within the rule may be followed.
44
45
Definition of Different Zones
Idealization Curves for Tangent Modulus History
46
IHYST = 25 TAKEDA with SLIP Hysteresis. [Kabeyasawa 1983]
This rule allows slip when the deformation reloads in the member strong direction.
ALFA Unloading Degradation parameter (see Takeda ALFA) (0.0 # ALFA # 1.0) F
BETA1 Slipping stiffness parameter F
BETA2 Re-loading stiffness parameter F
FC(i) Cracking Force for Component i ( > 0.0) F
RC(i) Cracking Displacement for Component i ( > 0.0) F
Notes:
1. The i refers to the different actions on the member, see the member data descriptions for the number
of actions and which action they refer to.
2. The initial elastic stiffness supplied with the section properties is the secant stiffness passing through the
origin and the yield points on the member hysteresis curve.
3. The yield points in the negative direction for component i is at ( -FC(i), -RC(i) )
4. Slip only occurs when re-loading towards the stronger direction.
47
The following example show the use of the Takeda with Slip hysteresis loop to model the behaviour of a pre-1970
reinforced concrete beam which is reinforced with plain round bar reinforcement [Liu, 2001]. The loops show the
observed experimental loop and the matching loop from the program HYSTERES using the loop parameters
provided below.
48
IHYST = 26 AL-BERMANI Bounding-surface Hysteresis. [Zhu, 1995]
This rule allows for the Bauschinger effects in steel members by using a bounding surface rather than the more
often used but more complicated Ramberg-Osgood functions.
ALFA BETA
Notes:
1. The bi-linear factor p must be greater then 0.001. (see section 11c or section 12a)
This limitation on p should prevent mathematical difficulties with the logarithmic and exponential
functions.
2 The use of this rule in Concrete and Steel Beam column members will require symmetry in the actions
at each end, i.e. ALFA = BETA etc.
Bounding-Surface Hysteresis
49
IHYST = 27 PEAK Oriented Hysteresis. - No further data is required.
This rule is similar to the Origin Centered rule except that on unloading the force-displacement relationship moves
along a line to the maximum force-displacement point in the opposite direction. If yield has not occurred in that
direction the opposite yield point is used as the target.
50
IHYST = 28 MATSUSHIMA Strength Reduction Hysteresis. [Matsushima 1969]
This rule represents the behaviour of short reinforced concrete columns failing in shear. The rule uses basically
a bi-linear hysteresis but that the elastic stiffness and strength degrade every time unloading takes place from
the post-yield part of the bi-linear force displacement curve.
Matsushima rule.
A B
Notes:
1. A and B are raised to then power N where N is the number of times the system unloads from the bi-linear
force-displacement hysteresis back-bone.
51
IHYST = 29 KATO Degrading Shear Model. [Kato 1983]
Represents the behaviour of a reinforced concrete member failing in shear. A tri-linear skeleton curve with a
falling bi-linear part is used.
Notes:
1. The i refers to the different actions on the member, see the member data descriptions for the number
of actions and which action they refer to.
2. The bi-linear factor p must be less than 0.0 and less than the tri-linear factor PTRI. (see section 11c or
section 12a)
3. The actions FU should less than the appropriate yield actions FY.
4. The use of this rule in Concrete members will require symmetry in the actions at each end, i.e. FU(i)+
= -FU(i)- etc.
5. For reinforced concrete members failing in shear the recommended values are ALFA = 0.4, BETA = 0.6
and GAMMA = 0.95.
6. For reinforced concrete members flexure dominated members the recommended values are ALFA = 0.2,
BETA = 0.0 and GAMMA = 0.0.
52
IHYST = 30 Elastomeric Spring Damper Hysteresis. [Pekcan 1995]
Represents a double-acting elastomeric spring which has resistance due to both displacement and velocity. The
stiffness properties are basically bi-linear elastic.
C DMAX ALFA
Notes:
1. The force FD in the elastomeric damper is given by
where the d is the displacement of the spring and where K0 is the initial stiffness of the device and is the
longitudinal spring stiffness of the member. The stiffness rK0 is the stiffness after the prestress Fy is
overcome where r is the bi-linear factor for the spring member. The prestress force Fy is taken as the
positive longitudinal yield force of the member and if the prestress is zero the spring stiffness is taken
as constant equal to rK0.
2. If DMAX is zero then the ratio of d to dmax is taken as 1.0. This has the effect of taking the exponent $
in the reference paper as zero.
3. The exponent ALFA was taken as 0.2 in the reference.
53
IHYST = 31 Composite Section. - Modified SINA Hysteresis.
This rule allows for the modelling of composite concrete-steel beams or concrete T beams where the behaviour
is different in the positive and negative flexural actions.
BETA Stiffness Factor Cracking to Yield in Negative direction (ALFA < 1.0) F
FCR(i) Cracking Force in Negative direction at component i (FRC(i) < 0.0) F
FCC(i) Cracking Closing Force at component i (FCC(i) < 0.0) F
Notes:
1. The i refers to the different actions on the member, see the member data descriptions for the number
of actions and which action they refer to.
2. The post cracking stiffness factor BETA must be greater than the bi-linear factor r (see member
properties sections)
54
IHYST = 32 Different Positive and Negative Stiffness. Modified Bi-linear Hysteresis.
This rule allows for different stiffnesses in the positive and negative directions. The basic hysteresis rule is a
modification of the degrading Bi-linear rule.
ALFA BETA
ALFA Negative stiffness factor of the positive stiffness (ALFA > 0.1) F
BETA Unloading Degrading Factor, see TAKEDA ALFA (0.0 < BETA < 0.9) F
55
IHYST = 33 Masonry Strut Hysteresis (Crisafulli 1997).
This rule allows for the modelling of masonry panels in framed structures. If the strut model is used with the spring
members then only the longitudinal stiffness is specified for the strut member and two struts are required to model
each panel, one strut across each diagonal of the panel. The masonry strut hysteresis is also used for the
Masonry Panel Element where four struts represent the panel together with a shear spring..
Stress-strain relationship
Strut data
56
Masonry Strut Strength Envelope
57
Notes:
FC The compressive strength f'm2 is the main parameter controlling the resistance of the strut. It must be
noted that FC does not represent the standard compressive stress of masonry but should be adopted
taking into account the inclination of the compressive principal stresses and the mode of failure expected
in the masonry panel. See Crisafulli 1997.
FT Tensile strength f't represents the tensile strength of the masonry or the bond strength of the panel-frame
interface, whichever is smallest. The consideration of the tensile strength has been introduced in the
model in order to gain generality. However, results obtained from different examples indicate that the
tensile strength, which is generally much smaller than the compressive strength, has no significant
influence on the overall response. Therefore, in the absence of more detailed information, the tensile
strength can be assumed to be zero.
UC The strain at maximum stress g'm usually varies between -0.002 and -0.005 and its main effect on the
overall response of the infilled frame is the modification of the secant stiffness of the ascending branch
of the stress-strain curve.
UUL The ultimate strain gu is used to control the descending branch of the stress-strain relationship. When
a large value is adopted for, example gu = 20 g'm , a smooth decrease of the compressive stress is
obtained.
UCL The closing strain gcl defines the limit strain at which the cracks partially close and compressive stresses
can be developed. Values of the closing strain ranging between 0 and 0.003 lead to results which agree
adequately with experimental data. If a large negative value is adopted, for example gcl = gu , this effect
is not considered in the analysis.
EMO The elastic modulus Emo represents the initial slope of the stress-strain curve and its value can exhibit
a large variation. Various expressions have been proposed for the evaluation of the elastic modulus of
masonry. It is worth noting, however, that these expressions usually define the secant modulus at a
stress level between 1/3 and 2/3 of the maximum compressive stress. In order to obtain an adequate
ascending branch of the strength envelope it is assumed that Emo $ 2 f'm2 / g'm .
GUN The unloading stiffness factor (un controls the slope of the unloading branch. It is assumed to be greater
than or equal to 1.0 and usually ranges from 1.5 to 2.5.
ARE The reloading strain factor "re defines the point where the reloading curves reach the strength envelope.
The calibration of the hysteretic model for the axial behaviour of masonry showed that good results are
obtained using values ranging between 0.2 and 0.4. However, higher values, for example 1.5, are
required to model adequately the cyclic response of the infilled frames. This is because other sources
of nonlinear behaviour, such as sliding shear, need to be indirectly considered in the response of the
masonry struts.
Four parameters are required to represent the cross-sectional area of the masonry strut. These are the initial area
Ams1 = AREA1 and final area Ams2 = AREA2 and the axial displacements at which the cross-sectional area
changes, )R1 = R1 and )R2 = R2. In a simplified model, it can be assumed that AREA1 and AREA2 are the
same using a low value of the strut area to avoid an excessive increase in the axial strength. In a more refined
analysis, a higher value of the initial area can be adopted, whereas the final area can be reduced by about 10%
to 30%. The displacement R1 and R2 can be estimated as g'm dm/5 and g'm dm (where dm is the length of the
masonry strut) respectively, at least until more precise information becomes available. Several empirical
expressions, which are described in section 6.2.1.3 of the reference, have been proposed for the evaluation of
the equivalent width of the masonry strut, whose value normally ranges from 0.1 to 0.25 of the diagonal length
of the infill panel.
IENV The descending branch of the stress-strain curve is usually modelled with a parabola instead of the curve
associated with Sargin's equation in order to obtain a better control of the response of the strut after the
maximum stress has been reached.
58
IHYST = 34 Hyperbolic Hysteresis (no further data required)
(Konduor and Zelansko (1963))
(also Duncan and Chang (1970))
This rule has been popular in representing the shear stress–shear strain relationships in soils subjected to
earthquake excitation.
Hyperbolic Hysteresis
59
IHYST = 35 Degrading Bilinear with Gap Hysteresis
This hysteresis was initially developed to model a strain-hardening behaviour which changed with increasing
cycle number. The members were used in parallel with a member having a more conventional hysteretic
behaviour such as Bi-linear or Ramberg-Osgood. The total member force was taken as the sum of the two
number forces.
60
IHYST = 36 Bi-linear with Different +/– stiffness hysteresis
The rule is to represent actions which exhibit different stiffnesses under positive or negative forces or moments.
This may be typical of reinforced concrete T sections for example.
61
IHYST = 37 Non-linear Elastic Power Rule
This rule is similar to the Bi-linear Elastic Hysteresis, IHYST = 15 except that it avoids the problems with the
sudden change of stiffness on unloading encountered due to the lack of energy dissipation in these non-linear
elastic hysteresis rules.
PFP(i) PFN(i)
PFP(I) Power factor in positive direction for component i (0.01 # PFP # 3.0) F
PFN(i) Power factor in negative direction for component i (0.01 # PFN # 3.0) F
Notes:
1. The i refers to the different actions on the member, see the member data descriptions for the number
of actions and which action they refer to.
1. Normally PFP and PFN are less than 1.0
62
IHYST = 38 Revised Origin Centred Bi-linear Hysteresis (No further data is required)
On unloading the path is back to the origin. In reloading the path follows the previous unloading path on that side
of the origin.
63
IHYST = 39 Dodd-Restrepo Steel Hysteresis Rule
This hysteresis rule is designed to allow for the Bauschinger effects in the steel hysteresis.
Reference:
Dodd, L.L. and Restrepo-Posada, J.I. Model for Predicting Cyclic Behaviour of Reinforcing Steel. J. Structural
Engineering, ASCE, Vol. 121, No. 3, Mar. 1995, pp 433–445.
64
Plot of Displacement History
65
IHYST = 40. Bounded Ramberg-Osgood Hysteresis. No extra data required.
This loop is similar to that for IHYST = 3 except that bounds have been applied to the forces so that off-sets to
the forces do not occur during small cycles of displacement.
66
IHYST = 41. Pyke modification to the Ramberg-Osgood Hysteresis. No extra data required.
This loop is similar to that for IHYST = 3 except that the small cycle behaviour has been modified to prevent off-
sets on the force in these small cycles. These loops were initially used to model the behaviour of soils.The small
loops indicate a greater rate of change of force with increasing displacement i.e. a greater curvature in the plots.
67
IHYST = 42. HERA - SHJ (Sliding Hinge Joint) Hysteresis.
This loop developed by the Heavy Engineering Research Association (New Zealand) is to represent the
behaviour of a sliding moment connection between steel beams with a concrete slab above them and connected
to at the joints to steel columns.
68
69
IHYST = 43 Resettable Actuator Hysteresis. No extra data required.
This hysteresis is to represent the behaviour of a semi-active damper member [Hunt, 2003]. The force is
proportional to the displacement until a saturation force is attained, Fy+ or Fy- (the yield forces for the
member) when the system appears to show a perfectly plastic response. On any reversal of displacement the
force is automatically reset to zero and the origin is moved to the existing displacement and the system will
then behave as an elastic member until either saturation is achieved or the displacement again changes sign.
70
IHYST = 44 Pampanin Reinforced Concrete Beam-Column Joint Hysteresis.
70
IHYST = 45 Degrading Stiffness RAMBERG-OSGOOD Hysteresis.
Note: The bi-linear factor in the section data is used as the Ramberg-Osgood factor r and must be greater
than or equal to 1.0
Ramberg-Osgood Hysteresis
71
IHYST = 46 DEAN SAUNDERS Reinforced Concrete Column Hysteresis.
This hysteresis is to represent the behaviour of Older Reinforced Concrete Columns where plain round
reinforcement is used.
Notes: 1. There is a Bezier curve fitted between the point (MTH+,(MTH+)/STIFF) and (YP,PHIY+) with
the initial slope STIFF and the final slope r*STIFF.
2. On unloading the rule is origin centered when the action is less than MTH+. When the action
is greater than MTH+ the unloading stiffness degrades with the factor ALFA as observed in
the Dergading Bi-Linear Hysteresis (IHYST = 10).
3. The behaviour in the negative Action-Deformation quadrant is identical to that in the positive
Action-Deformation quadrant.
4. The parameter BETA only comes into action if the deformation in the other quadrant has
exceeded the yield deformation.
72
IHYST = 47 Multi-Linear Elastic Hysteresis.
N F1 D1 F2 D2 F3 D3
Note: The F1, F2 and F3 factors should not be less than 0.0 for single degree of freedom systems.
73
IHYST = 48 Kinematic/Isotropic Strain Hardening Bi-Linear Hysteresis.
ALFA BETA
Note: The bi-linear factor ‘r’ in the section data must be greater than 0.0 if Kinematic or Isotropic strain
hardening is to occur
74
IHYST = 49 Kinematic/Isotropic Strain Hardening RAMBERG-OSGOOD Hysteresis.
Note: The bi-linear factor in the properties section data is used as the Ramberg-Osgood factor r and must
be greater than or equal to 1.0
Ramberg-Osgood Hysteresis
75
IHYST = 50 Flag-Shaped Bi-Linear Hysteresis.
76
MEMORANDA
77
MEMORANDA
78
MEMORANDA
79