Académique Documents
Professionnel Documents
Culture Documents
Problem
Input
1. Finite element mesh
Figure 1 shows the finite element mesh used for the concrete beam (four point bending) problem used
for discrete cracking. Interface elements are used only for discrete cracking analysis. These are element
81-81. Elements 1-80 are continuum elements. For the smeared cracking analysis we only used those.
2. Element type(s) (degrees of freedom, interpolation, dimension, stress state, )
Element type(s)
Degrees of freedom
Interpolation
Dimension
Stress state
Quadratic
25x25 mm
Plane stress
Constraints shown in red at beam ends. (ux=0 and uy=0). At the top a load displacement of -0.5 mm in
global y direction is applied seen in red.
Continuum (plane)
Concrete
40000 N/mm^2
0.2
-
5. Cross-sectional properties
Length
Height
Thickness
500 mm
100 mm
50 mm
Interface (line)
Concrete
1600000 N/mm^3
66667 N/mm^3
3 N/mm^2
0.125 N/mm
The linear static analysis does not take cracking into account as we expected. With a total load
displacement of -1mm in y-direction the stresses and strain reach beyond the failure criterion in the
beam (figure 3 and 4).
Analysis results
6. Displacements (e.g. as deformed model: scaled or non-scaled?; combined with e.g. stress
contour plots)
Figure 6 Deformed model combined with stress contour plot (SXX) at pre peak (step 4)
Figure 7 Deformed model combined with stress contour plot (SXX) at peak (step 14)
Figure 8 Deformed model combined with stress contour plot (SXX) at post peak (step 21)
Figure 9 Deformed model combined with stress contour plot (SXX) at post peak (step 80)
Figure 6-9 show the deformed models combined with the stress contour plots (SXX). The crack occurs
when the failure criterion is reached in the interface elements. This is the tension strength of 3MPa at
the bottom of the beam. Interesting to see are the stresses in figures 7 and 8 at the bottom of the beam.
Tensile stresses are larger than the failure criterion and should crack in reality. We used continuum
elements there so they only behave linear without cracks. Extra analysis with more interface elements in
the beam would be interesting.
7.
Force-displacements curve
6000
5000
Force [N]
4000
3000
2000
1000
0
0
0,02
0,04
0,06
0,08
0,1
0,12
Displacement [mm]
Discrete
Figure 10 Force-displacements curve of discrete cracking analysis
Figure 10 shows the load-displacement curve of the discrete cracking analysis of a four-point bending
problem. The load is applied with displacement control of 100 equally sized steps. The load in this graph
are the reaction forces in y-directions and the displacement is the nodal displacement in y-direction of
the node on which the load is applied. Qualitative and quantitative comments are given in section 10.
Figure 11 and 12 show the traction (stresses) distribution over the height of the beam at the line
interface elements.
6000
5000
Force [N]
4000
3000
2000
1000
0
0
0,02
0,04
0,06
0,08
0,1
0,12
Displacement [mm]
Discrete
Figure 13 Load-displacement curve of discrete cracking with stress contour plots (SXX)
Figure 13 and 14 show the load-displacement curve of the discrete cracking analysis of the four-point
bending problem of a concrete beam. Displacement controlled load is applied at the top of the beam.
This is done in 100 equally sized load steps. Four adequate stress (SXX) and traction diagrams are added
in this figures. As seen in the figures the force-displacement curve initially increases linear. The beam
has a linear stress distribution over its height (figure 14). Then pre-peak the failure criteria is reached at
the bottom of the beam in tension (in this case 3MPa). This causes the start of the crack and a change of
stress-relative displacement relation of the interface element namely the linear tension softening. This
happens consecutively until the peak where the beam reaches its bearing capacity (around 5700 N). The
post-peak shape of the curve is exponential. This is what we expected for the concrete beam with the
given fracture energy (=area underneath) and brittle fracture. The crack keeps growing and the tension
and compression zone of the beam decrease.
6000
5000
Force [N]
4000
3000
2000
1000
0
0
0,02
0,04
0,06
0,08
Displacement [mm]
Discrete
Figure 14 Load-displacement curve of discrete cracking with adequate traction diagrams
0,1
0,12
Analysis results
6. Force-displacements curve
Before discussing the Force-displacement graph, see below information (from the .out file) for load step
7 and 8.
From the screen shot above, while there was no crack at the 7th load step, we notice 12 cracks had
opened at Load step 8. This would play a role in the Force-Displacement relationship. Using a 0.01(100)
i.e. specified load steps, the Force-Displacement relationship illustrated below in Fig. 15 was obtained.
Figure 15 Load-displacement relation for prescribed deformation of 0.5 mm (in 100 steps)
From Fig. 15, the relationship is linear up till Load step 7, and non-linear from step 8. The appearance of
cracks is the reason for this. With more load steps, we notice the slope (and thus the stiffness) reducing
as more cracks appeared. More deformation is obtained with less force increment up till the 19th Load
step. By the 20th load step, further deformation even with small reduction in force. Thus this relationship
shows an initial linear relationship, followed by non-linear relationship as cracks appeared and grew.
After the 20th load step, there was divergence and the FEM software did not give results. After
modifying the maximum number of iterations, a modified plot was obtained illustrated in Fig. 16
Fig. 16 shows that a significant loss stiffness after the 20th load step, thus the force that can be
supported dropped significantly. Nevertheless, it does not become zero immediately. The fracture
energy play a role in the post-peak behavior (as seen in tension-softening in Fig. 16).
NB: I would use refer to Figure 16 subsequently. It was linear up till the 4th load step when 12 cracks
appeared. Non-linear relationship afterwards till the Peak (at the 12th Load step). Significant loss of
stiffness afterward.
7. Stress, crack and/or traction distributions
From crack initiation at Load step 4 (with 12 cracks), the cracks grew with increasing Load steps as
illustrated in Figures 17a d which shows the crack patterns at initiation, Pre-peak, at Peak and Postpeak respectively.
From the .OUT file, interesting facts on crack formation can be studied. From 12 crack at initiation (at
load step 4), the cracks grew to 86 at Load step 15. It is however interesting to note that no new crack
was formed afterwards. This behavior correspond to that expected on concrete subjected to imposed
deformation where there is a crack-formation stage (at which cracks are formed), and a stage of Fully
developed crack pattern (where no new crack is formed, rather the existing crack only grow). This
behavior was obvious from studying the .OUT file.
8. Displacements, stresses, strains etc. (e.g. as deformed model: combined with stress contour plots)
In Figure 21a - d below, I have presented Principal Strain data (E1) for Load steps 11 (pre-peak), 12 (at
Peak), 13 and 38 (both Post-Peak). I have indicated the elastic strain limit (
and the plastic strain limit
on the contour plots using a yellow and red coloured line respectively.
Tensile strength
Thickness beam
Height beam
Section modulus
Maximum moment
Maximum force
3 MPa
50 mm
100 mm
83333.33 mm^3
250000 Nmm
3333.33 N
Force [N]
5000
4000
3000
2000
1000
0
0
0,02
0,04
0,06
0,08
0,1
0,12
Displacement [mm]
Discrete
Smeared
discrete cracking starts peaking earlier than the smeared. Eventually the smeared overshoots the
discrete. A higher ultimate bearing capacity of the beam is reached. Reasoning for this behaviour is hard.
But it might have something to do that the discrete cracking with just one line of interface elements is
too conservative. More interface elements might be needed for more accurate behaviour.
13. Voluntary: comparisons between analyses (e.g. an increase of the fracture energy)
To study the effect of the fracture energy, we used other fracture energies and studied the effect.
Figure 23 shows the Force-displacement behaviour when a Fracture energy was used.
Force (N)
3500
3000
2500
2000
Series1
1500
1000
500
0
0
0,2
0,4
0,6
0,8
1,2
Displacement (mm)
Fig. 23 Effect of Fracture energy increase
The fracture energy had significant impact on the post-linear elastic behaviour of the structure. Unlike in
Figure 16 where there was sharp drop in stiffness and reaction force beyond a certain peak force,
increased fracture energy gave more capacity to the structure after the elastic region. The higher the
fracture the energy, the closer the structure get to an ideal plastic behaviour. Similarly, the lower the
fracture energy, the closer it tends towards a fully brittle material.
Behaviour in Unloading
When unloaded in several load steps, the outcome is presented in graphical form below in Fig. 24. It is
obvious that significant permanent strains have crept in. This is clearly demonstrated in Fig. 24.
Force (N)
Loading
Unloading
0,2
0,4
0,6
0,8
Displacement (mm)
Fig 24 Loading & unloading
1,2