Basic Tutorials

LS-DYNA / LS-PrePost
Ex. 2. Tensile test

2017-05-17 LS-DYNA / LS-PrePost

1 Introduction

Contents
1 Introduction ............................................................................................................................2
1.1 Purpose ..........................................................................................................................2
1.2 Prerequisites...................................................................................................................2
1.3 Problem Description........................................................................................................2
1.4 Data files ........................................................................................................................2
2 Part A - Explicit analysis .........................................................................................................3
2.1 Read geometry ...............................................................................................................3
2.2 Material properties ..........................................................................................................3
2.3 Element properties..........................................................................................................4
2.4 Boundary conditions .......................................................................................................5
2.5 Prescribed motion/displacement .....................................................................................5
2.6 Set the termination time ..................................................................................................6
2.7 Output ............................................................................................................................6
2.8 Save ...............................................................................................................................7
2.9 Run the simulation ..........................................................................................................7
2.10 Post processing ..............................................................................................................8
3 Part B - Implicit analysis ....................................................................................................... 10
3.1 Save the model ............................................................................................................. 10
3.2 Run the simulation ........................................................................................................ 10
3.3 Post processing ............................................................................................................ 11
4 Comments............................................................................................................................ 13
4.1 Explicit versus implicit ................................................................................................... 13
5 Optional exercises ................................................................................................................ 13

1 LS-DYNA / LS-PrePost
1 Introduction

1 Introduction
1.1 Purpose
• Get better knowledge in basic material modeling.
• Learn how to set up explicit and implicit simulations in LS-DYNA.

1.2 Prerequisites
• Basic knowledge in the finite element method.
• Understand the steps in tutorial 1 - Getting Started.

1.3 Problem Description

The task is to perform a tensile test on a flat specimen. The end of the specimen is constrained while
a prescribed motion is applied on the other end. The tutorial consists of two parts.

• Perform a tensile test with the explicit solver in LS-DYNA

• Solve the problem using the implicit solver in LS-DYNA.

The task is to compare the stress vs. strain curve in an element when the uniaxial tension test is
simulated using either the explicit or implicit solver in LS-DYNA.

Material properties

Density, 𝜌 7850 kg/m3

Young’s modulus, 𝐸 210 GPa
Poisson’s Ratio, 𝜈 0.3
Yield limit 250 MPa
Tangent modulus 1000 MPa
1.4 Data files
• The geometry that will be used - tensile_test.k.
• The final results - tensile_test_results.k and tensile_test_results_implicit.k.

2 LS-DYNA / LS-PrePost
2 Part A - Explicit analysis

2 Part A - Explicit analysis

The first part will show how to perform a tensile test with the explicit solver in LS-DYNA.

2.1 Read geometry

Open tensile_test.k in LS-PrePost, which contains the geometry of the test specimen.

2.2 Material properties

LS-DYNA expects, for most materials, input in terms of true stress vs. true strain, not engineering
stress vs. engineering strain. Normally an experimental uniaxial tension test is performed and
engineering stress and strain data are obtained. Before you use this data as input in a material model
in LS-DYNA, it must be converted to true stress and strain. The curves normally behave as in the
figure. How you convert engineering stress and strain to true stress and strain are showed below.

Engineering stress and strain

𝐿 − 𝐿0
𝜀𝑒𝑛𝑔 =
𝐿0
𝐹
𝜎𝑒𝑛𝑔 =
𝐴0

True stress and strain

𝐿
𝑑𝐿 𝐿
𝜀𝑡𝑟𝑢𝑒 = ∫ = 𝑙𝑛 ( ) = 𝑙𝑛(1 + 𝜀𝑒𝑛𝑔 )
𝐿 𝐿0
𝐿0
𝐹 𝐹 𝐴0 𝐹 𝐴0 𝐿 𝐴0
𝜎𝑡𝑟𝑢𝑒 = = = = (𝑐𝑜𝑛𝑠𝑡𝑎𝑛𝑡 𝑣𝑜𝑙𝑢𝑚𝑒 = ) = 𝜎𝑒𝑛𝑔 (1 + 𝜀𝑒𝑛𝑔 )
𝐴 𝐴 𝐴0 𝐴0 𝐴 𝐿0 𝐴

3 LS-DYNA / LS-PrePost
2 Part A - Explicit analysis

To create the material card do as follows: Click Model > Keywrd. Activate All in the Keyword
Manager. Double-click MAT > 024-PIECEWISE_LINEAR_PLASTICITY, which is an elasto-plastic
material. This is a widely-used material model and the user can define the plastic behavior in several
ways.

1. Define Yield stress SIGY and Tangent Modulus ETAN (see figure), which gives linear
hardening.
2. Define effective plastic strain EPS and the
corresponding yield stress value ES. At least two
points needs to be defined. This will give a piecewise
linear plastic behavior.
3. Define a curve of effective stress vs. effective plastic
strain.

In this tutorial, the first alternative will be used to define the

plastic region. Enter the title and the values as in the figure
below and click Accept, then Done.

2.3 Element properties

To set the element formulation and properties do as follows:

• Double click SECTION > SHELL in the Keyword Manager.

• Enter Specimen_shell as TITLE.
• Set T1 to 1.5 and press Enter, T2-T4 will then be changed to 1.5 as well and this defines
the thickness of the nodes in every element.
• Click Accept, then Done.
Now go to Part in the Keyword Manager and assign the newly created material and section. There
one can also change the name of the part, to Tensile specimen for example. Click Accept, then
Done.

4 LS-DYNA / LS-PrePost
2 Part A - Explicit analysis

2.4 Boundary conditions

Apply the fixed boundary conditions as follows:

• Click on the Bottom view in the

Floating Toolbar.

• Click Model > CreEnt.

• Double-click Boundary > Spc, click Cre. Make sure that Set
is selected in the Entity Creation window.
• Select Area in the node selection window (Sel.Node).
Make a box around the nine nodes as in the figure.
• Fix the nodes in X- and Z-translation and all rotations. This is
done by activating X, Z, RX, RY and RZ.
• Click Apply.

Now create another boundary condition:

• In the node selection window, write 80 in the ID box and press Enter. Write 87 and press
Enter. Two nodes will now be selected at the outer edges of the test specimen.
Fix the nodes in Y translation by only
activating Y.
• Click Apply.
• Close Entity Creation window.

2.5 Prescribed motion/displacement

As mentioned in the first tutorial, for motions and loads, a curve must be defined that states the
variation of the load/displacement/velocity etc. over time.

First create the curve defining the motion:

• In the top menu, click Application > Tools > CurveGen.

• Change Method to X - Y.
• Deactivate Smooth.
• Write X = 0 and Y = 0 (ignore the value of
smooth)
• Click Insert. Then 0.01 and 1, Insert.
• Finally, 0.011 and 1, Insert (the termination time
will be 0.01 s for this tutorial).
• Change Curve Name to e.g. Motion.
• Click Create, then Done.

5 LS-DYNA / LS-PrePost
2 Part A - Explicit analysis

Note that the curve could alternatively have been created using DEFINE > CURVE in the Keyword
Manager. A curve created in CurveGen will contain a larger amount of points, compared to a curve
created by DEFINE > CURVE.

Now apply the motion to the end of the tensile specimen:

• Click Model > CreEnt.

• In the Entity Creation window, double-click Boundary >
Prescribed Motion.
• Select Cre. Change Type to SET and activate Pick.
• Make a box around the nine nodes as in the figure.
• Set DOF = 1 (X-translational as degree of freedom) and VAD
= 2 (displacement as the prescribed nodal quantity, there are
possibilities to prescribe the nodal velocity or acceleration as
well).
• Click on LCID and select the previously created curve.
• Set SF = 20, which implies that the node set will be moved 20
mm.
• Click Apply, then Done.

2.6 Set the termination time

To set the termination time:

• Click Model > Keywrd.

• Double-click CONTROL > TERMINATION.
• Set ENDTIM to 0.01. The simulation will then last for 0.01 seconds.
• Click Accept and then Done.

2.7 Output
Here the output to be saved by LS-DYNA during the simulation is defined, first specify d3plot output:

• Create d3plots.
• Click Model > Keywrd.
• Double-click DATABASE > BINARY_D3PLOT.
• Set DT = 5e-4.
• Click Accept and then Done.
Since the task was to compare the stress vs. strain curve for an element with the material curve, we
want to obtain this data frequently. In the first tutorial (1 – Getting Started) we used the History
command to plot information. The time-history graphs created by History, use information found in
the d3plot files. To obtain data more frequently we can specify which type of data that are of interest
and how often it will be printed for selected elements. Do as follows:

• Click Model > Keywrd.

• Double-click DATABASE > ASCII_option. We are interested in the stress and strains in
the elements, therefore activate ELOUT. On the same row, set DT = 5e-6.
• Click Accept, then Done.

6 LS-DYNA / LS-PrePost
2 Part A - Explicit analysis

The element/elements that we want to gather data from must be defined:

• Double-click DATABASE >

HISTORY_SHELL in the
Keyword Manager.
• Write, for example, shell element
number 314 under ID1, which is
an element close to the center of the specimen.
• Click Insert, Accept, then Done.

Check the model to see that everything looks okay.

2.8 Save
The model is now ready to be saved. File > Save As > Save Keyword As. Choose a folder path
and name your file tensile_test_model.k for example. Note that the folder path cannot contain any
spaces. Close LS-PrePost.

2.9 Run the simulation

Run the simulation using LS-Run, see exercise 1. Getting Started for more information on how to
do this. The simulation runs in less than a minute to completion. Use the “SMP Single” preset.

7 LS-DYNA / LS-PrePost
2 Part A - Explicit analysis

2.10 Post processing

Open the d3plot in LS-PrePost from LS-Run or from the File menu of LS-PrePost.

Start to press the Forward button in Animate toolbar to see what happens with the specimen.

Now plot the stress-strain-curve from the elout file, first plot and save stress vs. time:

• Click Post > ASCII.

• Select elout* and click Load.
• Select Sh-314, Ip-1 or Ip-2 (doesn’t matter which integration point in this case) and 9-
Effective Stress (v-m).
• Click Plot.
• In the PlotWindow, click Save.
• Let the Output Type be Curve file.
• Enter Stress as the Filename (also set correct folder path).
• Click Save (located at the bottom toolbar, see red square).
• Click Quit

Now plot and save the effective plastic strain vs. time:

• Select 7-Yield Value (eps) instead of Effective Stress (v-m).

• Click Plot.
• Save the curve in the same way as for the stress, but name the curve Strain.
• Close PlotWindow and ASCII

Now combine the above two curves (Stress and Strain) to plot the stress vs. strain:

8 LS-DYNA / LS-PrePost
2 Part A - Explicit analysis

• Click Post > XYPlot.

• Activate Cross.
• Note that X-Axis is activated, select Strain in the top box and it will pop-
up under Curve Names.
• Click on Strain:1:Sh-314….. and it will be inserted in X:.
• When you do this, Y-Axis will be activated.
• Select Stress in the top box and click on Stress:1:Sh-314… under
Curve Names and it will be inserted in Y:.
• Click Plot.

The curve shows effective plastic strain vs. effective stress (von Mises). We
see that the yield limit is about 250 MPa as we stated in the material model.

Let’s check that the tangent modulus ETAN (1000 MPa) are correct as well. Click on two points on
the curve and calculate the tangent. Using the points in the figure gives:
377 𝑀𝑃𝑎 − 265 𝑀𝑃𝑎
= 999 𝑀𝑃𝑎
0.127 − 0.0149
Finally save the curve and name it Explicit_curve. Part A is now completed and you may close LS-
PrePost.

9 LS-DYNA / LS-PrePost
3 Part B - Implicit analysis

3 Part B - Implicit analysis

Run the tensile test using the implicit time-step method instead of using explicit time-step method
used in Part A.

Open the file tensile_test_model.k that you created earlier in LS-PrePost. Note: This can be done
also from LS-Run. Do as follows to change to the implicit time-step method:

• Click Model > Keywrd.

• In the Keyword Manager, make sure that All is selected, double-click CONTROL >
IMPLICIT_GENERAL. This keyword is mandatory to activate the implicit time-stepping
method.
• Set IMFLAG = 1, which implies an implicit analysis. As default in LS-DYNA, a non-linear,
static simulation will be performed when just defining an implicit solution. If you want to
perform a dynamic implicit simulation for example, you will have to specify that using other
control cards.
• Set the initial time step DT0 = 5e-4.
• Set IGS = 1, which can improve convergence in tensile.
• Click Accept, then Done.

3.1 Save the model

Save your model as tensile_test_model_implicit.k. Close LS-PrePost.

3.2 Run the simulation

Implicit analyses should always be run using the double precision version of LS-DYNA, as the implicit
solver is more sensitive to round-off errors.

Run the simulation using LS-Run, see exercise 1. Getting Started for more information on how to
do this. The simulation runs in less than a minute to completion. To use the double precision version,
make sure to use the “SMP double” preset in LS-Run, see the image below.

10 LS-DYNA / LS-PrePost
3 Part B - Implicit analysis

3.3 Post processing

Open the d3plot in LS-PrePost. Start to press the Forward button in Animate toolbar to see what
happens with the specimen.

The time step in this implicit analysis was set to 5e-4 s, which is larger than the time between the
output states in ELOUT (5e-6). Therefore, elout data will only be obtained every 5e-4 second.

Plot the Effective stress (v-m) from Post > ASCII > elout* as for Part A of the exercise.

Save the plot, name it Stress_imp.

The first time we obtain any data is after 5e-4 seconds (if initial state is ignored), visualized with a
red circle in the figure. We will therefore have no information about what happens between t=0 and
t=5e-4.

Now compare the results with the result from the explicit analysis:

• Click Load > Add File (XY Data) and find the curve file Stress that we created earlier.
• Select Stress and click Select.
• Note the difference at which time we reach the yield limit for the two curves. This is an
effect of that the explicit analysis uses a smaller time step size.
• Close the PlotWindow.

11 LS-DYNA / LS-PrePost
3 Part B - Implicit analysis

Now, plot and save the strain vs. time curve:

• Plot the Yield Value (eps) and save the curve to Strain_imp.
• Close PlotWindow and ASCII.
Finally create the stress vs. strain curve for the element and compare it that from the explicit analysis:

• Create an XY-plot of Strain_imp and Stress_imp.

• Click Load > Add File (XY Data) and add the curve Explicit_curve in the PlotWindow.
• Select Explicit_curve and click Select.
• We now know that the curves differ in the beginning of the simulation due to the time step.
• After the first time step, the curves are very similar, this is because we use linear hardening
in the material model.

12 LS-DYNA / LS-PrePost
4 Comments

4 Comments
In this exercise, you have performed a simulation of a tensile test both with explicit and implicit time
integration. It is always up to the user to decide whether the results are good enough. An
experimental tensile test is a slow process and can be considered as quasi-static i.e. not affected by
mass inertia. How much can the load speed be scaled while still maintaining minimal mass inertia
effects? How much kinetic energy can be accepted allowed in the analysis? It is difficult to give a
general answer and always up to the user to decide.

4.1 Explicit versus implicit

Explicit schemes give the configuration at time 𝑡𝑛+1 as an explicit function of earlier configurations.

𝑥𝑛+1 = 𝑓(𝑥𝑛 , 𝑥𝑛−1 , … , 𝑡𝑛+1 , 𝑡𝑛 , … )

Implicit schemes give the configuration at time 𝑡𝑛+1 as an implicit function of the unknown as well
as earlier configurations, i.e. solves an equation system.

𝑥𝑛+1 = 𝑓(𝑥𝑛+1 , 𝑥𝑛 , 𝑥𝑛−1 , … , 𝑡𝑛+1 , 𝑡𝑛 , … )

Explicit Implicit

• Computationally more efficient since • Solves an equation system at every

nodal accelerations are solved time step
directly and not iteratively • Cost of every time step is unknown,
• Conditionally stable, requires small due to variations in convergence
time steps. behavior
• Memory efficient • Unconditionally stable for linear
• Good for dynamic problems, e.g. problems, which means that there is
impact and shock problems no limit to the size of the time
increment
• Implicit time steps are generally
several orders of magnitude larger
than explicit time steps.

5 Optional exercises
4. Decrease the initial time step size in CONTROL_IMPLICIT_GENERAL to see how this
affects the results.
5. Refine the mesh. (EleTol > EleEdit > Split/Merge). Does this affect the results? (If the
simulation time becomes too long, decrease the termination time)

13 LS-DYNA / LS-PrePost