Week 11 Car Crash simulation
OBJECTIVE
To set up a simulation and carry out an explicit dynamic analysis of a car body crashing into a wall. A parametric study is to be carried out between the outputs of three different thicknesses of the car body. The outputs to be compared are total deformation and equivalent stresses acting on the model.
MODEL IMAGE
PROCEDURE
1. After opening ANSYS Workbench, we are met with the Project Schematic window. Here, we can select the 'Explicit Dynamics' analysis system on the left. Doing so creates a new project. Here, we can rename the project and also input the material. We will need to right-click 'Geometry' and select 'import'. The file provided for this project should be selected.
Now, we can add a material for the analysis. To do that, we need to double-click Engineering Data'. This opens up the list of inserted materials. We can then pick materials we need from the repository listed in the Engineering Data Sources. This challenge requires the use of Stainless Steel NL, which is available in the non-linear material source.
After that, we simply need to click the yellow '+' symbol on the material's corresponding 'add' column to add this specific material to the project.
Once we are done, we can simply close the tab.
We can then exit out of the engineering data tab and return to the project schematic window, where we can right-click geometry and select 'edit'. This will bring the model up in the Mechanical interface.
2. In the mechanical interface, in the outline, under geometry, we can rename each of the components. We also need to assign Stainless Steel NL material to the car body. A thickness needs to be assigned to the car body since it's a midsurface (0.3mm here). In addition to that, we need to check the parametric study box for thickness denoted in green in the following screenshot.
Parts other than the wall and the car body can be suppressed.
3. Then we can move to Connections > Contacts and delete all the existing contacts. We can leave body contacts as they are.
4. Moving on to meshing, we need to assign tetrahedron type meshing to the wall and triangle type mesh to the car body's midsurface. We can do this by right-clicking mesh > insert > method and assign the type to each body. This means we need to create two mesh methods.
In addition to this, we can select the regions on the car body that will most likely not be affected by the crash and assign a coarse sizing of 175mm to it.
Also, we can change the general mesh smoothing to low. This is to reduce simulation time (as well as the sizing done previously).
Final mesh:
5. We can assign the fixed support to the same regions we applied the mesh sizing on the car body i.e. surfaces that won't interact with the crash. (via right-clicking Explicit Dynamics > insert > fixed support).
Next, we need to right-click Explicit Dynamics > Insert > Displacement. This will be assigned to the surface of the wall facing the other direction of the car body. We can assign the z component a value of 500 mm and leave the other components as they are (unconstrained).
6. Moving on to the analysis settings, we shall be entering an end time of 0.001s as shown.
7. Also, we need to assign symmetry attributes to both bodies. To do that, we can go to model in the outline and right-click it > Insert > Symmetry. Then we need to right-click symmetry > select 'Symmetry Region'. In this menu, all we are to do are select the surfaces about which symmetry is to be applied, along with the axis that will be the symmetry normal. There will be two symmetry regions - one for the wall and the other for the car.
8. Now we can generate the outputs. To do this, we can right-click Solution > Insert > Stress > Equivalent (Von-Mises) (for stress) & right-click Solution > Insert > Deformation > Total (for total deformation). It is advisable to assign the stress output specifically for the car body to obtain useful results.
Now, we need to do right-click solution again and click 'Evaluate all results'.
When the analysis is done, we can view the results by simply clicking each of these solution entities we created, in the Outline menu. Then, we need to check the parametric checkbox for the maximum and average results of stress and total deformation, just as we did for thickness earlier.
9. We can then close the mechanical window and access the parameters section by double-clicking it from the project overview in Workbench.
Doing so brings up a table on the top-right. Here, we can enter additional thickness values to carry out the parametric study. I'll be analysing outputs for 0.6mm and 0.9mm thicknesses as well.
Now we need to select both attributes and right-click them. There should be an option to 'update design points'. Clicking it results in Ansys generating results for outputs requested as part of the parametric study.
Once that is done, we can see the table now shows values for stress and total deformation for each thickness value for our convenience, without having to set up and run simulations separately for each thickness.
OUTPUTS
TOTAL DEFORMATION
Maximum & Minimum deformation
EQUIVALENT STRESSES
STRESSES WITH WALL INCLUDED
Maximum & Minimum stresses
OBSERVATIONS
Since the wall was not given a rigid behaviour type, the stresses produced within it affected the stress outputs, so much so that the stresses generated in the car body were dwarfed by those generated within the wall. Hence, a separate stress output, measuring the stresses in the car body alone, was requested and generated. The maximum stress generated (in the parametric study) is the value of the aforementioned generated in the car body alone.
The following table is the result of the parametric study carried out for various thicknesses of the car body - 0.3 mm (original), 0.6 mm and 0.9 mm.
DP stands for design point and it denotes a particular parametric condition within the study. The comparison is between various design points of differing thicknesses.
The maximum and average values of the required outputs were included in the study. As we can see, the stresses produced in the car body increase with the thickness of the car body, the only anomaly being the maximum stress generated in design point 1 (0.6 mm). But the average stresses support this statement.
This also proves the increasing thickness improves the car body's capacity to absorb impact energy more. With that, the amount deformed also decreases, as seen with the average deformation values. They are almost similar but there is a slight downward trend from DP 0 to DP 2.
RESULT
The explicit dynamic analysis was carried out on a model of a car body crashing into a wall. The outputs of total deformation and equivalent stresses were generated and a parametric study comparing the outputs of the car body with varying thicknesses was carried out as well.