Comparison of improved shell properties with a base simulation on a crash beam using HyperMesh.

Objective:

1. The analysis runtime (Tstop) is changed to 55 ms.

2. The number of animation steps is decided as 45 given the minimum is 25 and maximum as 60. And the Tfreq is updated as 1.22 ms (Tfreq = Runtime / No.of Animation Steps)

The simulation is executed without any further changes, apart from the above the two changes in the respective cards under engine_keywords section. The analysis run file is renamed as Base_Element_Formulation-Shell-3_assignment_0000.rad and stored in a new directory named, "Base Simulation" .

Once the simulation is complete, open the Base_Element_Formulation-Shell-3_assignment_0000.OUT and check that the energy error and mass error values are less than 15%.

Energy error is at -10.3% which is acceptable.

5. The energy and rigid wall force values from the simulation results are plotted with respect to time. To plot the graph, open a new split window layout and open a Hypergraph2D session and open the output file Base_Element_Formulation-Shell-3_assignmentT01. Then, from the options deck, we can select the appropriate parameters to be plotted in the graph.

Now again another window layout can be opened or new plot can be created to plot the rigid wall force values.

To view the executed simulation, open HyperView in a new window layout. Open the Base_Element_Formulation-Shell-3_assignment.h3d file and select any field variable, to get its contour. The snap shots of the displacement and Von-Mises stress contour plots of the analysis are added below.

6. The properties of the meshed shell elements are changed to the recommended settings. Then, the file name is changed to Updated_Element_Formulation-Shell-3_assignment_0000.rad and is saved in a new directory "Updated Simulation". Now, again the similar procedure is followed to run the simulation and to plot the graphs.

Energy error is at -0.1% which is acceptable.

Conclusion:

A comparison table chart is drawn between the results of the both base and updated simulations.

9. In base simulation, the shell elements are assigned default properties of Ishell=1 (Q4), in which the hour glass effect is mitigated using viscous-elastic damping (Perturbation method). As a result, a damping effect is applied on the nodes, and this damping will slow down the deformation due to hour-glassing. This mode of reducing hour-glassing comes into effect, once the model detects the occurence or onset of hour-glassing, so the hour-glass energy will shoot.

But if hour-glassing happens, the physical model will try to counter it and as a result the total energy of the model will decrease, as in the case of our base simulation.

On switching to Ishell=24 (QEPH), an arbitrary stiffness in incorporated at the material level to counter the hour-glass deformations. So, the simulation had zero hour-glass energy, also the total energy of the model remained almost constant, which is expected in a good analysis.

Also, in the Rigid-wall resultant forces plot, a second surge in values is observed towards the end of the run time, after using the improved shell properties.