you are here:   / News & Insights / Engineering Advantage Blog / Why Preload Parts for Explicit Dynamics Analysis, and How – Part 2

Engineering Advantage

Why Preload Parts for Explicit Dynamics Analysis, and How – Part 2

April 26, 2016 By: Steven Hale

In Part 1 of this post, I discussed why preload analyses are needed, as well as three different methods for performing a preload analysis.  I also mentioned that it is important to verify that a steady-state condition is obtained before initiating the transient dynamic phase of the analysis. In this post, I’ll discuss how to verify that a steady-state condition has been obtained.

The best way to confirm that a steady-state condition has been reached is to run the transient dynamic phase of the analysis without applying any additional dynamic loads. Results should then be reviewed to confirm that there is no significant variation in the solution with time. This can be confirmed by plotting critical results such as energy, stress, velocity, and/or displacement and checking that there is little to no variation with time.

As an example, a simple blade model with shell elements was solved both with a preload phase and without. The preload phase was performed as a static analysis using an implicit solver (Option 1 described in Part 1 of this post). A rotational velocity was applied to the blade in the preload phase and as an initial velocity for the transient phase. The blade was then rotated a full 90 degrees in the transient phase without any additional applied loads. Figure 1 shows the variation in Von-Mises stress at two elements in the blade during the 90-degree rotation of the transient phase.

Figure 1: Von-Mises Stress History in Two Elements of the Blade During a 90-degree Rotation

The graph on the left shows the stress history for the case where a preload phase was not applied, and the graph on the right shows the stress history in the same elements for the case where a preload phase was correctly applied.The near-zero variation in stress shown on the right indicates that a steady-state condition was obtained for the case that included the preload phase. Comparing the two graphs suggests that the preload solution prevented the large stress oscillations seen in the no-preload case. Under conditions where a preload was applied but had not reached a steady-state, smaller stress oscillations relative to the no-preload case might be seen, but they could still be significant.

Figure 2 shows the stress solution in the blade at the zero and 90-degree rotation angles for the case that included the preload phase.The zero-degree solution represents the end of the preload phase and the start of the transient phase. Almost no variation in the stress contours is detectable between these two angles, confirming that a steady-state condition was reached.

Figure 2: Von-Mises Stress at Zero and 90-degree Rotation

To conclude, a preload analysis should be included prior to running a transient dynamic analysis whenever steady-state conditions generate stress, strain, and deformation in parts or assemblies that are then subject to dynamic loads. There are several different methods for performing the preload phase of an analysis, but in all cases, results should be checked to confirm that a steady-state condition has been obtained prior to running the transient analysis. This is critical to prevent artificial stress oscillations and to obtain accurate results in the transient analysis.

Please let me know if you have any experience or insight into methods for confirming that a steady-state condition has been reached at the end of a preload analysis. I look forward to your feedback!