Transient Stress Direct Inegration

Question

Transient Stress Direct Inegration

inaprocreate (Mechanical)

(OP)

29 Nov 07 12:44

I am attempting to analyze a small structure under a g-force acceleration on one axis, to determine the stresses developed in the parts during a sudden collision. I only have the linear package, so no MES.

I have been modeling this structure as a linear Transient Stress Direct Integration, using the translational ground motion load curves that can be applied to the model.

Initially I have the analysis set at 1 step with a step size of only 0.01 seconds, and load curves that reach the desired acceleration at 0.01 seconds.

What I am trying to clarify is if this is the correct analysis type to apply a g-force load.

Thank you in advance for your help.

GBor (Mechanical) · Answer 1 · 2007-11-29 13:25:53

Sounds like the answer is, "No".  And, if it is the right analysis, then your load curve is completely wrong.

I would suggest that you start by simply running a static analysis using a gravity multiplier along the axis of interest.  For instance, if in your load curve, your g-force reaches 3g's and you want it along your global x-axis, run a static analysis with gravity set to the standard and use a multiplier of 3 in the x-axis line of the gravity definition screen.

An alternative is to set gravity to standard, place the gravity vector to 1 in the x direction, and in the load case multipliers, put a 3 under acceleration/gravity.

With any dynamic analysis, you need to have multiple time steps within the analysis, so if your load peaks at 0.01 seconds, you need a time step of probably 0.001 so that you get at least 10 steps in the analysis.  I dare say that this won't be enough for quality convergence, but it's a start.

Garland

Garland E. Borowski, PE
Borowski Engineering & Analytical Services, Inc.
Lower Alabama SolidWorks Users Group
Magnitude The Finite Element Analysis Magazine for the Engineering Community

inaprocreate (Mechanical) · Answer 2 · 2007-11-29 15:54:27

Thank you for your help. Once I wrote the post, I realized that the way I was doing this was incorrect.

I am in the process of analyzing the case in a static load with the gravity multiplier as you suggested, decreasing the mesh size until the results converge.
Then I plan to compare those results to the results of a Transient Stress Direct Integration with 10 steps, increasing the number of steps until those results converge.

I would imagine that the two scenarios should have similar results.

Again, thank you for your prompt response.

GBor (Mechanical) · Answer 3 · 2007-11-29 16:29:31

Your results may differ by as much as 30%. If you extend your load curve out with a 3g load for an extended period of time so that all the "transient" characteristics settle out, you should then have something similar to the 3g "steady state" or static load. There is a reason why dynamic analyses are run, so don't expect it to be too similar to the linear static run.

Come Join Us!

Posting Guidelines

Transient Stress Direct Inegration