Altair RISE

A program to recognize and reward our most engaged community members

Nominate Yourself Now!

3sigma in Optistruct random fatigue analysis

Hello to everybody,

I found the way to perform a fatigue analysis based on a random input.

How can I impose to optistruct to use a 3sigma factor on the RMS value at the end of the random analysis to evaluate the damage level and life of the component during the fatigue analysis?

I try to explain me better, at the end of random analysis I can obtain a certain value of the RMS stress. I usually multiply this value by 3 to obtain the 3sigma value to compare with the material fatigue limit. How can make this operation in optistruct, so that the software use the 3*RMS stress value to perform the fatigue analysis?

Thank you

Find more posts tagged with

English

OptiStruct

Accepted answers

Adriano Koga_20259

Thank you, but I don't need to perform an optimization, so I cannot give a constraint on stress. I only need to perform a fatigue analysis using the results of a random analysis.

It is only one calculation, optistruct perform the random analysis and in the same run the fatigue analysis. When it perform the fatigue analysis I suppose it can take the RMS Stress values from the random anaysis, I'd like it could be take the RMS value multiplied 3.

I Hope to have clarified my question.

Hi Danillo,

OptiStruct can perform Random Response Fatigue directly. Take a look at this tutorial in OS help.

The solver itself will take into account the 1Sigma, 2Sgima, 3Sigma values accordingly, depending on what assessment method you choose (Dirlik, Lalanne, etc...). So you don't need to worry about this multiplication factor.

Now, If you want to compute by yourself, unsing Miner's for example, having the RMS stress value, in HV you can apply a simple multiplier in your legend to get the 3 Sigma.

OS-T_ 6060 Random Response Fatigue Analysis.pdf

Random Response Fatigue Analysis.pdf

All comments

Adriano Koga_20259

Hi, If you're performing optimization with OptiStruct, you can request RMS Stress as a response and add a constraint to it using 1/3 of your stress limit. Or also you could run some fatigue optimization, adding a fatigue constraint. If this just a matter of post processing you could go to HV and add a multiplier to your rms stress, so that you see 3sigma.

Danilo_21605

Hi, If you're performing optimization with OptiStruct, you can request RMS Stress as a response and add a constraint to it using 1/3 of your stress limit. Or also you could run some fatigue optimization, adding a fatigue constraint. If this just a matter of post processing you could go to HV and add a multiplier to your rms stress, so that you see 3sigma.

Thank you, but I don't need to perform an optimization, so I cannot give a constraint on stress. I only need to perform a fatigue analysis using the results of a random analysis.

It is only one calculation, optistruct perform the random analysis and in the same run the fatigue analysis. When it perform the fatigue analysis I suppose it can take the RMS Stress values from the random anaysis, I'd like it could be take the RMS value multiplied 3.

I Hope to have clarified my question.

Adriano Koga_20259

Thank you, but I don't need to perform an optimization, so I cannot give a constraint on stress. I only need to perform a fatigue analysis using the results of a random analysis.

It is only one calculation, optistruct perform the random analysis and in the same run the fatigue analysis. When it perform the fatigue analysis I suppose it can take the RMS Stress values from the random anaysis, I'd like it could be take the RMS value multiplied 3.

I Hope to have clarified my question.

Hi Danillo,

OptiStruct can perform Random Response Fatigue directly. Take a look at this tutorial in OS help.

Now, If you want to compute by yourself, unsing Miner's for example, having the RMS stress value, in HV you can apply a simple multiplier in your legend to get the 3 Sigma.

OS-T_ 6060 Random Response Fatigue Analysis.pdf

Random Response Fatigue Analysis.pdf

Danilo_21605

OK, thank you