Try with elemental initial strains, if available in your FE package... :-)

Thanks for the input Rahul.

But for the particular stress distribution from the literature , do I make plots and then insert it linearly somehow or do I input it as a continuous function ? A little detail would be helpful.

Hi,

In the following PhD thesis, they provided a simply way to define residual stresses.

STRUCTURAL BEHAVIOUR OF STEEL COLUMNS AND

STEEL-CONCRETE COMPOSITE GIRDERS

RETROFITTED USING CFRP

Dear Majid

Thanks for your time. I have hereby attached pics of the relevant information from the thesis which you have referred earlier, But I cannot seem to find the significance to my application. A little more detail would be appreciated , as I want to work around this problem.My residual stress distribution seems more complex. What does the author mean by shifting the origin axes of stress strain curves. How can I achieve that

Thanks

I had actually obtained residual stress distribution like yours (but one, without those sharp edges) when doing a plates' edge-to-edge butt welding simulation -- liquid metal solidification -- in FEA. And therefore I do not have much idea as to how to reproduce it via initial strains.

Your recent picture attached appears to be having residual stresses due to bending beyond yield stresses. In that case a non-linear analysis alone should do.

Anyway, coming back to application of Initial Strains: Irrespective of whether the initial strains abruptly changes from one element to its next, or varies gradually, finally there will be a redistribution of the initial strains, and consequently, the stresses too. So I presume any calculation on paper, regarding initial strains to be provided, will get a final re-distribution. If there's metal yielding (and the stresses crosses the yield stress) with material non-linearity options on, may be they will, the final stresses will be a smoothened out one.

So do the calculations and provide, but do expect a trial and error procedure in the process.

Dear Rahul

It seems I have to apply the inistate command using function based approach. there is not much literature available as per the steps involved. I seem to apply inistate in the following order , but this is not working. Can you please point out the error in the steps.:

INISTATE,SET,DATA,FUNC

INISTATE,SET,DTYP,STRE

INISTATE,DEFINE,,,,,LINY,YIELD_FLANGE,YIELD_FLANGE/WIDTH_FLANGE

With Regards

Inaam

I'm sorry I've never used the INISTATE command of ANSYS. Once when I provided initial strain, I did it via the R command. I did read some material on INISTATE, didn't get on well, but realised it is the way to go in your case...