Yield surface in sheet metal bends

Sebastianmaklary

Hi all

So far, I have only calculated sheet metal structures with the assumption that the material properties are uniform, so the material in the bends have the same yield strength and ductility as the rest of the part. I know that this is not the case and I know that in such things as stamped sheet metal parts in cars, i is quite important to get right.
I know of Gestamped who spent many years developing a new lower control arm for a Toyota Yaris and even got SSAB to develop a new grade of steel in order to use the control arm as an energy absorbing structure in case of an impact on the front corner of the car. Among the challenging things in that project was that different places on the structure has yielded different amounts during the manufacturing process and thus have different yield strengths and different amounts of damage and remaining ductility.

Now, I'm designing a crash structure made of bent sheetmetal, which I'm analyzing in Mecway with OpenRadioss and I wonder if any of you know of a way to change the yield surface locally as if the part has already yielded some.
I can imagine that it would be possible, but very complicated and out of the scope of my project, to simulate the bending process of each bend individually in such a way that the parts end up in the right place and then having to figure out how to assign bonded contacts as an additional step, but I hope that something a bit less explicit is possible.

Do any of you have experience with this or know if it is even possible?

Best regards,
Sebastian

Fatmac

I'm not sure it gives u any directions in OR but perhaps there is a similar approach in OR as to the work around I have been playing with here. In calculix I have added a custom card that defines temperature dependant plastic strain/stress strain relationships then simple defined a different temperature for chosen elements elements. However, having looked at it after playing with it, I guess if not achieved much -I have achieved little more than simply applying different material properties for each element!

disla

Maybe you could approach the problem in the other direction. I mean to make the real piece to resemble to your stress-free initial design conditions model.
That is the so-called Post weld heat treatment (PWHT). ASME Secction VIII, División 1 provides guidance for thick plates , some material numbers or parts exceeding certain bending radius during the fabrication process. It’s common practice in pressure vessel design and if the piece is not too large is not an expensive process.

Sebastianmaklary

@disla I is not a welded piece but aluminum plates which are bent, bonded and riveted. I guess that the individual pieces could theoretically go through a heat treatment to anneal and reapply the T6 artificial aging after bending but it is not something that we have ever done as we just use the sheet metal as-is. I doubt that it would be considered a good idea in our company to add a process like that to the final product because of shortcomings in the FEA model.

@Fatmac I guess that I could do something similar and test how much it influences the part. By selecting the bend surfaces and adding them to a named selection, I can then extend that selection to the elements themselves and move them to a different component that has different material properties applied to increase the yield strength and decrease the failure strain.