Residual stresses as result of cold-deforming

[gizmo85]

Hello,

I'm doing a static analysis of a cold-deformed L-shape (90 degree) wire. As a first step I did the analysis with the final shape of the deformed wire by simply meshing it and appling the force.
Because of the manufacturing process (cold-deforming, bending) there are some residual stresses through the wire.
My question is: How to take into account these residual stresses in the wire and how to calculate them?

Thank you,
Peter

 

[jorton]

I got into this once before. Please see the link below. I never really got my head around it. I hope you get some good answers here.

http://www.eng-tips.com/viewthread.cfm?qid=201635&page=155

 

[corus]

I wouldn't know how to calculate the residual stresses by hand, and would probably do it by using non-linear finite element methods as you'd have a change in section shape with the large displacements. I would suspect you'd get residual stresses though as the inner and outer radii must produce different stress concentration factors and hence produce a non-linear stress distribution through the thickness. I'd guess that on the inner radius you'd be left with a residual tensile stress as this region would see the greater strain and the stiffer outer region would dominate.

Residual stresses are often present in structural sections, often purely from the uneven cooling process after the section has been formed. Generally these tend to be ignored though, except for assessing for fatigue damage. My guess (yet another one) is that there is sufficient leeway in allowable stresses in design codes to allow for these 'unknowns'.

corus

 

[gwolf2]

Trickey, but interesting. The obvious solution is to just plug a stress-strain curve into an FE code and run a NL analysis to wrap the wire around a rigid surface radius.

I think you could probably do something with inner and outer bend radii, M/I=S/y=E/r, a Ramberg Osgood fit on your stress/strain, plus a consideration of bend radius and springback radius.

I did such a calc last year where the objective was to calculate the springback radius for a given bend radius. My metal was thin sheet which was quite a lot easier than wire. My material data was not very good and there was a 40% error in the predicted final radius vs actual (I made a number of test pieces). Unless you have good material data this will be difficult.

FE is probably the most accurate way.

gwolf

 

[truckcab]

Hi,
for some years ago I found an article about how to approximate residual stress and strain hardening arising due to bending of sheet metal. I cannot give you the exact reference since I simply dont remember, I am, however, sure that the name of the author was Sivakumaran. I guess if you check Science direct or something like that you will find the article.





Live Long and Prosper !