Sheet Metal Forming Analysis

[ekline]

My background is in linear structural analysis. I do have access to NEiNastran non-linear, and someone just asked me to look at a deep drawing operation and figure out why some cookware is coming out of the die warped. My instinct is that deep drawing of sheet metal requires a specialized analysis. I imagine it could be done in NEiNastran, but I thought this type of analysis was normally done in Abaqus or one of the specialized metal forming FEA packages. I will likely refer them to one of the consultants at Abaqus. Any input?

 

[rb1957]

yes, i agree with you that this is specialised FEA, and not for the inexperienced. just think of the plastic flows

however, i'd ask why is this production problem happening now ? presumably, these parts have been in production for a while ... isn't deep drawing a process that is "tweeked" through prototyping ?

 

[Drej]

If your cookware can be idealised as 2D i.e. plane strain, axi-symmetric, you could solve it as a deep drawn item using an implicit package (i.e. NASTRAN, ANSYS, ABAQUS/Standard) within a reasonable amount of time. It is probably a specialised analysis if you need anything more than this, an explicit code something like ABAQUS/Explicit, DEFORM or DYNA would be ideal for example.


------------
See faq569-1083 for details on how to make best use of Eng-Tips.com

 

[GBor]

There is a book entitled, "Metal Forming Analysis", by Wagoner and Chenot if you just want some information on this. I agree with everyone else...it is a very specialized, highly non-linear analysis. It probably could be done in NENastran, but the time it would take and the computer to solve it may be a bit more than you care to attempt.

 

[corus]

I wouldn't pass this on to anyone as next time you'll be just passed by too. It should be a simple problem of imposing contact from a moving rigid body on to a deformable body that is restrained by a third fixed rigid body. Approximate the part in 2D first to see how the analysis runs. The only difference from a linear analyis is that you'll put in a value for yield, have to define contact between surfaces, and use large displacement. A piece of cake.

corus

 

[Drej]

I think this type of analysis is anything but a piece of cake. For a seasoned non-linear FE professional, they could hit the ground running, but if you have little knowledge (and this guy's background is linear FE) then I think it would take time to get up to speed on the ins and outs of this type of analysis. Lots and lots of factors to account for and issues to deal with: non-linear geometry, non-linear contact, plasticity, friction, springback, accurate material data, computer resources, geometry of the components... the list is endless to begin with. And to boot I bet there's a tight manufacturing schedule on this as well.


------------
See faq569-1083 for details on how to make best use of Eng-Tips.com

 

[rkn40]

Deep Drawing is a highly non-linear phenomena. You will be working in the plastic portion of the material curve. Also, you have to understand the metal forming process itself(blank holding pressure, friction between die elements etc.) to provide proper inputs. However there are services which can do these analyses for you. You may start here

http://www.feadomain.com/e107_plugins/custompages/Metalforming.php

 

[analyst9]

One cannot ignore the effect of crystallographic texture in sheet forming processes. Crystallographic texture has a significant effect on warping.One should use a UMAT that incorporates crystal plasticity. I am working on this. Probably you can contact me after a couple of months, once I get my UMAT working !!

 

[Drej]

> One cannot ignore the effect of crystallographic texture in sheet forming processes.

Never heard of it. In industry I bet though that probably 90% of people involved in the modelling of this type of work (sheet forming) do not account for this crystallographic texture you mention, principally because of the UMAT development cost, which is astronomic.


------------
See faq569-1083 for details on how to make best use of Eng-Tips.com

 

[ekline]

Drej, isn't that how analysis? People treat most systems like they are linear, because... well... that's what they can analyze! I guess everything is linear given a small enough range and everything is non-linear given a large enough range. The most important part of an analysis is understanding its limitations.

 

[gwolf]

One can quite safely ignore the crystallograpic texture in sheet forming processes. This is because the effect is small compared to the other factors like trying to put too much plastic strain into a sheet with no lubrication. Fix that and the grain stuff pales into insignificance. Good quality sheet has good micro-structure, bad sheet has bad micro-structure.

Over-engineering whilst ignoring the simple big stuff is to be avoided.

I could say the same of high strain fatigue and fracture. I have yet to find a component, which cracks, which can't be fixed by re-design in such a way that the stresses become so low that you can throw detailed fatigue calcs out of the window.

 

[gwolf]

....... appart from thermal cracking in power station cooling pipes.

 

[analyst9]

"Good sheet has good microstructure".....yeah initial microstructure might be good, but microstructure continuously evolves during plastic deformation.As the microstructure changes during deformation, so does the mechanical properties of the material and hence influences the forming of a sheet.

Well, in industry people normally don't take texture into account as it is computationally prohibitive and much complex topic to understand. Moreover, texture causes anisotropy and dealing anisotropy in the analysis(keep in mind that it continuously changes during deformation) is not a trivial task.
The above info may not be useful to get immediate results for the problem at hand. Anyways just my 2 cents...