Skip to main content
R
rrakos1-Visitor
1-Visitor
March 31, 2017
Question

Plate bending

  • March 31, 2017
  • 20 replies
  • 11269 views

Hello,

I have a returning problem. I often have to simulate the process of different metal plate bending. There are a lot of cases when these processes include more steps. (Bending 1, Bending 2..Bending X). As far as I know, the material will include a residual stress after all bendings (in case of sigma > yield stress). So the result of current bending will depend on the residual stress field generated by previous bending step. Based on this, the simulation (result of STEP 2) will be correct and exact if the STEP 1 also were simulated.

residual_stress.png

At the moment I prepare the STEP 1 geometry (bended already) in Creo and a simulate only the STEP 2. Consideration of all STEPs would be too complicated and long. But I think this solution is not correct becasue the STEP 2 starts fomr a state which doesn't include the residual stresses.

- How can I consider the residual stress from STEP 1 with more simple method?

- Or how can I separate the STEPs? For example; Analysis 1: STEP 1 only than Analysis 2: STEP 2 only initialized from result of Analysis 1.

Thanks a lot:

Roland

20 replies

skunks
skunks19-Tanzanite
19-Tanzanite
April 1, 2017

sorry

it's not for simulate (mechanica) possible

ANSYS etc

regards

paul

skunks
skunks19-Tanzanite
19-Tanzanite
April 3, 2017

2d-example attached

regards

paul

skunks
skunks19-Tanzanite
19-Tanzanite
April 3, 2017

color plot:

plot_steps.jpg

R
rrakos1-VisitorAuthor
1-Visitor
April 3, 2017

Hello Paul,

it is OK but in your case, the STEP 2 is the bending tool back. In my case, an other tool (from other direction) should bend - initialized from the STEP 1.

Actually, it is possible to define the STEP 1 and 2 together in one analysis but the full simulation with 2 (or more) STEPs is too complicated. It would be easier if one simulation case includes one bending process. In addition, if the STEP x diverges then the STEP x-1 should not be started again...

1
1things1-Visitor
1-Visitor
April 5, 2017

The process of bending in the industry is basically a forming process where a sheet material is put within a die and then the upper side is pressed so that it can be bent well. This design step for bending has made easy operations as compared to Catia & Autocad..I loved the schooldays which was not containing anything of this.It was just a pure enjoyment.

R
rrakos1-VisitorAuthor
1-Visitor
May 4, 2017

I have created some simulation to investigate the behaviour of material (stress distribution) after the reach of yield stress. (My material model is defined by a measurement)

I have a simple measurement with a small part. After the bending of this part, I investiged with microscope the outside surface of part. I see the the material - along the bending radius - reach the yield stress. But the part was not broken...so I think that the stress doesnt reach the tensile stress. Insipte of this, in the simulation, as it can be seen below, the stress is in red area in full cross section along the bending radius....colored by red means stress is over tensile stress.

It seems to me that creo solver estimates over the stress distribution. It shows that the part will break but it did not happen in the course of my measurement.

What is the reason in your opinion?

Képkivágás.PNG

Thanks

Roland

skunks
skunks19-Tanzanite
19-Tanzanite
May 4, 2017

hello Roland

i used perfect plasticity:

plasticity.JPG

technical curve is not real:

true_curve.JPG

regards

paul

R
rrakos1-VisitorAuthor
1-Visitor
May 5, 2017

OK, I tried the perfect plasticity model, the stress value did not reach the ultimate stress, but actually it is not possible caused by the perfect plastisity model:

perfect_plast.PNG

If the stress value is higher than yield stress then it will not increase so it is evident that the stress doesnt reach the ultimate stress... In my opinion, this mode is ok for special materials but in my case it is not ok. I have a measured material curve and it is not similar to the perfect mat. model. The best case was the definition with the exponential law but it leads to too high stress value.... I dont understand it....

Terms & ConditionsAccessibility statement