Is it correct to change the mass density in some explicit analysis?

[ab110]

Hi,

I want to run an explicit skin pass roll forming analysis. As you can see in the attached file, the sheet passes through two rigid rolls and the thickness slightly decreases. I have checked the model and everything is correct. But when you run the model, it continues till time 4.01 and after that there is nothing (no convergence). The total time of analysis is 9.01. Does anyone know how can I solve this problem?

I tried to solve the problem by increasing the material density and considering 7.8e-6 instead of 7.8e-9 in this analysis. When I changed the density, the model run completely. But I do not know is it correct to change the density of material in this explicit analysis or not? I want to measure the residual stress on sheet in this process. If it is not correct, what is the problem of this model?

Cheers,
A

 

[MechIrl]

If you are not interested in capturing dynamic effects (i.e. a quasi-static analysis) then mass scaling is fine. Normally in that case a condition is imposed that kinetic energy remains below 5% of internal energy. Might be worth trying to figure out why the first density value isn't working too though, are you sure your assumed units are consistent?

 

[ab110]

Yes I am sure that the units are consistent. I can not understand why the model does not work with correct density.

 

[MechIrl]

Just looked at your model there...you have E and YS defined in MPa so I presume your base unit is mm. This would mean your mass density is in kg/mm3 which would mean 7.8e-6 would be a more appropriate value.

 

[corus]

Following on from what MechID said, the units for density is tonnes/mm^3 in a dynamic analysis, ie. 7.8e-9 is correct for steel without any mass-scaling.

Tara

 

[ab110]

As course mentioned, because the units of this analysis are mm, N, Mpa, Sec. So, the unit of density is ton/mm3 and for steel it should be 7.8e-9.

 

[MechIrl]

Apologies...yes, tonnes..

 

[ab110]

So, the units are correct. Anyone doe not have a solution for this problem?

 

[corus]

It is common practice to use mass scaling in 'rolling' type models in order to speed up the analysis. You need to check your results against some other measure to have confidence in the results, ie. with real life.

Tara

 

[paperboy84]

I have the same problem with my simulation. When I ran the simulation with the mass density 7.8e-9 my simulation doesn't complete. How did you resolve this problem?

 

[ab110]

Hi, I just increase the mass density. you can use 7.8e-6 instead of 7.8e-9.