关于两个大应变假定的区别的问题？

xiaomm09

小弟最近在使用软件的过程遇到两个大应变的处理方法：不知道具体问题中该选择哪一种，update lagrangian Hencky(ULH)
update lagrangian Jaumann(ULJ)
这两种算法各有什么应用范围或对象要求?
有没有相关的参考文献？谢谢，欢迎讨论。

tnt64

interesting, can you update several representative papers or documents about them?

tonnyw

1# xiaomm09

Jaumann can deliver erroneous results. In my experience, always choose Hencky formulation.

http://web.mit.edu/kjb/www/Publications_Prior_to_1998/Studies_of_Finite_Element_Procedures_Stress_Solution_of_a_Closed_Elastic_Strain_Path_with_Stretching_and_Shearing_using_the_Updated_Lagrangian_Jaumann_Formulation.pdf


You might set up a two-dimensional plane stress model and calculate J-integral using two formulations and you will see Jaumman results are wrong.

hillyuan

It is too strict to say Jaumann stress rate is wrong as you can also consider that the constitutive relation you used with Jaumman formulation is wrong. Jaumann rate is now become notorious because it gives rise to some physically unacceptable results. Hencky formulation seems much better. But in what means can you say Hencky is the right one?

Jaumann needs much less computation cost than Hencky. So if deformation is not that large, Jaumman is one choice although Hencky is the safty one.

tonnyw

It is too strict to say Jaumann stress rate is wrong as you can also consider that the constitutive relation you used with Jaumman formulation is wrong. Jaumann rate is now become notorious because it ...
hillyuan 发表于 2011-1-10 21:39

I agree with you. Even for large deformation, you can still use Jaumann formulation. But the load step size has to be very small.

hillyuan

5# tonnyw

If you consider Truesdall rate or Lie derivative of Kichhoff stress is the right one, Jaumman stress rate delivers quite different result in large deformation problems becauuse it does not consider the influence of deformation rate.

xiaomm09

多谢两位参与讨论：
To tonnyw :
“Jaumann can deliver erroneous results. In my experience, always choose Hencky formulation”，根据ADINA手册的解释：Large Strain Formulation [UL-FORMULATION]
Default means that the Updated Lagrangian Jaumann (ULJ) formulation is used if explicit transient dynamic analysis OR version 8.3 rigid-target contact algorithm is used. Otherwise, the Updated Lagrangian Hencky (ULH) formulation is used.

To hillyuan:
“Jaumann needs much less computation cost than Hencky. So if deformation is not that large, Jaumman is one choice although Hencky is the safty one.”应该是要考虑计算消耗，对于一般问题采用ULJ然后根据模型变形特征需要选择ULH，不知道我的理解对不？

tonnyw

多谢两位参与讨论：
To tonnyw :
“Jaumann can deliver erroneous results. In my experience, always choose Hencky formulation”，根据ADINA手册的解释：Large Strain Formulation
Default means that the U ...
xiaomm09 发表于 2011-1-11 10:29

That's what I am talking about. If you want to use Jaumann formulation, the load step size has to be very small which is the case in Explicit method. For Implicit method, if the load step size is too small, the computational cost will be prohibitive, which makes Hencky formulation a good option.

bbssbb

楼上哪位大虾普及一下，为什么本构关系用焦曼应力率表达会有问题。
问题在哪里？