Learning Forum

[Theo211]

Hello,

I am currently using a static structural analysis with Mechanical APDL as solver to calculate a metal bending process.

In this process a punch is pushing a metal sheet into a die. After releasing and forwarding the sheet another punch action is performed until a constant radius across the work piece is achieved. The contacts are defined as frictional (nonlinear), which are partly applied and released in each bending cycle.

When researching on LS-DYNA I found out that it is also useful for contact problem calculations and one of the core-competency is "highly nonlinear transient dynamic finite element analysis (FEA) using explicit time integration".

When using LS-DYNA instead of static structural analysis with Mechanical APDL:

Question #1: What are the differences regarding the calculation time (what solver is faster) ?

Question #2: Are there some other advantages of LS-DYNA?

[Saumadeep Choudhury]

To understand the difference between implicit and explicit time integration methods have a look at the following course - Time Integration - ANSYS Innovation Courses
To answer your second question, it depends on what you are trying to do as they both have different strengths. To get a better idea about LS-DYNA you can refer to the following link - Chapter 1: Introduction to LS-DYNA (ansys.com)
Regards Saumadeep

[Ram Gopisetti]

Hi LSDYNA and MAPDL has their tie-offs, Coming to your specific questions, its forming kind of simulation and dyna has some special type of contacts such as FOMING_SURF,NODE,ONEWAY TO SURF,NODE,SMOOTH,MOTOR which can handle forming simulation where the contact plays a crucial role ( check CONTACT DAMPING and 11-106 (CONTACT) in the dyna manual) and you can automate the forming by using FORMING application in prepost.
The answers for your questions are as follows :
1) DYNA solvers are faster than the MAPDL given the type of problem ( implicit might take time and need up to date solvers and often may require MPP)
2)You need to be specific to a topic for this question.
cheers, Ram

[Theo211]

Sorry for the late answer. To specify the problem further:
In the applied bending procedure the plates have an initial state of strain/stress. And therefore have to be meshed locally with a fine mesh of solid elements in some areas in order to allow to introduce initial high stress gradients over small lengths.
Since LS-Dyna might be good/fast in handling shell elements I am not sure about handling a lot of solid elements.

[Ram Gopisetti]

what do you mean by " I am not sure about handling a lot of solid elements"? can be specific in what terms you are not sue?
Cheers. Ram

[Theo211]

a collegue told me that LS Dyna is able to calculate 2D shell elements faster than the Mechanical solver.
But he is not sure about the LS Dyna solver calculation speed of solid elements that we need for our analysis. We don't know which of both solvers is able to calculate faster with solid elements in our example.

[Ram Gopisetti]

LSDYNA has similar capabilities on solving the solids as well.
Cheers, Ram

[vikaskaul]

It is very hard to do a fair comparison between two codes that have tens of thousands of features in them with particular strengths in a few key industries. In the tests that I have run, LS-DYNA was a beast when it came to solving physics. Both implicit and explicit codes are nothing short of amazing. However, there are no free lunches: LS-DYNA documentation, as an example, doesn't seem to work for everyone.
So, I would recommend doing a comparison between MAPDL and LS-DYNA (with the help of an LS-DYNA expert) on a few relevant problems but make sure you stage those out starting from very simple to something very realistic - if your ultimate intent is to add LS-DYNA to your toolkit. Alternatively, you could simply run LS-DYNA jobs that are available on dynalook.com from an implicit comparison between LS-DYNA and NASTRAN and recreate those jobs in MAPDL and see how it compares.