Cancel
Showing results for 
Search instead for 
Did you mean: 

advanced explicit analysis (no -24) time consumption

Builder
Builder

Hello everyone,

 

I have run advanced explicit analysis (no -24) on last Friday. Unfortunately it is still running. My model consist 134842 elements.

 

I have chosen 100 number of steps with 0.01 time increments and 1 – output every nth step. In addition, in the output, I have just selected displacement, stress and forces.

 

 Does anybody have any idea how long dynamic explicit takes time when element are more than 100,000? Attached please see the pic of analysis.

 

 Many thanks in advance.

8 REPLIES

Re: advanced explicit analysis (no -24) time consumption

Valued Contributor
Valued Contributor

Can you paste in the current Log Screen and the NASTRAN NXSTRAT Solver Parameters inputs you have used.

Re: advanced explicit analysis (no -24) time consumption

Builder
Builder

Hi Stuart,

 

Attached please find the pic of the log screen and nastran solver parameter.

Re: advanced explicit analysis (no -24) time consumption

Valued Contributor
Valued Contributor
Hi Varun

It has been a while since I have used SOL701. However, from memory the inputs you have used as the "time step" are not what you may think they are.

The actual computational time step is being calculated by the program and this is based of:

T = L/C

Where L is the characteristic material length and C id the material wave speed.

If you have a single small or excessively distorted element this will increase the solution time greatly. (I believe you should be able to find this element or length in the F06).

There are ways to reduce the solution time - global mass scaling - but I wouldnt try these methods until you know exactly how to implement them correctly. Do you need to use SOL701?




Re: advanced explicit analysis (no -24) time consumption

Builder
Builder

Hi stuart,

 

do you know any tutorial aboue global mass scaling.

 

well, i wanted to perform crushing analysis under short interval of time.

 

I have run already static and advanced static, So Now I wanted to run advanced explicit to see the difference in results.

 

Re: advanced explicit analysis (no -24) time consumption

Valued Contributor
Valued Contributor
Post a screenshot of the actual model, with mesh etc.

Re: advanced explicit analysis (no -24) time consumption

Builder
Builder

Attached please find the pic of model

Re: advanced explicit analysis (no -24) time consumption

Valued Contributor
Valued Contributor
I think you are "goosed"!

A single small tet is likely reducing the stable time step considerably, but even if your elements where uniform, with that complex geometry - it is always going to take a long time to solve.

Also, SOL701 only supports 4-node tets, which arent the most accurate elements - so even if the solution does solve I would spend a large amount of time querying the results.

I would start with a cylinder using uniform shell elements and conduct your comparisons to get an idea of the solve times, results etc. then do the same with solid hex elements and finally move onto your model - all being said I think you will still have a long solve time.

Re: advanced explicit analysis (no -24) time consumption

Builder
Builder

Thank you Stuart for being " UTTERLY BUNGLING".