Search
Close this search box.
Notifications
Clear all

Help with Abaqus Explicit

5 Posts
2 Users
0 Likes
477 Views
Posts: 29
Topic starter
(@nadlie)
Eminent Member
Joined: 14 years ago

I have a ball in socket contact problem from a seperation conditon under a load which leads to impact (hence I am using abaqus explicit). However, I have having trouble getting a solution. The analysis runs for a very long time and after about 12 hours of running abaqus just crashing to the Not Responding state. Can anyone offer any help or advice on this problem. What should I be looking at in the run monitor to assess if the analysis is likely to reach a solution or not. I really would appreciate any input, thanks.

Topic Tags
4 Replies
Posts: 23
(@asantis)
Eminent Member
Joined: 14 years ago

After 12 hours Abaqus Explicit should have produced many increments. Do you get any solution? If there comes absolutely nothing, you can kill the job after some minutes, because there is something wrong.

Do you get a message file or is there something in the .dat?

Topic Tags
4 Replies
Posts: 29
Topic starter
(@nadlie)
Eminent Member
Joined: 14 years ago

Crani, thanks for your feedback. I reach around 14 million increments (about half way through my time step) and the analysis does not progress any further. In the .dat file I recieve warnings but no errors.

Can I use the kinetic energy and total energy in the monitor to assess the analysis run?

Reply
Posts: 23
(@asantis)
Eminent Member
Joined: 14 years ago

It might be worth to have a look at the energies, if eberything seems logical and is in the expected range.

Why dont you split your process into more steps and produce a restart file? Then you can save time, when you process further. 14 million sounds a lot for just one step. Maybe you have trouble with the memory? For each increment the analysis saves the whole state, depending what output you request.

Reply
Posts: 29
Topic starter
(@nadlie)
Eminent Member
Joined: 14 years ago

My Kinetic energy seems very low (starts from 0.013 and drops to as low as 1.5e-05), and my total energies start positive and then become negative but again are very small. There is very little mass involved in my ball in socket type model and slow dynamic movement. The reason I am using abaqus explicit is because I have a seperation in my model which then comes into contact under a load. I know I can do this using abaqus standard and nudging the components together before applying the load, but I wanted to achieve this without modelling nudge.

In terms of memory, I have around 4GB hard drive space free and 8GB RAM, using a 2 processors (2.8 Ghz). During the analysis only around 50Mb or hard drive space is used and 100% of the CPU is used. Thank-you again for you input and I look forward to any further comments. My stable time increments are 4.15e-08 (total time of the step is 1s).

Reply
Share: