Simulation issue, is it a bug, feature request or me?

[Anonymous]

I have used the attached file to test the Event Simulation function and have come up against a serious problem.

When the fender is joined to the weight the simulation works fine and the solve is completed, but when I split the fender from the weight and use Bonded Contacts I get an error message and the simulation fails.

The reason I need the two components to be separate is because I would like to use different materials, is it possible?

[John_Holtz]

Hi Cx2,

That appears to be a bug. It is either taking time to download the results from the cloud, or it is not downloading the results for whatever reason. I will take a look at the model and see if I can find a solution.

[John_Holtz · Accepted solution]

Hi again Cx2,

Let me clarify a couple of things. I ran your model twice. For the first analysis, I got a message that the solve failed, but I could not find any details about why it failed. Is this the error that you were getting? The second time I ran the analysis (after doing some other things in another study), I got results, but the results are not what I would expect. This second issue is what I was referring to as "a bug".

It is definitely possible to use separate components (and therefore, different materials) in the analysis. I do not know why the results appear to be wrong when I ran your setup, but I made a copy of the study and changed these things. The results of the revised study look reasonable to me.

1. I put a 0.5 mm separation between the fender and the cages. This change probably has no effect on the analysis since the contact is setup as "Separation".
2. I assigned the initial velocity to the fenders in addition to the weights. I assume the fenders are also travelling at 1 m/sec, too This change is probably the difference between obtaining expected results and unexpected results.
3. I increased the simulation time to 0.003 seconds. (Even this is short compared to the impact since the weight is still moving into the cages at the end of the simulation.) This change probably has no effect on the analysis results.

So, give item 2 a try and see if the results are realistic. Let us know how it goes.

[Anonymous]

Hi John,

I tried adding initial velocity to the fenders and then when that failed I tried adding it to the weights but that also failed, I did not think of adding it to both at the same time.

The error is gone but the simulation does nothing, not sure if I have done something wrong?

[John_Holtz]

Hi Cx2,

Does it fail to solve, or is it taking a long time to finish? After looking at your new model, I see that there is one other change that I probably made that I may not have mentioned. Under "Manage > Settings > General", I probably turned off "Allow Element Deletion".

I also see that all of the parts are defined as separation contact. Is that intentionally? I would expect that the fenders would be bonded to the weight.

I have attached the version that I ran previously -- in case you want to continue from it.

[Anonymous]

Hi John,

thanks very much for your help on this, it now it works because of your suggestion about adding force to both components.

I used separation contacts intentionally in that version because it worked when I was using the same materials for both components.

I went back to the original modelling file, started the whole simulation process again, used bonded contacts between the weights and fenders and added force to both components.

The original file has had no other modifications made to it, all components are still touching and no time changes were made.

The lessons learned here are when multiple materials are used in this kind of test, each component must have its own force and bonded contacts must be used for the objects travelling together.

I have attached the latest file and a couple of images of 1mps and 4mps if you are interested, I plan to do a full car with the correct weight and shape when I have the time, I could not have done it without the ability to mix materials.

Your time is much appreciated.