Input a matrix into Abaqus (*MATRIX INPUT)

[Gramsey]

Hi all,

This is what I'm currently attempting:
-Run a static analysis with contacts and centifugal nonlinear effects
-Use the resultant displacements and stresses in a new analysis so the contacts disappear completely (not enough with contact change).

I have tried loads of different methods to overcome the contact problem (restart, different dummy MPC arrangements, activating-deactivating contacts...), unfortunately none of them is good enough for what I seek and the best I could manage was a new analysis without the centrifugal nonlinear effect taken into account, which is not ideal as the stiffening effect of the centrifugal load is considerable.

Recently, I came up with the following approach:
-Run the nonlinear static contact analysis generating mass and stiffness matrices (*MATRIX OUTPUT...)
-Use deformed mesh and stress initial conditions from .odb in a second analysis. In this second analysis I'm also trying to use the stiffness and mass matrices obtained in the first analysis as an input.

Reading the Abaqus documentation it seems like importing the mass and stiffness matrices from the first analysis would be possible by using *MATRIX INPUT and *MATRIX ASSEMBLE. However, when I go down this line I get the following error: ***ERROR: NO STEP DEFINITION WAS FOUND . This makes very little sense to me as I have obviously defined a Step in my analysis... I have no experience whatsoever using this command so I was wondering if somebody could give me a hand on this, as I could not find any examples on the internet.

Thanks in advance,

Gramsey

 

[FEA way]

Interesting approach. I think that the error is caused by the fact that matrices can’t be used in a model defined in terms of parts and assemblies. You can find exemplary .inp file template for analysis with matrices in the „Using matrices” section of Abaqus Analysis Guide.

 

[Gramsey]

Hi FEA way,

Thanks for your response. I actually thought the same and tried without part and assembly definition. In fact, I have taken that template that you mention as example but still not working for me :-(

 

[FEA way]

Could you attach your input file ? It doesn’t have to be solvable if you don’t want to share all the details about this model, you can delete matrix definitions. I wpuld take a look and try to find errors.

 

[Gramsey]

The model is actually quite simple. It is intented as a trial for a larger model. I commented the *INPUT MATRIX part

 

[FEA way]

Do you want to define the whole model using matrices or are there some additional parts of the model not included in matrices ? Keep in mind that parts defined using matrices must be connected to the remaining ones via shared nodes (these nodes need to be defined manually with *NODE keyword). Nodes that aren't shared and have no BCs and loads applied don't have to be specified, they are handled automatically. Also there is a *SYSTEM keyword in your model with no data lines. I guess it's misplaced.

 

[Gramsey]

Ideally I would like to have the possibility of using matrices+additional parts, but for the time being I would be happy with just matrices to keep it simple so that was my intention with that model. You are right about the *System command, I will delete it.

 

[Gramsey]

Hi,

I still haven't managed to make this work. I just want to successfully use the *MATRIX INPUT command. So I've got the .sim file from a previous analysis (same model I attached last time). However, the following error comes up:
***ERROR: SIM database file is not a valid SIM file:
/folder/example/job_matrix_trial_X4.sim
I have used the *MATRIX INPUT command as follows:
*MATRIX INPUT, NAME=MAT1, INPUT=job_matrix_trial_X4.sim, MATRIX=STIFFNESS

Something I'm missing/doing wrong? Why is the SIM file not valid? It comes from a successfully run analysis...

Thanks in advance

 

[FEA way]

That seems rather unusual. Isn't then SIM file empty ? Check .msg and .dat files generated in then first analysis, maybe they will give some hint about the reason of the error. You can also try generating matrix output in text format and then using it with *MATRIX INPUT keyword. Also make sure that the error is not caused by access permission issue. Is the second job run in the same folder as the first one ?

 

[Gramsey]

SIM file is not empty (1439 KB). Yes, the second job is in the same folder as the first one (could that be a problem?). The .dat and .msg files are totally fine with no errors:
S T E P 4 M A T R I X G E N E R A T E

THE FOLLOWING GLOBAL MATRICES WILL BE GENERATED AND WRITTEN TO job_matrix_trial_X4.sim

STIFFNESS
MASS

Regarding generating a text output, I have generated mass and stiffness .mtx too. Is there any way to use them in with *MATRIX INPUT?

 

[FEA way]

It should be possible, since *MATRIX INPUT keyword allows you to define matrix through data lines. So you could copy the content of .mtx file and paste it under *MATRIX INPUT keyword. Of course you have to make sure that syntax is correct.

 

[Gramsey]

Interesting. I have tried copying the .mtx into the .inp file and it partially works. I had to copy the whole text, which is fine for small models, but perhaps not very much so for bigger matrices. I've tried using an *INCLUDE to link the .inp to the .mtx file with the *MATRIX INPUT heading inserted. Unfortunately, it fails for some reason at reading the file it seems (it does not produce any .dat/.msg files).

Aside from that drawback, I'm trying to include a stress initial condition based on the previous .odb. It comes with this error:
***ERROR: OdbError: Cannot open file
/tmp/u5003396-2019.09.20-09.51.32/job_matrix_trial.odb. ***ERROR:
"/tmp/u5003396-2019.09.20-09.51.32/job_matrix_trial.odb" is not an
Abaqus database file.
Note that the stress initial condition works perfectly fine without *MATRIX INPUT as results look fine with no errors. However, when both are active in the .inp the above error appears. Any idea why this happens?

Thanks in advance,

Gramsey

 

[Gramsey]

Sorry, my bad, the first problem was due to a typo... But I'm still struggling with *MATRIX INPUT+*Initial Conditions