Radioss Implicit

gopal_rathore

Altair Forum User said:

Hi Datta,

In RADIOSS, both implicit and explicit time-integration schemes are available for the user. In an explicit scheme, velocities and displacements are obtained by direct integration of nodal accelerations. With this approach, the cycle time step is often small. Therefore, for static or slow dynamic computations where the duration of the study is long, many cycles are necessary to carry out the simulation. An alternative to explicit method in such cases is the implicit time-integration scheme.  Not all explicit features are available for implicit analysis. In loading only Concentrated load /CLOAD, pressure /PLOAD, and gravity load /GRAV are available.

The user should provide the required control cards to activate the implicit scheme. When the RADIOSS solver is running an implicit analysis it will be shown in the out file as  ** IMPLICIT OPTION USED **. The keywords required to launch a RADIOSS implicit analysis are defined in the Engine file and begin with /IMPL. In the Help Menu, please search for Activating an Implicit Analysis which will show the keywords required for an implicit analysis in RADIOSS. Examples are available in our Help Menu on implicit analysis. One tutorial and model files are attached along for your reference.

 

Hi,

It seems that we cannot use imposed displacement/velocity as loading condition while solving through radioss implicit. I was trying the same but was not working properly. I am using version 13 for the same. Kindly let me know about it.

 

Thank you

Unable to find an attachment - read this blog

Altair Forum User

Hi Gopal,

The following available kinematic conditions are supported in RADIOSS Implicit:

•Boundary conditions /BCS

•Imposed displacement /IMPDISP (velocity /IMPVEL, and acceleration /IMPACC)

•Rigid bodies /RBODY

•Interface type 2 /INTER/TYPE2 

•Rigid walls /RWALL (not recommended due to performance reasons. Use contact interface, if possible.)

•Interpolation constraint element /RBE3

•Rigid element /RBE2

gopal_rathore

Hi George,

Thank you for the information.

I was running an implicit simulation in radioss with imposed velocity applied on a rigid ram. The ram when just comes in contact with structure the solver stops generating output and non-linear convergence error is coming. Can you provide certain recommended implicit parameters to overcome this issue?

The contact is of type 7 between the rigid ram and the structure and gap min of 1 mm. In optistruct help, it is mentioned that contact in implicit does not happen if rbody is used with small gap value. Is this the reason for the above mentioned issue, if yes how to overcome the same?

 

Thank you

Altair Forum User

Hi Gopal,

Do you have any Istop id for the error in the out file?.

Please check if there any penetrations in the model. For Type 7 contact switch Iform=2 and put in a minimum gap value (GAPmin) for the type 7 contact.

Please share the engine out file.

gopal_rathore

Hi,

No Istop ID error, it just stops giving output as soon as it touches the structure and after longtime error comes non-linear iterations done 79 . terminated

No penetration in model. Contact is type 7 with IStf:4, Gap min:1 mm, Iform:2 , Inacti:6

Altair Forum User

Hi Gopal,

Implicit Solution is RADIOSS is more to complement the Explicit Solution.  So we normally suggest to use it only for specific cases like Spring Back.

This seems like an impact case which involves contacts. In such cases there can be some convergence issues.

Please share the engine out file.

gopal_rathore

Hi,

I have send the file through secure drop box link. Kindly check.

gopal_rathore

Altair Forum User said:

Hi Gopal,

Implicit Solution is RADIOSS is more to complement the Explicit Solution.  So we normally suggest to use it only for specific cases like Spring Back.

This seems like an impact case which involves contacts. In such cases there can be some convergence issues.

Please share the engine out file.

Hi,

Yes there is contact in model through which load is applied on the structure and it is a quasi static loading case. I have run the same using explicit but would like to compare it with implicit solver also. and then finalise the best possible way of solving such problems.

Altair Forum User

Hi Gopal,

Since the run was killed in between no much  information is available that led to termination. Typically for contact to work for implicit you need to use a larger gap.  

Another possibility is the function you have defined for imposed velocity. Make sure the load is ramped gradually in the model.

gopal_rathore

Hi,

I have sent the model file through secure link . Please check.

 

Thank you.

gopal_rathore

Hi George,

Have you gone through the file. Kindly let me know.

 

Thank you

Altair Forum User

Hi Gopal,

I got your files and will update you very soon.

gopal_rathore

Ok.

 

Thank you

Altair Forum User

Hi Gopal,

I have mailed you the modified model .h3d file with my observations.

gopal_rathore

Hi George,

Will check and let you know.

 

Thank you

gopal_rathore

Hi George, 

I have gone through the .h3d file. There seems to some intersection at the edges of the actuator. Please share starter and engine file also.

 

Thank you