Unit system in RADIOSS?

Altair Forum User

Hello,

 

I am running the same simulation in two different unit systems - inch-pound-second and mm-kg-ms.

Please note that everything is consistent in unit conversion.

I am observing tremendous stress/strain/deformation for the inch system compared to the mm system.

Please note that the difference is by multiple magnitudes.

What is wrong here?

I have attached the .hm files as well as the excel file with details on units followed.

As RADIOSS is a unit less solver, it should not matter what unit system we follow so long as it is consistent.

Unable to find an attachment - read this blog

Altair Forum User

Hi bro,

file was attached a little mistake

and i think inch, pound compare mm, kg only is approximation.

Altair Forum User

Hi Bongirs, 

Upload the inch,pound, sec .h3d model in HyperView for post processing and click on last time option and apply.

The Von mises stress value showing is 3.5e5 psi which is equal to 2.439 GPa (result from kg,mm,ms)

<?xml version="1.0" encoding="UTF-8"?>

Altair Forum User

Hello Duong and George,

 

You are comparing the end value of IPS with the max value of SI model. It is linear elastic, so this is not a good comparison as they occur at different stages of the impact.

Please see the excel I attached, the max stress values are provided in it. Clearly the IPS model has way too high stress.

If you still cannot see the difference then I would recommend you to choose a Johnson Cook model with your choice of consistent parameters.

You will find that the IPS model has like 3X plastic strain.

Altair Forum User

I am still expecting some kind of reply...

Do you get what I am saying George?

Altair Forum User

Hi bongirs,

 

I am a student working on academic projects.  For my problem I get different results when I use inch/lb and when I use mm/ton. I feel there is some problem with radioss with the units. It would be really helpful if the Radioss team deals with this problem. 

 

Thanks,

Datta 

Altair Forum User

Hi Datta1234,

 

Thank you for sharing your findings with me.

I spent months trying to find out fault in my simulations when I realized this bug in RADIOSS.

I don't understand why George is not replying.

If anyone else is facing the same issue, then please post it here...

Altair Forum User

Hi Bongirs and Datta, 

Please update the element size of your model with respect to the unit system you are following and try running the analysis.

That is, if you meshed the component for 5mm element size, convert the element size to inches and mesh the component with the modified element size in inches. Then try running the analysis.

Altair Forum User

George, I really didn't expect this kind of reply from you.

Please see the .hm files I uploaded previously.

I have already maintained consistency in geometry, mesh, material, boundary conditions, etc

 

Altair Forum User

Sorry Bongirs, I didn't notice it. These are some basic workarounds we suggest.

I'll take up this to the development team and will get it rectified, if its an issue. 

Altair Forum User

Your reply confuses me a little, so just want to confirm what I understood.

You acknowledge that there is some issue with RADIOSS for different unit systems and are checking with QA/SD teams?

Altair Forum User

Normally, the solver will provide the output based on the unit system we provide and I don't think there will be chance of a bug in such cases.
But I'll check with the respective teams for any issues related to this and if we find any bugs, we will fix it up.

Altair Forum User

George, any updates?

Altair Forum User

Hi Bongirs,

I will follow up and let you know.