Learning Forum

[Sebastien ARRIVET]

Hello, 

I'm building a model for a modal analysis, and since it has many differents parts with various scale I use 3type of elements:

• 2D square elements for tubing
• 3D hexa elements for extruded mass
• 3D tetra elements for complexe forms

My analysis goes like this:

• load mesh information for thermal study:
  • ET,2,87
  • ET,3,132
  • ET,4,279
  • (I have a total of 9 ET, but they only use those 3 elements, quadratique thermal elements)
• load some parameters (shell thickness for instance)
• give the shells some thickness with emodif, material attribution for all parts (9 differents materials) and TREF = 20
• load thermal boundaries from a given fluent analysis. For each node of the model:
  • d,'node_tag',TEMP,'temperature_value'
• solve the thermal analysis:
  • /SOL
    allsel
    ANTYPE,0
    NROPT,AUTO
    DELTIM,1.0,1.0,1.0,OFF
    CSYS,0
    SOLVE
    finish
• load mechanicals boundary conditions, starting with
  • /PREP7
    csys,0
    ETCHG, TTS
  • some nodes are rotated for cylindrical boundaries, some CERIG are used
• solve mecanical analysis:
  • /SOL
    !PreStress
    ANTYPE,0
    csys,0
    allsel
    LDREAD,temp,last,,,1,,rth
    allsel
    PSTRES, ON
    SOLVE
    FINISH

If I stop the job at the thermal analysis, the thermal field is correct. The analysis produces a file.rth.

At the end of the mecanical analysis, when I load the result, only the shells elements have the temperature from the thermal analysis, the rest of the model (3D hexa and 3D tetra) are at the TREF temperature.

I've tried many things, many selection combinaison, and I'm kinda out of ideas. Any chance someone spot an error in my input ?

Thanks a lot,

Seb

[Sebastien ARRIVET]

To add some information: I've tried to remove the shell elements, and it still doesn't work. I've also changed the ET 132 to ET 90, which seemed more appropriate, but it had no impact on the analysis.

I've looked into the elements KEYOPT to see some could impact anything, but I couldn't find something of value.

I feel totally stuck on this issue....

[Daniel Shaw]

Some suggestions:

1. Try explictly changing the element types rather than using the ETCHG command or at least list the element types after issuing ETCHG to ensure that they element types were properly changed.
2. Try listing the applied temperatures (BFLIST) to ensure that the were transferred properly before solving.

[Sebastien ARRIVET]

Thanks for the answer !