Hello.
I'm trying to move my welding simulation models from Ansys to Code_Aster. So far, I have a plate model, with convection and radiation as boundary conditions, properties as function of temperature, and a local volumetric heat flux. Therefore, it is a non-linear steady state model in Aster. Now, I would like to move that flux along the plate.
In Ansys, I usually don't define groups that represent the heat source locations. I use the position of the nodes only and the equation that represents the flux, inside a loop. Here is a pseudocode (i will move the source 10 times, in 1mm increments):
*do,iii,start,end,1 ! i will move the source over a direction, 10 times for example. So start=1 end=10
z0=iii*tame ! center position for source. tame=1mm
NSEL,S,LOC,X,X0-A,X0+A ! select nodes inside the desired volume, where X0, Y0, and z0 are the center of heat source
NSEL,R,LOC,Y,Y0-B,Y0
NSEL,R,LOC,Z,z0-C,z0+C
Get all node coordinates(X,Y,Z) from the ones selected above
*do ! for each node
q_node = equation !calculate the heat to be applied on the node, according to heat source equation
F,node,HEAT,q_node !apply the load on respective node
*enddo
SOLVE
*enddo
So, here are my questions:
1) I cannot do this in Asterstudy directly, as it only works with predefined groups in the geometry and mesh steps, correct?
2) Will I have to create all the path for the heat source in the geometry and mesh modules (using groups)? Is it possible to do this through a loop, and not manually? Is it possible to define them through positions, like a bounding box for selection?
3) After creating these groups, is it possible to extract the nodal positions (x, y, z) of each node in that group, to use in Aster to calculate the load using a equation?
Remembering that I will have to pass the current Aster model described at the beginning to the transient regime, since each SOLVE of the loop above lasts only a few tenths of a second (0.5s for example). This time is related to the travel speed of the source.
I find this format very productive for the way I simulate, but any insight are welcome. The biggest advantage of this way is that I only need to define the start and end position of the welding, without the need to define groups for each movement of the source and so on.
I've looked at these two posts below, but they weren't very useful at the moment (despite having good information):
ww.code-aster.org/forum2/viewtopic.php?id=20116
ww.code-aster.org/forum2/viewtopic.php?id=14799
I also looked at some test cases, like the tplv105a.
I would like to thank you for any comments that may help me or other colleagues interested in migrating to Code_Aster.
