Learning Forum

[John Miller]

Hello,

I’m scripting for an AM simulation (FDM Process) of the thermal behaviour of the additive manufacturing of polymer on several layers. That’s not difficult, but I try to enter the postprocessor during the simulation to select the contact elements to calculate the convection of the base plate (decreasing area due to deposition of the first layer).

The model consists of a workpiece volume, in which elements were activated a bit at a time and a base plate (two separate solids, with a contact area).

To achieve the collection of active contact elements I wrote this script:

*do

...

load, ...

... 

finish
  /post1      !enter postprocessor
  /GOPR
  set,last

  esel,s,enam,,174                                     ! selection of contact elements 174
  etab,contact_elem,nmis,41                     ! etable of selected elements
  cm,interface_nodes_bed_contact,node  ! building a named component

  finish            
  /solu            ! enter solution
  /GOPR
  nropt,full
  thopt,full
  antype,,rest

 ...select external nodes...                        !external nodes of path und bed (all)
  cmsel,r,interface_nodes_bed                 !reselect external bed nodes
  cmsel,u,interface_nodes_bed_contact  !deselect external bed nodes in contact
  cm,bed_ring_nodes_conv,node

  sfdele,bed_ring_nodes_conv,conv
  sf,bed_ring_nodes_conv,conv,hc,Talpha

*enddo

Something is wrong at this code, because I get always this:

The simulation consists of 3 steps, 1.) bed heating (1 s), 2.) Depsoition process (about x s) 3.) cooling (y s)

The simulations stops after 1st step, at the first iteraration of the second step.

 

The simulation is hovering in the nowhere

Without the entering of /post1 and re entering of /solu the AM process is working just fine.

Maybe someone has a clue whats wrong with my code.

 

Greetings

John

[Chandra Sekaran]

In the below commands, the elements are selected but ALL the nodes are selected at this point (assuming you did a ALLSEL before SOLVE). So the 'interface_nodes_bed_contact' has all the nodes? Can you try doing a NSLE before creating this component? 

  esel,s,enam,,174                                     ! selection of contact elements 174
  etab,contact_elem,nmis,41                     ! etable of selected elements
  cm,interface_nodes_bed_contact,node  ! building a named component

[Dave Looman]

In general, it's more robust to not leave solution.  While it wouldn't be possible to retrieve etable items in solution, many other results are available and perhaps one of them could be a surrogate for NMISC 41.   

[Bill Bulat]

The image you sent suggests you are using a command object in Mechanical. It might help to solve without using the distributed solver (use instead shared memory parallel processing). Also, include a command object set to be executed during the first step with this command: /config,noeldb,0 and also make sure you save "contact miscellaneous" results to the results file (this is set in Analysis Settings Details). Using /config,noeldb,0, it should not be necessary to use the SET command each time you enter /POST1.