Learning Forum

[Alina Niederberger]

Hi there, 

I´m trying to simulate a deep welding process with a moving heat source in Ansys Workbench. The heat source consits of a conical and a cylindrical part combined, depending on the Y coordinate. So far I´ve come up with two different solutions: 

 

1) Applying the heat load via BFE, using a APDL Code

    Basically a loop over all time steps and all elements, where h is the depth of the keyhole, re the radius of the conical part and ri the radius of the cylindrical part

 

*DO,jj,EMIN,EMAX,1    ! Selects centroid of every element, Iterates through element numbers

    Y=CENTRY(jj)

    *IF,Y,LT,h,THEN                                                      ! Heat source application 

        QF=0

    *ELSE

        X=CENTRX(jj)                                                      ! Centroid coordinates of jjth-element

        Z=CENTRZ(jj)

        CSI=Z+TempCSI                                                ! Moving coordinate

    *IF,Y,GE,Yi,THEN

         r1=re-TempR1*(Ye-Y)                                 

         QF=QF1*exp(((-3)*(X**2+CSI**2)/r1**2))             ! Heat source equation conical

    *ELSEIF,Y,LT,Yi,THEN

          QF=QF2*exp(((-3)*(X**2+CSI**2)/r2**2))                           ! Heat source cylinder

     *ELSE

          QF=0

      *ENDIF

    *ENDIF

    

BFE,jj,HGEN,,QF                                           ! Heat generation for selected element centroids; *ENDDO

2) Applying the heat load via BFE and the MAPDL Function Generator. In this case I used the same function for the heat source as in 1) and the primary Variables X,Y,Z,TIME (sattachement)

 

Even though I use the same function, the calculated results vary. What could be the reason fort hat? I understand that with BFE the load is applied at the nodes of an element. 

Are the differences due to different calculation points in the table, as the table might be using integration points for calculations instead of nodes? What else could be a reason fort he differences?

If I try the same but with BF, the calculated temperatures via those two ways are basically identical.

 

Thanks in advance

[Ashish Khemka]

Hi,

The heat generation rate loads specified with the BF command are multiplied by the weighted nodal volume of each element adjacent to that node. This yields the total heat generation at that node.

BFE defines an element body-force load.

Regards,

Ashish Khemka