Raise number of quadrature points

[FaWalte]

Dear all,

I am using tetrahedral elements with 4 nodes and 1 quadrature point during my simulation. However, within the scope of extrapolation from quadrature points to nodes I need an equal number of nodes and quadrature points.
In this way, I tried using the command "QUADrature order q_1" to raise the order, but however didn't achieve any difference. I tested this for a cubic element using the "block" command as well,  without success.

Code: [Select]

[QUADrature ORDEr 2]
Is there a different way to get a 4 node tetrahedron with 4 quadrature points or what do I have to consider when applying the "QUAD"- function? (I am using FEAP Version 8.6)

Thanks in advance!

[Prof. S. Govindjee]

What does it say in your O-file for the "Quadrature Order" in your two test cases?  (also post your material cards for the two test cases)

Lastly, can you try

Code: [Select]

QUAD TETR instead of QUAD order q_1 and post the output seen in the O-file?

[Prof. S. Govindjee]

Looking at elements/solid3d/quadr3d.f, I see

Code: [Select]

      if(nel.eq.4) then                       ! 4-node tet
        ttfl = .true.
        if(d(182).gt.0.0d0) then
          call tint3dn(nel,lint,el3)
        else
          if(stiff) then
            l = 1                     ! 1-pt quadrature
          else
            l = 2                     ! 4-pt quadrature
          endif
          call tint3d(l,lint,el3)
        endif
If you use nodal quadrature (quad tetr), then the call will be to tint3dn( ) and you will get 4-point quadrature.  To get an interior 4-point rule it looks like the flag stiff needs to be .false., but this only happens when computing damping and mass matrices.  You could force the issue by changing sld3d1.f, for example, to call quadr3d( ) with the flag set as .false., but that is a hack and not very satisfactory.  Better would be to check the value of d(5) and set the stiff flag accordingly.  Note d(5) stores the value of q_1.  Scrolling through the rest of quadr3d.f it seems that for the 3D built-in elements QUAD is ignored.

If one looks at elements/solid2d/quadr2d.f, however, one sees that d(5) is used.

Perhaps I am not reading the code correctly.  Prof. Taylor will hopefully interject with an authoritative response.

Also note that program/tint3d.f does have the capability to generate quite high order integration rules for tetrahedral elements.

[FaWalte]

Thanks a lot!

The command

Code: [Select]

QUAD TETR totally fixes my problem. The tetrahedron features 4 quadrature points now, which can be seen in the O-file also.