No Warping constant for 3D beam in COSMOS/M

According to technical support at SRAC, COSMOS/M BEAM3D element "does not take warping effect into account". I suspected this when I was doing linear buckling analysis of a frame made up of beam3d elements. I was getting eigenvalues less than I expected and after awhile I discovered the buckling load for the critical component in the frame equaled the critical torsional buckling load IF ends were free to warp. I guessed that if I defined the torsional constant as zero COSMOS/M might just ignore torsional buckling. Sure enough it seemed to work. However, this seems like an awfully kludgy fix.

Has anyone else come up against this? If so what are your thoughts?