Why isn't the new spectrum comination method being used by FSUM when issued in a WB Mechanical commands object at 13.0?
Therea are two reasons. Reason 1 is that to get the correct reaction via FSUM the modal results need to be written to the .mode file. If they are not, the special FSUM logic will not work. So that is a problem since by default Mechanical will not set that key since it just makes spectrum analysis slower unless you need FSUM. To force this on, set "Store Modal Results" to "For Future Analysis" in the modal analysis. If you do this in R14 you'll get the correct answer. Further in R14 if this isn't set, ANSYS will catch this with an error message to avoid getting wrong results like you get in R13. (In R13 you can tell the results are wrong because there isn't the note about the spectrum summation method)
FSUM in 14 without MSUP=0
*** ERROR *** CP = 3.728 TIME= 13:11:15
Element results are not present on the .MODE file for the calculation
of the sum of the nodal forces. The FSUM command is ignored.
FSUM in 14 with MSUP=1
***** SUMMATION OF STATIC FORCES AND MOMENTS IN THE GLOBAL COORDINATE SYSTEM *****
(Spectrum Analysis summation is used)
FX = 704.5789
FY = 713.0355
FZ = 5.311606
MX = 2.601548
MY = 2.257272
MZ = 31.89336
The 2nd reason is that at 13.0 this special spectrum summation logic only worked with loadstep 9999 but WB Mechanical writes the spectrum results to load step #4. MAPDL has been changed to support load step 4,5,6 now in R14. Thus your commands object with FSUM will be fine in 14.0.