This is a two separate question post regarding Braced Frame Beam Axial Forces.
PART 1: Looking to get some clarification on the correct axial load to be checking Braced Frame beams for with single diagonal braces. Per this Wiki article and some answers I've seen on the Forum, when checking axial forces in a Braced Frame beam with a Rigid Diaphragm we are to use the Story Shear Report to find the controlling seismic load and apply that Story Shear as the axial force in the beam(s). When checking this against chevron braces where the internal node is disconnected from the diaphragm to produce axial load in the beams, I am seeing some inconsistencies with reported Story Shear and the design axial force on the Member Code Check.
Example: I am checking one seismic load case on three-story, two-bay chevron brace frame where the stiffness of each frame on a particular level is equal. The Load Combinations Story Shear report is giving me a value of 499kips, which I would then split up equally between the two beams since the stiffness is equal, to hand check the beam axial/bending interaction for an axial force of ~250kips. However, when looking at the Member Code Check in the Steel Provisions, the member(s) are only being designed for an Axial Load of half that value, 125kips. Wouldn't we expect the beam to need to carry this full 250kip load before reaching the bracing? Seems that the members are only being designed for half of the reported Story Shear, per image below of Load Combination Axial Force Envelope.
PART 2: When checking these chevron BF beams in the Member Code Check, I noticed that the interaction equation for Eq. H1-1b (when Pu/Pn < 0.2) does not use the worst axis Pn for the equation. Pny will be less than Pnx, however RAM is checking only Pnx in the interaction equation when using Pny should control. Then another member being checked for Eq. H1-1a (when Pu/PN > 0.2) will use the smaller Pny even when Pnx is larger, which is to be expected. My understanding for compression members is the worst axis Pn value should be used and checked for design. Is there a reason RAM is not doing this?