Welds at Beam Bearing Plates

[mcc202]

When beams bear on CMU or Concrete walls, I have always shown the beam field welded to the bearing plate, with the bearing plate set in the wall with anchor rods. Recently I have ran into issues where the walls have cracked at the bearing pockets just below the bearing plate, most typically at CMU walls. I attribute this to the thermal expansion/contraction of the beam pulling/pushing on the bearing plate and anchor rods during construction prior to the deck being placed. This could be solved by eliminating the weld from the beam to the bearing plate, but then the beam would technically be unstable until the deck has been placed. I have also messed around with using oversized holes and hand tightened anchor rods, but the bearing plate still seems to bond to the wall via the grout.

Just curious what other engineers are doing. I have looked at several drawings from other firms and everyone seems to show the beam field welded to the bearing plate, so this must be a common occurrence.

Thanks