orsion in masonry lintels

huangyhg

torsion in masonry lintels
when working with cmu cavity walls, people in my office often specify steel i-beam lintels centered over the cmu.  bottom plates are welded to them to support the brick facade at an eccentricity dictated by the geometry involved.  
i'm concerned with the torsion that is introduced into the system with this design.  i've no problem with designing the lintels for torsion if necessary; however, i don't understand how the torsion in the lintels gets transfered out to the masonry walls at the bearing points.
it has occured to me that, once rotation occurs in the i-beam, the weight of the masonry above the beam would rest on the inside of the top flange and create some counteracting torque.  however, considering the typical geometries involved, this resisting torque hardly seems adequate.  
so what is left then for resiting the torque? all that i can think of is the tensile capacity of the mortar and i don't think that it would be wise to count on that. does anyone have any suggestions?  perhaps i a missing something significant in the way that i am visualizing the situation?  any help would be greatly appreciated.  thanks
adam
the weight of the brick coming down on the plate support can be considered as a triangular load due to arching in the brick (if and only if you have enough brick on either side of the opening).
this results in smaller loads on the plate and resulting smaller torsional loads on your beam but you are correct - you still have an issue with "where does the torsion go?".
in most smaller openings, i would suggest that the ends of the beams are pinched into the wall and this creates a bending moment on the wall at either side of the opening that is resisted by the wall's flexural capacity.  how much of the wall participates is a question to be considered in each case....usually use 16 to 32 inches of wall.
for larger openings, we have welded reinforcing bars to the top of the wide flange, extended up into the wall, to transfer some of the torsional bending into wall flexure above the beam.  in other cases, we use additional braces as the wf doesn't possess very much torsional rigidity.
extending reiforcing bars from the top of the beam into the filled cmu wall cells will also give the beam some extra strength to resist vertical loads applied after the wall is constructed, from the composite action.
thanks for your help guys.  could either of you recommend a rational method for determining when the reinforcing steel is necessary?  also, jae, could you tell me more about the "additional braces"  that you sometimes use?  again, thanks for your assistance.
i agree with the other posts.  i also take advantage of the trapezoidal bearing stress that can develop at each support to resist the torsion.  i have also, on occasion, extended to brick support plate 8" past each side of the opening, so that it bears on the brick each side of the opening.
thanks for your input dave.  i find your comment about extending the support plate into the bearing especially interesting.  if i understand you correctly, it sounds as though the torsion at the beam supports gets resolved by considering the small reactions developed between the plate and the brick on which it bears.  given the relative stiffnesses of the elements involved, i suspect that is a  fairly realistic model of how the forces would actually distribute.  
the additional braces are usually small angles (l2x2) that we attach to the inside face of the wall, vertically, and extend them down to the bottom chord of the wf.  then we add a diagonal angle from the bottom chord up to the floor/roof structure to create a triangular brace.
this must be coordinated with the ceiling/finishes so the "ugly" braces aren't seen.
also...the vertical rebar welded to the top flange creates a sort of tension/compression mechanism that resists the rotation of the beam.  this can be used instead of or in addition to the external braces.