Bookowski:
It seems to me that there are about seven (+/- a dozen) variables which enter into your question. Wood shrinkage and size variability are two of them; then you have the drilling, bolting and fit-up issues, plus slippage as bolts come into bearing on the steel and wood; finally, if the plate extends beyond the bearing edge of the wood for any reason, it acts more like a knife edge than a narrow, stronger, bearing surface. Most of the time, in this application, you are trying to make a wooden beam span, carry load and deflect better; you are working with wood framing and the beam bearings are not the real problem, but you make them work too, maybe an added stud under each end; and this generally involves compression parallel or perpendicular to the grain, which the small steel edge just overpowers. That’s some sort of a bastard bearing joint where the steel crushes the wood, either parallel to (on a col.) or perpendicular to (on a beam or sill pl.) the grain, and then the wooden side members pick up the rest, and this adds to total settlement in the joint. I’m sure you can find exceptions to the above: the knife edge (flitch pl.) on a steel bearing plate on conc. blk. or poured conc., but then you have a pretty tippy beam; maybe steel flitch pl. to steel beam or column. But, basically you are trying to strengthen and stiffen the middle 2/3rds of a wooden beam which should actually be deeper to span and carry the loads. Otherwise, you want to hide the steel within the wood so the carpenters can frame around it. You are correct that LVL’s shrink less across the grain, and they do have higher compressive strength perpendicular to the grain, but you’ll still get fit-up movement in the system.