Iv63:
Yes, it matters that the bolts are SC or bearing, and then mixed with welding. The whole idea of mixing bolts and welds in a connection (as a bad idea) is that the welds are very rigid and inflexible, while in the bearing connection, the bolts and pieces must move a bit before the bolts actually start picking up their respective loads. Until this small movement, bolt bearing, bending and minor yielding, and the bearing yielding of the plates and bolt holes has taken place, the bolts are not carrying their (any) loads. The welds and top pl. are carrying all of the load, and you have not designed for that. The welds and pl. will have to yield and strain enough for the bolts to move enough, or the bolts will not pick up their share of the loads. Our normal design approach assumes that some bolts will come into play before others, hole tolerances and all, and more yielding will take place at these to make all share the load. In the SC bolted joint, the connected pieces do not need the joint movement (we assume none for SC) for the loads to start to be transmitted immediately along the whole joint, so this is acting much more in unison with the welded part of the connection. Now, consider essentially the same problem, but rather than your welded top splice pl., make that weld a CJP weld of the top flgs., and consider how this reacts, yields, stretches, rotates, carries all the load, can it do this without failing.
In your case, you have to make an engineering judgement about the strength of the top splice pl. Right at the splice line, it will have to shear (move vertically) to allow the web splice pl. and its bolts to start acting. It will have to bend, stretch and rotate to allow the compression splice pl. and its bolts to come into play; at the very least the space in the comp. joint will have to close so the bot. flgs. are bearing on each other, end to end. Is that top pl. strong enough for this to happen, with the top pl. taking most of the joint loads until this movement does happen?
