I broke it down to make it easier to keep track.
Carl,
If the center tube is just a passage for a shaft as I suspect, then the only tolerances that matter will be the parallelism of the two faces and the relationship of the two sets of bolt holes...
On the tube itself, nothing really matters but the finished length. It will be assembled before the two plates are joined.
...Your tubing assembly will give you a few small problems, because the tubing will not be perfectly round, and the inside of each tube will have a small weld seam that will need to be removed to allow the next tube to slide in...
There is only 2 pieces of tubing, both are DOM. The difference in ID and OD is 1". The piece that connects the two is machined from the extra flat bar mentioned above. It can be made to fit whatever dimensions the tubes come to.
...When you weld on the flat plates, make the stackup a little thicker than the finished dimension to allow for machining both faces on the Bridgeport. Clamp the small face to the table (you'll probably need some blocking [or machinists jacks] on the overhanging end and will have to move clamps around as you cut) and take a skim cut until the surface cleans up. Flip the part, cut the small face to the correct dimension. We skim the large face and then cut the small face to the dimension to reduce the amount of material to remove. You could also remove equal amounts if you wanted...
The plan is to machine both plates, make the center tube, and weld them together. So long as the two plates are reasonably concentric with relation to the tube, the rotation of the 2 bolt patterns doesn't matter. Actually "clocking" the transfercase flat has advantages, and there's little disadvange to having this option.
Taking into account the thickness of the plates (.75" and 1") the size of the fillet (.25"), and proper control of the HAZ, facing both plates mounting surfaces shouldn't be required. A gasket or RVT will be used either way and unless there is serious bowing of the plates from the circular weld, I don't forsee an issue.
...While the small face is up, you can put in the bolt pattern on the mill, as well as any holes that can be reached on the lower large face. Flip the part, indicate the previously drilled holes, and drill the remaining holes. You really should spotface the drilled holes so the fastener heads sit perfectly parallel to the faces, but it probably is not real critical and a reverse spotface would probably be hard to find and expensive (MSC might have it)...
The plan (already mentioned I know) is to machine the plates before assembly. I don't see a disadvantage (provided bowing/cupping) is within spec, but I see plenty of advantage.
Facing the plate around the bolt holes is way more effort than is required for the application (transmission/transfer case adapter). The OEM adapter is left as cast in this regard.
If/when I build this, I might take several extra steps just to gain experiance, but this design doesn't need that level of perfection.
...Good luck. You're on the right track and based on your previous work, you should have no problems picking up the few new skills you will need to make this part...
Thanks, I agree it shouldn't be an issue. I appreciate input. I'm sure if/when I get a lathe/mill I'll have to join a new forum to get all my questions answered.
...If you have access to a fairly good size lathe with a good compliment of tooling, you might want to think about making the cylindrical section from one piece.
That would be cool, it would clean up the design considerably. However, it would increase the material cost of the project by 35% and increase machining time dramatically. Unless I'm missing a better, common source, the best option would be a 2.75" length of 5" round bar. I don't have (easy) access 5" bar's price, so I 4x'ed the 2.5" price.
Yeah, B, check out Sketchup. I used it before I got SW.