One did not go over the clevis given they are so fat, the washer turned out to be 15mm ID, there the cable spigot was exactly 14mm. I had my spacer made around 14.2mm so a nice easy slip fit. Well that was a bugger. But the clevis is so fat, I was more willing to file that than the nice spacer!
And as for the dowel rotation tool, worked like a charm! I tested its grab on the old dowel first, the ones I tapped the M14, with a bolt and even light clamp on it, I couldn't move it with the big shifter I had in hand.
But on the actual offset dowels, they were made of a much harder steel, so the original ones were soft and the tool could dig in good, but on the offset, they did slip. But once I tightened the clamp screw enough, it did turn it with some force on the handle without slipping.
The hard steel on the dowel was completely unscathed by the slippage of the first couple of attempts, so at least on the plus side there was no risk of damage slipping on them unlike the OG dowels which were marked from just clamping them.
I rotated the dowel initially to the score lines as intended and started off just seeing if that was all it needed and pretty much yes!
The error went down to 0.07mm. I just had to be within 0.127mm or 5 thou total indicated. The error also changed angle to about 45 deg of the ref which is what I expected if the 12- 6 was zero and the 9 - 3 was 0.15. I could only increase the 12 - 6 while reducing the 9 - 3 and being in the middle of the two would be the lowest overall.
Top right bellhousing bolt now binds a little due to the housing moving down quite a bit. I figured I'd recheck the reading with this eliminated. One thing it did was removed the tension against one side of the dowels. So now I had a very slight play side to side where I could go from that 0.07mm to about 0.10mm if I put some pressure towards the left during the bolt tightening. I can even feel that slight play. Since the left dowl with the housing fully preloaded to the right yeilds the 0.07, but to the left it gets closer to the previous error. I figured that if I slightly rotate the right dowel this time in the same anticlockwise direction, it should kick the dowel slightly to the right where it takes up this slack, that way, I should be able to just wack the housing on and get the same reading each time and not have to push it one way or another.
I'd say it only need a slight tweak there. That dowel was even easier to drive in, so it should rotate with ease.
I had already measured the rear parallel error and got these measurements. I will recheck them after I finish with the dowels.
My next step here is to get a engine machine shop to shim the housing on the surfacing machine such that they get the same or as similar reading as here. On that circle at each intersecting line as accurately as possible (values change fast when you move around that circle)
What this should do it zero the parallel out to my engine and crank by surfacing it deliberately wrong in the opposite way and amount. I've seen this done like this a few times on the Tremec FB group, when the block is wrong.
I had it machined parallel and it actually made is a little worse, probably becasue the error it had happened to be more or less in the opposite direction to the engine itself.
Before I get to that, I am going to put some fresh fuel in, prop the engine via the housing safely and start the engine with just the flywheel, so make sure the balance is good, the rear main seal I installed doesn't leak (flange was a bit rusty on the seal surface, I've cleaned is as best I can, but want to be sure!) and prime and flush the PS system of any potential overhaul and assembly contamination, these do not have filters so this is important if I want it to last!
After that and the housing rear surfaced having verified good, its the spigot and clutch side of things, then the box can finally go in
This dowel and also flywheel saga is something that was stressing me out and was a major roadblock for a long time, so I am so happy to be getting past it for certain now.
I also forgot to mention that the bellhousing cast iron brace once actually installed with the bellhousing and dust cover in between was not level again because the brace and or the block is not 90 deg exactly. So not only do you have to shim for straightness on the engine studs, but also for the angle on the block to get that bellhousing plane. I achieved this with 6 shims per stud for the front two, 2 shims per stud for the middle and none at the rear. This got me even contact across all studs and flat on the bell housing and studs at the same time, a lot of mucking around!
So the brace was all sorted and with dust cover acted as one unified engine block mating face which on the dial gauge showed basically no distortion when tightening the bolts on the brace/cover. I also recommend some blue loctite on those locking nuts. Had them come loose many times before, but likely from all the strain and misalignments that are evident now. But still, doesn't hurt to use some.