I would think that if I raise the blade and it moves a certain amount - lets assume it gradually moves from 0 to .005 - .001 per 1/8". When I lower the blade it should follow the exact same path and the indicator should move back to the original location at zero moving .001" per 1/8" down since that's what is did on the way up. This to me would indicate the blade is not flat or its flat but I'm not at 90 degrees to the table.
Concur. That is what I described.
This is not what happens - it jumps out of alignment within the first 1/8" or back into alignment depending on the direction its traveling and stays there throughout the test.
How is your dial indicator? Any backlash or friction? Do you get the same results with a different one?
Have you checked the arbor/flange for flatness and run-out? Are you always measuring using the same tooth? With an ATB blade like a WWII, the teeth are canted, alternating left-right. Did you make sure the blade is not warped in any way and that all the same-side teeth track in the same plane? Run it up all the way and check every other tooth, then do the same thing with the blade almost buried.
I have tested runout at the tooth on my WoodWorker II. After raising the blade, taking a measurement on the tooth, sliding the indicator to the back and rotating the same tooth back I get less that .001". I repeat the test by lowing the blade first and its .008" The further out to the edge you take the measurement the larger its out.
You mean you get less than .001" difference measuring from the miter slot to
same tip when the blade is rotated with the tip to the front of the table and then the back of the table with the blade full up, but .008" when you run the blade down then back up again? How do you get it back to the .001"? I am still not following. Something must be moving - the whole trunnion or the arbor is moving axially.
If it always changed by the same amount every time you ran it up and down it would mean there is some sort of twist in the whole mechanism- the plane of the blade actually rotating on its axis as it raises and lowers- it is twisting ever so slightly as it goes up and down- that could theoretically happen at all tilt angles. I can't visualize how that would happen without looking closely at the trunnion assembly.
Have you closely inspected the entire trunnion assembly including pinions and rack? Is there any play anywhere? Is the trunnion firmly attached to the cabinet? Is the cabinet warped? Any loose bolts or cracks anywhere? I can't say but it may be possible for something like that to happen if there is a problem with either the front or rear trunnion/mounts.
Then again it could be just a combination of the blade and miter slot not being parallel, and moving the dial indicator along the slot to line it up with the tooth as less of the blade is exposed. Remember, this a 3 axis situation and a problem in one direction can affect both of the others.
My recommendation- start from scratch:
1. Clean all parts, and make sure they operate smoothly.
2. Visually inspect the trunnion, racks, pinions, keepers, followers, cabinet, arbor, etc. for any obvious backlash or irregularities in shape or movement.
3. Inspect and verify your dial indicator is working properly.
4. Check the arbor for run-out and the arbor flange for flatness/run-out
5. Loosen the CI top attachment bolts and do a careful realignment of the blade - miter slot. Making sure they are perfectly parallel using a gauge plate, not a blade.
Then check to see if you get the same results. If you do, ask for a new saw.