Grizzly G0678 8x30: Repair, Scraping and Alignment

[Larry$]

Andy, Yes there are multiple ways of "stress relieving" but how do you know when it is fully relaxed. Traditionally casting were thrown out back to "rest." Some are now put through a high frequency shaking. FULLY??? If you are measuring to tenths , FULLY would seem to be a requirement.

 

[AndySomogyi]

Andy, Yes there are multiple ways of "stress relieving" but how do you know when it is fully relaxed. Traditionally casting were thrown out back to "rest." Some are now put through a high frequency shaking. FULLY??? If you are measuring to tenths , FULLY would seem to be a requirement.

Published empirical measurements show that thermal stress relief eliminates virtually all stresses with about a 5 hour bake, I don’t have the reports in front of me, so don’t quote me on the exact number. Conventional and mechanical aging do the same.

The castings on this mill are about 20 years old, and have been through countless vibration cycles.

Judging by the amount of rust I took off, I’d estimate the straight edge to be about 50 years old. Furthermore, I mechanically stress relieved it, and let it sit outside for about a year after I machined it.

I’d have to look up the papers that report on these methods and the exact published numbers, but I tried to follow published mechanical stress relieving procedures.

So I’m reasonably confident these are stable. Furthermore, I use many different measures for the same thing, i.e. three separate metrics to ensure ways are co-planar. Geometrically, the possible way the metrics all agree are if the planes are co-planar.

 

[Larry$]

Sounds like you have it under control. A lot of work. You must enjoy a challenge.

 

[AndySomogyi]

Sounds like you have it under control. A lot of work. You must enjoy a challenge.

I'm getting there.

A challenge, I don't know... When I bought this mill, I didn't realize how bad these were, and I figure, once you buy something, you're stuck with it. So, just trying to make the best of what I've got to work with.

 

[AndySomogyi]

Got a few hours of work in on thisUsing my version of Nick Müler’s dovetail alignment tool to line up the squareness of the knee.



It’s important that the square is aligned with the top flats so side motion of the block doesn’t throw off reading.

Here’s first print of the column, LOL, THREE points of contact. Not 3PPI, but three points of contact total.

 

[AndySomogyi]

Roughing in the back ways, they’re a freaking disaster from the factory. Started out with about 20-25 thousands of wallow.

Still can’t understand how could the factory have produced something this bad. Like I said, the machine ground surfaces aren’t bad, it the hand finished ones that are a disaster.

Tried an angle grinder and wasn’t having much luck, but your mileage will vary. If I tried to just dust the high spots, grinder wasn’t really moving more material than heavy scraper push. And I was afraid to push the grinder for fear of digging in. I’m going to try a big file as some have suggested. But down to about 10 thousands now, its contacting evenly, just still have that big dip by the oil grove that I want to get rid of.

 

[AndySomogyi]

After 20-25 some odd thousands, finally getting viable prints. Checking the squareness on the surface plate relative to the top flats every other print

At least getting a good workout lifting this 100 lbs knee around

 

[AndySomogyi]

using the box square as a master to scrape the knee square. I’m both inking the square, and indicating against it with a DTI, both confirm that I still need to twist the knee up a few tenths.

Yes, I realize that the usual method is to complete disassemble the mill, rotate the column horizontal and use that as a master, however I simply don’t have a way to move the column solo . One man shop with no forklifts, or anything, only my own muscle, and whilst I’m strong, I simply can’t move a column on my own.

 

[AndySomogyi]

Back flats are square to the top flats within 0.0002 / 10” with the front end tilting up by that amount.

Thinking this is good enough for a Grizzly ha ha, so calling it a day.

 

[AndySomogyi]

Put the knee back on the column and checking alignment. Reassuringly, the alignment about the X axis was dead, within 0.0003 over 8", which is what the surface plate showed. But twist about the Y axis is a disaster, this is what started me down this rabbit hole.

So, next steps is scraping the dovetails.