How should you secure & level a mill knee for scraping ?

[AndySomogyi]

Getting ready to pull the knee off this mill, all of the dovetail ways are badly mis-aligned. The Z dovetail is angled at like 0.015 over 8”.

I’m basically planning to pull it off and secure and level it to something solid, like my workbench. I just ordered an 8” square frame level, graduated to 0.05mm/m (~0.0002” / foot).

I’ve never done this before, so not really sure what to mount the knee to.

My bench is solid welded 2” box tube, top is 1.25” steel plate with 3/8 mounting holes drilled and tapped, whole thing weighs about 1100 lbs.

So I’m thinking get some all-thread and 2x4 to build a fixture to hold the knee, then I suppose I need a way to make fine level adjustments on it.

I have the top surface of the knee scraped flat, so I’ll use this as a reference to line up the column ways and dovetails.

I suppose scraping it will cause the reference to shift, so I guess I’ll have to re-check the level each scraping pass.

Also the knee doesn’t look that heavy, I’m guessing maybe 100 lbs, so I can easily lift that by hand onto my surface plate and ink it directly.

Does this sound like a reasonable approach?

 

[Richard King 2]

You should have asked for help before buying the level. The machine and knee do not need to be level. Just square. You can use the box level if it is square to .0002"/12. I wold clean up the column and scrape the flats using a 36" angle camelback and make a sled or home made King-way to make sure the ways are co-planar. I would also scrape the back of the knee to a surface plate prior to laying the column on it's back, then you scrape the column with both camel back and knee. Once you do the flats you flip it over one at a time and scrape the dovetails making them flat and co-planar. Them roll it so the column is still laying on it's back flats facing up and blue up the flats and dovetails and match fit the knee flats, dovetail and gib to the column, Then we stand it up with the Z axis feed screw and scrape the top of the knee top flats co-planar to each other and square to the column. Bot you don't make it a perfect 90 degrees. You scrape it out of square front of knee rising in front . 0002 to 5 tenths in 12".nee's Then pull the knee off and scrape the top doves co-planar to each other. Scraping should be pretty easy as you will be following the original surfaces on the ends that never wear. Also look on the top of this category and look at the Testing Machine Tools book, Nelson linked there. The specs for the Vertical mill as far as squareness will be there. Your lucky my company and I have rebuilt I would bet over 100 Bridgeport's knee type mills. I have also helped new machine builders make them in Taiwan.

 

[AndySomogyi]

You should have asked for help before buying the level. The machine and knee do not need to be level. Just square. You can use the box level if it is square to .0002"/12. I wold clean up the column and scrape the flats using a 36" angle camelback and make a sled or home made King-way to make sure the ways are co-planar. I would also scrape the back of the knee to a surface plate prior to laying the column on it's back, then you scrape the column with both camel back and knee. Once you do the flats you flip it over one at a time and scrape the dovetails making them flat and co-planar. Them roll it so the column is still laying on it's back flats facing up and blue up the flats and dovetails and match fit the knee flats, dovetail and gib to the column, Then we stand it up with the Z axis feed screw and scrape the top of the knee top flats co-planar to each other and square to the column. Bot you don't make it a perfect 90 degrees. You scrape it out of square front of knee rising in front . 0002 to 5 tenths in 12".nee's Then pull the knee off and scrape the top doves co-planar to each other. Scraping should be pretty easy as you will be following the original surfaces on the ends that never wear. Also look on the top of this category and look at the Testing Machine Tools book, Nelson linked there. The specs for the Vertical mill as far as squareness will be there. Your lucky my company and I have rebuilt I would bet over 100 Bridgeport's knee type mills. I have also helped new machine builders make them in Taiwan.

The column ways themselves do look pretty good, measured and they are parallel.

This is an 8x30 mill, so about half the size of a Bridgeport. It does have pretty low hours on it, so no wear, I’m just repairing what looks like horrible fitting at the factory.

The knee Y ways were ground with a bad twist and I scraped those flat, using the top as a quasi reference surface.

But the big issue, is the Z dovetails on the knee are not perpendicular to the top of the knee, in fact the knee sits off at an angle of about 0.015 per 8” (no wonder I couldn’t make a square part when moving the knee). See attached pic of how the knee sits.

I can fit the whole knee on my surface plate (barely), but I was following the Connelly book about using a level to align the knee dovetails with the knee top. And the box level I ordered cost about the same as a granite square, and seems more generally useful. Since it’s square, I can verify squareness on the surface plate also.

I suppose I can try to align the knee entirely on the surf plate, using the box level as a square.

Anyways, I’m trying to do a little as possible with this mill, repair wise, it just needs to make reasonably square parts. Hence I’m not planning on pulling the column, as it looks like it’s in good shape.

Not sure I understand why the front end of the knee should be scrapped higher than the back? I was taught to use the knee with loading on the screw, i.e. always lift the knee to dimension. But when lifting, isn’t there a drag on the column ways, so the knee tilts up slightly?

 

[Richard King 2]

Im am speaking to all the members now and Andy:

One has to be a detective and figure out what's going on.

I would set the box level (providing the bottom and sides are square (+ -) on the top clearance .0001")

and crank the knee up and down indicating the face sitting side to side and front to back. As the factory machined those surfaces at the same time as they did the flats. I use parallels setting on both flats and a blade square on top checking flats to column travel (+ - square spec. .0002") and then setting the square front to back and as I said leaving the front or operators side plus or rises up .0002 to 5 tenths (the Testing machine book says .0008 in 12", a bit to much for me) We scrape the knee out of square because of the sag the knee gets when we load the saddle and table on the top of knee. Another factor is you have to make these tests when the knee is supported by the Z feed screw as that could change the deflection or squareness if it's not in there. Trying to test that off the machine is impossible to transfer the measurements without id being installed " the way the machine operates when done"

As I said, I have scrapped machine for my whole life I am not guessing at this, I do not need to refer to the Connelly book. I try to teach on this forum to avoid our membership time and money. Do it the best way and fast way.

Before scraping the machine I would clean it and where the knee meets the column I would use some duct tape and cover that area both side and middle to keep scraping chips from entering there.

Pic's top to bottom Testing book page 48 showing squareness of table top, but this reflects to top of knee too. Specs on page 49 look at 9 - 11 telling you the front rise spec. Pic from Connelly book.
4th from top Connelly book page 337 showing sag factor scraping. also can see it on page 343, Metric to inch conversion chart and cover of the Testing book.

 

[Richard King 2]

Also when making the test with the knee assembled you tighten the lock and when testing the saddle and top of table the locks are on. Knee tests are always done when raising the knee so the pressure or backlash on feed screw is pushing up. If it were a Bridgeport oor a USA made machine I have only seen factory error 2% of the time...I would guess 1 in 5000. If it's made in China machine the odd's are 1 in 5 machines. Your machine looks like a copy of a Atlas mill. What is it a Rung-Fo?

 

[AndySomogyi]

Thanks, first thing I did when it was assembled is measure squareness against the column with a certified square.


The right way on the knee (when looking at the mill) was basically square, off by ~0.0005/12”, but the left side was curved down, about 0.008/12”. I don’t know, maybe it shifted when grinding?, who knows.

Then I measured angle about the Y axis (sideways tilt of Z knee dovetails), with a precision square on the top of the flats, test indicator in spindle and moving knee up and down — this is 0.015 over 8”!!!.

With the knee on the bench, I’m able to confirm with a square that the fixed dovetail is indeed at an angle relative to top.

This is a Grizzly G0678, a re-branded Enco 8x30. The strange thing is it’s a Taiwan made mill, guess I got one of the rejects

The measurements I made referring to the attached figure from "Testing Machine Tools", with the knee attached, and spindle trammed to table are:

Fig. 9, spindle over 3": 0.0005
Fig. 10: spindle over 3": 0.0001
Fig. 11: table Z motion over 8": 0.0010
Fig. 12: table Z motion over 8": 0.0150

 

[Richard King 2]

Over in Taiwan there are classes of machines. Enco and Grizzly are a class 1 of 3 classes. Back when it was made most of the factories made crap.

You may want to call around and see if anyone has a precision horizontal mill or boring bar. Devlieg Jig Mill would be best. and get it machined.

 

[Richard King 2]

I've been thinking of how you can test the knee and scrape it.. I understand you have some experience, but I want to share a tip that can save you or the readers a few hours of time. Many times when someone who is new to scraping doesn't scrape evenly. On flats like you have on the top of the knee one scrapes the flats more on the outside, or nearest you. That is what you said your knee top looks like. I see from one of the photo's you have scraped it. Another issue rookie scrapers have is under the dovetail there are high spots that are super hard to see and scrape / stone let alone scrape. So the outside is easier to see and scrape the blue and the flat gets lower on the outside edge. This is when it is important to look at the straight edge (SE) after you rub the flats. When you blue up the flats rub for a count of 10 and when you lay the SE into the flat leave it off the ways and set it down in the bottom or back of the flat first and only rub the SE back and forth or toward the column and then toward the front of the knee, don't pull the SE toward you. If you keep it keep it back there and rub it and then lift the SE straight up, look in there with a flashlight for high spots, also look at the SE and see where all the the blue has rubbed off. Another tip . Once you get the flats co-plnar and it is still out of square with the column I would scrape the narrow way on the back of the knee to get it square, Less work. To find out exactly how much you need to scrape off there. slide in a narrow strip of shim stock on the low side of the dove. I use plastic or steel and bend over the end sticking out....Ive even super glued a small spot so as you crank the knee up and down it doesn't slide out. Just a tip. Andy you can do it your way.

 

[Richard King 2]

I would leave the spindle alone as it can be turned to get it straight or sweep the table top. Do that last . Nothing to do with the squareness of the ways. If it was a fixed spindle like a horizontal spindle then its important.

 

[AndySomogyi]

I've been thinking of how you can test the knee and scrape it.. I understand you have some experience, but I want to share a tip that can save you or the readers a few hours of time. Many times when someone who is new to scraping doesn't scrape evenly. On flats like you have on the top of the knee one scrapes the flats more on the outside, or nearest you. That is what you said your knee top looks like. I see from one of the photo's you have scraped it. Another issue rookie scrapers have is under the dovetail there are high spots that are super hard to see and scrape / stone let alone scrape. So the outside is easier to see and scrape the blue and the flat gets lower on the outside edge. This is when it is important to look at the straight edge (SE) after you rub the flats. When you blue up the flats rub for a count of 10 and when you lay the SE into the flat leave it off the ways and set it down in the bottom or back of the flat first and only rub the SE back and forth or toward the column and then toward the front of the knee, don't pull the SE toward you. If you keep it keep it back there and rub it and then lift the SE straight up, look in there with a flashlight for high spots, also look at the SE and see where all the the blue has rubbed off. Another tip . Once you get the flats co-plnar and it is still out of square with the column I would scrape the narrow way on the back of the knee to get it square, Less work. To find out exactly how much you need to scrape off there. slide in a narrow strip of shim stock on the low side of the dove. I use plastic or steel and bend over the end sticking out....Ive even super glued a small spot so as you crank the knee up and down it doesn't slide out. Just a tip. Andy you can do it your way.

Thanks, good idea about the shim in the dovetail.

I've got the top almost flat and co-planar (I first scraped the saddle, and found that rotating the saddle 90 degrees and upside down makes a nice long straight edge, and also used the bottom of the saddle to get them co-planar, yes, I scraped the saddle first against my surface plate).

The knee top flats are almost square with the knee back flats, so it'll take very little scraping of the back flats.

I'll verify the angle of the dovetails a couple ways:

(1) level the mill with leveling feet, using the column as a reference, with the box level, then use the box square on the knee flats.
(2) test indicator in the spindle, box level on top of flats, and indicating off side of box level while moving knee up and down.

I'm totally planing on leaving the spindle alone. As you said, it adjusts in the tram direction. I know the nod direction is a bit off, but it's good enough for now, and I just need to get it back to working. Maybe sometime in the future, I'll pull the head off and scrape those surfaces square, but as long as the knee is square, it's good enough.