Making a gib for my lathe-

[ErichKeane]

SO, Full backstory: https://www.hobby-machinist.com/threads/how-do-i-make-this-gib-materials-order-of-operations.79423/

For those who don't want to read, my 14" Reed Prentice lathe's compound slide gib was pretty messed up, so I opted to make a new one! I am following some of the advice in that thread.

First, I ordered a piece of Class 40 Cast Iron from Speedy Metals.

Next, I cleaned my table and retrammed the head (I used this https://www.edgetechnologyproducts.com/pro-tram-system-01-000-10-000-09-000/). I'd prevoiusly had the whole machine apart, so the head was quite a bit out of tram, despite aligning it with a square.


While that was happening, I also cut my 24" piece of stock in half. I ordered 1.25" square stock, 24" long. I need ~10-11 inches (and much less than 1.25"!), but ordered 2x as much so that I have 2-4 chances in case I mess up.


The stock was actually REALLY flat from speedy metals, surprisingly so! On the table, I couldn't get .003 shim stock under it anywhere! I flattened 1 side to be a reference surface anyway, but I had to swap 3 clamps around. I ended up taking ~.008 off, but the 2nd pass was likely unnecessary.


I flipped it over and did it again This will be the new 'top' of my gib.


Next, I indicated along the back side to minimize the amount of material I needed to take off. I'd initially done this with a combo square off the front of the table and was only off ~3 thou over the whole length! A little bit of tapping, and it was just about perfect.


Next, I cut enough material off the back to remove the rounded top. Initially, I'd planned on this being the dovetail side, but it isn't going to work out that way, so cutting this much material off ended up being unnecessary.


Next, I was going to cut the dovetail, however ran into a huge problem: I have a riser block, so even with the spindle all the way extended, the table all the way up, and quite a bit of stickout, I'm still about 3/8" too high!



SO, I'm going to cut the "flat"/"overhang" side on the back. I can use the same endmill as before and just take the 0.130 off the back instead. I am just going to use 1-2-3 blocks under the stock when I flip it around to do the dovetail side.

I came up with ANOTHER significant issue however that I wouldn't mind feedback on. The nose of the dovetail (the acute side) is rounded over a bit, so measurements off of it aren't terribly accurate. Any amount that I miss this 'radius' is going to significantly alter the size. At the moment, my solution is going to be to just make it 'close' but a bit oversized. Then, I plan on just doing a set of test/fit cycles. I don't have a good, repeatable way to put it back on the mill (since I'll have removed most of the material from the 'bottom' as it sits on the mill now), so my ideas are:

1- If I end up too large, put some ~600 grit sandpaper on my surface plate, and take a few swipes at a time off the dovetail.

2- If I end up being too 'small', take some brass shim-stock and loctite it to the flat (which is not a moving surface).

Thoughts?

 

[RJSakowski]

What type of gib are you making? I have seen two types of gib design, one where the gib is adjusted by a series of set screws and a second where the gib is tapered and adjusted by moving the gib lengthwise in a similarly tapered space.

The first type really only has one critical surface, the side that contacts the moving dovetail. It isn't even necessary that the gib have parallel sides as the back side is held away from contact by the adjustment screws.

The second type is considerably more critical as for any section along its length, the two contact faces have to be parallel and the taper has to exactly match the taper of the ways. Given that the taper is usually not specified, this becomes a trial and error exercise. If the original gib was not in too bad a shape, it could be used as a backing plate to set the taper angle, at least to rough the new gib in but scraping would be needed to properly seat the new gib.

The two non contact faces are not critical. I would mill them using an angle vise or sine plate rather than using a dovetail cutter.

 

[ErichKeane]

What type of gib are you making? I have seen two types of gib design, one where the gib is adjusted by a series of set screws and a second where the gib is tapered and adjusted by moving the gib lengthwise in a similarly tapered space.

Its the second, it is a tapered gib adjusted by a screw that pushes it lengthwise.

The first type really only has one critical surface, the side that contacts the moving dovetail. It isn't even necessary that the gib have parallel sides as the back side is held away from contact by the adjustment screws.

The second type is considerably more critical as for any section along its length, the two contact faces have to be parallel and the taper has to exactly match the taper of the ways. Given that the taper is usually not specified, this becomes a trial and error exercise. If the original gib was not in too bad a shape, it could be used as a backing plate to set the taper angle, at least to rough the new gib in but scraping would be needed to properly seat the new gib.

I have at least a good 6" of the original gib that I intend to use as the backing plate to set the taper angle. I don't know anything about scraping, so I was hoping I could "scrape" it in with sandpaper.

The two non contact faces are not critical. I would mill them using an angle vise or sine plate rather than using a dovetail cutter.

I have neither of those. I have a sin bar, but I couldn't come up with a way to use it to mount this. The dovetail cutter is cutting a contact surface, so I'm hoping it ends up being close enough to work.

 

[machPete99]

I have not done this, but you should probably check out this video on the subject by Stefan Gotteswinter:

 

[ErichKeane]

I have not done this, but you should probably check out this video on the subject by Stefan Gotteswinter:

I'd watched that a few weeks ago. Is there something in particular you think I've messed up? I only have a 1 direction taper, which makes a lot of what he does unnecessary.

 

[machPete99]

I think all of these gibs are only tapered in one dimension, and I thought that was what he was doing.
The geometry is still a bit tricky to get your head wrapped around, but I figured that he is a pretty good machinist so is probably doing it the right way.
You typically end up with a parallelogram shape in looking at the end, and getting those angles layed out correctly can be tricky.
Like I said, I don't have the answers here, but may go through this on one of my machines one day, so if you figure out a simpler way please document it!
--mP

 

[machPete99]

On you measurement issue, I would try to use a dowel pin or small length pf drill rod, diameter of which should be a bit under the total height (.542) probably around 0.500 diamter. Put the dowel pin against the angle surface, that way you are measuring somewhere in the middle, and reference from the bottom surface. Do this on the original and new part in similar locations to compare. That way you are not measuring at the sharp (or not so sharp) edge. You could probably also achiece this using a similar size ball bearing for a single contact point, which might even be better.

I would also be tempted to arrange for the part to be held at the correct angle and use a straight end mill to machine, similar to Gotteswinter. That also gets the part up into your work envelope.

If your table is level you might be able to use a sine bar and machinist level to read the original angle, as well as setting the work holder to cut the new one. You would need a square bar with a vee notch along its length cut with a straight end mill, with head tilted at a 30* angle. This bar would then be tilted in the vise to get the proper taper.

 

[ErichKeane]

On you measurement issue, I would try to use a dowel pin or small length pf drill rod, diameter of which should be a bit under the total height (.542) probably around 0.500 diamter. Put the dowel pin against the angle surface, that way you are measuring somewhere in the middle, and reference from the bottom surface. Do this on the original and new part in similar locations to compare. That way you are not measuring at the sharp (or not so sharp) edge. You could probably also achiece this using a similar size ball bearing for a single contact point, which might even be better.

I just don't have a good surface to reference the dowel pin off of though. Typically you do that on the inside of something, but I don't really have that. I'm at a bit of a loss.

I would also be tempted to arrange for the part to be held at the correct angle and use a straight end mill to machine, similar to Gotteswinter. That also gets the part up into your work envelope.

If your table is level you might be able to use a sine bar and machinist level to read the original angle, as well as setting the work holder to cut the new one. You would need a square bar with a vee notch along its length cut with a straight end mill, with head tilted at a 30* angle. This bar would then be tilted in the vise to get the proper taper.

My table isn't particularly level unfortunately, Its close, but not enough to measure off of. My floor is wavy enough that it makes it pretty tough. My head only spins left/right, not titlt up/down, so I'd have to do it on the Y axis, where I don't have that travel unfortunately.

Also unfortunately, the way I had to line it up (and the only way I had a machined surface) was to do it like below. I used 1-2-3 blocks under my material and clamped it down. Then I used 2 woodworking clamps (1 removed in the photo) to hold the old gib against the side on top of a parallel (so it would clear). I used a straightedge to keep them as even as possible al the way across.

I got it dialed in about 5 tenths over about 10 inches, which is about all the good material in the old gib. I didn't make cut, but I plan on re-indicating tomorrow and trying a cut.

 

[Cadillac]

Using your old gib to reference the angle is what you want that's how you do it.

 

[ErichKeane]

Using your old gib to reference the angle is what you want that's how you do it.

Yep! That's what is on the top of that stack!