What's new
What's new

Studer RHU 450 from the 1960s

Hi Richard,

I decided to "prove" my findings by adding shims to compensate for the missing material under the swivel table. Here are the three shims, thickness 130 microns = 0.005", 75 microns = 0.003" and 25 microns = 0.001". None of these needed to be "forced" into place. This did improve the geometry, but not quite in the way that I had expected, so it was worthwhile.

Here are the shims:

attachment.php


This fixed the problem with the table drooping by 30-70 microns on the left. Here's a video showing that with these shims in place the swivel table is level. Indicator is 2 microns = 0.00008" per division (a bit less than a tenth).


Next I aligned the swivel table parallel to the travel (you have to click on the link since PM only allows one embedded video per post!)

Testing swivel table for parallel to travel

Then I aligned the workhead so that the spindle axes was parallel to the travel. This is WITHOUT the tailstock (you have to click on the link since PM only allows one embedded video per post!)

Testing workhead spindle for parallel to travel

Finally in the next clip I engage and disengage the tailstock. The tailstock is off in horizontal position by about 7 microns = 0.0003" (you have to click on the link since PM only allows one embedded video per post!)

Checking tailstock alignment front-to-back

BUT the tailstock is still too high, by about 70 microns = 0.003" (not shown).

Summary of the effect of shimming the swivel table to compensate for the missing material and make it flat:

(1) The workstock spindle now droops down rather than pointing up. This is completely consistent with what I see on the surface plate. Before shimming the swivel table, the tilting of the swivel table top was making the workstock spindle point up!

(2) The tailstock is now almost in the correct front-to-back alignment. Apparently the swivel table was twisting the workhead spindle axis when the table was clamped down.

So it looks like the right order to fix things is:

(a) Get swivel table ground flat and parallel

(b) Scrape the bottom of workhead to get the spindle axis tilting slightly up

(c) Scrape bottom of tailstock to lower it to the same height as the workhead spindle axis.

I need to do step b before step c. Step a can happen anytime.
 
Last edited:
It sounds like your machine is not worn in the Typical way.

In most cases the TS is low and one scrapes the HS 2nd. But if it is lower now, you can just scrape TS. Also leave the Head-stock a couple of tenths low so as the TS wears the machine gets better, other then making them the same and as the TS wears, it gets worse. The height of the Workhead and Wheelhead all depends on the parts OD your going to grind. The bigger the part the spec doesn't need to be to precise. Be sure to scrape one side before grinding anything so it is pulled down solid. Shimming is not the best way. Rich
 
Last edited:
Hi Richard,

It sounds like your machine is not worn in the Typical way.

Yes, I completely agree. And I think what I have found on the swivel table can not be "wear". It's ground, just not ground correctly! I think that the machine was restored or reconditioned, but at least a couple of mistakes were made in that process.

In most cases the TS is low and one scrapes the HS 2nd. But if it is lower now, you can just scrape TS.

I need to scrape both, because the axis of the HS spindle is pointing downwards, about 25 microns (0.001") in 150mm (6"). According to the Studer spec, it should be somewhere in between level and a maximum of 8 microns (0.0003") upwards-pointing in that distance. So first I need to fix the axis of the HS spindle, then I can do the TS last.

Also leave the Head-stock a couple of tenths low so at the TS wears the machine gets better, other then making them the same and as the TS wears, it gets worse.

OK, makes sense.

The height of the Workhead and Wheelhead all depends on the parts OD your going to grind. The bigger the part the spec need to be to precise.

I was thinking that it was the other way around, a small part would suffer more if the HS and wheelhead heights differ. Need to give this some more thought.

Be sure to scrape one side before grinding. Shimming is not the best way.

I am talking to a grinding shop in Germany that seems to know exactly what they are doing. So they may do this for me, because I don't have a surface plate big enough to print the table top.

Richard, I really appreciate your advice. I am sure that you are scrambling to teach and travel and are squeezing this in "on the side" under time pressure. Nevertheless, if at some point you do have the time for it, I would be very grateful for answers to the questions in my earlier post #140.

I'm happy with the progress that I am making. I think the machine will be in good shape very soon now, and I can spend my time using it rather than fixing/improving it.
 
Yeah I left out a word on my last one...I read it and fixed it...big difference... lol...

Yeah I am exhausted and didn't check for typo's before hitting enter...lol. After the CA classes, 2 in a row..back to back with one day to drive to new city. Then we flew to Austin TX when my Brother-in-law sold us some of his $$ furniture as he is moving out of the country. Alex and I (my son) rented a truck and drove it home. 1700 miles in 1 1/2 days. I am happy to have helped you. :-) ...going to go read #140. Rich
 
Hi Rich,

Thanks for taking the time, after two weeks on the road that's a scarce commodity!!

I started scraping the HS to shift the spindle angle from pointing down to pointing up. The flange is 160mm (6.5") in diameter and about 20mm (3/4") wide. The starting print was OK, contact most of the way around, not great. I split it into four parallel zones. Made three scraping passes on the zone farthest from the spindle nose, two passes on the next zone, one pass on the next zone, and no passes on the zone closes to the spindle nose. That reduced the downwards tilt by about half. Contact print is not as good as when I started. I figure I'll make another 3/2/1/0 pass to see if that gets the tilt to zero or a bit positive. If so, then I'll start scraping for better contact and worry about tuning the angle close to where I want it. Is that a good approach? Or after each 3/2/1/0 pass should I try to refine for good contact before doing another "tilt" pass?

It's great to have all the tools at hand from scraping the oil pockets a few months ago. Having blades of different radii and the power hone to quickly sharpen them makes all the difference. Plus this is easier scraping, good access, a much smaller area, I can turn the HS around to get a good scraping angle or better lighting. I don't think it's going to take me more than a few hours spread over a couple of days to get the HS into reasonable alignment.

Cheers,
Bruce

PS: You mentioned the weight of the test bar and "lift". I've calculated that the weight of the cylindrical portion outside of the taper is 2.3kg. So in measuring the tilt of the HS spindle, I hold a small digital scale under the end of the test bar and lift up until the scale indicates 1.15kg (half the weight). I think this should remove the effect of the test bar's weight "twisting" the spindle axis. If there is a better way, please tell me!
 
Last edited:
When the grinder was delivered, it was on a pallet which was almost exactly the same size as the base of the machine. Since I don't have lifting equipment in the right place, this made it a challenge to get the grinder off the pallet gracefully.

Now that I am getting down to the fine points of adjusting the geometry, I thought it was important to get the machine level and standing on the correct points. So after some careful work last weekend it is safely down off the pallet and standing as designed. (Note that the swivel table is off being ground and I am scraping the underside of the work-head to remove the droop.)

attachment.php


On the two front corners are adjustable screws with "ball ends" that go into ball sockets on the round feet. For a concrete floor that would be the end of the story, but my shop has 25mm OSB on 50mm of hard foam, so I have 300mm x 300mm (12" x 12") steel plates that are 15mm (5/8") thick spreading the load.

On the two back corners, there are the same adjustable screw feet, but they are only there as a "safety" to protect against tipping:

attachment.php


The load is carried (as per the manual) by a 70 x 100mm (3" x 4") steel block in the center. So the screws at the two rear corners are adjusted hand-tight then backed off a quarter turn.

I have the machine level according to my antique-but-still-good South Bend 12" lathe level. Since the machine is on a 3-point support, that should be good enough.

I wonder how long it has been since the machine was standing as it was intended to stand. Hopefully this will give the castings a chance to relax back into their original shape.

One nice thing about this foot design is that it makes it easy to get a pallet jack under the machine if I want to move it. I tighten all four jacking screws to lift the machine 25mm (1") then hammer a couple of wooden wedges under one corner to support the load, back off the screw in that corner, and add a 25mm thick wooden square underneath the square steel foot. Repeat for the other three corners, then start the cycle again. In 15 minutes the machine base is 85mm = 3 1/2" off the ground and I can push a pallet jack under it to move it. To lower the machine, just reverse the procedure.
 
Last edited:
I also teach about 3 Points as I believe their use is also a "Lost Art" Talk to an old timer and they will show you how they used 3 points to mount parts in mills and grinders and blocked around it. All the precision machines of old and many now use 3 points to stabilize and self level machine bases using 3 main points and smaller stabilizers much like your grinder base. I am so happy you have it off the skid...:-) Keep up the good work. Rich
 
Hi Richard,

Keep up the good work.

Thanks! Things are going well.

I got the bottom of the swivel table ground by a local place called Hoppeschliff. They did a good job. No more need for the shims on the left side and now when I traverse the table left/right I get about 3 microns of height variation (slightly more than 0.0001") over 500mm (24").

I also step-scraped the bottom of the workhead to get the spindle level. I got to do it twice, because the first time I finished it I had the spindle level and good contact over 3/4 of the mounting ring, but none on the other 1/4 :angry:. I made a classic beginners mistake, and started refining too soon. So I had to rough-scrape the 3/4 region a few times until I got bearing all around then redid it. I found that if I sit close and get the light right, I can see the high points because the blue is thinner and they are surrounded by a darker blue ring.

I also ground the mounting plate that goes between the workhead and the swivel table because it was not parallel. That also came out well, it's now parallel within 2 microns (better than 0.0001").

attachment.php


Should I scrape oil pockets into the bottom of this mounting plate (the side shown above) where it bears on the swivel table?

Now I am turning my focus to the tailstock. It's badly out of alignment in both axes. The good news is that the taper is concentric with the bore and with the barrel, and the barrel is a honed fit with no play. I found a serial number stamped under the tailstock (602) which is 15 larger than the rest of the machine (587) so they were not factory-matched. Fortunately I have a good MT2 test bar and it was easy to measure the tailstock both on the surface plate and on the machine itself. I was able to confirm the measurements on the machine by shimming. Here it is with the shims in place:

attachment.php


To get the tailstock parallel in the horizontal and vertical axes will require removing 0.08mm = 0.003" from the left side of the bottom (tapering to zero on the right) and removing 0.24mm (0.009") from the left side of the front (tapering to zero on the right). It's a lot to scrape off so I am thinking about putting together some kind of clamp to hold the tailstock body in the surface grinder. On the other hand the bottom of the tailstock is small and accessible, so step scraping might not take so long. But the surface grinder clamp might pay off because after I get the tailstock parallel in the two planes, then I may need to remove more material to bring it into alignment with the workhead.

Cheers,
Bruce

PS: I think I'll step scrape the bottom of the tailstock rather than grinding it. I think that will be faster than making a fixture to hold the tailstock. For those who don't know what it is, here is a nice short video by Jan Sverre Haugjord about step scraping.
 
Last edited:
Jan was my first European student. He flew over to the 2013 Columbus GA class. He was the first one to talk about me on You Tube....he sometimes calls me the professor....lol

Ballen it looks good....I am a bit concerned with the index pin alignment though. Why not talk about that in here to keep the instruction of the rebuild all in one place?
 
Hi Richard,

Thanks for sticking with me on this.

Jan was my first European student. He flew over to the 2013 Columbus GA class. He was the first one to talk about me on You Tube....he sometimes calls me the professor.

You are creating quite a legacy!

I am a bit concerned with the index pin alignment though. Why not talk about that in here to keep the instruction of the rebuild all in one place?

I had the impression that no one else but you is reading this thread anymore (Probably it should have been in the "Machine Reconditioning, Scraping and Inspection" forum). Also, I was afraid that I have overloaded you with questions. (You did not reply to my questions about "checking the lift" in post #140 or about scraping oil pockets in #149. I know that you have classes to teach and a very full life and am afraid to impose so much!)

Getting back to the index pin, I have the impression that the alignment hole is a bit deformed but will take a good look at it and in particular measure how far off it is. If the right thing to do is to scrape the workhead so that it's correctly aligned with that pin in place, I will do that. First will shim and measure as you suggest.

Cheers,
Bruce
 
I had the impression that no one else but (Richard King) is reading this thread anymore
Nope, I've been following all this with interest. Very few projects have been presented anywhere near as thoroughly as you are showing us this Studer project.
 
Ballen it looks good....I am a bit concerned with the index pin alignment though. Why not talk about that in here to keep the instruction of the rebuild all in one place?

OK, I'm moving new stuff back into this thread. I've made more progress. I've machined and scraped the HS and TS into the proper alignment. Vertical and horizontal tilt and relative alignment are now all within the Studer "new machine" specs. Quite a bit was needed, including removing about 80 microns = 0.003" of height (and a similar amount of tilt) from the TS and moving the HS back about 240 microns = 0.009". I'm getting better at scraping. What I've been doing is first removing the tilts by step scraping and then lowering the part and finally finish scraping for good contact. I now have a good feeling for how much I need to leave behind after the lowering step so that there is enough remaining to scrape for good contact.

I still need to check the height of the grinding wheel spindle and adjust it if needed. But before I get to that I want to address one more thing I noticed along the way. The rotation pin for the swivel table has a lot of play (about 30-40 microns (0.0012-0.0016"). This is bad because I want to set the swivel table for grinding tapers using gauge blocks. Play here means that one has to account for backlash, which complicates getting the settings right.

Here is how the rotation pin is built. The traversing table has a 16mm pin sticking up in the middle:

attachment.php


I removed the pin:

attachment.php


Here is the pin and the underside of the swivel table:

attachment.php


The pin is slightly worn (10 microns = 0.0004") but the matching hole in the swivel table is worn oblong about 30 microns = 0.0012". So I am going to bore out the hole in the swivel table, enlarging it by 100 microns = 0.004" and then I'll turn a new pin from prehardened 42CrMo4V (ANSI 4140) and lap it to fit. This area is protected from coolant and grinding dust and air, so if I lubricate it properly it should not rust inside and lock up.
 
The repair of the swivel pin went well. First I bored out the hole to 17mm (0.669") and turned down the 16mm pin to 15mm, and then I made a Delrin plastic bushing to fit in between.

attachment.php


Also shown is a threaded 6mm (1/4") thick "puller disk" with a diameter of 16.9mm. This goes behind the bushing, and is threaded M6 so that it can be pulled out with a slide hammer. This disk provides an easy way to remove the bushing. I made the plastic bushing so that the ID was a loose fit over the rotation pin, but so that the OD had negative clearance in the cast iron. When I tapped the bushing into the hole, this compresses it to get a slight interference fit over the pin. I added a drop of cyanoacrylate (super) glue to ensure that the bushing does not move in the cast iron hole.

attachment.php


For good measure I filled the cavity behind the puller-disk with way oil and plugged the threaded hole with a felt disk. This should do a reasonable job of lubricating the pin.

attachment.php


There is no longer any "lost motion" in the swivel table. If I put an indicator at both ends and push or pull either end, the indicator at the other end follows without any hesitation.
 
Last edited:
Are you not concerned about the squish factor with the Delrin? ,I have the same problem with a 6ft machine but the wear is more like 10th.
 
Are you not concerned about the squish factor with the Delrin? ,I have the same problem with a 6ft machine but the wear is more like 10th.

If I see backlash or free play developing in this bearing, then I can swap out the Delrin for a standard "solid bearing" insert. These are available with 15mm ID and 17mm OD, so in principle it's "drop in".

My understanding is that Delrin does not cold flow, and has an elastic modulus of about 3 GPa = 400,000 psi. Since the bearing insert is 1mm = 1000 microns in thickness, this means that to make it compress by 1 part in 1000 (1 micron) will require about 400 pounds per square inch. The bearing surface is about 0.7 square inches, so to compress the Delrin shell by 1 micron requires about 280 pounds of force. Now when I push the end of the swivel table, I am applying few pounds of force, well below this value. (The "leverage" increases this somewhat but the fulcrum of the lever is not at the pivot point, so it's not enormous.) So the pressures being applied are well below the yield limit of the Delrin, and the deformations should be small, even on the micron scale.

I have tested this bearing for backlash as follows. I have a precision dial indicator permanently mounted at the right end of the swivel table (divisions are two microns = 0.00008"). I also mounted a test indicator on a mag base at the left end of the swivel table. I set both indicators to zero. Then I pushed the table to +2 microns at the right end, and saw that the indicator moved to -2 microns at the left end. Then I pushed the table to -2 microns at the right end, and saw that the indicator moved to +2 microns at the left end. Then back to 0 microns at the right end, and the indicator at the left end moved to zero microns. So at least right now, the swivel table has zero play. It's a lot better than it was, where I could move the table tens of microns at one end and not see motion at the other end.

I'll test this from time to time in the future. If this bearing develops play or backlash, then I'll replace it with a manufactured solid bearing or in the worst case just make up a new metal pin which is a precision fit in the (new) 17mm bore. A good solid bearing choice would probably be an Iglidur GSM-1517-10 (see this catalog and page 70 for details).
 
I think the Delrin idea is a good one that I might have to steal. As you say if it does not work there are other options ,thanks Bruce!
 
I think the Delrin idea is a good one that I might have to steal.

The idea of a Delrin insert came from some other people here on PM, so I can't claim credit.

An interesting fact from the Iglidur catalog is that for lowest friction the pin should have a surface roughness of about 0.8 microns Ra (rather than a mirror finish). See the plot on the bottom right of page 66. I don't know if that's also the case for Delrin, but would not be surprised.
 
Dear Richard,

Here is a video showing the HS and TS alignment. I found it fun doing the last bit of alignment tuning because it's quick and easy in comparison with the large adjustment phase that comes before. I now understand why you normally do the HS first, because removing it and putting it back is a hassle. So better to do the "final tuning" on the TS, which is easier to get on and off.


The last thing I need to do is to get the grinding wheel spindle at the same height as the HS/TS centerline. According to the Studer spec these should be within 100 microns = 0.004" of each other. This is also what Schlesinger calls for (No 21, Chart 3, Axis of wheel spindle and headstock at same height with respect to swivel plane.)

I was preparing to make a pair of precision rings which were cylindrical on the outside and tapered on the inside when I realized that I could do a precise enough measurement using the wheel hub for this. Here is the setup:

attachment.php


The cylindrical portion at the outside of the wheel hub is 75.00mm in diameter and has 2 microns (0.00008") TIR , so the top is 37.50mm above the grinding spindle centerline. The collet holder in the HS is 40.00mm diameter on the outside, so it's top is 20.00mm above the HS spindle centerline. I set a dial indicator stand on a pair of 18.00mm high parallels and zeroed the indicator on the wheel hub. Then I moved to the top of the HS collet holder (after removing the 18mm parallels). If everything were correctly aligned the indicator would move up 0.50mm = 0.020". In fact the indicator moved down 0.02mm = 0.001". So the grinding spindle centerline is 0.52mm = 520 microns = 0.021" too high.

(Note: this proves that the machine has had work done to it after leaving the factory but before I got it. In my adjusting process I scraped about 30 microns from the bottom of the HS to correct drooping of the spindle axis, ground about 50 microns from the mounting plate to get it flat, and had about 140 microns ground from the swivel table to get that flat. So in total I lowered the HS axis about 220 microns. Since now it is about 520 microns too low, I conclude that before I got the machine, someone must have removed about 300 microns from these. That also explains why the TS was about 250 microns high compared to the HS. The TS must have been replaced with a new one, but was not adjusted to the HS.)

So now the question is, where should I remove 520 microns = 0.021" of material to lower the grinding wheel spindle height? This is enough that I want to machine off (say) 480 microns and grind or scrape off the last 40 microns. The four possible locations are shown on the next two pictures, labeled 1 to 4 moving upwards. In fact there are 8 possibilities since each location has an upper and lower surface.

attachment.php


attachment.php


Location 1 is the fixed V and flat slideway on the machine base
Location 2 is the circular rotation table above that
Location 3 is the dovetail+gib slideway above the rotation table
Location 4 is the bottom of the wheel spindle casting

I believe that the best solution is to mill then grind the bottom half of location 4 or the top half of location 2, because the pieces are fairly easy to handle, align, and hold, and are simple: they do not need to be scraped to be sliding bearing surfaces. (I need to check that removing a half millimeter won't interfere with the clamp/lock mechanism, or that I can trim those if needed.)

Comments please. Where is the best place to remove 520 microns = 0.021" of height?

PS: these old photos from January are also relevant here, since they show this area of the machine:

attachment.php


attachment.php
 
Last edited:








 
Back
Top