CMT Ursus 250 Repair: 1) Clutch

[RobinHood]

Peter, here is a sneak peek at a bushing I have already done - it is part of a compound gear that lives in the apron feed gear train. The oil enters the bearing through two holes in the root of the gear teeth. The mating gear acts like a pump and forces the oil through the hole into the oil groove in the bushing. Very clever design. The old worn bushing is on the left.

 

[RobinHood]

Success! (on the bushings)

Set the baskets up in the 4J and took some time to indicate them as best as I could to minimize final runout. Radially I got them down to ~1/2 needle width on a half thou indicator and about 1 needle width axially. Then I carefully bored them out. Sneaking up on the final dimension. I opted for shaft OD 32.99 + 30um for the large journal bearing and OD 27.99 + 25um for the small end.

Then I moved to the mill to broach the oil passages. This is the tool I previously ground.

Did the end ones first, then the bore grooves.

Decided to have the grooves clocked by 90* between the front and rear bushings on each basket. Oil enters in the center radial groove (That was produced by making the bushings “too short” so that they did not touch in the middle).

And a neat little pile of chips left after broaching an axial groove. 1.5 thou DoC per stroke. 40 thou deep. Bronze cuts like butter - love it.

I then took some time to remove any burrs and round over the transition from the groove flanks to the inside bore so that the oil film will not be scraped off the shaft by a sharp edge.

I tried the finished part on the shaft with some oil and I can feel a thin film is forming between the moving parts. Heavy way oil produces more friction than the lighter hydraulic oil (which the manual calls for, and I’ll be using).

Very happy how the bushings turned out.

The heavier wave spring washers have arrived. I’ll be working on them next...

 

[RobinHood]

Great news: the new spring washers work!

I just ground two to size to try them out (they fit into the space available - but are still over on both the ID & OD when compared to the OE washers). They are as strong or stronger than the OE. They are separating the plates perfectly and when the clutch is closed, they close with the OE ones. So they must be very close. I’ll do some fine tuning and testing tomorrow and take pictures.

 

[RobinHood]

I’ll do some fine tuning and testing tomorrow and take pictures.

Tomorrow ended up being two days of messing about...

The testing revealed that yes, the clutch closed just fine, but the driven plates that had the stiffer, new washers were not being clamped. What gives? Adjusting the clutch did not really help either as once the two plates with new washers were being driven, the 5 others were dragging when the clutch was open. So that won’t work.

I suspected it had to be the spring rate of the washers being different from the OE.

So I used the method suggested by @cuslog with the bathroom scale on the mill table.

I used the nuts off the grinding wheel hubs - they are surface ground flat and parallel on both sides.

Worked very well. The springs were surprisingly consistent amongst the same type. I won’t bore you with the numbers, suffice it to say that the OE and new were not close at all. The fully compressed force for the OE is about 60#. The new, heavy ones require about 20-30 thou more travel to reach 60#. Plus when they deform, the geometry is such that they only compress to about 50 thou (a driven plate in which they nest is 44 thou thick). One can get them flatter, but then the force goes >150#s! So that was what prevented them from closing properly.

This is a OE spring, not compressed & compressed

Note how nice and flat it ends up

Here is a new, heavy spring. Note how the edge bends up and rotates all the way to the top plate when it bottoms out. More force can be applied to flatten it, but that will never happen when installed in the clutch pack.

You can even see some light coming through in other areas.

The solution I ended up with was using the weaker new springs, I ground seven of them + 1 spare. I installed them on the REV (B) side. They deform in the same way as the stiffer ones, but being of thinner cross section, they close enough before bottoming out. That is why they seem to work - plus all 7 of them have the same constant, so deformation is evenly spread across the clutch pack side. Final test will be with the clutch installed and properly adjusted, but I am very confident that it’s going to work.

Here is the finished clutch ready for install when the time comes: all new bushings, and one side (B) has all new wave spring washers.

 

[PeterT]

It good you could do those tests. Nothing like real life data to see what's going on.
I'm not sure if those springs have a reference free length but generally something you have to pay attention to. It might be approaching or exceeding yield as they bottom out / flatten & that might explain non-linear behavior, but I actually don't know what the base curve is supposed to look like. If they returned to their original thickness then you are OK.

 

[RobinHood]

The height is listed as 0.197“ and the height at working load of 65# is listed as 0.097”. That is for a spring as purchased. I ground about half of their widths off (between opening up the ID and reducing the OD). So no official data at that point. Only what I measured/observed. I think they will be fine as they will be “normally open” when the clutch is not engaged. They will only be compressed during a cut with the spindle running. And the weaker ones that I am using are not maxed out. Fingers crossed, I guess.

 

[RobinHood]

Other Clutch Components: shift pads & fork pin

Did some more work on this lathe. The old shift fork pads had wear on one side each - probably from the clutch not working properly because of the broken springs. So they just adjusted the mechanism to always put pressure towards the ”closed” side using the lever. That’s not how these things work: the lever “feathers” the clutch towards engage only until the shift block reaches the “overcenter” point on the shift ramps, then it snaps in and locks the clutch plates. So there is no more relative movement between the driver and the driven plates ==> no wear/abrasion at all at this point. Thus the clutch also does not heat up. Once the overcenter point is reached, there is also no more side pressure applied to the shift block by the shift fork pads. The activating lever mechanism has both a FWD and REV spring loaded ball detent. It happens to be located at the right side of the QCGB where the clutch lever shaft enters (there is also a secondary lever at this location - it is missing on my machine - so will need to make a new one). All linkages from that point on are adjustable. You set the lever into a detent and adjust the throw so that the shift block reaches and goes beyond center on the clutch pack. It is fiddly work and takes time. The clutch itself is adjusted as well during this process to give the required acceleration time. I do not have the factory specs for the CMT, but on the Colchester Master it is 2 to 2.5 seconds from standstill to 2500 rpm with an 8” chuck and no work piece. I expect it to be about the same with a 12” 3J to go from 0 to 1000 rpm (top speed of the CMT when run at 50 Hz - I will get a bit more than that as I’ll run it at 60 Hz). So proper adjustment of these clutches is crucial for long life.

Here are some pictures:

The pads are bearing bronze. Test fitting the turned pads in the fork

Off to the mill and spindexer to produce the required width. A collet block would have worked as well for this op; spindexer is faster though.

You can see the worn sides on the old pads.

Trial fit on the clutch shift block

The next part on the list was a new fork-to-activating shaft taper pin. This is a little unique as it needs to have the thick end threaded for future removal as the thin end is never accessible because of the shift block. These pins are not hardened. So a 3” long 3/8” GR5 bolt was the raw stock. The taper is 1/4” per foot included. So I set the TA to 1/8” and adjusted for a good, full contact fit in the fork. Could have used a shorter bolt as the pin is only about 30mm long (plus the threads). I wanted enough stock to be able to sneak up on the dimension.

Here we are getting ready to part off the excess. Yes, I did grip the bolt by the threads. 3/8” is small enough and fits nicely into the “hollow ground” chuck jaws - almost like using a collet.

This is how things go together. The shift fork is discoloured because I needed heat to get the old pin out. It was badly mangled from uneven wear and did not budge. I ended up drilling it out. Oh, the hole is drilled at an angle from factory.

Installed

Looking in the top - only the FWD clutch is visible. The REV is on the other side of the rib. The piece of wood is holding the REV clutch gear temporarily in place. This gear needs to be placed in position even before the lower end of the head stock is assembled as it is trapped in this location. The last shaft to go in is the brake shaft (to which the rev gear mates) - it is still out in this picture.

 

[johnnielsen]

You have quite the analytical mind. Well done.

 

[John Conroy]

Another machine that will be better than new when you are done with it. Nice work!

 

[PeterT]

Coming together nicely!

Tell me about the holes in the chuck jaws. Do they hold dowel pins for special clamping purpose or?

 

[RobinHood]

Do they hold dowel pins for special clamping purpose or?

I “drilled” those holes using a carbide end-mill (because I don’t have any carbide drills; the jaws are hard) to insert dowels. The dowels then held a special pre-load ring so I could re-grind the inside of the jaws.

This is a Röhm 3J chuck that came with the Colchester. It was in really bad shape and was stuck on the spindle. The cams would not even turn. Probably took the better part of a week to get it off. Once off, I was able to rebuild the chuck; luckily, the scroll is in really good shape and very accurate in its entirety.

 

[RobinHood]

Another machine that will be better than new when you are done with it.

I probably would not go that far.

There are items that will be / have been corrected and improved to hopefully mitigate wearing out as they did (details to follow later). But then again, I don’t know how hard this machine has worked in its life.

My goal is to get it as accurate as when it came from the factory back in the ‘70s; this will be a tall enough order.

Here is what the manufacturer claims this lathe is capable of (units are mm):

 

[RobinHood]

You have quite the analytical mind.

Thanks John.

 

[YotaBota]

Morning, I know this is a bit late but I came across this site when looking for other stuff and wondered if you'd been here.
http://www.keithprice.com.au/index.html
He's advertising Ursus parts and pieces.

 

[RobinHood]

Thanks a lot for the link, Mike. Much appreciated. It‘s always good to have another source of parts.

A quick update: if I can get the electrical sorted out on the lathe, I will be able to try the clutches under power. They work great when I turn the input shaft by hand.

The wiring diagram I have is for a newer version of the lathe (one with electric rapids in both Z and X) and does not match what I am seeing in the control box. The relay coils seem OK when measured, but they don’t energize when I supply the control voltage to the control circuit. I need to chase each and every wire to see what’s up. There are no shorts and no magic smoke left the box when I hooked power to the lathe. Could be some sort of interlock perhaps or a bad momentary switch (although each component seemed fine when I had them out on the bench for cleaning/testing.

We’ll get there…

 

[YotaBota]

You're welcome. I was cleaning up bookmarks and thought maybe they had the wiring diagram to match your machine.
Does your machine have the 4" spindle bore?

 

[RobinHood]

Yes, it does. Main reason I bought it. Most of my work involves “sticks” 4” and smaller.

 

[Tom O]

I had a Modern tool lathe with a 4 1/8” spindle bore in the early 90’s I miss that lathe!

 

[YotaBota]

I started with the craftsman 109 - no spindle hole, then the Altas 618 - 1/2 inch, Logan 400 and SM9" - 3/4 inch and now the SM1120 at 1 3/8 inch bore, I hope this hobby doesn't end up like boats and 2footitis,,,,,, lol

 

[hotrod1442]

I I'm trying to locate a supplier for the clutch plates for a CMT Ursus lathe, could you point me in the direction of your contact in Italy?