Ideas needed for problem wth Tom Senior E type Mill

[Mike CrossfieldParticipant @mikecrossfield92481]

I recently finished a major rebuild of a Tom Senior E type milling machine.  This is the light round column mill with the S type vertical head.

The machine as purchased was in reasonable condition mechanically, but with rough paintwork. It had good slides and no major wear anywhere.  Nevertheless I rebuild the head very carefully with new oilite buses and taper roller bearings.  I also fitted a 3 phase motor and VFD.

The machine was set up very carefully. Accurately trammed and with gibs adjusted for no shake. It runs very sweetly, with no strange noises.

Now the problem:

If I take a cut in the y direction everything seems normal.  For the sake of argument, with a 150 thou height of cut on the side of a EN3 mild steel block, with a very shallow 30 thou deep cut, everything goes very smoothly and quietly. (new 3/8in 4 flute cutter running at 600 rpm and about 3in/min feed.  Cutter held in MT2 taper ER25 chuck).

However, when I switch to an x axis cut, without changing any settings, I get a painful high pitched squeal for the duration of the cut. It makes no difference whether I traverse left to right or right to left (conventional not climb milling)

I have tried changing the depth and height of the cut with little effect.  If I reduce the spindle speed to 400 rpm, the frequency and level of the noise drops, but it is still there.  Increasing the speed to 700 rpm increases the frequency of the noise, and it is still painfully loud. Doubling the feed rate reduces the noise but doesn’t stop it. Changing drive belt tension does nothing.  The quality of the cut is not noticeably different in the 2 directions.

I cannot for the life of me think of a reason why it should cut quietly the y direction, but emit this load squeal in x.

Any suggestions would be very welcome indeed.

 

Thanks

Mike

 

 

 

 

[PeteParticipant @pete41194]

Did it mill normally before your rebuild Mike? Are you locking the Y axis slide before making a cut using the X axis? Then there’s X,Y feed screw adjustments if they are adjustable, or excess clearances from wear, even loose feed screw attachment bolts or screws. There could be other reasons, but cutting vibration and noise only when switching from the Y to the X axis seems to indicate it’s an issue with something loose or out of adjustment. And the Y axis can have an effect even when it’s not being used.

Did you disassemble the table, feed screws and Y axis? With a used machine there’s no way of knowing it’s past history or what may have been done in the past. I don’t blindly trust anything on a used machine as ever being 100% correct without checking. I think I’d also try indicating each axis by trying to twist and pull on the table in various directions while checking for any excess clearances. What are your X,Y axis back lash numbers?

[Mike CrossfieldParticipant @mikecrossfield92481]

Pete: Thanks for your message? The machine was partially dismantled and crated in anticipation of the previous owner emigrating. I bought it due to a change in his plans.He claimed it had been running perfectly, but I never saw it running. I completely dismantled the machine and checked everything for wear. No significant issues. Slides very good with signs of original scraping still present. Backlash on x and y 10 thou and 13 thou respectively. Gibs set very carefully after rebuild. Absolutely no shake or tightness at the ends of travel. Head rebuilt with new oilite bushes on pulley spindle and new Timken taper roller bearings on spindle. With new 3 phase motor and VFD it runs very sweetly. Until I cut!

I am pretty familiar with small lathes and milling machines and how to use them. I own both, and other machine tools as well. So issues such as locking unused slides etc is second nature. But I have never come across a problem like my current one. A friend has the same mill in good order, similarly rebuilt with new head bearings, and his machine does not behave like mine. The only obvious difference is that he uses morse taper collets directly in the spindle, whereas I am using an ER chuck plugged into the mt2 spindle taper, so the stick out on mine is much greater. He is lending me some morse taper collets to try out in the next few days. He suggested the problem might be related to the spindle bearing preload, which is quite tricky to set correctly, but if it were that, why would the noise only be in one axis?

???!

[noel shelleyParticipant @noelshelley55608]

My first thought would be the taper roller bearings, do they need more preload ?  As things come under load they can flex or move though seem fine when static. A matter often not appreciated. Noel.

[Rod ClemettParticipant @rodclemett60985]

I’m inclined to think that when cutting with the Y axis, the spindle is in line with the column, so the forces are  just pushing/pulling against the round column.

When using the X axis,  the spindle is offset from the column, so the cutting forces are tending to apply a twisting force to it.

Could the noise be related to some ‘bounce’ (for want of a better word) in the head, caused either by an issue in the column clamping arrangements, or a lack of  stiffness in the column itself?

Maybe using 2MT collets allows the head to be a few inches lower, so there is less ‘exposed’ column – and an increase in effective stiffness, so to speak?

Clem

[blowlampParticipant @blowlamp]

What happens if you carefully try climb milling?

Martin.

[old martParticipant @oldmart]

With the head bearings, you need to adjust the preload so that the headstock just reaches lukewarm after running at top speed for about 20 minutes and minimum grease should be used, there should be a minimum of 50% air space inside the quill, we run 90%. If you can reduce the tool overhang from the end of the spindle, it may help, and also try changing speed slightly in case resonance is a factor. Locking any unused axis helps and changing cutters may work. I assume you are using HSS cutters because we normally use solid carbide at max speed in our mills even up to 16mm diameter. Some soluble oil would be good for tool life and may quieten things down and trying different feed rates and cutting depths is worth trying.

Just because you use a VFD is no excuse not to choose the optimum belt ratio in the drive. Our Tom Senior light vertical has been converted from MT2 to R8 WITH A 1HP three phase motor set to run between 27 and 75 Hz from a VFD, but the four belt ratios are still used when needed.

[Mike CrossfieldParticipant @mikecrossfield92481]

Noel: preload on the bearings is high on my list of things to check since I’m running out of ideas.

 

Rod: might be something along the lines you suggest, combined with bearing pre-old perhaps? Everything is clamped up pretty solidly, and the problem is the same with light or heavy cuts, so I don’t think it’s a rigidity issue. And I have never seen a problem like this on my much more flexible Dore Westbury mill. It will be interesting to see if things change with the MT2 collets. If the noise is still there, then bearing preload is the next port of call.

 

Martin: I haven’t tried climb milling yet. If the noise changes, will it tell me anything I wonder?

Old Mart: bearing preload keeps cropping up, and though I can’t immediately see why the noise in only in one direction, it may be simpler to just try increasing the preload. I thought I had set it properly when I first rebuilt the spindle, but maybe things have slackened off after a short period of settling down? Running at high speed for a few minutes, the inside of the morse taper is barely warm. As I said in earlier messages, changing speed over range 400-700 rpm changed the pitch of the noise but didn’t stop it  and all unused axes were locked. All tests with new hss cutters. Brush applied cutting oil made very little difference. Spindle speeds were on the optimum pulley setting, with VFD frequency between about 33Hz and 60 Hz. I have VFDs on all my machines, so I I’m used to keeping the frequency in sensible range by changing pulleys.

 

Thanks all for the suggestions.

 

Mike

[blowlampParticipant @blowlamp]

Mike.

Conventional milling ‘rubs’ as each tooth enters the cut, whereas climb milling allows the teeth to take a clean bite into the workpiece. I would expect your machine to have an easier time climb milling, so it might be worth trying to see if there’s a difference.

As Clem pointed out, milling along the X axis means there is no solid metal behind the spindle to directly absorb the cutting forces, and this is fundamentally different to milling in the Y axis.

 

Martin.

[old martParticipant @oldmart]

It’s a difficult thing to assess as you are doing everything correctly.

Perhaps the way the work is held is a factor, if you are using a vise, it might be possible to clamp it down at right angles to it’s present position.

If you climb mill, you may have to put a little friction into the bed movement with the lock, and adjusting the friction might also work with conventional milling.

[HuubParticipant @huub]

Check for play in the quill sleeve. If the quill is locked by a bolt it probably still can wobble in 1 direction. The quill sleeve play of my brand new Paulimot F207V milling machine was 0.06 mm (0.002″) and even with the quill locked the spindle moved around when milling.

[Kiwi BlokeParticipant @kiwibloke62605]

We’re all floundering about in the dark, really, without being able to hear and see the machine in action, aren’t we? Two lines of suggestions have been offered: the noise is coming from the milling head – bearings perhaps – , or it’s the cutter chattering – possibly because of play somewhere. First step is to decide which…

I’d suggest that you rig up an arrangement whereby you can apply controlled(ish) forces to the spindle, and see what happens. Set up something like a short shaft in the chuck, and run it in a bush or ball bearing, fixed to the table somehow. Then, with the feed screws, apply force to the chucked shaft, and see how it deflects in different directions. Then run the spindle and again apply force, and see if the noise occurs. I’m sure you can improvise some set-up that allows this sort of test; it doesn’t need to be sophisticated. Also, apply different axial forces.

Please let us know what the solution turns out to be.

[Mike CrossfieldParticipant @mikecrossfield92481]

Kiwi Bloke

I didn’t mention it earlier, but one of the first things I tried was to mount a ball race in the milling vice and apply pressure on it in all direction using a piece of 10 mm silver steel in the chuck. No noise over a wide speed range, even when applying so much force with the lead screws that the silver steel was visibly flexing. So I don’t believe the noise is associated with the head.

Other things I’ve tried since my last message:

Increasing preload slightly on taper roller spindle bearings. No significant effect on noise, though bearing ran a little warmer. Maybe 10 degrees above ambient after 15 minutes at 2900 rpm.

Replacing ER chuck with MT2 taper collet.  So much less overhang. Only had one size to try – 8mm- but there was no noise over a range of cuts and speed. Not conclusive because all my previous tests had been with larger cutters.  When I tried this same 8 mm cutter in the ER25 chuck there was only the slightest trace of a noise, and only with tall side cuts. I have ordered up some larger sizes of collet to explore this further.

Out of interest I tried running the same ER25 chuck and original 9/16 cutter in my Dore Westbury mill with exactly the same set up that I had been using in the Tom Senior. When I took a shallow (20thou) but tall (0.25 inch) cut on the side of my test block I got a similar noise to the one I got on the TS. Not as loud, but same characteristics. No noise with 0.1 inch tall cuts.

Mike

 

[Grindstone CowboyParticipant @grindstonecowboy]

Probably teaching grandmothers to suck eggs (apologies if so), but do you have the ER collet properly snapped into the nut?

Rob

[Mike CrossfieldParticipant @mikecrossfield92481]

I have just made a short video of me taking a cut in the x direction. The painful noise is clear. Video is on the streamable hosting website at http://www.streamable.com /37bfq2. It will only be available for 2 days.

 

Mike

[blowlampParticipant @blowlamp]

It’s not going to be bearing preload because the machine cuts OK along the Y axis.

I think you are probably at the limit of what the machine can do with that type and diameter of cutter, depth of cut, and milling in the conventional direction.

What happened when you tried climb milling?

Martin.

http://www.streamable.com/37bfq2

[Mike CrossfieldParticipant @mikecrossfield92481]

Hi Blowlamp

 

I just tried a climb milling cut with identical conditions to those in the video, and there was no nasty noise.

What does this tell us about the source of the noise?

[Julie AnnParticipant @julieann]

Looking at the video the noise didn’t seem that bad to me. But two things struck me, the feedrate seemed slow and the swarf was very fine.

Both of those indicate that the cutter wasn’t being worked properly. When conventional milling the initial depth of cut per tooth is zero, increasing as the cut progresses. Since the machine, and cutter, are not infinitely rigid the cutter initially deflects and rubs rather than cuts. Eventually the cutter will start cutting, only to repeat the sequence as the next tooth comes round.

How conventional milling proceeds will depend on the stiffness of the cutter and of the machine plus how sharp the cutter teeth are. Even a new cutter has a radius on the teeth. It might only be microns, but it is there. Carbide cutters are better in this respect as they are stiffer than HSS cutters.

In summary the noise is a consequence of the cutting action; rub and cut, rub and cut and repeat. It should be possible to tie in the fundamental frequency to the spindle rpm and the number of teeth on the cutter.

Julie

[Mike CrossfieldParticipant @mikecrossfield92481]

Thanks Julie. Interesting observations. But can you explain why the noise only occurs with cuts in the x direction ?

[Julie AnnParticipant @julieann]

It’s a question of looking at the forces from the cutting process. These can be resolved into a force parallel with the cutting direction and a force perpendicular to the cutting direction. For endmilling with a reasonable depth of cut the force in the feed direction is normally the larger of the two forces.

When cutting in Y the larger force is primarily trying to bend the column fore and aft. The column should be fairly stiff, and the leadscrew behind helps stiffen the column in bending. When cutting in X the larger force is trying to twist the column as well as some bending. The combination of twisting and bending means the column is likely to move more for a given set of cutting parameters. Coupled with a small chip load this is likely to lead to cyclic movement and relaxation of the column; equals periodic noise.

Julie

[Mike CrossfieldParticipant @mikecrossfield92481]

Thank you Julie, that’s very plausible.

 

Mike

[old martParticipant @oldmart]

As it is not difficult to remove the quill, you might try degreasing the surfaces of the parts and lubricate them with spray chain grease which will damp any looseness that Huub mentioned.

[howardbParticipant @howardb]

“With the head bearings, you need to adjust the preload so that the headstock just reaches lukewarm after running at top speed for about 20 minutes”

I doubt that “lukewarm” temperature after 20 minutes is caused by bearing preload, it’s caused by the rollers and cage skidding around in the outer race.

You should be able to feel a temperature rise within a couple of minutes depending on mass of metal etc.

When setting a preload on bearings, you need some sort of visual indication that the  slack take-up adjustment has ceased and pre-load has begun,  ie when the inner race, cage and rollers are snugged into the outer race, the increasing reading will abruptly stop.

Using a sensitive dial gauge from the bearing housing to the end of the quill will tell you where this point is, then you can start estimating how much preload to add.

New bearings need more preload than used bearings, some large assemblies i have worked on actually specify this.

My experience tells me that you always need more preload than you think you need.

Build it up, run it hard it should get warm quickly, adjust preload accordingly.

 

[Julie Ann]

On Julie Ann Said:

It’s a question of looking at the forces from the cutting process. These can be resolved into a force parallel with the cutting direction and a force perpendicular to the cutting direction. For endmilling with a reasonable depth of cut the force in the feed direction is normally the larger of the two forces.

When cutting in Y the larger force is primarily trying to bend the column fore and aft. The column should be fairly stiff, and the leadscrew behind helps stiffen the column in bending. When cutting in X the larger force is trying to twist the column as well as some bending. The combination of twisting and bending means the column is likely to move more for a given set of cutting parameters. Coupled with a small chip load this is likely to lead to cyclic movement and relaxation of the column; equals periodic noise.

Julie

My immediate thought when first reading the post is basically what Julie said in her detailed reply above, to my mind it’s nothing to do with the quill or bearings.

Tony

[Mike CrossfieldParticipant @mikecrossfield92481]

I think I have now bottomed this one out.

I was able to set up exactly the same set-up that I had been testing on my mill on my friend’s identical mill. Same ER chuck, cutter, test piece, spindle speed, cut details etc. and the noise was exactly the same. Which would seem to rule out a specific problem with my mill such as bearing preload.

Yesterday I received the set of MT2 taper collets I ordered a few days ago, and when I tried the largest of these (12mm) with my biggest metric cutter (16 mm) I was pleased to find that the noise issue was absent under all the cutting conditions that had previously been problematic.

So my conclusion is that Julie’s explanation was correct: the noise was rub-cut-rub- cut and repeat. With the rubbing being exaggerated by slight deflection of the ER chuck/cutter combination, which protrudes quite a long way out of the mill spindle, and a low feed rate. The more rigid arrangement with the cutter in a collet inside the morse taper spindle socket essentially eliminated the deflection (and the noise).

When my machine was designed (in the 1960s I believe) ER collets weren’t around, so the issue I was having would never have arisen.

My thanks once again to all those who contributed their ideas.

 

Mike

 

[Author]

Posts