What's new
What's new

DIY inside grinding attachment for cylindrical grinder, what design?

ballen

Diamond
Joined
Sep 25, 2011
Location
Garbsen, Germany
I'm giving up on getting hold of a "factory original" inside grinding attachment for my cylindrical grinder. So am thinking through various ways of making my own.

The holder for the spindle or spindle motor is not difficult. I'll probably weld it out of some steel plate. This will mount on top of the grinder head, hinge up out of the way.

I could try to reproduce the original design. This had a 3000 rpm 400 Watt (1/2 HP) induction motor with a 250mm (10") diameter pulley driving a thin flat belt, then (for examples) spindles driven with 25mm (1") pulley to get 30,000 rpm.

Or I could use a more modern design: integrated motor and spindle unit with a VFD controlling the speed. One choice might be a 4,000 - 24,000 rpm spindle motor with (say) an ER16 or ER20 collet setup. Another way would be a Precise/Fischer spindle motor with an SRP6 collet setup (max 7.5mm ~ 5/16"). Or the same thing with a threaded arbor interface rather than collets.

For grinding shallow bores and tapers, I think any of these would work fine. But one of the things I want to be able to do is to clean up the short MT5 workhead taper on the grinder. This goes from 38 - 43mm over a bore depth L=100mm. So I could grind this with a 30 or 35mm wheel, or something a bit smaller. Question is, what's the minimum diameter of an arbor that goes this deep? About the longest I have seen in catalogs has L/D=6, which means something like 16mm diameter.

Has anyone here put together an internal grinding setup like this? How did you do it? Are you happy with the result? Can you grind a 38-43 mm diameter taper 100mm deep and get a mirror finish?
 
Hi Ballen:
I've tried it both ways; belt driven from the main spindle and integrated NSK electric spindle with collet.
Each has advantages and disadvantages, the NSK was wimpy and the belt drive unit was clumsy.

Moving on to getting a "mirror finish"; as others will point out too I'm sure, there's a lot involved, especially internal cylindrical grinding, because unlike any other precision grinding process the wheel has a bigger zone of contact with the work, and the size of the zone depends on the ratio of wheel diameter to bore diameter.
The consequence is a greater tendency to glaze, to overheat, and to burn with all of the bad things that come from that.

So what you're looking for in the spindle is rigidity and concentricity as well as reach, but there is much, much more to it than that in order to get what you want to get.

However, if we're talking only about the influence of the spindle, then there are two mutually exclusive considerations that will impact your choice:
1) Can I get a spindle small enough to have the spindle go into the bore so the front bearing is close to the wheel.
2) Can I make an extension strong enough not to flex too much if the wheel needs to be hung out a long way from the front bearing.

I have seen it done both ways, but obviously if you can go with scenario 1) you have a better chance to get a good result if the bore is very long, and for that scenario the smaller the unit the better, so the NSK wins out because it's only 25 mm diameter.
However, a 30 mm wheel on my NSK is pushing it, and the bearings are not robust enough to handle it if the wheel is imperfectly balanced or has a misadventure.

The vast majority of internal grinding spindles use extensions and gazillions of bores are ground every day that way.
Unless your work is very small (mine is) I suspect you'll be happier using a more robust unit.

Something like a Precise direct drive spindle is very nice except for one thing; the damned motor hanging off the back.
It makes the unit tail heavy and long, so you may find it a nuisance.
It has to be small enough in diameter that you can set up the centerline of the spindle at the same height as the centerline of the workhead without crashing into the table or the tailstock.
You can get past the tailstock issue by taking it off to internal grind but it's a PITA to do so.

The belt drive units do not have that problem but they're a minor pain to set up: even the swing-down kind is a bit of a nuisance, but compared to humping an integrated spindle into place and bolting it down without banging it around, maybe it's not so bad after all.

At a guess, I'd say you'd be happiest for general occasional internal grinding with a swing down unit and belt drive from the main spindle like everyone else does too.
Your deep bore is well within range to grind with an extension.
You can buy complete engineered internal grinding spindles from Duplex, or Themac or Dumore among others to save you having to build your own or bodge something together, and they are nice spindles that will do what you want.

On Edit: I've attached a couple of pictures to show what I mean about the two types of setup for deep internal grinding.

Cheers

Marcus
Implant Mechanix • Design & Innovation > HOME
Vancouver Wire EDM -- Wire EDM Machining
 

Attachments

  • long spindle.jpg
    long spindle.jpg
    4.6 KB · Views: 387
  • short spindle with extension.jpg
    short spindle with extension.jpg
    5.6 KB · Views: 360
I've started putting together a drop-down inside grinding attachment. This will hold a grinding spindle of up to 60mm (2.4") diameter, and a drive motor, connected with a flat belt. It will swing down for grinding and up to get out of the way for normal outside grinding. It's very conventional with one exception. Rather than a ~3000 rpm motor with a large pulley (250mm/10") I am using a brushless DC motor with hall-effect sensors and a variable speed drive. The motor goes up to 12000 rpm and uses feedback to control its speed. This means that I can get by with a smaller 60mm (2.4") pulley on the motor side and have variable speed control without changing pulleys. So the whole setup will be smaller and lighter than a factory original one.

Before doing a design on paper or with a CAD program, I am spending a few hours making a wooden mock-up. For this project that's a better way for me to check fit/interference and play around with proportions.

I have a question about the location of the ID grinding spindle relative to the OD grinding wheel (300mm/12" diameter). This is from the point of view of the operator. When I drop down the ID spindle, the business end of the ID spindle will be just to the right of the OD grinding wheel. My question: how far forward of the OD wheel (towards the operator) should the center of the ID spindle be? Obviously I can remove the OD grinding wheel if I need a lot of clearance for a part, but for most work I would like to avoid it. So I was thinking that for most purposes if the center of the ID spindle is about 30mm (1 1/4") forward of the front of the OD wheel, that should be reasonable clearance. Does that sound right? The logic is that this amount of clearance would allow a long (er) 60mm diameter ID grinding spindle to completely clear the grinding wheel and "stick out" to the left of the grinding wheel.

Could someone with an ID grinding spindle on a similar size machine (100mm/4" center height, 300mm/12" grinding wheel) measure that for their machine?
 
Hi B,

I've been wrestling with this one for a while, now. Marcus' reasons are precisely why, too. I've thought about making a VFD speed controlled motor ( like a Precise with VFD ) and using that as basis, and also considered using a DuMore spindle and driving with belt. The problem for me has always been my aversion to spending more on the ID attachments available than I spent on the grinder itself, and balancing that against the time and materials I'll spend ( and waste ) on the effort to cobble something together. I'm still wrestling the decision, so am interested in what you come up with and how it works. Our early Myford is already a very frequently used machine in our shop, and having ID spindle for it would be a very welcome addition, indeed.
 
I'm still wrestling the decision, so am interested in what you come up with and how it works. Our early Myford is already a very frequently used machine in our shop, and having ID spindle for it would be a very welcome addition, indeed.

When it's finished, I'll post some photos of my wooden mock-up.
 
I haven't had much time to spend on this, but thought that I would post a few pictures of the current status. I am first making a wooden mock-up of the internal grinding attachment. For me this is the best way to make sure that the proportions are OK, that it does not interfere with other parts, and to figure out some of the details of the design. When I'm happy with it, I'll make drawings or sketches and then make it out of aluminium, steel and cast iron.

The most unusual thing is the motor. Rather than the traditional approach of using a 3000 rpm 1/2 HP (400W) induction motor with a 250mm (10") diameter pulley, I am using a 12000 rpm variable speed 400W brushless DC motor with a 60mm (2.5") pulley. Here is the test bench:

attachment.php


The motor drive/controller is powered by 48VDC, and is drawing about 3.5A from the power supply, so about 168W. The motor contains 3 Hall-effect sensors, and thus has a feedback loop for controlling the motor speed as the load changes.

I recently regreased the spindle (40mm SJF/Fischer UJ40) and am monitoring the temperature, here about 42C. The LED display on the motor drive is showing 12,000 rpm. Since the spindle has a 25mm pulley, this corresponds to a spindle speed of 12,000*60/25 = 28,800 rpm.

Here it is in place on the machine:
attachment.php
.
The 40mm spindle is sitting inside a 60mm adaptor. I have sized this so that I can use up to a 60mm spindle and clear the front of the grinding wheel. I'm probably going to flip over the big clamping block so that the slit is downwards rather than upwards, to help keep coolant out of the spindle area.

Here is a side view. In this photo and in the last photo, you can see some wooden blocks underneath which I was using for support while messing around with different heights and offsets.
attachment.php

One of the tricky parts is a clamping arrangement. If you look closely by the machine head you'll see I have cut out a v block with a matching V slot. The V block will be attached to the grinding head either with epoxy or with a couple of bolts. The next step is to make a clamping bolt that goes down diagonally to engage the V block. The final piece to work out is a clamp for the hinge along with a retainer hook for when the inside grinding attachment is hinged back.

attachment.php


Once I have these clamps in place I'll post a couple of more photos.

attachment.php


Anyway, I have 4 parts left to mock up: 2 clamps, a retainer hook, and a shield/cover for the belt. I might also remove the cooling fan from the motor and fit a separate small 12V cooling fan with a shroud, because I think it might significantly reduce the noise and motor vibration.

I need to run the cabling and figure out where to mount the 48V power supply and the speed control/display. I also want to figure out a simple mechanism for tensioning the beit adjusting the angle of the small motor.

The idea is to sort out all of these details with the mock-up before I start to make drawings/cut metal.
 
I haven't had much time to spend on this, but thought that I would post a few pictures of the current status. I am first making a wooden mock-up of the internal grinding attachment. For me this is the best way to make sure that the proportions are OK, that it does not interfere with other parts, and to figure out some of the details of the design. When I'm happy with it, I'll make drawings or sketches and then make it out of aluminium, steel and cast iron.

The most unusual thing is the motor. Rather than the traditional approach of using a 3000 rpm 1/2 HP (400W) induction motor with a 250mm (10") diameter pulley, I am using a 12000 rpm variable speed 400W brushless DC motor with a 60mm (2.5") pulley. Here is the test bench:

attachment.php


The motor drive/controller is powered by 48VDC, and is drawing about 3.5A from the power supply, so about 168W. The motor contains 3 Hall-effect sensors, and thus has a feedback loop for controlling the motor speed as the load changes.

I recently regreased the spindle (40mm SJF/Fischer UJ40) and am monitoring the temperature, here about 42C. The LED display on the motor drive is showing 12,000 rpm. Since the spindle has a 25mm pulley, this corresponds to a spindle speed of 12,000*60/25 = 28,800 rpm.

Here it is in place on the machine:
attachment.php
.
The 40mm spindle is sitting inside a 60mm adaptor. I have sized this so that I can use up to a 60mm spindle and clear the front of the grinding wheel. I'm probably going to flip over the big clamping block so that the slit is downwards rather than upwards, to help keep coolant out of the spindle area.

Here is a side view. In this photo and in the last photo, you can see some wooden blocks underneath which I was using for support while messing around with different heights and offsets.
attachment.php

One of the tricky parts is a clamping arrangement. If you look closely by the machine head you'll see I have cut out a v block with a matching V slot. The V block will be attached to the grinding head either with epoxy or with a couple of bolts. The next step is to make a clamping bolt that goes down diagonally to engage the V block. The final piece to work out is a clamp for the hinge along with a retainer hook for when the inside grinding attachment is hinged back.

attachment.php


Once I have these clamps in place I'll post a couple of more photos.

attachment.php


Anyway, I have 4 parts left to mock up: 2 clamps, a retainer hook, and a shield/cover for the belt. I might also remove the cooling fan from the motor and fit a separate small 12V cooling fan with a shroud, because I think it might significantly reduce the noise and motor vibration.

I need to run the cabling and figure out where to mount the 48V power supply and the speed control/display. I also want to figure out a simple mechanism for tensioning the beit adjusting the angle of the small motor.

The idea is to sort out all of these details with the mock-up before I start to make drawings/cut metal.

I'm kinda interested in this but usually your images don't appear unless I make a comment lol... Then they lay up on screen later ?

Initially I don't see your images you are referring to, I wonder if they are large files etc ?

¯\_(ツ)_/¯
 
yup , now they pop into existence whereas before if I hit refresh and wait etc. the images don't 'Lay up" to the page ?

Anyhoo lol.
 
Our early Myford is already a very frequently used machine in our shop, and having ID spindle for it would be a very welcome addition, indeed.
Pardon me for lecturing you, Mr Z, but the better thing for a gear guy to do is bite the bullet and scrounge up an ID grinder. If you do carburized gears you're going to use it - a lot. And you'll find that other people will come to you for ID grinding, after they discover you have one. It will pay for the space it takes up.

The internal attachment thing is fine for occasional use but can't match the accuracy, ease of use, speed, or any other characteristic of a real ID grinder. If you do much of it, just cut to the chase. You'll be glad you did.

(If you want to go crazy, add a thread grinder to your abrasive department. Good for woims. Again, something unusual that few other people have can be a pleasant little money-maker. Plus ground threads are tits :) )
 
A bit more work on the mock-up. I've added a mechanism for snugging up the swing-down hinge:

attachment.php


I've also rotated the spindle clamping block so that the slot is at the bottom, and added the lock-down mechanism (head of the bolt visible in the center)

attachment.php


and a belt shroud/shield

attachment.php


I am now thinking that instead of just covering the belt and pulleys, it would make sense to have a cover which encloses everything and just leaves the spindle sticking out the end. I'll then modify the locking mechanism so that I could access it with the enclosure cover in place. I don't see any downside to this, and it would be good to have better coolant and grit protection for the innards. The only reason I can see not to do this is that if the cover is sheet metal then it might have a tendency to resonate/vibrate.

This is the nice thing about wooden mock-ups. An hour or two with some masonite and hot glue, and I can try this out.
 
Hi Ballen:
I've tried it both ways; belt driven from the main spindle and integrated NSK electric spindle with collet.
Each has advantages and disadvantages, the NSK was wimpy and the belt drive unit was clumsy.


Marcus
Implant Mechanix • Design & Innovation > HOME
Vancouver Wire EDM -- Wire EDM Machining

If you grind a lot of small diameters and the NSK was not strong enough more than likely you are trying to grind too much too fast. It is all about number of passes. This helps getting great finishes and perfect roundness.
 
Good morning Frank:
You wrote:
"If you grind a lot of small diameters and the NSK was not strong enough more than likely you are trying to grind too much too fast"

Yes, that is certainly true when the goal is significant stock removal and you just want to "git 'er done".
So my commentary was not to disparage the NSK unit although it may have come across that way.
Rather, it was to point out the limitations of the approach, in spite of the fact that is looks like a really attractive solution at first glance.

My experience for general internal grinding of the kind the OP describes, is that it is not really suitable for that kind of work, but it is a great fit for the small stuff that I do.

It is an excellent unit...very expensive, very precise, very fast (60K), and very delicate.
That makes it intolerant to even smallish mistakes and as you point out, one of those mistakes would be pushing it too hard.
It likes 1/8" shank mounted points for jig grinders, and I start to get nervous when those mounted points are anything bigger than about 6 mm diameter.
My concern is mostly glazing, unbalanced cutting and damage to the spindle bearings, or God forbid, a physical smash with the obvious consequences.
I don't think it could survive something like that, and it's well over 5 grand to replace.

So I brought it up mostly to discourage the OP from exploring it too deeply as a possible solution unless his work is much like mine, where a big bore is 0.250" diameter and 1/2" deep.

Cheers

Marcus
Implant Mechanix • Design & Innovation > HOME
Vancouver Wire EDM -- Wire EDM Machining
 
Pardon me for lecturing you, Mr Z, but the better thing for a gear guy to do is bite the bullet and scrounge up an ID grinder. If you do carburized gears you're going to use it - a lot. And you'll find that other people will come to you for ID grinding, after they discover you have one. It will pay for the space it takes up.

Not all, EG. I'm always open to listening and learning. In fact, I'm in complete agreement with you on all counts. Unfortunately, that last bit that I added the boldface to is the hitch. Until we move, there is no more space. Period. Very ESPECIALLY since I have somehow ignored myself and am trying to shoehorn our new to us Gleason/M&M Analytical Gear Inspection machine into the envelope, now. :rolleyes5: :willy_nilly: Stupid, really, but it kind of fell into our lap and there was no possible way I could resist with what it brings to us and adds to the shop.

For now, we have been doing fine with hard turning and honing, but it would be very helpful to add ID grinding to our already existing machine until such time as we move and I can add a machine dedicated to the duty.


Ballen - Very nice work so far.
icon14.png
 
Don't enclose the motor too well, it needs to be cooled!

I'm going to add a pair of 12V computer fans, one to push cold air in near the bottom and one to suck hot air out at the top. Need to find locations that are well projected against splash. Will probably use temperature regulated fans to maintain a constant temperature inside. Ideas?

According to the manufacturer the motor can run at up to 100C / 212F. I could add some thermal compound to the aluminium mounting clamps and make sure that those clamps make good thermal contact with the base. Then just let all the mass of metal dissipate the heat. Problem is that heat will also make the ID grinder base plate expand, and that's not good for precision. The whole thing is going to be about 300mm (12") square, so a 10 Celsius temperature rise would make that expand by 30 microns = 0.0012". Fortunately most of that growth would be in the vertical direction, which is fairly insensitive. Even so 30 microns is 20 times larger than the sort of precision that I can hold with the OD grinder. So keeping temperatures under control in this will be important.
 
I've always thought that I would get the most use out of a "sensitive high speed quill".

A plane bearing quill fitted with taper or collet on the business end, driven by an overhead gut rope and manipulated across the work with finger and hand motions.

Sensitive!quill.jpg
 
What about X- Cel- O.....all mine are XLO,and some even had rusty bearings from coolant ......I just replaced the balls with new ,and they work 100%......In their wisdom XLO used odd size (metric) cup and cone bearings,rather like magneto bearings.....the housings are incredibly soft steel ,and easily damaged driving out bearings.
 








 
Back
Top