Issue machining driving wheels

[kipford]

 I am machining the driving wheels for the Aspinal and have an issue which I am not of the cause. So I have mounted each wheel in the independent 4 jaw with the back face outwards gripped on the tread. The 4 jaw was adjusted to give the best concentricity on the inner edge of the wheel tyre. I then turned the outer diameter to + 0.5 mm and then machined the back face to give the best axial position of the spokes. Then the centre axle hole was drilled, bored within 0.3 mm of size, then reamed. I was very careful only taking 0.25 mm cuts with a 6mm diameter insert tool. Having read lots of threads on machining wheels I decided to mount the first wheel back in the 4 jaw (my 3 jaw is not quite big enough) gripping on the turned outer diameter and with the back face against the chuck. The wheel was tapped into place against the chuck and clocked into the centre axle hole for good measure. Again taking small cuts I cleaned up the front face of the tyre, I still have 2 mm to take off the front face. This where I found a problem. I check the tyre width all round and found its running out by .155 mm or 6 thou. So what is causing this I would have expected it to be a lot closer. I stopped at this point although the wheel is still in the chuck. 

Questions:

1. Is it just bad set up by myself. If so what can I do to reduce the error?

2. Is it the chuck?

3. Is it something else?

I am seriously considering the following route. Put the face plate. Take a skim cut across it just to ensure it is square, then mount the wheel on that and face the front features. Is that a reasonable approach?

For the tread I will use an arbour to rough nearly to size and then finish turn after the wheels have been assembled to the axles.

The slide shaft I machined seemed easy compared to this!

Dave

[Roger]

So if I understand you correctly, the issue seems to be that you assumed that tapping the wheel back against the jaws would mean they were running true? You simply can't assume that. The problem is two fold. Firstly, you don't know if the fronts of the jaws are running true in the first place. Secondly, you don't know if the face is touching all round.

In my opinion, the only way to be sure something is running true is to clock it. Sometimes that's not very convenient, you may have to set the clock on the cross slide to touch the back of the previously machines face, then wind clear, turn to miss the jaw and then go back in to see if it's true. It can be very time consuming, but sometimes that's what you have to do. A DRO on the carriage can be a useful means to measure the differences more accurately and find your way back to the same position if you have to move the carriage to miss the jaws as they turn.

I find that it's usually easier to lever the part away from the chuck with a big screwdriver rather than by tapping it in with a mallet. It's easier to control the amount that way.

[kipford]

Roger
Thanks, I was wondering about the accuracy of the jaws. Although I have been in Engineering all my life, I am not a machinist and am still very much at the beginning of this particular journey. I had read in other threads of people simply reversing in the chuck or using a 3 jaw to do the same, so assumed it was an acceptable method. I will check the back face tomorrow to see what is going on. Would you agree perhaps an easy set up once I have sorted it would be as I said to use the face plate instead?

Dave

[uuu]

I would mount on an arbour or mandrel first, just to check that the back face is running true. Having the back face and front faces running true to one another is of no use if the back is not true to the bore.

Wilf

[kipford]

Wilf
Good Idea, I have a mandrel already set up in my collet chuck. So I can easily swap over the chucks and check that tomorrow.

Dave

[Roger]

May 26, 2020 19:20:08 GMT kipford said:

Roger
Thanks, I was wondering about the accuracy of the jaws. Although I have been in Engineering all my life, I am not a machinist and am still very much at the beginning of this particular journey. I had read in other threads of people simply reversing in the chuck or using a 3 jaw to do the same, so assumed it was an acceptable method. I will check the back face tomorrow to see what is going on. Would you agree perhaps an easy set up once I have sorted it would be as I said to use the face plate instead?

Dave

Hi Dave,
Personally, unless accuracy isn't important, I clock things. That way you can't be wrong and you're not assuming anything. In the end, it all depends on what level of risk you're prepared to take.

A face plate ought to give much less of an error, but it isn't a quick job to set it concentric. You could turn a mandrel with a large diameter to pick up the rear face I suppose. That ought to give a good result, but you need an decent undercut to be sure it's sitting flat. 
You rarely get a really sharp shoulder on a piece that you then turn round and hold in the chuck. It's easy to slide a large diameter up to a shoulder, thinking the shoulder is pressing against the jaws when in reality it's resting on a small radius at the root of the shoulder. That's unlikely to be sitting flat.

These things are perennial problems when reversing large parts that need faces to be true to each other. 
In the end, I think it takes less time to do it as accurately as possible rather than try to fix it when it isn't good enough.

[dscott]

I like to get the front face running true first and will spend quite some time on this.
Then take a cast iron cut over all the surfaces including a parallel tread for gripping later.
During this turning Getting a datum point set up so that everything is taken from this point.
Even if the rear of the spokes are now moved into another position.
Machine the back. bore and ream.

Once mounted on the axles a finishing cut can be taken over the front. The rear being concentric with the axle.

David and Lily with Cast iron hands as usual. 20 second wash of hands more like 5 minutes!

[springcrocus]

Dave,

I covered two different ways of machining wheels; the driving wheels using soft jaws HERE and the tender wheels using a mandrel HERE. I don't have a faceplate so couldn't go that route.

Both methods gave acceptable runouts of a couple of thou or less.

Regards, Steve

[ettingtonliam]

Setting the part machined wheel in the 4 jaw and tapping back against the jaws will work if: its a good quality chuck thats not had much use/abuse, and the chuck body itself is running true. Otherwise it won't. A small fragment of swarf (we all know how cast iron swarf gets everywhere) trapped between the wheel and the jaw can easily cause your amount of run-out. Wear of the slides of the jaws and/or the chuck body can cause 'lift' just like it can in a vice. I always go round several times with my 'tapping hammer' to make sure the workpiece is properly seated.
You could always try giving the casting another few firm blows at the thinnest point to see if it seats better. At least you've still got another couple of mm to go yet.

[stevep]

I'm with David. I always get the front running true (the inner rim) first, then take the facing cut and a parallel section on the tread, up to a step. I also drill through with a small drill (1/8"), and when I then set up to machine the back, I 'clock' this hole. Then the hole can be opened up, bored and reamed, knowing that it will be pretty concentric with the previously set inner rim.

[JonL]

Sorry to digress, can I assume by "clocking" you mean checking with a DTI and then rotating it in the chuck until you get the smallest error? I still find some terminology a bit confusing.

[stevep]

May 27, 2020 10:26:43 GMT JonL said:

Sorry to digress, can I assume by "clocking" you mean checking with a DTI and then rotating it in the chuck until you get the smallest error? I still find some terminology a bit confusing.

Yes - 'clocking' means using a DTI. I mount the wheels, when I am going to do the back, in the 4-jaw. I then put a centre in the tailstock, and put another centre between it and the hole in the back of the wheel. You then rest the lever of the DTI on the centre against the wheel, and use the jaws of the 4-jaw to get the hole running true.

Here's a picture to show the set-up



As the DTI is calibrated, you can see how much run-out there is, and by careful adjustment, halve the error. Do this for each pair of opposing jaws. It only takes a couple of adjustments to get it true.

[ettingtonliam]

Similar, but I miss out the middle centre. I use the tailstock centre, in the pilot hole, to push the wheel against the jaws, then, at first gently, then more firmly, tighten the jaws onto the wheel. Which doesn't help the OP, because his problem was the wheel, when pressed against the jaws, not running true in a longitudinal direction.

[kipford]

Quick update. I have checked the wheels on my arbour and they have just on a thou run out. I am happy with that

[JonL]

Thank you Steve, I had assumed you mean't you rotated it a few degrees to see if it got better or worse, of course the 4 jaw I had forgotten about.

[andyhigham]

I like to use soft jaws for this kind of set up

[Roger]

I think it's important not to get into a mindset of 'this is how to do wheels' etc. When you're in General Engineering, you rarely come across the same job twice. I think it's far more important to grasp what's required to be sure of obtaining the required accuracy.

In general, I'd say that people are too quick to make unrealistic assumptions, such as...

1) My 3-jaw runs true... I doubt if it's better than 50 microns
2) My 3-jaw repeats... It might to 20 microns or so
3) My faceplate runs true... Maybe, but why not check it?
4) Turning to a stop is very accurate... unless you'd got swarf on it
5) My lathe turns parallel... it should, but probably not as good as you think
6) I don't need to turn the soft jaws, they're already at the right size... you do.
7) I don't need to clock it, I know it's close... no you don't

I could go on, but I'm sure you get the point.

Accuracy doesn't just happen, you have to make it happen. That means I clock everything unless it's the roughest of the rough jobs where it really doesn't matter. Sure, it takes a little longer, but I rarely have to make something twice.

[kipford]

So we have a result as far as I am concerned. After digesting the above and having a good think about best way I was going to get an acceptable result this is the method I am now using, a bit long winded to get there but the time was not an issue.


!. Turn up a steel shouldered arbour a couple of mm over axle size, a couple of mm shorter than finished wheel width with an M12 thread on the other end.


2. I had a couple of 6" or so diameter offcuts of L165-T6 Aluminium Alloy about 60 mm long which had been lying around at home for about 30 years after being thrown out from work. I knew they would come in useful some day. This was to form the basis of a turning jig. I faced up the off cut in the 4 jaw and then bolted it to my faceplate, this was done on the mill.

3. The face plate was then mounted in the lathe and the other end faced and the outer diameter turned to about 10 mm below the wheel flange diameter. The faceplate was removed from the lathe and more holes put in to enable the wheel to be clamped to it.

4. Back to the lathe for the final time. A 12 mm thread was tapped in the centre of the offcut and a counter bore machined to clear the shoulder of the steel pin.

5. The steel pin was wound in and tightened. A quick skim across the face of the jig to ensure all was square, I then machined the pin to a close fit on the reamed axle hole. 

6. So now I had a substantial mounting face and arbour as accurate as I could get them. 

So the first wheel was mounted on the arbour and bolted back to the fixture with 3 equi-spaced bolts. I clocked the back face, I am learning Roger , half a tenth movement on my crap clock. So here we go a couple of cuts to clean up the front face, a the clock showed around half a tenth of movement again, result. The picture shows the set up.  

A by-products from this is  with a small mod the fixture can be used to drill and bore the crankpin holes and it can be used again on different axles sizes.

Dave

IMG_2023 by Dave Smith, on Flickr