Stuart Twin Victoria (Princess Royal) Mill Engine

[Dr_GMJNParticipant @dr_gmjn]

Thanks Ramon. I assume the theory is that the piston rod should be constrained radially by the piston at one end, and the cross head at the other, and that the rod itself will then centre the cap before final tightening? Having said that I think the gland packing itself would allow a bit of rod misalignment, so long as it’s constant (ie the bore and cross-slide rails aren’t sloping)?

[JasonBModerator @jasonb]

Like Ramon I like a firm fit of spigot into cylinder, easy enough to push together by hand but able to stay in place, see this post on my current Real thread. I ream the piston rod hole at the same chucking as turning the spigot so they are concentric so the piston should therefore be concentric to the cylinder bore. Gland I usually make a slightly looser fit in it's hole.

On full size and in particular when the layout was horizontal the cylinder cover was fitted with a "neck bush" as gravity would want to male the piston end droop and possibly rum on the lower face of the cylinder but the bush prevented that by adding support. My Fowler traction engine has that feature.

[Ramon WilsonParticipant @ramonwilson3]

I've never actually thought of it that way Doc but yes it could be construed as such.

Whilst I certainly agree with Jason on full size practice the masses involved in small engines such as this can be fundamentally ignored but the alignment is important

I've always thought of it as the bore, cover/gland and cross head need to be perfectly in line in both planes and the piston travels the bore freely – the packing taking any microscopic misalignment within the bore. This means all surfaces must be true to each other to achieve a perfect fit on assembly – something which, even with extreme care, is not always the result on bolting up. A small means of making allowances needs to be there then.

My thoughts are alignment tolerances first, then minor gland packing ease of tolerance, and finally the piston packing taking up the the last, if any, discrepancies. When using the PTFE packing the piston can be a good 2-3 thou down on bore size but if the rod isn't in perfect axial alignment with the bore then tightness issues will occur on piston travel.

I would stress that these are my thought on engines to be run on air for display only but something that worked well on the three in line cylinders plus cross head on the Corliss engine build.

Keep up the good work

[SillyOldDufferModerator @sillyoldduffer]

Posted by Ramon Wilson on 12/06/2022 18:18:00:

…

I have to say after a lifetime in engineering I'm not an advocate of using an edge finder to find the centre of a hole. I'm not saying it can't be done just that it isn't correct practice (for the edge finder) A DTI in the spindle is far more the correct way

…

Although edge-finder centre finding works well enough for me, I've always wondered how on earth it could be accurate. Perhaps the edge-finder error is hidden by my equipment's other errors, which don't matter because I rarely work to better than 0.02mm!

Next time I have a spare hour in the workshop I'll experiment with an edge-finder and DTI to see if I can detect a problem.

More light on my darkness, very educational this forum, thanks!

Dave

[Dr_GMJNParticipant @dr_gmjn]

Ramon, Jason – yes, now you describe it another way, the theory might be that the cover and cross head slides are supposed to centre the piston in the bore. I doubt it matters at small scale though.

Re. Thinking some more about the edge finder on cylindrical parts: Assuming you’re not touching at the exact quadrant points to begin with, the spigot will be either climbing or moving off the surface due to it’s rotation being clockwise, but the curved surfaces being mirror images.

But if you assume the contact that causes the spigot to throw off is a point (or line), and it’s instantaneous, why should it matter if the surface is curved or straight? I look for the smallest ‘throw off’ from concentric spinning, zero the DRO, then repeat several times to make sure that point is consistently zero. I can see how it might get sketchy for very small internal diameters, but for others I can’t see a particular issue.

Ill do some checks tonight with the edge finder and then check with the dti to see how far out I am.

[Ramon WilsonParticipant @ramonwilson3]

Okay lets approach this from a different direction.

The component part is placed on the table/vise/rotary table and held. You put the edge finder in the chuck/collet and bring the part somewhere near centre and bring the edge finder into the hole.

Move to one side till the finder runs off and take a reading – this is with DRO of course – using dials would be very difficult – then move in the opposite direction and do the same. Split the difference and move to the centre line on the DRO then repeat in the opposite axis. At no point do you KNOW (apologies for the capitols but it requires stressing) if the edge finder has run off equally but assume it has and carry out the operation. It's only when something doesn't fit that you realise you were not exactly on the centre – too late.

It's waste of time to do this then check your position after setting with an EF with a DTI if you can't be sure – why not just use it in the first place

If you use a DTI you know right from the outset you are as concentric to within whatever tolerance the DTI reads.

 

Sorry guys but I'm not convinced – as said above a quick superficial alignment of the part with a cone in the drill chuck as mentioned previously (it is not running BTW) then replaced with DTI to align the centre accurately. I use a small mirror to see the reading away from you.

An edge finder is what it is – yes you can use it but it does not give a reading and was not designed for centering and as such cannot be totally relied on – well not in my opinion

Like everything else yer pays yer money etc depends what you want of the end result I guess

Ramon

Edited By Ramon Wilson on 14/06/2022 12:41:29

[JasonBModerator @jasonb]

Run off up or down the curved surface is not an issue with an electronic finder so that can be taken out of the equation as it does not need to run off. Electronic one I use is also said to be good for 0.0002" compared with a mechanical one at 0.0005"

The correct way is to do say Y by eye and then ctr the spindle to take the x which should then be near as dam the true diameter. You should then place spindle central on X and do the Y again so that that will be a true diameter reading but I don't find that is really required but if it is something very critical may just do it for peace of mind.

As I said I've checked a few times to see if my edge finder method is accurate by then running a dti and what little movement there was on the needle would not be worth worrying about so don't use the dti now as I know my edge finder method works.

When I do use the DTI on the CNC I have a holder that puts the dial face upwards so no need for mirror tricks. Also got one of those co-axial finders but that stays in the cupboard, only coming out where the long probes reach where the edge finder and dti won't

[Ramon WilsonParticipant @ramonwilson3]

Digital (light emitting) edge finders are very good Jason, just one fault in my book, very easily broken! Don't ask!!

I can't actually use a traditional Starrett type edge finder or wiggler on my Linley mill due to not being able to get the shank diameter in the drill chuck but I have both types to use for the Amadeal – I'd still not use either for this task I'm afraid as despite the accuracy you describe – which I assure you I don't doubt for one second – you still don't have a visual confirmation that you are directly concentric.

There's a co ax DTI for use too but again winding the head up to get the room to use it is off putting. A creature of habit I guess – shortcuts yes but not if the accuracy cant be verified.

Regards – R

 

 

 

 

Edited By Ramon Wilson on 14/06/2022 15:23:01

[JasonBModerator @jasonb]

Yes they don't flex or spring much which is why I don't tend to use it on the CNC, I've even sent the end of a mechanical one across the workshop with that

I suppose after comparing edge finder and DTI results a few times early on I just have to trust the DRO. And all my videos of running engines are the confirmation that I need that it works for me

[Ramon WilsonParticipant @ramonwilson3]

Posted by JasonB on 14/06/2022 16:36:11:

And all my videos of running engines are the confirmation that I need that it works for me

[Dr_GMJNParticipant @dr_gmjn]

Drilled the cylinder end caps this evening:

I didn’t check it with the dti in the end – to be honest I couldn’t figure out how to do it! I think it would be easy if the o/d was unobstructed by the jaws.

Anyway, the edge finder and DROs resulted in them fitting to the cylinders, so all good.

[Ramon WilsonParticipant @ramonwilson3]

Well I can't disagree with you there Doc, not the ideal set up to use a DTI on. Whatever, what matters is that you got there in the end – more good progress

Just as a matter of interest do you have any scraps of flat aluminium plate to call on? Ideal for making quick one off fixtures for jobs like these and easily recycled for something else down the line.

[Dr_GMJNParticipant @dr_gmjn]

Ramon – yes, I’ve got some aluminium off-cuts, in fact this evening I’ve been using it to make a fixture to hold the cylinder caps to machine the gland housings (in the R/T, as per the valve chests).

Part of them is a milled square spigot to clamp in the vice (which slides side-to-side so I can mill the end radii (again as per the chests).

The spigot forms the stem of a “T” shape, with the horizontal top of the T intended to sit firmly on the top surfaces of the vice. I notice a small corner radius which is stopping contact between the T and the top of the vice. Is there a good way of removing this tiny radius and getting a small corner undercut? Can be by hand for this – it doesn’t have to be pretty.

I could use thin parallels to get around this, but it’s more stuff to arrange and potentially come loose during machining.

Thanks.

[Ramon WilsonParticipant @ramonwilson3]

Doc –

I think I follow what you are describing – taking the corner out with a triangular file would seem best by hand or possibly a slitting saw by milling? I take it you are making oval/elliptical shaped glands as opposed to the diamond shaped style?

It's always good to have scraps of ali to call on especially if you can get some tooling plate – very flat and stable but anything is better than nothing.

[JasonBModerator @jasonb]

Hacksaw held at 45deg will also clear a corner fillet. What are you cutting with Insert tooling?

[Dr_GMJNParticipant @dr_gmjn]

Ramon – it's the same housing form (radiussed diamond type) and method as outlined in pages 21/22. I think I'll remove the internal rad from the fixture with a triangular file.

Jason – not sure what you mean by insert tooling? I was milling the spigot with an end mill.

I'll post some pictures later on for a bit more clarity.

Cheers.

[JasonBModerator @jasonb]

Insert milling cutters tend to have a corner radius 0.4 or 0.8mm typically in out sizes

Should not get a significant radius when using standard milling cutters but a rougher/ripper or one specifically with a corner radius will leave one.

You could save time and just sit the jig down on parallels rather than lugs that sit on top of the jaws, bottom of vice is also more likely to be flat than the tops of the jaws.

[Dr_GMJNParticipant @dr_gmjn]

This is the beginnings of the fixture for holding the end caps in place on the rotary table – for machining the diamond shaped bosses:

Slides along the jaws of the vice to re-set the ends under the spindle centreline – as per the valve chests, but of course they weren’t circular:

Intention is to drill and tap the mounting holes and screw the cap to the fixture.

Puzzling over how to make sure the gland clamp holes are perfectly in line with the 9 and 3 o’clock mount holes. Currently machining a clamp out of aluminium as a test – Intention is to use that as a template to spot through onto the cast iron cover when it’s a best fit, then drill all four holes on the mill in one y axis setting.

[JasonBModerator @jasonb]

Two small flats would have done for going against the jaws. and left you more metal for other uses of the jig

As for setting the position of the holes place your jig in the vice and get it directly below spindle. Then loosely put the cover in place and move the mill table to the hole offset. Put a bit of something round in the chuck approx dia of the rounded end and bring it down so you can eyeball the position over the end. Move to other offset and double check that end adjusting to get the best for both and then tighten cover in jig. You then drill the gland and PCD ring of holes at one setting using co-ordinates from that centered position before altering the rotary table angle to mill the glands.

But before all that get your glands turned and drilled for the two stud holes. That way once the cover is drilled you can screw it into place and mill to shape at the same time as tidying up the gland boss.

Or forget the above and simply screw the turned gland on and file to the shape of the boss which will act as your guide.

Edited By JasonB on 17/06/2022 07:32:56

[Dr_GMJNParticipant @dr_gmjn]

Jason – not sure I understand your first point? It can’t be done in one setting because the central radius and end radii have different centre positions? The sliding jig is for milling the gland boss profile (exactly the same as on page 21/22 for the valve chests).

Ill try several methods for marking the end holes before committing to drilling them.

The issue I’m having at the moment is setting up the R/T. I can get it centred under the spindle so there’s no error when a dti is rotated with the spindle in the central taper. As soon as I centre a part in the vice or chuck that goes on top of the table, I can get no better than within 0.007” runout (either on a mill shank or on the cap gland bores). Must be a setup issue, but can’t understand how. I seem to remember a small error on the valve chests, but nothing like that bad.

Anyway it’s been a while since I did this, so I’m still getting back into the swing of things.

cheers.

Edited By Dr_GMJN on 17/06/2022 07:46:17

[JasonBModerator @jasonb]

I edited out the first bit as I remembered the ends.

You should certainly be able to set the part true under the spindle, have the vice slightly loose on the table and tap it into position as you rotate the dti in the spindle and then tighten down. Any error after that will be the table not rotating true to the tapered hole. I'd run the Dti on the piston hole if you can get that far in, larger gland hole if not.

When I did the Victoria which came without the gland castings I measured what I had on the iron boss and then drew glands that would fit just within. Machined the glands to drawing and screwed them into place so I could scribe around them and then filed the boss to the line blending it out to the fillet.

[Dr_GMJNParticipant @dr_gmjn]

Thanks Jason. I should have mentioned I also rotated the table 360 degrees and again it was still fine. I’ll re-check everything later.

I have already drawn the diamond profile on CAD and printed at 1:1 to make sure I can get the profile in the casting. That’s fine.

I guess I could stick the drawing onto the face and mark the end hole positions, then set them up on the mill, then drill them using offsets from the centre axis.

Anyway, first job is to trial cut the profile in aluminium to familiarise myself with the method and make sure I turn the wheels the right way…

ETA I take the point about two flats being enough to hold the circular jig, but I needed the spigot to contact the screw end stops, and I thought the overhang on the upper side would automatically give me a consistent height for when I un-clamp and slide the fixture side-to-side.

Edited By Dr_GMJN on 17/06/2022 08:09:38

[Ramon WilsonParticipant @ramonwilson3]

Whilst I certainly compliment you on the thought you are putting into this Doc I do think that you might be over thinking the process from the fixture point of view.

I've found these image from a couple of build that might be of help in the future.

This is the type of ali split clamp fixture I was referring too. It's being recycled from previous use and is for holding the eccentrics. The drill diam is the amount of throw of the eccentric, the parallel being set to the clamp not the other way round.

The initial set up is as described before by 'cone'

Once set up the fixture is bored to take the part then released and moved over the 'throw' amount. Drilled and reamed for the shaft and the boss turned

Whilst the above is on the lathe the principle is easily transferred to the mill on the rotary table.

Incidentally setting the work piece with a cone from tiny to big is pretty accurate but don't use a pre – tapped hole as a location – unless you have counter drilled the first the first thread out before threading (best practice anyway)

In early days of my journeyin ME I was told by an old and much missed 'mentor' you ought to make your self a toolpost drilling/milling spindle. He was absolutely right, but you do need means of dividing in the lathe for it to be of best use. I think you have an ML7 in which case the 65T bull wheel is not of use but an expanding arbour inside the rear of the spindle to hold a gear and a firm detent is well worth making.

Sorry for more pics on your thread but I do feel they are relevant.

Best – R

[Dr_GMJNParticipant @dr_gmjn]

Thanks Ramon. I might be over complicating it – probably am, but from experience I know that the slightest simplification or short cut I take usually results in re-doing something! Same with plastic models.

I could have used a split bush, but to get the hole array centre, and to centre it on the r/t I wanted to have free access to the o/d of the cap. Hence surface mounting it.

Here is the test piece I made today to check hole positions and centralisation on the castings:

I want to remove the minimum cast material, so it looks ok. The issue (what I’ve been struggling with) is how to align the two gland screw holes exactly with the x-axis of the milling vice. Once I’ve figured that out, it should be a trivial matter to drill the cover mounting holes and the gland screw holes, then profile the boss.

You can hopefully see that a tiny bit of angular misalignment with those gland holes could easily mean I’m milling fresh air when I come to profiling. I could reduce the central gland radii a bit to give more tolerance, but then the gland overhang gets perilously thin (see the paper template).

Edited By Dr_GMJN on 17/06/2022 17:33:45

[JasonBModerator @jasonb]

Doc, take a look at what I have just posted on my real thread. If you put a bit of rod or the reverse end of a drill bit into your chuck/collet and then offset to the gland hole positing you can then bring that down into the gland hole and turn the rotary table until the cast boss looks to be at it's best position. Note the reading on the rotary table and use that as your zero.

[Author]

Posts