Piston Valve Liner Clearance

[runner42]

Hi all,

I have produced for LMS Black 5 (Doris) the piston valve and liner in accordance with LBSC's requirements, viz used a 5/8th" spiral reamer for the valve liner and used 5/8th" stock stainless steel rod for the piston valve. However, trying to push 5/8th" piston valve into a 5/8th liner produces running fit issues, so I have had to remove some material from the stainless steel, particularly when machining the bobbin part of the piston valve invariably produces small burrs at each end of the waisted part.

I have convinced myself that I have achieved a good fit by fitting the piston valve into the valve liner and shining a light at one end and looking for light shining through. No light is detected even by placing only one half of the bobbin into the liner. But I may be fooling myself.

Measurements are not practical since I only have a digital caliper that has a resolution (not neccessarily accuracy) of 0.0005" really only 0.001" and I can alter the reading by altering the pressure on the caliper jaws.

An aside micrometers have clutch mechanism that controls the pressure applied to the jaws, but they are no where as flexible as digital calipers.

Also measuring the diameters only doesn't take into account possible ovality.

What sort of running fit is required for piston valve/liner to work without leakage and how can I verify that I have achieved it?

I have read on M.E.C.H that a general rule of thumb is 0.001" per 1".

Thanks in advance,

Brian  

[runner42]

Now that was a disappointing response, even Justin (jma1009), who normally is so helpful passed on this one.

Perhaps I need to change tack.

I suspect that stainless steel is a diffulcult choice for machining and that LBSC selected it because he was advocating the use of standard bar stock that required no machining assuming that 5/8ths" stainless steel would fit snugly in a 5/8ths" reamed gunmetal liner. But that does not take into account working fit tolerances. 

Today I made two more stainless steel piston valves, but used only fine abrasive medium to slowly reduce the diameter and using the valve liner as a guage, hoping to get the best possible fit. There is a gnats whisker between being too tight and too loose. 

In selecting the choice of materials to make the piston valves what is the best material to use in conjunction with a gunmetal liner?

Brian

[Roger]

That's a really tough problem to solve without any way of measuring things accurately. I have no experience making piston valves but have been involved in precision engineering. The comment I wanted to make was that stock material is usually pretty hopeless when it comes to geometry. Even things like dowels what you would think are round are often far from being so. Is there any way that you can set up the piston between centres on the lathe and use a toolpost grinder to produce the outside diameter? I hate the thought of grinding on the lathe, protecting the slides is vital, but sometimes these things have to be done.
When it comes to seeing what touches, at this level of fit, marking blue isn't much us in my opinion. I use a soft pencil and rub that all over the surfaces. It's pretty easy to see where things touch using that method.
I have no idea how round or parallel the bore will end up by using a reamer. Maybe you can track down a friendly local engineering company who will let you use their internal bore micrometers to see what you've got. I know from what Julian has said that the clearance needs to be very fine indeed. I guess you could always get a company to hone the bore for you. We used to be able to get bores for Air Bearings parallel to within 2 microns without much difficulty on a a manual honing machine. At least if you knew that the bore was good, you'd have a reference to fit the piston to. At the moment there's no way of knowing what you've actually got.

[jma1009]

hi brian,

the bobbin must be steam tight but not mechanically tight. that sounds a bit glib, but there isnt a lot more to say! follow LBSC's instructions carefully. don young has a different method based on what he did on his K1/1 with a solid bobbin - make it mechanically tight then force the valve backwards and forwards with moly slip till it becomes a sliding fit - in the course of which treatment the theory is that machining marks and the surfaces become smooth.

the main problem with stainless is it's different expansion. ive seen at least 2 new locos with stainless bobbins seize up on the loco's first steam up. if damage is at the same time caused to the piston valve liner bore then very disheartening.

don young never specified stainless for piston valves, and i dont think LBSC used stainless on his own locos. good quality PB rod turned down to size seems to me a much better engineering method than using stainless. stainless can do strange things sometimes. for example stainless piston rods actually wear quite badly in service and can score. we dont have much alternative for piston rods, but it is a fact that stainless wears badly and does strange things in bronze particularly hard drawn PB.

it should be quite easy to turn up some plug gauges to check the piston valve liner bore size after reaming. though if you havent got a 'mike' this is probably a waste of time.

it is very difficult to get a large reamer to make a long parallel bore with the average size model engineer's lathe.

how are you going to fit the liners to the cylinders?

john baguley (baggo on here) has an excellent method of making piston valves arrived at after much experimentation on his website.

for DORIS with gunmetal liners i would used phos bronze bobbins not stainless.

cheers,
julian

[pault]

Hi.
If I remember correctly the valve bobbins are just plain with no rings or even oil grooves in them. This means even if you get the fit correct to start with as soon as they start to wear a little blow by will increase causing increased water consumption. Normally this is compensated for by piston rings which allow for wear.
As Julian has said there is a potential issue with the different rates of expansion between stainless and bronze. This is further complicated by the different rates of heating due to the different masses of the valve and cylinder block/liner. I’m in two minds as to whether stainless or bronze is the best material for this application bearing in mind the close fit required and the transient temperature issues
Personally I would recommend going with PTFE sealed valves as a number of people on here have described. Going down this route will give you an excellent chance of a first time success and a long working life. Another alternative is a bobbin a few thou under the bore size sealed with cast iron rings although finding them in the required size could be difficult.
Regards
Paul

[runner42]

Thanks striplar, you have suggested some good approaches to the problem like using local engineering companies to assist in measuring the bore diameter and/or honing the bore. However this need is predicated on achieving a running fit tolerance in the 0.0001" range, which maybe detrimental if the coefficients of expansion are different, which causes the locomotive to seize when hot. Notwithstanding, knowing what you have got in terms of bore diameter, parallelism and ovality is an advantage in making suitable piston valves. I shall use graphite powder instead of a lead pencil as a means of identifying high spots.

Thanks Julian, that was an eye opener statement about LBSC not using stainless steel for piston valves in his locomotives. I guess by his words and music that he was keen to take his followers down the path of using piston valves, instead of slide valves and this was on the face of it a simple way of getting them interested in changing in 1948. I am not aware of any post fallout and that all Doris builders followed his instructions and achieved remarkable success. He did caveat his approach by the statement that if the stock stainless steel rod can be inserted in the bore without shake then OK, if not use PB. Maybe reamers are produced to be a few tenths oversize so that there is an already in built clearance? Your suggestion to use Don Young's approach of using molyslip to act as a fine abrasive to get the piston a sliding fit I shall use irrespective of the material I choose. 

I have a 0 - 1" micrometer. You suggested making plug gauges for measuring the bore, I assume that these are GO/NOGO gauges, if so what should be the diameter of each?

The liners are already fitted to the cylinders and they were fixed using soft solder as LBSC suggested.

I have had a quick peruse of John Baguley's website and haven't yet managed to find the area for making piston valves, but I did look at the place where he with trepidation made piston valves that were a good fit at room temperature and then heated them with boiling water and removed metal so they were a sliding fit at this temperature and foreshadowed the need to take perhaps another 0.001" off to ensure they don't seize when really hot. Are we talking in terms of a few thou, instead of tenths?

Thanks Paul, yes you are right the piston valves are plain bobbins having a width of 7/32". I am sure some skillful artisan can fit rings to this size of bobbin, but I don't think I can. Your recommendation of using PTFE sealed valves, what is a sealed valve?

Brian

[Roger]

Ok, here's a completely 'off the wall comment' that might make an interesting experiment, but first an apparently unrelated story...

One of the 60,000RPM ball bearing spindle designs that I used to repair, required the top bearing to be a super precision fit in a steel housing so that the wavy washer underneath it could maintain the preload by allowing it to move. The trouble is that the bearing could then tramp round slowly and they often bore what looked like rust marks where they'd been fretting. The design had been modified at some point to add an 'O' ring in an attempt to stop this from happening, but it's didn't work.
My solution was to bore the housing oversize, press in a PTFE sleeve, then bore the sleeve with a razor sharp boring bar (I used a 2 flute carbide PCB slot drill). All the work was carried out on a Harrison Lathe without a DRO, and as the final fit was approached, the tool tip just showed what looked like a tiny ball of flour on the tip of it.
What's interesting is that it was possible to make the fit virtually size for size with the bearing being a nice sliding fit without any shake, but with a very definite dragging. The bearing could be initially fitted well before the right amount of drag was achieved, and several tiny cuts could be taken to get it right. Obviously the cost of the material was very small, and if the bore went over size I could just start again. These spindles ran the quietest of any of that type and lasted longer.

So my slightly outrageous suggestion is to wonder whether it's possible to make the whole of the piston bobbin out of PTFE? The material is possibly compliant enough to withstand the differential expansion and strong enough to do the hob? If not PTFE, then perhaps some other exotic high temperature plastic. My impression is that the forces on the valve are not great, that's surely one of the advantages of piston valves? The fit could be made so that it dragged just like my bearing, it's natural lubricating properties preventing it from scuffing.

I'd be interested to know other members thoughts. It's such an easy thing to try, I'd love to know the answer. Maybe my SPEEDY will be the first?

Cheers,
Roger

[pault]

Hi Brian
Basically I was suggesting valves sealed with PTFE rings or similar, if you have a look at about half way down this
modeleng.proboards.com/thread/7932/introduce?page=9
you will see the sort of thing I am talking about. There are write ups on here by various people about doing this I think baggo is one of them.
Roger has an interesting idea ref making the valve totally out of PTFE, my only reservation would be the strength of the wasted part between the heads would it be strong enough especially when you consider the temperature of the steam surrounding it.
On a different note a firm I used to work for used to put plastic “bearing sleeves” into electric motors used on reciprocating pumps, it was the only way to prevent fretting in the aluminium bearing housings
Regards
Paul

[Deleted]

Brian - the piston valve saga is documented in the Helen Longish section with all the trials and tribulations! I also made a set of PTFE headed valves for a friend at the club that were very successful.

Piston Valves



One thing I have found is that although the valves seal perfectly under steam they will leak when cold e.g. if you want to run the loco on air again - They shrink when they cool down after the first steaming. You could possibly reduce this shrinking on cooling effect by making the PTFE heads as thin as possible. Those on Helen Longish are only 0.0625" thick.When you initially machine the heads you just make them a push fit into the liner and they sort themselves out when they get hot.

I tried various ideas for the valves for Helen including the Don Young method of knocking the valve back and forth using MDS as a lubricant and this did work well. Unfortunately, I did use stainless for the bobbins this time and they did score badly as Julian mentions.

Roger - I've tried the solid bobbin method with Helen using bearing grade PEEK and gave up on it in the end. To get a running fit at working temperature the bobbins had to be a really sloppy fit when cold. I did have two things against me though - the piston valves are made to the double ported Greenly type and the valve heads are very long which increased the area of contact between the head and the liner and made them more prone to seizing. Also the level of superheat on Helen is very high which increases the working temperature of the cylinders to over 200°C which made the expansion problem much worse. You could try using Fluorosint (Mica filled PTFE) which has an expansion similar to bronze. Jim Ewins used this but it's very expensive!


Original solid valves:



New PTFE version in present use and still perfectly steam tight after 5 years use:

[Deleted]

Brian - I personally wouldn't rely on a reamer finish for a piston valve bore. I would finish it off by lapping with an expanding lap or a hone taking care to keep the bore parallel. You need as good a finish as you can get if the valves are to seal over a long time. Also the valve heads, if metal, need a very good finish as well unless you can fit them with rings.. Ideally, they could be ground or lapped to a good sliding fit in the liner. The problem with just a turned finish is that no matter how good it looks, the surface will still be like a ploughed field and as soon as the high spots wear off, the valves will leak. All in all, it's much easier to make the valve heads from PTFE! Just fit and forget.

John

[Roger]

Hi Baggo,
Those are really interesting designs, but how do you fit the PTFE onto the second design? It doesn't look like the flanges are removable.
Roger

[Deleted]

Hi Roger,

Yes, the bronze bobbins are solid and turned in one piece including the end flanges. The thin PTFE heads were turned to size and then split lengthways with a sharp knife. They could then be sprung over the bobbin. The split closes up and reseals when the bobbins are fitted. This should work for the normal type valve I would think.

On the top valves ( for an LBSC 75000 class loco ) the ends of the bobbin are threaded into the centre part so the plain heads can be fitted and secured with Loctite. I didn't take any photos of the bobbin in pieces unfortunately.

In the photo of the normal type valves you can see my way of making the adjusting of the valves dead easy. The bobbin floats on a sleeve which threads onto the valve spindle. Adjustment of the valve is then a simple matter of turning the threaded sleeve to move the valve on the valve spindle and then locking it with a nut when the adjustment is correct. Much easier than the LBSC, Evan's etc. method using two nuts either side of the valve or shims which is a pain. Adjustment is done with a couple of concentric box spanners with the front valve chest cover removed. I don't claim any originality for this idea as I probably saw it in a copy of ME or somewhere.

John

[Roger]

Hi John,
I have to say that I prefer the first one that doesn't require slitting the PTFE sleeve. For one moment I thought you'd somehow managed a long helical slit in the second one until I realised that it had grooves presumably for holding lubricant. Now there's an ideal 3D printing application if ever there was one!
The concentric box spanners seem to be a neat way of adjustment, I'll make a note of that too.
Anyway, I'm pleased this topic came up, I'll do something like that on SPEEDY when I get to that part.

Roger

[Deleted]

Jan 18, 2014 12:44:10 GMT John Baguley said:

One thing I have found is that although the valves seal perfectly under steam they will leak when cold e.g. if you want to run the loco on air again - They shrink when they cool down after the first steaming. You could possibly reduce this shrinking on cooling effect by making the PTFE heads as thin as possible. Those on Helen Longish are only 0.0625" thick.When you initially machine the heads you just make them a push fit into the liner and they sort themselves out when they get hot.

Hello JOHN....given BRIAN has a Bronze cylinder then your Bronze design is the way to go I would think....I must admit that PTFE is one material I've not had any experience with re}-- valves and piston rings etc...Have you ever considered a "Snap Valve" like the GWR type but using PTFE as the ring ??------------ Also, with the design shown do you use the valve head for setting or the edge of the valve itself

[Deleted]

Just to add to the above, for those "not in the know" here are some GWR Snap Valves

Attachments:

[Roger]

As a bit of an aside, how do you decide when the port is just opening so that you can set the valve? Is that done by eye? Does anyone apply compressed air at low pressure to one of the drain cocks and then listen for the exact opening point?

[miken]

Jan 18, 2014 20:23:01 GMT Roger said:

As a bit of an aside, how do you decide when the port is just opening so that you can set the valve? Is that done by eye? Does anyone apply compressed air at low pressure to one of the drain cocks and then listen for the exact opening point?

To set the valve, assuming that you have the dimensioned drawings ( you should if you have just made the parts). You just leave the front cover off and depth mic down to the bobbin and calculate the opening point.
As another aside, Am i the only one who fits Iron bobbins in iron liners with iron rings to the piston valves to miniature locos?
Mike

[Deleted]

PETE--------- this might help with tech. terminology on your Speedy ?/>>>>>>>> en.wikipedia.org/wiki/Walschaerts_valve_gear

[Roger]

Thanks Mike, I should be able to do it that way.

[Deleted]

Jan 18, 2014 20:09:47 GMT @4930hagley said:

..Have you ever considered a "Snap Valve" like the GWR type but using PTFE as the ring ??------------ Also, with the design shown do you use the valve head for setting or the edge of the valve itself

I don't see why you could not fit PTFE rings to the valve head. I use PTFE rings on pistons with no problems so far. The carbon filled PTFE might be better for thin rings like that as it's springier. However, the majority of what I have done so far has been with 2½" and 3½" gauge stuff so the valves are not that big and fitting rings could be a bit fiddly. Also, the ports on Helen are just machined grooves so any rings would just jam in the ports.

I've always assumed that the edge of the ring controls the opening of the port as it's the ring that does the actual sealing, not the valve bobbin. Certainly, that is the case with the valves I've made.

Roger - yes, I've used the air and draincock method for setting the valves but the valves need to seal well otherwise you get a blow all the time and it's difficult to determine the exact point of opening. Another way I've used is to measure the position of the port and the valve from the end of the liner with a depth gauge/digital caliper.