Valve leakage

[Deleted]

I'm building the Modelworks Britannia and have tested successfully on compressed air. However, there's a fair amount of leakage on the right-hand piston valve - by blowing hard into the inlet pipe with the valve in the mid position I can hear the air leaking to exhaust. The left hand valve seems pretty good, and if I swap the valve pistons over the problem remains on the right hand side which suggests that the problem is the liner rather than the piston rings.

The liner is cast iron, in two halves, pre-assembled into the cylinder block from either end by Modelworks and apparently done as a press fit. The valve piston is a steel bobbin with two split rings at each end - they appear to be made of bronze.

The problem could be leakage past the rings or possibly leakage between the two halves of the liner and then between the outside of the liner and the cylinder block. Apparently a lot of Winson builders had the latter problem, but they had to assemble the liners themselves with Loctite, and a factory-fitted press fit ought to have reduced the risk of this.

The piston rings have uniform rubbing marks all round, suggesting that they are making good contact at least in some part of their travel, but the cylinder liner has some areas highly polished and others rather matt (around the same cross-section), which is perhaps where the problem lies. The bobbin has reasonable sideways float on its rod, so I don't think it's being forced against one side of the liner.

I could obviously ask Modelworks to replace the complete cylinder assembly, but I've spent a fair amount of time polishing the exterior so I'd rather not do this. Is it likely that the liner will improve with running in - it has only run on air for about 10 minutes so far - or can it be reamed by hand to a better finish? Also, I'm not sure how good the seal has to be - if the regulator is opened with the reverser in mid-gear would you expect steam to leak to exhaust?

I'm not planning to paint the loco until I've steamed successfully and solved this problem, so replacing the cylinder at a later stage would still be an option if running-in doesn't improve it.

Regards, John

www.britanniabuilder.co.uk

[baggo]

Hi JJ,

a few ideas:

it sounds as though the valve liners are not a constant diameter all along the length. They may have been 'squeezed' in slightly when they were press fitted. I would have thought though that they would have been reamed or honed after being fitted which would have sorted out any problem like that.

The best way to 're-finish' the bores would be to hone them (if you could find a small enough set of stones) or lap them with an expanding aluminium lap and fine lapping compound. That would not be easy unless you have the facilities to turn one up yourself or could get some-one to do it for you. The problem is that if you have to remove a lot of metal to true up the bores then the valve and rings may then be too small.

If you can hear leakage by just blowing down the inlet then I would think the problem is quite serious. Have you tried oiling the valve with some thick oil such as proper steam oil and trying again? If the valve fit is the problem then the leakage should reduce. If it doesn't make any difference then it may be the fit of the liners in the cylinder block that is the problem.

It is possible that the problem will disappear when the loco is run under steam as the valves and rings will expand quite a bit and take up any gaps and the seal will improve with running.

John

[Deleted]

Hi John,

I'd previously just smeared a little steam oil on the valves before inserting them, but I've now tried sloshing it all over the liners with a toothbrush and this does improve it - in fact it blocked the leak completely for a few seconds and then the air broke through again. Also the sound of the leak now varies a little as the valve is moved around the midpoint, which I guess proves that it's getting past the valves rather than through the joint in the liner.

Perhaps I'll wait and see how the first steam test goes.

Regards, John

[mott]

Hi John, I'm interested in this string because my valves as at present have no rings, the bobbin is just a sliding fit. In asking about as to what to do for rings, the use of PTFE is often mentioned. It does have a very high rate of expansion but can be made into very successful rings provided an overlap joint is machined in and there is also Flourisent which is PTFE with added mica which has a much lower rate of heat expansion and whilst expensive is the "bees knees" for rings. There have been several strings about this problem under general chat and it might be worth doing a search to see what turns up. The other source is Britannai owners and I will ask around at the club tonight to see what they have done. Mott

[mott]

John, the only opinion I got at the club last night when we had a Night Steaming Session was that if the valve liner and bobbin are cast iron then it would be best to fit cast iron rings as well so as in the running in process the bore and ring would bed in together.They would have to be fitted to very tight tolerances, correct grove size etc. Dont suppose this helps but ....
Mott

[Shawki Shlemon]

Many years ago there was an article in ME on the use of PTFE with details . I used the information and followed the instruction to use PTFE on three engines successfully the engines 5" g NSW C 38 , NSW C 36 and GWR King .I am sure I have that issue somewhere , if you want the info send your e mail and I will send them to you . You can see these engines with short videos on my web site .

[Deleted]

Mott, Shawki,

Thanks for your comments. I know that a lot of Winson Britannia and 9F owners ended up using PTFE rings, and I may need to do the same eventually. I think I'll stick with the current rings for the time being and see how they behave under steam and with a bit more running in. With the Brit's big boiler and grate I should have plenty of steam in reserve, so the leakage will most probably be an embarrassment rather than a show-stopping problem. With two outside cylinders it only takes 10 minutes to remove the running boards and then remove the valves, even after the loco is painted.

Regards, John

[pondok]

Hi John,

I realise this post was started some time back, but I thought I'd put in my ha'penny's worth. Even very moderate leakage past the valve 'bobbin ends' can affect the running of an engine and the enjoyment of it. Essentially if the blow-past up the chimney equates to the blast of a fairly gentle blower, that is usually enough to lose control where you need it most - starting and very slow running.
A complete seal allows you a slow controlled start with a heavy load, as the regulator need only be cracked open and the steam chest pressure will slowly rise to boiler pressure and the wheels start to move - once moving, the steam chest pressure drops again as the steam is admitted to the cylinders, steam chest gauge going up and down markedly and so avoiding slipping more easily. Essentially wire-drawing put to advantage...
To make a start with leaky valves, the regulator has to be that much further open to get the wheels to begin turning at which point there is little drop in steam chest pressure and often causes a slip...
I absolutely recommend PTFE rings, having used them on 2 locos , one of which has had them for 3 years with no perceptible wear. Plus of course there will be no wear of the valve liners at all, and the exhaust beat is much crisper, especially if you design then such that the entire 'bobbin' end doing the job of controlling admission and exhaust is a solid PTFE ring, held in place by a bronze 'body' having chamfered 'cheeks' either side of the ring. The ring can be machined to size and cut with a 'step' to allow it to be slipped into position. Once closed again over the bobbin body, it will make a complete seal with the 'step' and should be a tight push fit in the liner. PTFE expansion will make it even tighter under steam, but will not sieze.

Good luck!

Andy

[Deleted]

Hi Andy,

That's an interesting point about starting and slow running - I was expecting any valve leakage just to be an embarrassment, but I can see now that it might affect control as well.

Unfortunately I've made no further progress since my last post above - still waiting for Modelworks to deliver the final kit, although a friend visited them last week and saw the 50 boxes lined up in dispatch, so it may be imminent.

Regards,

John

[pondok]

Hi John,

Another poster here asked me about the PTFE story so I'm taking the liberty of posting my reply to him here too, in case you or anyone finds it useful.

My slight prejudice against following the conventional full size practice in models of metal valve bobbins and, say, cast iron rings etc reveals itself...:-) but is probably more to do with me being someone who has never has the chance to try his hand at building an engine from scratch, so the PTFE is such a great option to sort out older valve sleeves with wear in them from having metal rubbing on metal for years...

___________________________________

The whole idea is basically that the bobbin ends, ie the bits that do the job of opening and closing the ports have to be replaced by PTFE.

You can achieve very good metal valves that will perform as well as PTFE, but it basically means grinding the valves in with the valve sleeve to make sure there is next to no leakage.

That's hard work but anything less will be a waste of time because even with careful reaming of the valve sleeves and careful machining of the valve diameter to be a nice tight push fit almost always
leaves tiny gaps for steam to get past - it's very hard work to get that seal.

Usually, if you can push metal valves into the sleeve by hand, there are going to be big enough gaps to mean fairly poor control with heavy loads.

The lovely thing about PTFE is that when you machine it with a tight push fit, it will heat up and expand to a total seal and take up any slight imperfections in the sleeve. So its plastic properties make the perfect seal and zero friction but at the same time is hard enough to
last a very long time (one of my engines has the same ptfe rings for the last 3 years with no wear whatsoever).

Some have made the entire valve bobbin out of solid PTFE, but that's maybe not ideal because tightening up lock nuts on either side of it on the spindle, even with washers will have the effect of squashing the valve and with the linear expansion and contraction the lock nuts seldom stay locked. Also there is a whole lot more PTFE to expand laterally and make them very tight with steam, although they're unlikely to sieze it will be very hard to move the reverser lever when not running.

So the idea is to make up a bronze (even brass) bobbin that will support ptfe rings in the correct position.

The most important thing is that the rings are not just replacing, say, cast iron rings set into grooves in the bobbin ends - they become the bobbin ends themselves...

This requires the new bronze bobbins to have metal "cheeks" to hold the pfte in place.

Imagine one end of the bobbin - in end-to-end cross-section the face of the valve is flat around the spindle hole to correspond with the lock nut diameter, then it tapers ("cones" is probably a better word) forward at a suitable angle to as close to the bore of the valve sleeve as you can.

At that point, the metal drops square back in to form the one side of the slot that takes the ptfe ring. The ring (and therefore the slot) should be at least 3/16" deep (could be as much as 1/4") for a valve bore of around, say, 7/8" to one inch.

So of course you have the other 'cheek' sandwiching the ptfe and it also has its sharp apex close to the valve bore diameter - after which it 'cones' back in to a suitable diameter for the
'steam chest' middle of the bobbin - could well be around 1/2" diameter for this size valve sleeve.

I machine them by starting off with bar exactly the same bore as the valve sleeve,
I measure up exactly where the slots/grooves should be that will take the ptfe ring and use the parting off tool to cut the slots/grooves and make sure those are exact - nothing else is quite so crucial.

If you have a running engine that's being modified, then the proof the new bobbins are correct is if you hold the old bobbin side by side with the new one, the old bobbin ends should match up exactly to the slots in the new ones, and with pressure they should be able to slide into the slots.

Now the goal is to make ptfe rings that will slot into these recesses and make a tight push fit into the valve sleeve.

As you'd expect, first bore out the correct inside diameter to match the diameter of the bottom of the slots of the bobbins, leaving maybe a few ten thou to make sure the ring will slip over comfortably. Then face off the end and neaten up with a craft knife edge to get rid of the burr without rounding the edge. Then part off the correct thickness - exactly to match the slots as a tight 'tap' fit. It's best to neaten up the other edge burr with the craft knife before parting it off completely.

There will then be another burr on the inside bore after parting off and this can be rounded over by hand no problems.

Now of course it needs to be cut so it will slip over the bobbin into its groove. A straight cut with the craft knife will
be ok but it's best to slice a "step" by making a 1/2" or so cut into the face of the ring parallel to the edges before
making alternate cuts from the "rims" in toward each side of the middle cut.

They should now split open and go over the slot and be a bit too tight to push in by hand but should go in with a gentle tap from
a small mallet or what have you.

Having been bent open to slip over the bobbin, the edges should meet again to make a good seal but will probably need some squashing in the lathe to encourage them to stay there - bend up a brass strip to a short tube close to the diameter of the bobbin that will protect the ptfe from being deformed by the lathe jaws - squash it in the lathe (the brass strip must have a small gap to allow for the squash).

It will probably be easier to have a tube or pipe reamed to the same bore as the valve sleeve, allowing you to push the complete valve into the tube and then into the sleeve without having to worry about an edge of the ptfe ring catching on the valve sleeve edge - of course not necessary if the valve sleeve end has a shallow champfer.

The complete valve should go in to the sleeve with a light coating of oil as a tight push fit or require gentle tapping.



Of course if you've got the valves out you may as well double check the dimensions of the valves before doing any of this to be sure they are exactly right for the ports and valve spindle travel. Usually with inside admission piston valves and walchaerts gear, the 'end-to-end' measurement between the outside edges that will be opening the ports to exhaust should be the same as the distance between the outside edges of the ports.

The thickness of the actual bobbins ends (or in this case the ptfe rings) should be the width of the ports plus half the total travel of the valve spindle between front dead centre and rear dead centre of the wheels.

If this is right (and assuming the eccentric rods are the right length), you only need to then make sure the bobbin is in the right place on the valve spindle. The easiest way is if you're lucky enough to have sighting/ doping holes on the side of the cylinder casting that lead to the ports. compressed air will soon tell you whether the ports are opening at the right time. With the engine slightly forward of mid-gear, the ports should crack open very slightly just past front and rear dead centre.

The end result should allow you to literally drive on the reverser with perfect timing, notching back seamlessly to mid-gear while still having almost full pressure in the steam chest - good for efficiency and superheating.

The good seal with the ptfe should allow you complete control over slipping while starting heavy loads because the ptfe seal will allow steam chest pressure to build up to close to boiler pressure whilst the regulator is only cracked open - this means once the wheels start to turn, steam chest pressure drops quickly as the cylinder is filled and avoids the need to over-open the regulator to compensate for valve leakage which invariably causes a slip.

[Deleted]

Hi Andy,

Thanks for this - it will come in very useful if I eventually need to make PTFE rings to replace the existing bronze ones. I've got the final kit now, so hopefully will be ready to steam by the start of next season. Incidentally the Modelworks bobbins have shoulders at each end to carry the rings, and disks on the ends to retain the rings, so it's not necessary to stretch the rings to fit over the bobbins.

Regards,

John

[Deleted]

As expected, the first steam test of my Britannia revealed a fair amount of leakage past the piston valves and so I plan to replace them as soon as possible with PTFE or Fluorosint rings on the bobbin ends along the lines that Andy describes above.



I think at least part of the reason for the leakage may be that the existing rings are quite narrow, probably hardly wide enough to cover the ports, and hence probably allow steam to flow from inlet to exhaust when the rings are in line with the ports. The steel bobbin and end disks are a reasonably close fit in the liner but certainly not steam tight, so their edges don't really define the valve events as they should. In asking around for advice on PTFE rings I've come to realise that these should be the full width of the bobbin ends, held in place by conical sections on the bobbin.

Regards,

John

[mutley]

Hi John
In my understanding valve events are determinded by the rings and not by the bobbins. I'd do some measuring first and compare to a set of drawigns.

Andy

[paultomlinson]

This is my last post on this forum. There are too many theorists and armchair engineers on this forum, 99% of which is absolute garbage. There are only a couple of members on this forum thats worth listening to. just read the post before this one,this certainly isn't one of them.
Any amateur reading advice on this forum could make some potentially very expensive mistakes. I do not have time to read such rubbish, I have been a member for just a year but I shall not be looking at this site again.
Please any serious MODEL ENGINEER be very careful when you read this forum, really ought to come with a health warning!!
Paul
ps No point in replying to me personally as I will not be reading it.

[Deleted]

John,
Interesting picture of the bobbin and rings in their original state. I am building the 7&1/4" Winson Britannia and was advised to replace the brass rings with PTFE ones (thanks BB). I have assembled the cylinders and will start on the motion next week and hope to airtest within a couple of weeks. I will post a picture of the rings I made next time they are out. I replaced the 1 Winson ring at 12.4mm width with 2 PTFE ones at 6.2mm but never cut any grooves in them. Its trial and error at this stage.
Regards
Jim

[Deleted]

Andy,
The overall length of the bobbin including the steel shoulders agrees to the Spink plans. If I remember rightly the Spink plans show a plain bobbin without rings.

Paul,
That's a bit of a conversation-stopper, but thanks for all your help in the past!

Jim,
Someone else has told me that the Winson rings on the 5" Brit covered about twice the width of the 2 thin ones in my photo. I guess that would be a big improvement, but I'll go for rings the full width of the bobbin end.

Regards,

John

[mutley]

Hi John
Check were the ports are in relation to the perrier drawing.

Paul before you throw your toys out the pram you are not captain of this board!
Each model design tends to be diffrent. Looking at the drawings for Tony Allcocks Jubilee, a model you have built, the dimension over the outer edges of the ports is 2.30" . The dimension over the edges of the two outer most rings is also 2.30" The bobbins have a further 0.050" in length, with according to the drawings, a 'shamfer' on the edge. Now you must know as well as I do that you need a sharp edge on the valve for clean opening and closing of the ports which you will not get with a shamfer. So what, I am left wondering is controling the port openings if it isnt the rings?. Any thoughts somebody? I'm prepared to be proven wrong.
I also find it rich that you sit there and decry the posts of other members when you yourself have only made 61, well over 50% of which have been about your Modelworks Duchess. If you have so much information and advice to pass on then why havent you been more forthcoming in correcting some of the incorrect theorys and try to prevent some of the expensive mistakes happening?

Andy

[baggo]

Looking at this objectively, I would tend to agree that the port opening probably is controlled by the rings rather than the valve bobbin. If it is determined by the edge of the bobbin this implies that the bobbin is a perfect fit in the valve liner and therefore there is no need for the rings at all

As Andy (Pondok) says in his excellent post above on PTFE rings, it is very, very difficult to get a plain bobbin to seal perfectly so there is always going to be a certain amount of leakage past the bobbin which is going to affect the valve timing.

John

[mutley]

Hi Baggo
I came to a similar conclusion last night. The Peier Brit was alway designed for the experinced engineer and called for tight tolerances. I made a new set of bobins for my farthers Brit a few years ago and I had a few goes getting the tolerances close then. Usually a good supply of steam oil is relied upon to effect a seal.
I think rings were adopted to try and make things easier for the modeller. Bore valve guide to 0.994" and then run a 1" reamer through or hone to size. Machine bobbin as close as you can, however 1" piston rings will do the sealing for you.
The other point that maybe worth pointing out is that although the Modelworks Brit appears to be a copy of Perrier's Brit unless all the diffrent parameters are carfully measured you cant assume that it was built exactly to the same dimensions and to work in the same way. Did Arthur, the desginer, intend the rings to do the opening or closing of the ports or the bobbin face.

Andy

Andy

[Deleted]

Andy, John,

The ports also agree to the Perrier/Spink drawings.

Doug Hewson has sent me a photo of the PTFE valves that he made for Winson Brit customers. They are to much the same design as Andy (Pondok) describes above, and the PTFE covers the full width of the bobbin ends.

I steamed on the rolling road again today and the valve leakage didn't seem quite so bad this time, so I'll try it out on the track before changing the valves.

Regards,

John