|
Post by Doug on Apr 21, 2015 15:08:52 GMT
dug007red : I think we've identified most of problems with using ptfe in steam cylinders but not yet come up with any complete solutions . For the remaining part of this running season I strongly suggest that you do a quick fix using soft packing or O rings . My preference would actually be soft packing - at least it is a known quantity . There are now more uncertainties over use of ptfe than ever . Perhaps we should all be looking in another direction . I'll have to have a think about that . MichaelW Hi Michael I dont have too many issues with the rest of the running season i will be getting very severe hay fever in the next couple of weeks which will stop all running till Aug, anyway Butch is my runnig test bed for Speedy I am learning whats important and an issue with running a loco once Speedy is built Butch is going to retire.
|
|
|
Post by Roger on Apr 21, 2015 15:34:20 GMT
This is where I started with my design but rejected it because it leaks so badly when cold. You need a really large ring gap and I didn't think it was acceptable for it to leak that much until it warmed up. If that doesn't bother you, then it's worth pursuing. Roger, it bothers me. And we are all looking for the best -yet simpler- possible approach. You keep the lead by actually having tested things, so: Why do you think that this leaks when cold?. If there is no radial gap -the oring sealing that- and the two contiguous PTFE rings are set at different angles, where is the passage that causes it to leak?. I can not see it. I am referring to main piston, not the valve. Thanks. I'm not sure we're talking about the same thing, there has to be a radial gap ie there is a gap between the open ends of each ring. In that case, steam can go all the way along the first ring gap, past the 'O' ring that's underneath it, and then up through the gap in the next ring. ie there's a path that goes over each ring so there's no seal. It leaks a lot because I've tried it. That's why I ended up trying to block the path from under the rings. This is the same for piston rings or piston valves, the problem is the same.
|
|
|
Post by Roger on Apr 21, 2015 15:42:42 GMT
Hi Foulersfury. I have already referred to that article in this same thread. See this. modeleng.proboards.com/post/122177/thread Roger suggested that one possible issue was an unappropriated shape of the valve ports, and/or no rings internal clearance. The author also seems to assume that ring expansion would small and thus no provision for it is made in the design. As already posted before the original link for this article is found here: www.modelengineeringwebsite.com/PTFE_valve_rings.html Thanks to all who read through my long-winded post; a few additional points.... (1) I was the author of the Model Engineering Website article, so no question of plagiarising "Crankpin"! David Carpenter doesn't have a 'Letters' page of course but he agreed to put the article in should anyone have any input but none was forthcoming. The glass-filled Fluorosint piston valve 'rings' are shown on there. I modified the design of Geo Thomas for the valve itself as his idea of easy axial adjustment along the valve spindle seemed very good. (2) On this question of thermal expansion of virgin PTFE and glass or mica filled PTFE, I hadn't meant to imply the data sheets indicated little or no expansion. The data sheets showed that under 150 oC the expansion was significantly less than above that temperature. Accordingly, under conditions of superheated steam at 150psi in the compound, the Fluorosint expansion was likely to problematic to say the least. I found the most instructive information on the filled PTFE polymers came from Quadrant Plastics. linkThe comparative data on thermal expansion is:- Fluorosint 207: Coefficient of linear thermal expansion: – average value between 23 and 100 °C m/(m.K) 85 x 10-6 – average value between 23 and 150 °C m/(m.K) 90 x 10-6 – average value between 150 and 250 °C m/(m.K) 155 x 10-6 (3) I did think seriously about using graphited "yarn" since like many of the posters in this thread, I believe it has many desirable properties. In fact on an earlier engine build, I had used 1/4" square graphited "yarn" in GM cylinders. I did find though, maybe because I'd packed it too hard, the engine ran very tight when cold. Once warmed through, it ran freely enough. Whatever, I didn't have enough remaining in stock for this build and at the time, thought it was no longer available. (4) My reference to Dupont's Kalrez O rings. On the basis of their data sheet, Kalrez would fulfill all the demands of HP superheated steam, GM cylinder bores, presence of oil and longevity. linkBut being a tight-fisted model engineer, the very high prices couldn't be justified. Furthermore, whenever I searched on Fleabay for cheaper Kalrez O rings, they were never available in the sizes I needed and couldn't be sure they were genuine DuPont and not grey imports. Yet....I may live to rue my decision because the propsect of a future, total strip down to replace the Clupet piston rings and Fluorosint valve 'rings' fills me with dread, never mind the cost already incurred for the 6 Clupet rings and 12" of Fluorosint . Just two points... 2) I think it's going to be really difficult to get thick solid rings to work in any of the engineering plastics because of the large thermal expansion. I can't see that they would ever seal when cold as well as hot. 4) When you talk to manufacturers of 'O' rings, they point out that 'O' rings aren't really suitable for this kind of dynamic application. Having said that, they would probably work better than Viton ones that appear to give a couple of seasons life.
|
|
|
Post by andyhigham on Apr 21, 2015 15:58:41 GMT
I think solid rings are a non starter. I just did a quick calculation using a bore diameter of 40mm, a linear expansion of 1%, a ring depth of 3mm and disregarded the fit of the piston in the bore. For a solid ring the difference in diameter would be 0.4mm giving a leak area of approx 25 sq mm ((40/2)*(40/2)*3.14))-((39.6/2)*(39.6/2)*3.14)) for a split ring the difference in end gap would be 1.256mm giving a leak area of 3.7 sq mm (1.256*3)
|
|
jma1009
Elder Statesman
Posts: 5,901
|
Post by jma1009 on Apr 21, 2015 22:01:34 GMT
if it is any help, the late Jim Ewins' groundbreaking 0-6-2T in 5"g had gunmetal cylinders and soft (square braided graphite) packing fitted, and is one of the best locos ive ever driven.
Jim always challenged everything that was accepted practice, and i am sure had he thought there was a superior alternative to soft packing to the pistons he would have used it.
i think it is a mistake to think that problems with piston valves apply to pistons.
cheers, julian
|
|
|
Post by Roger on Apr 22, 2015 6:41:34 GMT
if it is any help, the late Jim Ewins' groundbreaking 0-6-2T in 5"g had gunmetal cylinders and soft (square braided graphite) packing fitted, and is one of the best locos ive ever driven. Jim always challenged everything that was accepted practice, and i am sure had he thought there was a superior alternative to soft packing to the pistons he would have used it. i think it is a mistake to think that problems with piston valves apply to pistons. cheers, julian Agreed, piston valves and piston seals have very different requirements and pistons are much less troublesome. I wouldn't accept that just because someone else wasn't able to come up with something superior to soft packing, that it can't be done though. It's a great fall back position to have that solution but I'm sure there's a more subtle one.
|
|
|
Post by fowlersfury on Apr 22, 2015 12:03:04 GMT
For what it’s worth, here are some truncated exchanges I had a couple or so years ago with specialist ring suppliers and the Copper Development Association regarding the most suitable material for valve and piston rings in GM bores with HP (150+psi)superheated steam and uncertainties about lubrication under those conditions:- <><><> DuPont Kalrez O rings. Data sheets suggest they have all the properties necessary for use in such an application. Quotation for piston rings only ! “6 off 31.7x3 Kalrez @ £43.19 each with a 4-5 week lead time. All prices are subject to delivery and VAT”). www.polymax.co.uk<><><> Copper Development Association; following an enquiry for their advice about GM bores:- “We are not experts in piston ring materials and manufacture. I have looked at the details of a specialist in the manufacture of piston rings, namely, Precision Products,Unit 1, Cobnar Wood Close,Chesterfield S4 9RQ email sales@ringleader.net. They use phosphor bronze- I think that this would be PB102 or PB104.” Their comment and quotation:- “GM bores are not common. Normal piston ring and cylinder material is alloyed cast iron. I have come across gunmetal piston rings working with a cast iron liner, but not the other way round. However the application of superheated steam is not conducive to easy lubrication. In view of the lubrication problems, A PB ring would be recommended in a cast iron liner, but PB rubbing against GM is an unknown. The rings would not be a problem to manufacture, but for a quantity of 6 would cost approximately GBP45.00 each ex works and with a delivery time of 4-5 weeks PP Ringleader”. <><><> A further and more reasonable quotation:- “ We are pleased to quote as follows:- 6 - Phosphor Bronze Piston Rings, 1.490" dia. x .123/.122" wide x .058" radial thickness with butt joints. Price: £18.96 each. Delivery: 4 weeks P&P £8.00 All prices are plus VAT Bradford Piston & Piston Ring Co.Ltd.” <><><> And another quotation:- “You are correct meehanite would be too hard in GM Liners. We have these solutions: 1.490 x .125 " Compression Rings Material Phosphor Bronze 6 pcs GBP37.85 each Delivery 4 weeks. HP Rings Ltd, www.hprings.com ” <><><> So….with all that lot to consider, especially the prices, I examined again all the published articles in M.E. and E.I.M on polymer rings. By this time I had no idea how to proceed other than scrapping the carefully machined GM cylinder castings and buying CI ones. I considered making CI liners for the cylinders and valve chests but the differential expansion of the 2 metals would represent another potential problem. In the end I phoned the ex-metallurgist Alan Smith who runs the Clupet Piston Ring & Gauge Co. He was of the opinion that centrifugally cast iron rings in a GM bore should not present a problem. I ordered 3 sets from him and used glass-filled PTFE for the piston valves, more in hope than expectation. As before, I’m far from convinced there’s not going to be trouble ahead.
|
|
|
Post by Deleted on Apr 22, 2015 12:29:23 GMT
Hi fowlersfury ,
As mentioned in a much earlier post chrome plated conventional rings will work perfectly in bronze cylinders .
MichaelW
|
|
|
Post by fowlersfury on Apr 22, 2015 16:47:25 GMT
Michael, Thanks for your response. I haven't been able to find your reference thro' the Search box. I had, & still have, reservations about Cr plated rings with GM, not necessarily based on personal experience but from when doing some research into valve guides made from Colsibro etc against Cr plated valve stems in high speed ICEs. Undoubtedly hard chrome plating is extremely wear resistant and using Cr plated rings will wear the GM bores. My probably false reasoning was that I'd rather eventually replace the rings than have to rebore the cylinders. The aforementioned metallurgist Alan Smith of Clupet was adamant that the softer metal (GM) would outlast the harder one (his centrifugal CI rings) due to particles being embedded in the softer GM. I couldn't envisage that happening with hard Cr plated rings. However it's appreciated that an oil layer at the interface could have a profound effect on relative wear. But then I've yet to identify a lubricant that won't char in the superheater wth steam at 150+ psi. Adding powdered graphite or molybdenum disulphide to a high temp steam oil might be a solution.
Ian C.
|
|
|
Post by Roger on Apr 22, 2015 17:09:29 GMT
This is where the old fashioned packing or PTFE solutions are to be preferred in my opinion, because they are unlikely to damage the bore. Unless I was really sure about the compatibility of the ring material with the cylinder, I wouldn't risk using metal rings. I suppose a small test cylinder and rings could be made and an endurance test set up to run day and night to prove that a material combination is viable. Maybe something simpler like two materials in question loaded while in an oil bath and moved against each other.
|
|
|
Post by Deleted on Apr 22, 2015 18:30:07 GMT
(1) The rings are not Chrome plated to make them hard they are Chrome plated to make them very smooth and to provide a rubbing surface coated in one of the very few metals that will not pick up when rubbed along a gunmetal/bronze cylinder bore .
(2) There is very little radial ring to bore load to cause wear with any type of properly designed ring .
(3) The bearing properties of Chrome come in the excellent class when mated with a whole range of common materials .
(4) The old folklore about which hard/soft members of a bearing set wear more/less is just that - folklore .
(5) The ideal Chrome plated ring would be broad in the axial direction and relatively thin radially to make it truly flexible .
|
|
jma1009
Elder Statesman
Posts: 5,901
|
Post by jma1009 on Apr 22, 2015 19:28:41 GMT
i dread to think the cost of chrome plating phos bronze (or cast iron if that is possible!) piston rings.
there is guy selling the old fashioned square section braided graphite packing on ebay at approx £3 per 12".
cheers, julian
|
|
oldnorton
Statesman
5" gauge LMS enthusiast
Posts: 693
|
Post by oldnorton on Apr 22, 2015 19:38:14 GMT
Dug007red,
If you still have the 3/16 square groove in you Butch piston then I recommend 3/16 square woven PTFE packing from Reeves. It is what I used last year in the engine pictured with me on the left and it seems to work very well. The 3/16 material was very accurate and although compliant and 'shapeable' required a precise groove to allow the pistons a tight sliding fit. I had to adjust the grooves as mine were a few thou undersize. I cut the woven material at a diagonal on a mandrel that was the same diameter as the base of the ring groove. I made a tapered bore adapter to compress the rings and enable them to slide in.
I was concerned that the PTFE would expand and the cylinders would tighten up when hot, but no, it was completely free and smooth when hot. I guess the woven material has a fair proportion of airspace in its construction that takes up the expansion.
After several tens of hours running last year, this winter I had it on the bench running on air, i.e. cold. The rings still appeared to be sealing cold as it ran fine on a light regulator.
I have always felt that solid pure PTFE rings are the wrong application of an otherwise useful material. The filled PTFEs are better, but still the coefficient of expansion works against them. Only when the PTFE is a thin ring sitting on top of an o-ring will it seal at a range of temperatures.
I am constructing my future engines with cast iron cylinders, pistons and rings - it works and lasts so why mess with anything else?
Norm.
|
|
|
Post by Roger on Apr 22, 2015 20:01:22 GMT
Dug007red, If you still have the 3/16 square groove in you Butch piston then I recommend 3/16 square woven PTFE packing from Reeves. It is what I used last year in the engine pictured with me on the left and it seems to work very well. The 3/16 material was very accurate and although compliant and 'shapeable' required a precise groove to allow the pistons a tight sliding fit. I had to adjust the grooves as mine were a few thou undersize. I cut the woven material at a diagonal on a mandrel that was the same diameter as the base of the ring groove. I made a tapered bore adapter to compress the rings and enable them to slide in. I was concerned that the PTFE would expand and the cylinders would tighten up when hot, but no, it was completely free and smooth when hot. I guess the woven material has a fair proportion of airspace in its construction that takes up the expansion. After several tens of hours running last year, this winter I had it on the bench running on air, i.e. cold. The rings still appeared to be sealing cold as it ran fine on a light regulator. I have always felt that solid pure PTFE rings are the wrong application of an otherwise useful material. The filled PTFEs are better, but still the coefficient of expansion works against them. Only when the PTFE is a thin ring sitting on top of an o-ring will it seal at a range of temperatures. I am constructing my future engines with cast iron cylinders, pistons and rings - it works and lasts so why mess with anything else? Norm. All good, but the answer to the last question is..... Rust!
|
|
jma1009
Elder Statesman
Posts: 5,901
|
Post by jma1009 on Apr 22, 2015 20:18:35 GMT
you would actually need 5mm square section braided packing in a 3/16" slot on the pistons. 3/16" stuff (it varies in size) would simply 'fall in' and do nothing at all. i mike up the square braided stuff of the old fashioned variety of soft graphite packing after a soak in steam oil, tightening the jaws of the mike in the process, then machine the recess in the pistons to suit with a few trial fits whilst the piston is still set up in the lathe.
the initial fit (soaked in steam oil) should be very tight in the cylinders, but after awhile moving up and down well lubricated with oil the fit will have compressed to a working fit requiring just palm pressure. by the time the loco has been tested on air (with the correct oil) the packing has settled down and will be long lasting and very durable. soft packing has the advantage in gunmetal or bronze cylinders of absorbing oil and so helps create an effective oil film in what are otherwise quite harsh conditions for oil notwithstanding use of modern steam oils.
there was a very interesting contribution to ME in 1968 that steam oil wasnt suitable for gunmetal or bronze cylinders! just to throw another spanner in the works!
cheers, julian
|
|
|
Post by Doug on Apr 22, 2015 20:53:30 GMT
I can see myself trying graphite square section again, my "string" did work very well even if it did get cleaned by the steam so it ended up as just string there is a solution in this some where just have to keep trying things till I am happy it works. I am am getting a very good selection of photos of my Pistons with various sealing materials.
|
|
jma1009
Elder Statesman
Posts: 5,901
|
Post by jma1009 on Apr 22, 2015 21:15:42 GMT
hi doug,
all i can say is that one of my locos with gunmetal cylinders and phos bronze pistons and soft braided square section graphite packing of the old fashioned asbestos (or perhaps not asbestos in light of Reg's posts) came 3rd in IMLEC in 1995.
as LBSC would say...
nuff sed!
cheers, julian
|
|
|
Post by ejparrott on Apr 22, 2015 22:41:33 GMT
An interesting (maybe) note on hard chrome plating - we have a number of things that are plated and ground for one of our customers in the power turbine industry. Whatever finish you want on the plating, needs to be on the parent metal under the plating. We've noticed that despite plating and grinding, the surface finish beneath always shows through.
|
|
|
Post by fowlersfury on Apr 22, 2015 23:30:56 GMT
(1) The rings are not Chrome plated to make them hard they are Chrome plated to make them very smooth and to provide a rubbing surface coated in one of the very few metals that will not pick up when rubbed along a gunmetal/bronze cylinder bore . (2) There is very little radial ring to bore load to cause wear with any type of properly designed ring . (3) The bearing properties of Chrome come in the excellent class when mated with a whole range of common materials . (4) The old folklore about which hard/soft members of a bearing set wear more/less is just that - folklore . (5) The ideal Chrome plated ring would be broad in the axial direction and relatively thin radially to make it truly flexible . Regarding your point (4), IF I understand your point correctly, I don't accept it's "folklore". There is too much published evidence (in peer-reviewed journals) to claim otherwise. To quote just one article in Tribology International Volume 81, January 2015, Pages 258–266 "The influence of surface hardness on the fretting wear of steel pairs — Its role in debris retention in the contact" by J.D. Lemma, A.R. Warmuthb, S.R. Pearsonb, P.H. Shipway. The Abstract states:- " The influence of specimen hardness (between 275 kgf mm−2 and 835 kgf mm−2) in an AISI Type O1 steel-on-steel fretting contact was examined. In equal-hardness pairs, a variation in the wear volume of around 20% across the range of hardnesses examined was observed. However, in pairs where the two specimens in the couple had different hardnesses, a critical hardness differential threshold existed, above which the wear was predominantly associated with the harder specimen (with debris embedment on the softer specimen surface). This retention of debris provides protection of that surface from further wear and also results in accelerated wear of the harder counterface due to abrasion by the oxide debris bed which has built up on the opposing specimen."
|
|
|
Post by Deleted on Apr 23, 2015 8:32:33 GMT
What actually happens as regards levels of friction and wear between two elements in a sliding bearing situation depends on a large number of factors .
Some combinations of factors will produce one result and other combinations of factors a different result .
|
|