Talking Thermodynamics

[paul gough]

Hope I'm not deluding myself on the figures I presented: Valve dimensions, 0.457 x 0.185 = 0.084 sq. ins. x 60 psi = 5.04 lbs.
Underside of valve (or load bearing surface): 0.084 minus area of steam passage (0.285 x 0.091 = 0.026), thus 0.084-0.026 = 0.058, rounded off to 0.06 sq. ins. Thus approx. 5lbs or 2.2kg load on 0.06 sq. ins. If I'm wrong, please help, my old brain can't see where.

Don't know that I could get inside admission type into the space available, the outside admission type seems just possible, maybe, subject to achieving satisfactory sealing. I am thinking of one of the engineered 'plastics', perhaps a graphite embedded teflon or nylon or some such for the piston heads and valves. Regards, Paul Gough.

[Admiral_dk]

Human Power - I kind of thought I knew, but ..... ended Googling it and according to Wikipedia https://en.wikipedia.org/wiki/Human_power

A trained cyclist can produce about 400 watts of mechanical power for an hour or more, but adults of good average fitness average between 50 and 150 watts for an hour of vigorous exercise. A healthy well-fed laborer over the course of an 8-hour work shift can sustain an average output of about 75 watts.

Quite a range - from not very much to rather impressive.

[Maryak]

Last time I was looking at where the heat goes in a boiler, this time let's look at the losses in the engine. 

Heat losses Marine Steam Plant



Sometimes a picture is helpful.

Regards Bob

[derekwarner]

Whilst a little simplistic on labyrinth grooves [for model steam engine piston valves] , I posted the following on August 6/2017 on Mammod Steam Models site
____________________________________________________________

To take this further here is some light reading

https://www.google.com.au/url?sa=... FQjCNEkXGWRYQuZJVuLfNb_X_LrJiyyrw

However for some further understanding we go back to Gas Laws..

 P1 V1 = P2 V2...

 P1 = steam pressure of the system
 V1 = the annular volume in the piston valve spool
 V2 = the annular clearance volume between the valve spool and the valve body

So we have P1 & V1 doing their work  with each stroke created by the eccentric movement

Then the same P1 travels along the minute [a ~~0.001"] annular clearance path V2, then the steam literally falls into the larger volume of the groove

This pressure drop then creates a reduced value of P2 in the groove........

Since we are talking relatively low steam pressures here, the value of P2 may well approach that of atmospheric pressure [1 Bar] and hence the body of steam in V2 will partially condense and simply shuttle back & forth as a lubricant and being contained by atmospheric pressure

The reason this steam does not spray out is that the new pressure P2  is in ~~ balance with atmosphere 

Of course, a little condensate will migrate out along the valve spool .....thus maintaining the Gas Law balance

Full sized higher pressure steam turbine engines utilize similar multi Labyrinth sealing arrangements
______________________________________________________________________________________________

Derek

[paul gough]

Derek, I keep getting an invalid URL notice if I try to go to your 'light reading' site. Could you check there is no typos etc. Regards Paul Gough.

[derekwarner]

Sorry Paul........here is the full extension thread link....I don't understand why copy & paste did not function???....

As you can see, the actual link at the very end ...FQjCN is the same leadin  ........Derek

http://redirect.viglink.com/?format=go&jsonp=vglnk_150433320089612&key=190cd6dc27c66f8edf4cd0d943583f3b&libId=j72x4w9101000abj000DAb8zr06z2&loc=http%3A%2F%2Fmodelsteam.myfreeforum.org%2Fabout97187.html&v=1&out=https%3A%2F%2Fwww.google.com.au%2Furl%3Fsa%3Dt%26rct%3Dj%26q%3D%26esrc%3Ds%26source%3Dweb%26cd%3D1%26cad%3Drja%26uact%3D8%26ved%3D0ahUKEwi-pam4ocHVAhXGG5QKHfMdCBoQFggmMAA%26url%3Dhttps%253A%252F%252Fen.wikipedia.org%252Fwiki%252FLabyrinth_seal%26usg%3DAFQjCNEkXGWRYQuZJVuLfNb_X_LrJiyyrw&ref=http%3A%2F%2Fmodelsteam.myfreeforum.org%2Fforum7.php&title=The%20Unofficial%20Mamod%20and%20Other%20Steam%20Forum%20%3A%3A%20Piston%20Valves&txt=https%3A%2F%2Fwww.google.com.au%2Furl%3Fsa%3D...FQjCNEkXGWRYQuZJVuLfNb_X_LrJiyyrw

[Steam Haulage]

Hi All,

Labyrinth seals
Please note my liberal use of ‘may’ and ‘perhaps’ in the following.

May I suggest that the concept of the labyrinth seal goes back further than we might appreciate. Perhaps even before Jerónimo de Ayanz y Beaumont, Thomas Savery, James Watt et al were conceived.

It may be that the idea goes back into the mists of time in the joinery trade of our forefathers.

I have seen many examples of grooves in timber door frames. Before the adoption on a large scale of plastic as a construction material almost every frame would have a semi-circular groove cut into the hinge, lock and top of the frame. Apprentice joiners were instructed by the journey man they worked under that this was to allow winds to enter any gap and then reverse its direction and so prevent the draught from entering the inside of the building.

When I was a lad this groove was in common use, and visible in every house in which my family and relations lived, as well as in every door-frame at the school, built in 1909, which I attended in the fifties.

Steam Haulage

[MJM460]

Hi Paul, now I see where you are coming from on the area and force on your slide valve.  I don't see myself as in any position to arbitrate on right and wrong, but I would tackle the problem slightly differently.  I will leave it to you to decide on whether it is helpful.  The slide valve has two sides, just like the piston.  The steam chest side is easy, steam chest pressure over the area of the valve gives your 5.1 lbf forcing the valve against the valve face.  The other side of the valve is more complex.  The pressure in the valve cavity is a bit higher than exhaust pressure, higher by the pressure required to drive the exhaust flow through the port to the exhaust system, so we can estimate the net force on the valve over the exhaust cavity area as a bit above exhaust pressure.    However the sealing faces of the valve are more complex.  I suspect that there is a thin film of steam and oil separating the faces however slightly, and the pressure in this film varies perhaps roughly linearly between the valve chest pressure on the outer perimeter, and the valve cavity pressure around its periphery.  On this basis, I would estimate the average pressure over the sealing face is about half way between steam chest pressure and valve cavity pressure, tending to lift the valve, so tending to partially balance the pressure on the steam chest side.  Then, as I mentioned yesterday, the valve rod does not have to resist the normal force, but only the friction force, so now we have to estimate a suitable friction coefficient.  I hope this helps.

Yesterday, I mentioned inside admission, because the steam pressure forces on the piston valve are better balanced in the inside admission arrangement and the rod sealing and unbalanced rod forces are determined only by exhaust pressure.  However the challenge of making piston valves with sealing rings on such a small engine would be totally beyond me.  I suspect that you would be better to stay with the simplicity of slide valves, and if necessary, increase the piston diameter slightly to provide the extra force.  Avoiding the issue, I know.

Thanks for joining in Admiral_dk, glad to have you on board.  I think you have answered yesterday's question on human power.  Now that you have put the figures in, I seem to remember being on the tread mill for a stress test after a heart attack many years ago, closely monitored by a cardiologist and all the instrumentation, and being told that the load, which was provided by a small generator was about equal to a 60 watt globe, but fortunately I only had to keep it up for about 15 minutes.  But definitely at the low end as you might expect in the circumstances.  For an 8 hour day, I suspect your 75 watts is about a reasonably good effort when it has to be kept up all week.

Hi Maryak, thanks for another of your excellent diagrams.  It is very helpful in showing all the losses in proportion.  Interesting to note that the final figure driving the ship is only 6%, but I suspect in model sizes, the mechanical losses are a bigger proportion.  It is one of those wrinkles in the maths, that if we say halved the mechanical losses, it would not make much difference to the overall efficiency, but all the reduced loss would appear as extra output power so would have a big percentage effect on the out put power.  That is what is behind my intent to look at some of the losses that we can do something about.  At the end of the day however, as your diagram shows, the major losses are in the heat contained in the exhaust, followed by the heat carried up the stack.

Hi Derek, it is good to see someone else having a go at explaining some of the many little puzzles we come up against in every design.  If we can nut out the theory, it will definitely help inform out experimentation.  As with Paul's problem, I would take a slightly different approach, and would start with a diagram showing where conditions are known, specifically the top and bottom edges of the piston.  This pressure drop forces the flow past the piston.  Each time the flow encounters a groove, it slows down, but with this geometry, no pressure recovery.  There is then a small pressure drop to supply the energy to reaccelerate the flow into the next section of the annulus, followed by a steady pressure loss towards the next flow.  The pressure cannot get below exhaust pressure, but is always a little lower when there are grooves, so restricting the leakage flow a little.  I hope that helps your thinking a little.

Hi steam haulage, good to have you aboard also.  A truly interesting little piece of history around those grooves.  I expect the journeyman learned the explanation when he was the apprentice and so on back to time immemorial.  I suspect neither he nor his apprentice or their forbears really understood the explanation.  I suspect that we can prove using the law of conservation of energy that the wind cannot be turned back with enough energy to flow back on itself, however that does not mean there is not a labyrinth effect there that means the groves might reduce the draft a bit, and that would always be welcome.  I hope that I have continued your most appropriate use of perhaps and may.

Once again, time and words have run away, so perhaps next time for looking at what we can do on our engines.

Thanks for reading along,

MJM460

[steam guy willy]

Hi I have now spent all day looking at doors in Medieval city of Norwich and have not found any of these grooves yet !!! In the previous posts there is mention of grid iron valves these were used as there was less movement in the valve so contributing to greater efficiency ,however this is from the 1881 book...so would this be correct. A local engine builder uses PTFE  'rings' on his piston valves and found that the diameter should be slightly less than normal rings due to the expansion of the PTFE. lots of good info coming on this site. keep it coming !! Also this was the first i have heard of gridiron valves in the last 70 years !! How would the efficiency compare with a diesel engine in a ship ?? or a steam turbine ?

[derekwarner]

Gents......my earlier posting in Mamod Models from August the 6th, was in answer to a question asked by another member as to what the two fine grooves on a model steam engine piston valve spool by way of the gland area could be?

I simply offered a thought and basic explanation of the pressure drop that occurs when a series of labyrinth grooves are used in fluid sealing applications

As you have noted MJM, the pressure reductions contained/constrained by a series of labyrinth grooves cannot exceed 'exhaust pressure', so again considering the model steam engine application, and from my recent trials of excessive making of condensate  one could expect exhaust steam pressure [within the system] to be at the physical exhaust point be approaching atmosphere

In this low pressure [say 2 to 3 Bar] application, the resulting steam/condensate [as liquid] trapped in the labyrinth grooves will tend to act as a lubricant and buffer against physical displacement outside of the gland area

As mentioned, on my return to NSW, I will conduct the series of tests with the larger flow capacity exhaust system.... Derek

Derek

   

[Maryak]

Hi All,

In full size turbines the labyrinth glands on the turbines are balanced to atmosphere with gland steam.

Often LP turbine admission is at the centre, (to balance axial thrust), with steam flowing out to the ends hence the glands are not much above condenser pressure. Without sealing steam being supplied to the glands condenser pressure would tend to rise towards atmospheric pressure causing a drop in plant efficiency.

Basically all the glands in the turbine set are interconnected at their inner and outer pockets and the gland steam system is balanced to maintain minimal leakage of gland steam from the the outer pockets. The fine adjustment valve shown below is in modern plant connected to a controller which automatically maintains the system in balance over varying load/speed ranges.....stop, ahead and astern



Regards Bob

[MJM460]

Hi Willy, I wonder if the door detailing, being passed down from master to apprentice, became a regional thing.  I am not sure just what a grid valve looks like, however if the design enabled a shorter stroke, it would reduce the work done by the valve rod.  Remember the definition of work, force times distance.  Even with the same normal force on the valve and the same friction coefficient, half the stroke means half the work dine by the rod and half the parasitic power lost to the engine output from this cause.  A possible answer to Pauls dilemma is to reduce the valve travel, though there are practical limits on such small models.  Valve rod forces have always been known to be high however.  The early joy valve locomotives were I believe known for breaking the con rod due to the side force from the valve linkage.  Or perhaps it was just poor design.  My text book has worked examples for the efficiency of steam plants with 2 and 4 MPa operating pressure, turbine drives and 10 kPa exhaust pressure, quite good vacuum.  2 MPa gave about 30%, 4 MPa gave 35%.  Interestingly, the reheat example gave less than 1% more, but is favoured in steam plants because it gives much drier steam at the low pressure turbine exhaust, so less blade erosion by the condensate droplets.  I think reciprocating machines are a bit higher efficiency than turbines, but they are limited to much lower speeds and power outputs.  My somewhat unreliable memory recalls large industrial Diesel engines could achieve about 30% when I was using them for compressor drives.  In a simple open cycle gas turbine, about a third of the energy goes out the exhaust as tonnes of hot gas, a third is consumed by the compressor at the air inlet, and a third is available as output power, very roughly.  Again this is from using them as compressor drives.  As jet aeroplane drives, there is no mechanical output, apart from that consumed by the compressor on the front end and the fuel pumps, the power inherent in that mass of gas exhausting at the back end provides the thrust. 

We do not get this performance from our steam models due to the low boiler  pressure and normally atmospheric exhaust.  I am not advocating that we try and emulate these pressures, let alone the much higher pressure used in modern industry.  A typical petrochemical plant has steam systems operating at 10 MPa, while utility scale power plants are built to supercritical pressures, that is above 22 MPa.  Steam at any pressure is dangerous, and these very high pressures are extremely so.  Fuel costs for a model are very reasonable as the running time is generally low.  We do not need to go to those extremes in the pursuit of fuel efficiency, but it is interesting to work out just what we can do to increase the power output within the limits of our operating conditions.

Hi Derek, my little Mamod engine is a single acting oscillating engine.  I am not familiar with the design of piston valves for such small engines.  I am sure the small size makes them quite different from the common designs for the larger locomotives.  I really can't picture the grooves in the gland area that you are talking about.

Hi Maryak, gland systems can get quite complex as shown by your diagram.  But I am sure that conditions in the boiler room were unpleasant enough without extra steam leakage from shaft seals.  At the end of the day however, you have reminded us that labyrinth seals still leak, just a bit less than a plain shaft or cylinder.  A well made mechanical seal of appropriate material is still the far better solution.  I am sure that perhaps carbon rings from that material used for Stirling engine power Pistons could be made for this size, but until my skills develop, I turn the grooves as they should be better than plain pistons.

Back when we were all sidetracked by the mention of labyrinths, I had talked about some of the ways heat is lost before it is turned into work, so next I think it worth talking about some of the ways we lose work output by consuming it in the engine before it gets to the output shaft.  It always seems a bit unfair when there are so many limits to how much of our energy can be converted to work, that we then lose a significant fraction of the work we do produce to friction within the engine.  The labyrinth discussion was really prompted by the desire to reduce friction between the piston and cylinder without losing too much work due to steam bypassing the piston.  Piston rings are intended to reduce the flow, but at a cost of causing friction.  Making good rings is a challenge for a beginner, for me, anyway.  Of course a packing can also be used with a bit of judgement on how much can be added without making it too tight.  Some people advocate using o-rings, and whatever works is helpful, though it was not what O rings were designed for.  My guess is that they are initially a touch tight, and some rubber wears away to leave the clearance "just right", when they are better on a small model than any poor fitting alternative.  The concept of an abradable seal is actually used in full size compressors as an alternative to machining the seal to perfectly fit each shaft, so there is even full size precedent, even if it is in shaft seals.  Cylinder lubrication helps reduce friction, and Derek has already mentioned that the moisture in wet steam helps lubrication.  That is why it becomes more critical to use the right oil if we have a lot of superheat, so dryer steam at the exhaust with less condensate.

Then, in no particular order, we can look at bearings, main bearings, big end, and small end.  The cross head guides, and all the pivots and slides of the valve mechanism, or port face and pivot on an oscillator, all involve friction, which takes some of our valuable work and turns it into heat.  I am not suggesting that this friction can be eliminated, but attention to alignment, clearances and lubrication all leave more work available at the output shaft.  In the extreme, some modern compressors have magnetic bearings where the electronic controls vary the current to electromagnets so the magnetic forces to keep the shaft cantered in the bearings despite fluid load and unbalance of a high speed machine.  Incredibly low loss, but the electronics are mind bending, and I doubt they will be practical for model engines in the near future.  I don't know if anyone is experimenting with improved bearing design or lubrication for models.

That's enough for today, I still want to summarise the effect of the exhaust system, though the detail has probably been well enough covered.

Thanks for dropping in

MJM460

[steam guy willy]

 Hi MJM, this is the pic of the gridiron valve......It does seem to have a lot of metal to metal contact, so although it has a shorter stroke does this cancel out the friction losses? also in the pic the copper expansion joints ?? thanks for the input .........

[paul gough]

Just quickly, part of my reasoning for outside admission piston valves is to have very short exhaust passages straight out of the top of the block and into the blast pipe which is co-incident with its centre, an initial attempt in concert with reducing the piston clearances to about ten thou to reduce exhaust and compression volumes. The block cannot be enlarged nor can boiler out put be increased without an extra wick on the metho burner, whilst this has been done previously it is forcing the tiny boiler, so larger dia. cylinders with larger steam demand is out. There is no room to move, so to speak with the Lion, it is about as small as one can go in Gauge One and still have a loco that is not a pain to operate. Putting twin inside cylinders and an eccentric driven water pump off the main axle underneath was quite a task, so I am left with trying to come up with ways of improving performance within the existing dimensional constraints. The 6 1/2 lbs of force on the 3/8" piston working against the 5 lbs load on the valve, plus any other imposts does not give this little engine much in the way of reserve or very good slow running. Labrinth piston heads and piston valves are an exploration into possibilities and I had considered combining these with the use of engineered plastics, if that is the correct term, for these components if an appropriate grade can be got. I understand Aster used Rulon for their piston rings in their piston valve engines, I understand it is a teflon with powdered mica filler, but is a bit soft for a piston head. So not only is the physical design of the engine important but trying to find alternatives to traditional materials is part of the game in situations where things are approaching functional limits. Anybody who has any ideas or experience with teflon, nylon, or like derivatives for use as pistons with steam is very much encouraged to assist. I had not considered carbon and know nothing of its characteristics for use with steam, again any info. welcome. Thanks MJM for reminding me that reducing the travel of the valve is also a valuable thing to chase if achievable. Sorry about this not so quick post. Regards Paul Gough.

[MJM460]

Hi Willy, sorry that I missed the significance of the grid iron valve drawing in your earlier post.  If I understand the system, there are separate inlet and exhaust valves, so presumably two cams and linkages.  But I can see how it reduces the stroke of each valve.  Friction is interesting.  The conventional model of friction is independent of area, just depends on the total normal force.  More area means less pressure on the surface but the total friction is assumed to be the same.  With the grid plates on edge like the drawing, even the weight is hardly a big contributor.  It the steam force pressing the valve on the face is F, then the friction force to move the valve along the face is friction coefficient times F.  It does not depend on area.  As I have mentioned to Paul, the friction coefficient for dry steel on steel is generally given as 0.3, so the friction force is only 0.3 times F.  For a very smooth, well lubricated surface, the friction coefficient might be 0.1 or less.  Now friction is hard to measure, because it tends to have a stick-slip motion.  It takes a bit bigger force to start the movement, but then a smaller force is enough to keep it moving.  So the coefficient is at best an approximation, but it does allow design to proceed.  It is possible that the area has a small secondary influence, but it is usually ignored.  Sometimes the friction coefficient is quoted as both a static friction value, the value to get movement started, and dynamic value, the value once movement begins.

Hi Paul, now I can see what you are doing.  Thank you for your clear description.  Sounds like a great project.  I can only admire your watch making skills.  Will you be posting a build log or at least some pictures?  It looks like a project where, if you can use a millimetre better than others, you have made a major breakthrough.
Remember though, the valve rod does not have to work against the 5 lbs.  The rod force will be the friction coefficient times that force, and with a little oil or even condensate lubrication, together with a nicely smooth surface, the rod force can be expected to be nearer 0.5 lbs.  It may be worth rigging up a little test rig to get a rough measurement.  You can't simulate the temperature, or pressure on the sealing face of the valve, bit you will get an idea.  Regarding sealing rings, I can imagine that you have very little room.  Generally in the applications I am more familiar with, pure teflon is considered too soft, with too low a melting point.  It softens and spreads rather than take the load.  A filled teflon, is normally used for higher load bearing properties.  Graphite or other fill materials give different properties, but availability tends to override for hobby applications, so some experimenting with the available materials in a small stationary engine rather than risk damaging your amazing locomotive.  Carbon is quite fragile and so would require a two piece piston (or ring)  construction if that is practical in your design, but is used in pump and compressor seals to quite high temperatures, so may be worth a try.   At least the experiments could be carried out with much less effort in a separate single cylinder stationary engine, allowing more trials of less mainstream ideas.

The last area of loss that I wanted to list was, as Maryaks recent heat energy diagram showed (see post #242), the heat that goes straight through to the exhaust as latent heat of the exhaust stream.  We have to condense all of the exhaust totally to water before we can pump it back to boiler pressure, or simply reject it to the atmosphere, either way, the heat is not converted to work.  That diagram nominated a figure of 70% of the fuel energy.  Regardless of the accuracy of this exact figure, it shows that it is a major factor.  Can we do anything about it?  It is low temperature heat, so not easy to use elsewhere.  Perhaps as a boiler feed water heater if we have a continuous boiler feed pump.  Perhaps a little in preheating our fuel, or even the incoming air, or for avoiding the pressure loss in butane, but the majority is unavoidable loss.  The giant cooling towers attached to modern power stations attest to the limited alternative uses for this heat apart from some very specific circumstances.  I have for example worked in plants where the turbines exhaust to a low pressure steam system, around 100 kPa(g), sacrificing some work output, but using the exhaust heat for process heating purposes.  These processes tend to be relatively constant, only minimally dependant on weather conditions, but if the exhaust steam was not available, fuel would have to be burned anyway to supply the process needs.  When you do the calculations on how much extra fuel you have to burn to drive the turbine, you get the turbine power at an efficiency of around 80%.  Much higher than can be achieved by any isolated power plant, no matter how sophisticated.  But it is really only an accounting exercise, the process requirements reduce the fuel consumption attributed to the turbine power requirement, but do not actually increase the conversion of energy in the fuel to work. I am sure the process people attribute a very poor efficiency to the turbines and take credit for using the "waste" heat to reduce the fuel costs attributed to their process.

I think that covers most of the directions energy goes other than conversion to work, so I would like to look at the thermodynamic limits on how much of the fuel energy can possibly be converted to work in an ideal power plant, which leads to different measures of efficiency.

Thanks for looking in,

MJM460