Author Topic: Talking Thermodynamics  (Read 111761 times)

Offline MJM460

  • Full Member
  • ****
  • Posts: 825
  • Melbourne, Australia
Re: Talking Thermodynamics
« Reply #825 on: April 09, 2018, 12:54:03 PM »
Hi Admiral DK, Good to hear that you look in every day.  Like you, I read the latest posts on nearly all threads every day.  I find some (or more accurately, many!) that are clearly way beyond my skill levels, but most are still interesting, and often still have something I can even use.  But I suspect that what we all have in common is an appreciation and admiration of the skill demonstrated in so many builds and an ability to understand and admire the skill necessary to do some of the amazing builds we see here.

Sorry I get a bit esoteric at times, not surprising you skip a bit sometimes.  Don't hesitate to point it out, have a dig at me if you wish, it is the only way I can understand where people are up to.  Of course, sometimes I start out on a calculation because the data is available and the calculations straightforward, but without being sure where it will lead, so like a ship heading out into a fog, just hoping the way forward will clear before we reach the edge of visibility.  Sometimes it surprises me with just what results, sometimes it leads nowhere.  But my basic intent is to keep to areas that help us understand our engines and hopefully sometimes even to improve them. So please let me know if there are some areas you would like me to include, and particularly any you could help with the discussion.  I am sure that I am not the only one with anything to contribute.

Hi Willy, no need to apologise.  I hope you do not mind me having a go at your amazing library.  I am really quite envious.  But I don't think you have ever mentioned a book written even as recently as last century.  But there are too many books to collect them all and I admire your ability to focus on a special area of interest.

I am not sure how you would change the basic concept of a piston in a cylinder with a crank to provide continuous mechanical power output from the heat in a gas.  I suspect the main area of progress on this basic theme is around materials and manufacturing methods.  Followed closely by valve mechanisms and drive methods.  I don't know why we haven't seen much in the way of steam driven Wankel form engines, possibly sealing issues, or computer controlled, electrically operated valves, though possibly it comes down to complex mechanisms with many moving parts being replaced by the simplicity of turbines, basically one perfectly balanced moving part, and a much more favourable power to weight ratio, space to power ratio and capable of producing much higher power in each unit, though the boiler is a downside for transport application.  And of course reciprocating internal combustion engines and gas turbines are the more significant areas of advancement in mechanical power production, especially in transportation and very large scale. 

For smaller applications, and even quite large, electric motors are leading the field.  The power can be produced by one large, very efficient generator plant, with redundancy for reliability, and the power distributed by wiring.  Alternatively, and increasing in importance, power generated by solar or wind power, spread over a wide area to reduce distribution costs.  Of course renewables need not only generating capacity, but also storage for load characteristic matching, but politicians and the media do not seem able to grasp the basics of these things.

I have mentioned that I am getting back to the question of steam vs. air, and having another go at those calculation I started way back.  There is no point in the question if you are not wanting develop some real power, so I have assumed a pressure of 450 kPa and atmospheric exhaust so 100 kPa, about 50 psig.  Should have used 500 kPa so I could use the superheat tables directly, but the interpolation was not too complex.  Obviously steam has to be hot.  At atmospheric temperature and that pressure, the water would be liquid.  So 450 kPa saturated steam is 148 deg C.  I also looked at superheated steam, same pressure but 200 deg C.  I don't believe anyone would be using air at 148 degrees, more likely exiting the air receiver at around 30 deg C.

Now there are a few steps in making the comparison.  First, it is worth calculating how much work  could be produced by a ideal adiabatic engine.  This is the limit to what any engine could achieve, but any real engine will produce significantly less.  (Remember adiabatic means no heat transfer in or out.)  We can only demonstrate how much less the real engine produces by a test run.

Second, it is necessary to understand how the steam is actually used in the engine, and how this differs from that ideal engine.  It is important to remember that while the steam inlet valve is open, heat is being supplied by the incoming fluid.  Work output is equal to pressure times volume change, no difference between steam and air.  I don't believe there was any disagreement on that part.

When the inlet valve closes, the steam trapped in the cylinder starts to expand.  Depending on the cut off intended and how accurately this is achieved by the valve timing, the piston is probably around 30 - 60% of its stroke.  The expansion involves the resultant trapped steam including the clearance volume.  Realistically, in a single expansion engine, does not expand the fluid, whether it is steam or air, to much more than double its volume at cut off.  Then, near the end of the stroke, the exhaust valve opens and the remaining steam is exhausted to exhaust pressure.  Because valves actually open slowly under the action of the eccentrics and valve linkages, it is difficult to be sure exactly what the pressure in a model cylinder would be, but an idea can be obtained from the indicator diagrams that Maryak and others have posted.  But remember it is only expanding fluid that will show any difference between steam and air.

An adiabatic engine calculation assumes the whole volume of the steam is expanded through the full pressure range.

Reciprocating engines are volumetric machines, that is, each revolution of the engine admits a fixed volume of the working fluid, so instead of calculating work on the more usual mass basis, I have used the specific volume to convert the figures to a volumetric basis, so kJ/m3.

So what is the theory?  Saturated steam gave the output of an ideal engine as 618 kJ/m3.

For superheated steam, the figure is 570 kJ/m3.  That will require more explanation later.

For air, the figure is 537 kJ/m3.  I used the standard integral table for air, but the results seem identical with using the ideal gas laws, not very surprising for air.  For steam I used the steam tables which is the recognised best model for steam properties.

If you expand air in an adiabatic engine from 450 kPa to 100 kPa, starting from 30 deg C, the final temperature is -78 C.  A real engine will produce less power, and will have a higher outlet temperature, but if you do real work with your engine on air, you will get a very cold exhaust.  If you are just running unloaded with low pressure, you might have to use thermocouples to detect how much the air is cooling.  I normally only run on steam so I have no experience of running on air other than a simple and quick test on completion of engine assembly to prove I have a runner.  So I have to leave it to others to do some tests to demonstrate the performance of their engines on air.  Perhaps I will make up some fittings and try.

If I assume that the adiabatic efficiency of the engines is the same on air and steam, then we can see that saturated steam has the most "oomph", followed by superheated steam, and not really far behind, air.

Getting to be a long post, so I will pause there, and continue tomorrow.

Thanks for following along,

MJM460
The more I learn, the more I find that I still have to learn!

Offline steam guy willy

  • Full Member
  • ****
  • Posts: 2110
Re: Talking Thermodynamics
« Reply #826 on: April 09, 2018, 05:07:19 PM »
Hi MJM ,thanks for this ..i thought that may have been the case but for different reasons !.....saturated steam has lots of particles of water that act like the shot from a shotgun pushing against the piston ,unlike air that is quite 'soft' ! and also superheated steam that is also absent in :::mass"""!!!Willy ????
PS please correct me if this is an incorrect and silly surmization !!!

« Last Edit: April 09, 2018, 05:10:29 PM by steam guy willy »

Offline MJM460

  • Full Member
  • ****
  • Posts: 825
  • Melbourne, Australia
Re: Talking Thermodynamics
« Reply #827 on: April 10, 2018, 01:35:36 PM »
Hi Willy, partly right, never silly, but some extra explanations required.  And you already know I am not good at the short answer, so here goes.

Saturated steam is by definition dry, but it is also the boundary boundary between wet steam and superheated steam.  So as soon as you start extracting heat by doing work on the piston, it starts to condense.  But with the piston also moving the pressure is falling, which should limit condensing.  And condensation releases the latent heat, so contributing to maintaining the temperature and pressure.  But condensation reduces the specific volume of the steam so the pressure falls faster than if condensation did not occur.  A bit of a race between falling pressure keeping it dry and loosing heat causing condensation. 

The steam tables tell us the outcome, especially when combined with the second law of thermodynamics.  The second law says that for adiabatic expansion the entropy is constant.  This allows us to calculate the other steam properties using the steam table data, including just how much of the steam condenses.  For a real engine, less heat is converted to work, the entropy increases, and there is less condensation than for an adiabatic engine, but only a test run can tell us exactly how much work the real engine produces.  The quantity of water condensed is quite small, less than 10% for an ideal engine and less again for a real engine! and the velocities in the cylinder are around the same as the piston speed, both for the dry steam remaining and for the droplets.  But it is this condensation which means we cannot use the ideal gas laws for saturated steam, or anything only slightly superheated, we have to use the steam tables as the more accurate model of steam behaviour.  (This is the mistake I made way back in my first attempt at this question.)

When we expand air instead of steam, "soft" is not really a concept that translates to action on the piston.  The piston reacts to pressure.  Pressure over the area of the piston makes a force, and as the piston moves, that force moving through a distance does mechanical work.   Pressure is pressure whether exerted by steam or air.  The density of steam is less than air, but the specific heat of air is less than steam and when you work through all the equations, that ends up meaning you get less work from air.  The higher density of air just reduces the difference.  I did the calculations for air at 27 degrees C, mainly because that is 300 K, and appears directly in the air tables, but also that is realistic for air from the volume tank of a compressor.  Steam of course had to be at the saturation temperature or above so 148 deg C.  I will get back to try the calculations again with 148 for the air temperature, or at least the nearest temperature that appears in the tables to see if that makes much difference, or just makes the exhaust temperature more moderate.  It depends on how much the specific heat varies with temperature in that range, and of course it also varies the density.  No avoiding the equations and specific heat, it is not an intuitive result.

For superheated steam, you are spot on, it is the lower density that means there is less energy per unit volume available for conversion to work.  So that opens the question of why do we bother?  Part of the answer is that our reciprocating engines are volumetric machines.  That means each stroke takes in the same volume.  If we want to use lower density motive fluid, we need to use a larger engine to take in the same mass of the fluid for equal power output.   So if we design an engine for superheated steam instead of saturated steam we would design a larger engine.  So pluses and minus perhaps?  The clincher is that the extra energy to make superheated steam produces more extra work than the extra fuel energy from heat input to the boiler would suggest, so it is more efficient.  Less coal burned to travel a given train journey.  Less coal to cross the Atlantic in the same ship.  If we don't need the extra work output, then the smaller output from the smaller engine uses less coal than just the difference due to the difference in work output would suggest.  So, complex questions to answer, Willy your questions are never silly.

So far I have only discussed a single cylinder engine.  There is a limit to the degree of expansion that is practical.  I suppose with valve gear fully notched up, the steam is cut off very early and there could be a good degree of expansion, with just a short pulse of steam or air admitted each power stroke.  That would mean reduced power output but reduced steam consumption and increased efficiency.  If we have a compound engine, no extra steam is admitted to the lp cylinder, and the steam from the hp cylinder is expanded during the transfer to the lp cylinder.  With a triple we carry this process further with further increase in expansion of the fluid.  I seem to remember people reporting that those very low air temperatures do cause problems when running these engines on air, even though the extreme low temperature predicted by the adiabatic calculation is not reached in a real engine.

We could build a simple twin cylinder engine with equal power output to the triple expansion engine.  It would be much smaller, but would require more fuel.  We trade off engine size and efficiency.

One of the factors reducing the efficiency of a real engine is the heat loss from the cylinder walls.  It is also a departure from the no heat transfer definition of the adiabatic calculation.  You would expect that expanding air, the low temperature would mean there would be a heat gain, which might increase the power output.  The reality is that while the heat gain would be important, the area for heat transfer is very small and there will still be significant cooling due to expansion.  However that heat gain will reduce the cooling somewhat.  Just difficult to separate from other sources of inefficiency which reduce the degree of cooling during expansion.

And just to re-emphasise, the differences between steam, whether saturated or superheated, and air only occur due to the behaviour of the fluids on expansion.  While the admission valve is open, there is continual addition of energy through the incoming fluid, and the engine work output is the same for any fluid at the same inlet pressure.

And in case you are wondering, the reason we don't normally use air to drive a steam engine for producing work is that it takes more work in the compressor to produce the compressed air than the engine produces in expanding the air.  Generally it is easier to store electricity in a battery than air in a pressurised tank to produce power when steam is not practical for some reason.

I hope that all makes sense.

Thanks for following along,

MJM460

The more I learn, the more I find that I still have to learn!

Offline steam guy willy

  • Full Member
  • ****
  • Posts: 2110
Re: Talking Thermodynamics
« Reply #828 on: April 10, 2018, 08:25:08 PM »
Hi MJM, yes that makes sense and may i use your comment about not being silly for my epitaph  ? !!!!including your initials of course !!..I have a new question that i have been saving up though.....would it be possible to make a steam engine more efficient by using the fuel to just heat the cylinder and then inject water into the inlet sides of the valve  thereby dispensing with the boiler completely ? this would sort of be like a diesel cycle !!

Willy

Offline steam guy willy

  • Full Member
  • ****
  • Posts: 2110
Re: Talking Thermodynamics
« Reply #829 on: April 11, 2018, 12:02:15 AM »
Hi MJM, Also just found a diagram of the Willans engine ! I have not seen this before and also not seen a model of it ...apparantly there is a model in the Science museum, There is an extensive write up in the 1910 Model engineer mag and it looks very complicated and a bit like the Hargreaves type that i posted earlier....So this was an example of the Thermodynamic engineers having a go at getting more power and efficiency from steam

Offline derekwarner

  • Full Member
  • ****
  • Posts: 478
  • Wollongong ...... Australia
Re: Talking Thermodynamics
« Reply #830 on: April 11, 2018, 07:21:34 AM »
Hullo MJM...and all......I have been following on with each posting.....generally understanding, however sometimes confused  :facepalm:

We see you have chosen 148 degrees C for the point of saturated/dry steam, and from the tables, this equates to 3.5 Bar

This is a point of quandary in that in does not answer for those with steam plants operating at lower pressures

My Saito Y2DR 9cc horizontal twin engine has a recommended WP of 2 Bar, my boiler design is 3 Bar however I intend to trial & run the engine at   somewhat gently & progressively  :hammerbash: higher WP's than recommended

From the tables we see...........

2Bar = 134 degrees C [Saito recommendation]
3Bar = 143 degrees C [my boiler relief setting]
3.5 Bar = 148 degrees C [as used for your calculations]

So again whilst I understand much of the last postings......I cannot necessarily apply these questions/points/& answers to my plant

Derek

Derek Warner - Honorary Secretary [Retired]
Illawarra Live Steamers Co-op - Australia
www.ils.org.au

Offline MJM460

  • Full Member
  • ****
  • Posts: 825
  • Melbourne, Australia
Re: Talking Thermodynamics
« Reply #831 on: April 11, 2018, 12:22:21 PM »
Hi Willy, no problems with your epitaph, though I hope and expect it will be a long time before you need it.  Your questions are quite insightful and very clearly identify, in non-technical terms, the point that is puzzling you.  So much better than meaningless "techno-babble", even when they are a bit colourful.  As a result you have made a great contribution to this thread, and in answering your questions, I have often clarified things to myself that were really a bit fuzzy in my mind.  It has been and continues to be great fun.

I reckon I have seen that Willans engine before, though I can't remember where.  The concentric cylinders, with pistons not on the same piston rod, make for an interesting crank and cross head arrangement.  I assume the design was about improving efficiency, but I would need more information on the arrangement to understand how it worked.

Hi Derek, great to hear from you again.  I chose 450 kPa (absolute) as a pressure very close to 50 psig which seemed reasonable for an engine working hard, in the absence of other information.  The actual figure was not important as the purpose was to explore the difference between steam and air at the same pressure.

For your Saito engine at 2 bar gauge, say 300 kPa absolute, and atmospheric exhaust, say 100 kPa, the analysis goes like this.

First we look up the steam tables and pick out the enthalpy of dry saturated steam (2725.3 kJ/kg) and the entropy (6.9919 kJ/kg.K) at 300 kPa.  The corresponding temperature is 133.55 C.

Then we look up the properties of steam at the exhaust pressure.  I have assumed 100 kPa, but the tables also include a line for 101.3 if you prefer it.  We pick out the enthalpy if saturated liquid, hf, and of dry saturated steam hg and, because I am using a calculator, and basically lazy, the difference between the two, which is listed as hfg.  So hf = 417.46, hg = 2675.5 and hfg = 2258.0.

Similarly we look up the corresponding figures for entropy.  We find sf= 1.3026, sg= 7.3594 and sfg = 6.0568.  I suspect you are comfortable with the steam tables, but the detail might help some other readers follow the process.

We now calculate the exhaust conditions for an ideal adiabatic engine with those steam conditions.  The second law of thermodynamics says for an adiabatic process there is no change in entropy.  This means the entropy of the exhaust steam would be 6.9919, the same as the engine supply steam, so now we have two independent properties of the exhaust steam (pressure and entropy) and so can calculate all the others.  That entropy lies between the value for saturated liquid and dry vapour at the exhaust pressure of 100 kPa, so we know it will be wet steam.

We calculate the dryness using the entropy,

Dryness = (sexh - sf) / (sg - sf) = (6.9919 - 1.3026) / 6.0568 = 0.9393

Using the dryness, we calculate the exhaust enthalpy as

 hf + 0.9393 x hfg = 0.417.46 + 0.9393 x 2258 = 2538.4

And the change in enthalpy, or work produced per kg of steam is 2725.3 - 2538.4 = 186.9 kJ/kg

Now all that is for an ideal adiabatic engine.  A real engine will produce an enthalpy change of about 70% (the adiabatic efficiency).  I can't prove this figure from theory, it has to be determined from a test run with careful power output measurement.  Without other information it seems like a reasonable estimate, a model may have lower adiabatic efficiency but I have only rarely seen even a degree or two of superheat in the exhaust of my engines so dry exhaust is a reasonable limit condition.

So 70% adiabatic efficiency means our real engine produces an enthalpy difference of 0.7 x 186.9 = 130.8 kJ/kg.

You can see, the potential work output is only a tiny proportion of the heat used to just evaporate the steam, let alone heat up the water from cold.  So the steam cycle is not high efficiency.  It is limited by the low pressure, and the fact that most of the heat used to evaporate the steam goes out in the exhaust and gets lost to the atmosphere or rejected in the condenser of a full scale plant.  And you can see why engine manufacturers might prefer to talk about adiabatic efficiency.

We can subtract that enthalpy change from the supply steam enthalpy, and find our real engine exhaust enthalpy is more likely to be 2725.3 - 130.8 = 2594.5 which we can see still lies between the enthalpy values for saturated liquid and dry vapour so the exhaust is still wet, and further illustrates that most of the heat from the boiler is rejected in the exhaust. 

We no longer know the entropy, which will have increased, but we can use the enthalpy to calculate the real engine exhaust dryness.

So dryness = (2594.5 - 417.46) / 2258 = 0.964, somewhat dryer than the adiabatic engine exhaust.

This means that just 3.6% of your steam supply appears as condensate in your exhaust.  I doubt if that is enough to explain your condensate issues.

Realistically, there is not enough heat transfer area around the cylinders to produce much more condensate by heat loss once the cylinders have heated up.  And likely even some of the 3.6% will go up your exhaust stack as that vapour mist.

That leaves condensate due to the initial heating of the cylinders, which should soon cease as the cylinders warm up, or carryover due to priming when the boiler is initially at maximum level.

It might be worth trying a run with less than half the normal fill of water, (making sure that it is enough to cover all the heating surfaces) to eliminate priming, or even try firing the boiler with the steam pipe discharging to atmosphere instead of to the engine, so explore the possibility of priming.  Obviously keep an eye on water level and don't run the boiler dry.

Then, if you run the plant out on the bench, perhaps without the separator, just an exhaust line run so you don't get scolded, or make too much mess, and see if the condensate stops after a short time or not.  If it is condensate due to warm up, you will need to instal cylinder drain valves, perhaps like the ones described recently in the Conway thread, or a simple screw down type, or even a plug type if you like a challenge.

I hope that adequately answers your question.  I have raced ahead a bit, based on some of your previous questions, so I hope I have made a reasonable guess at what is behind your question.  Apologies if that issue has been solved long ago.  Would you like me to continue and show the difference if you had a superheater which reached say 150 deg C, which is probably around the limit of what you could achieve in your boiler without going overboard, but importantly, directly in my superheat tables so easy to use.

The procedure for calculating the work output when air is used as the motive fluid is a little different, as the ideal air tables have a very different form from the steam tables, so I will not confuse things by returning straight into the air calculation.  However I will return tomorrow with a calculation based on air at the same temperature as well as the same pressure as the steam in yesterday's example, to see if that explains the difference.  But it is a bit of a theoretical calculation, as we would not normally have access to high temperature air to run our engines.

Thanks for following along,

MJM460

The more I learn, the more I find that I still have to learn!

Offline derekwarner

  • Full Member
  • ****
  • Posts: 478
  • Wollongong ...... Australia
Re: Talking Thermodynamics
« Reply #832 on: April 12, 2018, 06:35:40 AM »
Goodness MJM.....thankyou and that is quite a number of paragraphs and sentences for me to read, re-read and adsorb 

One of the important points noted in non-technical terms is that 'relatively lower' pressures are clearly far less efficient than 'marginally higher' pressures in producing work....[and totally discounting using air as a pressure medium fluid]

Digressing a little, I suspect in 1956, Mr Saito [Senior :old:] in Japan was very conservative in his design calculations data for his steam engines

Derek
Derek Warner - Honorary Secretary [Retired]
Illawarra Live Steamers Co-op - Australia
www.ils.org.au

Offline MJM460

  • Full Member
  • ****
  • Posts: 825
  • Melbourne, Australia
Re: Talking Thermodynamics
« Reply #833 on: April 12, 2018, 01:00:03 PM »
Hi Derek,  you are most welcome.  I really only needed your operating pressure to get started.  To give you a chance to take that all in, I will avoid adding too much more tonight.

You are quite right about low pressures being less effective than just a little higher pressures.  With the same calculation procedure as yesterday, we can actually put figures on that.  Now I like round figures to work with, and as a survivor from slide rule days you probably feel much the same.  So I will do the calculation for 400 kPa, (or 300 kPag), just 100 kPa above your rated 200 kPag operating pressure.  As this is equal to your boiler safety valve set pressure, you would not want to operate at this pressure, but would need to keep enough of a margin to ensure the valve was not weeping.  However, the figures do clearly illustrate the point.  I will just summarise rather than go through the complete detail again, it is exactly the same procedure.

We need just two extra figures, hg for 400 kPa (2738.6 kJ/kg), and sg (6.8959 kJ/kg.K).  The saturation temperature will be 143.6 C.  It is worth going back to yesterday's post to compare those values with the figures for 300 kPa.

Now, imagine heating the boiler from cold up to 133.55 C.  The enthalpy of the water is the saturated fluid enthalpy of 561.5 kJ/kg.  Then, instead of evaporating that water to make dry steam at 300 kPa, keep the regulator closed and increase the pressure to 400 kPa before opening the regulator.  The enthalpy of the dry saturated 400 kPa steam is 2738.6 kJ/kg, just 13.3 kJ/kg more than it took to make dry saturated steam at the lower pressure from that same starting point.  But at that condition, the entropy is  actually lower than it was at at 300.

Now, we follow the same procedure to calculate the work produced by an adiabatic engine,  the second law says the entropy for an adiabatic engine is same at inlet and outlet.  From that we calculate the dryness as 0.9235, a little lower than yesterday.  We use that to calculate the enthalpy of the exhaust and subtract this from the enthalpy of the dry saturated inlet steam to get a work output for that adiabatic engine of 235.9 kJ/kg.  Now it is worth looking back to the figure for yesterday.  For that input of just 13.3 kJ/kg, we get an extra 49.0 kJ/kg as work output.  Now that is efficiency!  Even when we look at our real engine, again assuming the same 70% adiabatic efficiency, we get an output of 165.1, compared with 130.8 at 300 kPa.  The exhaust dryness is 0.955, a little more wet than yesterday, that is where the energy for the extra work came from, but still mostly a misty vapour.

As always when I do these calculations, I do check them carefully, however I am always a bit nervous sticking my neck out and posting the results.  But we will get nowhere without some courage.  I have given you all the information necessary to check the result, an exercise which will really help you understand the process.  Even better if you use your own steam tables.  Please let me know the result.

Unfortunately we can't get that sort of result as a total efficiency, but obviously the extra work output at the slightly higher pressure improves the overall result, as you have mentioned.

A short but perhaps heavy post tonight, I will let that sink in before I have a look at the effect of superheating tomorrow.

Oh, by the way, regardless of the conservative design, I don't recommend increasing the safety valve set pressure.  It will be a real pain getting a new design pressure through the boiler inspectors at this stage.  I have had to do this on full size vessels, and the first time it convinced me that for all future designs I would base the design pressure on the actual plate thickness, rather than the minimum required design pressure, so that further raising the design pressure later was not a possibility.

It is also worth remembering that the engine in a ship can only put out the power the propellor can  absorb.  If you increase the steam pressure the propellor will of course turn faster, and produce more thrust, and the power (if the propellor does not cavitate) will mostly go into an unrealistic bow wave.  So it is not necessarily practical to take advantage of the extra output at the higher pressure.

Thanks for following along,

MJM460
The more I learn, the more I find that I still have to learn!

Offline MJM460

  • Full Member
  • ****
  • Posts: 825
  • Melbourne, Australia
Re: Talking Thermodynamics
« Reply #834 on: April 13, 2018, 01:02:40 PM »
Those who have followed yesterday's post may be asking themselves, " If it is so much more  efficient to use steam at 400 kPa than 300 kPa, then, even when we can't use the extra power, should we still raise steam at 400 then throttle back to 300 kPa?"

Once again, we can work through the numbers and see what the thermodynamics says.  The procedure is much the same as the last two problems, the key is simply how the first law of thermodynamics applies this time.  Throttling occurs in a very small space, so there is effectively no heat transfer in or out, and in the throttle, (think a simple orifice plate, or perhaps a throttle valve partly open,) no work is done.  For this case, the first law reduces to no change of enthalpy.  This gives us the necessary two independent properties for the throttle outlet (pressure and enthalpy), from which all the other properties can be calculated.  We can then use these properties as the inlet steam to our adiabatic engine operating at 300 kPa.

The calculations are very similar to the preceding ones, so I won't bore you all by setting them all out again unless someone would like to see it.  The result of throttling dry saturated 400 kPa  steam to 300 kPa is that steam enters the engine slightly superheated, to 140 deg C, so we have to use the superheat table, to find that the adiabatic engine produces just 1.3 kJ/kg more work from that extra 13.3 kJ/kg required to produce the higher pressure steam.  The real engine, still assuming 70% adiabatic efficiency, produces 131.7 kJ/kg or 0.9 kJ/kg more from that higher temperature steam.  The real engine exhaust is still wet steam, but just slightly drier than from the dry saturated 300 kPa steam.
We can see that from the efficiency point of view, there is minimal advantage in generating steam at a higher pressure, then throttling it.  The total heat to evaporate the steam is about 2200 kJ/kg to produce 131 kJ/kg of work, so 0.8 from 13 is slightly better than average, but not nearly as good as just using the higher pressure.  In fact, if the steam is generated at higher temperature, the flue gases will also be discharged at a higher temperature, so increasing the losses, both as heat in the flue gas and as heat loss from the boiler.  In reality, these theoretical differences are almost insignificant compared with the heat input to the steam or even the work produced.

On the other hand, when steam is raised at a higher pressure, the specific volume of the steam is less, (0.46 at 400 kPa, compared with 0.61 cubic meters per kg at 300 kPa) which makes it easier to separate from the liquid, so there may be less carryover and perhaps less tendency for priming, especially when the boiler is at maximum level.  This is a practical issue which may well outweigh any theoretical efficiency considerations.  But, like the efficiency of the engine, it is not really quantifiable by calculations.  However, if you are experiencing difficulty with boiler carryover, increasing the boiler operating pressure and then throttling to the operating pressure you need, may be helpful.

I hope that little diversion has given an idea of the things that can be better understood with a few calculations.

You might have noticed that I am wavering a bit on the pressure terminology.  The preferred SI terminology is to use kPa for absolute pressure, and only qualify it when gauge pressure is meant.  However this is a world wide forum with many people much more familiar with imperial units and psig for pressure, but often just psi, and generally only qualifying it if absolute is intended.  I have tried to stay with kPa as absolute pressure, but to make the mental conversion easy, I have generally tried to add in brackets the gauge pressure in kPa and psi, with the appropriate qualification.  Generally if there is any doubt, absolute pressure is used in all my calculations and steam table references.

Similarly, there is no practical difference between the standard atmospheric pressure of 101.3 kPa and the easier to remember 100 kPa.  As both are listed in the tables, they are equally easy to use until you need to consider superheat, when generally only the 100 kPa table is provided.  As 100 is within the range of atmospheric pressure which occurs due to the normal progression of weather patterns around the planet, occurs directly in the superheated steam tables and is also a nice round figure, it is the value I prefer to use as atmospheric pressure.

Thanks for following along,

MJM460
The more I learn, the more I find that I still have to learn!

Offline MJM460

  • Full Member
  • ****
  • Posts: 825
  • Melbourne, Australia
Re: Talking Thermodynamics
« Reply #835 on: April 14, 2018, 01:42:07 PM »
Now for the last step in the calculations for Derek's engine, the effect of superheating the 300 kPa steam to 150 deg C.  Remember 300 kPa in absolute pressure is 200 kPa on the gauge, about 30 psig, the recommended pressure for his engine.

Now that is not much superheat, but it is realistic for a small boiler, based on the temperatures I have seen from my minimal superheater coil on my Meths fired boilers.  It is not a lot more than obtained from throttling 400 kPa steam per yesterday's calculation.  And, importantly, it appears directly in the superheated steam tables, so requires no interpolation to get the required data.

The calculation is the same as previously so I will just give the results and talk about what they mean.

The enthalpy of the superheated steam is 2761 kJ/kg, which is 35.7 more than saturated steam at that pressure.  The work from an ideal adiabatic engine is calculated as before, based on the second law for an adiabatic engine, so no entropy change.  The exhaust steam would be wet steam despite the superheat at the engine inlet, and would have a dryness fraction of 0.953.  The adiabatic work out would be 190.5 kJ/kg, only 3.6 more than for output from saturated steam.  Certainly not much reward for the effort to add a superheater to the basic boiler.  It is even slightly worse efficiency than the efficiency based on the heat necessary to raise saturated steam, and only 2.3 kJ/kg more than was obtained by throttling 400 kPa saturated steam without the superheater.  So no incentive to add a superheater, at least to this temperature level, if you can achieve a higher temperature the result may be different.

You will remember that when I did the calculations earlier for a somewhat higher pressure, it all worked out as a positive benefit, even if not large.  Really I expected the same again, so I was quite surprised at this result.  Despite careful checking, I have not found any error, but if you decide to try it for yourself, please let me know if you find any errors.  There is nothing like actually trying the calculations to help understanding.  So how does it come about?

If you look back at the pictorial representation of the steam tables (I have attached them again so you don't have to search), and follow the vertical lines of constant entropy on the T-S diagram and constant enthalpy on the P-h diagram.  In the wet region, temperature and pressure are not independent, so you can read higher temperature as higher pressure, and vice versa.

You will see in the T-S diagram, that expanding in an adiabatic engine at constant entropy, the vertical line, always gives an increase in entropy and wet steam at exhaust.  However, when you look at the P- h diagram, the enthalpy of the wet steam increases at higher pressure to a maximum at about 3000 kPa, then starts decreasing with further pressure above that. 

Below 3000 kPa, higher pressure steam always has higher enthalpy, so if you reduce the pressure by throttling, a constant enthalpy process, it becomes slightly superheated.  And of course, all our modelling applications are below this level.  Industrial steam plant of any size, generally works well above this (it's about 420 psig of you are thinking in imperial units). Unless the steam is used mainly for heating and the higher temperatures are not wanted.

Also, remember back to the earlier efficiency discussion, and the Carnot efficiency limit, determined by the ratio of the maximum and minimum absolute temperatures in the cycle.  When we are operating at only 2 bar, exhausting to 100 kPa (133.5 degrees to 99.6) gives us a very low efficiency limit, and efficiency increases more rapidly with increase in temperature in this area.

Unfortunately the graphs I have posted use MPa for pressure, 1 MPa is 1000 kPa.  And the logarithmic scale makes the top of the wet steam area appear closer than it really is.  Similarly the T-s diagram uses absolute temperature, so you need to add 273 to your temperatures in C to find your temperature ion the diagram.

Clearly it is worth doing the calculations for your actual operating conditions before assuming too much from conclusions based on significantly different conditions.  And with 2 bar gauge operating pressure, not unreasonable for a small model, it is clearly not worth adding a superheater.  But by  400 kPa, the benefits are starting to appear, even if still very small in practical terms.

If you want to chase the higher efficiency associated with higher pressure, a good direction to consider would be to make a smaller engine for the same load, so it needed the extra pressure to provide adequate power.  You would calculate a piston diameter which would experience the same force at 400 kPa as the larger one at 300 kPa.

Well, Derek, I hope that has adequately answered your question, and I hope it has been of interest to everyone else.

Thanks for looking in,

MJM460
The more I learn, the more I find that I still have to learn!

Offline MJM460

  • Full Member
  • ****
  • Posts: 825
  • Melbourne, Australia
Re: Talking Thermodynamics
« Reply #836 on: April 15, 2018, 01:31:31 PM »
Thinking about the discussion on air powering our models, I had a look in my box of excess tube fittings, spare plugs and so on, and sure enough found an adaptor that allowed me connect the air compressor hose to the engine.  Basically it fitted the boiler instead of my normal thermowell fitting.

I set the compressor pressure control to 30 psig (it is a machine distributed by a major US compressor manufacturer, though almost certainly made in the Far East), and connected the tyre connector to the fitting on the boiler.  When I started the air flow, the engine ran quite nicely, starting at about 900 rpm and soon accelerating to 1220 rpm. 

My engine is fitted with temperature thermowells on the inlet and exhaust, so I monitored the temperatures.  Ambient temperature was 15.6 C, unfortunately the days of 20+ seem to have departed, at least for the moment, if not for the season.

The temperature monitor on the engine inlet was showing 15.9 C.  Side by side the readings of the ambient temperature and the engine inlet thermocouples were the same, but even putting it into the thermowell well before the run, seems to introduce a small variation.  Something to think about on another occasion.  During the engine running, this slowly reduced to 15.4.  Again this is not easy to explain so also back to that later.  Ambient temperature reading did not change.

The exhaust temperature reading started at 18.  That meter does not have 0.1 degree resolution and always seems to read a little higher than the others.  During the runs it reduced to 13 degrees.  Not quickly, probably due to the heat absorbed in cooling the cylinder which is a large lump of bronze.  But eventually a 5 degree reduction from the starting point.  I was a bit worried about cylinder lubrication, as the displacement lubricator does not work with air, so I did not run very long.  The temperature may have dropped a bit lower had I run for longer, but it was moving quite slowly, so perhaps near the minimum.

The pressure on the tyre gauge was about 150 kPag, a bit higher than the steam pressure I usually see based on temperature measurements, but I think we would all agree that apart from the probably inaccurate gauge, the pressure on the gauge before the connector to the tire is not an accurate measurement, pressure at the engine would be much lower.  As the engine speed was about the same to a little lower than when running on steam, I suggest the pressure was around the normal 50 to 75 kPag.

So where was the -78 I talked about earlier?

First, -78 was the adiabatic engine exhaust temperature.  For a real engine with 70% adiabatic efficiency, -55 would be a better estimate, but still way lower than the observed temperature.

Next the pressure ratio.  I assumed 450 kPa, or 350 kPa gauge, exhausting to 100 kPa for the earlier exercise on the basis that there is no need to compare air and steam unless the engine is required to do some real work.

So my little mill engine, running unloaded had, let's say 70 kPa gauge max at the inlet, so 170 kPa  exhausting to 100 kPa.  With a smaller pressure ratio, the exhaust temperature will be much more moderate.  Remember also that the adiabatic engine takes all the steam at inlet pressure and expands it all adiabaticaly to the exhaust pressure.

Rather than calculate a new pressure drop, I took the valve chest cover off and had a close look at the valve timing.  It was my first attempt at a slide valve, and it soon became apparent that I had not provided enough lap.  The outside end of the cylinder opened quite well on top dead centre, and closed just before bottom dead centre, I estimate, with a protractor beside the crank, perhaps 175 degrees.  Now with the sine form of the piston movement, the remaining volume change is minimal.  Almost no expansion at all.  On the inner end, or crankshaft end, it was slightly better, the cut off was about 170 degrees.  Slightly better, but not much better.  So my little mill engine is basically taking in full pressure fluid, whether air or steam, throughout most of the stroke.  This is not very close to the ideal adiabatic expansion.  When the exhaust valve opens, the remaining pressure is basically throttled to exhaust pressure, with no further work production, and perhaps a small temperature reduction.  Surprising to see 5 degree temperature drop, from such a tiny expansion ratio.

I carefully measured the valve and the steam chest, and I think I could make a new valve with a bit more lap.  Then I could reset the eccentric to give the necessary lead, and I should get some more expansion.  Now Chris would have the new valve made and installed by now, but I live in that alternative universe where things move more slowly.  It's on the list, but for now, lets assume I achieved about 40% cutoff, then allowing for the clearance volume in the cylinder and steam passages, say one to two expansion, still less than for that fully loaded up engine. 

I suppose it is possible to set the cut off very early to get a much greater expansion, but the steam inlet flow, and hence the power developed would be much reduced with such a setting.  However, a compound engine could be expected to achieve greater expansion, and hence lower exhaust temperature, with enough steam flow to provide reasonable output.  Alternatively, with Stevenson's valve gear, or Willy's Allen gear, I assume notched back, would give earlier cutoff and more expansion, and would see lower exhaust temperatures.

Clearly life is more complex than being a matter of doing some theoretical calculations and demonstrating the results on a model engine.

Thanks for following along,

MJM460
« Last Edit: April 15, 2018, 01:38:57 PM by MJM460 »
The more I learn, the more I find that I still have to learn!

Offline steam guy willy

  • Full Member
  • ****
  • Posts: 2110
Re: Talking Thermodynamics
« Reply #837 on: April 15, 2018, 04:38:22 PM »
Hi MJM , interesting info on using compressed air for your engine... is the temperature drop partly because of the draft occurring  and with steam is there a similar cooling effect with this "draft" as in standing next to a slightly open door. ? And rather than making a new valve could you just make the buckle  a rather loose fit in the valve slot to get a similar valve action  as with the lap ? ie if there is a gap it will make the valve move a bit later on the stroke ??!!! also a 1902 engineers pocket book with various tables for different steam engine builders....

Willy
« Last Edit: April 16, 2018, 01:30:19 AM by steam guy willy »

Offline MJM460

  • Full Member
  • ****
  • Posts: 825
  • Melbourne, Australia
Re: Talking Thermodynamics
« Reply #838 on: April 16, 2018, 02:11:27 PM »
Hi Willy, I almost didn't do the run on air, but when I found the necessary fittings sitting with my spares, I though I might as well.  As always, actually doing the experiment, especially with an idea of what you are looking for, usually seems to reveal unexpected detail.  So seeing the relatively small temperature drop, I had a new look at the valve, and looked again at K N Harris' book with new understanding of lap and lead.  Now I need to look at Martin Evans on valve gear as I am sure he mentions how the lead changes, or doesn't, with various valve gears.

I didn't think to take photos yesterday, and it was too late when I was typing the post, so they are attached today.  Yesterday I was running the mill engine.  My own design, first attempt at a slide valve, driven by a simple eccentric.  I was a bit vague about lap and lead at the time, I have thought a lot more about them since.  I was just delighted to achieve a working engine.

Today I had a closer look at the fittings I had, and found I could also connect my air supply to the diagonal engine with a lever operated variation of Joy valve gear, again my own design, with a very similar temperature result, in both forward and reverse.  And even as the gear notched back, a bit of a surprise.  But back to that later.

By draft, I assume you mean around the boiler.  It almost certainly is important, but the boiler is on the inlet side of the engine, so has no effect on the temperature difference across the engine.  The temperature difference across the engine is due to converting energy in the gas into work.  The enthalpy of air is directly proportional to its temperature.  If you extract energy, whether by heat transfer or by converting it to work, the temperature will fall.  The amount of fall when work is done, is dependant on the expansion ratio.  Once the inlet valve closes, further movement of the piston results in expansion and a significant temperature drop.  The late closure of my valve means there is only a very small expansion in this phase of the cycle. 

But what about before the valve closes?  Before the valve closes, work is done as the pressure force acts on the moving piston, but as more air continues to be admitted so the pressure does not appear to drop.  And as air is consumed, more energy is put in by the compressor.  So the heat converted to work is continually replaced by the compressor motor, or when steam is the motive fluid, it's replaced by the heat from combustion of fuel.

When we look at the compressor and the compression process, also an adiabatic process, but the pressure rise means the temperature rises, as you know if you ever put your hand near the compressor head while it is working.   A large industrial machine will have a heat exchanger to cool the air.  In our small shop compressors apart from fins on the compressor, the main cooling surface is the air tank if you have one.  My small compressor as you can see has no tank.  Pressure control is achieved by an adjustable relief valve which is adjusted to vent air at the required pressure.  I then have about 5 metres of air hose.  And of course my little boiler and its superheater  will provide some cooling through the draft that you mention.  But none of that explains what I observed during my test runs.

Ambient temperature in the shop was very steady at 17.9 C.  When I put the thermocouple I was using for the engine inlet, next to the room air device it agreed within 0.2 C.  When I then put it in the thermowell at the engine inlet, it dropped to 15.9 C.  This was before I started the compressor, so it had all had plenty of time to settle at ambient temperature.  Much more difference than the 0.2 degrees when the thermocouples were side by side.  Not sure I totally understand that.  May be something about differences in absorptivity and emissivity to radiant heat that slightly alters the heat balance.  On conduction alone there should be no difference from the air temperature.

When I turned on the compressor, the mystery deepened.  As the engine ran, temperature at the engine inlet dropped.  It seemed to settle out around 14.4, but could have been still dropping slowly.  The head of the compressor was hot, as was the hose connection at the compressor end.  There was definitely heat lost along the hose as the connector on the gauge end was only slightly warm.  The boiler shell and superheater should have been helping the air reach atmospheric temperature, especially with some draft through the furnace, but 14.4 was distinctly cool, when I would have expected it to be still warm from the compression process.  Could have been some heat loss along the hose, and usually some pressure drop through those cheap tyre inflation gauges.  But I have not been able to come up with a convincing explanation for actual cooling.  Even put the thermocouple back beside the room temperature thermocouple and the reading slowly increased back to very close the the ambient temperature.  Very mysterious.  Any suggestions, anyone?

The exhaust temperature decreased, only 1 degree below the engine inlet temperature but definitely cooling in the engine.  Again I removed the valve chest cover, and watched carefully as I turned over the engine.  Again, my radial gear seemed to be cutting off very late in the stroke.  Seemed close enough to 180 degrees admission.  Obviously, I can't see the exhaust cavity of the valve so have no information on that apart from measurements of the valve.  Clearly not enough expansion for any real temperature drop.  As air enters the cylinder, it is replaced by more from the compressor via hose.  I thought there might have been earlier cutoff with the valve travel,notched down, but I was not able to demonstrate this.  I suspect a new valve with more lap may be required but that may require remaking one of the vibrating levers as well to adjust the lead.  I have been thinking about your suggestion of slack in the buckle.  It is getting too late now so I will reply on that  tomorrow.

The Joy valve gear is a radial gear and not easy for me to see how to adjust the valve lead, it is built in to some of the link dimensions, so these links would have to be remade after the lap was carefully measured.  Again, I was just delighted to have a running, reversible engine. 

And one more thing, in case you are wondering why we don't see many full scale application for air motors, that second law of thermodynamics prevents any motor from producing enough power to run the required compressor, you get much less power from the engine the you have to supply to drive the compressor.  Though air has a place in pneumatic tools where the temperature of steam would not be helpful, and in potentially flammable atmospheres where the possibility of sparks must be avoided.

So some interesting results from short runs on air.  I was a bit worried about cylinder lubrication, so did not run for long.  I am not sure what others do, but I suspect I would need to try some sort of pressure lubrication if I am to do much running on air, along with oil suitable for the low temperature.

Photos not posting, may be too big, I will try again then if no success will resize them tomorrow.

I hope you found that interesting, thanks for following along.

MJM460
The more I learn, the more I find that I still have to learn!

Offline MJM460

  • Full Member
  • ****
  • Posts: 825
  • Melbourne, Australia
Re: Talking Thermodynamics
« Reply #839 on: April 16, 2018, 02:15:23 PM »
I tried the photos, one at a time and no success.  Will resize tomorrow.

MJM460
The more I learn, the more I find that I still have to learn!