Talking Thermodynamics

[Gas_mantle]

Hi MJM just a few technical questions........... on my electric boiler the output pipe is 5/32" ,and the inlet pipe to my new engine will be larger , about 1/4"  if i want to connect the two together should i have the connection as close to the boiler as possible or to the engine ?.... when the two pipes are joined is there a significant pressure drop  into the larger pipe ?
Willy.

Hi Willy, I experimented with different sized pipes on my boiler and found that it made little difference so I think the position of the joint with have no noticeable effect. I could be wrong but the way I view it is that when the steam leaves the pipe and enters the steam chest it is effectively entering a pipe with a huge increase in size (albeit a rectangular pipe / chamber).

I'd be interested to hear what the experts say though as I've pondered similar questions in the past.

[paul gough]

Hi MJM, Could you give some guidance on the addition of a reflective layer to our insulation sheeting. Is it always superior when added than just using the 1, 2, or 3mm ceramic sheeting; does it matter whether it is inward or outward facing; and what materials might be effective that we could easily access, eg. aluminium cooking foil??? Regards, Paul Gough

[paul gough]

It just stuck me that at or near the 1st anniversary of this thread you will have something near 50,000 reads!!! This shows there is quite 'demand' for knowledge/information on thermodynamics and the associated ramblings by modellers. I wonder if you would be interested in distilling all that has presented thus far and publishing it?? I think it would be pretty attractive to both the publisher and reader of A.M.E. magazine or some such. Paul Gough.

[MJM460]

Hi Willy, good to see you are still following along.  Tent flys don't help much with flies as they are not enclosed, but generally open at the ends and lower sides to allow plenty of air flow.

Thanks for posting that formula for the model boat rating formula, unfortunately I can't clearly read the second page which has some of the important details.  Despite its length, it is actually quite a simple formula.  It adds length and sail, area which are known to make a boat faster, forcing the designer for each class to find that balance between longer for less wave making resistance, and sail area to provide a higher driving force.  You cannot add length and area, but the square root of area has dimensions of length so can be added.  The 1/3 factor is an attempt to equalise the relative importance of sail area and length from the potential speed point of view, probably arrived at by careful observation of the performance of many existing designs.  The other factors are simple "taxes" on factors that the rule makers want to discourage, to avoid designs with undue advantage.  Basically those girth measurements are smallest for a hull with vertical ends so the waterline length is the same as length overall.  When the yacht has long overhangs and that graceful form of older designs, if the beam at the deck line is is significant, there is the possibility of hull shapes that have a longer waterline length when heeled over under the pressure of wind on the sail.  The rule designers wanted to discourage this factor, so the resulting designs would be narrow at the ends of the waterline so minimal extra length when heeled.  I think the second page has some words about why extra freeboard was also discouraged.  Beam is indirectly implied by that G measurement, as a wider hull can carry more sail area for its length. 

Of course the formula is still not dimensionless and the measurement system has to be part of the definition.  Obviously a boat rated in meters is a much bigger boat than one rated in feet.  Depth was probably not penalised enough if I look at the latest designs in model yachts, which by the way I have read do tend to lead the way for full size practice.  The trend is to very deep keels, as the better righting moment from a deep keel allows more sail area, and the torpedo bulb at the bottom of a deep fin keel does not have as much surface area drag as the traditional long keels the rule designers were no doubt imagining.

Regarding your boiler, I would make the change to the larger pipe diameter so as to have the minimum length of the smaller diameter pipe, as the resistance to flow varies with the velocity squared, so larger diameter same mass flow and steam conditions means lower velocity, less friction loss, but not a huge difference, particularly as you then have a throttle at the inlet.

At the join there will be one velocity head of pressure lost.  When the units are correct it will not be a very large number.  The velocity can in principal be converted back to higher pressure in accordance with the well known Bernoulli formula to reduce the loss, however, the diverging walls for the size change must be less than about 15 degrees, and nearer 12 would be better.  Remember that delivery cone on the injector we looked at a while back.  You could make a proper diverging adaptor, but the pressure gain in this case would not be really worth the effort.  Velocities are so much less than in an injector, and you could probably gain the pressure elsewhere a lot easier.

It is easy to work out the mass of steam from the electric element rating and the steam conditions using the steam tables.  The steam tables also tell us the volume of that steam.

Similarly, we can easily work out the swept volume of the cylinders, even allowing for the piston rod diameter if you wish.  So for a given rpm, the maximum amount of steam consumed by the engine can be calculated.  We don't really know how much the valve timing will reduce the total steam taken by the cylinders.  Nor do we know how much the throttle will lower the pressure at the cylinder when it is partly closed by the governor.  The lower pressure increases the volume of the steam, so reduces the mass flow consumed by the engine.  And of course it is pretty hard to estimate steam losses at valve and piston rod demands and piston ring blow by, or even exhaust valve leakage.  However, that basic calculation of steam mass and volume give a good idea of whether the boiler will do.  And if we have underestimated the volume the engine uses it will go a bit slower, while it will go faster of we have more volume available with sufficient pressure.

It is quite late here now, so I will look at the calculation tomorrow.  By the way, I am following your build and admiring your skill in both the design and bench work with the files.  I did go to the wood working tool shop yesterday.  They had a whole cabinet full of brand new planes of various designs, all those little ones that you would have put to good use in your previous life.  Retail therapy at its best.  Now I now what women are talking about, they just go to the wrong shops.

Hi Zephyrin, thanks for coming in again.  Please be assured that I am not criticising your design in the least, in fact I am admiring both the design and your skill in soldering it together.  I am quite confident that you will have tested it and that it is quite safe.  And of course, no code will protect against spilled fuel, derailments etc.

We have our own code here.  I am not inclined to be too critical of it, as I am aware that it is a balance of the politics of self regulation vs government authorities stepping in, and I am not aware of all the reasons some code requirements were introduced.  However I have heard that some well proven designs do not comply and have been rejected after many years of successful and safe running.  A section for subminiature boilers has been introduced recently, so for interest I looked at how your design would compare.  I don't know what the inspectors would say, but it looks to me like they might call for a modification to the firebox width to meet the requirements for ligaments around the fire tubes.  However, the sub miniature code also relies heavily on hydro testing and steam testing, so it is also possible that it could be totally acceptable based on testing.  Like some other forum members, I am not in a club, and don't really know much about how the rules are actually applied.  My experience is more with full size pressure vessels.  No way of understanding the impact of applying a different code without checking a design against the code, so the exercise was interesting, but I probably should have kept my mouth shut.

I am quite interested to read that gauge 1 book and the designs it includes, however it is down my list for the moment.  But I will have a closer look at the links you have provided, they are quite interesting, thank you.

I guess gas is better contained, hence less likely to be spilled, providing the fuel is transferred to the tank with plenty of ventilation.   My background makes me well aware of the consequences of igniting a gas leak, so I am more inclined to Meths, which is easily extinguished with water.  However, I also have a commercially constructed boiler and gas burner, and will be happy enough to use it when I eventually get a boat built. 

Do you have any information on the steam production and fuel consumption of your little boilers?  It would be interesting to calculate the performance against heating area for comparison with other arrangements.

I notice three more replies arrived while I was typing, thank you Paul and Gas Mantle.  It is after midnight here, so I hope you will understand if I respond on those tomorrow.  Really great to have comments and contributions.

Thanks everyone for following along.

MJM460

[MJM460]

Quite a backlog from yesterday, there were even three replies while I was writing.  So let's back up a bit and try and catch up on all the issues this time.

First Willy, you were asking about your boiler and the new engine.  At this stage, you will obviously hook them up and try it, however, the purpose of this thread is to demonstrate that you can make a sensible estimate before you build the boiler, so you have a better chance of a successful result.  If that sounds a bit like hedging my bets, remember that even full size equipment usually needs a performance test to prove that it is meeting the design intent.  Your boiler is an excellent test subject, as the heat from the electric element all goes into the boiler and it contents, apart from those losses to the atmosphere we spent some time discussing.  So let's see what the calculations say. As one who has followed your thread from the start, I am guessing it will not be long before you can test it all out.

First a bit of a reminder of the results we have obtained from your testing.

The electric elements are each nominally 500 watts, but we found as best we could determine from the measurements you were able to make, that the actual heat output was about 430 watts for a variety of reasons when the boiler and its contents were heated up to 135 degrees C.

We also found, from your cooling tests, that the heat loss from the boiler after you added that rockwool insulation was about 150 watts, compared with 310 with the basic timber strip insulation. 
Lets assume the rockwool is still in place.  So 430 - 150 = 280 watts.  Remember, 1 watt is one joule/sec, so the net energy input at 135 deg C so 280 watts or 0.280 kJ/sec.  This is all available for steam production, as the copper was heated to 135 C with the water during heating up, so now raising steam at constant temperature, it absorbs no further heat.  The heat just transfers through.

Now let's look at the steam tables.  We use the section with temperature in the first column, and look at the line for 135 C.  We find the pressure should be 313 kPa which is 212 kPag or about 30 psig, a good check for your pressure gauge.  We also find the specific volume of the steam is 0.5822 m^3/kg and the enthalpy for evaporation (hfg) is 2159.6 kJ/kg.  We divide the heat input per second by the hfg to get 0.00012 kg/s.  Now this is quite a small number and it is tempting to multiply it by 1000 and work in grams, but mistakes are less likely if we don't change the units, and the calculator has no trouble keeping track of the zeros.  We now multiply that mass of steam by the specific volume at those conditions to get m^3/s and by a further 60 to get 0.00453 m^3/min, breaking my own rule to work with rpm!

I am going to keep to the rule on the volumes of your cylinder, but rather than be totally consistent and use revs/s, I am going to carefully stay with revs/min.  So for an engine 1 1/2 inch stroke x 3/4 inch bore the swept volume in m^3 is Pi/4 x (0.01905)^2 x 0.0381 x 2 in each revolution.

That last 2 is for the double acting cylinder. Strictly we should subtract the piston rod cross sectional area, but unless your rod is really huge diameter, it is not a big error.  Let's keep it simple.

A little maths and multiply the volume by your rotational speed of 150 rpm and we get 0.00326 m^3/min.  At first glance we can see that the engine could potentially run at about 210 rpm, which looks promising for a satisfactory performance.

Before we get to excited, we should look at the sources of error in making this calculation.

First, we don't really know at what point your valve cut off will occur, and in any case, the motion of the slide valve driven by eccentrics, even through the Allen reversing gear essentially involves some throttling while the valve opens and closes, so the engine should take in a smaller volume of steam.  Secondly, we do not know just where the exhaust valve will close, and how much recompression of the remaining steam will occur, which could affect the actual steam flow either way, depending on whether the inlet port opens to a cylinder of steam recompressed to above or below the supply pressure.

Third, we do not now about losses around the engine, valve rod packing losses, piston rod packing losses or how much loss due to blow by past the piston, or, dare I mention boiler fittings?

Finally we need to think about the governor action, as the governor throttles the steam inlet to the engine.  With the engine stationary, or at relatively low speed, the weights do not fly out and the throttle valve should be full open.  As the engine accelerates to the set speed, the weights start to fly outwards, and the linkage uses this action to start closing the throttle.  And if the weights, springs and linkages are all suitably in proportion, the governor finds a throttle position that keeps the engine speed approximately steady.  I assume you know all that, but have you thought of how this affects the boiler?

As the throttle closes, the steam flow reduces, but the heating elements are not connected to the throttle, and so the energy input to the boiler continues.  The pressure and temperature start to rise.  With the higher temperature the losses increase, and to make steam at the higher pressure more heat is required.  Either the extra heat required by the steam plus the extra losses is enough to just allow for the reduced steam production, or the heating element control system sees the higher temperature and cuts the power, reducing the heat input to the boiler.  Eventually the system is in balance again, with just the right amount of heat input for the steam the governor allows to the engine.  If all else fails the safety valve should limit the pressure rise.

If you add an engine driven boiler feed pump, the engine will require more steam to supply the power required by the pump, and the boiler will need more heat input to heat the cooler water up to steam temperature.  So the final equilibrium will be at a slightly different point. 

The next step is to complete that engine, connect it to the boiler, and see what speed it runs, but it looks encouraging as a suitable boiler.

I am not familiar with "Bacofoil", I assume it is a form of aluminium foil perhaps with some reinforcing fibre.  Three layers could easily be added using some packing rope or similar flexible material to support them off the boiler.  The spacing is not particularly critical.  And preferably shiny side in, as you said.  If the spacer material minimises air movement out of the gap, it should be quite effective, though most of the foils will conduct heat quite well, and air between the layers will transfer additional heat between the layers by convection.  Hard to prevent this except by evacuating the space between the layers, like a glass or stainless steel thermos flask.  Perhaps better to mount your boiler inside one of those wide mouth thermos flasks, with some suitable flexible spacer/support material!

Remember that the calculations for Paul's firebox showed about  3 Joules lost by convection for each one by radiant heat transfer, so reducing convection is the key.  Also, while foils start quite shiny, so the low absorptivity and high reflectivity increase their effectiveness, but as the surface oxidises, that advantage is rapidly lost.  So for best effect it would be best to use stainless shim.  Lower conductivity and reflectivity maintained better.  Would be a very interesting experiment, particularly of weight was a significant factor.  But it would be easily crushed flat, so not easy to maintain.

Hi Zephyrin, I followed that link to your photo page.  That tank locomotive is truly amazing.  If you are ever building another I think your build log would be very popular.  Absolutely beautiful miniature engineering.  I liked those animations of the valve gear as well.  I am particularly interested in the Joy valve gear, and have built an engine with radial valve gear like the Joy, but with radial links instead of that characteristic curved slider.  I thought it might involve less friction and easier construction.  But you have built the complete slider beautifully. 

I am not surprised the engine frames get hot.  Conduction from the cylinders, radiation from the firebox, mechanical friction in bearings and even churning of the air all contribute to the heating.  That strong draft would certainly help with the cooling, and besides it is a good heat recovery system, using those losses to preheat the combustion air.  I guess the insulation of the boiler is a compromise between external outline and the size of the boiler drum, but I am sure it all helps.  The cork probably blocks all the radiant heat transfer but the tin plate wrapper would look a whole lot better as the outside layer.

Hi Gas Mantle, if we look at the steam flow from Willy's boiler above, the steam velocity will be 14.7 m/s for 5/32 tube, and 4.65 m/s for a 1/4 tube.  The Bernoulli formula can be used to calculate the pressure loss for the velocity change, which turns out to be 167 N/m^2, or 167 Pascal, more usually expressed as 0.167 kPa or 0.024 psi.  Not really surprising that you couldn't tell the difference, it would take a really accurate instrument, if possible at all given the pulsating nature of the flow.  So minimising the loss is correct in theory, but in practice, as you have pointed out, other considerations such as ease of manufacture are almost certainly much more important.

When the steam expands into a large space such as the valve chest, the velocity becomes very low, and the total velocity head is lost.  The velocity head in 1/4 inch tube is nearly zero anyway.

I took it a step further and tried 1/8 inch tube.  I have a couple of short lengths, and the ID was between 2.4 and 2.5 mm, not in accordance with the standard wall thicknesses.  However, the velocity is about 16.6 m/s, so the velocity head is only about 20% more so still not very important.  It certainly justifies the tiny tubes used on the little Mamod and similar engines, but to me it would look too small on a larger engine.

Personally I still use slightly larger tubes than smaller, but I can't really say that smaller will not work.  On the exhaust side, I definitely use the larger size when practical, as any resistance reduces the flow from the exhaust stroke, so increases the exhaust back pressure and reduces the engine power.  There is so little, I don't want to loose any.  But I can see that smaller tubes are necessary on the little locos that Paul and Zephyrin are building.  And they clearly work well.

Hi Paul, I have probably answered your question on radiant heat shielding in the replies above.  Remember that your ceramic fibre probably shields most of the radiant heat which travels in line of sight.  However, fibre is porous so there will be extra convection transfer.  The outer non-porous layer prevents that air flow so is important for heat conservation as well as appearance.  Use the thickness you can accommodate.  The system used by Zephyrin is probably a good guide as to what experience recommends.

I looked up the start date, 11 May last year, so a bit early yet.  I don't dwell on it but it is helpful to my confidence to see that people are still reading, a factor that others have also clearly felt.  But as always, I particularly appreciate the replies by regulars such as you and Willy, but also from all the many others who periodically come in.  So thank you for pointing that out.  I have wondered from the start what the response would be.  I have written in the spirit of developing a knowledge base, building up the application of thermodynamics to our hobby as I gather data and apply the maths that I know from my working life.  I have been surprised by how far it has gone and pleased with the progress.  Though certainly it is a work in progress that would require some editing for publishing.  But I am not against the idea if there is interest from the publishers, and would certainly be prepared to put the effort into tidying up the text.  It is pretty hard to find stuff from early posts despite the excellent search function on the forum, so I am considering celebrating the anniversary by taking some time to assemble an index.

Definitely too long a post, but I hope that I am now up to date with the questions.  I will make it a lighter day tomorrow to allow people to catch up.

Thanks for looking in and for all the responses,

MJM460

[Gas_mantle]

As someone who is interested in running engines on steam the subject of pipes sizes etc is something I've pondered in the past.

My guess is that ultimately the steam has to negotiate the port in the cylinder and any pipe larger than the port is unneccesary. I accept that different sized pipes will have different thermal losses etc but if we stick to steam flow considerations surely larger pipes achieve nothing ?

[steam guy willy]

Hi MJM , Thanks for all this info, There is a lot more to it than i first imagined. I suppose with a larger pipe as there is more surface area in contact with the air, there is also more heat/cooling transfer ! so if the pipe is not insulated that loss can also be part of the calculation. the length of the 5/32 pipe is about 28" !! this was to reach the connection to the Beeleigh engine that was on quite a large plinth .The Bacofoil that i mentioned is the AL foil that is used to wrap chickens and turkeys in when you roast them. Another question .. when the steam enters this 28" pipe with a valve quite close to the boiler and the valve on the engine closed how does the steam and the air in the pipe interact ?? as the engine block is steam jacketed the steam valve is opened and the drain cock on the cylinders opened and drains the jacket to heat up the cylinders . Once pure steam issues from the drain cock ,this is closed and the valve to the steam chest opened . the governor linkage is then manually lifted and the flywheel given a push. the engine then starts and the governor handle released to do its predetermined duty. this i initial procedure does use up quite a lot of steam initially but the boiler can cope with it . new question ....with a loco there is always steam escaping from the safety valves so is this a waste of  energy ? In my boiler there is a pressure switch that turns off the currant before the safety valve lifts. quite easy with the electrical heating system but not so easy with solid fuel boilers !!  so.. i shall find the rest of the sailing boat article and give you the rest of the formulae. Thanks Gas_mantle for the comment about port size. On this Stuart Turner engine block the ports are cast in and rectangular in shape so the cross sectional are can be calculated and used for the round pipe..... we will see what happens later..
Willy

[MJM460]

Hi Gas Mantle, I am starting to feel that you are right, and much smaller pipes could be used. 

Certainly energy losses due to change in velocity at size changes directionally lose pressure, but when the calculations are completed with realistic model engine data, the losses, even though they occur over and over, do not add up to much.  The same energy losses occur at changes of direction, and in driving the steam around the valve outline but in total for a the typical length of tube used on a model again is not much.

My initial thoughts on velocity are certainly over conservative, and much higher velocities are definitely acceptable. 

All the discussion so far has been about energy considerations and the effect of velocity changes.  In addition, we need to take friction into account as a source of loss, which is especially significant on long pipelines.  I have looked to the old fluid mechanics book, and carefully checked the calculations.  For drawn tube, which has a very smoothe wall, the friction factor results in about 0.3 kPa for a 300 mm length, and still only about 2 kPa for Willy's 28 inch long tube, so this again does not explain much.

I do keep thinking I am missing something, as there are some inconsistencies.  I am sure that I have read recommendations based on a proportion of piston diameter.  And most models seem to have at least 5/32 tubes.  Also, if you look at scale sizing, I am not sure what a typical full size engine of the type Willy is constructing would be.  Perhaps 4" bore, so 3/4 inch bore is about 1/5 full size.  I don't know what size steam line would have been used, but perhaps around 1 1/2", so proportional diameter around 7.5 mm, say 5/16" tube. 

Certainly exhaust has additional considerations.  First a much higher degree of pulsation means acceleration losses at every exhaust stroke.  And as I mentioned earlier, a little back pressure reduces flow from the exhaust, meaning higher back pressure on the piston and reduced work output.  I suspect this effect increases the importance of those small pressures enormously, but I am rapidly running out of ideas for calculations that would prove it.  I might just have to accept that much smaller pipes would be satisfactory unless more information turns up.  A bit of a surprise after working for forty years with full size piping, which is sized to reduce pressure drop, though the pipes are much longer.

Of course the obvious answer is to try it.  The steam pipe on a simple test setup, (not after the full diorama treatment) is not a very big task. I would be very interested to hear the result if anyone just tries a very small tube, and I think I now have to add to my project list a few experiments on pressure drop.  I think that measuring the pressure drop would require better instruments than I can afford, however if I do something simple, like say setting up my boiler and engine about  a metre apart, and make up two or three steam pipes say 3/16 that I currently use, 5/32 and 1/8 if I can get a suitable length.  I think it is used in refrigeration so it should be available, but the lengths I have at the moment are only about a foot long.  If I get sensible results with a very free exhaust, it would be interesting to continue by trying a smaller exhaust with some length as well.

The advantage of this approach is that the engine will make the flow pulsations realistic, and I can calculate the flow based on water consumption.  I am thinking some plastic tube will either melt or provide suitable insulation.  Don't hold your breath, but I will get there eventually.

Hi Willy, what do you think would be a suitable scale steam line size for your engine?  I have probably exhausted that topic in discussing Gas Mantles comment.  While in principal increasing the size earlier is best, in practice I think I finally have to agree that in practice you are unlikely to see a difference, especially with the governor in operation, so let availability of the tube and appearance dictate.

When there is air in the boiler, and steam is introduced by evaporation over the whole liquid surface area, the system is well mixed and we have liquid and vapour and a phase change and the air affects the boiling temperature.

In your steam pipe, it is very different, you are just mixing two gases, the air does not affect the liquid boiling temperature.  When you first open the boiler valve and steam enters the end of the pipe and the small cross sectional area does not allow good mixing.  There is a tendency for the steam to act a bit like a piston, and compress the air to the same pressure as in the boiler.  However, without a solid piston to clearly separate the two, there is some mixing and the interface will involve a more or less gradual change in concentration from steam at the inlet end of the pipe to air at the other, with most of the change occurring over a short length.  If the pipe is still blocked at the engine end, that random molecular motion we have talked about before will result in mixing over a longer length, but I am not sure how long it would take to be completely mixed.  More likely you open the valve to the jacket and engine, and the flow is initially air but quite soon the entire mass of air has gone and you then have steam from the boiler.  Of course, if the boiler already had air in it from when it was filled, the process will involve the air in the pipe being pressurised by that steam/air mixture.  There will still only be slower mixing with the extra air initially in the pipe, so the engine end will still be mostly air until it is purged out.  As there should be no air being admitted unless there is a lot of air dissolved in the feedwater, the mass of air initially in the boiler is soon purged out and the engine is run on steam.  The initial air in the steam jacket will affect the condensing temperature, which will not be 100 degrees until the air is purged out.  Makes that initial heating a bit gentler I suspect, ideally an air vent as well as the condensate drain valve, but a bit to theoretical to worry about unless you find a problem.

Yes, the steam leaking from the locomotive safety valve wastes energy, as the heat in the steam cannot be recovered and used in the engine, but as you say, the coal fire is not as easy to turn down as your electric element so they don't have much choice.  Another advantage of your electric boiler, the heat input is fully controllable between zero and 100%.

In the talk about pressure drops, I have glossed over the heat transfer issue.  There is the same flow through the pipe, what ever its size.  Clearly, a larger tube has a larger surface area and a longer residence time for heat loss so probably looses more heat.  On the other hand, a smaller tube has higher velocity so higher convection film coefficient on the inside thus increasing the heat transfer.  So pluses and minuses.  I think the main thing is that both require insulation, and if the outside temperature burns if you accidentally touch it, try more insulation thickness.

I thought I was aiming for a shorter post tonight, just as well I was not expecting a long one.

Thanks for looking in,

MJM460

[steam guy willy]

Hi MJM , the steam pipe on the Beeleigh engine is 4" outside so a scale pipe would be 1/4' as it is 1/16th scale ....however one can not scale nature  so do the calculations get adjusted with very large and very small thermodynamic installations ?? thanks also for further info....
Willy

[derekwarner]

Guys....we have some confusion here...steel pipe according to American Standard ANSIB36.10, is measured by nominal bore [NB],

Steel tube is manufactured to an OD dimension

3" NB = 88.9 mm OD...this dimension is constant with the schedule [Pressure Rating or wall thicknesses] hence reducing the ID
This same 3" NB pipe could be schedule Standard [5.49 wall], Extra Strong 7.62 wall] and XX Strong [15.24 wall]

[So in round figures, a 3" NB diameter** pipe could be as little as 60 mm actual bore]

The next size up from this is 4" NB which is 114.3 mm OD

There has been no such product produced as a 4" OD steam pipe

Steam machinery produced at the turn of last Century was invariably produced with the heaviest schedule steam pipe irrespective of the steam pressure

The only reason I offer this is  the great possible variance even before scaling down, hence we really need to understand the scale variance when referenced back to scale sized OD tubing

Derek

** word should not have been used

[MJM460]

Hi Willy, the issue of scale is a perplexing one, but it always makes more sense with a real example.  So let's have a look.  Now 1/16 scale means 1/16 full size.  From the appearance point of view, linear dimensions are all scaled by that same scale factor, so the 1/4" tube will look right on your engine, while a 1/8 tube would look ridiculous, regardless of friction losses or lack of them.

Then, your 3/4 bore by 1 1/2" stroke engine corresponds to a 12" bore by 24" stroke full size engine.  We need to know the power rating, operating pressure and running speed of the full size engine to compare with what might happen in the model.  But we can say the piston area of the full size and  model can each be worked out using the formula for the area of a circle, Pi / 4 x d^2.  If you work them both out, you will find the area of the model piston is not 1/16, but 1/256 of the full size.  You need to carry all the decimal places to get the exact answer, or you can take my word for it.  So the steam pressure works on a much smaller area to provide torque relative to the linear scale.

The volume of your model will will be 1/16^3 or 1/4096.  Now the model is probably made of very similar density materials to the full size engine.  So the mass of the model will be 1/4096 of the mass of the full size engine.  This is very convenient, as it will make it a lot easier to carry to an exhibition than if it were 1/16.

When we get to performance, things get more tricky.  First, the model cylinder volume, will be 1/4096 of the real engine cylinder volume, so it will only take in 1/4096 of the volume of steam.  To work out what that means, we need to know about the steam pressure used by the full size engine and the speed it normally ran at.  I expect that you have that data, and probably a good idea what speed you think the model should run at, so rather than my wild guesses, let's wait until you post that data before I look further at performance.

Of course, most of our models tend to be run unloaded, including mine.  This is a pity, because it means they are not doing any of the work they were intended to.  I have made a tentative start with driving a DC motor as a generator, so now I have to add an electrical load.  I have a standard Meccano size end on the shaft of each engine in preparation.  A generator is probably wrong period, unless you build up an open frame machine with visible carbon brushes, so a pump or mine lift, or perhaps a timber saw would be more realistic.  Even an overhead shaft and a full workshop per J.L.   Another whole branch of our fascinating hobby. 

Hang on, I was thinking of your freelance mill engine, but I just read your post again, you said the Beeleigh.  If I remember correctly, that is a compound engine with condensing and condensate/air pump.  That will be a whole new level of test run.  Need a manometer for vacuum, perhaps better a gauge until you have some idea of the vacuum you will achieve, and perhaps a load so you can get reasonable pressure in the hp cylinder.  And of course a method to measure the condensing water flow and temperature rise.  Or am I mistaken again?

However, running unloaded means regardless of the boiler operating pressure, the regulator and governor valve will both be nearly closed, and only minimal pressure is seen by the piston, as you would see if you had a good pressure gauge on the steam chest.  Temperature is no help for pressure once you are out of the boiler.  So that is another thing to remember when assessing the model running conditions.

Scaling of the engine to continue as data is made available.

Hi Derek, good to see you back again.  Did you ever get sorted on that condensate issue?  I must admit that I never even thought of using steel tube in full size, as there are not so many pipe fitting available to suit.  To me, pipe implies to ASTM ANSI Standard B31.3. Or occasionally the similar pipeline codes, and rarely to BS (but mostly only for plumbing) or DIN, but that just reflects my oil industry background.  I guess the one thing for certain is that the Beeleigh engine would not have used ASTM standard piping which is as you say, though the OD standard dimension results in and ID somewhere in the region of the nominal diameter depending on the schedule selected.  However, these calculations are at best very approximate, so to use the nominal diameter results in easier maths, and not large errors in the grand scheme of things.  Obviously for a detailed calculation of the discharge pressure required of a compressor for a long pipeline requires knowing the exact specified diameter and even the manufacturers tolerance.  More likely an early BS piping code might have been used, (I can look out the ID for more recent versions if you wish), or perhaps even manufacturers standard prior to that.  It might be a bit like early thread "standards".  I really don't know the history, my experience only goes back to when I started work in 1967, though the codes were a few years old by the time we got them in the pre-Internet age, as you know.  But which ever way, use of the nominal diameter as the inside diameter gives an idea of the difference between pipe sizes without the complication of many significant figures in the calculation, especially when comparing real and model sizes.  When I calculated the tube velocities for the model sizes, I did use the wall thickness of the tubing I have on my shelf, but the actual numbers still depend on the wall thickness you are using as tubing at least under the AS codes is it is manufactured to a standard OD and several wall thicknesses.  In fact I probably have a mixture of wall thicknesses as when I get to the shop, they rarely have all sizes available, and anyway, I am not sure which would be most suitable anyway.  They are all adequate for the pressures I use, it is about bending them without undue flattening.  So I hope you will forgive this approximation for the sake of simplification.

Finally managed another test run of my engine today.  Basically quite a successful run in terms of data gathered, though the temperature really rocketed up between 40 and 80 deg C, I presume when those water tubes started to produce steam, and I missed a few readings.  Forty degrees in less than a minute caught me by surprise, just didn't even see it happen.  I melted the nice plastic handle on the thermocouple I was using on the stack temperature, should have installed an insulating heat shield on it, but it still seems to work ok.  Also, when the engine got up to 2000 rpm by the non-contact digital tachometer, burner moved around quite a bit due to the vibration, and ended up in the wrong place.  I also had to tighten a safety valve fitting leak, so there was a little steam leak until I found the right size spanner. 

For the cool down, I removed the stack thermocouple and put a block of wood to block the stack air flow.  Then the cooling is just due to heat loss through the insulation.  Of course the temperature inside the furnace is no longer raised by fuel combustion, so I will have to think about how to allow for that.  I have now made a base to allow the burner to be fixed in place for next time.

During the cool down, all proceeded normally until 55 deg and I was waiting for 50, when the temperature suddenly jumped to 65.  This happened once before, so I decided to just continue the cooling until it again reached 55 and continued to take readings while it cooled until I eventually stopped the timer at 39 deg C.  I have swapped over the thermocouples since last time, so it was the same instrument but a different thermocouple to last time it happened.  The instrument takes in two thermocouples.  All the while, the second one read totally rationally.  I have no idea how to explain that.  We will see what shows up, if anything, when I do the calculations.  Any ideas are welcome.

No calculations as yet, but at least I now have the data, so can make a start.

Thanks for looking in,

MJM460

PS corrected ASTM to ANSI, should not have forgotten that one already. 

[steam guy willy]

Hi Derek  , the pipe i think is a cast item as can be seen in the picture with a cast in flange and blobs for the risers during the casting process ..so how thick the walls are i don't know. also it was made about 200 years ago !!!

[MJM460]

Sorry that I missed last night.  I knew I would be late, as we had to deliver a granddaughter to the theatre prior to her show, but it was after midnight when we got home after the show.  And we had to take her brother to the basketball at 08:30.  Both parents were tied up with other activities that have them quite busy, but they don't usually all coincide like that fortunately.  Life was easier when I had to go to work, but not as much fun.

The actors put on a wonderful interpretation of the Shrek musical, plenty of youngsters coming in to keep up the numbers.  Basketball was the grand final.  In the end, they lost in the last 30 seconds, but no disgrace to come runners up , especially in such a close game.  Never more than 6 points between them.  As the sign on the wall says, they are kids(under 10's), it is a game, the coaches are volunteers and the referees human.  And it is not the NBA!  But a wonderful game.  When we arrived, two previous games were both just in the final few minutes.  Both were a draw at the final bell, so went into an extra 5 minutes.  One of them, the draw was achieved in the final 3 seconds.  So a kids league, producing some wonderful games and heaps of fun and skill development for the players.

Willy, do you know what pressure they ran that cast iron pipe at?  I suspect not very high.  These days cast fittings such as valve bodies and so on have to be cast steel.  But then, cast iron was probably as good or better than the boilers.

Not many calculations done today.  Definitely in rest up mode.  Perhaps tomorrow.

MJM460

[derekwarner]

Willie.....you may be surprised at that cast pipe being steel and not iron...[if you had access to it, one minute with a triangular hand file could determine the material]

In the first year of my apprenticeship [1966], each Saturday [overtime $ ] my task was the mechanical check of two huge Broom & Wade twin cylinder [HP & LP without intercooler] low speed air compressors.....each flywheel was taller than me.......

The were manufactured in the UK & from memory .....Year Manufacture - 1890 on the name plate

All interconnecting pipework was cast steel...........the complete compressors entablature, casings & pipework all painted gloss light cream colour.....with polished copper lubrication tubing

Derek

[MJM460]

Hi Derek,  that's interesting, that cast steel was being used so early.  Do you know when it first came into use?  I don't know the history at all.

MJM460