Author Topic: Talking Thermodynamics  (Read 194831 times)

Online Kim

  • Global Moderator
  • Full Member
  • *****
  • Posts: 7863
  • Portland, Oregon, USA
Re: Talking Thermodynamics
« Reply #165 on: July 25, 2017, 02:39:07 PM »
The Biot number is defined by Bi = h x d/ k, where h is the convection film coefficient, and k is the conductivity of the fin material, and d is half the fin thickness (please don't ask why the half, it's in the book).

Intuitively, it seems that the 1/2 the fin thickness makes sense because the maximum path for the heat transfer would only be 1/2 the thickness of the fin.  The heat can conduct to the outside on either side of the fin, it doesn't have to travel all the way through the fin.

Might be wrong, but this logic makes sense to me :)
Kim

Offline paul gough

  • Full Member
  • ****
  • Posts: 457
  • Tropical Queensland, Australia
Re: Talking Thermodynamics
« Reply #166 on: July 25, 2017, 02:49:00 PM »
Thanks very much for offering to provide some ideas which might resolve the question at issue. Please don't feel obliged to provide explanations that labour the point, it is my responsibility to work at understanding what is presented then to seek further assistance if needed and as this is not paid work for you please engage with my enquiry at your leisure and when you think it appropriate to the discussion. My comment about 'labouring just as intensively' was not meant to be literally the case, more a bad metaphor to indicate I/we are paying attention, working at keeping up with the prodigious information flow, and attempting to assimilate it as best I/we can, I guess  something akin to diligent students.   

As a Far Nth. Queenslander,I don't regard Rockhampton as being in the tropics, just on the edge of it, a bit of a conceit I admit. I reside in the hills behind Cairns. Regards, Paul Gough.

Offline jadge

  • Full Member
  • ****
  • Posts: 503
  • Cambridge, UK
Re: Talking Thermodynamics
« Reply #167 on: July 25, 2017, 03:21:16 PM »
I had a most interesting chat with steam_guy_willy a couple of years back at the Forncett Model Engineers day about his electric boiler. I was interested as I wanted a simple boiler to provide steam for injector testing.

The heating elements used are obtainable from commercial distributors such as RS in the UK. They are basically resistance heating elements running on 240VAC, at least the ones I was interested in. Although they contain a helical wound element I doubt that the inductance will have any impact at 50Hz. One could get an estimate of inductance from the Wheeler formula, but even if it was 1mH that's only about 0.3ohm at 50Hz. If the heater is 500W then the DC resistance is on the order of 115ohms when hot.

The internal insulating material is magnesium oxide; average thermal conductivity but high dielectric strength. The outer case is usually sealed stainless steel.

The heating element depends upon a good heatsink to work. If you have an element about the size of your finger dissipating 500W and relying primarily on convection in air it's going to get mighty hot, mighty quick! Hence the need for a reamed hole and thermal grease. The manufacturer of the elements sell a high temperature grease, but rather stupidly RS do not stock it. Since I need the boiler to run at 170psi (~190°C) that rules out most of the general heatsink compounds. In high power electronics any sort of air gap, or even air pockets, is death to heat transfer, and thus to the semiconductor device.

As an aside I decided not to follow the steam_guy_willy design on the electronics side. For £18 I got a PID controller, thermocouple and solid state relay from Ebay. So it should be simple to implement a closed loop control system, based on temperature, and thus indirectly steam pressure. Mind you it might be prudent to add a proper safety valve, and possibly a water level detector.

The main thing I know about thermodynamics is that, despite doing a course in it at university, I still don't really understand entropy.  :noidea:

Andrew

Offline steam guy willy

  • Full Member
  • *****
  • Posts: 3239
Re: Talking Thermodynamics
« Reply #168 on: July 25, 2017, 03:22:24 PM »
Hi MJM, These are the RS Components  Cartridge heaters that i used in this boiler. They are the 500 watt 100mm x10mm ...250volt ones. the boiler dimensions are 3" x6". The boiler takes about 7 mins to get to 2 bar (30 psi) , this is to get over the attention span problems of the youth today !!!. I do not use any sealant just push them in slowly as the air needs to come out. Interestingly when i demonstrated it at Beeleigh mill because it was outside they had a large extension cable un wound and it took about half an hour to heat up !! I think the inductance of the coil had something to do with it. This boiler was featured in the Engineering in Miniature magazine a few years ago......No Questions at the mo ,so you can catch up !
.

Offline steam guy willy

  • Full Member
  • *****
  • Posts: 3239
Re: Talking Thermodynamics
« Reply #169 on: July 25, 2017, 03:32:26 PM »
Hi Just seen the above comment that overlapped mine. when i talked to RS they did not mention the high temp grease so i did not use it.!! I used the control system designed by a friend that used a 9 volt battery for the low water safety circuit to trigger the 250 volt circuit, incidentally you do need a fairlly good battery ,as although the led's light up the circuit will not function properly !!

Offline Admiral_dk

  • Full Member
  • *****
  • Posts: 3752
  • Søften - Denmark
Re: Talking Thermodynamics
« Reply #170 on: July 25, 2017, 09:01:53 PM »
The first thing that springs to mind when considering a high temperature insulator that is a good thermal conductor, is aluminium oxide as the material is used inside certain electronic components as such. The thermal conductivity of a few um is almost the same as aluminium and it will withstand thousand volts or more (depending on layer thickness). Aluminium oxide is extremely strong / tough, but still it can be damaged and then the insulation properties are gone, so ....  :noidea:

Best wishes

Per

Offline jadge

  • Full Member
  • ****
  • Posts: 503
  • Cambridge, UK
Re: Talking Thermodynamics
« Reply #171 on: July 25, 2017, 10:14:50 PM »
Some power semiconductors, especially RF power devices, used beryllium oxide as an electrical insulator, but also with a very high thermal conductivity. There were dire warnings about not cutting open such devices and creating dust, as BeO is carcinogenic.

Andrew

Offline steam guy willy

  • Full Member
  • *****
  • Posts: 3239
Re: Talking Thermodynamics
« Reply #172 on: July 26, 2017, 01:28:45 PM »
Hi,  Entropy !!  Is this the elephant in the room ?? James Clark Maxwell certainly thought so........!! Found this reference in quite a good book that should be on every model engineers shelf? !!!!!!

Offline MJM460

  • Full Member
  • ****
  • Posts: 1647
  • Melbourne, Australia
Re: Talking Thermodynamics
« Reply #173 on: July 26, 2017, 01:50:34 PM »
Thank you everyone for the helpful comments and contributions.  I will try and do justice to each of you, however it is getting very late due to an unexpected water leakage problem we discovered this afternoon and that had to be attended to first.  Have just about everything dry, and back in place and will attend to repairing the fault tomorrow.

Kim, I believe that you are on the right track with your suggestion of the two sides of the fin meaning there will be a two somewhere, alternatively a half will have to appear in a number of other places.  I think it is just a matter of convenience which approach is used.  On the issue of conduction path, remember that heat has to flow from where the fin attaches to the cylinder along the the complete length of the fin to the tip, not just through the fin thickness.  This is why there are diminishing returns on making fins ever longer.

Paul, FNQ makes sense, thank you.  Unfortunately turning left at Rocky, but if you ever visit the Deep South, please do yell out.  I am trying to find that balance between too many words and too superficial.  So please do say something any time you need a bit more explanation.  Life does not have to be that hard.

Thanks jadge for the extra information on the heaters.  I totally agree with you that the inductance would not significantly change anything.  Also, only the resistance is responsible for the heat, and we only measure resistance with a normal multimeter.  Good information on the MgO and the stainless steel sheath thank you.  Interesting that the manufacturer does sell high temperature grease.  However, I also wonder if there is any problem inserting a close fitting sheath in a reamed hole with a coat of grease, could the air get out?  Or would it make a very effective pneumatic ram?  The PID controller is a great solution for precise pressure control.  I think Willy's system is a simple low level protection,  and he expects to use all the heat, so temperature control is not really an issue.  And I would always recommend having the safety valve.  And low level protection is also a good idea.  While pressure inferred from temperature measurement is probably more accurate than little pressure gauges, when the system is first heated there is still air in the boiler, so the pressure can be higher than you would expect based on the temperature, until the air is lost with the first steam production.  Remember the mountain top experiments?

Willy, thanks for the element data.  Looks like a good range available for many boiler sizes and capacities.  If you are using those little rectangular 9 Volt batteries, they will not last very long with an LED and a relay.  The relay should be energised to make contact in the 240 V circuit, so that in case of a fault the power is isolated.  The ones I am thinking of would not last very long as they have very little energy capacity.  I would recommend a 9 V plug pack with a reasonable current capacity.  After all you already have 240 V power for your heater.  Alternatively a 12 V motorcycle battery with the appropriate regulator to supply your 9 V would have enough capacity to last a reasonable time.  An unusually long boiler heat up time at your display venue could be due to excessive voltage drop in the long extension lead, especially if there is a poor contact somewhere along the way.  Remember the power is V^2/R, so a small under voltage does significantly impact on your heater power.  I do not advocate exposing mains power terminals to measure the voltage unless you are an electrician, but you could check each plug and even the power switch with one of those infra red temperature devices.  They are excellent in such applications.  A hot spot indicates a poor contact, and may indicate a faulty cord or switch.

Thanks also to jadge and Admiral_dk for suggestions of further materials with suitable combinations of thermal and electrical properties.  Clearly there are many options.  Obviously the carcinogen one would be well avoided, but I suspect that enclosed in a stainless steel sheath it would be well enough protected from damage in normal use.  I suspect to find a material that could be bent or deformed without consequence would be more of a challenge.  I hope most users would understand that the element should be treated with reasonable care.  Above all it would be interesting to know which insulating material the manufacturer has selected.

Now recognising the importance of a proper heat sink to the element, and the need to avoid impediments to thermal conduction, we are back to Willy's conundrum of whether to machine the grooves in the outside of his pressure boiler insert.  In most situations, we could measure all the inlet and outlet temperatures on two samples and easily deduce which is most effective.  But in this specific electric heater case, the element temperature varies to ensure rejection of all the heat, and we can't measure that temperature.  With the heat input to the boiler always equal to the element power rating, there is no expected variation in time to heat, whether the outer surfaces has fins or not.

I would suggest that the mathematics to calculate the difference is certainly too hard for me, and I am not sure if the heat transfer coefficient can be calculated with sufficient accuracy to give a clear answer.  I suggest the most definitive way to answer the question is to make two element carriers, one with grooves and one without.  And conduct an experiment.  But what experiment?

I have implied that the element resistance is constant, however resistance of a metal is temperature dependent, and does change as the temperature rises.  I seem to recall that the resistance would increase as the temperature rises.  Increased resistance means reduced current.  So in principal, if we had an ammeter in the circuit, we might expect to see a different current in each case, and the lower current implies lower temperature, hence better heat transfer.  But I am not sure how much the temperature changes, and whether the measurements would be sensitive enough.  Also I am definitely not advocating putting your multimeter in the mains circuit to measure the current.  Perhaps one of those clamp meters or a Hall effect current detector could be used if it is sensitive enough.

Another possibility is to make the two element carriers say 12 mm longer than required, assuming the boiler shell is long enough to accommodate the longer element mounting.  A thermocouple could then be tucked in after the element is inserted, and perhaps a little kaowool or similar fibre  insulation pushed into hold everything secure.  The thermocouple would then measure a temperature which might be close to the element sheath temperature, and more importantly, might vary in the same way as the element temperature.  Again the lower temperature implies better heat transfer to the water, and hence gives our answer.  I think I would find this method the most practical, and probably most likely to provide a definitive answer.  But I would have to try it to find out.  What do you think?  Is there another method that might resolve the issue?

I hope the plumbing issues will be resolved tomorrow and that at last I can not only address that condenser, but also think a bit about Paul's locomotives.

Thanks for following along

MJM460

PS Willy I saw your post just as I was about to post today's essay.  Looks like I will have to have a go at entropy sooner rather than later.
The more I learn, the more I find that I still have to learn!

Offline steam guy willy

  • Full Member
  • *****
  • Posts: 3239
Re: Talking Thermodynamics
« Reply #174 on: July 26, 2017, 02:23:57 PM »
Thanks for the latest info.......... New question  ........would the boiler produce steam quicker if the filler cap is left loose to allow the air to escape and not be compressed by the expansion of the heating up water ??or could you increase the pressure in the boiler first with a bike pump ???.... and ........if there is a device to vibrate the boiler at a certain frequency would this also speed up steam production ?? please only answer these questions at your leisure !! These cartridge heaters were never meant to be used in my model boiler i suspect, but used primarily to heat up large chunks of metal used in industry where one could have an exit hole to release the air. Also the grease would enable the item to be removed easily if it needed changing.
« Last Edit: July 26, 2017, 02:29:43 PM by steam guy willy »

Offline paul gough

  • Full Member
  • ****
  • Posts: 457
  • Tropical Queensland, Australia
Re: Talking Thermodynamics
« Reply #175 on: July 26, 2017, 04:10:50 PM »
I am sitting here at 1am pondering the interaction of the 'dancing' molecules of steam, (a gas), interacting with the 'buzzing' somewhat constrained molecules of the piston surface, (a solid). Now, one can easily comprehend the affect of the force(s) of the steam molecules on the piston as analogous with the force(s) of a ball thrown against a wall or some such. But is our mechanical or engineering description/analogy all there is to it and sufficient??? After all most volume of the piston is in fact the space between the atoms, this leads me to think that some sort of other interaction(s) is/are occurring rather than just our day to day and somewhat 'gross' explanation. Are we in fact talking of molecular/atomic forces interacting and if so there would presumably be losses involved in any energy transfer between them, are these significant or relevant??? I don't want to side track this thread into the dense and dark regions of physics or expect any technical explanation, however I would appreciate having a laymans handle on the phenomena and if there is any connection or relevance to our discussions on thermodynamics. Regards, Paul Gough.

Offline steam guy willy

  • Full Member
  • *****
  • Posts: 3239
Re: Talking Thermodynamics
« Reply #176 on: July 26, 2017, 08:22:03 PM »
Hi, Paul, I an see what you are getting at and are some metals that may be a bit elastic or have large spaces between the molecular structure be better at transferring the energy in the steam propellant. err ,um, i think what i am saying are some (things) better than others when used as a piston in a steam engine ???

Offline steam guy willy

  • Full Member
  • *****
  • Posts: 3239
Re: Talking Thermodynamics
« Reply #177 on: July 26, 2017, 08:24:29 PM »
Hi MJM, Saw this and thought of you !!! I am a member of the Newcomen Society and they have in there magazines quite a lot of info on Australian engines...........

Offline paul gough

  • Full Member
  • ****
  • Posts: 457
  • Tropical Queensland, Australia
Re: Talking Thermodynamics
« Reply #178 on: July 27, 2017, 12:42:06 AM »
Hi Willy, The physical properties of metals at the molecular level and their various capacities to exchange energy is not something I would claim any understanding to, so whether an 'elastic' metal had different behaviours to an inelastic one I have not a clue, to me it would be like trying to speculate on what happens in a black hole. We can only hope there is a higher authority amongst us who can deliver enlightenment.

I also found the Newcomen society a valuable storehouse of very interesting information, I spent many weeks reading a large proportion of their articles going back to Volume 1 of the transactions, (Journal), while I was researching disk engines some time ago,  they are available to members via the web. Regards, Paul Gough.
« Last Edit: July 27, 2017, 12:49:40 AM by paul gough »

Offline MJM460

  • Full Member
  • ****
  • Posts: 1647
  • Melbourne, Australia
Re: Talking Thermodynamics
« Reply #179 on: July 27, 2017, 01:12:47 PM »
Well at least the plumbing issue is resolved and fixed.  No amount of theory will predict that the tube had not been pushed far enough into a quick-lock fitting.  But something made me look very carefully into the fitting, make some measurements and stick a little bit of tape on the tube to mark the correct insertion.  I had just not used enough muscle.  I had to get that out of the way before discussing entropy.

I have avoided mentioning entropy up until now but Willy, your post last night suggests it is on your mind, so let's have a go.  The page you scanned from your book made a quite suitable introduction. 

You may remember that I described enthalpy as a property that only depended on pressure temperature pressure and specific volume, all things we can measure.  It did not depend on the previous history of how the substance came to have that particular pressure, temperature and specific volume.  That sentence is basically the definition of a property.  I noted that enthalpy is calculated from the pressure temperature and specific volume, we don't have an enthalpy meter.  But the calculation of enthalpy comes into many significant problems and thermodynamics (particularly problems involving heat and work, so obviously interesting to anyone designing and operating heat engines).  It is so useful that it is convenient to have the value tabulated in steam tables, and the tables for other fluids such as refrigerants.

Entropy is another such calculated property, given the symbol S.  No such thing as an entropy meter, it is the outcome of a calculation.  I think the main reason that entropy is so mysterious is that it is not a simple calculation involving things we can measure, but it requires a bit more complex calculation that has a surprising applicability.  The calculation involves summing the quotient of heat input divided by temperature as a substance undergoes an ideal reversible process, which goes in such small steps that the temperature can be considered constant within each step.  In symbols it is written
dS = dQ / T at each step, and leads to the result that e change of entropy with the process (S2-S1) is found by integration  of dQ/T through the process.  Now none of us want to have to do integral maths to understand our engines, but we don't have to.  The unexpected result is that this change in entropy, even though it is calculated by considering an ideal reversible process, is exactly the same, whether the process was reversible, or ideal, or was not reversible.  The reversible process allows the change to be calculated, but then the result also applies to any real irreversible process as well.  So we can leave it to a few boffins to calculate the values  and include them in our steam tables, and tables for other fluids.

Now apart from the interesting unexpected nature of this property, why do we care?  It comes down to the second law of thermodynamics.  If we first note that the first law calculates the heat exchanged in a process based purely on conservation of energy, but gives us no idea of whether the process can actually occur.  For example we can calculate the heat lost by your tea to your teaspoons when you plunge them in to the hot brew.  The first law tells is that it is the same amount as is gained by the spoons.  But we can also calculate how much heat would be gained by the coffee if it became hotter by cooling the teaspoons.  Now we know that can't happen.  The second law of thermodynamics tells us it can't happen, but does not really quantify why it can't happen.  This is where entropy comes in.  It turns out that the only processes that can happen (without heat or work input) are processes that result in an increase of entropy.  This is relatively easily extended to mean that the entropy of the universe is increasing.  Philosophers ponder if there is a limit, if so what happens when the limit is reached, or is there a mechanism somewhere in the universe that reverses the increase of entropy.  But I find all that way too esoteric.  I would rather leave such discussions to someone who cares.

However, accepting the concept of entropy, and having it tabulated in steam tables is very useful.  So let me illustrate by showing how it helps us understand a steam engine.  Let's assume we have a boiler, and the steam outlet pipe then loops back through the fire box a couple of times as a superheater, then on to the engine.  We can measure the temperature and pressure at the engine inlet. We would all like to know how much work our engine can produce.  We need to know the exhaust pressure, but we don't know the exhaust temperature.  We know the amount of work is given by the change of enthalpy, but we need to evaluate that change.  Now we remember that ideally steam expansion in an engine is an adiabatic process, meaning no heat transfer in or out.  And it turns out that an ideal or adiabatic process also means no change in entropy.  So if we look up the entropy of our steam at the engine inlet, it will have the same value at the exhaust of our adiabatic engine.  With a bit of interpolation of the steam tables, we can find the temperature, and enthalpy of the exhaust steam, and also the exhaust steam quality or dryness if the exhaust is wet steam.  Now we can easily calculate the change in enthalpy by subtraction.

Using entropy, just using the tabulated values in the steam tables, has allowed us to calculate the power output of an ideal engine.  The second law of thermodynamics tells us that any real engine will produce less power than an ideal or adiabatic one.  This leads to a definition of adiabatic efficiency, but that can come another time.

Now your other comments, letting the air out when your element is inserted.  Your explanation seems likely.  It would take some ingenuity, but it should be possible to devise a way to solder a return bend on the end of your reamed tube, and use say 3 mm tube vent back to the mounting flange. This would allow you to use the grease.  But your other suggestions to reduce the time to raise steam?  I am tempted to stick my neck out and suggest the only way to raise steam quicker is to put in more heat, either use a bigger element, or put in two or three elements.  You see your heater power rating is the amount of heat generated by the element per unit time.  A 500 Watt element means 500 Joules/second.  We have previously looked at how to calculate how much heat is required to heat water from cold, say 15 deg C to saturation temperature, and we can look up the specific heat of copper and calculate the heat required to increase the copper temperature from 15 to our operating temperature.  You have an advantage over fired boilers, you know how much heat is produced and you can insulate it well to limit loss to the atmosphere, though we should make an estimate of the heat that will be absorbed by the insulation.  So you only need to weigh the empty boiler, and the quantity of water you fill , and you can calculate the time required.  Similarly, once you start steam production, you can calculate the heat required per kg of steam, and so you can easily determine how much steam you can make with 500 watts, as no more heat is absorbed by the copper or insulation once steady temperature is reached.  All your effort to improve the heat transfer coefficient only reduces the temperature the element must reach to transfer the rated heat.  It does not affect the time to heat or the the amount of steam you can raise.  So long as you insulate the boiler well!  I will hold the air questions until I get to condensers, and try and address Pauls comments tomorrow. 

Paul, you have obviously understood well my explanation of how heat is changed to work in our engines. I will address your questions on this next time.

Oh and thanks for the notice about the Mildura conference.  Mildura is about 600 km from Melbourne, quite a solid drive.  Unfortunately I have too many other commitments in October, but it's good to see these events being held "locally".

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

 

SimplePortal 2.3.5 © 2008-2012, SimplePortal