Thoughts on Newcomen engine.

[John Hill]

Newcomen's atmospheric pumping engine was one of the very earliest practical and useful steam engines. The operating principle was easy enough, steam entered a cylinder where a spray of cold watet caused it to condense and the resultant loss of pressure caused atmospheric pressure to push the piston down.

The engine was very inefficient and part of the reason was that steam entering the cylinder was prematurely condensed by contact with the cold cylinder walls. James Watt improved on that by making a seperate condensing chamber.

However, I am wondering if cylinder walls made of low conduction materials would enhance efficiency of the basic Newcomen engine, and what might be a practical material to use?

My intention is to build a Newcomen engine for amusement and not a faithful copy of any that were made and I will use a crank and flywheel rather than the pump rod and beam that Newcomen used.

I hope someone can answer my question, thanks.

John

[chrismb]

Funnily enough, I've never been very convinced by that argument for the machine's inefficiency either. The cylinder should operate pretty much at 100C where both water and steam can exist, because both fluids are passing through.

The main inefficiency, to my mind, comes from the fact that as the water causes the steam to condense so the steam releases heat. The water actually has to sink all of the latent heat of evaporation during the phase change. That heat has nowhere to go, except into the water, so unless yet more and more water is poured in, which can't really happen quick enough, you only ever get partial condensation during the power stroke itself. Whereas in the 'Watt' machine, because the walls of the condenser cylinder are already cold, as the water changes phase so there is already a thermal sink, the cooled cylinder wall, for that thermal flow.

So I don't really buy that the inefficiency comes down to 'pre-condensation' in the Newcomen design, per se*, simply that the 'Watt' design provides an appropriate heat flow for the heat being rejected by the condensing steam (just take a look at a Carnot diagram!). Where I see the relative inefficiency coming from is because the 'Watt' design can draw out so much more of that energy per stroke (as the steam is not so fully condensed during the power stroke in the Newcomen), and in turn this means that the energy losses for a given flow of condensate around the relative systems are, per stroke, lowest in the 'Watt' machine, and it is this condensate that is the main route of the thermal energy losses.

I reckon if you could recover all the condensate out of a Newcomen engine and keep it thermally insulated to be re-heated and re-fed, then the efficiency would nudge up towards the Watt design. But there again, if you could do that then you could equally bolster a 'Watt' machine by the same means.

*I'm not particularly knowledgeable about the details here, but these are broad-brush strokes on why the usual argument doesn't seem to quite hang together - it strikes me that as long as the steam is admitted after TDC then the 'condensing' argument falls apart, because that's exactly what you want to happen after TDC in a Newcomen!! Conversely, it looks to me like there isn't enough condensation going on in Newcomen, rather than too much!!!

[John Hill]

Hmmm, my understanding of the supposed ineffeciency of the Newcomen engine is that as steam is being inducted through the entire up stroke it is during that time that premature condensation takes place.

It is a bit like the 'flame licker' engines which draw hot combustion gases in then the piston is drawn back as the gases cool. They are very inefficient too and wont even work if the cylinder is too cold presumably causing premature cooling of the gas.

So I am thinking if I could insulate the cylinder bores somehow I could avoid premature condensation and dump in as much cold water as it need.

BTW, I have tried to understand the Watt condensing engine but none of the diagrams on-line make sense to me, just what makes the steam in the main cylinder rush into the condensing cylinder to be condensed?

[chrismb]

Hmmm, my understanding of the supposed ineffeciency of the Newcomen engine is that as steam is being inducted through the entire up stroke it is during that time that premature condensation takes place.

Obviously, mine is an armchair view of the engine, as I've never worked on one. But I hope that doesn't exclude me from discussing its physics!

Well, all the same, where steam is sucked into the cylinder by the upstroke, there's no particular reason that cylinder should have dropped much below 100C - most of the energy is in the phase change, not in the temperature delta in one phase. So maybe what you are proposing might well make a critical difference to efficiency - if the cyl is maintained at 100C then the phase change is 'all in the cylinder' rather than on it. I'll go with that, the question is simply what contribution that all makes to inefficiency. Look at it another way, the 'sucking' pressure in the cylinder must be lower than atmospheric, so that would also help slow down condensation a little?

BTW, I have tried to understand the Watt condensing engine but none of the diagrams on-line make sense to me, just what makes the steam in the main cylinder rush into the condensing cylinder to be condensed?

Simply, as the steam condenses in the cold cyl, its volume drops suddenly so sucking yet more steam through, to further condense.

[Doug Coulter]

Try an obtain "Steam, it's generation and uses" by Babcock and Wilcox. This is a very thick, well written book that has more information than you'd think possible about water's phases, how to make efficient boilers and turbines (and nuclear reactors). It's "the bible" of the field, or nearly. They print a new version every few years, but not much really changes in that business, so the older ones are good.

Remember, at 99c water still has significant vapor pressure. To really get it to act like water ~20 mbar at room temp), you gotta cool it way down below the boiling point. This might figure into your efficiency calx.

I also have a couple of really cool old engineering books from the era when steam was "it" - lighting was all carbon arc (very clever rod feeder designs too) - but sadly, it's way too much to scan. Details of steam engine efficiency are worried to death in these books, with long pedantic arguments about how this or that design of valving gets the most, and how many stages are best (yes, there were multi stage piston steam engines and they actually did get better efficiency). If I can put my hands on a few relevant pages, I'll scan those.

But go ahead and buy the B&W book (used, of course, they show up often that way, but are over $100 new) - you'll be amazed how much there is to know on this topic - it's not just the physics in the usual oversimplified theories - it's the entire shebang and includes all the nasty details from all the loss sources - and there are quite a few you might not have thought of.

For quite awhile, and maybe still - the idea of a steam engine made of modern parts - say stainless steel air-hydraulic cylinders - has attracted me. Say a 3 cyl radial (guaranteed to be self starting). We have stainless and teflon - they had cast iron and horse hide piston rings. Surely we can do better than they did, and they did really well.

http://www.amazon.com/s/ref=nb_sb_ss_i_ ... Caps%2C310

Found a few used copies.

[John Hill]

Thanks guys..... I cannot claim to have any real understanding of the physics of water, heat, steam et al and my eyes tend to glaze over if I ever open a text book! OK, that probably sounds pretty bad?

Now, without getting too deep and confining myself to the basic Newcomen engine, I am thinking that a cylinder and piston made of heat insulating material and staying with Newcomen's water spray condensor would be moderately effective as an engine for amusement purposes.

My simple thoughts are that as steam and a bit of cold water goes into the cylinder and and what comes out is warm water which goes into the boiler it must also be a reasonlably efficient engine.

Of course, I understand that a pure insulator is not going to be available so my question is, what commonly available material could I use? The cylinder temperature is never going to be above 100C as the steam going in is at atmospheric pressure. I suppose I could use a metal cylinder amply lagged.

John

[Doug Coulter]

You'd probably be best off using a conventional cylinder material and insulating it. Bimba (and others) make air-hydraulic cylinders often available cheap as surplus. Stainless cylinder, sometimes aluminum ends, teflon as piston rings. Air cylinders are designed to work wet, as most shop air is wet. I made a "hydraulic ram" that used one large diameter cylinder driving a smaller diameter one, with about 2" stroke and a valving system hooked to the rods in an over-center arrangement. This let me take water with a 20' drop and push it up a 120' hill to my house. It operated many millions of strokes without wear issues, but the local raccoons kept unhooking that spring on the thing that went clack-clack near their home.