The Double Belpaire firebox.

[7j43k]

The Double Belpaire firebox has been talked about, occasionally, for a goodly many years (about 70).

When I first read about it, it was in an article in Trains magazine, discussing the final un-built steam locomotive designs of Lima.  And I still do read about it:  most recently in Withuhn's new book, "American Steam Locomotives, Design and Development, 1880-1960".

But I have recently run into a "road block" in my understanding of it.

Withuhn says: "It had the traditional Belpaire bulge at the top, but it was mirrored by the novelty of a bulge at the bottom, resulting in more tubes and flues, and thus, more heating surface."

That sentiment has been voiced a number of times, elsewhere.  I have also heard mention of greater superheat surface area.

If we look at what appears to be the announcement of the concept by its designer:

http://www.rypn.org/forums/viewtopic.php?f=1&t=36360

we see a drawing and a photograph that shows the rear tube sheet.  It does, indeed, show more tubes and flues.

 

 

Here's my problem:

The rear tube sheet for a steam locomotive is placed just at the front of the firebox or, as in this case, the front of the combustion chamber.  The tubes and flues go forward from this sheet.  But there is no place for the tubes/flues to go forward from the lower corners of the Double Belpaire--they will immediately run into the transition from the Double Belpaire to the typically round cross-section of the rest of the boiler.

Or so it seems to me.

Thus there would be NO increase in the number of tubes and flues.  And this would make the photograph in Figure 5 of the article quite suspicious.  Or misleading.

If I am in error, I do hope I will be corrected.

 

On a more encouraging note, it is easy to see that there is an increase in surface area for the firebox/combustion chamber.  But.  How much?

For a ballpark approximation, I started with a grate area of 100 sqft, and a combustion chamber length of 6 feet*.  And did a lot of rounding.  And approximating.  

I came up with 656 sqft of surface area for the Double Belpaire.  And 633 sqft for a single Belpaire.  That's a difference of 23 sqft, or an increase of 3.5%.  One could get the same increase by adding 8" to the length of the combustion chamber of the regular Belpaire.  Decisions, decisions.

 

Ed

 

*also:  

width of inside of firebox = 9'

height of inside of firebox = 11'

length of inside of firebox = 11'

 

[Overmod]

7j43k

Withuhn says: "It had the traditional Belpaire bulge at the top, but it was mirrored by the novelty of a bulge at the bottom, resulting in more tubes and flues, and thus, more heating surface."

Withuhn has said a great many things, not all of which turn out to be strictly correct with respect to technical matters.  Perhaps the great 'nominal' improvement of the double Belpaire is in the increase of radiant-uptake area in the combustion chamber combined with the (prospective) better nucleate flow from steam generation in that geometry.  But there are improvements in flow due to geometric concerns (see more below) that account for more advantage in the convection section itself.

I have also heard mention of greater superheat surface area.

You know better.  The flues for superheat are elsewhere in the boiler shell.  What the double-Belpaire does here is optimize gas flow in the lower portion of the tube nest sufficiently that a larger number of flues (which are generally considered to have less convective steam-generation capacity compared to an equivalent number of tubes in the same location) can be used to produce the same convection-section steam generation ... and more or better elements then accommodated in those more/larger flues, and more relative gas flow directed through them in proportioning induced draft characteristics.

If we look at what appears to be the announcement of the concept by its designer:

http://www.rypn.org/forums/viewtopic.php?f=1&t=36360

we see a drawing and a photograph that shows the rear tube sheet. It does, indeed, show more tubes and flues.

The first key to this situation is to look at the little dotted line that shows the shell ID -- it's important to consider the geometry of the rear tubesheet in comparison with this datum. 

The second key is to read the accompanying discussion Col. Townsend provides, while looking at the different configuration of tubes and flues in the designs as drawn.  Now, the sections are drawn to show the chamber region, but (as you noted) the tubes continue straight into the dotted-line-indicated shell, but in different proportion (in part because of the nominally-improved gas flow action down near the bottom in the double Belpaire chamber). 

If you want to backstop Townsend's analysis you could tot up the flow through the different arrangement of tubes and flues and see whether this approximates the percent improvements he obtained (probably with much more intricate and empirically-enhanced analysis that Lima could conduct and pay for at the time!) but I see little reason to believe things are not as the comparative diagram implies.

[Paul Milenkovic]

David Wardale in "Red Devil" disrespects combustion chambers, claiming the South African Railway tests were inconclusive, the great German steam locomotive designer Wagner didn't believe in them, and the Chinese disregarding advice from the Russians on the futility of combustion chambers were engaging in hopeful thinking.

Any thoughts as to why Wardale could be wrong?  Along with the source he quoted?

[Overmod]

Paul Milenkovic

Any thoughts as to why Wardale could be wrong? Along with the source he quoted?

It is not so much that he is 'wrong' as that scale and firing efficiency are important.  As part of your education, review the Le Massena article on Niagaras that was published in Trains in the 1980s; this contains some valuable information (both negative and positive) on chambers and their sizing.

I am a firm believer in the value of good chambers in large locomotives, particularly when run with large amounts of induced draft at high speed.  Remember that radiant uptake is far more effective with the power spectrum from 'luminous flame' than from relatively transparent gas, and keep the time of flight of evolved particles from practical combustion, from the firebed up through the elongated passage created by the arch, in mind.

Wardale of course thought in terms of GPCS, usually in terms of reducing the amount of PM actually carried in suspension in the firebox radiant volume (substituting gas production and late oxidation/combustion, for example of CO to CO2, well above the bed in good secondary air.  To him, a BTU produced that way was more efficient than one from 'soot' particles levitated in what would become the exhaust.  And he may well be right -- just that it's highly difficult in the first place to run true GPCS proportionally in a locomotive plant at high but variable power.

In the meantime, consider the Big Boy, which explicitly burns much of its friable and low-grade fuel fully levitated, almost as if it were PC co-fired.  Each combusting particle has a nice, distributed hot-carbon emission profile, close to blackbody emission spectrum, with the surface area decrementing as the carbon burns and the CO/CO2 is scrubbed off, to the point there is no more to burn and you have a hot ash particle and no soot.  Consider the time (in the generally reducing and sub-atmospheric-pressure gas flow in an induced-draft locomotive boiler setup with limited secondary air) that these particles take to burn down that way, and consider how beneficial the longer time granted by a good combustion chamber is.  (What you want, theoretically, is for the luminous flame to go fully to the point that only 'afterheat' is radiant as the gas goes into the tubes, with no carbon 'wasted' after that point)

Now, most conventional locomotives when worked hard, let alone to near their grate limit, are observed to throw considerable unburnt fuel, let alone produce prodigious smoke, so there's evidence of nominally incomplete combustion.  That in itself is of comparatively low importance (although railroad officials tasked with the fuel bill might splutter at such an observation!) as the practical uses of incomplete combustion outweigh the theoretical advantages of operating at stoich or its equivalent at all times.

There are also considerations of gas distribution across the rear tubeplate, which factor largely into the analysis of the double Belpaire itself.  Some consider this the greatest 'advantage' of chambers, noting that the term 'combustion chamber' itself is a misnomer -- that additional 'combustion' contribution is almost vanishingly slight even aggregated across a large mass flow of particles due to the short TOF in traversing the chamber length.

In my possibly incomplete and certainly biased opinion, Wagner didn't even appreciate the contribution of radiant uptake to steam generation, let alone what the actual function of a chamber in contributing to radiant uptake might have been.  Certainly he didn't even come close to designing a boiler capable of functioning adequately on what was a relatively small 4-8-4.  So I'd dismiss him fairly briskly as any sort of authority on what works with North American power combustion, although he is correct about proportioning effective water-tube length.

In short if you are not anticipating good GPCS-optimized combustion 'and all that implies' in locomotive design, I think you're more likely than not to see benefit from a proper chamber on a big locomotive.

[7j43k]

Thanks to Overmod's prompting, I re-examined the drawings, and now see how things "fit together".  It is difficult to navigate the 3-D space of a locomotive boiler all in my head (which is already pretty full of something--don't know what, though).

So my objections to the layout for the Double Belpaire have been resolved.

The writer of the article does mention deflection (see Figure 7).  That might be a problem for the bottom part of the Double Belpaire.  In the upper part, there are cross-ties across the width of the boiler.  There's no room for them at the bottom.  He said the model didn't deflect.  I wonder how a full-size would have reacted.  

 

Ed

[Overmod]

7j43k

In the upper part, there are cross-ties across the width of the boiler.  There's no room for them at the bottom.  He said the model didn't deflect.  I wonder how a full-size would have reacted.  

At least part of this is an artifact that I think you'd appreciate better with a phantom 3D view or even isometric of the arrangement.

Remember that the bottom "bumps" of the double Belpaire aren't there to do what the upper ones do -- look at their construction in the cross-section and note that there's actually partial vacuum, not pressure, in the space inward from the double wall of their construction.  The actual 'force' involved in any pressure deflection of the bottom bumps is largely taken by the chamber-like staybolting between inner and outer sheets (and I leave it to the reader to read the 'rear boiler knowledge' article that was printed in German in Glasers Annalen post-WWII and translated by Wasatch Railroad Contractors and determine the extent to which these could be welded) and I think any net 'deflecting' forces of the actual bump overall would be of minimal significance as all the compensating forces would be taken in the plate and stay alignments, to a smaller extent as they are elsewhere in the firebox waterspaces under thermal variation.

[sgriggs]

Paul Milenkovic

David Wardale in "Red Devil" disrespects combustion chambers, claiming the South African Railway tests were inconclusive, the great German steam locomotive designer Wagner didn't believe in them, and the Chinese disregarding advice from the Russians on the futility of combustion chambers were engaging in hopeful thinking.

Any thoughts as to why Wardale could be wrong?  Along with the source he quoted?

 

 

If there is any doubt about the value of a combustion chamber, all one has to do is compare the horsepower difference between a J1 and J3 Hudson.  The J3 had 7% less evaporative heating surface, 10% less superheater surface area, and 8% less Net Gas Area through the tubes and flues, but produced 20% more power than the J1.  Now, I grant you that some of that increase was due to the higher boiler pressure of the J3, but the J3 also had a 43" combustion chamber whereas the J1 did not have any combustion chamber.

One possible explanation for South African and European experience with combustion chambers being different than American experience might be related to firing rate, grate size, and combustion gas velocities.  The larger American types employing combustion chambers tended to have much larger grate areas than European (and I suspect South African narrow gauge) locomotives, but the net gas area through the tubes and flues was not proportionally larger.  I would expect combustion gas velocities in large American locomotives to be significantly higher because of this.  The American locomotives were also invariably stoker-fired, while the European and South African engines were probably a mix of stoker- and hand-fired (maybe predominantly hand-fired?).  At the higher power outputs and smaller coal particle size present with stoker firing, I would argue that the benefits of a combustion chamber (increased direct heat surface area and incrementally longer combustion time) would be more apparent.  It's not clear to me if the locomotives used to evaluate the effectiveness of combustion chambers in Europe and South Africa would behave in the same way.