Tube to Flue Area Ratios

[hendersonsteamworks]

Hi all,

Starting with the short story, I have a Rob Roy that steams badly, lots of looking around the net and stuffing my face in books has been undertaken and I see that the more recent discussions here were coming on for 9 years ago.
I understand that hard data doesn't age, but opinions and experience accrue, and thus I felt it acceptable to drag up this old topic.

Without getting into figures immediately...

Does anybody have any experience of upping tube numbers in boilers such as a Rob Roy?



With that over with, here's some of my data.
Current boiler was built by the previous owner / builder and deviated from the drawings. Flues went up in diameter and fire tubes went down, though up in number. 3/4" flues substituted for 22mm and 7 x 3/8" fire tubes became 9 x 8mm with a 5.6mm bore. Highly prone to blocking.

Free area ratio of tubes to flues with superheater considered is 0.85:1 for a standard boiler, Flues 298mm2 / tubes 351mm2
My non standard boiler swings the opposite way dramatically at 2.17:1, Flues 551mm2 / tubes 237mm2
The most telling ratio of my difficulties is the single firetube to flue open area ratio - Standard is 2.97:1 where current is 9.75:1 !!! As good as leaving the front door open with the heating on.

So I say to myself no wonder it steams so poorly, the heat really is going straight out the chimney!

Other supporting evidence of this is that a DIY extended steel venturi chimney has turned a dark blue at its base, some 350 degrees C, which is as I'm sure you'll agree far from ideal. Club members at SwanseaSME have suggested 180 should be more normal range.

Based on these figures I have constructed a spreadhseet to asses the performance of a boiler constructed with a greater number of tubes, a single superheater flue and a double pass radiant superheater.
With some trickery I'll skip for now I have managed to get 13 x  3/8" firetubes and 1 x 3/4" flue to fit into a Rob Roy firebox tubeplate and the ratios suddenly become a lot more favourable.
All tubes to flues ratio drops to 0.09:1 and single firetube to flue ratio is near par at 1.17:1
I gain 85% in firetube open area, 9.5% total open area and this is a bit more fuzzy due to book figures on firebox area, but have gained in the region of 34% in total heating area.

My second question therefore is - Can somebody tell me if this has any downsides?

Going through the Baguley spreadsheet it would suggest that my new arrangement has a good engine factor Ee of 0.166 (all factors fixed at design), but it puts boiler factor at Eb 36.72 (80 ideal) and Overall factor at Eo 6.09 (12 ideal).
Footnotes would suggest that this would be a good steamer though may suffer with fire lifting, but a large blast nozzle ( already in stock from prior experimentation) and a rosebud grate (Same applies) would I imagine offset most, if not all of these ill effects.

Opinions please!

I have already started manufacture of said boiler, so I do hope you all mostly agree that it won't be a complete disaster.

[millman]

Rather than mess around with calculations why don’t you make a Rob Roy boiler to the correct Martin Evans design, thousands of Rob Roy’s have been made and the boiler as designed works OK.

[hendersonsteamworks]

Feb 13, 2023 13:07:31 GMT millman said:

Rather than mess around with calculations why don’t you make a Rob Roy boiler to the correct Martin Evans design, thousands of Rob Roy’s have been made and the boiler as designed works OK.

Granted, they work well enough, and in some respects that would still be a worthwhile experiment as presumably I would be left feeling a degree of vindication that my suspicions were correct. However, where would the fun be in that?

I admit, I'm possibly the rarer kind of model engineering breed. I'm not patient enough for fine scale or rivet perfect prototype construction, nor blind drawing following, fun as that may be; what interests me more is the attainment and understanding of efficient steam generation and mastery of a tiny fire, shovel in hand, soot in the eye kind of stuff (and satisfying my inner egotistical maniac's desire to trounce all before me!).

Sure, Stephenson could have just run a return flue boiler, Watt put his feet up after repairing a model Newcomen, and while I hasten to make clear air between myself and the heroic engineering greats of the past, I think we have that very human trait of innate pursuit of improvement in common.

I might add that my background includes a good dose of motorsport, where great is always sought but never attained.

[chris vine]

I am not fully understanding your ratios.
Are you saying that there is too much, or at least more than normal, free area through the superheater flues compared to the normal tubes?

If so, as an experiment, can you block off some/all the superheater flues (bung up with something), and try steaming the engine. You won't have any superheat, but you might get more steam. It might tell you something...
If it looked like a good avenue for experiment, then you could make little baffles to restrict the gas flow through the superheater flues.

Chris.

[hendersonsteamworks]

Hi Chris, tried my best to be clear. sorry if I failed!

Yes, current "not built to the drawings" boiler has a vast flue to firetube ratio, each flue has 9.75 times more open area than a firetube, where a standard Rob Roy is far less at 2.97 time greater. My alteration would make that nearly imperceptible a 1.2:1, the fire would effectively only "see" 14 firetubes (as opposed to 13 firetube and 1 flue), and not take any easy routes through the boiler barrel.

It's equivalent to a parallel electrical circuit, if one resistance is much lower than the rest, the current will choose that path. In balancing these out fairly closely I would hope that maximum heat ought to be extracted.



Yes, blocking them up would tip the balance, the tiny firetubes I fear would just block up with coal particles immediately!

[flyingfox]

Greetings, interesting discussion. Over the course of the years, I have built 2 Robroys, and experienced locos that were not easy to steam.
Apart from the tube discussion as above, I am convinced that the ashpan was not allowing enough air into the fire, so eventually, I cut large parts of it away, leaving enough to catch ash, and support the grate, which made some improvement. Opening up the firebar spacing also gave a noticeable improvement, and the final thing I wanted to do, but never did, was to have a single superheater flue tube, with a radiant superheater pipe, all other tube sizes remaining as the design.

I am convinced that Robroy, built as suggested in my text above would have been a good performer, and knowing Martin Evans, discussed it with him, and he thought so too.

I would be interested to know about the outcome of your rebuilt boiler.

Regards

Brian B

[hendersonsteamworks]

Thanks Brian, great to hear that you've had similar thoughts, and even spoken to Mr Evans, amazing!

I'll keep you all updated. (Tried adding a picture of only 168mb, but refused to upload)

[44767]

I think the first thing to get correct is the diameter of the fire tubes versus their length; there is a fluid flow law which must be obeyed! To use the electrical analogy, if the wire diameter is decreased, the resistance gets larger. If you say the tubes are smaller than designed, that is probably the main problem. This is the reason we cannot have the scale number of tubes in a scaled down boiler. Once you have that correct, the next thing to get right is the ratio of grate area to the tube area and then the blast nozzle diameter relative to a cylinder diameter. Draughting in the petticoat area is important too but the first thing is the fire tube diameter, ignoring the flues.

Others on here know what the figures should be.

Cheers,Mike

[hendersonsteamworks]

Hopefully this'll work...


New phone, new camera to learn, apologies for the slight out of focus edges!
My first ever boiler plate, quite pleased with the outcome so far. Firebox backplate in production and hammer forms for backhead and smokebox tubeplate in the planning stages.

[GWRdriver]

I don't believe I've er heard a discussion where a distinction was made between a fire tube and a flue. To me those are one in the same and as Mike says are, or should be, governed by a formula.

The formula for flue/fire tube diameter to length ratio is d² = L/65
where d = the inside diameter of the flue (in Inches.)
L = overall heated length of the flue (in Inches.)
and 65 is an average numerical value between 60 and 70 which will result in an optimal balance between free flow of combustion gas and heat absorption through the flue tube walls. For our purposes, any value between 60 and 70 will give a near enough optimal result.

As has been mentioned, the next relationship is grate area to flue area, and after that it's the "Front End." In my experience Front End problems (air leaks, poor draughting, etc) are the cause of more steaming problems than anything else.

BTW HSW, that's a very well done job on your first plate.

[jma1009]

I think that the 'Rob Roy' design can be improved. I would be inclined to consider all aspects that affect steaming characteristics of a loco such as this, as others have already wisely commented upon.

I believe there is no suspension offset on the launch type links in the valve gear. A simple correction to make that will vastly improve the valve events and 'notching up'.

I definitely did a re-draughting of the smokebox innards via email with one of our members in February 2020. The petticoat pipe 'choke' is rather big, and the distance between it and the top of the blastpipe far too long.

A cramped ashpan is always going to present a challenge, and it will need to be kept clear of ash especially at the sides close to the underside of the grate.

[hendersonsteamworks]

Feb 14, 2023 15:50:07 GMT GWRdriver said:

I don't believe I've er heard a discussion where a distinction was made between a fire tube and a flue. To me those are one in the same and as Mike says are, or should be, governed by a formula.

The formula for flue/fire tube diameter to length ratio is d² = L/65
where d = the inside diameter of the flue (in Inches.)
L = overall heated length of the flue (in Inches.)
and 65 is an average numerical value between 60 and 70 which will result in an optimal balance between free flow of combustion gas and heat absorption through the flue tube walls. For our purposes, any value between 60 and 70 will give a near enough optimal result.

As has been mentioned, the next relationship is grate area to flue area, and after that it's the "Front End." In my experience Front End problems (air leaks, poor draughting, etc) are the cause of more steaming problems than anything else.

BTW HSW, that's a very well done job on your first plate.

Thanks GWR Driver.

With 3/8" tube giving a 5/16" bore I've got / the std Robb Roy has a tube factor of 73, so I'm fairly happy that that's good enough and not worth attempting to refine - Would I go up or down anyway? 

Thanks for the boiler plate compliment, I hope I can do similar justice to the rest of the plates. Updates will follow.

[hendersonsteamworks]

Feb 15, 2023 13:03:24 GMT jma1009 said:

I think that the 'Rob Roy' design can be improved. I would be inclined to consider all aspects that affect steaming characteristics of a loco such as this, as others have already wisely commented upon.

I believe there is no suspension offset on the launch type links in the valve gear. A simple correction to make that will vastly improve the valve events and 'notching up'.

I definitely did a re-draughting of the smokebox innards via email with one of our members in February 2020. The petticoat pipe 'choke' is rather big, and the distance between it and the top of the blastpipe far too long.

A cramped ashpan is always going to present a challenge, and it will need to be kept clear of ash especially at the sides close to the underside of the grate.

This is where my research first started, as the front end left quite a lot to be desired.

I made a new petticoat pipe and messed around with the blast nozzle and smokebox sealing. The superheater had a mild reshaping to clear the petticoat pipe. Months later and I discover the exhaust Tee was barely hanging together by a thread. I then fully rebuilt that with my own swept Tee and specially shaped nozzles and on the bench was able to draw a vacuum in the opposite branch with an exhaust blast, so that aspect is working really well - Infact far far too well for the boiler now, because I can set it to suck the fire out and block the tubes in half a lap!
The chimney can be shaped to improve efficiency and I've also tried longer venturi chimneys with great effect too, once again to the degree that it works too well, and a much larger blast nozzle can be substituted.

Short story is, Rob Roy front end can be improved upon massively and the 1in3, 1in6 rule for petticoat pipe setting, as well as blast nozzle diameter rules of thumb are way off the mark.

My valve gear is worn out, and some work in this department may help a little too. There is a sweet spot when screwed back a few turns. I will have to force myself to address these later this year when the old boiler comes off and there's a bit more space to work.

[GWRdriver]

Feb 17, 2023 10:51:57 GMT hendersonsteamworks said:

With 3/8" tube giving a 5/16" bore I've got / the std Robb Roy has a tube factor of 73, so I'm fairly happy that that's good enough and not worth attempting to refine

I quoited that constant from memory (which I haven't needed to pull up for a donkey's age), so correcting myself, Harris and Keillor quote 60-80 for the constant so you're in good shape in either range.