Optimal number of boiler tubes.

[Raygers]

Too clever for me Richard... I can follow your train of thought, but you simply pose better questions and answers than my brain! - Thanks!
Back to the number and size of Flues in Rayger's Boiler.
is it still to be 55 off 7/16" flues?
Or the scheme Martin proposed?
Maybe I missed something along the way, when I wasn't here....?
K2

I think I'm sticking with the original plans, 55 7/16" flues. The only deviation, at the moment, is to use an oval firebox door after reading of failures with rectangular openings. I do like the idea of a superheater whether smokebox or firebox, I didn't quite understand tenor's description of his proposed firebox "I first tried 4No. x 1" superheater flues with a 3/8" diameter superheater with a single spear - i.e. "out and back" as far as the firebox tubeplate."

Ok I went for a nap, I'm old so allowed, and it came to me in a dream what tenor meant. Replace a number of tubes with a 1" one, run the steam pipe from the top down the 1" tube, run around the firebox a few times and then back up the tube to the steam supply.
Do I have this right? I don't do well with words, I'm more of a blueprint/picture/visual guy.

One caveat though, the melting point of copper is 1083.4 whereas tenor's calculations predict firebox temperatures are 1265.3, would this not be a problem/

​

​

​

 

[Steamchick]

Hi Raygers.
To address your points individually.

• "I went for a nap, I'm old so allowed," - Join the club. Random Napping is a part of the joy of retirement, I think?
• "and it came to me in a dream what tenor meant." Take more naps! - Brilliant. You are correct - ish.
• Replace a number of tubes with a 1" one, run the steam pipe from the top down the 1" tube," - CORRECT.
• "run around the firebox a few times" No, just take the down-tube to a "spear-point" sort of joint with the return tube and then back up the tube to the steam supply. This Spear-point can project maybe an inch or so into the firebox. Here is a picture of spear-point superheaters for a horizontal (locomotive) boiler, but will show you better than a wordy explanation. Imagine (in your dreams?)the spear end projecting into the firebox, not to prevent the coal and shovel, but maybe to the sides and back away from the fire-door, so you can see the spear-ends and still get coal all around the fire-box. (or a gas burner beneath the spear-ends).

• Do I have this right? - Yes, mostly. "I don't do well with words" - You berate yourself. As clear as everyone else, "I'm more of a blueprint/picture/visual guy". - I think we all are really... 50 years of Engineering profession and thousands of small sketches - on whatever was at hand. Sometimes just a stick on a rusty /dirty surface was all there was, but enough! = More usually, a pen on paper - or *** packet! Nowadays a phone with camera?
• "One caveat though, the melting point of copper is 1083.4 whereas tenor's calculations predict firebox temperatures are 1265.3, would this not be a problem?" - OK, a good question, the answer (not something from a dream) is "no". What happens is that the gas closest to the metal forms a relatively insulating protection where the heat is flowing into copper with "cooler" water at the other side of the copper tube wall. Here's a picture of temperatures in a firebox, at the wrapper/water jacket. Personally, I would have a thick insulation not "20.C. atmosphere" around the outside. But you'll get the gist.

You can see from the "graphical line" that the temperature adjacent to the copper surface drops very rapidly from 1000C to metal temperature. This is only "schematic", but close to the truth. It is because of the high "heat" needed to melt copper, as well as the "high conductivity" that carries the heat very quickly to the wet-side, that makes it work so well. Aluminium has similar properties, but is rapidly corroded by steam at steam temperatures, so is no good for our boilers. (but OK in air conditioning systems with refrigerant gases).
Finally, a picture of a section through a fire-hole door - as recommended in "the books". - Sorry it is not "CAD"!

Turn a copper ring with a step on either end, compress the "top and bottom" (in a vice?) to make it oval, than form it around a solid 5in dia former to make it fit the dished holes and cylindrical shape of the dished hole in the firebox wrapper and Shell outer. - I think you get my meaning...
Not easy copper forging! Anneal frequently while forging the copper to shape. (it hardens considerably at each bend).
Hope this helps?
(That was much easier to explain the Quantum mechanics!).
K2

 

[Steamchick]

While considering the strength of the shell under internal pressure, and Rayger's fire-box wrapper under compression, with the fire-door aperture, I have been reading various papers on the web...
One finite element study explains how a thick-walled tube penetrated for a side-tube has a maximum stress concentration inside the material of the side connection.
Further study - because I find it of interest... - leads me to a thesis written in 1975 by a PhD student at the University of Virginia:
As a part of his dissertation he quotes ASME of that time:
"... - the ASME Pressure Vessel Code requires in a stress or fatigue analysis the stress concentration factor (SCF) to be not less than 3.3 for a "well designed penetration" in a cylindrical shell unless positive evidence is available to the contrary. This evidence usually means a separate analysis to confirm the peak SCF."
While I have not (yet) found this in ASME, it crops up in various other quotations, so I presume it is still applicable today.
By "positive evidence to the contrary" - I doubt that my scribblings and estimates of stress concentration factors would be considered by any US lawyer, so I should revert to this SCF =3.3 in any designs I study.
Unfortunately, it "Kills" most of the designs I study for model boilers, where there are side penetrations in such things as Boiler shells and fire-tubes.
The inference that I have found mentioned refers to "fatigue" failures at stress concentrations.
One loco boiler that I know had failed - while being used to haul folk - simply could not maintain steam pressure, and extra steam appeared to be coming from the smoke stack. The failure was a tube to firebox plate joint, and not due to a low water level (with consequential overheating) but simply put down to the 20-plus years of satisfactory steaming - I.E. "Fatigue at the stress concentration of a tube in compression, where it was silver soldered into a hole in a plate". NOT the same scenario as the tube being penetrated and the major tube failing, but the minor tube failing in compression, with stress concentration at the silvered soldered joint where it was rigidly supported. Similarly, I dismantled a repeatedly repaired boiler fire tube with cross-tubes, where the cross-tube to firetube joints had repeatedly failed - I think due to the stress concentration not covered in Boiler design books? It does appear to me that many "model designs" fail to appreciate such stress concentrations, even though catastrophic failures do not occur.
I am particularly concerned with Rayger's boiler that there are 8 penetrations which significantly reduce the axial strength of the shell in tension, as well as the hoop stress. In addition, the fire hole has a major effect of increasing stress in both the shell tube and firebox wrapper, requiring a stress concentration factor that may not have been considered by the original designer. I have also repaired and de-rated a boiler with a firebox aperture that had failed in service at the firebox joints,
Tenor is assisting me in my deliberations (with British Standard advice), but other experience will also be welcome, particularly of how ASME analyses these stress concentrations? - And how Raygers should consider these deliberations in Canada? (I know nothing of "managing" model boilers there).
Thanks,
K2

 

[Steamchick]

Hi Raygers,
I have not forgotten your original question, or how it has developed...
But I have just found this Regulation, which may or may not be applicable to where you are?
https://www.ontario.ca/laws/regulation/010220Download it and see if is relevant - or not?
Have you discussed any of this with your local club?
Most Model Engineering clubs have a lot of "like to think they are Experts" and a few real experts, - who are the only ones anyone really listens to. But they are a valuable source of information, and in the North-East of England, many people join the clubs specifically for their boiler and testing expertise. - I did.
K2

 

[Steamchick]

Raygers,
I shall draw your attention to the Ontario Regulation:
- see the last bit - point 8... - But don't ignore what has gone before about a registered design, changes to design, etc.... Your local club should have experts to advise you better than I.
Cheers,
K2

Design registration requirement​

4. (1) Subject to subsection (2), no person shall manufacture a boiler, pressure vessel, fitting or piping for use in Ontario unless its design is registered with the director. O. Reg. 220/01, s. 4 (1).

(2) A person who submits a design submission for registration may commence construction of the boiler, pressure vessel, fitting or piping before the submission is registered if the person assumes all risks related to the construction, whether for an installation or alteration. O. Reg. 220/01, s. 4 (2).

(3) The design for a boiler or pressure vessel shall bear the signature and seal of a professional engineer who is experienced in the design of boilers, pressure vessels, piping or fittings. O. Reg. 220/01, s. 4 (3).

(4) Where the designer, manufacturer, installer or owner of a boiler, pressure vessel, fitting or piping proposes a change to its registered design, as determined in accordance with the code adoption document, they shall submit the design and specifications of the change to the director and obtain registration before beginning to make the change. O. Reg. 220/01, s. 4 (4).

(5) If an inspector finds, after its manufacture or installation, that a boiler, pressure vessel, fitting or piping for which a design registration has been issued is defective, the inspector may, despite the fact that the registration has been issued, permit the boiler, pressure vessel, fitting or piping to be operated or used within such limits of safety as the inspector considers adequate in the circumstances and shall require the manufacturer or installer to correct the defects within such period as the inspector may allow. O. Reg. 220/01, s. 4 (5).

(6) If the defects found under subsection (5) are due to the design and specifications of the boiler, pressure vessel, fitting or piping and, in the director’s opinion, they cannot be remedied, the director shall cancel the design registration, and no additional boiler, pressure vessel, fitting or piping shall be manufactured or installed based on that design. O. Reg. 220/01, s. 4 (6).

(7) Where a boiler, pressure vessel, piping or fitting has not been manufactured or installed in conformity with its registered design but nevertheless may be used safely at a lower pressure than its design pressure, the inspector shall fix its maximum allowable working pressure having regard to its design, condition and installation and the purpose for which it is to be operated or used. O. Reg. 220/01, s. 4 (7).

(8) Where an unused boiler or pressure vessel has been manufactured and its design and specifications have not been registered, the director may cause it to be inspected and, if satisfied that it may be operated or used safely, may issue a certificate of inspection for it as a used boiler or pressure vessel. O. Reg. 220/01, s. 4 (8)."

 

[Steamchick]

Next bit of Canadian regulations stuff:

Boiler and pressure vessels directive​

This says that Raygers' boiler is simply too small for this regulation:

1.1.2 This standard does not apply to:

(b) a pressure vessel that has a capacity of 40 L (1‑1/2 cubic feet) or less;

(d) a pressure vessel that has an internal diameter of 150 mm (6 inches) or less;

BUT I suggest you browse this search as I don't know whereabouts you are in Canada, and there seem to be regional regulations e.g. for Nova Scotia, Ontario, Prince Edward Island, etc... as well as National regs.

Enjoy!
K2

 

[HMEL]

I think I'm sticking with the original plans, 55 7/16" flues. The only deviation, at the moment, is to use an oval firebox door after reading of failures with rectangular openings. I do like the idea of a superheater whether smokebox or firebox, I didn't quite understand tenor's description of his proposed firebox "I first tried 4No. x 1" superheater flues with a 3/8" diameter superheater with a single spear - i.e. "out and back" as far as the firebox tubeplate."

Ok I went for a nap, I'm old so allowed, and it came to me in a dream what tenor meant. Replace a number of tubes with a 1" one, run the steam pipe from the top down the 1" tube, run around the firebox a few times and then back up the tube to the steam supply.
Do I have this right? I don't do well with words, I'm more of a blueprint/picture/visual guy.

One caveat though, the melting point of copper is 1083.4 whereas tenor's calculations predict firebox temperatures are 1265.3, would this not be a problem/

​

​

​

I think it would be wise to stay with one diameter tube for all. The reason is the draft will change because the tube will have different pressure drops for the air flow through them. This may cause the fuel to burn differently depending on where they are located in the tube sheet. And the gas velocities will change in the tubes which will affect heat transfer. There is a rule on how close you can space the penetrations in the tube sheet and you want to optimize both the heat transfer area with the hot gas flow through the tubes. Using the original is probably a very good solution.

 

[Steamchick]

HMEL, the main reason (I think?) Martin had 2 sizes of tube was to enable the fitting of "spear-ended" down and back superheater tubes. This significantly improves the total heat efficiency of the boiler, as although there is a bit less steam available, it is superheated, so the heat reaching the engine on full power is increased. Also, the hotter and dry steam carries less/no free water when it gets to the engine, so less water at the exhaust. - which also means less water is needed to maintain the boiler water level, and less "lost power" to the water pump (if "Boiler powered"). You have raised an interesting question though about how the air-flow through the tubes would affect the fire. I do not know, but I could hazard a guess at anything between "significantly" and "not a lot"... as the last coal fire I managed was about 50 years ago as a teenager at home.
But if gas fired, the air-flow mixing with gas is mostly determined by the burner internal configuration before the flame holes.... so for gas firing it will be "no problem" in my experience.
An interesting question though. I wonder from a "stressing" perspective how best the mix of large and small tubes should be best arranged? - If insignificant, is there a better location from the perspective of the gas temperature entering the tubes so for superheater tubes, you may want to capture exhaust gas from the hottest zone of the fire-box combustion zone? - Which (I GUESS) is in the middle of the fire? (Away from the cooler walls of the firebox).
Perhaps Tenor can comment?
Interesting.
K2

 

[user 52078]

Steamchick,
My program is not sufficiently advanced to look at different flows and differing temperatures down different tubes or flues - that would require seriously good CFD software. So the implicit assumption it is the same across all tubes, and a different value across all superheater flues. That is sufficient for the purpose the program was designed, which is for outline design of a boiler to look at the many different variables (tube size, number, superheater geometry, firebox size, etc. etc.) and their effect on draught requirement and steam generation.

My own thinking is that provided the tube bank gives a reasonable resistance (draught), the flow across the bank will be fairly even, hence flow through the firebed will be fairly even. In fact, putting flow through a whole load of parallel resistances such as honeycomb or mesh is a classic way of producing smooth flows in applications like wind tunnels.

HMEL
I have found that balancing flows through superheater flues and firetubes is quite important to the overall boiler performance. You are right in that if you get it wrong you will either:

• Get all the gas flow going through the firetubes, while the superheater flues are dead and do very little superheating. (plenty of steam, very little superheat)

or:

• Get all the gas flow going through the superheaters, while the firetubes are dead and not evaporating much water. (great superheat, just not enough steam)

That's a bit of a simplification, but I hope it shows the problem.

The superheater flues have superheater elements in them, so that blocks part of the flow and the extra surface area pushes up the friction. So my program puts the numbers to that conundrum. If you look at the summary spreadsheet I did back in post 33 (page 2 of this thread), you will see that I work out the draught through both firetubes and superheaters (known as max & min draught) and also the flue gas exit temperature from both firetubes and superheaters. You will see that for both the superheater options I propose, the draught is very similar and the spread between exit temperatures is reasonable (about 30 deg. C) showing that there is a reasonable balance between the two. You can optimise that balance by changing the superheater flue diameter, superheater element size and number of elements (assuming a constant firetube set up). The degree of matching is limited by working with commercially available tube sizes.

Hope that helps,
A Peaceful and Happy Christmas to all,

Martin

 

[Steamchick]

Thanks Martin, I knew you could explain it much better than I...
I understand "the implicit assumption it is the same across all tubes". That is what I expected. But in my "imagination" (where anything is possible) I figured that the heat losses by conduction to the firebox wrapper - water-cooled - would cause a temperature variation between combustion gases towards the middle of the firebox being hotter than those feeding the outer ring of flue tubes. - Assuming it is not so turbulent in the fire to cause complete mixing and a relatively uniform temperature - which is the necessary simplification for your programme. NOT a criticism of your programme, just an idea that maybe the best flues (hottest gases from the coal fire) for superheating would be in the middle of the array, not at the outside, for a vertical boiler of this nature.
Just revisiting the spreadsheet from post #33, it says a lot about the advantages of the superheater array you propose.

• The Average smokebox temperature is at its highest, suggesting least restriction to the flue gases ( => lower back-pressure in the firebox preventing good draughting).
• Steam temperature at the engine is highest (201deg.C) - best for engine performance, and early cut-off efficiency.
• Lowest evaporation rate - which means less demand on the feed pump.

So in conclusion : based on Martin's calcs. and my postulation as to the "optimum location for superheater flue-tubes", I would suggest a central square of the 4 superheater flues, surrounded by the 18 off 1/2in tubes.
I assume the gas flow in a loco horizontal boiler to be quite different, possibly why superheaters tend to be placed towards the top of the array? - But that may be just "packaging?
An excellent response from you anyway.
Thankyou.
K2

 

[Steamchick]

Both 1" flue tube should be 12 gauge and 1/2" tubes should be 16gauge (or thicker) if you can get it.
K2

 

[Raygers]

Both 1" flue tube should be 12 gauge and 1/2" tubes should be 16gauge (or thicker) if you can get it.
K2

That may be difficult to get if at all, the thickest I've seen is 16 gauge 1.6mm. I have a book by Alex Wiess "Making Small Gas Fired Boilers for Steam Models" in which he uses 1.6mm copper for everything. I can get 10ga for the main boiler tube and the firebox as well as the tube plates.

 

[Steamchick]

Sorry Raygers,
in my haste I input NWP =100psi to my spreadsheet... - copied across from an earlier calculation.
MY ERROR. = Stupid me!
You are planning 60psi NWP if I remember correctly?
So 16 SWG (0.064in. thick wall) tube is OK for that pressure.
K2

 

[user 52078]

Both 1" flue tube should be 12 gauge and 1/2" tubes should be 16gauge (or thicker) if you can get it.
K2

I hope you mean all 4 flue tubes?
Martin

 

[Raygers]

If I were to go with 4 superheater tubes would they be connected in parallel or in series?

 

[Raygers]

Hi Raygers.
To address your points individually.

• "I went for a nap, I'm old so allowed," - Join the club. Random Napping is a part of the joy of retirement, I think?
• "and it came to me in a dream what tenor meant." Take more naps! - Brilliant. You are correct - ish.
• Replace a number of tubes with a 1" one, run the steam pipe from the top down the 1" tube," - CORRECT.
• "run around the firebox a few times" No, just take the down-tube to a "spear-point" sort of joint with the return tube and then back up the tube to the steam supply. This Spear-point can project maybe an inch or so into the firebox. Here is a picture of spear-point superheaters for a horizontal (locomotive) boiler, but will show you better than a wordy explanation. Imagine (in your dreams?)the spear end projecting into the firebox, not to prevent the coal and shovel, but maybe to the sides and back away from the fire-door, so you can see the spear-ends and still get coal all around the fire-box. (or a gas burner beneath the spear-ends).

View attachment 151994

• Do I have this right? - Yes, mostly. "I don't do well with words" - You berate yourself. As clear as everyone else, "I'm more of a blueprint/picture/visual guy". - I think we all are really... 50 years of Engineering profession and thousands of small sketches - on whatever was at hand. Sometimes just a stick on a rusty /dirty surface was all there was, but enough! = More usually, a pen on paper - or *** packet! Nowadays a phone with camera?
• "One caveat though, the melting point of copper is 1083.4 whereas tenor's calculations predict firebox temperatures are 1265.3, would this not be a problem?" - OK, a good question, the answer (not something from a dream) is "no". What happens is that the gas closest to the metal forms a relatively insulating protection where the heat is flowing into copper with "cooler" water at the other side of the copper tube wall. Here's a picture of temperatures in a firebox, at the wrapper/water jacket. Personally, I would have a thick insulation not "20.C. atmosphere" around the outside. But you'll get the gist.

View attachment 151991
You can see from the "graphical line" that the temperature adjacent to the copper surface drops very rapidly from 1000C to metal temperature. This is only "schematic", but close to the truth. It is because of the high "heat" needed to melt copper, as well as the "high conductivity" that carries the heat very quickly to the wet-side, that makes it work so well. Aluminium has similar properties, but is rapidly corroded by steam at steam temperatures, so is no good for our boilers. (but OK in air conditioning systems with refrigerant gases).
Finally, a picture of a section through a fire-hole door - as recommended in "the books". - Sorry it is not "CAD"! View attachment 151995
Turn a copper ring with a step on either end, compress the "top and bottom" (in a vice?) to make it oval, than form it around a solid 5in dia former to make it fit the dished holes and cylindrical shape of the dished hole in the firebox wrapper and Shell outer. - I think you get my meaning...
Not easy copper forging! Anneal frequently while forging the copper to shape. (it hardens considerably at each bend).
Hope this helps?
(That was much easier to explain the Quantum mechanics!).
K2

Here's what I'm proposing to do, a ring of copper soldered around the door between the inner and outer wrapper. Probably hold it in place with screws until it gets soldered..

 

[Steamchick]

Hi Martin, by "both" I meant 1" dia, & 1/2" dia, flues.
Raygers: 4 superheaters connected in parallel is normal. The fun part with superheater is making both inlet and outlet manifolds... so you have a whole assembly to fit into the boiler- then connect to the boiler steam outlet point, then pipe to engine....
Loads of pipe engineering!
K2

 

[user 52078]

If I were to go with 4 superheater tubes would they be connected in parallel or in series?

Paralell is what I worked the calcs. for.
Happy Christmas,
Martin

 

[Steamchick]

It's the eccentric engineer who looks at regular shapes - with right-angles, parallel lines, circles and flat faces and tears up the rule book / drawing board and starts using non-right angles, curves, etc... I upset more than a few when I worked in a "traditional" design office and doubled the performance of their existing design by being a bit eccentric....
So I may well be wrong.
K2

 

[HMEL]

Steamchick,
My program is not sufficiently advanced to look at different flows and differing temperatures down different tubes or flues - that would require seriously good CFD software. So the implicit assumption it is the same across all tubes, and a different value across all superheater flues. That is sufficient for the purpose the program was designed, which is for outline design of a boiler to look at the many different variables (tube size, number, superheater geometry, firebox size, etc. etc.) and their effect on draught requirement and steam generation.

My own thinking is that provided the tube bank gives a reasonable resistance (draught), the flow across the bank will be fairly even, hence flow through the firebed will be fairly even. In fact, putting flow through a whole load of parallel resistances such as honeycomb or mesh is a classic way of producing smooth flows in applications like wind tunnels.

HMEL
I have found that balancing flows through superheater flues and firetubes is quite important to the overall boiler performance. You are right in that if you get it wrong you will either:

• Get all the gas flow going through the firetubes, while the superheater flues are dead and do very little superheating. (plenty of steam, very little superheat)

or:

• Get all the gas flow going through the superheaters, while the firetubes are dead and not evaporating much water. (great superheat, just not enough steam)

That's a bit of a simplification, but I hope it shows the problem.

The superheater flues have superheater elements in them, so that blocks part of the flow and the extra surface area pushes up the friction. So my program puts the numbers to that conundrum. If you look at the summary spreadsheet I did back in post 33 (page 2 of this thread), you will see that I work out the draught through both firetubes and superheaters (known as max & min draught) and also the flue gas exit temperature from both firetubes and superheaters. You will see that for both the superheater options I propose, the draught is very similar and the spread between exit temperatures is reasonable (about 30 deg. C) showing that there is a reasonable balance between the two. You can optimise that balance by changing the superheater flue diameter, superheater element size and number of elements (assuming a constant firetube set up). The degree of matching is limited by working with commercially available tube sizes.

Hope that helps,
A Peaceful and Happy Christmas to all,

Martin

Ok, I researched locomotive boilers and how superheaters were incorporated to the steam circuit. The larger tubes contain the superheater elements. Also the superheaters are designed to minimize header lengths so in order to do that they are located closer to the steam dome. Also you want the superheater to see the hottest part of flue gases. -- My observations of why things are put where they are. Now I would also bet that when the larger tubes are sized they take into account the area the superheater tubes take up so the total flow area in the larger tube is approximately equal to the smaller tubes so the flow is balanced even though tubes are larger. Now I cant prove this unless I had access to the design prints but it is certainly how I would have approached the design. And I agree superheat is good on most systems. The exception being process equipment requiring saturated steam.

Merry Christmas Everybody