Boiler calculations, end plates

[Benedict WhiteParticipant @benedictwhite51126]

Clive, I am not looking to reinvent the wheel, just extend slightly its reach.

The formula in KN Harris's book and others is derived from mathematical models, one of my books goes through the derivation of the maths for the shell. What it does not do is the end plates.

(Another one goes through the maths for valve gear including Joy valve gear).

[SillyOldDufferModerator @sillyoldduffer]

Posted by Clive Brown 1 on 21/09/2023 15:51:54:

Whatever the reason, model loco boiler construction in the UK hasn't changed much for a century. Could be time for a change!

Dave

Why, many thousands of perfectly satisfactory model boilers have been made over many years in home workshops Their safety record is first class.

…

Can't disagree with that, but one opportunity might be steel welded boilers,. They're potentially much cheaper than copper and silver solder constructions. I think an assured safe design would be popular because lots of people do home welding now.

Dave

[Benedict WhiteParticipant @benedictwhite51126]

Posted by SillyOldDuffer on 21/09/2023 19:13:44:

Can't disagree with that, but one opportunity might be steel welded boilers,. They're potentially much cheaper than copper and silver solder constructions. I think an assured safe design would be popular because lots of people do home welding now.

Dave

The fed code does include a section on steel. There are issues with steel instead of copper though, for example steel is a poor conductor of heat compared to copper and as a consequence a too low water level is an instant disaster to the firebox crown whereas you have time to refill the boiler in a copper one before the crown goes. Also harder to steam.

These are much less of a problem at 7 1/4" and above than 5" and below.

[JasonBModerator @jasonb]

It's the quality home welding that is the issue not the design of plates, tube and suitable welds

[duncan webster 1Participant @duncanwebster1]

This is just going round in circles. The Aus code and BS both have formulae which allow one to calculate the stay circle diameter for any thickness of flat plate once you've decided on a safe working stress. The Aus code has this for copper, and it is possible to derive one from the BS as I have done, coming to a figure remarkably similar to the Ausy one. If not confident of coming up with this stress, just buy the Aus code and have done with it.

Just because our boilers are smaller than big industrial ones doesn't make the calculations invalid. Some time ago I outlined a layout of tubes which would meet the Aussy code with 16g end plates, a couple of bits of copper about 3" diameter isn't going to break the bank.

[PeroParticipant @pero]

Going back to the beginning, which is a small copper boiler of under 3'', some of the older construction articles simply instructed the builder to slit and flatten out a piece of the boiler tube material to make the end plates – no calculations presented.

And for a small, low pressure multitube boiler none needed. Properly built, failures did not occur, the staying provide by the tubes being more than adequate.

The AUS code for small (copper) boilers provides a more technical approach for those who feel the need.

Pero

[noel shelleyParticipant @noelshelley55608]

As Pero has said "going back to the begining " and as Duncan has said ( in so many words ) this is destined to disappear up it's own fundamantal orifice. The O/P has a bit of scrap copper water or gas pipe and wants to build a boiler using this ? He has done the calculations for the stress in the shell and it works out safe, the end plate is under the same pressure but is flat – it needs stays, be they solid or tubular. A rough calculation at even 70 psi is that it will be under 2 bar/litre and not much bigger than a Mamod and only requires a safety valve. The calculations are in many books, 2 of which he has, After 3 pages of answers and other comments is this one NOT resolved ? Noel.

[JasonBModerator @jasonb]

Posted by noel shelley on 22/09/2023 09:51:06:

The calculations are in many books, 2 of which he has, After 3 pages of answers and other comments is this one NOT resolved ? Noel.

Noel, can you say which page the calculation is on in Harris's book as like the OP I can't see one.

As far as I can see it tells us how to calculate stay diameter , gives a table for spacing of stays for various material thicknesses but not the thickness the OP wants to use. No indication of how the table was arrived at.

[Benedict WhiteParticipant @benedictwhite51126]

Thank you Jason. That was roughly the original problem I had faced. Also my machinery's has some odd rules under stress which involves formula but with rules of thumb for some constant whose derivation and units are unclear. (1966 edition)

I do have a "Fowlers Mechanical Engineers Pocket Book 1955" Which does have a formula bit it uses a value called "fibre stress" for which I am unable to find a definition. However my next port of call is to use the rules of thumb in K.N. Harris's book to reverse calculate "fibre stress" of copper for one plate thickness and then see if the formula then produces results at other thicknesses consistent with the rules contained in the K.N. Harris book.

[Benedict WhiteParticipant @benedictwhite51126]

Whilst I am talking about stays and plate thickness, the other problem is that the K.N. Harris rules of thumb make no allowance for pressure which is not satisfactory. I will have to assume that it is on the basis of around 90PSI for the rules of thumb.

[JasonBModerator @jasonb]

Harris book says the stay table is for pressures of 100-120psi, last line above the butt strap drawings

[Benedict WhiteParticipant @benedictwhite51126]

Many thanks JasonB. I missed that in all the excitement of picking through many books. Covers the area for most model boilers.

[JasonBModerator @jasonb]

Just had a look at John Haining's "Countryman's Steam". Much like the others he goe sinto detail about barrel calculation sbut not flat plates but he does say

"As a rough Guide for those that do not want to become too involved all flat stayed surfaces should be at least 1/3rd thicker than the barrel material"

So why not take that as a fact that your 1mm material is not ideal. You can then just use Harris's table for 16swg material, or go with Duncans suggestion of tube spacing in 16swg material etc. It will be a lot easier to drill and solder without the risk of overheating

If you are not careful using very thin plate will result in having to use so many tubes to act a sstays that you will start to reduce the water volume within the boiler so much that you will end up with  a flash steam boiler

Edited By JasonB on 22/09/2023 11:18:33

[Benedict WhiteParticipant @benedictwhite51126]

I can do that JasonB, and that puts me within the K.N. Harris book so I need look no further. It's just not that satisfactory to work from rules of thumb without seeing where they come from.

Still, I have enough places to look now to solve that problem.

[noel shelleyParticipant @noelshelley55608]

Jason I'm sorry that I did not make myself clear, that the end plates are under the same stress as the shell and that the end plates will be the same thickness as the shell, then the stay calculations have to be done to deal with loading, these calculations are in several books. Looking at drawings for Lion in 3.5" the tube plates Etc are the same as the shell but for Butch in 5" some flat plates are thicker, outer fire box while the inner is the same as the shell. My 6"vertical has shell and tube plates at 10G.

But looking at Model boilermaking by E.L. Pearce, Page 9, like John Haining, he suggests using 25% thicker plate than that for the shell.

K.N. Harris states (pages 29 & 30 ) the stay spacings for as jason says of 100-120 psi but then follows it up with the calculations and TableVI at the back gives stay strengths for various BA sizes and materials.

I hope this will be helpful. Noel.

[JasonBModerator @jasonb]

Although under the same stress being flat the same calculation for the barrel can't be used so there is still no calculation for flat plate thickness which is what the OP wanted.

If you just worked it out based on what diameter and number of stays were required and populated them equally about the surface there would still be a risk that the material could deform as the spacing may be too large even though there are enough stays.

It is interesting to do Harris's stay calculations based on 110psi being the average of what he says the table relates to, Only problem is 5 or 6BA stays at 1/2" spacing would not be strong enough! and there is certainly a bit of pick whatever safety factor you fancy.

 

Edited By JasonB on 23/09/2023 07:27:00

[Benedict WhiteParticipant @benedictwhite51126]

Posted by JasonB on 23/09/2023 06:55:18:

Although under the same stress being flat the same calculation for the barrel can't be used so there is still no calculation for flat plate thickness which is what the OP wanted.

If you just worked it out based on what diameter and number of stays were required and populated them equally about the surface there would still be a risk that the material could deform as the spacing may be too large even though there are enough stays.

It is interesting to do Harris's stay calculations based on 110psi being the average of what he says the table relates to, Only problem is 5 or 6BA stays at 1/2" spacing would not be strong enough! and there is certainly a bit of pick whatever safety factor you fancy.

Edited By JasonB on 23/09/2023 07:27:00

There are a number of issues with the end plates. Firstly as you say the nature of the stress on a flat plate is not the same as for a circular barrel, and the flat plate may end up plastic deforming.

The formula I have from one of my old books suggests that the relationship between stay spacing and plate thickness is a linear function of thickness whereas the numbers in K.N. Harris are less than linear which makes more sense as the area between stays and therefore the force on that area goes up as a function of radius squared. On the other hand some people argue it is like a beam where it would be linear.

I may have to see if I can find a book on full size practice and see how it is approached there.

[duncan webster 1Participant @duncanwebster1]

There is not and cannot be a relationship between tube plate thickness and barrel thickness. Require tube plate thickness is dependant on material properties, working pressure and stay (or tube) pitch.

I have given references to where formula for working out stay pitch can be found. There seems to be some reluctance to follow these up.

[Benedict WhiteParticipant @benedictwhite51126]

Posted by duncan webster on 23/09/2023 11:56:58:

There is not and cannot be a relationship between tube plate thickness and barrel thickness. Require tube plate thickness is dependant on material properties, working pressure and stay (or tube) pitch.

I have given references to where formula for working out stay pitch can be found. There seems to be some reluctance to follow these up.

The relationship between end plate thickness and stay pitch is linear in those and is not linear in the K.N. Harris table.

[duncan webster 1Participant @duncanwebster1]

so who are you going to believe, national standards (both UK and Aus) or KNH? I've no idea where KNH got his figures from.

Reading KNH again after many years, he quotes stay pitch for pressures of 100-120 psi, and then goes on to give a range for the required pitch. Perhaps this is meant to mean take the lower for 100 and the higher for 120, but he only quotes one figure for 16g, which makes no sense. In any case, if you plot the upper and lower figures for stay pitch against material thickness you get this

Bearing in mind that KNH only quotes to nearest 1/16" I'd call that fairly linear. I wouldn't expect to get close agreement with the Aus standard or my figure derived from UK standard.as KNH uses an acceptable working stress of 3125 psi, which is somewhat lower than UK or Aus

[JasonBModerator @jasonb]

The problem is that if you work out the loading on a stay threaded 5BA @ 120psi it is not strong enough at 1/2" spacing and comes below Harris's suggested factor of safety

Using his calcs a stay supporting 120psi @ 1/2" spacing needs a Cross section of 0.012 sq in. but the core dia of 5BA is 0.0073 sq in this falls below his suggested average safety factor of 8 ( 25000/8 = 3125). The 6BA at the same 1/2" spacing would be outside his safety factor range of 6 to 10 completely.

His spacing table does seem fairly linea going down 1/8" spacing for each 1/32" drop in plate thickness. So you could say that for 1mm / 0.040" plate the spacing of 1/8" stays would need to be about 10mm, so 7mm between stay or tube edges. Just need to arrange enough of your 10mm tubes to meet that spacing withing the 50mm or so dia of teh end plates. Probably needs a ring of eight tubes around a central one so 9 tubes in total, they will be stronger than stays.

[JasonBModerator @jasonb]

Just tried his calcs for my 2" fowler boiler which has 1/8" solid rivits @ 7/8" spacing and WP of 100psi. His calcs say I would need as cross section of 0.022sq in using a factor of safety of 8 but an 1/8" stay only has a CSA of 0.012sq in

[duncan webster 1Participant @duncanwebster1]

I hadn't even looked at the actual stay diameter in KNH book. He actually recommends 6 or 5ba for 1/2" spacing, using 6ba and 120 psi gives 5286psi stress which is well over KNH, Aus and my figure. Even using 5ba we get 3977 psi, again above Aus and me.

[Author]

Posts