Do the ends need to be thicker?

[Bob PerkinsParticipant @bobperkins67044]

Dear all

I'm building a couple of Tubal Cain's "Polly" vertical steam engines, from plans in that well known book. The plans specify for the boiler shell 24SWG. It says the following with regard to the end plate material "These are thicker than the shell, for strength reasons. You need 22SWG"

I understand that there is more stress on the end plates, and there is a risk of deforming and weakening and I have read that end plates can be domed to increase their strength. (I'm building up a bit of an interest in the topic)……

Curiosity leads me to ask the following……

Is the strength or integrity of a boiler comprimised if the shell and end plates are the same thickness, or should they be different thicknesses?

is it just a material cost thing, i.e. why use thick when you can use thin?

Will the world stop rotating if I use 22SWG for my shell and end plates?

Comments and advice please.

Thanks

Bob..

[Bob PerkinsParticipant @bobperkins67044]

[Sub MandrelParticipant @submandrel]

It's an engineer thing. If tyou don't need 22swg for the shell, why waste material (and money) doing so?

The only difference will be a tiny reduction in boiler capacity – copper conducts heat better than water so it won't make it steam less efficient.

I made a modified version of the 'Henry' boiler and used well over-thickness copper for the central flue to no ill effect.

Neil

[MICHAEL WILLIAMSParticipant @michaelwilliams41215]

Hi Bob ,

I'm afraid that the complete answer would take many pages but here is some of it :

(1) The strongest boiler possible is spherical and made from constant thickness material .

(2) A reasonable working approximation to (1) is a cylindrical centre section with spherical ends and all made from constant thickness material .

NB : Neither (1) nor (2) need any staying – the steam pressure load is resisted by what is properly called a homogenous tension field in the boiler shell – basically works in same way as a toy balloon .

(3) In practical boilers the ends have to be much flatter , the tension field becomes a minor factor and the end plates now have to resist steam load in bending – basically like putting a weight on a circular disc supported at the edges . Bending strength of a flat plate is very poor compared to the balloon strength of curved plates mentioned above and for the same thickness of plate large deflections can occur .

These large deflections can be controlled either by making end plates significantly thicker or by using auxilliary stays .

(4) In boilers with cylindrical main barrel and flat ends there is a secondary problem in that the stresses and deflections at the joint between the two sections don't match and this can give rise to local stresses wich are significantly higher than elsewhere in the boiler .

This mismatch can be managed by using flat end plates but ones designed with a relatively large radius flange where they join the barrel .

There is much much more to it – please ask any questions you like .

Regards ,

Michael Williams .

PS:

Bit hard to get your head around but any pressure vessel which is not already balloon shaped is always trying to become balloon shaped . If you can visualise how any arbitrary design of boiler would have to distort in order to become balloon shaped it is a very good illustration of how it is resisting the internal pressure forces and also how design could be modified to be more structurally efficient .

There are two alternative principles used in more sophisticated pressure vessel design :

(1) The constant stress principle – all parts of vessel stressed to same level and all 'working' to same extent . This is the most weight efficient design and is used in aeroengines .

Usually has to be an approximation due to many conflicting requirements and constraining geometries .

(2) The variable stress principle – all parts which can be easily analysed are relatively highly stressed and all parts which are difficult to analyse are relatively modestly stressed . Not very weight efficient but very conservative and used for many purposes . This fundamentaly is how model boilers have always been designed historically .

Edited By MICHAEL WILLIAMS on 11/12/2012 15:33:57

[nigel jones 5Participant @nigeljones5]

You can also brace the end plates to help with the stresses. Locomotion no.1 uses bracings.

[Robert DoddsParticipant @robertdodds43397]

Michael Williams,

An elegant explanation of a complex subject.

I particularly appreciated your balloon analogy. I might even have gone up a grade in Strength of Materials if I'd had that proposition before me.

Bob D

[Ian PParticipant @ianp]

Agree wholeheartedly with Bobs compliments. I have no interest in steam or locomotives but Michaels explanation was pleasure to read in its own right.

Some people have a gift of using just the right analogy to explain something complex, Michael certainly has it.

Ian

 

Edited By Ian Phillips on 11/12/2012 21:44:59

[Bob PerkinsParticipant @bobperkins67044]

Wow. Thanks for the comprehensive feedback Michael. I agree with the above comments , your explanation is both informative and easy to understand . I've been e-thumbing through K Harris's model boilers and boiler making, which explains the basic principles and calculations and then moves way above me. I've read mentions of a Tubal Cain book that goes into the same. Does anybody have details of this book? I'm interested in the factors that affect the calculations e.g. Riveted or silver soldered joints. I can find the stress figure for copper and the recommended safety factors but can't find anything for brass boilers. Can anybody lint me towards these please? The Polly boiler is brass, and I know that the design pressure will have been calculated, and that there are probably thousands of these been made, and that this is probably all un warranted …… But I would like to understand an make the calculation for myself.

Thanks for the comments

Bob..

[Robert DoddsParticipant @robertdodds43397]

Bob P,

You seek info regarding brass boilers. Without being disrespectful to Tubal cain or Polly it is usually considered more as a toy type of engine rather than one built to do real work. Brass is not generally adopted for working boilers. It has serious limitations in its failure modes if a hot fire is set under one.
The melting point of brass is substantially lower than copper or steel, it is weakened by "dezincification" under prolonged high temperatures and is far less predictable when it comes to annealing after any cold forming operations to give rolled flanged ends.

Brass is not a No1 choice for silver soldering either, again due to the lower melting point of brass, and while I admit to not knowing Polly in detail I'm guessing the design is for a soft solder boiler assembly with a spirit burner to provide the heat.
By all means use the thicker wall tube if thats what you have, but don't deviate too far from the build method that Tubal Cain specified if you are working in brass.

Many tomes have been written on boiler construction, dimensions etc over the years. Most of the early ones were based on empirical data gathered from boilers that had not blown up or burnt out. The info gradually got better as faiure modes were analysed but it was 20th century before stress analysis was extensively used to detail the design of boilers.
Most of the data is for steel, some for copper but very little would be for brass built boilers.
Tubal Cain was a well respected engineer and model engineer, you won't go far wrong sticking to his design.

Bob D

[SteamshyParticipant @steamshy]

Hi Bob, The formula for copper Boiler plate thickness is T= P X D /7. This would give you the plate thickness in Thousands of an inch. The formula would give you a safety factor of 10. T refers to thickness, P is Pressure, D is the plate diameter. Brass is not mentioned often because simply it is not a good thing to construct boilers with.

Andy

[Bob PerkinsParticipant @bobperkins67044]

I guess that "Polly" is a toy engine, which is ok as the three I'm building are for my grandson, nephew and one for me, who is probably the most immature of the three of us! Copper is given as an option for construction, and would be my preference , if only for aesthetic reasons, however I seem to be unable to find a supplier of modellers amounts of 22 and 24SWG copper sheet. Any suggestions would be welcome. The construction instructions, as per other articles I've read calls for silver solder, and warns against the problems and hazards with using soft solder on any boilers.

Bob P

(I,ve picked up a P which is good)

[Bob PerkinsParticipant @bobperkins67044]

Re my previous post I see I can get 22 and 24 swg copper from m-machine.

Bob P

[Ian S CParticipant @iansc]

I have made little boilers from 1 1/2" copper pipe, and used a bit of the same pipe flattened out for the endsilled a hole in the center of each end, and used a bit of 1/8" brazing rod threaded at each end with nuts on, through the boiler, then silver soldered over the nut. Ian S C

[Author]

Posts