Model boiler safety calculations

[Andrew TinsleyParticipant @andrewtinsley63637]

I have made quite a few copper boilers, some to my own design. Never had a problem certifying them until this latest one. The boiler is for a 3.5" gauge Quarry Hunslet. Looking at a couple of published designs, show them to be very similar (as perhaps one might expect)

I have now been asked to provide boiler safety calculations, which is quite reasonable. However to do a proper stress calculation for the complete boiler looks as though it could turn int quite a complex exercise, if done correctly.

Can anyone point me to what is required and how straightforward or otherwise these calculations are?

I am quite relaxed about the results, as comparing my designs to published ones. I always err on the safe side and my boilers are probably stronger than most.

Andrew.

[Andrew TinsleyParticipant @andrewtinsley63637]

[noel shelleyParticipant @noelshelley55608]

There are many books on boiler building and the calculations. Speak to those who will test the completed boiler first, and if you can use a published design then this work should not be needed, though may prove interesting. Who wants these calculations ?  Read the latest copy of the boiler testing code ! It was posted on here today.

Model engineers handbook – Tubal Cain, Model boilers & boilermaking – K.N.Harris Model locomotive boilers – Martin Evans.

Put simply, Basic maths on the strength of the boiler shell, tube plates, firebox and the staying of flat surfaces. Good luck, Noel

Edited By noel shelley on 18/01/2021 15:58:08

[Phil H1Participant @philh196021]

Andrew,

It is a blue Monday – so please forgive my opinion but here goes;

I will echo the 'good luck' from Noel because I have searched a few times for this without finding anything that is convincing. I have at least one of the referenced books and you will get the following results;

Shell 3/32" thick (1/8" if it is a larger boiler)

Backhead 1/8" thick.

Other plates 3/32" for a smaller boiler or 1/8" for a larger one.

The equation for the side stay spacing is wrong in the book by Martin Evans (probably a printing error)

What on earth your inspector is going to do with the results that are nearly always the same – I have no idea because nearly all of the 3 1/2" gauge boilers I have seen have the same plate thicknesses.

I repeat – good luck.

[Andrew TinsleyParticipant @andrewtinsley63637]

None of the books on model boilers give any indication of how to calculate the strength of the shell. Even the older ones going back to Greenly's time, don't give such calculations.

Also never seen a published boiler design that gives the calculations.I suspect that the calculations are simple minded ones coupled with using "normal good practice".

Doing a proper strength calculation on a model boiler isn't trivial and I am not too sure that many boiler inspectors would be able to follow the maths anyway.

Please tell me if any books or designs do have the calculations, it would be interesting to see.

Andrew.

[Brian HParticipant @brianh50089]

Andrew, if you are in Australia then there is a published standard that covers model boilers.

If you are in the UK then there is an equally good standard:

**LINK**

Or an excellent book is "Model Boilers & Boilermaking" by K.N. Harris.

Brian

Edited By Brian H on 18/01/2021 17:36:09

[duncan webster 1Participant @duncanwebster1]

+1 for the Aussy code, the SFED document (Brian's link) is not really helpful on how to do calcs, it just says you might have to provide them.

If you have access to a library which keeps such things then PD5500-2003 covers unfired fusion welded pressure vessels in copper, this is probably as near as you will get to a national standard for silver soldered copper boilers. The old standard was BS5500, this didn't cover copper, but it did cover aluminium and some logical adapting can make it cover copper

[Andrew TinsleyParticipant @andrewtinsley63637]

Hello Brian H,

The UK link you give does not give the basis of the required calculations and KN Harris' book, although an excellent one, gives no information on the required calculations either.

I will peruse the Australian code with interest, something that I have not done. Thanks Duncan too.

Andrew.

[Nick Clarke 3Participant @nickclarke3]

As far as I recall there were a few formulas in later editions of the Model Engineer Plans Handbook 3 but these seem to go for very silly money online.

But I suspect that formulas were less important than ‘rule of thumb’ and experience. Here are some instances from old ME magazines – and through old ‘grandfather’ rights they could possibly still be built today (not that I would!)

‘The boiler (for a 3 ½” loco) was made from 18g copper as I had some in’ (LBSC Canterbury Lamb, describing Grosvenor’s boiler)

‘The stays were made from 1/8” copper unwound from electrical cable I was given’ (LBSC again)

Another author made alterations to a 7 1/4" boiler design in the very next instalment and then published a design with further revised staying in the instalment after that (Martin Evans, Highlander design)

Edited By Nick Clarke 3 on 18/01/2021 18:51:48

[Dave HalfordParticipant @davehalford22513]

That sounds rather like an 'I don't want to do the job'. He knows very well you won't be able to provide the calc's.

[Dave Smith 14Participant @davesmith14]

Andrew

I have a copy of the Aussie code if you want a copy. It details how a person without the necessary engineering expertise can demonstrate a boiler design is structurally sound. I am modifying the design of my Aspinall boiler to delete the crown girder stays and substitute rod stays. As this is a substantial deviation from the original design a structural report is required to show no regression. I am approaching the analysis in two to ways. First I am using the Aussie code and applying it to both the original and revised design. In actual fact if you have all the data the calculation part only takes a few minutes as most of it is using tables to establish things like stay diameter and spacing. The second approach is using Finite Element Analysis (FEA) the same as Luker did in the recent ME. I have already done a steel welded boiler using FEA about four years ago, which was backed up with reference to the Aussie Steel boiler code and hand calculations. The FEA analysis was validated by my senior stress engineer at work, if you are not careful it is S!"T in gospel out. I am running two analysis of the boiler, the revised design and the existing design to give a base line and show no regression. Not sure how our boiler inspectors will react to the FEA, but they have to come in the real world, hence the Aussie back up.

Dave

[Tony Wright 1Participant @tonywright1]

Best thing ask a boiler inspector who has a bit of common sense ! My club has a Ken who fills the kettle with water pumps it up twice the working pressure . If it doesn’t pop you get an hydraulic certificate ! So simple has been used for years never any problems . To many what if’s nowadays.

[noel shelleyParticipant @noelshelley55608]

Andrew, In all 3 books I listed there ARE the calculations for boiler design, certainly enough to show a boiler is sound. Martin Evans and Tubal Cain both give formulas enough that should satisfy most boiler inspectors.

Martin Evans page 28, K.N. Harris Page 31, Tubal Cain Page 10,2   Noel

Edited By noel shelley on 18/01/2021 21:44:42

[Paul LousickParticipant @paullousick59116]

I am in Australia and build boilers to our design code for model boilers and by doing so, calculating the stresses is not necessary because they have already been done to produce the code. If the boiler materials or dimensions were outside of the code, I would have to provide calculations to satisfy the inspector.

Paul

[BazyleParticipant @bazyle]

While I have both the Evans and Harris books I haven't cross-checked but I venture to suggest it is a fairly trivial calculation that any first year mechanical engineering or Physics student should be able to work out from first principles. Even A-level schoolkids as it is just a mechanical calculaton and the syllabus does cover things like Young's modulus.

[Paul KempParticipant @paulkemp46892]

Bazyle,

I would be very happy to see your formula from first principles to provide the pitch of stays. Agreed most of the shell calculations and size v number of stays is pretty basic but I think the calculation to prove resistance to quilting between stays of a flat plate is more complicated?

Paul.

[JasonBModerator @jasonb]

There is also Hainings "Countryman's steam" although not so much on copper boilers he does give a very simple formula for Firebox stay spacing and an even simpler way to arrive at diameter. You would have to check with your inspector if the use of these formulas was acceptable.

You might also want to think about contacting someone like Paul at Southern boiler works who I think will produce a drawing and calculations for a fee. He has a spreadsheet that he can enter details in and it will show if they fit within the required factor of safety or not.

[IanTParticipant @iant]

Andrew,

I attended one of the Southern Feds Boiler Inspector seminars a few years back. They provided guidance 'notes' in this area which I think may be sufficient for your needs. Most boilers are inspected on the basis that they are a "recognised" design (e.g. one that's been published) – and of course they are normally over-spec'd anyway.

If you'd like a copy of the notes, I'll scan them for you (assuming that I can find them). Please PM me.

Regards,

IanT

[Phil H1Participant @philh196021]

So have we concluded that calculations are not necessary yet?

Just to repeat;

Tube section in front of the firebox – about diameter 4" at 13g – no calculation required..

Firebox outer and inner wrapper 3/32" or 13g – no calculation required..

Throatplate, backhead and smokebox tube plate in 1/8" – no calculation required.

Firebox plates 3/32" or 13g – no calculation required.

Firebox stays on a 3/4" square grid at a decent size – say 4BA or equivalent copper rivets- no calculation required.

Then you have some nice choices regarding tubes but that is based on performance only assuming you put enough in to 'stay' the tubeplates.

Crown stays – ooh tough – 3/32" sheet if you are in Britain or more rods if you are Australian.

Please somebody tell me which bit needs a calculation?

[Andrew TinsleyParticipant @andrewtinsley63637]

It is pretty obvious that my ideas of a safety critical calculation is miles away from what is understood by model engineering boiler makers.

I agree with Phil 's comments above, the so called calculations are simply trivial or non existent. If I had realised this in the first place, I would not have posted.

It would appear that most things are decided by past practice and no real calculations are involved. In a past life I was responsible for some safety critical calculations on nuclear reactor pressure vessels and I was expecting a derated version of this type of calculation. My apologies for misunderstanding the context of safety calculations in model boilers.

Thanks everyone,

Andrew.

[SillyOldDufferModerator @sillyoldduffer]

Posted by Andrew Tinsley on 19/01/2021 15:20:55:

…

It would appear that most things are decided by past practice and no real calculations are involved. In a past life I was responsible for some safety critical calculations on nuclear reactor pressure vessels and I was expecting a derated version of this type of calculation. My apologies for misunderstanding the context of safety calculations in model boilers.

…

That's my perception too. Model Boiler Inspectors seem to look for established best practice and don't have to do the sums, understand materials or the engineering. The system is a little rough round the edges! Tony mentions an inspector whose hydraulic test consists of pumping a boiler to twice working pressure. OK except applying excessive test pressure to a boiler weakens it. Does this chap know what he's doing or is his twice working rule guesswork? If guesswork, the test could unnecessarily take a few years off the boiler's working life. Hundreds of pounds of the owners money down the drain on an Inspector's whim. Perhaps 1.5x working pressure is good enough? Or maybe I'm completely wrong and test pressure should be 3x? I suspect no one knows!

Apart from the possibility of quirky inspections, I feel the boiler inspector system works well though. By only allowing time proven designs clubs get the necessary insurance, engines steam in public, and everyone is happy.

Almost everyone! The loser is the guy who wants to innovate and experiment. The chap with ideas, improvements, and cost efficiencies. Today's LBSC! To get approval the builder has to prove he knows what he's doing, which involves a professional design approach that's hard to explain to anyone other than a suitably qualified engineer. Andrew has my sympathy!

Dave

[Phil H1Participant @philh196021]

Andrew,

I think it is a very interesting post and well worth it. Somebody might pop along later today and explain why a particular sum is required and we will all be educated and better people for it.

[Oven ManParticipant @ovenman]

I was reading the Boiler Test Code 2018 referred to above just for interest last night. I know nothing about boilers or their testing but it says in the code that an initial pressure test is 2x WP and subsequent tests with bits attached is 1.5 x WP. It seemed pretty well defined to me.

Peter

[duncan webster 1Participant @duncanwebster1]

Working to the Aussy code is not difficult, just look up some tables for most things. They are at least based on national standards rather than something written in a book without any justification. Why are people so resistant to it? For PhilH1's example you could get away with a thinner barrel and save some money. I can't find anything banning girder crown stays, they are more difficult to justify but not impossible.

Yes if it passes the pressure test it is OK, but if it fails because of a design error you'll be a bit fed up, it has been known!

Edited By duncan webster on 19/01/2021 17:21:10

[Dave Smith 14Participant @davesmith14]

Duncan

I agree the Aussie code is well thought out and written. It is very easy to work through a new design using it. The only issue some people may have with it is the use of metric units and having to convert from imperial. Surely it should be adopted here as the boiler test code is purely that, a method of testing boilers with no guidance on what is an acceptable design. That is left to the whims of the boiler inspectors (no offence intend to them by the way) which as a professional engineer I find an odd situation.

On girder stays, para 3.4.4.1 of the Aussie code says you can use Girder stays. However para 3.4.4.2 then goes on to say they can only be used if longitudinal rod stays cannot be fitted. So whilst permitted I reckon there are not many situations where you cannot use rod stays, so effectively banned!

Dave

[Author]

Posts