Model boiler safety calculations

[Paul KempParticipant @paulkemp46892]

I started typing a contribution yesterday but this dumb site logged me out in the process so I lost it and then lost interest.

For Andrew you should ask your inspector for the specific calculations he requires – as he asked he must know? I suspect he has asked due to the requirement in the test code which is fair enough but you could then ask him against what standard he will be assessing the results? Assuming he is competent to assess the results then he should also be capable of doing the numbers, do maybe ask him if he wouldn't mind doing them for you?

There seems to be an assumption that for the published designs calculations have been done? Has anyone seen the supporting calculations for these designs? In fact do any of the published drawings bear any evidence they have been assed or approved by a notified body or in fact any body outside the publisher?

Jason, I don't have that particular book, any chance you could pm me the formula therein for stay spacing? I used to have one from the Indian regulations of all places that seemed to give sensible results for larger scale miniature boilers but I have lost it and the web site I nicked it from has been deleted!

I don't disagree with the comments regarding the Australian code but not convinced we just need a direct copy as it is restrictive on working pressure for steel boilers and I understand prohibits some of the established boilers in use here with proven safety records such as the Simplex design.

SOD, 2 X wp and 1.5 X wp tests are industry standard derived from stress fos.

Paul.

[Tony Wright 1Participant @tonywright1]

Just to say , all the boiler designs of Tubal Cain ,LBSC , EVANS .Have always said initial hydraulic test 2x working pressure .Re test 1.5 x working pressure . It’s good enough for me I’ve never had a boiler failure .If it ain’t broke don’t fix it .Dont try to over complicate things.

ps has anyone ever known a model boiler actually explode !!!

[duncan webster 1Participant @duncanwebster1]

Who's trying to change the test procedure? Not me or any other contributors.

[John BaguleyParticipant @johnbaguley78655]

Quite a few years ago I found this spreadsheet for designing boilers which I've found very useful:

Boiler Design Calcs

It allows you to see the effects of varying the shell thickness, the stay diameter and spacing etc.on the safety factor of the boiler.

It's always struck me that the stay spacings given by LBSC on many of his boilers are a bit wide and I tend to use more and put them closer together. The existing spacings obviously work as there have been hundreds of boilers made to his designs but it does lower the safety factor. I always try and work to a safety factor of 10 whenever possible which is probably a bit over the top.

John

[duncan webster 1Participant @duncanwebster1]

In my copy of aussy code (1994) para 3.4. 2 is relevant to staying of the backhead and tube plate not the firebox crown.

[JasonBModerator @jasonb]

John, I know of one commercial builder that uses the same spreadsheet so that would be a good starting point for anyone wanting to get some figures. I did not know it was in the public domain so had not been willing to share my copy.

[nigel jones 5Participant @nigeljones5]

Theres a spread sheet which covers every aspect of safe limits and thickness…ive shared it with several people over the years and now no longer have a copy! If anyone could send me one back I would be grateful.

[JasonBModerator @jasonb]

Fizzy, look at John's link, bottom of the page that comes up

[nigel jones 5Participant @nigeljones5]

Sorry the one at the bottom of Johns link is the one to use – Can you please send me a copy olde friend?

[JasonBModerator @jasonb]

Click on "download the program" 4th line down the page.

Best make that two pros that I know use it.

Edited By JasonB on 20/01/2021 09:23:45

[SillyOldDufferModerator @sillyoldduffer]

Posted by Tony Wright 1 on 19/01/2021 22:53:14:

ps has anyone ever known a model boiler actually explode !!!

Model boilers are unlikely to explode, and the few attempts to do it deliberately have been disappointing. They split and leak rather than go bang, though one soft-soldered example popped the end off and rocketed into a tree. (One of my earlier posts has the reference to an ME article: one of the clubs destroyed 4 or 5 boilers, including at least one Inspector reject. )

Steel boilers are more likely to go bang than Copper, but in model loco sizes even they aren't scary. Bigger boilers are another matter entirely. Mythbuster's have a couple of episodes featuring what happens when a large US style domestic water heater is completely sealed and all the safety measures are bypassed. Spectacular and very dangerous. Don't try this at home folks!

Despite chemicals, electricity, cutting tools, burners and steam boilers Model Engineering is a safe hobby. Far safer than woodwork! Amateur radio is far more likely to be lethal than boilers because people fall off roofs or touch power lines whilst erecting aerials! Apparently real men play with radios, not trains!

Dave

[Dave WoottonParticipant @davewootton]

Thank you John for posting the link to the spreadsheet, i'm bumbling my way through re- working the boiler for my 2 1/2" gauge President Washington B&O pacific that I've been posting about. It was designed in 1928 using cast gunmetal throatplate and foundation ring, so needs a little updating. spreadsheet will be a great help.

Dave

[Dave Smith 14Participant @davesmith14]

Posted by Paul Kemp on 19/01/2021 21:20:41:

I don't disagree with the comments regarding the Australian code but not convinced we just need a direct copy as it is restrictive on working pressure for steel boilers and I understand prohibits some of the established boilers in use here with proven safety records such as the Simplex design.

The introduction of any code should allow grandfather rights on any existing design unless it is deemed dangerous. however if it is substantially modified and I would consider the change from crown to rods stays a substantial modification then, the modification would have to be in line with the code.

Duncan

You are correct my about para 3.4.4 it is about longitudinal rod stays, my apologies. The only stays it mentions for the crown are rod stays, so is it by inference that crown girder stays are forbidden?

Dave

[IanTParticipant @iant]

Yes, of course model boilers are unlikely to explode Dave – because they are designed (over-engineered) not to.

But don't under estimate the ability of a small minority to do dumb things given the opportunity. I often hear folk raging on about the 'dumbness' of the current boiler regulations and why they don't make sense for 'this' or 'that' reason. They seem to miss the key point that (these days) it is crucial to have Public Lability Insurance. My Society became Incorporated (a Limited Company) a few years ago, simply because it helped to lessen (but not remove) the liability risks that Society Members (potentially 'Partners' previously) might otherwise be subject to, should an accident occur.

That accident of course is not limited to a boiler explosion, in many ways that is the least worry. I have seen a large live steam loco nearly exit a high level track with a child sitting under it's path. Fortunately, it didn't come completely off (just overhanging) but it would have certainly done harm if it had. In these ambulance chasing times, a compensation claim would have been a high possibility.

What's this got to do with the Boiler code? Well there is only one insurance company here in the UK that will 'actually' provide Clubs/Societies with insurance cover (try getting it anywhere else in practice) and they have accepted the Orange Book as the necessary guidance in this area. It's been negotiated between the various parties involved and is a pragmatic solution to the problem. If they required all boilers to be painted with polka dots – it would be simpler to just do it than risk not being insured for public liability.

Of course none of this applies to anyone operating a boiler in a private setting but should an accident occur (of any kind) then they'd need to be very sure that their private insurance covered them.

So let's be clear. Society/Club boiler Inspectors make sure that the boilers they examine (within the scope of their competence) are safe for use within the given guidance. In doing so – they are not trying to work on the limits of engineering safety but within the limits of established practice set by practical experience. Their work is well documented and is therefore an important part in demonstrating good faith/observance of the contract made between the modelling organisation and the Insurer.

All this refers to the 'smaller' boilers generally covered by the UK's Orange Book. Anyone contemplating designing/building larger pressure vessels (steel boilers for instance) clearly need both professional help with design, build and test.

Btw – a Shell Test to 2 x WP done properly (e.g. the pressure is increased in gradual stages) should have no adverse effect on the working life of that boiler.

Regards,

IanT

[Nick Clarke 3Participant @nickclarke3]

Dave Wooton – PM for you next time you log in.

Nick

[Tony Wright 1Participant @tonywright1]

Just been watching a you tube video on compressors going pop! If you have an old comp please check it out for rust and the safety release is ok ! Don’t want to lose any contributors ☠️

[duncan webster 1Participant @duncanwebster1]

Dave, if girder crown stays are not banned they are allowed, there is no room for inference in standards. You just need to justify them

I can't make the spreadsheet work, it downloads but then I can't enter anything in the boxes. However it still doesn't reference back to any national standard, which is my gripe all along. Note the get out clause on line 4 'All of these are for guidance only and should not be the sole basis for any designs. It appears OK as far it goes but it makes no mention for instance of reinforcement around big holes in the barrel for domes.

BS5500 is based around 0.2% proof stress at the appropriate working temperature and requires a factor of safety of 1.5 between working and proof stress. However it also requires a test pressure of 1.25 times working pressure. This means that at cold pressure test the stress would be proof stress/1.2. According to the Copper Development Association the cold proof stress of annealed copper is 9370 psi, so stress at pressure test should be < 9370/1.2= 7808 psi, and therefore stress at working pressure should be <7808/2=3904 psi, this is less than that suggested in the spreadsheet. Does this matter? Well probably not, the stress at test pressure (8000 psi) would still be comfortably below proof. The proof stress at the temperature corresponding to 100 psi is 8909 psi, so using my 3904 figure the factor of safety at working pressure is 2.3, which is well above the 1.5 required and so there is no need to have any arguments about whether to make design pressure higher than working pressure to allow for safety valve accumulation. The Australian code quotes 3771 psi. In the interests of international harmony I'd plump for this figure, then if you emigrate and take your loco with you it might still be acceptable. And yes I do know that BS5500 is superseded, but it's the only one to which I had access. I haven't checked the spreadsheet values for stay spacing, but have no reason to doubt them.

There are some aspects of the Aussy code which I find restrictive, for instance the ban on end to end longitudinal stays silver soldered both ends, but I can see why, it is very easy to finish up with such stays slack due to differential heating. Don't ask me how I know. They also ban connecting the top and bottom water gauge fittings by anything other than the gauge glass. I do see the argument, but I'm not sure how valid it is.

The current SFED position is illogical, it requires justification in some circumstances but doesn't mention what stress is acceptable. All I'm agitating for is for the SFED to come out of the woodwork and tell us. Complicated calculations are not needed, the tables in the Aussy code or a simple spreadsheet would suffice. It could be produced in both imperial and metric very easily. I would offer to do it, but it would need to be endorsed by SFED, I don't have Professional Indemnity Insurance any more.

This is all aimed at designing boilers which will pass the 2*wp test, you'd be a bit sick if having invested hundreds of £ and a lot of effort it failed. Again I know this doesn't happen often but we had one in our club years ago where the crown came down.

Edited By duncan webster on 20/01/2021 16:14:26

[nigel jones 5Participant @nigeljones5]

Duncan – this is the spreadsheet which those of us in the industry have used for a long time – I did know who originated it but have lonf since forgotten. You cant use your download as it is Locked for Editing. Assuming you have excel as soon as you open it it should tell you this at the top – click unlock and your good to go. I have used it on every boiler I have designed (quite a few!) and it has never failed to produce a boiler which did what it was supposed to do. That said there is an anomoly with one of the stay calculations but it over compensates so is fail safe. I think the biggest problem is, even with the data/formulae that knowing how to implement it requires considerable knowledge and understanding of boiler making. Once I have the data required I apply it in a cad drawing using stress zoning for every stay and flat area. Only then can you clearly see if youve covered all areas of the build – sounds complicated but its very simple if you can use basic cad.

[Nick Clarke 3Participant @nickclarke3]

Posted by duncan webster on 20/01/2021 16:11:26:

Dave, if girder crown stays are not banned they are allowed, there is no room for inference in standards. You just need to justify them

Edited By duncan webster on 20/01/2021 16:14:26

A long time ago when Simplex was first being described in ME and the suggestion that girder crown stays were an issue was being discussed a letter was published in the magazine from Reg V Wood on behalf of the AMBSC basically saying that they did not disagree with plate stays as such – rather that they were unhappy with the factor of safety used in the boiler design. (vol 143 issue 3558)

Of course that was years ago, and things have almost certainly changed since then, but the type of stay was not originally an issue.

Edited By Nick Clarke 3 on 20/01/2021 18:13:10

[duncan webster 1Participant @duncanwebster1]

I've got it to run, it is to do with macro security setting in Libre calc.

Fizzy, I'm not trying to argue that there is anything wrong with this or any other definition of acceptable stress. I repeat yet again, none of them (apart from Aussy) say what they are based on.

I'm not even trying to argue that a code should be mandatory, as I've said before if it passes 2* wp test there can't be a lot wrong with it, but how is a boiler inspector supposed to assess the calcs which SFED want without a recognised criterion? if designed up to the Tubal Cain limit, it would fail the spreadsheet limit, if designed up to the spreadsheet limit it would fail the limit suggested by Keith Wilson. The Aussy code tells you how to do all the sums in a small number of pages, with tables to avoid most of the sums. It is not terribly restrictive, I'll put money on it that Fizzy boilers would pass without even having seen one. If we want to use this spreadsheet just let the SFED endorse it.

I've now said my piece repeatedly, so I'll retire gracefully from the field.

[Roger BestParticipant @rogerbest89007]

My professional specialism is Nuclear crane design, so you would have thought that I would have something useful to say about design codes and standards.

Not a hope.

They are written by committees, so satisfy no-one. The committee rarely wants to take full responsibility so they leave lots to the designer, and its usually impossible to correlate between a prescribed design feature geometry, a degree of stress and the probability of failure.

At best any design code is a hint of Relative Good Practice for the observance of idiots to reduce risk to the public.

Its great to see people talking about FEA and parallel calculations. That way any anomalies and errors are likely to come out.

One more point, you need a four times safety factor against failure for good safety in a variable load environment. Buildings can make do with less as their self weight is the dominant load. Aircraft fly with this sort of factor to allow for a blustery flight. A 2X hydraulic test should be well away from causing damage if the design is that sound.

[Anonymous]

Posted by Roger Best on 23/01/2021 16:14:44:

One more point, you need a four times safety factor against failure for good safety in a variable load environment. …………………..Aircraft fly with this sort of factor to allow for a blustery flight.

Not so; if they did they wouldn't fly as they'd be too heavy. For sailplanes limit loads, including gust loads, are calculated and then a factor of 1.5 is applied to get an ultimate load. The structure should not permanently deform at the limit load and not break until the ultimate load is reached. The loads are also speed dependent so each aircraft will have a flight envelope and speed restrictions such as maximum rough airspeed and maximum manoeuvring airspeed. Below maximum rough airspeed a sharp edged gust will stall the wing before the limit load is reached. Above it something may deform or break. Likewise below maximum manoeuvring speed full control deflection can be applied without reaching limit loads. Above it maximum allowable control deflections reduce to about a third of full deflection at never exceed speed. Applying full up elevator at never exceed speed will most likely pull the wings off – it has been done.

Andrew

[Anonymous]

That's odd, there was a following post made that simply said "deleted", presumably by the poster. But it now seems to have disappeared. Is it policy to remove posts where the poster has thought better of it and deleted the text?

Andrew

[Nick Clarke 3Participant @nickclarke3]

This bit from Keith Wilson's description of his Bulldog/Dukedog boiler may be of interest (ME 16/5/80)

A subsequent letter suggested the value of 5000 be substituted for the 7000 above.

I wonder if the relative safety of model boiler designs is down to the high factor of safety, rather than detailed design.

[SillyOldDufferModerator @sillyoldduffer]

Posted by Andrew Johnston on 24/01/2021 09:18:53:

That's odd, there was a following post made that simply said "deleted", presumably by the poster. But it now seems to have disappeared. Is it policy to remove posts where the poster has thought better of it and deleted the text?

Andrew

Not policy, but I occasionally delete empty posts. I often remove duplicates and rotate photos for the same reason – it makes threads easier to read.

It's a bit risky, so I don't always intervene. No mistakes allowed whilst deleting because it's permanent and rotating photos can mess up the text.

Dave

[Author]

Posts