Why we test boilers!
|Nigel Bennett||19/07/2019 14:30:56|
|296 forum posts|
Some years ago I bought a part-built 2" scale Fowler BB1 ploughing engine to John Haining's design from the estate of a deceased model engineer. He'd bought it from somebody else. The workmanship wasn't brilliant, but the boiler had been commercially built in about 1975 and came with a certificate, saying it had withstood 175psi without leakage or distortion. The boiler itself looked lovely, so I was quite happy with the purchase. It was always going to be a long-term project, so I was in no hurry to get on with it. (Ignore the bloody great hole in it in the picture below!)
There were a lot of further leaks from stays (screwed and silver-soldered) so I applied more Comsol. A couple of days ago I pumped it up again, and it withstood 150psi with only some very slight weeps from the odd stay. One or two responded to a bit of gentle caulking with a hammer and punch. Then a small leak manifested itself in what was apparently a plain section of boiler barrel, just in front of the firebox. Curious... I lightly tapped the area with a hammer and the leak became a torrent. I filed the area and a brass colour became exposed. The barrel had been brazed to the firebox, and then smoothed over before copper-plating the boiler to disguise it.
That piston ring was only serving as a location for brazing the two halves together!
These are the two sections of barrel/firebox wrapper removed by hole-sawing - and the braze had given way during the operation.
The whole thing had been a potential bomb. The scary thing is that if the joint had been only a few percent better than it had been, it wouldn't have leaked and I would have been blissfully ignorant of the potential. And it's a ploughing engine - it's not just pressure stresses on the boiler, but driving it over rough ground imparts who knows what additional loads on the joint?
There were other concerns as well, such as very thin fire tubes - they'd have worn through very quickly - and the internal pad under the cylinders was not in the right place (or too small) so the front ring of cylinder bolts were only tapped into 2mm of copper.
I'm not prepared to disclose the name of the maker. Technically I can't prove anything because there is no marking on the boiler or certificate. Suffice it to say that he was later bankrupted. And I hope he had a really miserable time of it.
At least there's no chance of it being used - it's well cut now and I shall weigh it in for scrap.
Moral: Take care when dealing with older boilers built by unknown "professionals", and look at them very carefully before parting with your money. If a boiler has been copper-plated, walk away from it. It's hiding something. And always, always test a boiler you wish to use.
|Nigel Graham 2||20/07/2019 23:49:35|
|363 forum posts|
Or when dealing with drawings from unknown professionals....
I once helped prepare for testing a newly-built boiler whose wide firebox shape suggested an LNER locomotive, in 3.5" or 5"g. The metal-working including the silver-soldering was immaculate, but inverting it revealed the inner firebox looked more MacDonalds than Gresley, so badly crushed-in was the crown.
Closer inspection revealed only about 8 or so stays, and only in the back-head and throat-plate.
The builder was a retired coppersmith on his first foray into model-engineering, so he had decided to start with the boiler as the simplest part for him, whilst admitting not knowing much about locomotives.
He assured us (club members) he'd tested it hydraulically but only to working pressure. We believed him, but realised he'd not have known to look for more than just dripping water.
Fortunately he had the drawing with him. It was a low-quality photocopy significantly lacking any designer's or publisher's name or other source details. The poor chap had followed it accurately, not knowing it was so flawed. We had explain to him why the design was dangerously poor, and break it to him that the boiler would not be repairable.
We wanted to write to ME with a warning; but for some reason he would not reveal where or how he had acquired the drawings. They might have been made privately by some past model-engineer for his own project, perhaps from railway-literature drawings, but who had a weak knowledge of design principles.
We never saw him again.
|599 forum posts|
A newly constructed silver soldered boiler usually has the strength integrity of a piece of chocolate until it has been fully work hardened, normally by carrying out the hydraulic test in stages and slowly building up to the required pressure.
Pumping a new boiler up to working pressure in one hit will almost certainly cause deformation in all sorts of places.
|Dave Halford||21/07/2019 11:29:25|
|441 forum posts|
Makes you wonder if boiler and certificate got married sometime since 1975, I've seen boilers that bad on ebay, they can be a handy source of cheap castings or gears when the builder realises they do not have the boiler skills required.
|696 forum posts|
The applied pressure shall be TWICE Working Pressure (2xPW) for the initial hydraulic shell test and ONE and ONE-HALF times Working Pressure (1.5xPW) for system hydraulic tests. The pressure shall be applied gradually and increased in steps of not more than 10% once the pressure exceeds the Working Pressure. For an initial hydraulic shell test on a soldered or welded copper boiler, the pressure should be increased in small steps from zero, with intermediate relaxations, to allow stress redistribution and work-hardening of the very soft annealed metal.
|Kiwi Bloke||21/07/2019 11:55:37|
|227 forum posts|
May I suggest that the above extract from the boiler regs is published regularly in ME? Whilst I would expect boiler inspectors to know this, how many builders would know to first pressurise their newly-built boiler this way? Not me, certainly - although I'm not into steam.
|old mart||21/07/2019 18:57:44|
|453 forum posts|
My only experience of live steam, was as a teenager, getting hold of a Mamod oscillating, very basic, stationary engine. The first thing I had to do was replace the rubber seal in the filler/safety valve, the blow off seal, that is. There was no way of checking the actual pressure the valve would work at with my carefully cut 1/8" long section of bicycle tyre inflator hose. I bent up a length of copper tube and soldered it to the exhaust port and through a hole drilled in the dummy chimney.
I'm pretty sure the boiler was made of brass, I thought that de-zincification only occurred in a salt water environment, which is why bronze is used.
|Neil Wyatt||21/07/2019 19:26:43|
16449 forum posts
The problem with publishing extracts is that there's an awful lot more boiler builders should be aware of; if they read a brief warning someone will say 'well I did 'X' so it should have been OK!'
The effect of tales like this should be to persuade people to seek DETAILED guidance on construction and testing.
Edited By Neil Wyatt on 21/07/2019 19:27:32
|696 forum posts|
Neil is obviously correct, the whole document should be read, so apologies for my post.
Please do not read it in isolation.
It was the part about work hardening that I was interested in as was not aware of this aspect.
|old mart||22/07/2019 14:24:50|
|453 forum posts|
The work hardening aspect is interesting. Copper work hardens by stretching, so how much smaller do you have to make the boiler for it to be the correct size after testing?
|599 forum posts|
Never had a boiler fail to fit its intended purpose yet, then again if you build and test the boiler first, everything is machined to fit.
And you are only talikng a few 'thous' of growth if you test properly.
There is always a difference in sound after testing too - before the metal will sound dull if you gently tap it, afterwards it will 'ring'
|CuP Alloys 1||26/07/2019 07:42:12|
192 forum posts
The strength of a brazed joint in a boiler is not going to be influenced much by testing upto 150 or 200 psi.
Check out the strength of annealed copper. It is way above the operating stresses of a boiler.
The model engineer has a far greater influence. In order to achieve high integrity joints he has to ensure that all parameters are met to create a sound joint. And that means creating the right conditions for capillary flow.
Is the joint gap and length correct?
Is the joint design appropriate for the operating conditions?
Is the right flux being used to meet the melting range of the filler metal, to remove all the oxides present and the time taken to melt it?
Is the joint being heated correctly to promote capillary flow and ensure penetration of the filler metal into the joint?
Is the appropriate grade of copper being used and not a cheap equivalent?
The responsibility for the quality of a brazed joint lies solely behind the torch, not in front of it!
Isn't the technique of gradually increasing stresses during testing simply one of prudence? Does anyone, when testing any new machine,start by winding it up to full bore immediately? If anything goes wrong the ramifications are often less severe. Increasing the loads gradually usually leads to increased confidence.
That is certainly the case with a boiler. It has nothing to do with increasing the strength of a brazed joint. But, I repeat, if it makes you feel good......
PS The source of my information is a metallurgical degree and 50 years spent trying to ensure that every brazed joint is a sound one.
Please login to post a reply.
Want the latest issue of Model Engineer or Model Engineers' Workshop? Use our magazine locator links to find your nearest stockist!
You can contact us by phone, mail or email about the magazines including becoming a contributor, submitting reader's letters or making queries about articles. You can also get in touch about this website, advertising or other general issues.
Click THIS LINK for full contact details.
For subscription issues please see THIS LINK.