Why are copper plates in boilers flanged?
|Bob Worsley||29/06/2020 12:21:38|
|57 forum posts|
Re-starting model engineering and need a boiler.
Got and read the various ME books such as Tubal Cane, Martin Evans etc. First gripe, haven't they heard of indexes? What is wire drawing?
It is clear from these that the clearance between copper surfaces in the boiler are of critical importance when silver soldering, something like a few thou.
Never made a boiler but done some limited, evening class, copper bashing and seems to me that the hammering needed for a 3mm copper plate to flange it results in anything other than a few thou close fitting joint.
One example I am working on is a part complete Clayton by Robin Dyer. Nice and simple with just two plates, top and bottom. But they are both, particularly the firebox end, with the tube holes very close to the flange with consequent pulling of the metal and hammer damage.
Why not just use a piece of 5mm or 6mm copper plate and just bore the holes, turn to exact size on the lathe, and solder? Ok, the copper cost will be higher but in the whole model that isn't really very much, so ignore cost.
Yes, copper is sticky and not the easiest metal to drill. But use a lubricated step drill, I find these are amazingly useful drills and solve endless problems, probably the greatest one being the end result is circular! Or could bore the holes on a mill.
The soldered joint needs to be in shear, from books, so the wide plate joint will be in shear. In fact exactly like the foundation ring. The actual area of solder will be the same for plate and ring, so no hazard there.
Inspecting the joint is equally impossible for flange or plate, no extra hazard there.
The next model wants to be a traction engine with a loco type boiler. Here again I can't see why the plates need to be flanged, a solid 6mm plate will be just as good. Yes, there will be a reduction in the water space around the firebox, but just make it slightly smaller to maintain the 3/8" or whatever gap. There is the throatplate, double flanged, but could be two 6mm plates brazed, not soldered, together with one for the boiler, one for the firebox. Again, cost but seems a cheap solution to hammering and bashing a poor piece of copper sheet.
Any one already doing this?
18885 forum posts
Quite common to put your flanged plate onto the lathe and skim the diameter down to a suitable fit in the barrel and no hammer marks left.
Flange before drilling the holes then they won't get pulled or damaged.
Speak with your boiler inspector to see if he is willing to accept thicker non flanged plates and if so also get a bug burner to heat them when soldering.
6329 forum posts
'Wire drawing' is an effect caused by anything impeding flow in a pipeline. The idea is a thick wodge of gas going in is repeatedly swaged down like a wire making machine to produce a thin weak stream of gas at the output that starves the cylinder.
Superb boilers are wasted if steam can't freely get to the piston due to a badly designed or built steam delivery system. Small diameter pipes, dirt, rough surfaces, tight bends, long runs, kinks, joints, valves and small or badly shaped steam ports all contribute to 'wire drawing'. The regulator throttles steam flow deliberately, everything else is bad. 'Wire drawing' covers accidental throttling which can be considerable and is well worth reducing as much as possible.
|J Hancock||29/06/2020 13:17:29|
|452 forum posts|
Assemble your boiler in the wrong sequence and I will guarantee it will fail a pressure test.
Alec Farmers book is the only one to read.
|Nick Clarke 3||29/06/2020 13:25:16|
884 forum posts
Jason's advice, to talk to your boiler inspector is the correct one, but he would be entitled to see new drawings and all of the calculations for the boiler as it is not to a published design (but in practice he might not)
I don't know what the joint between sheet and your thick plates would be like - would there be a tendency to crack? - an expert needs to comment here.
|Paul Kemp||29/06/2020 13:37:53|
|560 forum posts|
Bob, main problem I see with your plan is as you identify cost but in more practical terms unequal expansion. Whilst it is true expansion is linear and to a degree not vastly important on thickness but over the area of the plate, when heating up (raising steam) the thinner sections will heat up quicker and thus expand quicker and when cooling down the thinner sections will lose heat faster and contract quicker. Whether this is a problem in very small boilers I am not sure but certainly in full size boilers by raising steam too fast (or forcing as is usually the term) considerable damage can be caused by the stresses to stays and plates. This is the reason on a full sized boiler steam is raised over many hours, often with a small warming fire being set the day before to bring the temperature up slowly and as evenly as possible. Whilst a flanged joint is double thickness at the joint it is a relatively small area of the whole (and the plates need to be joined somewhere somehow. The technique of thick plates is sort of the approach taken in a miniature steel boiler although there most of the plates will be a very similar thickness and the joint will be a full penetration weld with root and capping runs and may even be back gouged and capped on the inside as well depending on access and sequence of assembly. Foundation ring is usually the same section all the way round and is at the cooler end of the boiler (below the fire) anyway. Just my thoughts.
As Jason says it is normal to adjust the flange faces either by machining, filing or other hand work to maintain the requisite clearances for the solder. It can be a mistake to have clearances on say tubes too tight because the silver solder if the clearance is tight will not flow through the joint. Seen several loco boilers fail at the smokebox tube plate / tube joints where they have been built like this and there is only a fillet of solder outside the tube. The tubes being thinner than the shell and heating up quicker by being exposed to the "fire" want to grow quicker than the barrel which is heated by the water and the stress on the joints of the tubes to firebox and smokebox can be quite high.
|Andrew Johnston||29/06/2020 14:28:03|
5665 forum posts
Not necessarily. Wire drawing occurs when steam is throttled and the input and output velocities are similar. This means that the expansion is isenthalpic, ie, enathalpy stays constant although the pressure drops. Since the pressure has dropped but the enthalpy stays the same the output steam is drier, or even becomes superheated. That may be an advantage as condensation in the cylinder will be reduced and hence less energy lost due to condensation and re-evaporation.
With Stephenson's valve gear, at short cut offs, the input ports are only partially open so wire drawing is almost certain to take place, which may, or may not, be an advantage.
Losses due to poor passage design and sharp changes of direction are different.
|duncan webster||29/06/2020 15:12:40|
2795 forum posts
To amplify Paul Kemp's point on expansion, the innards of a boiler get hotter than the shell, not by a lot, but there nevertheless. Unless this expansion is accommodated high stresses can be built up. This is why full size boiler design codes require 'breathing' space in the flat end surfaces. Making the end plates thicker makes them stiffer. Not a good thing.
Its easy to skim a smokebox tubeplate, less easy to file a backhead but worthwhile as the outer profile will be slightly tapered due to circumferential compression as the plate is formed round the curved profile.
|Bob Worsley||06/07/2020 20:05:52|
|57 forum posts|
Short delay while I bought and read Alec Farmer's book as suggested. Yet another person who is incapable of creating an index!
Thanks for the comments on wire drawing.
I see that it uses oxy acetylene for all soldering work. The other boiler books are all very wary of this, and some ME articles positively say no. Basically the flame is too hot, and can/will result in the alloy metals being boiled out. Possibly of more importance with cadmium alloys.
To reply, not answer, comments. Copper is a very good conductor of heat, and I don't have the expertise to say that heating the water will result in an x deg C temperature drop along the copper. Many experiences with a soldering iron and far thinner wire than a 3mm sheet of copper suggests this speed of conductivity is high. So I really don't see that using a 6mm tube plate without flanges is in any serious way different to a 3mm flanged plate. What about around the firebox? Again don't see that it is important, especially since the foundation ring could be 1/2" square copper. Also the firehole ring is a piece of far stiffer metal than the surround, don't get fracture failure around them.
The shell will expand. But all the descriptions of boiler testing are that the whole boiler can, probably will, move slightly as the pressure comes on. The metal will move until enough is there to take the stress, the strain in the metal is minimised. With firing of the boiler hopefully this movement won't reoccur, but if it does then the copper is as soft as it can be, and will adjust. Again, don't see this as a problem.
Other comments are that the instructions to solder a boiler very frequently say that it is the firebox plate where the leaks are. Could be for a number of reasons. 1 - There is very little metal left after the tube holes are cut, so could be partially melted or other nasties. 2 - It is always a tricky shape, not like the smokebox plate, so can't just stick in lathe and machine the flange. 3 - Because it is hand work to get the plate to match the firebox then who is going to claim their filing is within 5 thou over a distance of 10" and a complicated shape. 4 - And how are you going to justify that claim.
This is why using a thicker piece of metal, no hammering and banging, has got to be an improvement.
18885 forum posts
The other problem with thicker plate that has not been mentioned is the room it takes up. I don't know what size boiler you are thinking of but at 3mm sounds like 125-150mm dia which is around the point where steel and copper overlap and one of the big disadvantages with steel on boilers of this size is that to maintain the outer dimensions you either have to reduce the width of the water space or the grate area. won't be quite so bad on your copper boiler as its is only front and back that are affected and not the sides but you could be looking at a loss of 15% grate area which will affect the steaming ability of the boiler
Edited By JasonB on 06/07/2020 20:34:38
|Paul Kemp||06/07/2020 22:12:43|
|560 forum posts|
oxy acetylene; no real problem with this if you know how to apply the flame to the job, yes you could overheat either the parent metal or the filler but if used correctly you won't.
Diferential expansion is less about the speed of conduction and more about the absorption. A thicker section takes more heat (energy) to raise the temperature a given value in a given time than a thinner section. You don't need to be able to calculate it you can do an experiment with your soldering iron, a thermocouple and a stop watch.
Older copper boilers do develop leaks around the fire door ring / section and around the foundation ring. Agreed model boilers less likely to develop cracks in the backhead / inner firebox due to the malleability of the copper (although copper does work harden even in a boiler). Usual source of leaks in firebox tube plate, smokebox tube plate, firehole and foundation ring is joint failure.
Yes the shell will expand, but not as quickly as the firebox, only source of heat to the shell is by conduction through the water. The whole boiler will expand and yes both as a function of heat and pressure, my 4" scale TE boiler expands about 3/32" between 0 psi and 120 psi but that is steel and only an infestiminal amount is due to pressure, the rest is heat.
If the tube plate does get overheated in my experience as per cracking and quilting of the firebox sides and crown it is due to a boiler not having been maintained correctly and a layer of scale deposited which is a good insulator and stops the heat conducting to the water or the water level has been allowed to get too low!
I will claim to be able to file a piece of metal to within 5 thou, one of my first jobs as an apprentice was to make an engineer's square calibrated to a master, manufactured using only hand tools in fact the calibration was less than 2 thou and as to irregular shapes we had to produce other pieces that fitted inside each other supporting their own weight and with no light visible through the joins.
Dont get me wrong, there is absolutely nothing wrong with questioning a process that has always been done in a particular way as to whether that is the correct way. However there is usually an underlying reason why it has been done the same way for many years! Feel free to construct / redesign your own boiler in any way you wish, there is nothing stopping you. However to get it certified under either the MES boiler scheme or a commercial examination as it will not be to a published or established design you will be expected to present calculations to justify your method. I would suggest you run your ideas past whichever boiler inspector you intend to use for certification before and during construction to avoid any disappointment on completion.
All the best,
|Brian Baker 1||07/07/2020 07:46:17|
134 forum posts
Greetings Bob, as a general comment, I would say that indexes cost money to prepare, and take a lot of time.
Re boiler plates, the practice that you propose is done on small boilers, ie pot boilers, but you must remember that the boiler regulations, custom & practice, have produced a system of building that is safe for amateur construction. The flanged plate method produces a joint, which even if not completely covered in solder, will be safe, and if there is not enough contact, the boiler fails the pressure test.
It is generally accepted that the flange length needs to be three times the plate thickness, so assuming that you are using 3mm plate, then the tube plate that you propose needs to be 9mm thick at least, which may a bit pricey, as well as more difficult to silver solder. If you are buying that size material, then it might be easier to turn the plates from the solid. This last remark was said "tongue in cheek" but I have seen a firehole door ring made that way.
The whole point of this rambling is the word "amateur".
|Phil H1||07/07/2020 09:45:14|
|315 forum posts|
To me, your main issue seems to be related to your technique for the flanging of the plates and as I read your messages etc - it simply sounds like you are a bit frustrated with the flanging process and you are looking for what you think will be an easier alternative.
For example - your mention of hammer marks? You can flange the plates easily with a mallet and a few bits of wood for internal corners if the copper is annealed frequently. I have seen a hammer being used to finish the flange edges off but again, the aim is to get a nice flat edge without prangs.
The right sized steel former and a bit of patience will soon achieve a close fit.
As somebody has already mentioned, you bore the holes after the flanging - so no stretched holes.
Oxy acetylene is perfectly suitable for silver soldering. By far the best boiler I have ever seen was made by one of my apprentice instructors. It was a work of art (the soldering) and all done with one type of solder and oxy acetylene. I only saw the finished job but I wish I had asked hundreds of questions.
0.005" over a length of 10"? Yes, like Paul, I will claim that I can file far closer - easily. It is just a bit of patience that is required.
|135 forum posts|
Hi all. I agree with Phil, If the flangeing is the issue then it is easily enough resolved. I have had to double flange the firebox front plate and the throat plate for a design that uses an elliptical combustion chamber.
Firstly,flange before drilling for any holes . make up inside and outside formers to the shape you require, easy lathe work for circular. on the mill for the elliptical, and a third former to go inside the flange and sandwich the plate between them. If you are double flanging, which it doesn't seem you are; you will need four formers in total. If you have access to a hydraulic press just put the whole sandwich of the two inner formers and the copper in it and press the outer former over it or through it as the case may be. . You may have to anneal a few times but the initial bend will keep the formers in the right place. place I use simple hydraulic press made from an old truck bottle jack and they come out nicely formed , and with the minimum of dressing work needed to clean up to an edge ready for soldering.
If you can find it, I suggest you read an article " boilermaking made easy" by Jack Coulson I think it was in MEW or EIM but not sure, It adds to, but also gives some very good alternatives to Alec Farmers view. However one of our members will no doubt be able to tell you where to find it. I may have an electronic copy somewhere but it may take me a while to find it. If I do I will try and send it to you.
Bear in mind that having Acetelyne at home will completely invalidate probably 90% of all house insurance policies, which is why I got rid of mine a long time ago.
Happy flangeing !
|Bob Worsley||09/07/2020 10:54:27|
|57 forum posts|
Thanks for the continued interest, or exasperation!
If you are skilled then, as people say, use it. If you are not skilled then steer well clear. After reading books and articles on the use of oxy then that is my conclusion. The problem is that the flame temperature is so hot, 3,300C, and the boiling point of copper is 2560C. So the chances of overheating rises to, what, 100%? Look at the tubeplates, the firebox one is hard to see with lots of very thin copper. Gas is 1900C so can't melt and the flame size spreads the heat so even oxidising is unlikely.
The snag with reading older books and articles is that things change. Silfos was recommended ears ago, now it isn't, with large warnings of sulphur effects in coal fired boilers if the temperature exceeds 200C.
What troubles me is providing the guarantee that the flanged joint is consistently the required 5 thou. If you are making some steelwork then you use bluing paste to make sure that the whole area that needs to be in contact is in fact in contact. This is easy on the tubeplate, but that can be machined anyway, it is the firebox tubeplate and the springiness of the joint. This is why it is suggested you tack the joint with rivets, but then how do you know if you still have the 5 thou. At this point you have filled the joint with flux and any gaps are now invisible.
Solder is strong, far stronger in tension and shear than copper. This means that with a 3mm plate then the solder only has to be 0.5mm for the same strength. So the 6 or 8mm flange has plenty of redundancy, the snag is you have no idea how much of that redundancy you are using. Is the flange width so wide simply because they are so poor and the maximum redundancy is needed? My suggestion using thick plates might bring some clearer idea of how much is being used. Many years ago I found Frank Price's book Right First Time in the library, and bought a copy, and that has been a guiding principle ever since in design.
As mentioned, using a sandwich of formers seems a very good idea to improve the quality, however defined, of the flanges.
The last jobs I had to do that were completely new to me were felling trees. Standing at the bottom of a 90' tree smartened up you risk assessment. But even more so did the cutting of wind throw and hung up branches, they look small, lots of them, just hack away with the saw. One tree about 70' high and only 5" DBH I felled, oblivious to the squirrel damage, invisible, half way up that caused to tree to fold in half as it came down. If I hadn't obeyed the rule, watch it, then it would have hit me. Likewise with hung up branches, you wouldn't believe how much weight a 2" branch can support without breaking, cut it and twang. You hide behind the stem and cut from there. Similar situations is repairing valve audio amplifiers, 600V HT is nasty. Don't really want to have a boiler experience.
|Stuart Smith 5||09/07/2020 11:08:57|
|125 forum posts|
You might find these videos interesting. They show how Blackgates Engineering make the boiler plates for their Sweet Pea boiler kit.
|Phil H1||09/07/2020 11:46:06|
|315 forum posts|
As pointed out - oxy acetylene will be difficult if you are a home user (insurance etc) - so the realistic alternative is likely to be propane.
Flanged plates? There is an excellent series of youtube clips showing a chap building a 5" gauge Lion locomotive boiler and he goes through every imaginable detail regarding the flanging of the plates. That might be worth watching before you go much further.
|Paul Kemp||09/07/2020 19:39:28|
|560 forum posts|
On the acetylene question I have often heard tell of insurance exclusions but does anyone have any specific info to back this up? Not that I have acetylene at home or any real intent of doing so in future but out of interest I dug out my home policy and there is no specific or related exclusion that I can see? Interested to hear if anyone has a concrete example of this being the case?
|paul rayner||09/07/2020 19:51:35|
|160 forum posts|
as is happens I've just been watching this guy building his boiler, fittings etc
doing a good job too (in my inexperienced opinion)
|Bob Worsley||09/07/2020 19:57:17|
|57 forum posts|
Just been given a large pile of MEs, and quickly scanning my way through them.
One from the 1980's I think had an article by Keith Wilson about how great silfos is to soldering up boilers.
Then I have just read Wardle's article in 1st Feb 1980 about a couple of silfos boilers that literally fell to pieces after about 1000 hours of steaming. I found that incredibly frightening.
The solder had been attacked by the sulpher in the coal causing it to turn to black mush. This is one of the reasons I hesitate about ME flanging plates, that the 5 thou gap might be 50 thou and I do have a bucket full of silfos (used to solder coax cables in the days before fibreoptics) and was tempted to use it. I still look at the reverse curve on a loco firebox and think, without rivets, how do you keep those pieces of copper at a 5 thou distance. Note, not more than 5 thou, or less than 5 thou, but at 5 thou. If bits of the gap are 5 thou then what percentage?
Yes, I worry, but my professional engineering was like that too, eliminate by design known possible failures.
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.