Boiler thickness and pressure

Im fast aproaching completing my 71/4 invicta. Now moving to the boiler design phase. Im planning on using copper for this, the thinner the better as far as construction ease goes (TIG Welded) so I put the following for discussion:

ID is 5.625"

Using 3125lb/in (safety factor x8) and 16g tube I can run WP 70psi.

Is 70 psi likely to be enough pressure or will I need to calculate this also?

Can i justify a lower safety factor?

Thanks

Just worked it out quickly using the formula John Haining gives in Countryman's steam and get 0.128" which seems about right as most engines in the 5" dia barrel tend to be 10swg or 3mm

(OD x WPx 8) / 16000 x 2 x 0.8

Where the 0.8 is temperature allowance, 2 constant and 16000 his tensile for copper, same 8 safety factor, 70psi pressure and took OD as 5.75

J

EDIT or work it back using your 16swg and his formula gives 35psi WP, this is the same formula that Greenly suggested.

Both based on drawn tube not roll your own

Edited By JasonB on 20/10/2014 17:43:09

so did harris get it wrong or me?

I also did it with Harris's figures which use the same formula and similar stress as Tubal Cain and get 0.062"

The 16000psi is a minimum to use so once you start to increase that the thickness comes down. There is also the fact Haining is a Traction engine man where we have no frames to take the load off the barrel. Harris used 25000psi

On the other side the one Loco book I have by Martin Evens suggests 10swg or 1/8 for a boiler with a barrel of 5-6" dia.

Is tube that size available in 16swg? if not then you are going to have to figure in an allowance for the barrel joint which may lower your WP or up the thickness which would need to be done as you will probably have to buy 2mm sheet.

Edited By JasonB on 21/10/2014 07:55:51

I can get 16swg @ 51/2 inch. If I couldnt then I would roll it and tig the joint so there will be no need to add in any other stress factor. I seriously dont want to use 10swg as I cant weld it, and although I have oxy in the workshop I am loathed to pay out for silver solder. Obviously I need to justify why I think 16swg is safe to the inspector before I start and I dont think it will get far on 35psi! 62 thou is exactly what I wanted to see for WP 70 psi, I now need to source deoxy copper plate?

Yes I would like to know what the inspector says. After all his word is final. Would hate to think of someone reading this thread if it stopped here and going into their shop and making a thin boiler, please let us know the outcome.

When I first saw your post it just seemed way too thin to me but why do we have two formulas that give 0.062 and another two that suggest twice that?

Interesting that you say you won't use a factor to derate the barrel if it is welded, haining suggests 0.8 for welded copper which would take you from 0.062" upto 0.077" and anything as much as 0.35 for a single butt weld on a steel boiler.

J

Indeed Jason, and I greatly apreciate your inut - I have to question why would one add a strength factor to a joint which will undoubtedly be at least as strong as the parent metal, dare I suggest that the person writing the instruction had no idea about welding copper (which will be ndt and certified before use), or was being cautious as testing was far less strict when first written, hence the possability for it to be welded by a complete goon? Since the Harris formulae has been in print for a very long time and must have been used on hundreds of boilers I fail to see how any inspector could refuse to accept it. Sure it looks thinner, but it either conforms to accepted standards or it doesnt - however, I produced a steel boiler a while back which did not require any crown stays yet was told to add them 'so that it looked right'! Im spending the rest of the day flint napping !!

Having designed boilers for Cammell Laird and other boiler makers I think its fair to say Haining new a bit about boilers, his book is also the only one I know of that deals with model steel boilers, weld prep, joint design etc.

Does the NDT only show that the weld is sound even if the joint design could be totally wrong? Haining gives several different factors to be used depending on type of weld and whether heat treated afterwards

One would also have thought that by doubling the dia of the original Canterbury Lamb you should at least be looking at double the boiler thickness unless that was thicker than theory for some reason.

J

Edited By JasonB on 21/10/2014 11:52:38

Using Martin Evans numbers I get minimum 0.082" wall allowing a .9 factor for the Tig butt weld.

Silver solder lap joint would be 0.8

Steve

thanks all. I refer to my point about using harris formulae...used many times by many people. I used to weld on BNFL projects and they didn't add a safety factor for joints, but that assumed you were qualified, which I was. Hence is the safety factor there to compensate for non professionals welding? All very interesting by the way.

Check with your boiler inspector. A lot of wasted time and money if he will not certify it.

For information - taken from PED guidance notes (as an indication of industrial practice).

For welded joints, the joint coefficient must not exceed the following values:

for equipment subject to destructive and non-destructive tests which confirm that the whole series of joints show no significant defects : 1

for equipment subject to random non-destructive tests : 0.85

for equipment not subject to non-destructive testing other than visual inspection : 0.7

So basically, more thorough testing gives the potential to use higher joint factors (hence thinner material).

I used to work on a major chemical site, for much of the pressure piping 100% radiography of all welds was quite common (in addition to certified materials and coded welders).

I don't know anything about model boilers, or clubs, although I have built a couple of small boilers ( just for myself). My point is that when I worked on real boilers and pressure vessels at different companies, the design criteria (eg. safety factors ) and testing regimes varied. But at the end of it all it was the insurance companies that dictated what was done. In effect this sounds just the same.

I have some tube, 0.043" wall thickness, nominally 2.625 diameter I bought as 'boiler tube', without a specific project in mind.

Tubal Cain's formula gives a working pressure of well over 100psi (it gives 124psi if you use 3800 as the 'working stress' at 100psi)/350fahrenheit), which seemed very high to me. I note that the likes of LBSC always seem to specify 16 gauge tube for the smaller diameters.

My worry is not that this tube is too thin for a pressure of, say, 60-90 psi which would be the max I would use it for, but that it wouldn't be strong enough to stand up to normal handling, or even silver soldering, especially for a 'long' boiler.

My thought was to use it to make a small, gas fired scotch boiler suitable for a small steam boat.

Any comments?

Neil

Surely, if it isn't built in accordance with the specifications of the design for the engine it isn't going to pass the boiler inspection.

What do the drawings say?

Lofty, Its a one off so no existing drawings at that scale.

Neil as you say the theoretical thickness and what is practical do seem to differ, you don't want to end up with something like a Coke can that can stand a lot of pressure when shaken up but will collapse in your hand

Neil - which takes me back to the fact that if you use the equation I have which has been in print and used for many years then you have no problem. A lot seems to be down to perception and (dare I say) old school thinking. I have in the past experimented with ultra thin boilers with some interesting results. I made a 4" dia boiler shell out of 22g and tested it to 600psi. It sort of resembled a rugby ball when at full pressure but the silver solder joints held 100%. Ive also recently tested Screwfix plumbing tube with an external pressure of 600psi (thats as high as I can go) with no effect. Im sure your .43 tube would solder well and would be perfectly fine for this application but I would opt for much thicker end plates as in my experience this is where the weakness will show.

That should read 0.043!

If your model is to be insured and steamed in public, should it not be the case of establishing what design stress, joint factor and safety margins are acceptable to the boiler inspector and insurance company. Then the design can be worked out to meet their requirements and subsequently approved by them. In other words what code of practice do they accept?

Although you have done some testing to justify to yourself the safety of your design these are unlikely to be taken into consideration by your boiler inspector or insurance company unless very well documented and maybe also witnessed.

I am not familiar with the formulas stated above, the following table shows calculated tube thickness using a standard hoop stress formula (also typically used in pressure vessel codes). I have used some values suggested above for yield stress and joint factor and included a range of safety margins for comparison. Normally you would establish a design stress based upon material mechanical properties and operating temperature and use this in the formula.

Other considerations concerning boiler tube thickness will be a) does the boiler tube support any other significant structures like cylinders etc. b) what reinforcement is applied to openings/nozzles.

One method concerning openings/nozzles is to use an 'area replacement method', basically you take a section through the opening/nozzle, calculate the area of material removed then ensure that any reinforcing ring has at least the same section area. If the boiler tube is thicker than theoretical then the extra thickness above theoretical can be included in the calculation, for example, if the theoretical thickness is 0.05" and the actual thickness is 0.15" then there is a good chance that a reinforcing ring is not needed as the actual tube thickness already has sufficient strength. There is a specified method for determining the width of any reinforcing ring, I would have to look this up.

Here's a useful source of information.

I hope this is useful info.

Nigel

From one nigel to another, thanks! So from this table I could use 60thou tube with a x3 safety margin? So many variations on this, that one has no way of knowing what is correct.