Thanks Jason- does the article state N&W or just 'a large prominent eastern' road?

Those square mudring corners must have worked like a charm - not!

Looks like p. 89 & 90 of the MBMA...

Robert Yuill

The NWHS website just uploaded these photos of Y5 and Y6 fireboxes and front tube sheet. I don't know construction details, but am merely posting the photos. I'm curious about the "bracing" at the crown/tubesheet knuckle. Is that for assembly or permanant?

That's it Mark.

The gussets at the rear tube sheet were a permanent part of the assembly.

The front tube sheet photo shows a flat (un-flanged) tube sheet attached by fillet weld, to a ring, which is riveted to the shell. The diagonal gussets were to provide some support to the large areas of unsupported plate between the tube hole and shell. With a flanged sheet, the knuckle contributed to the support of the sheet, so with out that knuckle the diagonal gussets were required.

Robert Yuill
http://www.historicmachineryservices.com

Sir Yuill,

The preceding pages spoke of the N&W as I recall..... I'm fairly certain of this as my reaction when initially reading through was, "well of course it was the N&W developing this, who else would be so progressive? " . In reference to the gussets, in yet another MBMA book (40's era) there is an article which details their use as well as the short pieces of round stock along the inner throat joint.

Cheers, Jason

Was just talking about square corners with someone the other day...

Besides the comments about weld quality control, the design issue at hand with square corners is the fact that they form a stress concentration point as was also previously mentioned. Assuming the joint was initially designed for the expected pressure stresses, if those stresses were above a certain level, over time the joint can still fail because of cyclic fatigue.

All materials have flaws in them. Some materials (steel is one) have a crystallographic structure which makes them susceptible to fatigue failure in that those flaws or cracks will tend to spread if the joint is subject to a certain stress level. Think of a paper clip. If you bend it just a little, you can bend it a little bit for a large number of times and it will still function. If you start bending it a lot, it will fail after only a few bends (cycles).

We mechanical engineers will often characterize this on a plot of number of cycles vs. stress. The curve starts off as you would expect, a large stress will fail a material with a small number of cycles. As the stress level decreases, the number of cycles to failure increases and the curve generally will flatten out so that if the stress is low enough, the material will not fail regardless of the number of cycles of loading and unloading.

Let's say railroad A operates a locomotive with square corners in a gentle manner; slow fire up times, firing practices on the road that don't expose it to quick thermal changes. It is quite likely that said square corner could last a while - perhaps even the life of the boiler. Now, if railroad B operates the same locomotive with fast fire up times or really forces the boiler when out on the road, you subject the joint to higher amplitude or frequency cyclic stresses producing a situation where the fatigue is higher and the joint will fail.

An important point here is that those stresses can be surprisingly low compared to the ultimate strength of the material because of the nature of crack propagation.

A lot depends on how well the design engineer understands the stresses that the joint will be subjected to. In design, we try to avoid square corners because they do have a tendency to concentrate or focus stresses as someone stated earlier in this thread. Rounded and smooth surfaces tend to distribute loads more evenly. Since we can never fully anticipate every source of stress or the cyclic loading expected without extensive and expensive testing, the smart move is to use geometry that is known to inherently be better suited for the loading case...in boilers and such the more cylindrical or spherical, the better as those are the geometries that have the lowest stress when subjected to pressure...

Thank you Wolf, great explaination of the nature of response to concentrated repetitive loads. As for poor little #12......I don't think you could find an adhesion railroad anywhere that is as prone to a flogging 6 times a day. She was a bit lightweight also for the service and really didn't have a chance. The Shays and consolidations are more up to it.

Good work, Jason.....trying to make it as good as you could within the shell. Sorry I didn't get to see it.

There's never been a boilermaker on staff there, just contractors from Denver and one other guy who built the botch in the first place. Come to think of it, he does manufacture small scale boilers as a business. Probably explains a lot.

dave

On the side view of the completed firebox, can anyone tell me what the series of "lumps" are that cross the butt weld seams at the sides of the combustion camber? They can be seen where the throat sheet joins the combustion chamber and where the side sheets join the throat sheet, and again where the throat sheet wraps up and joins the lower/forward portion of the combustion chamber?

Are they intended to be some sort of reinforcement for the butt weld seam?

Exactly!

From what I have read and what I have calculated, the normal range of temperature difference between the firebox sheets and the shell ranges from 200F to 300F. Obviously, this can vary widely depending upon scale and insulation condition. If you want to get an idea why bigger engines can have more problems due to stress, calculate how much bigger the firebox gets in relation to the shell at different points in the firebox.

I took it one step farther and did an FEA simulation of the movement at the extreme corners of a mid-sized firebox. The FEA is still a work in progress, but suffice it to say the stresses were off the chart simply because you have a steel box attached to a steel box with steel rods and the two boxes aren't doing the same thing. Something has to give and it is primarily the staybolt, followed by the thinner of the two sheets. The box will move far enough to take the stays past yield with every cold-hot cycle. Hence the desire for rubber staybolts. Flexible stays help, but they only lessen the amount of bend. They do not eliminate it entirely, which is why they break as well.

I have also seen scenarios where not only are the stays taken past yield to the point of becoming permanently distorted, but also crown sheets taking a permanent set bent upward at the forward most row of crown stays. The crown sheet to tube sheet knuckle is probably the least restrained portion of the firebox due to its distance from the most rigid attachment points of the backhead and foundation ring and therefore would tend to suffer the greatest amount of distortion and stress concentration. Throw in a very rigid square corner and a weld that may or may not be stress relieved and you have a potential for a very concentrated occurance of stress corrosion cracking (as Linn mentioned, the stress is multiplied as the size of the boiler goes up). We know that this occurs on flanged joints in this area (what locomotive boilermaker has not seen at least one rear tube sheet knuckle pad welded over SCC), but the result is somewhat lessened by the flange allowing for an even distribution of the accumulated stress making the cracking easily detectable while it is still superficial. Cracking at a square corner with that level of stress concentration may become very bad very quickly. It is just my opinion, but I think that SCC is what the N&W was trying to avoid with the welded gussets at the crown to tube sheet joint. The cycle stresses will still go somewhere regardless of how rigid that knuckle is, but they will not be concentrated at the weld by the vertical expansion of the tube sheet and the forward expansion of the crown.

Also, did anyone notice the snake shaped vertical weld at the rear corner of the side sheet? I hope they laid the box on its side to do that. Running it vertical would sure be a pain.

Several years ago, three of us went to an unnamed boiler manufacturer to observe some of their methodologies. The boilers were various sizes and all full pen bolts and square corners. I asked them what the expected lifespan of the boilers was and I was told, “In about 15 years you come back and buy another one”.

From my experience in the modern commercial boiler world, about 15 years is the normal lifespan of a boiler. Most are replaced in 20 years, due to one reason or another. Some manufacturers of Scotch Marine boilers use flat sheets and some use flanged sheets, but I've not paid attention to see which one last longer.

No commercial modern boilers suffer the "box in a box" expansion issues that a loco boiler does, but they have similar problems for the same reasons. It is quite common to make crack repairs on Morrison Tubes where they attach to the tube sheet due to repeated thermal cycling. The more constant the load and temperature, the less problems. The design of the Morrison (corrugated) Tube is meant to deal with the expansion, but it still is not completely mitigated.

Here is a pix of a square cornered door sheet to crown sheet joint. Notice how the corrosion on the waterside of the crown is in line with the tube sheet fillet welds.

Robert Yuill

Were it not for the fact of the crown looking to be way to small.... I would swear this to be a crown/doorsheet bit out of a certain mexican 4-8-4.

Cheers, Jason

My personal experience is that you can get away with square corners but its not the best. You always in the approximately 10 year time scale get problems with cracking in the lower corner of the foundation ring and cracking in both fireboxes along the mud ring starting at the corner. The tube plates suffer vertical cracking through the tube ligaments next to the sides as a result of the stress in the square corner. As for the thickness of the plates being increased to allow for the extra stress of the squareness it increases the load on the stays in the braking areas and stays there only lasted a few wears. The tube plate thickness is critical to preventing heat build up. If you make it to thick the heat is not distributed and creaking will start. The thickness is defined by how you fit the tubes which if you weld them in can be very thin compered to expanding.

We later replaced the square welded box with a round cornered one. It must be said though that the sheets where life expired regardless of the other problems because of lack of boiler water treatment.