Having spent some of my free time over the course of the past two weeks in the search for documents related to boiler stresses PRIOR TO THE FIRST MATCH BEING LIT IN THE FIREBOX; I have come to find something interesting about the stay bolt topics currently at hand. Additionally, there has been an interesting discussion with-in another committee that I attend about stress and values that has led me to consider my above bolded topic even more.

PRE-FIRE stress values in boilers built by LIMA, Baldwin and ALCO in the last days of steam were treated FAR (when I say far....I mean FAR differently) than they are today. In fact, I have come to find in discussions between a few of my best professional allies that our view on STRESS is far different than it was in say, 1940. A few examples include, but are not limited to;

In the early years of welding, welds were stress relieved by peening or hammering the weld. This was later found to be mostly noneffective and also created a crack sensitive area in or near the weld.....however, as an early practice, it was considered the best they had.

Riveted areas (in some cases not all) were stress relieved by heating the area both before and after rivets were installed. There have been discussions on why and how this is done and there are today discussions toward the effectiveness of this habit.

Fireboxes were built on the floor and THEN after they were complete, they were installed. Flanged areas were often "normalized" as a heat treat-stress relief practice.....however, I struggle to find a WIDELY accepted standard and practice for this art. I BELIEVE from my readings that this type of post bending or post flanging treatment was not ALWAYS the case....but was practiced when practical. (What I mean is; there were roundhouse shops that did not have the ability to treat a fully flanged sheet.....so.....possibly they simply let it go as it was flanged.)

Fast Forward now to our current practice:

The Germans and Swiss and others who have accepted the EU standards do not apply post weld heat treatment or REQUIRE stress relief of parts the way we do in the States.

It appears that we (the USA) are one of the few that require full heat treatment of finished all welded boilers.

What am I really getting at, you may ask?

I really wonder if we are eating ourselves up on topics of STAY BOLT STRESS when the real topics that we are overlooking are those related to OTHER aspects of the boiler that clearly contain stresses, however, for the lack of knowledge or full understanding....we consume our time with ONE aspect of the boiler construction to which there are ample studies and examples.....without diving fully into those topics which were not as well studied or practiced, mostly for lack of technology to accomplish the end goal.

My goal is not that of in-sighting a detailed debate about practices (which I am fully prepared to see this thread turn into.). I am however interested in knowing how many of us have actually sat back to consider WHY it is that so many stress impacted ares of construction in 1940 were NOT addressed as we address them today. Additionally, why is it that others view the prevention of such stresses so DIFFERENTLY than we do in the USA, today? It will be interesting to see how long this discussion can be maintained on the topic of WHY, not HOW.

I struggle personally to find the WHY's of my question.......not the HOW's!

Kindly yours,

Good and intersting information, John, thanks. All welded boilers (or ships, or tanks, etc) were very new in the 1940's and we did have a couple other national engagements that took priority - winning that world war and all........I think the best they could do at the time was very good indeed. Our ability to evaluate and test fabrications, and even to know more about how welded steel behaves, has improved dramatically in the past 70 years or so, as have our test instruments. It only makes sense that our practices would change as we learned more about the technology in general, and developed the ability to better test our results. I'd be surprised if we had left the 1940 standard alone given what more we have learned........probably the nuclear power industry forced critical evaluation and testing of welded steel pressure asemblies to the fore, and good thing they did.

As to why we are different......culture? Much has been said about how our design was different based on our loading gage, desire to move mountains of freight rapidly with one power unit, and heavy rigidity to counteract stresses rather than frequent light small trains operating with more delicate albeit efficient boilers that absorbed, distributed and accepted stresses. Unless we compare how they tested and qualified their standards and practices with our approach, we won't really know the story. If they needed other ways in their experience no doubt thay would have found them.

Tross documentation is available and comprehensive - so we start where we have information. Practically, how many of us have problems with cold boilers failing through fabrication stress rather than stay problems caused under load? We work on our most pressing problems first, especially if there's easily available data to start with.

My take on it is based on that practicality insofar as answering your questions. I wish I had more knowledge of foreign practices, but more educated people on this list can no doubt chime in.

dave

John, is this still true? I recall that the first U.S. all-welded locomotive boiler, built by ALCO for an older D&H 2-8-0, was heat treated after completion of fabrication and that ALCO was the only company in the U.S. with a furnace large enough to allow this for several years afterward which made them the sole source for all-welded boilers for several years.

I was under the impression that this practice was dropped by the early 1950's. Can anyone confirm this?

John,

I think you are exactly right in that we only discuss what we know, and try to bunch all of the symptoms we see into that one possible cause that we understand. Why do staybolts break - it must be tension stress, right? That is the stress everyone calculates for because everyone understands if you pull enough on something it will break, and pressure pulls on those bolts for sure.

"No one knows what he doesn't know until he knows it" - LDP

I was just speaking with some other boilermakers in addition to yourself and we were discussing how old sheets spring out of place when they are cut out and how the rear boiler changes shape (despite bracing) over a few months to a year afterwards. Is this stress from initial construction or years of operation?

How can we learn more despite limited resources in our industry for innovation? There are so many neat tools out there that we occasionally hear about from some of our more mainstream ASME friends that post here. We just need to figure out how to get some of our hardware in the lab.

Until we can figure out how to preserve innovation, the preservation of the technology of the steam locomotive is incomplete.

Matt,

You used the word symptom, which, to me seems to be the heart of my question. I struggle to find in my studies hard proof that the stress that actually causes a stay bolt failure (for example) is 100% related to internal pressure (or the pulling that you speak of) rather than being caused by movement of sheets at fire-up or cool down that are under natural stress from the construction process.....accentuated by the expansion and contraction of the plates as they heat and cool.

If this is true of stay bolts, we must then consider failures in the form of a crack in a knuckle, or corner or even side sheet that COULD be caused by already existing stress, accentuated by expansion and contraction of heating and cooling. Rear tube sheets knuckles are a great example. A rear tube sheet knuckle fails over time as this area endures great expansion and contraction (primarily (we think) at heat up and cool down).

It has been my own observation that not ALL rear tubes sheets end up pitted or cracked. It is my observation that some locomotives are hard on stay bolts, while others go years with no failures. My own observations show boilers with multiple corner repairs, regions of broken bolts, but, nice looking rear tube sheets or even front tube sheets. Though not all boilers are created equal......it is clear in my mind anyway that stress values and natural stress regions on any given boiler are also not equal one to another.

Without saying it directly Matt; Does the stay bolt issue and the need for new designs of bolts address a problem with the stay bolts themselves.....or does it really address an unknown or unquantified problem of un-relieved stresses that could not, or were not accounted for? This is now becoming the soul of my question. However, I think the discussions at hand are still addressing stress on the BOLT DIRECTLY, exclusive of the discussion of other natural stresses.

Hugh,

In discussions with others over the course of the past two weeks, ALCO's all out heat treatment of a few all welded boilers was discussed. Indeed, this was done and ALCO did do it and we can document that it was done. Today, all of the boilers that I build are heat treated, 100% as part of the code process. Yet, the same boiler that I would build here in the States, built in say, Germany would NOT be required to endure a full body heat treatment. So again, WHY is it that we have determined that this is a requirement, when others do not see it the same way? I know WHAT we are achieving in a full heat treatment, I struggle to fully document the science behind WHY we do it to the EXTENT that we do it.

To paraphrase Matt J. We simply do not know why we do what we do as we have not innovated enough technology to fully understand it so we cover our symptoms and problems in discussions and explanations that we think we do understand. Right, Matt? (This is what made our Durango Project so much fun is that we actually got to see things that we had "talked about" but had never physically seen or proven.)

Yet, I hold out hope that somebody out there can shed some never before presented science behind the WHY's of how we got to where we are today and why others are where they are today.

John- I was not aware that full boiler heat treatment was still done, nor even that the facilities were available to do it.

Those are very interesting questions regarding stresses and their effects. I guess this goes to show that firetube boiler design remained (and remains) a largely empirical process. Most of what's been learned has been via trial-and-error, and many unanswered questions remain.

This would be one way to improve our understanding of the processes:



I ran across this image on a model engineering forum. The guy is building a 5" gauge live steam locomotive with a copper boiler and was able to do at least limited finite element modeling (FEM) of it. The complete thread is here:

http://www.homemodelenginemachinist.com/index.php?topic=9804.0

I imagine it would be quite an effort to completely an accurately model a full-sized locomotive boiler, taking into account the stresses from heat, thermal expansion, pressure, and external forces on the boiler, but it seems like it'd be worthwhile. I've never done any myself, but I did work closely with a consulting mechanical engineer on a project a few years ago. He said that in his experience he found that eliminating excessively stressed areas in a design using FEM is as much art as it is science. The natural inclination is to beef up overly stressed areas, but this often simply concentrates excessive stress in other areas. Sometimes the solution is to reduce the size of the structure adjacent to the highest-stressed areas which spreads the stresses out. I can imagine we might find similar issues in examining firetube locomotive boilers.

This seems like it'd be a good college senior (or graduate level) engineering project. Anyone know someone who might be interested?

I'll give an observation that may or may not be related....

Struthers -Wells corporation in Warren, PA was a manufacturer of high-strength pressure and heat exchanger vessels up until about the md 1980's. While they weren't stayed construction, the applications included nuclear.

One of the odd features of the plant was a HUGE heat treating/annealing furnace - you could roll a railroad flatcar into it, and it had big roll-up doors on either end with massive insulation on them. They were quite proud of it, had it painted on the side of the furnace.

From the time I was aware of what it was (lets say late 1960's) up until the site was razed for demoltion, I never saw those doors move. Trees grew through the chains and cables. The plants high-water mark was about 1977, and even then, that annealing furnace never budged.

For a minute I thought my memory had failed me until I found this abstract online:
http://pubs.acs.org/doi/abs/10.1021/cen-v010n003.p036

What does that tell you? I'm inferring a lot here, but the value of annealing the life out of an entire pressure vessel or boiler seems to have been all the rage in the 30's and fallen completely out of practice by the 60's, at least in my observation. I'm not sure 'why' either, but I think this lends credence to the discussion here.

The design group still lives even if the plant doesn't.

.

John,

I think that's good that you find that staybolts don't just break because of tension stress in your studies because that's my point. Most staybolts show cracks which suggest they have tried to be bent into a "S" shape due to movement of the sheets - primarily shrinkage in the hot zones - however, you and I have seen some pretty poorly fit sheets, too!

Installation stress was another topic Dr. Tross addressed in the form of applying the firebox and wrapper welds of a staybolt at the same time and also having clearance on the staybolt to sheet fit - among more elaborate suggestions of counter-stressing the sheets to operational stresses.

Mr. Austin addressed installation stress and operating stress rather extensively in his specification for the installation of the firebox on UP 844 - perhaps the greatest extent actually addressing the problem and not the symptom has been considered.

I think the least installation stress imparted and highest level of stress mitigation in construction will give the best operational results, especially in designs that more closely approach the tensile strength of the material at operating temperature. As we continue to lower the factor of design in many things, including non-locomotive boilers, these stresses certainly become more critical. What factor of design is sufficient to ignore it? I don't know.

The Tross staybolt does not know installation stress from operational stress - it only knows how to better distribute bending stress across its entire body rather than have it concentrated near the sheets. It is a very simple, but misunderstood technology much like the locomotive it is part of.

This is so perfect...may I use this quote Mr Janssen?? Seriously.

Stephen

Hugh and Randy,

Wasatch Railroad Contractors has an annual service project where-in we provide the instruction and judging for the Wyoming State SKILLSUSA Welding Competition. I myself an a two time Utah State Welding Champion and I love welding with the kids and running the Wyoming State Event. The event is held annually in Casper, Wyoming, the heart of the open pit coal fields in Wyoming and the largest single producer of rail tonnage in the United States (and disputably the highest volume rail tonnage producer in the world).

Many very large equipment manufacturing companies are located in Casper, Wyoming. The worlds largest single operator dump truck is produced in Casper and shipped world wide. Last year WATCO produced well over 100 units that hold the world record for largest production unit in the world. At any rate, nearly EVERY piece of the machine is subjected to full heat treatment and stress relief as part of the fabrication process. I have seen a single truck bed frame that was well over 10 feet high, 16 feet wide and almost 40 feet long in one oven at one time. The process of heating, maintaining and cooling the unit is an 18 hour process!

Most of our large boilers go to Casper as we have well over 6 options for BIG heat treatment. Our boilers have been in ovens with 6-8 other items at a time. One oven in particular is so big, it has doors on both ends and they use very large front end loaders to get items in and out. Another is so large, it has a huge table on tracks that roll items in and out. BIG HEAT TREATMENT is a BIG Business in the world where I live.

Mr. Andrew Cook of Sheffield England, owner of William Cook Cast Products has looked at the large castings that make up some of the locomotive beds from ALCO before the end of steam. Cook, a world renowned castings specialist is amazed that they made locomotive beds as large as they did back then. Though he could do the same thing today, he claims the challenge to be sizable, but doable. His biggest concern is not the casting itself.....but controlling the cool down in a manner slow enough to keep the casting from cracking. His answer is a large heat treatment oven. The casting would be poured in a pre-heated box and then quickly installed in the oven and heated as fast as possible to retain heat and then allowing the casting to be brought down over an extended period of time, controlling contraction during the solidification process. Can you even begin to imagine the size of box that would be needed for a FEF-3 Union Pacific 4-8-4 bed casting? That casting alone is 66 feet long, 8 feet wide and at some points nearly 6 feet high! Double that for the box in which the bed would be cast. That box would be at best 70 feet long, 12 feet wide and at least 10 feet high! Now, an oven to fit such a box and keep the box well over 1200 degrees.....WOW. (side note: Kratvilles Book, Mighty 800's notes cracking issues with General Steel Castings 4-8-4 FEF UP locomotive beds).

My point is this;

WATCO heat treats the machines it makes as they are generally alloy steels of high friction nature (AR rated plates), the welding of which that is not treated has grave repercussions. They know WHY and how to do this and they have specific reasons for it.

Cook when casting large items has a very specific reason for heat treatment units to prevent casting flaws and or cracks. He knows why he does it and has a specific prescription for such work.

Kelly has noted the Bla, Bla, Bla of the code and HOW we do it. We are still left with the specific of what it is exactly that we are doing in our process, what it is we are curing or solving and what the end impact (good or bad) really is.

Hugh, I love the picture and the explanation. Mr. Janssen and our team were well on our way to doing the same thing in a firebox.....but were not given the chance to go that far on our project. We are convinced (I am anyway) that if given the real chance to see what a firebox is doing in terms of movement during fire-up, during idle times, during full fire and during cool down.....we would almost be shocked at the amount and severity of the movement ALL OF THE TIME.

I am still hoping for a really good engineers analysis of the unknown and un-calculated stresses involved in the fabrication process of a heavily stayed firebox in a generally rigid shell body. I still think we have addressed the symptoms, not the true cause of the problem.

Kindly,

Anyone who doubts that boilers move during the cold to hot cycle need only sit on one and listen.

Frisco, Well said!

John E., I have done experiments regarding weld peening and would say that it most definitely does work! Give me a day or so and I shall sketch something up. I have a feeling the cracking issues you reference would have been the result of "over peening", repeated impacts and excessive "working" of the metal having effectively (to an extent) work hardened the material.
Regarding rivets, it has long been my understanding that the practice of preheating the hole was not for the purpose of stress relief but rather for the purpose of slowing the rate at which heat is drawn out of the rivet as it is being driven. Is it really necessary, does it really help? Don't know....but that is my understanding of the practice.
When it comes to why we stress relieve, well that is a good question in some areas. Why are flush patches relieved even when below the ASME standard, that is the ASME states that such welded joints need not be post weld stress relieved if they are below either 7/8" or 1" ( don't remember which ).
What are your thoughts on this type of a device/process? It/they are called an Ultrasonic Impact Treatment, I know the navy uses these systems to prolong the life of certain crack repairs. I have in the past read that such systems eliminate about 80% of stress in wleds....as you weld, sounds almost to good to be true.

All,
There is no doubt in my mind that indeed the breakage issues are the result of flexural stresses versus those of tension. The stresses are imparted not just at initial heatings or coolings but during changes in operational demands and the sudden manner in which they occur. Think about a locomotive that has been hot for days, perhaps loafing around the hard, going from a nominal to idle demand to that of pulling an 800 ton train up a 1.5 percent grade. A Flannery drawing Mr. Austin had posted within the last year or two regarding actual temps of fire side sheets while the boiler is under heavy load comes to mind.
Case in point, it has become practice of one mainline locomotive to not be banked overnight. Rather, the keep a lazy fire on the grates all night. The yearly breakage (whilst the locomotive saw very regular service) was reduced from nearly 20 bolts a year to 1 or 2.
Additionally, how many of the issues today are the result of incompetent engine crews? I have ridden the Western Maryland three times, each time we had to stop 1/2 to 2/3s the way up the hill because the boiler pressure had dropped to 130-140 and they were utterly low on water. What kind of stress does that sort of operation cause? While I may not be able to quantify that with any sort of engineering lingo I can assure all that the firebox sheets and bolts were SCREAMING as the issue had not been getting progressively worse from the time of departure.... It occurred within just a few miles as the crew lost control of the fire. This is far from the only example, but I can't give them because this is not a board that allows realities of this industry to be brought to life if they are unpopular.
No matter what minute changes could be made, IMHO, it comes down to better care, maintenance and ( MUCH, MUCH, MUCH ) better traing. For as long as flat plates of different lengths (and heights) are exposed to grossly unequaled temperatures then we will always have "Issues" with stresses in our furnaces. It seems that the best line of defense is that of slow heatings (possibly using a torpedo heater overnight prior to fireup), slower coolings and fireman whom can maintain pressure.

There would have been more but I am beat.

Cheers all, Jason

This continues to be an interesting discussion.

I posted a story to the flimsies yesterday you may find of interest- the Golden Rock workshops in India have just completed their 2nd oil-fired rack-and-adhesion steamer for the Nilgiri Mountain Railway, and it's the first with a welded boiler. The first new locomotive, completed about a year ago had a riveted boiler. It appears both locomotives are basically copies of the SLM X-class locos built in the 1950's.

Another article (here: http://www.thehindu.com/todays-paper/tp ... 812985.ece) mentions that fabrication of the boiler was farmed out to an Indian boiler manufacturer. It'll be interesting to find out how this project turns out long-term.

John, I make no claims to the contrary, I'm just puzzled by the fact that apparently the oversized furnace I drove by twice a day for probably 15 years never seemed to be used at all, in a company that made heat exchangers for the nuclear industry in a busy plant. Even by your standards it was BIG, as in 18'x18'x87' and capable of 2100 degrees. That could anneal a Superliner or any locomotive-size component. You'd think something that unique would have a line out the door even by by the 60's. I suspect however, that the answer was that they still managed to heat-treat the welded vessels they made without resorting to the ultimate 'big gun' to do it.

Being no stranger to heat treating (worked at National Forge, the crankshaft people) there was always something amazing about watching the heat-treating and surface hardening (subsequent oil bath quench) of big forged castings - particularly stuff like 20-throw SD45 cranks with welded-on counterweights. They didn't do boilers though. We had, and have, some first-class metallurgists around here.

The new replacement boilers for the TSRR - particularly the Dixon one on the #300 - is an ASME-spec'd boiler that passed the Form 4 without issue, but is all-welded from the start and amazingly non-complex- few interior braces or add-ons to the essential design. I'm assuming that it HAD to have been fully heat-treated under ASME, certainly looked that way in retrospect as there weren't any signs of unanticipated stresses since construction.