I bring no specific documentation to the table.

I will state my semi-educated opinion:

1) "If you turn on the injector and put cold water into the boiler, it will explode". FALSE - The water you're putting in may be slightly cooler than the water already in the boiler, but it is NOT "cold". As already mentioned, the newly introduced water is nowhere near the crown sheet.
2) "Water sloshing on an overheated crown will cause it to crack". FALSE - Boiler steel is mild steel, and is commonly hot forged in the flanging process with no ill results.
3) "The water on the red hot crown instantaneously flashes to steam, causing the boiler to explode". FALSE - While the red hot crown will rapidly evaporate water, I cannot believe that enough water could be INSTANTLY converted to steam to overpressurize the boiler.

Again, my opinion:

Nearly all of the photographic evidence of crown sheet failures due to low water (or any other cause) shows the crown stays separating from the plate. The stays are intact, and for the most part the plate is intact. Yes, the generally intact plate has tears throughout from the explosion, but the plate is not "shattered" as it would be if it failed from a microcracking or brittle standpoint. The plate simply pulls off the ends of the stays.

Standard boiler construction during the steam locomotive era is the treaded staybolt (not welded). While the threaded design is strong, it could possible be prone to failure at overheated temperatures.

The threaded joint relies on a fairly small amount of surface area of thread to support the load. This thread area is subject to shear loading as designed and constructed under normal operating conditions.

If the crown sheet becomes soft due to overheating (red), then the sheet/plate is soft and the very ends of the crown stays are soft. The threaded joint is soft, and is most likely the weak link. The stays are acting as "heat sinks" in an overheated situation where the sheet is increasing in temperature.

If water from the boiler (at boiler temperature, via sloshing or adding) then covers the crown, it attempts to cool whatever it touches. I would think that the stays, being cooler, and having a lot of surface area in contact with the water, would cool more rapidly than the sheet.

As the stays cool, they contract. This contraction would cause a decrease in diameter at the threaded joint. While very small, this contraction may make the stay-to-sheet contact area in the threaded joint become a very loose joint, and cause the threads to then experience a bending moment instead of a pure shear load. Due to a little bit of "wiggle room" in the threaded joint, the joint then fails.

The design of crown stays that are tapered and/or have button heads is an attempt to put a margin of safety in this scenario, and works at least in the example of the Gettysburg incident. I would speculate that in our modern world, the use of full penetration welds on crown bolts might be a better practice than threaded installation, due to the ability of the weld to stay intact at higher temperatures.

In my current work, I've seen overheated boilers that have failed. In one instance, a scotch marine Morrison tube collapsed, but the boiler did not explode. In another instance, a steam drum went dry, and all of the tubes pulled loose. The drum was deformed, but did not explode. However, if the tubes had been staybolts and the drum a crown sheet, it would have been a goner. The tubes are a mechanical connection.

Maybe I'm way out in left field, but it seems plausible to me.

The closest I've had the misfortune to watch happen was an incident of firing up a boiler with low water from warm. The fireman was a retired roundhouse clerk......didn't try the glasses or cocks. The locomotive had been run the previous day,filled to over the top of the glasses, and left. A front corner had cracked open during the night and let most of the water out of the boiler.

Roaring blaze, 90 minutes later, maybe 15-20 PSI? By this time, another retired guy had showed, and hauled the locmotive to the fire plug and introduced really cold water through the overflow pipe, through the injector, into the boiler, until it filled the glass about halfway.

The firebox now looked like an Amazonian rain forest. Pouring water emerged from every crown stay hole, and the ends of the stays were visible in the plate. The sheet and stays had in fact shrunk from each other. Were the leaks there before the cold water hit it? I don't know....but it wouldn't surprise me either way.

I'd much prefer not to be in this situation in the first place, since there's probably no right way other than to kill the fire and get away from it.

dave

I can't think of any locomotive I've ever seen in person or in photos where cold water from the tender is injected anywhere close to the crown sheet. Being mostly familiar with the FEC Pacifics through involvement (once upon a time) with Savannah & Atlanta 750 (ex-FEC 80), and more recently with Georgia Northern 107 (ex-FEC 88), water from the injectors was routed along the sides of the boiler and up to a top-mounted check manifold. Others I've seen have the injection points on either side of the boiler, but always toward the front of the boiler and not in proximity to the firebox/crown sheet.

I don't make a living working on boilers, but in theory, the newly injected, colder and more dense water would sink and convection currents would move this water back toward the firebox where it would heat and rise. The colder water would draw heat from surrounding water as it mixed, lowering the overall water temperature until that water reached the firebox area and the entire volume absorbed more heat. By the time the "new" water reaches proximity of the firebox it is no longer cold, nor would it be appreciably discernible from the "original" water.

Wasn't that kind of the idea behind thermic siphons, to improve circulation and provide more heating area in the firebox? I see no way you would get a crown sheet failure simply because of turning on an injector, without other factors coming into the equation. If the crown sheet is exposed, it's just a matter of time...and I wouldn't want to be anywhere near! If anything, I might consider a flue failure before a crown sheet failure in this instance.

Mark,

Your points are well taken, but there is another major factor at work here:

Boiler pressure is pushing on the crown (and all the other sheets as well). With the uncovered crown at red heat, the crown sheet becomes plastic (in the "old" sense of the word) and starts to bag and deform downward, off the ends of the crown stays. If it happens quickly enough, the sheet can rip at a bolt hole and open up instantly. That's when all the water (2500-4000 gallons or so) turns into many more (12X?) gallons of steam. Slowly enough (stay by stay) and you have what happened to 1278.

Those old ICC reports usually show that is how the event (BLEVE-- boiling liquid, expanding vapor event) takes place. I'm inclined to doubt the "putting water in and she blowed up!" stuff.

Once again, the water glass is "the most importatnt window on a train".

Howard P.
Steam Novice, CT

As for the "where" of injected water entering the boiler, top checks de-aerated the water a bit and also raised the temp a bit as well. The PRR was known for backhead-mounted injectors and check valves on many classes of locomotive. Even more modern PRR power had backhead checks. Those checks fed into delivery pipes that ran all the way over the crown and tube bundle to just shy of the front tube sheet; the theory was that the delivery water would be warmed to almost full boiler water temp.

Good theory, but when the pipes broke off or developed leaks right at the backhead, you were putting rather cool water right into the firebox end of the loco. Probably not good for side sheets and door sheets....

Howard P.

Kelly, Dennis and Mark have made some good points for me already. As I expected, there are a number of stories relating to the subject, but in no case has any actual scientific evidence been provided to support the theory that feeding water to the boiler in a low water situation will cause a crown sheet failure. My intent is not to sound condescending here, but when making decisions regarding boiler operating practices when dealing with life and death circumstances we should be looking to hang our hats on something more than a story that Uncle Orpheus heard from Cousin Jethro about boiler operator Bernie who blew himself to kingdom-come when he turned on the injector. Especially considering that it is usually very difficult, as Kelly mentioned, to conduct a post-mortem interview with Bernie. I think this is a very common misconception and an emphasis needs to be placed on removing it from common teachings.

In lieu of documented evidence supporting this scenario, I wanted to bring up a few points that have not already been covered. First, the temperature of the water being fed to the boiler is hardly cold when using an injector, inspirator or feedwater heater. Regardless of the temperature of the water going to the injector, the water coming out is going to enter the boiler at approx. 160 degrees, getting preheated by the action of the injector. 160 degrees is hardly cold. An inspirator will deliver water at or just over 212 degress and a feed water heater can achieve temps approaching 240 degrees. Considering where the water inlets are located on the boiler, by the time the newly introduced water reaches the crown sheet I think it is safe to say it is not cold. Second, an often overlooked fact is the feed rate of an injector. Think about how fast an injector will raise the water level in a boiler? An average injector might feed at a rate of 100gal/min, but when you consider the total volume of even a smaller size locomotive boiler this isn't much. We have already determined that water levels will drop well below the top of the crown sheet before the sheet becomes overheated to the point of failure. If that be the case, an injector would have to be run for several minutes before the water level would even REACH the crown sheet. At the feed rates we're talking about the rising water would hardly create the sudden shock described as it creeps up the slope of the crown from back to front. Even if the water did suddenly surge onto the crown sheet, I would imagine as Mark mentioned, that the damage would be minimal. The sheet may crack due to the uneven thermal stresses (carbon steels used in boilers do not contain enough carbon to harden by quenching) but a cracked sheet is far superior to having your boiler do its best impression of a Saturn V rocket.

The point of my original post was to get everyone thinking about where are our teachings and practices come from and whether or not they are rooted in sound logic and engineering. In this case, I think we have identified an old wive's tale that still plagues this industry.

Amen and bravo for Brendan's last paragraph.

I know of one locomotive that suffered for years at the hands of "that's how we did it back in the day on the _______". Practices such as no water treatment and 45-minute fireups from ambient to full working pressure, did bad things to this engine.

Maybe we are at a far-enough remove from the "old days", and we have enough young(er) guys who bring logic, analylitic skills and sound engineering to steam locomotive care and operation. This can only be a good thing, in my opinion.

Howard P.

I cast my vote as this is one of the more informative threads on RYPN.

Upon adding additional water to the boiler the level would rise, but the water that would then start covering over the crown sheet would be water well above the boiling point as it is the water that already in the hottest part of the boiler.

I have never seen a piece of red hot steel obtain additional ductility after being quenched. The time when the crown sheet would be most likely to pull through the stays is before being quenched, not after and not during.

Similar to nukes... steam is not nearly as effective as a coolant, but it has some effect. The question is whether there is enough convection of steam to remove heat from the sheet (or fuel rod) enough to prevent failure. Overheat is not instantaneous, and whether you can get away with a moment's uncovery really depends on the physics of the situation.

I would say that if there is uncovery, cover it!!! Also, it goes without saying that you SCRAM or drop fire.

I don't see where there'd be any room for the wives' tale in emergency response. If you really want to explain the wives' tale to a bunch of raised-eyebrow NTSB investigators, then, well, I don't think you do.

The topic situation is never going to actually arise in the real world for any of us, not even the few who do run steam. It's dreamin about steamin... all noise and no signal AFIAC. None of it helps any of us do anything about railway preservation.

it's just the classic RyPN situation. A discussion actually relevant to boots-on-ground railway preservation gets 12 views, and a foaming hypothetical arm-waving nonsense question about "steam engines" gets 12 pages.

Once I started a real thread. And then, as a test, I started a "troll" thread about steam engines. The result was exactly as I describe. It's like going to the Yahoo homepage to do a search, and there's all those articles you just can't resist clicking, even though really you shouldn't care about Lindsay Lohan or some cat.

As a mechanical engineering student, we are always to support our claims with solid evidence. This evidence being in the numbers. That's what I'm not seeing here, even though some claim they are using an engineering approach. Though some good claims have been made whether or not the injectors are or are not a direct or indirect cause, I just cannot be persuaded without seeing the numbers.

Though I can't fully support this claim either, it may be good food for thought. But if you're at the point where the injectors would affect the crown sheet's material properties, you're probably at the point of no return already.

The main cause of a boiler explosion is a failure in which an almost instantaneous significant pressure drop occurs. I opened up my thermodynamics textbook and threw around a few numbers real quick from the steam tables. At 200 psi (absolute pressure, which in Imperial units is not too off gage), you have a saturation temperature of 381.80°F and the specific volume is between 0.018 and 2.28 cubic feet/lbm (depending on quality of steam, which in locomotive boilers will not on average be at the extremes). If pressure were to nearly instantaneously drop to atmospheric pressure, which is about 14 psi, the temperature will be nearly the same and the water/steam mixture will be in superheat. This drastically increases the volume as the specific volume of superheated water at atmospheric pressure is about 32 cubic ft/lbm. Again, depending on the quality of the steam, this is anywhere between 10 and 1800 times the previous volume. Now that's the science behind the explosion, I'll leave it up to you all to debate whether or not, the injector can cause the boiler to shock to this point of failure and allow such an explosion.

Kyle

Edit: I'm no particular expert and I applied knowledge to the best of my current education level. Some numbers were approximated quickly using less than ideal techniques and there may be some small errors, but not enough to affect my point.

Been away at a meeting, and at the moment I can't offer too much to the evidence so far, except two observations:

1) I can't say for back then, but contemporary NTSB policy is never to determine outright blame, but instead "probable cause" and "recommendations". Even when it's brazenly obvious what caused the accident, they're not going to say so outright.

2) A simple thought that occurs to mind when evaluating this stuff, based on my time in both the kitchen and the foundry: Let's remember that the actual temperature of the steam and boiling water is dependent upon the boiler pressure. I'd have to go dig out the textbooks, but as my foggy memory recalls, when you have 200 PSI in the boiler the temperature of the boiling water is somewhere around 382 degrees F., and at 250 PSI about 400 degrees F. Now, throw an exposed crownsheet into the mix, not necessarily glowing red but still much hotter than 500 degrees. How effectively, and for how long, will an exposed section of crownsheet "superheat" the steam, and will that increase in temperature have a negative effect on the boiler?

EDIT: I see from the post above mine that my memory wasn't that far off!

I guess you missed the post that almost certainly created this whole discussion....

Back in about 1982, I was conductor on a run with Hillcrest Climax #10. The blowdown valve on the engine broke off due to faulty piping. The engine crew bailed out of the cab since it was full of steam. They were all injured to some degree, and were tending to those injuries.

I had an engine in front of me that was still venting a massive amount of steam. It was also smoking, and it was unclear if the fire was totally extinguished or not. I was concerned about a pan fire, or possible firebox explosion. I wanted to get on the engine to activate the emergency fuel cut off.

Part of my brain said "It's safe to do so, the engine is not in danger of exploding..." Another part of my brain said "You do realize that you're literally betting your life on that information being correct, right?" (Though it said so in a much more primal manner.)

As somebody has already pointed out. Let's say you're on a coal burner and you are in a low water situation, for whatever reason. You have two options: 1) Run like hell screaming "She's gonna blow!". 2) Dump the fire. While dumping the fire, would turning on the injector be helpful or harmful to your situation? Explain why. Explain whether or not you're willing to bet your life and the lives of those around you on your answer being correct.

Tell me again how this sort of discussion isn't relevant? I honestly hope it never will be, and that nobody will ever be in a low water situation to start with. But knowing what to do if the unthinkable does happen could prove very useful.

Also bear in mind that nobody is forcing you to read this.

That is a very good point, and I would certainly agree. In fact, this morning I posted that I thought this whole discussion was simply a theoretical discussion at this point in steam railroading. Water out of sight? Kill the fire and get everyone out of the area.

Then somebody brought up the point that on a coal or wood burning locomotive, it's not that simple. You don't just slap a valve shut and exit stage left. Leaving the fire burning would not be the safest course of action either.

So you're dumping the fire, and while you do so, what, if anything, should you do with the injector? Should you turn on both injectors? Would that help matters, or is it dangerous? The general consensus here seems to be it could actually help, but you have to wonder why the Old Wives tale is so pervasive. Why did they warn everyone that turning on the injector just about guaranteed instant death? Why did so many railroaders believe this, and pass it down?