Some of the photos are fairly modern heavy locomotives. In one photo, the "hairpin" superheater elements are laid out and warped like over-cooked spaghetti, while the first course of the boiler and the flues which encased the superheater elements is blown away.
A number of the photos show explosions due to failure of the barrel courses, rather than due to firebox/crownsheet failures. In one photo, the failure ripped the barrel open along a longitudinal seam. This would point to possible weak or inconsistent material (old as the boiler in that photo appears, could have been wrought iron, and maybe a lapped seam).
In another photo of an older locomotive, the barrel course is blown away, and it looks like the dry pipe is the only thing holding the remaining portions of the boiler together.
From my own work with locomotive boiler engineering, the more common failures were crown sheet failures due to low water. Bad water chemistry adds its own failure mechanisms such as grooving and oxygen pitting. Interestingly, when a locomotive boiler is first having steam raised from cold, a phenomenon occurs within it: steam bubbles forming in the water surrounding the firebox tend to circulate and when they contact the cooler "wrapper" (outer shell surrounding the firebox/waterspace), these bubbles condense. In condensing, carbonic acid is produced, and this attacks the sheets of the boiler wrapper (and sometimes the barrel sheets). The result is grooving due to acid attack. This type of attack increases in areas of higher stress, such as along seams. Riveted boilers had a "calking line" along the seams, where the seams were made steam-tight by running a calking chisel along the lapped edges of the sheets. This calking produces a shallow groove in one sheet and upsets the metal against the overlapping sheet's edge. Higher stress area, and a groove to start with, this is often a place where accelerated corrosion or acid attacks occur.
Another interesting phenomenon I've come across is referred to as "halo'ing" by the boilermakers. This happens on the crownsheets. A ring of heavy corrosion/lost of metal surrounds each staybolt on the water-side of many crownsheets. If the sheet is ultrasonically thickness gauged, the UT's will generally show sufficient metal to handle the maximum allowable working pressure. Get inside the boiler when the tubes are removed and take a look across the crown sheet, and there is often this "halo'ing" around each crown stay. To my thinking, crownsheets are the most critical sheet in a locomotive boiler. If a crownsheet fails, it usually collapses into the firebox (either the sheet got red hot from low water/no water covering it, or the halo'ing around the stays finally caused the sheet to "unzip" from the staybolts). When a crownsheet fails, it is usually death to the engine crew. On heavier locomotives, a low water/crownsheet failure would literally blow the boiler right off the engine's frame, tearing it free at the smokebox seam. Locomotive boilers were anchored to the locomotive frame by the "saddle" on the cylinder block. The smokebox course of the boiler was bolted to the saddle with reamed and fitted bolts. When a crownsheet failure occurred, the blast of steam acted downwards into the ashpan, with some blasting out the firehole into the cab. The boiler tended to rocket up off the frame, and tore free at the smokebox seam.
In the 1940's a mallet type locomotive on the Delaware & Hudson RR had a boiler explosion near Central Bridge, NY. Low water was the culprit and the boiler blew right off the frame of that locomotive. The engine crew was killed. I did the boiler calculations for the "Royal Canadian Hudson" a good 20 + years ago. A package of records and drawings for the boiler on that locomotive was sent to me. In the records was the fact that locomotive got a new firebox sometime in the 1930's. The note is somewhat laconic and read: "low water explosion, Caron, Saskatchewan... 5 men killed."
I've been working on the engineering to design repairs to a boiler on a smaller Alco/Cooke works locomotive built in 1920. The boiler shop cut some portions of the inside corners on the firebox tubesheet away and we got a look at the insides of the water leg. More work required. The crownsheet has the classic haloing. The rest of the crownsheet has plenty of steel left. I recommended going to larger diameter crown stays and using full penetration welds from the firebox side to put them in. This will remove the halo'd areas. The inner door sheet had too many thin spots to be repairable, so a new inner door sheet has to be "flanged up". There are a number of other areas to be repaired, and it is nothing I have not seen before on other older boilers. Ultrasonic thickness gauging is a wonderful thing, but it can only do so much. Visual inspections and a knowledge of how locomotive boilers were built as well as various types of corrosive or erosive attack are all needed to properly evaluate a boiler's fitness for service. Having fiberoptic 'scopes to take a look inside mudlegs and other inaccessable areas is also a great advantage the older generations of boilermakers, boiler inspectors and engineers did not have.
Most of the bigger railroads recognized the importance of maintaining proper water chemistry in their locomotive boilers. Water chemistry varies widely with the sources and even with the seasons of the year. The bigger railroads had water chemistry labs, and took water samples for analysis, then formulated the chemical treatment. Some of the later heavy steam locomotives actually had instrumentation to detect foaming (due to water chemistry) and pans within the boiler to try to capture the foam for blow-down.
Two boiler explosions in more recent times in the USA come to mind with this thread. One was at Gettysburg, PA, and was a low-water explosion on a locomotive boiler. The root causes were a lack of consistent maintenance, lack of water treatment, and gauge cocks (supplying water to the level glass in the cab) were clogged shut with "mud" from the boiler. The gauge glass had a trapped volume of water in it, so gave the false reading that the boiler had a safe water level. Had anyone properly blown that glass down, they would have known immediately that the glass was not connected to the boiler due to the clogged gauge cocks. It cost two men their lives. One survived for some years but had many reconstructive surgeries, talked with an electronic voice box, and eventually died of the long term effects of his injuries. The other boiler explosion was a locomotive type boiler on a traction engine, at Medina, Ohio. It was a case of people having a false sense of security since the boiler had passed a hydrostatic pressure test. When the traction engine was started in motion to be moved on the fairgrounds, the boiler blew up. Low water on the crownsheet due to a slight gradient that the traction engine was travelling onto, and a number of other causes including inaccurate pressure gauge and (possibly) a faulty safety valve. That explosion killed a few people.
I enjoy doing the engineering on the steam locomotive boilers, and each time I get into another boiler job, I realize how little I know and how much more there is to learn. When I evaluate a locomotive boiler, I run a "full set of numbers" on it, calculating the minimum required thicknesses of the different parts for a given maximum allowable working pressure with a minimum factor of safety of 4 on the older boilers. On new work, such as firebox replacements, I use a minimum factor of safety of 5. A set of calculations on a basic locomotive boiler (no superheaters, no combustion chamber) will typically run about 120 pages of hand-written calculations. These accompany the "Form 4" (boiler registration document) and the "Form 19's" (boiler repair or alteration reports) and are sent to the US Federal RR Administration for review/approval. The calculations are based on a current set of ultrasonic thickness gauge readings as well as as-found measurements and inspections. Measuring brace rod diameters, getting inside the boiler with a slag pick and banging away at any suspect areas are all part of the process.
I often use a combination of riveting and flanging as well as welding to repair or alter the older boilers. It's a kind of engineering that I enjoy doing, and I do it in the old ways, no computer software, no CAD, just get the drawings and "as found measurements", get the ASME boiler codes and maybe a stress analysis textbook, and a pad of graph paper and a calculator and go to it. I mentally put myself into the boiler, starting with the front tubesheet and work my way back through the boiler, evaluating each sheet, brace, stays, seams, and all else. Detail sketches accompany my calculations, and I "show all units", whether pounds per square inch, inches, feet, types of stresses (tensile, shear, crushing), and all my figures are neat and legible. Each locomotive boiler I've done engineering on has its drawings and calculations on file in my house. Rolls of drawings and ring-binders of calculations and folded up drawings and submittals in file folders all have accumulated over the years. For variety, I've also run calculations on Scotch Marine Boilers, Horizontal Return Tube (HRT) boilers, and Vertical Firetube Boilers (VFT), along with a number of compressed air receiver tanks. I am an old-time engineer and working on boiler engineering suits me fine. Proper evaluation, proper engineering, and properly done work are part of the puzzle. Operators of boilers have to be on top of their game with water chemistry as well as operating and firing practices. As an engineer, I can only do so much, but once a boiler is in service, the rest depends on the people entrusted with its maintenance and operation.
All too often, on the railroads, low water explosions resulted from either injectors failing to "pick up", or a fireman not realizing that an injector, instead of "picking up" was putting the water on the ground via the injector overflow. Sometimes, a crew would be low on water and would gamble on making it to the next location where there was a water tower or other means of getting water in the tender. Towermen or men working along the line would hear the low water whistle (an alarm in the later/heavier engine cabs) shrieking as the engine and train went past. Sometimes the crew made it and took on water, and many times they didn't and there was a classic low water explosion. Being out on the mainline with a locomotive and train and realizing there was insufficient water in the tender tank to keep the boiler water at a safe level if the locomotive was working had to have been the worst imaginable feeling. The alternative was to foul the mainline with the train (setting out fusees and torpedoes) and cut off the engine and run it light for water was something no crew wanted to do. No radio communication in those days, so a crew was on their own once out on the road. If a tender had next to no water in its tank and there was a heavy train behind the engine, the only safe alternative would have been to stop right there, cut off the engine and move it light (assuming a water source was in reasonably close distance), take on water and get back to move the train. This meant placing fusees (flares) and torpedoes (explosive warning devices fastened to the rails) and having brakemen or flagmen sent out to flag trains coming along that same track. Most crews took the gamble and made a run with the train for the next place to take on water. Most made it and survived, but plenty didn't, hence some of the photos posted here.