What's new
What's new

Thanks to Maynah

Joe Michaels

Diamond
Joined
Apr 3, 2004
Location
Shandaken, NY, USA
Thanks to Maynah for providing us with a Number 3 (Male) x Number 4 (female) MT extension. This week, we started the repairs to the outside cranks on a GE 50 ton diesel locomotive. Each truck of the locomotive has two axles. One axle in each truck is driven by an electric traction motor. Power is transmitted to the other axle by side rods (like a steam locomotive). Heavy cast steel cranks were pressed onto the end stubs of each axle. A typical press fit for this kind of application would involve using a wheel press and about 800-1500 tons of force. Over the years, some of the cranks worked loose on the axle stub ends. Square shaft keys and keyways got battered and deformed. Prior owners of this locomotive tried a variety of sketchy repairs, none of which secured the loose cranks to the axles. I was asked to come up with a repair that would work. I first suggested we take the cranks off the axles in place, most of the cranks being so loose you could move them along the axle. Send the cranks to a machine shop for boring the hubs oversize & recutytting the keyways. Then, bring in a contractor on site with an oribtal lathe (one such tool being a "Journal Squirrel). Mark the keyway locations to keep the 'quartering' (orientation) of the cranks, then turn the axle stub ends undersized. Build up with a MIG gun on the orbital tooling, then turn to a shrink fit in the re-bored crank hubs. Mill the keyways in place.

This idea, while a Cadillac fix, was shot down for reasons of time (needing the locomotive to be moved to accomodate train operations) and cost. The ball got tossed to me again.
So much for "retirement". It was time for the "Chevy Nova" repair.

I decided to go with two 1 1/4" diameter steel dowel pins set in the end face of the crank and axle, at 120 degrees, half in the hub and half in the axle. There was enough real estate on the crank counterweight to set a magnetic base Milwaukee drill. The drill takes a Number 3 MT. Getting into the realm of a 1 1/4" machine reamer and reamer drill, the shanks are number 4 MT. Maynah sent me a 'care package' of the 3 x 4 MT adapter (or extension), two of them, and some other goodies which I am using in my shop on other work. We drilled the first dowel hole and worked up in drill sizes. I scored a 3-flute core bit on eBay in some metric size that is 0.009" under the 1.250" reamer diameter. Using the extension Maynah sent us, we sailed right along with the drilling and reaming. 1 1/4" nominal diameter x 4" depth.

The reamer cleaned the hole to 1.2505" and 1.252" (0.0015" out of round given the fact there was a a few thousandths gap at the interface between axle and crank hub). I was happy with that. I machined the dowel pin from 4140 HT. I allowed 0.0055" for the shrink fit. The end of the pin was tapped 1/4 -20 for an eye bolt. We hung the pin on baling wire in a flask of liquid nitrogen for about 30 minutes. I heated the hub and axle with a propane weed burner until I got a light straw color, then we added an oxacetylene rosebud and got a blue temper color in the reamed hole (600-700 degrees F). The dowel pin went in with a bit of light hammering. As we started the pin into the reamed hole, I hollered: "It's honeymoon night- drive 'er home , quick !". The pin drove to full depth. I had put a healthy chamfer on each end of the pin. I did some dentistry with a die grinder and put a weld groove around the pin on the face of the crank and axle end. We got a 400 degree preheat back on, and ran a couple of passes with super-duper high strength repair electrodes. I normally use MG 600 for this sort of thing, but am out of MG 600 stick electrode. It is PRICEY but does work on mystery alloy steels and tool steels. I had a handful of Cronatron 330, roughly equivalent.

Next week, we play it again with the second dowel pin in that same crank/axle. I figured the shrink fit was going to move the crank hub radially on the axle due to the clearance between them. Once the first pin expanded in its hole, it would stake the hub to the axle. So, being "retirees" we are doing one pin a week. We have at least two other cranks to dowel, so Maynah's MT adapter is going to be getting some real good use. Thanks again !
 
John.K

The GE 50 ton locomotives were designed for switching service. These locomotives were often used as industrial switch engines. Having all axles powered was a necessity since these locomotive had to work on tight curves and switches and move cuts of loaded freight cars. GE initially used side-rod connected axles on each truck (bogie). At some point they went to using a heavy roller chain and sprockets to connect the drive axle to the driven axle on each truck. This locomotive has side rod connected axles. This design allows the use of one traction motor per truck, with both axles powered. Truth be known, on a tourist railroad, having a locomotive with exposed cranks and siderods is 'part of the experience'. People like seeing the siderods moving as the locomotive shuffles along.

The other reality is tourist railroads such as the Catskill Mountain RR tended to have limited funds. As such, they often bought locomotives that were beyond being hand-me-downs. I do not know the history on this particular GE 50 tonner. It was built in 1946 and went thru a number of owners, winding up on a tourist/short line operation in New England before we got it. I have no idea what the 'designed service life' for this locomotive was. I am sure it was a lot less than 78 years (its current age). I am also sure the locomotive was repowered at some point. This was simple enough to do as Cummins 6 cylinder diesel engines, same as were used on heavy trucks of that era, were used in this locomotive. In my experience as a mechanical engineer, I did some very limited design work using Timken Roller Bearings. Timken has excellent engineering data. From this data, a projected service life for the roller bearings can be calculated. I made the mistake of playing with these calculations once when we were working on a similar diesel locomotive (an H.K. Porter, center cab, side rod connected trucks, Cummins power). When I figured the service life using Timken's engineering data for the roller bearings in the gearcases of that H.K. Porter locomotive, it was a case of "ignorance is bliss". The roller bearings in the gearcases on that locomotive were well over several times the service life.

These industrial switch engines were likely built for a relatively short and hard service life. They are handy little locomotives as they can be hauled over the roads on low-boy semitrailers. The Cummins engines are something maintainable with readily available parts and something local mechanics used to truck and heavy equipment engines can work on. In tourist train service, these engines run further and likely faster than they ever ran in industrial switching work. I am guessing that starting cuts of heavy freight cars put quite a pounding into the side rods and started the issues with the cranks loosening on the axles. Then, running with the throttle pulled wide open over runs of some miles with tourist trains hammered the cranks on the axles. One crank actually has been drilled and bolts tapped thru its hub into the axle. Thiswas done ages ago, and even the crankpin on that crank has been bolted. That crank flops around on the axle end like a fresh caught fish on the boards of a dock. The slop between that crank hub and axle end is so bad I stuck the blade of my jacknife into the gap. We will see how the dowel pin repair holds up on this crank/axle we are doing now. By spacing the dowels at 120 degrees, we figure the dowels will act like the jaws on a lathe chuck, using the force from the shrink fit (and resulting expansion of the dowels) to lock the axle in place in the hub.
 
Joe Michaels ,
It was interesting to read about your repair on the locomotive.
There is a similar one that is operated by the Bytown Railway Society in Ottawa Ontario
Here are some links about it .
They have been doing ongoing repairs and upkeep to it .
These are links to their public Facebook site so you don't have to log in to see them
Some more pictures of it here.
1713561667724.png
1713561933972.png
 
Jim Christie:

Thank you for posting the pictures & links about the GE 50 ton locomotive. The locomotive on the Catskill Mountain RR is a twin sister to the one in your post. It was also built in 1946, and like the one you posted about, went to work on an industrial site. The CMRR locomotive went to work at The Stanley Works in New Britain, Connecticut and switched cars there for a number of years.

In the pictures you posted the outside cranks & side rods are plainly seen. It is those outside crank webs that worked loose on the stub ends of the axles. Hopefully, Bytown's Number 10 is not suffering from the same malady. In the case of CMRR number 42, I believe the damage was caused by using the locomotive a lot harder than it was ever intended by the designers. These were primarily industrial switch engines. If they moved 2-3 loaded freight cars a short distance at slow speeds, that was the normal service. Short Line/Tourist RR service is often a lot harder than what this type locomotive was designed for. By the time the GE 50 tonner arrived on CMRR property, at least two (2) of the outside cranks had already worked loose on the axle stub ends. Half-assed repairs made by ringing the end of the stub axle/crank bore with weld had been done a few times over. This welding had cracked thru ages ago and I could stick the tip of my jacknife blade in the crack as well as the gap between the axle stub end and the bore in the crank hub. One hub had been bolted with bolts tapped radially thru the crank hub and partway into the axle.

Another issue that came up with the CMRR 50 tonner was the engine alternators. The original 32 volt generators were long gone and alternators had been retrofitted to the Cummins diesel engines. Railroads (like farm lighting) used 32 volt DC current for lighting on locomotives. On the GE 50 tonners, the 32 volt system did more than charge batteries and run lighting. It also provides excitation for the traction generators. Over the years, people kept running whatever 32 volt alternators they could find. These alternators were too light for the job, not enough amperage, and usually fried before too long. I was able to source some 70 amp alternators for charging 32 volt systems from a marine electric supplier. These are capable of 100 amp output if turned fast enough.

Hopefully, Bytown has level track with little to no grades, has lighter trains and does not have hotshots jamming the throttle right to the stop.

Members of this 'board will recognize the name Bullard- the maker of vertical turret lathes, formerly in Bridgeport, Connecticut. Bullard also had their own switching locomotive. The difference was Bullard bought their in the 1920's or 30's and went with a steam saddle tank engine. It is a unique little locomotive in that it was ordered with an oil fired boiler, and cab setup for one man operation. It has an enclosed all-weather cab rather than the somewhat open cabs on traditional steam locomotives in the USA. I seem to recall another unique feature of the Bullard locomotive was that it had roller bearings. It may well be the only steam tank engine in the US to have roller bearings. The Bullard switch engine is (or was) sitting at a museum in Connecticut.

The Stanley Works, the original owners of CMRR number 42, also had a 'fireless' locomotive on their property. Fireless locomotives were rolling thermos flasks. The boilers were filled with water at or above the 'saturation temperature' for steam at the corresponding pressure. When the throttle on a fireless locomotive is opened, the hot water at the surface in the boiler barrel flashes to steam and enters the dry pipe to the cylinders. Fireless locomotives worked well in powerplants and industrial plants where there was a stationary boiler plant. Heated feedwater was drawn from the stationary plant to partially fill the 'thermos flask'/boiler barrel on the fireless locomotive. These locomotives otherwise resembled steam locomotives and could usually do a good day's work of switching cars on one fill of the thermos flask. Fireless locomotives saw extensive use in munitions and chemical plants where no combustion or source of ignition was allowed. They also saw service where the locomotive had to move freight cars in and out of buildings. Other than exhaust steam, there was no smoke or diesel fumes.

It is nice to daydream about operating a steam locomotive on CMRR, but the realities of steam locomotive operation are something we know quite well. These was good reason industry as well as the railroads dieselized. The locomotives like these GE center cab engines took things to the next level. These locomotives are small and handy for tight track curves, and with what amounts to truck diesel engines, can be easily maintained. I believe GE offered this same locomotive in a 45 ton version. The reason for the 45 ton version had to do with some rules (possibly unions) requiring a two man crew on any locomotive weighing more than 45 tons. This was a carryover from steam locomotives, when a fireman was a necessity. The era these center cab locomotives came into was a time when the railroads in the USA were predominantly steam powered. Going to diesel was going to lay off a lot of firemen. GE made the 50 ton version by adding more steel ballast, real havy plate welded onto the locomotive to add the extra tonnage. These little locomotives have surprisingly high drawbar pull for the HP ratings of engines and traction motors. They can break loose and haul a surprising tonnage, which leads to issues such as we are dealing with on the cranks.
 
Joe ,
Its always interesting to read about your experiences.
I don't know how much work Bytown had to do to get their locomotive running and haven't found much on line about their work on it.
I do know some of the people involved with it so I will post some more later if I learn something .
I did find more about the Bytown's locomotive and it's time on the Railway at Thurso Quebec
I posted these in this older Singer Thread some time ago

Here is the link to the picture below from Colin Churcher's site.
1713747163015.png
Jim
 
Last edited:
Joe,
If I correctly understand your repair, you used what I have called a "dutch key". I have applied a version of this with some success, that achieves interference fit without reaming or heat.

After making the hub fit the shaft as well as convenient, I drill and tap for standard pipe plugs at the shaft OD/hub ID interface. I have used a 1/16" pipe plug at the bottom, with a 1/8" plug on top of it, to reach down through the thickness of the hub. Headless (allen-socket) pipe-plugs, of course, lubricated, and torqued as tight as I dared.
 
Magnetic Anomaly:

Some time back, my buddy Earl did pretty much the repair you describe. The crank hub was already loose on the axle, the keyways battered and deformed beyond any real use. Earl took his magnetic based drill and drilled a hole half in the hub, half in the axle. He tapped the hole 7/8"-9 UNC. Into the hole he ran a railroad track bolt. This bolt was run in with a pipe wrench & cheater on the wrench. The bolt was cut off flush and ring welded with E 7018. In almost no running time, the E 7018 weld had cracked thru, and what had been tightly screwed in was loose in the tapped hole. The forces acting on an 'overhung crank' on even a light diesel-electric locomotive's drive axles are more than the nominal HP of the traction motors would seem likely to produce. When these axles and cranks are assembled, a very heavy press fit is used. A locomotive wheel press of somewhere around 1200-1500 tons is used to press the wheels onto the axles, and the cranks onto the stub ends of the axles. The keys are usually an interference fit on locomotive drive axles, and are driven in with a rivet buster. As I noted, by the time we got the locomotive on the Catskill Mountain RR, it had gone thru a few owners and was about 60 years old. During this time with previous owners, the cranks had worked loose on a maybe 4 out of 8 of the stub ends of the driver axles. People had vainly attempted to weld things together, had drilled and tapped in bolts on one hub to try pinning the hub to the axle, with no success. What had been a heavy press fit had opened up to where I could stick a jacknife blade in the looser locations, and a 0.020" feeler in most others.

The shrink fitted pins are at 120 degrees to each other, with the shaft key making the third point. I machined a shaft key based on readings taken with adjustable parallels, telescoping gauges and micrometer. The keyways in the axle stub end and hub were so battered and deformed I milled a stepped key with a tapered entry end. I milled this key oversized, and used some manganese alloy steel from a railroad switch track part. We drove this key in using a 16 lb beater and cut it off flush as it started to buckle before reaching full depth. We welded this key in place. By the time we got set to put in the shrink fitted dowels, the weld on the end of the key had cracked thru. Fortunately, this little locomotive does not go very far, and about 10 mph is all the fast it is run.

In an ideal world, the locomotive would be taken off its trucks, and those would go to a railroad backshop. Aside from the issues with the crank hubs worked loose on the axle stubs, the original Timken Roller Bearings from 1946 are on the axle journals. The journals are inboard of the cranks, and one runs thru a gearcase with a bull gear on it. A traction motor in each truck drives one axle via gearing. This axles drives the other axle via side rods. I suggested bringing in a contractor with an orbital lathe and welder and portable keyway miller. We'd be taking a chance on getting the 'quartering' (orientation of the keyways/cranks on opposite ends of each axle) right. But, as sloppy as things are, I figured we could layout the key locations on the ends of the axles before field machining and buildup welding. Once we had the axles built up and field machined, we'd send the crank hubs out for reboring/and freshening up the keyways. We'd use a shrink fit to set the crank hubs on the axles. This idea was shot down as too time consuming and too costly for what use the locomotive is. It is not the locomotive used for hauling tourist trains and is mainly in work train service or 'helper' service as there is a steep gradient on part of the line. When "Polar Express" is run with a full consist of coaches, this 50 ton center cab engine is tacked on the tail of the train as a helper.

My repair idea with the shrink fitted dowels was a few steps up from the normal backwoods ways of repairing things like this. I am sure you know how farmers, loggers, and backwoods types keep trucks, tractors, sawmills and similar held together and working. I went to a memorial service a few weeks back for a mechanic from the old crew at the powerhouse. He was a logger before he came to work at the powerplant and kept at it while employed there (about 30 years). In retirement, he resumed logging. At the service, his family put a roll of 'duct tape' amidst the pictures of him displayed at the funeral home. We joked that they'd forgotten the "JB Weld" and the baling wire (aka "mechanic's tie wire"). The shrink fitted dowels, with calculated shrink fit allowances and reamed holes were a quantum leap from that school of doing repairs.
 
We agree on the "right, right" way to fix such things. Pre-load is essential, and shrink-fit is the practical and traditional way to get it.
I remember when I was slowly upgrading some explosive-test pressure vessels (chasing failures), more by trial and error than by calculation, also constrained by the fact that there was no money for proper re-design. A rule of thumb sticks in my mind, that mild steel, or 304 stainless, will tolerate 30,000PSI in bearing...that is to say that if the bearing face under a bolt head had to carry more than that, it would indent the flange and pre-load would be lost. At that point, grade 8 bolts were worth no more than wrought-iron.
That is a failure-mode for shaft keys and keyways, too. A standard taper-key does not really address this, because the taper is radial, not tangential. As it is driven in, it tightens the diametral fit, increasing friction, but there is no tightening between the flanks of the key and the sides of the keyway unless the key is cold-forged in place. Probably it would buckle before that happened to a useful degree.

Some bulldozer (earthmoving bulldozer) sprockets are mounted to the axle with tapered splines....But the ones I have worked on are not REALLY tapered splines, which would have to be tapered in all dimensions, like bevel gears. That would be a trick, machining a 10 degree angle internal bevel-gear form inside a hub. The "tapered splines" I have seen are straight splines, cut on a tapered shaft, or in a taper bore. The splines fit pretty well, and then the smooth tapers lock up..........and after umpty--thousand hours, they wallow out and the logger comes to the guy in the woods with the welder and the chisels and files and soot.

A COTS "solution" offered for high-torque reversing shaft connections is various kinds of tapered or distorting bushing assemblies. "Ringfeder" was the brand I first became aware of. I think they are used instead of stud-mounted flanges to secure semi-floating drive axles to hubs on some European heavy trucks. I do not know how well they work. They were always too expensive for my customers. I did once make a double-tapered shaft coupling for the main shaft in a sawmill. I used keys also, so not a fair test of tapers alone, but I heard recently that it is still running 20 years later.

I almost had the chance to try the tapered-pipe-plug repair on a steam loco near here (Bull-gear-to-drive-axle, IIRC), but the owner's own mechanic had ideas of his own, so I did not get the chance to try.

Welding on top of press-fits I have always thought of as a recipe for failure. The heat cycle anneals the material in the load zone, relaxes the interference, and leaves residual tensile stress. Three strikes, as I see it, and one would be enough to cause failure. After all, heating or welding are the tools-of-last-resort we use to disassemble press fits, AKA The Blue Wrench.

In theory, the tapered threads of pipe-plugs create a lot of interference/preload. Since it exists all around the circumference of the plug, some exists at right angles to the shaft-hub interface, so it should resist fretting by torque load of that fit. An overloaded bolt-head will indent the part of the flange it bears on partly because the metal can flow out of the way, into the clearance around the bolt-shank or up around the head. But the pipe-plug threads tend to prevent upward flow of metal around the plug (as an unthreaded taper-pin would not) , and there is nowhere else for metal to flow to, so it may be possible to exceed the 30KPSI load(or whatever it actually is in bearing with only one axis of constraint).

Of course I do not know if it would work on a crankshaft-to-crank-web joint. Few things stay tight forever...unless you are trying to get them apart.
 
Magnetic Anomaly:

You bring back memories of my early years with the NY Power Authority. I was assigned as a construction superintendent on some small hydroelectric projects. Small meaning units producing 1 to 3 megawatts apiece. In the case of two different turbine builders at two separate plants, both used Ringfeders in various forms in their turbine designs. One vendor used what amounted to a 'giant Swagelok' of a Ringfeder to couple a turbine shaft to a generator shaft. The turbine shaft was hollow, made of maybe 12" diameter steel tube. The end that coupled to the generator shaft had a steel end section, spigotted fit, but welded to the tubular shaft. This end section was hogged out of solid bar stock and necked down from 12" to maybe 10" OD x 8" ID. The 8" ID was bored to a very close fit on the generator shaft. No shaft key was used. The Ringfeder choked around the 10" OD section, and when made up, created the equivalent of a shrink fit.

The other application used smaller Ringfeders which expanded in an annualr space to make a friction coupling. These were made up with their bolts torqued to a specified torque, enabling the Ringfeders to slip under excessive load. These were used on wicket gate arms, the idea being that if debris jammed a wicket gate on closing, the Ringfeder would allow the gate arm to slip on the gate stem.

Ringfeder is the German for "ring key". A 'Feder' in German is a feather, which is how they refer to shaft keys. A key for a door lock is a "Schlussel" if I remember my German rightly.

On our railroad, we never had the luxury of a shop or locomotive wheel press. We had another small diesel locomotive, also with outside cranks and rod connected drive wheels. This was an H.K. Porter locomotive built for the Navy. It had a chronic problem with failures of one sort or another in the trucks. We got pretty adept at jacking that engine up with air-driven jacks (using the locomotive's airbrake compressors to run the jacks). We'd slide the damaged truck out on scrap rails coated with grease using a backhoe. I know at least once, we had to remove one of the outside cranks and replace an axle bearing (roller bearing) and make a new bearing spacer. We rigged a puller with the heaviest Porta Power ram we could borrow, heated the crank hub with rosebuds and weed burners and beat on it. It popped off. I made a new bearing spacer out of an un-used large Falk coupling hub (from the local surplus store). It was a job I ran on an ancient Niles vertical turret lathe. We put the crank back on using more heat and a lashup with a couple of Porta Power rams and some sledging. That locomotive had traction motors at 90 degrees to the axles, and used a bevel pionin and ring gear + pinion and bull gear (on the axle). Periodically, the locking nuts holding the bevel pinions on the motor shafts would work loose. This with a locking taper fit, key, and locknut and cotter pin thru the shaft end and locknut. We got pretty adept at removing the trucks, and removing the traction motors for repair. This usually consisted of building up the tapered fit on the shaft with welding, recutting a new taper (blued into the fit in the pinion), cutting a new keyway, and building up/recutting new end threads. We made locking nuts out of 4140. We gave up on cotter pins and after socking up the lock nuts, would ring weld the exposed threads on the shaft end to the nut with E 309L or MG 600. The problem was likely due to our taking a locomotive built for industrial switching and moving cuts of passenger coaches over a couple of miles of track at more speed than was healthy for it. Eventually, that locomotive shed a pinion off a traction motor shaft and ran it thru the ring gear. Bent the ring gear and damaged the teeth. That locomotive has been sitting ever since, no chance of repairing it. The traction motors are 400 volt Westinghouse, so using trucks from more common center cab locomotives (GE) won't happen. The more common traction motors are wound for 600 volts. The 50 ton GE locomotive we've been working on is in service, though not used for pulling event or tourist trains. The HK Porter locomotive will sit until who-knows-when. We had a belly full of jacking it up, sliding the damaged trucks out from under it, and making repairs, knowing it was a matter of time until we had another such repair.
 
We agree on the "right, right" way to fix such things. Pre-load is essential, and shrink-fit is the practical and traditional way to get it.
I remember when I was slowly upgrading some explosive-test pressure vessels (chasing failures), more by trial and error than by calculation, also constrained by the fact that there was no money for proper re-design. A rule of thumb sticks in my mind, that mild steel, or 304 stainless, will tolerate 30,000PSI in bearing...that is to say that if the bearing face under a bolt head had to carry more than that, it would indent the flange and pre-load would be lost. At that point, grade 8 bolts were worth no more than wrought-iron.
That is a failure-mode for shaft keys and keyways, too. A standard taper-key does not really address this, because the taper is radial, not tangential. As it is driven in, it tightens the diametral fit, increasing friction, but there is no tightening between the flanks of the key and the sides of the keyway unless the key is cold-forged in place. Probably it would buckle before that happened to a useful degree.

Some bulldozer (earthmoving bulldozer) sprockets are mounted to the axle with tapered splines....But the ones I have worked on are not REALLY tapered splines, which would have to be tapered in all dimensions, like bevel gears. That would be a trick, machining a 10 degree angle internal bevel-gear form inside a hub. The "tapered splines" I have seen are straight splines, cut on a tapered shaft, or in a taper bore. The splines fit pretty well, and then the smooth tapers lock up..........and after umpty--thousand hours, they wallow out and the logger comes to the guy in the woods with the welder and the chisels and files and soot.

A COTS "solution" offered for high-torque reversing shaft connections is various kinds of tapered or distorting bushing assemblies. "Ringfeder" was the brand I first became aware of. I think they are used instead of stud-mounted flanges to secure semi-floating drive axles to hubs on some European heavy trucks. I do not know how well they work. They were always too expensive for my customers. I did once make a double-tapered shaft coupling for the main shaft in a sawmill. I used keys also, so not a fair test of tapers alone, but I heard recently that it is still running 20 years later.

I almost had the chance to try the tapered-pipe-plug repair on a steam loco near here (Bull-gear-to-drive-axle, IIRC), but the owner's own mechanic had ideas of his own, so I did not get the chance to try.

Welding on top of press-fits I have always thought of as a recipe for failure. The heat cycle anneals the material in the load zone, relaxes the interference, and leaves residual tensile stress. Three strikes, as I see it, and one would be enough to cause failure. After all, heating or welding are the tools-of-last-resort we use to disassemble press fits, AKA The Blue Wrench.

In theory, the tapered threads of pipe-plugs create a lot of interference/preload. Since it exists all around the circumference of the plug, some exists at right angles to the shaft-hub interface, so it should resist fretting by torque load of that fit. An overloaded bolt-head will indent the part of the flange it bears on partly because the metal can flow out of the way, into the clearance around the bolt-shank or up around the head. But the pipe-plug threads tend to prevent upward flow of metal around the plug (as an unthreaded taper-pin would not) , and there is nowhere else for metal to flow to, so it may be possible to exceed the 30KPSI load(or whatever it actually is in bearing with only one axis of constraint).

Of course I do not know if it would work on a crankshaft-to-crank-web joint. Few things stay tight forever...unless you are trying to get them apart.
Those Ringfeder couplings work really well. However I have never used them on any thing larger than 2".They claim to be better than a key. Slow tapers are about the best, hard to beat but if the oe design has no provision for a clamp bolt or nut then that is another problem. All of those pin / plug interference patches don't make up for the lost surface area contact of the oe fit. They are all point loaded which start to fail as soon as they start to rotate under a load when every thing starts to flex ever so sightly.... at first.
I understand fully the situation they put you in when there is no money for the correct repair and expect the best of the worst to get back running!
I used to get that al the time when I first started here. The difference is not lack of money but time. However I have resisted or played dumb when asked to patch any thing to get by. The patch means at least 2 down times and possibly more collateral damage the next time. So its cheaper to fix it once right.
 
It is interesting that a lot of equipment, presumably designed by competent people employed by large, reputable companies, nonetheless fails. Of course I do not expect anything to last forever, but we expect and predict things like rolling-element bearing failure, steady wear of friction materials, slow wearing-away of teeth of worm-wheels and poorly-lubricated bushings. But it seems to me that we ought to know enough about stresses and materials that failures of static load-transmitting fits, should not have to happen.
I have never worked in an environment where products are designed for mass-production and sale. It may be that a machine designed for indefinite life would simply be too expensive to be built at a profit. But there are many load-transmitting parts of mass-marketed machines that simply do not fail, at least in my experience. One silly handy example is car and truck wheel attachments. As long as they are not GROSSLY overloaded,and are more-or-less properly torqued on installation, normal tapered-seat lugnut mounted wheels, Budd or Dayton-style wheels on trucks, seem to stay on, I cannot remember ever seeing a failure of axle-drive-flange-to hub connection, whether with or without split tapered bushings. Tapered-shank tie-rod ends do not (again, if properly installed) work loose.

I am going to guess that the problem with railroad equipment, is the shock loads occasioned by steel-wheel-on-steel-rail, which must propagate back through the drivetrain, and must multiply the calculated service loads by a large factor.
 
Magnetic Anomaly:

Thank you for your insightful posts which are based on hard experience combined with an excellent working knowledge or sense as to how materials behave and stresses develop and act on them.

With the GE Center Cab locomotive and the 44 ton HK Porter locomotive, the failures can be attributed to two (2) simple, obvious root causes:
-use of these locomotives for much harder service than they were ever designed for
-use of these locomotives for many times whatever their original design service life

Getting to the first root cause, these locomotives were designed for industrial switching. Moving a few freight cars at relatively low speeds over short distances.
In "tourist train" service, these locomotives were run to pull possibly heavier consists (trains having more tonnage) and run further and faster than in industrial switching jobs. In the case of the HK Porter center cab locomotive, it was delivered to the US Navy at Portsmouth, NH Naval Shipyard sometime during WWII. It spent its working life moving freight cars within the shipyard. Likely very slow, relatively short moves of one or two freight cars. At some point, the Navy replaced the original Cummins diesel engines with newer engines. This is where the story gets interesting. The locomotive was sold to us as surplus in the 1990's. Shipyard workers told our guys the story when our guys went to inspect the locomotive. After the new engines were installed, someone operating the locomotive never opened the battery disconnects or some other 'main switch' after securing the locomotive following working it. The locomotive was by then considered as a 'secondary' unit, so was parked on a siding or spur track and left for some time. When it was needed next, some months had elapsed. The batteries were flat, and the locomotive would not crank nor had it any cab lights or electrical systems working. Someone looked at the hour meters for each of the 2 Cummins engines and saw several thousand hours. This was due to the battery switches being left 'on', not actual run time. This hour meter reading caused the engine to be considered as not worth repairing (not that it needed repairs at that time), and it was surplussed. We bid a low price and got it. We ran it onto a lowboy trailer under its own power with temporary rails and cribbing to make a ramp track.

Once the engine went into service on our railroad, it was run hard and some engineers ran it with the throttle pulled wide open. A contributing cause is the fact that each engine runs independently of the other with ballhead mechanical governors used to set engine speed. The throttle linkage is sloppy, and works the speed control on each of the Cummins engine governors. Plenty of lost motion in that linkage, and mechanical tachs (speedometer cable drives) on each engine are not real precise in terms of reading rpm. I know the traction motor in one truck often got ahead of the other in service and we'd tinker with the throttle linkages. I lost count of how many heavy repairs we did to the traction motor armature shafts on that locomotive. I attribute the failures we saw to heavy shock loads transmitted back into the motor armature shaft. The bevel pinion on the armature shaft has a tapered fit, key, and heavy locking nut. The shock loads were such that the locking nut would bust welds and spin off the end of the armature shaft, then the pinion would break loose of its fit, sometimes shear its key, and sometimes freespin on the shaft. One traction motor armature shaft actually snapped between the pinion tapered fit and somewhere at or in the bearing journal in the motor end bell. We had a machine shop repair that armature shaft. We scavenged some traction motors from a similar locomotive derelict in a closed cement mill. My other guess as to a contributing cause of the failures is the fact the locomotive drive wheel axles are siderod connected. As things wear (shoes and wedges), movement of the axle boxes (bearing housings) in the bearing pedestals in the truck frames fore-and-aft occurs. The axle boxes are working against spring suspension, so vertical motion is designed for. Wear on the side rod end bearings also happens, and this will affect the 'quartering' of the cranks (position of the crankpins across an axle from one side to the other).
Get the crank quartering messed up and more shock loads result.

As I wrote, the H.K. Porter locomotive finally spit a bevel pinion gear into its ring gear. The ring gear is bolted to a heavy 'center', or disc that is pressed and keyed onto the axle within the gearcase. This ring gear and the heavy forged steel disc actually bent and jammed the bevel pinion. Usually, the bevel pinion would be spit off the armature shaft and land in the sump at the bottom of the gear case with the worst damage being a totally destroyed tapered fit and stripped end threads on the armature shaft. This last failure was a true "FUBAR" for the H.K. Porter locomotive. Hard use, way harder than it was designed for, and way longer than it was designed for finally killed it for good.

The GE Center Cab locomotive uses traction motor gear drives which have the armature shafts parallel to the axles, and no compound reduction (as was used on the H.K. Porter locomotive). Since the traction motor/axle gearing is much more rugged, the outside cranks bore the brunt of the damage. On steam locomotives, the side rods which transfer power to the driving wheels are trammed to check rod lengths. Over time, in service, particularly if there was a derailment or some other accident, rod lengths get out of tram- the rod dimensions may lengthen or shorten. The rods would be sent to the blacksmith shop to be straightened and either lengthened or shortened (upset forging was used for this) to restore them to meet the tram dimensions. Similarly, when a steam locomotive was 'shopped', the 'quartering' of the crankpins was checked. Crankpins often wore to the point that the angular relationship of the crankpins from one side of the axle to the other was changed. Any rod connected locomotive will have a 'push-pull' set of forces working on the crankpins and axles. Over time, these forces can cause the axle to move fore-and-aft in an angular motion as shoes and wedges in the axle bearing pedestals wear. The GE center cab locomotive, being rod connected on its drive wheels, is a prime example of this. Lots to study and reflect upon, but it comes down to time and harder used for way longer than those locomotives were ever designed for and no heavy shopping to replace axle and rod bearings or rebuild shoes, wedges and pedestals.
 
Thanks for the details, Joe! I have spent enough time ogling the rolling-stock at Cass to put a mental image to each piece you refer to, and I see how accumulated slop everywhere multiplies the loads.
My father exposed me too early to "The Wonderful One-Hoss Shay, or The Deacon's Masterpiece" I am just programmed to believe against all logic and experience that things should last (almost) forever.
 








 
Back
Top