Welded track and steam operations

[Anonymous]

I understand that large steam locomotives, even the so called "balanced" ones, could really "pound" a track because of the up and down movement of the side rods and counterbalanced drivers. Track then was the segmented, bolt joint "clickety clack" rail.

Nowadays track is welded in long segments. It seems (might be a perception on my part) that it's heavier, too. My primitive understanding is that diesel and electric locomotives still bring stresses to the track, but that it's much different than a steam locomotive.

So, my question is this: Would today's heavier, welded rail withstand the stresses of yesterday's road freight steam engines better, or worse?

My notebook and pen are out awaiting the physics and mathematics answers.

Erik

[Anonymous]

CWR is also a lot easier on equipment in general, everytime a wheel would hit a joint it would produce wear and tear, resulting in more maintenance all around.

CWR also provides a much smoother ride to those few people who still partake in passenger rail service.

[BaltACD]

Riding in a locomotive or on a passenger train, outside of the clicking from jointed rail, one can't appreciated the ride differences between jointed and welded rail.

Ride a caboose over territory that contains both and the differences are brought home in spades. Caboose trucks are sprung as stiffly as freight car trucks and the ride over jointed rail is jaring to say the least. Traversing segments of both welded and jointed rail, you pray that the next welded rail segment is quickly approaching.

[cefinkjr]

Erik:

Your "perception" of heavier rail is basically correct. Although steam pounded 155 lb (per yard) rail on PRR's Horseshoe Curve, I don't believe anyone else has ever used rail that heavy. On the other hand, 133 lb (or thereabouts) is much more common today than it was when steam was the dominant form of motive power.

Chuck

[jeaton]

Mudchicken would know more about this, but word is he is on the road. He had told me that I would be riding on jointed rail on in Kansas and Colorado on the Santa Fe "pasenger" route, now the route of the Southwest Chief.

Perhaps not as dramatic as the ride in a caboose, I can assure the the difference in the ride quality between jointed and welded rail in a SuperLiner is big.

Jay

[Anonymous]

N&W had about ten miles of a test application of 155 Lb. rail east and west of Webb, W. Va. on the Kenova District. Their normal standard at the time was 132 Lb.

The steam locomotive is going to pound the rail, whatever its length or weight. The heavier the rail, the better it will withstand the pounding.

The side of the engine that had the lead (the drivers being "quartered" - the piston on one side on dead center and the other side at mid-stroke - the piston on one side will therefore begin its stroke before the other, and is said to have the lead) will tend to pound the rail harder.

Most US engines had right-hand lead. Pennsy used mainly left-hand lead.

Old Timer

[jeaton]

OT

In the category of one is never too old to learn something new, your note about the lead side being harder on the rail falls there.

I am not very knowledgable about the details of steam locomotives, and it was only a few years ago that I learned about the drivers being quartered. I think I picked that up in a story about an N&W engine having lost the cylinder or rods on one side in a wreck and being run light to a shop for repairs. It was noted that if the engine was stopped with the piston dead at either end of the cylinder it would take a bit of a roll to get going again.

Jay

[M636C]

The problem with diesel locomotives is the "Unsprung Mass". The normal US locomotive has traction motors which are described as "axle hung", the motor being supported on bearings directly on the axle with the other end, the "nose" being carried on rubber bushings on the truck frame. The frame is carried on springs, but the axle of course sits directly on the rail, so at least half the motor mass is "unsprung". This is an important cause of impact at rail joints, and for high speed locomotives particularly, reduced unsprung mass is important. The British Rail (now GNER) Class 91 electric locomotives intended to run at 140 mph, have body mounted motors driving through cardan shafts and right angle gear boxes.

Peter

[hmftrs]

Follow up question; Why was the JingPing line in China built with jointed (and even joints) instead of cwr?? This line was built in 1992

[dknelson]

If you check the engineering standards of some railroads, some used "suspended" rail joints where, as Mark H said, the rail joint was a break in the beam. That is, there were ties on both sides of the joint but not underneath it. (What they would do however was use tighter tie spacing.)
Other railroads would put a tie directly under the joint, although how the rail was fastened to it, assuming it was fastened, is not clear to me. Whether this had any practical effect I do not know but it would seem there should be some beneficial effect.

Having said all that, two weekends ago when the Soo Line 1003, a 2-8-2, was running photo excursions on the Wisconsin & Southern for the late Dave Goodheart, a welded rail snapped under the locomotive (fortunately not failing entirely until the historic wood Soo Line caboose had passed over it), cancelling the first planned runby of the day. The rail is welded but it has rolling dates in the 1920s. The right of way itself dates from 1856 and can get rather rough.
Dave Nelson

[Anonymous]

Thanks for the answers!
It seems that CWR came into use sometime in the 70's- maybe the 80's? If the improvement in ride and (I assume) maintainability costs were improved as significantly as everyone seems to say they were, how come railroads didn't use CWR much earlier? Was there a change in welding technology or in metallurgy that enabled CWR?

Erik
"Still blinded by the science"

[jockellis]

I remember looking up railroads in our new 1957 Compton's Encyclopedia as if it were yesterday and the photograph of CWR being installed. That was the first I'd ever heard of it, but, of course, I was only 9 years old. When did the long rail begin to be used? Somehwere I thought I read that in the days of steam, fairly light rail was used because it tended to move with the heavy drivers and didn't get beaten up so badly. Did I misread that?
Jock Ellis

[dldance]

Like any other technology, CWR required more than just the welding development to become industry practice. Some of the other items included:
- Development of long rail transport and delivery
- Mechanization of rail laying as opposed to a track gang
- Control of expansion joints
- Inspection technologies that recheck the welds
Coupled with the long life of stick rail in most locations, developing these methods slowed CWR installation for many years.

dd

[Overmod]

There are some other considerations that may be of interest.

I think it may be valuable to separate the potential track damage from steam locomotives into several categories (mudchicken, when he reads this, may revise this to a better set):

Railhead damage
Rail deformation
Rail breakage through shock or defect propagation
Lining/surfacing defects caused by 'knocking' the rail in various planes
Lining/surfacing defects caused by pumping rail against the ties, and track against the ballast

Note that I'm leaving the joints out of this discussion. Comments so far are accurate on that subject, but I think the original question involved continuous track structure more than joint effects. One comment: it might be considered that the frequent shocks and high forces caused by steam locomotives might keep bolted joints a bit more free -- helping, for example, eliminate sun-kink problems -- than smoother operation with well-suspended diesel-electrics...

OT might have mentioned that N&W was one of the more advanced railroads in the country -- I might argue in the world -- with respect to steam-locomotive balancing theory and practice. The J-class 4-8-4s had an interesting method of eliminating many of the 'problem' areas of dynamic balancing.

One consideration that hasn't been mentioned is that modern rail has a different metallurgical composition, and a different required method of 'head hardening'. I believe that much of the theory and adoption behind current rail steels has presumed the absence of high shock loading (as would be generated by conventional large 2-cylinder steam locomotives with 'normal' cross- and overbalancing). I would expect to see aggravated problems with crack propagation in the martensitic railhead-surface layer, with gauge-corner cracking problems in general, and perhaps with catastrophic crack propagation on chilled sections of LWR under tension.

One problem with conventional steam is that it's impossible to balance a locomotive dynamically (at the main driver) over a wide range of speeds. That is because the main rod gyrates with the crankpin at the driver end, but is constrained to move only longitudinally at the crosshead end; the section of the rod is also changing (normally becoming heavier toward the crankpin end) and of course increases dramatically in the region of the big-end and its bearing. While it is possible to calculate or measure quite accurately what the effective rotating equivalent of the rod's mass is -- and therefore possible in theory to balance the engine "perfectly" dynamically [gotta love those double adverbs!] -- the problem is that the kinetic energy / momentum of this equivalent changes with speed in a way that is not proportional to the kinetic energy / momentum of the (rotating) balance weights on the drivers. This force also varies as the square of the rotational speed, and (for example) doubles between 60 and 80mph. The problem is not limited just to hammering, though: the axles of a locomotive are suspended on springs, and normally have no active 'shock absorber' damping other than the viscosity of the lubricant on the axlebox faces and whatever pressure the self-adjusting wedges (on modern power) exert. The vertical component of the big-end momentum is free to act on the springing... and it's easy to see that there can be 'critical speeds' where there is resonant pumping of the suspension (this is complicated somewhat by the equalizing system, but can be determined quite accurately by empirical testing on, for example, a properly designed rotating test plant).

Another consequence of steam-locomotive design is that some of the rod thrust is not longitudinal, but is directed in the vertical plane. This is a relatively small resultant, but with large cylinders can be a significant effect. Note that this may either add or subtract from the inertial force due to the main-rod geometry.

It should not be surprising to find that some express locomotives (I remember this described explicitly for C&NW 4-6-4s) would actually bounce the main drivers high enough that you could see "see daylight" under them -- i.e. even the flange was clearing the plane of the railhead!

Heavier rail ought to be capable of absorbing a bit more impact energy without damage, but at least some consideration needs to be given to the shape of the web and base. More metal in the head would allow more regrindings for the damage that slipping and hammer-blow might cause, including increased martensitic-derived cracking.

I would be interested to see any data AREMA, Pandrol, etc. might have derived for the behavior of spring-clip equipped track under the rhythmic and dynamic loads from steam locomotives. I would expect that many of the potential problems from short-duration high peak impact loads on track geometry might be relieved somewhat, if not actually solved, by having a sprung clamp on the rail and a resilient pad between it and the tie.



With respect to JingPeng, I can only suppose that the Chinese did not have equipment to fabricate and position LWR, and did not care to invest in it at the time. Be interesting to see if this has changed, and if so, what current practices are.

[Anonymous]

Having been a part of the locomotive crew on AWP 290 when in operation in the mid-1990's I can say from personal experience that CWR gives a much better ride than jointed rail, especially at speeds over 30-40 mph.

However, that is not an indicator of the kind of wear that is occurring at rail level. The dynamic augment (the pounding that occurs to the rail) occurs with all steam locomotives with side rods and pistons varies based on cylinder bore diameter, steam pressure, side rod weight, and counterbalancing combined with the overall weight of the locomotive applied to the drivers. Some locomotives were easier on rail that others due to lighter reciprocating forces or better counterbalancing. Also, dynamic augment was largely something that occurred at different rates at different speeds, again varying from one locomotive to another. Heavier rail would naturally withstand the effects of dynamic augment better than lighter rail; CWR would not have the inherant weaknesses that jointed rail possess, no matter what the effects of dynamic augment.

One other area of wear that occurs to rail is in curves. In the days of steam the roadbed was super-elevated in the curves, that is the outside rail of the curve was elevated higher than the inside rail, much like the banked turns of a race track. While there is still some superelevation used today, it was greater in the days of steam on most US mainlines. The present track dynamics used for todays diesels would need to be changed to reduce wear in the curves if steam were still in wide use. This would be due to the longer continuous wheel base of steam as compared to diesel.

Essentially, there will be wear to the rail and to the wheels of all locomotives over time. CWR at the present standard of around 130 to 136 lbs would have improved wear over the jointed rail of equal or lesser weight. More modern steam would undoubtably have improved the overall nature of the characteristics of dynamic augment by using stronger and lighter materials for the side rods and pistons along with better technology of steam utilization combining the relationship of bore, stroke, driver diameter, cutoff, and steam pressure.

Bill

[M636C]

QUOTE: Originally posted by hmftrs

Follow up question; Why was the JingPing line in China built with jointed (and even joints) instead of cwr?? This line was built in 1992

I think the location of the Jing Peng line in a high altitude desert area with a very high temperature range may be a major reason for the use of jointed rail. The risk of rail breakages in continuous rail in an isolated area may have been regarded as too high compared with the need to supply (still relatively cheap) labour to check the bolted joints.

Also, the "pounding" effect of the steam locomotives mainly used on this line is independent of rail joints, unlike the unsprung impacts of diesel locomotives which are directly caused by rail joints.

The Jing Peng was a privately financed railway, and their use of steam locomotives was related to the ready availability and low initial price of used steam locomotives compared to new diesels. Building the line with bolted rail in an area of relatively low labour costs may have resulted in lower up front costs to get the project under way.

Peter

[nobullchitbids]

Several of the writers mentioned experience on the N&W; today, the one railroad with experience of how steam behaves on welded rail would be UP. Does anyone who actually has worked for this road know what UP has found re how its "historic" engines treat their modern plant? I know UP was an early user of the 131-133 lb. rail in mainline applications and considered 2-10-2s and 4-8-4s to be "small" steam power. No doubt the UP has company data on the difference between 131-lb welded and 131-lb. bolted, and no doubt such a safety-conscious company would be well aware of the effect use of something as big as a challenger would have on track not laid with such locomotives in mind.

[oltmannd]

I'd have to go way back to my college days to be able to write the equations of motion for a reciprocating steam engine and calculate the forces. And, since Mr. Peabody is already using the only known "way back" machine, I'm stuck.

But, on the surface of it, I think you can pretty much balance the rotating mass, but the reciprocating side rod, et. al. is the real problem and most of the resultant force would longitudinal and could cause the locomotive to want to "waddle" or yaw more than any vertical "pounding".

I guess you could configure a lever driven balancing mass the would move opposite to the driving rod to counter balance the forces, but the complexity would probably exceed the benefit.

[jockellis]

Mark, you're probably right about the ravages of time on a memory. Maybe you could help me with something else apropos to this post. Now working in magnetic and flourescent non destructive testing for GE, I have become interested in the history of the discipline. But the history seems to be almost nil. Has Trains or any train mag ever done stories on NDT? I know it started as a result of railroad needs but that is about all I can find out.
Jock Ellis

[daveklepper]

Don't forget that today's pounding of the rails is by equipment with wheel flatspots, and the criterion for most freight cars is not sufficient to eliminate the bang bang bang we sometimes hear while a freight passes by. If there is noise, there is also mechanical energy.

[Overmod]

Jock, perhaps a useful entry point for your research would be:

http://www.magnaflux-online.com/overview/timeline.stm

My suspicion is that either car wheels or car axles would be the primary application of NDT, but I can't say whether one was adopted before the other. Note that a date of 1938 is given for Magnaflux testing of axles.

Might also be mentioned that a form of "NDT" was practiced on air reservoirs in the old days -- a fellow tapped every square inch of the exterior with a little hammer and listened to the resulting 'ring' -- if it sounded dead or otherwise 'off key', there was enough of a crack or defect to call for maintenance. I seem to remember reading about a similar procedure for steam-locomotive rods.

[Anonymous]

QUOTE: Originally posted by oltmannd

But, on the surface of it, I think you can pretty much balance the rotating mass, but the reciprocating side rod, et. al. is the real problem and most of the resultant force would longitudinal and could cause the locomotive to want to "waddle" or yaw more than any vertical "pounding".

I guess you could configure a lever driven balancing mass the would move opposite to the driving rod to counter balance the forces, but the complexity would probably exceed the benefit.

I have often wondered about that kind of lever mechanism myself. Does anybody else have any information on that?

-Daniel Parks

[Overmod]

Sounds like what the Russians were trying... I repeat, TRYING to do with the levers on the OR class... uzurpator, you might have some distinctive information not otherwise on the Web regarding these engines.

Note that the 'salmon' rods on Cossart valve gear were intended to do this sort of thing.

Probably the 'best' solution is balanced four-cylinder simples, with receiver-balanced (a la Chapelon) four-cylinder compounds closely behind. An interesting geometrical take on this idea is the Withuhn conjugated duplex (employed on the ACE 3000 design, and described in a 1974 issue of Trains Magazine) which has the net thrust couples balanced across the engine bed, but still has an unbalanced lifting component acting on the suspension of individual axles.

[Overmod]

Sounds like what the Russians were trying... I repeat, TRYING to do with the levers on the OR class... uzurpator, you might have some distinctive information not otherwise on the Web regarding these engines.

Note that the 'salmon' rods on Cossart valve gear were intended to do this sort of thing.

Probably the 'best' solution is balanced four-cylinder simples, with receiver-balanced (a la Chapelon) four-cylinder compounds closely behind. An interesting geometrical take on this idea is the Withuhn conjugated duplex (employed on the ACE 3000 design, and described in a 1974 issue of Trains Magazine) which has the net thrust couples balanced across the engine bed, but still has an unbalanced lifting component acting on the suspension of individual axles.

[M636C]

Overmod,

There was some spring loading involved in the Cossart rods to reduce the effect of the heavy counterbalanced rods at the valve gear end. Apart from the drive, and the use of vertical piston valves instead of poppet valves, the Cossart gear itself wasn't that different in layout to Caprotti. The two Nord Pacifics weren't well thought of, but the suburban tanks seem to have been a success. Whether reduced rail forces played any part at all in the use of this gear seems doubtful.

I assume the Russian locomotives you refer to were those derived from the steam/diesel experimentals of the 1930s, which terminated in the frightening "Stalinets" with its steam/diesel propulsion and a condensing tender with an inbuilt coal pulverising plant to fuel the compression ignition part of the operation. As they say, what could go wrong?

The OR "October Revolution" prefix applied to a number of early post war locomotives built at Voroshilovgrad, some of which had quite conventional drive. I think the OR23 was the unit with the lever drive from four cylinders (or was it two with opposed pistons)on each side. But the Russians were always hampered by light track, and the elimination of pounding might have been useful if diesels hadn't taken over. The Chinese QJ, with a 200kN axle load, uses a boiler that required a 2-6-6-2 simple articulated to carry on Russian tracks allowing only 180 kN axle load.

Peter

[Overmod]

I didn't intend to conflate the purpose of the salmon rod with reducing the tendency of the locomotive to nose under thrust! What I understood it was intended to do was to balance out some of the 'unbalanceable' component either of the big end of the main rod or of the eccentric crank. That accounts for the differential weight distribution at sections from end to end of the rod. I wasn't aware that there were springs in those rods! Shades of tuned mass dampers... ;-}

The reference to the rocking-lever "OR" locomotive is taken from a note in the Ransome-Wallis Encyclopedia of World Railway Locomotives, which I believe was written not long after the time the engine in question was being tested (circa 1949?). It is NOT one of the terrifying Teploparovoz designs... I have trouble when looking at the OR23 in concluding that it uses the same mechanical 'principle' that Ransome-Wallis was implicitly mentioning. See picture of mechanism, down the page a bit at

http://www.dself.dsl.pipex.com/MUSEUM/LOCOLOCO/russ/russrefr.htm

I still think that one of the great ironies is that Britain, the nation that gave the Russians advanced turbojet engines, didn't give them Sentinel engine technology -- probably the only "right" sort of answer for large numbers of powered axles on very light and uncertain trackage is the use of gear or shaft (or even direct a la Paget) drive from individual steam motors or cylinder blocks.

I suppose we shouldn't forget the 'horse's necks' and other weird expedients tried back in the early days of steam locomotives, either... not all that technology is as silly as it first appears, when the weaknesses of contemporary track are properly considered!

[M636C]

Overmod,

The diagrams in the link are drawn from Le Fleming's book which draws on Rakov as its source. I have both, although my Rakov is a translation into German by Slezak.

I had in my memory over-complicated the OR23, but the basic principle of rocking levers driving cranks seems to be the same, even if you have only one cylinder rather than two driving the lever. I have some recollection that steam was applied to both steam and compression ignition cylinders to start those Steam -Diesel units. I did suggest that the OR23 was "derived from" the Steam Diesel units, not that it was one of them.

I can check, but I think Le Fleming did suggest the Steam-Diesel units as the origin of the lever system for OR23. I think the 23 referred to a 230kN axle load, so since the track wasn't ever upgraded to that level, that unit was doomed regardless of how it may have run.

I may have misled you on the Cossart rods. I don't think the masses themselves were sprung, but that there were springs further on in the drive from the rocking arm to the valve cams to reduce shock loading.

The B&O W-1 4-8-4 (or 4-2-2-2-2-4) was to have multicylinder drive. It was announced about the time EMC delivered the first EA+EB units to B&O, so we don't have to guess why it didn't go ahead. The Emerson boiler was used on a President class pacific which became class P-9b. The SNCF built 232P1 with that sort of multicylinder drive and a very high pressure boiler, and it wasn't just WWII that shut that down.

Steam locomotives worked best if they were simple, in most meanings of the word.

Peter