Trains.com

Low axle loadings on the 4-12-2 Union Pacific-type locomotives (9000 series)

5121 views
31 replies
1 rating 2 rating 3 rating 4 rating 5 rating
  • Member since
    July 2004
  • 2,582 posts
Low axle loadings on the 4-12-2 Union Pacific-type locomotives (9000 series)
Posted by Paul Milenkovic on Sunday, January 24, 2021 7:51 PM

Anyone know why the driver axle loadings on the 4-12-2's were substantially lower (below 60,000 lbs/axle) than on the 4-6-6-4 Challengers?

I wouldn't think with the long rigid wheelbase (yes, with lateral motion devices, but still) would be amendable to anything else but main lines with gentle curves?  Was the UP using lighter rail on their network into the 1920's when the 9000-series locomotives were introduced?

If GM "killed the electric car", what am I doing standing next to an EV-1, a half a block from the WSOR tracks?
  • Member since
    September 2003
  • 16,282 posts
Posted by Overmod on Sunday, January 24, 2021 11:02 PM

I believe Juniatha had a particular interest in the Nines and would be an ideal source for this kind of information on them.

  • Member since
    January 2019
  • From: Henrico, VA
  • 6,475 posts
Posted by Flintlock76 on Monday, January 25, 2021 5:42 PM

I went looking through the archives here at the "Fortress Flintlock" and all I could find out about the 60,000 pound axle loading was in "Alfred the Bruce's"  Wink  book "The Steam Locomotive In America."  Apparantly the 60,000 pound axle load (and not above that figure) was specified by the Union Pacific, but Bruce doesn't say why. 

Another book I've got on American steam development doesn't say why either.

Maybe it was  due to the track it was going to run on.  Those 4-12-2's were smooth-running locomotives though, however due to that long rigid wheelbase they were restricted to mainline runs and trackage with very gentle curves. 

I'll keep looking.

I did.  Can't find nuthin' else.  Sorry.

  • Member since
    May 2019
  • 1,768 posts
Posted by MMLDelete on Tuesday, January 26, 2021 11:17 PM

I had not heard of this wheel arrangement before, so I looked for info on it. I found it was a 3-cyl. engine. Is the 3rd cylinder in the middle? Are the valve-gear and rods exactly the same as those for the other two cylinders?

  • Member since
    January 2002
  • 4,257 posts
Posted by M636C on Wednesday, January 27, 2021 2:56 AM

Lithonia Operator

I had not heard of this wheel arrangement before, so I looked for info on it. I found it was a 3-cyl. engine. Is the 3rd cylinder in the middle? Are the valve-gear and rods exactly the same as those for the other two cylinders?

 

This is explained in the Wikipedia entry:

4-12-2 - Wikipedia

Note that the inside cylinder has a stroke one inch less than the outside cylinders.

The outside cylinders drive the third axle while the inside cylinder drives the second axle.

The Gresley gear, in fact developed by Harold Holcroft, had an inherent defect since the inside valve was driven by the mechanism attached to the outside piston valve rods. (I prefer the term pivoted to the description "hinged" used in the Wikipedia entry). But as the outside piston valves and rods expanded due to the heat of the steam, moving relatively forward, the inside valve, fixed at the front. moved relatively backward. This resulted in more work being done by the inside cylinder, which increased the load on the already vulnerable inside big end.

So a number of these had a third set of Walschaerts valve gear fitted on the right side (like Baldwin 60000 and other Baldwin three cylinder locomotives).

I suspect that the original intention was to limit the axle load to reduce the lateral forces on the rail in curves, since the lateral force on the rail is related to the axle load, and high forces might be expected with such a long rigid locomotive, even with lateral motion devices on the end axles.

Peter

  • Member since
    September 2003
  • 16,282 posts
Posted by Overmod on Wednesday, January 27, 2021 9:10 AM

M636C
The Gresley gear, in fact developed by Harold Holcroft...

If I recall correctly, Holcroft's gear was originally developed for four-cylinder engines and only incidentally could be extended to three-cylinder 'line-abreast' engines with double-pivoted asymmetrical levers; personally, I think Gresley did enough work on the two-lever system and its promotion to qualify the particular variant as 'his'.  Note that the Australian versions that (somewhat imperfectly) tried to implement the ratio with 'rotating' rockshafts rather than inertially-challenged levers are not generally referred to as 'Gresley' when perhaps he (and by extension Holcroft) should be credited for the geometry and principle.  (P.S. -- all the Australian designs needed, really, was to use a large-diameter hollow or 'pipe-based' shaft to improve the shaft torsional resistance and 'rate'.)

... had an inherent defect since the inside valve was driven by the mechanism attached to the outside piston valve rods. (I prefer the term pivoted to the description "hinged" used in the Wikipedia entry). But as the outside piston valves and rods expanded due to the heat of the steam, moving relatively forward, the inside valve, fixed at the front. moved relatively backward. This resulted in more work being done by the inside cylinder, which increased the load on the already vulnerable inside big end.

I'm not sure why on a proper DA 2-cylinder this effect would occur.  All that changes with this is the relative timing, and since both ends of the valve conduct both admission and exhaust, it is difficult to see how the 'net' result of the differential expansion would have an effect other than transient (e.g. in torsional couples or in rod stresses between crank and outside-driven axles) in 'work done'...

The explanation I've always heard is that the increase in 'work done by the inside cylinder' was the result of 'lost motion' due to some combination of wear and lever whip, which in the Gresley design add cumulatively to cause overtravel of the center valve relative to the other two.  This is obviously a greater cause of cumulative change in effective cutoff.  Were the Gresley gear changed to be timing-only (with the valves driven by some external means like those on Corliss engines) this effect could be at least partly 'compensated for'.

The case often comes up 'why weren't rolling-element or needle bearings used on these critical pivots' as if Gresley were so ignorant he had overlooked the essential point of preventing or reducing the effect of lost motion in service.  Sometimes an excuse like 'wartime maintenance' or poor lube quality is brought up either to explain away or accuse.  In actual fact, Gresley paid special attention to hardening the bearings and using some forms of rotating-element bearings -- but like Bulleid with the chains, he didn't quite understand the mechanical magic involved.  A rotating-element bearing depends in part for its integrity on continued smooth rotation; if it is used in partial-arc service -- as, specifically, with the various pivots in Gresley levers -- both wearing and fretting 'corrosion' can increase, at which point the clearances will suck just as they would with plain bearings, and just as cumulatively.  As I recall, part of this could be addressed with Multirol bearings (which aren't really 'rotating-element' bearings as they have no cages, but can be thought of as 'renewable hard-surface bearings' with integral lube space in the 'packing') and a perhaps better solution would be to use mutual-taper Multirols periodically tightened with something like a microcastellated nut to take up the 'end-shake' and thereby restore precise lever geometry as required.

I have always found it interesting that, while everyone loves to trot out the inside-main failures on Mallard and others, not one in a hundred can tell you the detail improvements made on the inside-main construction after the Mallard failure.  Which extended the practical life of the A3s and A4s for decades, in service likely well over 100mph peak, without big-end overheating or overstressing being a major cited disaster thereafter.  

At least presumptively this would have given Union Pacific access to some of the analysis and solutions involved in England, some of which might have been useful in improving the center big end on the Nines if that were an issue.  It would be unfortunate if some form of NIH or 'their stuff won't work here' kept UP from improving that set of components -- perhaps by that time it was assumed the Nines would be supplanted by Challengers and the 'three sets of valve gear' would expensively solve the 'issue' at its source.

I suspect that the original intention was to limit the axle load to reduce the lateral forces on the rail in curves, since the lateral force on the rail is related to the axle load, and high forces might be expected with such a long rigid locomotive, even with lateral motion devices on the end axles.

But this could easily be addressed, as it was in so many other classes of engine, by using more aggressive tapered loading.  

One thing about the Nines was the flange gauge reduction on some of the axles not fitted with controlled lateral motion (which I think was still fairly crude in the mid-Twenties).  I think that as with Stroudley's 0-4-4s there was some careful design involved in progressive steering of the long rigid wheelbase, and that at least with regard to the lead driver pair UP could have installed better controlled-lateral-motion devices -- as on many of the more-recent Alco locomotives, even if roller bearings were not installed in the cannon boxes involved.

My understanding -- quite possibly defective, as I'm not a specialist on the Nines -- was that perceptions of the civil department on perceived peak augment of these very long and powerful engines did not fully account for the supposed smoothness increase of the three-cylinder 'divided drive'.  Whether or not this might have been based on experience with the preceding three-cylinder 'Overland' 4-10-2s I don't know, and I don't think I trust Kratville on the issue (in part since it appeared he thought the 4-12-2s were compounds).  

  • Member since
    July 2004
  • 2,582 posts
Posted by Paul Milenkovic on Wednesday, January 27, 2021 2:52 PM

Holcroft gear for 4-cylinder locomotives?

The usual 4-cylinder setup has the cylinders on each side working in opposition - much like the "boxer" engine on the rear-engine Porsche and VW Beetle or the front-engine 4wd Subaru.  This allows using one valve gear per side, where the inside-cylinder valves are actuated by rocking shafts as on Churchward's Castle class?  I thought the Holcroft gear was "a thing" with 3-cylinder locos as this way for pairs of cylinders to share a valve gear covers the 4-cylinder case, that is, if you don't mind the 2-cylinder uneven torque curve but want the reduced track pounding.

The British had a Lord Nelson class with one valve gear for each of 4 cylinders where the cylinder strokes were spaced out to give a smoother torque curve than either the "boxer" type 4-cylinder or the 3-cylinder layout.  Even though it lacks pistons that balance by moving in opposition, it was said to be so smooth a locomotive that it was hard to fire -- the coal shoveled didn't spread across the grate, and coal didn't feed itself from the tender from the usual shaking on a steam locomotive.  

If GM "killed the electric car", what am I doing standing next to an EV-1, a half a block from the WSOR tracks?
  • Member since
    May 2019
  • 1,768 posts
Posted by MMLDelete on Wednesday, January 27, 2021 8:08 PM

Was the inner cylinder at the same height from the ground as the side ones? I've seen one photo which seems to show that the middle cylinder was up significantly higher, and perhaps angled downward. ?? So I'm wondering if that's true. If so, was that to make it easier to have the main rod connect to the the second axle without encroaching on the first? I assume that a connecting rod linked the second axle with the first, right?

  • Member since
    September 2003
  • 16,282 posts
Posted by Overmod on Thursday, January 28, 2021 12:00 AM

Lithonia Operator
Was the inner cylinder at the same height from the ground as the side ones?

No -- but the valve cylinders were (because the Gresley lever arrangement required this) and consequently the location and inclination of the inside cylinder is 'relative' to the location of this third valve bore, in order to have comparably short port and passage characteristics with the outside cylinders given the slightly shorter stroke.  The third cylinder is to the side of its valve, whereas the two outside cylinders are below, and this will give you a measure both of the elevation and, pointing back at the centerline of the cranked second axle, the angle of inclination.

It was not unusual in these 'line-abreast' cylinder arrangements to find the first driver axle 'cranked' (sometimes by forging it with an intentional 'bend' in the middle relative to quarter) so that the inside rod would clear the axle as it rotated.  I believe the Nines handle this without explicitly cranking the first driver axle, by adding to an already long rigid wheelbase by inserting space between the first and second driver axles.  The drive is divided: the two outside cylinders have elongated piston rods and crosshead guides, and even so, and with 67" drivers, the outside mains bear on the third driver axle. 

I assume that a connecting rod linked the second axle with the first, right?

You can see the outside connecting rods (which are in the same quarter as the outside mains and other rods) in pictures.  They are longer than the other side rods, and you can gauge their mass by looking at the size of the counterweights in the first driver pair.

  • Member since
    April 2001
  • From: Roanoke, VA
  • 1,892 posts
Posted by BigJim on Thursday, January 28, 2021 3:41 PM

Overmod
I don't know, and I don't think I trust Kratville on the issue (in part since it appeared he thought the 4-12-2s were compounds).  

It has been a while since I have read Kratville's books on the 4-12-2s, but, I don't remember him ever mentioning anything about them being "compound". Can you cite where this was mentioned?

.

  • Member since
    September 2003
  • 16,282 posts
Posted by Overmod on Thursday, January 28, 2021 6:37 PM

BigJim
I don't remember him ever mentioning anything about them being "compound". Can you cite where this was mentioned?

I don't have any of the relevant Kratville books on the Nines; I'm going by a couple of posts, where a couple of (as I recall, reliable) people mentioned the specific references.  

As I recall they were relatively early works, from the era when booboos like this might well enter the 'enthusiast press' as easily as Arnold Haas saying both Hudsons and Niagaras regularly exceeded 120mph.  

  • Member since
    May 2019
  • 1,768 posts
Posted by MMLDelete on Friday, January 29, 2021 12:34 PM

In this photo, you can clearly see the third cylinder.

https://www.facebook.com/photo?fbid=10225035216217807&set=gm.788472115088734

This is the only pic I've come across in which it's visible. Was there generally a panel there that concealed the cylinder?

  • Member since
    January 2019
  • From: Henrico, VA
  • 6,475 posts
Posted by Flintlock76 on Friday, January 29, 2021 12:52 PM

That's the third cylinder all right, and just what they looked like, no cover. 

Must be right out of the shop, the whole engine's clean as a whistle!

  • Member since
    September 2003
  • 16,282 posts
Posted by Overmod on Friday, January 29, 2021 2:41 PM

Flintlock76
That's the third cylinder all right, and just what they looked like, no cover. 

Well, there is a cover, a shiny passenger-engine cylinder-head cover to match the ones on the outside.

Locomotive cylinders do not have shiny heads only held on with one stud in the middle, you know.  Wink

Note there is even a shiny 'ring' cover on the inside valve cylinder, as on the outside.  If you need further proof this was considered passenger-grade in its day... 

There might have been something over the Gresley levers to protect them from road damage at some point, but I don't know of any.  You will note that if the cylinder has more normal black lagging over the head it would be more difficult to see in the shadow under the smokebox and behind the levers.

What I'd like to see is a clear low ¾ shot of one of the engines modified with the three sets of valve gear, to see how this changed the view into the front deck...

  • Member since
    May 2019
  • 1,768 posts
Posted by MMLDelete on Friday, January 29, 2021 3:09 PM

In the other shots I saw, the bright sun and high-contrast light caused deep shadow in that area. I couldn't see any detail; the area was just dark. I thought maybe there was a black panel to keep grit and debris out.

  • Member since
    September 2003
  • 16,282 posts
Posted by Overmod on Friday, January 29, 2021 3:33 PM

I just looked at a bunch of Nine pictures, and even on a phone you could make out the angled lagged cylinder end fairly readily.  It helps to know you're looking specifically for a large inclined circle with a dot in the middle.

The center support for the lever has three hoses or lines running over it, and this blocks the view 'back' to the cylinder head behind.

  • Member since
    January 2019
  • From: Henrico, VA
  • 6,475 posts
Posted by Flintlock76 on Friday, January 29, 2021 4:01 PM

Overmod
Well, there is a cover, a shiny passenger-engine cylinder-head cover to match the ones on the outside. Locomotive cylinders do not have shiny heads only held on with one stud in the middle, you know.  

Well jeez, if you wanna split hairs!  L-O wanted an answer, not a thesis!

It's cool, love 'ya bro!  Smile, Wink & Grin

  • Member since
    May 2003
  • From: US
  • 20,523 posts
Posted by BaltACD on Friday, January 29, 2021 5:46 PM

Lithonia Operator
In this photo, you can clearly see the third cylinder.

https://www.facebook.com/photo?fbid=10225035216217807&set=gm.788472115088734

This is the only pic I've come across in which it's visible. Was there generally a panel there that concealed the cylinder?

I see the air pumps hung on the front of the smokebox ... what is the machienry that interrupts the walkway on the side of the locomotive.

The shiney covers are exactly that - covers for the cylinder head, the cylinder heads themselves would be held down by multiple bolts to withstand the pressures developed during operation.

Remember, auto engines have 'valve covers' that are attached to the cylinder heads and are used to contain the lubrication that circulates through the valve train.

  • Member since
    January 2019
  • From: Henrico, VA
  • 6,475 posts
Posted by Flintlock76 on Friday, January 29, 2021 6:13 PM

BaltACD
I see the air pumps hung on the front of the smokebox ... what is the machienry that interrupts the walkway on the side of the locomotive.

I believe that's the feedwater heater.  A Worthington BL type if I'm not mistaken.

  • Member since
    January 2002
  • 4,257 posts
Posted by M636C on Sunday, January 31, 2021 6:52 AM

Overmod

 

 
M636C
The Gresley gear, in fact developed by Harold Holcroft...

 

If I recall correctly, Holcroft's gear was originally developed for four-cylinder engines and only incidentally could be extended to three-cylinder 'line-abreast' engines with double-pivoted asymmetrical levers; personally, I think Gresley did enough work on the two-lever system and its promotion to qualify the particular variant as 'his'.  Note that the Australian versions that (somewhat imperfectly) tried to implement the ratio with 'rotating' rockshafts rather than inertially-challenged levers are not generally referred to as 'Gresley' when perhaps he (and by extension Holcroft) should be credited for the geometry and principle.  (P.S. -- all the Australian designs needed, really, was to use a large-diameter hollow or 'pipe-based' shaft to improve the shaft torsional resistance and 'rate'.)

 

 

 
... had an inherent defect since the inside valve was driven by the mechanism attached to the outside piston valve rods. (I prefer the term pivoted to the description "hinged" used in the Wikipedia entry). But as the outside piston valves and rods expanded due to the heat of the steam, moving relatively forward, the inside valve, fixed at the front. moved relatively backward. This resulted in more work being done by the inside cylinder, which increased the load on the already vulnerable inside big end.

 

I'm not sure why on a proper DA 2-cylinder this effect would occur.  All that changes with this is the relative timing, and since both ends of the valve conduct both admission and exhaust, it is difficult to see how the 'net' result of the differential expansion would have an effect other than transient (e.g. in torsional couples or in rod stresses between crank and outside-driven axles) in 'work done'...

 

The explanation I've always heard is that the increase in 'work done by the inside cylinder' was the result of 'lost motion' due to some combination of wear and lever whip, which in the Gresley design add cumulatively to cause overtravel of the center valve relative to the other two.  This is obviously a greater cause of cumulative change in effective cutoff.  Were the Gresley gear changed to be timing-only (with the valves driven by some external means like those on Corliss engines) this effect could be at least partly 'compensated for'.

The case often comes up 'why weren't rolling-element or needle bearings used on these critical pivots' as if Gresley were so ignorant he had overlooked the essential point of preventing or reducing the effect of lost motion in service.  Sometimes an excuse like 'wartime maintenance' or poor lube quality is brought up either to explain away or accuse.  In actual fact, Gresley paid special attention to hardening the bearings and using some forms of rotating-element bearings -- but like Bulleid with the chains, he didn't quite understand the mechanical magic involved.  A rotating-element bearing depends in part for its integrity on continued smooth rotation; if it is used in partial-arc service -- as, specifically, with the various pivots in Gresley levers -- both wearing and fretting 'corrosion' can increase, at which point the clearances will suck just as they would with plain bearings, and just as cumulatively.  As I recall, part of this could be addressed with Multirol bearings (which aren't really 'rotating-element' bearings as they have no cages, but can be thought of as 'renewable hard-surface bearings' with integral lube space in the 'packing') and a perhaps better solution would be to use mutual-taper Multirols periodically tightened with something like a microcastellated nut to take up the 'end-shake' and thereby restore precise lever geometry as required.

I have always found it interesting that, while everyone loves to trot out the inside-main failures on Mallard and others, not one in a hundred can tell you the detail improvements made on the inside-main construction after the Mallard failure.  Which extended the practical life of the A3s and A4s for decades, in service likely well over 100mph peak, without big-end overheating or overstressing being a major cited disaster thereafter.  

At least presumptively this would have given Union Pacific access to some of the analysis and solutions involved in England, some of which might have been useful in improving the center big end on the Nines if that were an issue.  It would be unfortunate if some form of NIH or 'their stuff won't work here' kept UP from improving that set of components -- perhaps by that time it was assumed the Nines would be supplanted by Challengers and the 'three sets of valve gear' would expensively solve the 'issue' at its source.

 

 
I suspect that the original intention was to limit the axle load to reduce the lateral forces on the rail in curves, since the lateral force on the rail is related to the axle load, and high forces might be expected with such a long rigid locomotive, even with lateral motion devices on the end axles.

 

But this could easily be addressed, as it was in so many other classes of engine, by using more aggressive tapered loading.  

 

One thing about the Nines was the flange gauge reduction on some of the axles not fitted with controlled lateral motion (which I think was still fairly crude in the mid-Twenties).  I think that as with Stroudley's 0-4-4s there was some careful design involved in progressive steering of the long rigid wheelbase, and that at least with regard to the lead driver pair UP could have installed better controlled-lateral-motion devices -- as on many of the more-recent Alco locomotives, even if roller bearings were not installed in the cannon boxes involved.

My understanding -- quite possibly defective, as I'm not a specialist on the Nines -- was that perceptions of the civil department on perceived peak augment of these very long and powerful engines did not fully account for the supposed smoothness increase of the three-cylinder 'divided drive'.  Whether or not this might have been based on experience with the preceding three-cylinder 'Overland' 4-10-2s I don't know, and I don't think I trust Kratville on the issue (in part since it appeared he thought the 4-12-2s were compounds).  

 

I have Holcroft's autobiography Locomotive Adventure in two volumes. I recommend it to anyone. His original patent used two 2:1 levers connected to a central oscillating lever with the central valve driven directly from the centre point of the centre arm.  This required that the three valves had to be parallel at the same level and equally spaced, placing the centre valve above the centre cylinder which would be impractical in Britain with their limited clearances.

He built two 2-6-0s for the SECR (one was a 2-6-4T as built) which used the ame arrangement as Gresley at about the time Gresley built his first locomotive, also a 2-6-0 No 1000 with his version of the gear. Holcroft was careful enough to drive the conjugating gear from  the valve gear directly, using rods outside the piston valve chambers to avoid the expansion problem.

Gresley's 1000 suffered from excessive valve travel on the centre valve such that at speed (despite being a 2-6-0 the operators were impresssed by its speed and power, and put it on heavy passenger trains.) I understand that a combination of dynamic flexing, slack in the pin joints and the effect of valve stem expansion knocked the inner piston valve cover off. Holcroft didn't experience that problem, but felt that slack in the pin joints reduced the effectiveness of the gear (and because of the direct drive, Holcroft had a couple more pin joints). So the two prototypes were converted to three separate Walschearts gears and further locomotives were built with three valve gears.

SECR N1 class - Wikipedia

SR U1 class - Wikipedia

I understand Gresley used "needle roller bearings" on the main two to one arm pivot, which were expected to work better than conventional roller bearings in the essentially oscillating application, to prevent "brinelling" of the races.

Overmod referred to Australian conjugated valve gear...

There were 43 three cylinder locomotives in Australia.

29 used Gresley/Holcroft Gear. One used a 1915 vintage Henschel conjugated gear. The remaining 13 used a rack and pinion version of Gresley/Holcroft gear.

I have photos of this gear and it used about a six inch diameter tube for the main cross shaft. The Henschel gear used about a ten inch diameter hollow shaft for the main cross shaft. The locomotive is fortunately preserved, so I was able to check it for myself.

I have a good photocopy of an internal report on the testing of NSW locomotive 5711 in 1929, including many of the indicator diagrams. Sadly, the indicator on the centre cylinder didn't work as well as those on the outside cylinders.On many occasions they just assumed that the centre would be the same as the outside in the absence of a diagram.

But in one test at the highest speed reached, 45 mph, the centre indicator worked and gave a very strange diagram suggesting that the centre cylinder was producing 50% more power than each outside cylinder, which has been discussed as a problem for Gresley A4 class. Thompson lined up the centre cylinder on the A4 from 18" to 17" to cope with this problem. 

Peter

  • Member since
    February 2021
  • From: Germany
  • 142 posts
Posted by Sara T on Thursday, February 25, 2021 1:43 AM

Wasn't the 9000 an early type of locomotive, I believe constructed in 1925 or 6 and following years? The Challenger was built later. The axle load could have been limited and that was the reason Union Pacific went to six coupled axles?

In Russia there was an even larger locomotive being constructed with seven driven axles, the AnAn20 type. It seems then they got afraid of their own boldness and left the type with one built. They didn't even preserve the locomotive, they just let it slip away in silence, I don't know if the designers were later arrested by the KGB or if they were allowed a peaceful pension.

In Austria Karl Gölsdorf had one mountain engine with six coupled axles, the 100.01. Again the Austrians later had it scapped unseen in a quiet place but the steam friends still know they had once been reaching for the stars.

Today they don't even have six driven axles under the new electrics but only the eternal BoBo type for everything.

Only Union Pacific wanted a high number of driven axles under one diesel loco, I'm sure one of you will know what it was, I don't have any details.

But the 9000 type must have been quite a sight and a show too!

Sara  05003

  • Member since
    December 2017
  • From: I've been everywhere, man
  • 3,358 posts
Posted by SD70Dude on Thursday, February 25, 2021 6:35 PM

Sara T

In Russia there was an even larger locomotive being constructed with seven driven axles, the AnAn20 type. It seems then they got afraid of their own boldness and left the type with one built. They didn't even preserve the locomotive, they just let it slip away in silence, I don't know if the designers were later arrested by the KGB or if they were allowed a peaceful pension.

Didn't it have trouble negotiating switches and yard trackage with that long rigid wheelbase?  If that were the case I could understand why it was not duplicated.

Did the Russian railways ever use articulated steam locomotives?

Also, welcome to the forum!

Welcome

Greetings from Alberta

-an Articulate Malcontent

  • Member since
    September 2003
  • 16,282 posts
Posted by Overmod on Thursday, February 25, 2021 7:04 PM

The Nine was an evolution of the three-cylinder 4-10-2 Overland; it was not 'one driver too far', in fact you will note that the rigid wheelbase is actually 'longer than it needs to be' to make a little more room for the throw of the inside crank, and there was a remarkably sharp curve (as mainline curves go) outside Lawrence, Kansas that they were specially designed to run through.  Juniatha has a special interest in these locomotives and is the ideal one to PM for specific detail.

It has been a while since I looked at the AA20 (named for one of Stalin's master murderers; I don't remember seeing that the designers wound up in wrecker trials or sharashkas, so that may have worked); what I recall is that it was only intended to work a relatively straight line with mineral trains assembled behind it; it had particularly light axle loading for an engine so large and was not expected to have extraordinary lateral load (e.g. to overturn rail) in the intended service.  The locomotive is covered in a book on modern Soviet steam power which I cannot access due to a hard-drive crash.  I believe Peter Clark has written about this design in some detail.

Improvements in Mallet-style chassis, starting with proper simple articulateds and then incorporating knowledge from Baldwin, then N&W, then Alco on how to stabilize the forward engine for high speed are what revolutionized engines with more than 5 driver pairs.  Alfred Bruce of Alco was proud of the arrangement on the Challengers which essentially turned them into near-equivalents of a divided-drive 4-12-4 except for a tightly-controlled lateral hinge.

Remember that the highest horsepower recorded for an American locomotive of any number of drivers was a ten-coupled -- the divided-drive PRR Q2.  Which had a shorter rigid wheelbase than an ATSF 5001 or 5011 class 2-10-4...

IIRC the Russians had the world's largest Garratts.  Elsewhere in the world, that was one definitive answer to running relatively large locomotives on light and often indifferently-kept track.

  • Member since
    July 2008
  • 754 posts
Posted by Juniatha on Thursday, February 25, 2021 7:25 PM

 

Hi SD70

First, yes, it did

and

second, no, they didn't.

The 4-14-4 did have trouble in tight curves - but only because of two reasons:

a) the track was arguably too light and not well anchored in ballast - any engine that really meant it could have shifted those curves.

b) the arrangements of the lateral motion devices were less than ideal to keep it mildly. 

In an early stage of design they had approached German technicians about this curve negotiating problem and Eckhardt had proposed his three-axle design of a leading 'bogie' - that is: leading idler combined by a double-arm lever with the leading coupled axle and that one with the second coupled axle by a second double-arm lever. This way lateral forces would have been divided ~ equally, or exactly as desired between the three. Back in Russia, they rubbed their eyes, cleared their noses, and then made a decision to dump that complex German arrangement and just put a proper bogie in front. On the rear two carrier axles were needed for the firebox, so all in all one had 'progressed' to a well wheel equipped 4-14-4.

The coupled axles were not well arranged as concern lateral play, they didn't even copy the trick in the UP-9 4-12-2 with the hind coupler also being laterally movable against centering force as was the front coupled axle. 

Now, the front bogie could still have been counter-lever connected to the front coupled or the two front coupled axles - but there was non-such German gimmick.

So, the awkward locomotive did what it had to do: it bent its backbone in curves - and that wasn't really 'Russian' - so: bye-bye ill-cast giant.

I come back on my old note: to use big locomotives you first need to get your track in shape!

Side remark:

... and the cylinder volume also was sort of fantastic, not the sort to get steam in and out easily. She (he?) performed some 3500 ihp on test runs - modest in view of another ill curve carver: the DR 2-10-2 45 class which had 2750 ihp originally and 3200 ihp reboilered. The Germans made a concise construction of it, not using too many axles only to bend curves, it was enough to limit lateral motion to a not enough displacement. 

Still, on DB the engines were first reboilered, then management lost interest, and finally, the depot saved lubricant on the firebox gliding surfaces which served perfectly to crack the inside cylinder of each and every engine until one - 45 010 - got a cast steel inside cylinder. To redesign the rather freezing than gliding stays to ensure free boiler movement cold / hot was beyond the amount of thinking assigned to be investigated in that engine class. Finally, they found a 44 class could - more or less - also do what the 45s were used for: as brake locomotive for the Minden testing station. So, when the 45s had all been used up with broken inside cylinders, they got a couple of 44s. It was all coming to an end anyways - why waste any thought ..?

If that doesn't sound typically German you must take into account that DB was a fully state-owned railway and this made itself known now and then.

It makes itself known incomparably more though, with DBAG officially briskly on its way to 'private enterprise' - a process which as a result of long and thorough after-thought has now been silently dropped ...

Much could have been done - little was done.

 

Juniatha

  • Member since
    December 2017
  • From: I've been everywhere, man
  • 3,358 posts
Posted by SD70Dude on Friday, February 26, 2021 5:11 PM

Fascinating!

Your line about track conditions made me chuckle, that simple fact has been repeatedly learned and forgotten over here as well, only to be learned again the hard way. 

Greetings from Alberta

-an Articulate Malcontent

  • Member since
    January 2019
  • From: Henrico, VA
  • 6,475 posts
Posted by Flintlock76 on Friday, February 26, 2021 6:31 PM

SD70Dude

Fascinating!

Your line about track conditions made me chuckle, that simple fact has been repeatedly learned and forgotten over here as well, only to be learned again the hard way. 

 

Reminds me of a story:

A roadmaster had gotten himself a cab ride on a big ol' steam engine.  While they were barreling down the 'road the engineer turned to him and said:

"You know, it takes a hell of a lot of nerve to sit up here and run one of these things!"

"Really?" said the roadmaster.  "Takes a lot more nerve for ME to sit up here and ride with you!"

"How so?"

" know what you're runnin' on!"

  • Member since
    February 2021
  • From: Germany
  • 142 posts
Posted by Sara T on Monday, March 1, 2021 9:43 AM

Paul,

Smooth running Lord Nelson - I'm sorry but that sound like the usual British self-applauding. We all know that without the British we would still be in the stone age. I doubt that a locomotive without the counter balancing of drives could be smoother than a locomotive with it. 

I may not be your all-knowing technician but that sound logic, doesn't it?

Valve gear: I am for the set-up with one valve gear for each one cylinder and drive. How otherwise could you make sure all cylinders work the same power? And I think that should be important to have that torque go round smoothly and without hops and dents, if I may say so.

The 05 class locomotives had this type of valve gear, drive of the middle was to the first coupled axle because in this the middle cylinder could be left horizontal like the outer ones which drove to the middle coupled axle of course. 

0S5A0R0A3

  • Member since
    September 2003
  • 16,282 posts
Posted by Overmod on Monday, March 1, 2021 10:32 AM

Sara T
I doubt that a locomotive without the counterbalancing of drives could be smoother than a locomotive with it. 

Part of the concern is that one aspect of reciprocating-locomotive balancing interferes with another.  If you look at the kinematics of the main rod, the back end is revolving with the wheel, but the front end is accelerating and decelerating in a straight line.  You can't balance these effectively with wheel counterbalance at the same time.  Ralph Johnson describes one method of 'best compromise', which is to measure the center of percussion of the rod and balance the big end as revolving and the small end as reciprocating.  Very late American thinking, since large locomotives have a relatively long rigid wheelbase and controlled lateral and have a relatively large polar moment of inertia at both ends, and could use firmer lateral compliance and better damping in leading and trailing trucks, was to restrict or even eliminate reciprocating overbalance entirely (combined with different compression management) which can nearly eliminate vertical augment force; the Australians actually implemented zero overbalance on at least one class of small locomotives on light rail.  (Perfidious-Albion Riddles, or one of his minions, had a highly interesting alternative, applied to the 9F 2-10-0s...)

The Withuhn conjugated duplex (as featured in the stillborn ACE3000 locomotive) deals with overbalance in what is basically a couple of quartered 2-cylinder engines by fixing the reciprocating moments 'equal and opposite' (which on the ACE3000 was complicated by the compounding arrangement proposed).  If you use Deem-style conjugating instead, you can (in theory at least) get the benefit of zero overbalance by locking the two engines in antiphase with a detent when the Ferguson-clutch accommodation is not needed -- but (and here is where the Lord Nelson example comes in) you can get whatever advantage in high-speed adhesion from torque-peak reduction at very short cutoff but high admission mass flow the 135-degree phasing of the sets of quartered cylinders offers by setting the detent accordingly.  Between these two you're going to get about the smoothest 'torque with minimal hops and dents' that a four-outside-cylinder engine with roller bearings on lightweight rods will produce.  To do better I think you'd need to go to a motor locomotive, like Roosen's with external 90-degree V motors or Besler's with geared motors, but physically conjugating these would be highly difficult, the arrangements at the time to handle wheelslip did not appear fully successful.  (I personally think an analogue of traction control with fast-acting independent on driver rims or cheek plates is a workable answer, but it's efficacy in addressing high-speed slipping is yet to be simulated still...)

If you have not looked at the possibility of Cossart valve gear and valves for a high-speed design, there are some interesting possibilities.  It has been difficult for me to find sources here that adequately explain the development and sequential teething troubles of that system, but we see progressive (and seemingly radical!) lightening of the 'salmon rods' in practice...

  • Member since
    January 2019
  • From: Henrico, VA
  • 6,475 posts
Posted by Flintlock76 on Monday, March 1, 2021 10:57 AM

Getting back to Pauls original question, has anyone  figured out why the UP wanted a 60,000 pound axle loading, and no higher?

still  haven't found out anything.  

Maybe it's one of those mysteries of the universe we're not supposed  to know?  Hmm

  • Member since
    April 2001
  • From: US
  • 81 posts
Posted by sgriggs on Monday, March 1, 2021 2:58 PM

All of the UP locomotive types I checked on Steamlocomotive.com up to the 1936 CSA-1 Challengers had axle loadings under 60,000 lbs.  This included the UP 4-12-2, the predecessor 4-10-2 3-cylinders, the 7000 class 4-8-2's, and the 2-8-8-0 'Bull Moose' compound articulateds.   It wasn't until the CSA-1 Challengers in 1936 that allowable axle loadings were increased to 66,000 lbs or so.  The 800 class 4-8-4's, debuting one year later, also had axle loadings around 67,000 lbs.

 

Without doing a small research project involving skimming many books on UP motive power history, I would say that the railroad upgraded their physical plant at some point in the early- to mid-1930s.  Rail or bridge ratings must have been a limiting factor up to this point.

Join our Community!

Our community is FREE to join. To participate you must either login or register for an account.

Search the Community

Newsletter Sign-Up

By signing up you may also receive occasional reader surveys and special offers from Trains magazine.Please view our privacy policy