MAIN ROD TO DRIVER CONNECTION

2585 views
27 replies
1 rating 2 rating 3 rating 4 rating 5 rating
  • Member since
    May, 2015
  • 21 posts
MAIN ROD TO DRIVER CONNECTION
Posted by Loco2124 on Thursday, November 16, 2017 4:24 PM

Was curious as to how the connection of the main rod to the driver wheel was determined. Some locos use the second driver and some the third. 

  • Member since
    September, 2017
  • 48 posts
Posted by msrrkevin on Friday, November 17, 2017 10:34 AM

Interesting question.  I know the D&RGW had some narrow gauge 2-8-0s of each type.  I remember reading once that those that connected to the 3rd driver were more stable and could be run faster.  But I would be curious to hear the advantages to having a shorter main rod.

- Kevin

Check out my shapeways creations! HOn3 and railroad items for 3D printing:

https://www.shapeways.com/shops/kevin-s-model-train-detail-parts

  • Member since
    May, 2015
  • 21 posts
Posted by Loco2124 on Friday, November 17, 2017 11:12 AM


The only thing that comes to mind is the weight of the rod which would contribute to larger counter weights and more track wear. Am supposing it could be similar to long vs. short throw pistons in an internal combustion engine but can't remember what the advantage/disadvantage was. Maybe max RPM?

  • Member since
    February, 2005
  • 1,708 posts
Posted by timz on Friday, November 17, 2017 4:53 PM

One disadvantage of the steam locomotive is its uneven tractive effort. During each turn of the drivers, at low speed and maximum tractive effort, that tractive effort peaks when the left crankpin is at its 45-degree-upper-forward position and the right crankpin is 45-degrees-lower-forward. TE peaks again (a slightly lower peak) when the crankpins are lower-rearward and upper-rearward.

For a given stroke, the shorter the main rod, the higher the peak in the torque diagram is compared to the low point. I assume that's why the main rod on 0-6-0s usually connected to the rear driver rather than the middle; I assume that's why 2-8-2s and 2-8-4s (and a few 4-8-2s) drove #3 instead of #2.

  • Member since
    December, 2005
  • From: Cardiff, CA
  • 2,818 posts
Posted by erikem on Friday, November 17, 2017 10:03 PM

Loco2124

Am supposing it could be similar to long vs. short throw pistons in an internal combustion engine but can't remember what the advantage/disadvantage was.

Shorter connecting rods mean less mass, but more of an offset to the center motion of the piston versus the crankpin. Longer connecting rods have more reciprocating mass but the center of piston motion occurs closer to where the crankpin is directly above or below the center of the driver axle. Longer connecting rods also translate to lower side thrust at the cross heads.

  • Member since
    September, 2003
  • 5,628 posts
Posted by Overmod on Saturday, November 18, 2017 10:01 AM

erikem

... Longer connecting rods also translate to lower side thrust at the cross heads.

I’m waiting to get to a computer to post most of the details I want to, but this is going to induce some confusion as stated (I think bacause the terminology comes from a vertical-engine model, like using TDC instead of FDC or R/BDC).  That would be a matter of semantics EXCEPT that a key issue with lightweight rods is lateral buckling, which everyone interprets natively as ‘sideways’ on locomotives.

What erikem is describing is the VERTICAL reaction forces at the crosshead, relative to the crosshead guide(s) — similar to the forces that caused the practical demise of the original form of Vauclain compound.  And those are as he says a principal reason to get the rod angularity right.

Goes hand in hand with absolute reduction in the portion of the rod that is in reciprocating rather than revolving motion, and the (relative) reduction of main-rod circle relative to side-rod circle as seen in some English practice and on the T1 duplex, achieved by grinding the main journal eccentric to the rest of the pin.

More on this later.  But a locomotive with a 4-wheel Adams pin-guided lead truck has more available ‘room’ from rear cylinder head to back of the trailing truck wheel and inherently will be built with higher drivers, so mains on second pair.  A Berk even with 70” drivers needs drive on the 3rd pair to get the angularity right...

  • Member since
    December, 2007
  • From: Georgia USA SW of Atlanta
  • 8,569 posts
Posted by blue streak 1 on Saturday, November 18, 2017 12:29 PM

Have no idea if this is revelant.  Was in New Orleans a few years ago on a river boad steamer.  The drive rod was at least 20 feet long to the paddlewheel.  Of course the rotation speed was very slow maybe 10 RPMs ?  Others can interpert the significant.

  • Member since
    May, 2003
  • From: US
  • 13,926 posts
Posted by BaltACD on Saturday, November 18, 2017 4:38 PM

blue streak 1
Have no idea if this is revelant.  Was in New Orleans a few years ago on a river boad steamer.  The drive rod was at least 20 feet long to the paddlewheel.  Of course the rotation speed was very slow maybe 10 RPMs ?  Others can interpert the significant.

Notice that the stern paddle wheel of the Mini Ha Ha continues to rotate through the 'concert'

         

Never too old to have a happy childhood!

  • Member since
    September, 2003
  • 5,628 posts
Posted by Overmod on Saturday, November 18, 2017 5:28 PM

Remember that one component of rod angularity is crank circle, which is usually related directly to stroke.  A Mississippi steamboat might have a stroke of many feet.

Meanwhile, the Maudslay side-lever engine and some ingenious contemporary designs were intended to work side wheels with ‘minimum footprint’ (no long tunnels in the superstructure as for sternwheeling).  Some of these represent almost an origami-like folding of a rod-drive engine into least space.

At least some of these boats only had one cylinder, and needed a ‘starting bar’ or other assistance if the engine stopped on a dead center and there were no desirable way to ‘roll’ the wheel, say by having the boat moored and letting the current do it.  Under such conditions it would be no surprise to keep the engine turning net of all paddle resistance ... keeps the cylinders as free of condensate as they will get, too.

  • Member since
    December, 2007
  • From: Georgia USA SW of Atlanta
  • 8,569 posts
Posted by blue streak 1 on Saturday, November 18, 2017 5:36 PM

Overmod

At least some of these boats only had one cylinder, and needed a ‘starting bar’ or other assistance if the engine stopped on a dead center and there were no desirable way to ‘roll’ the wheel, say by having the boat moored and letting the current do it.  Under such conditions it would be no surprise to keep the engine turning net of all paddle resistance ... keeps the cylinders as free of condensate as they will get, too.

 
The one we were on had two cylinders but did keep the paddls running although slower.
  • Member since
    September, 2003
  • 5,628 posts
Posted by Overmod on Saturday, November 18, 2017 5:39 PM

Then probably either avoiding condensation in the cylinders or keeping the cylinder and valve lube warm and properly spread

  • Member since
    September, 2003
  • 5,628 posts
Posted by Overmod on Monday, November 20, 2017 9:27 AM

Overmod
I’m waiting to get to a computer to post most of the details I want to ...

Rod angularity is important in part because the greater it is, the more there's a vertical component of the main's reciprocating inertia force and thrust which acts on the suspension.  The only reasonably 'exact' number I have for this is on the N&W J class as Voyce Glaze balanced it, with lightweight rods in their original plane, which is given as about 80lb (I presumed at steam and cutoff conditions corresponding to 100mph with train), that being the amount of overbalance incorporated in the counterweighting of the main driver (with the rest he used being distributed in the coupled wheels).  The situation would be more pronounced with non-Timken rods.  N&W had more than usual experience with the flip side of low rod angularity, having chosen third-axle drive on the K3 4-8-2s designed before either lightweight rods or advanced balancing methods were in common use, and suffering endlessly with the resulting augment until they were able to shuck the dogs to a mark ... ahem, another railroad.  Arguably if these had been built as 2-8-2s of similar size otherwise (with appropriate weight distribution over the axles) the main could have been of proportional size to, say, the T&P 2-10-4s and therefore a later balancing program would have relieved the augment with at least equal success as on those locomotives.

An interesting compromise is found on the MILW A class, which for high-speed stability has its main rod on the leading driver pair.  This makes the engine much longer to get acceptable rod angularity (but there are advantages that in Alco's opinion at least made the arrangement worthwhile, and the 'real' Canadian Jubilees used it as well).  The F7 class has normal drive on the center driver pair, which is the most reasonable stable method for Hudsons.

As noted in the discussion on balancing the British 9F class, there can be significant advantages even when there is a high amount of nominal overbalance or component of inertial force if the balance is arranged so the augment comes on and off equally for both wheels in the pair.  The 9Fs were overbalanced at 40%, not only enormously high but well over the nominal amount so notoriously unsuccessful on the ACL R1s as built -- yet were renowned for smooth riding at speeds over 90mph on comparatively very small drivers.  It is the absence of cross-level augment and its potential for resonance that is the only real explanation for this. 

  • Member since
    May, 2015
  • 21 posts
Posted by Loco2124 on Thursday, November 23, 2017 1:46 PM

erikem
Longer connecting rods also translate to lower side thrust at the cross heads.

 

So considering the side thrust on cross heads... I guess that explains why some locomotives had canted cylinders.

  • Member since
    March, 2013
  • 402 posts
Posted by Dr D on Saturday, November 25, 2017 6:43 PM

Loco2124

You have a good question here and it has considerable engineering implications that might not be obvious a first look.

Sir Isaac Newton gave us the formula for the laws of motion - namely FORCE equals MASS times ACCELERATION.  The key concept behind the heavier an object is the more force and acceleration are effected.  These weights and forces can get massively out of control with speed of movement.  A piston rod and crank at one speed can generate astronomical forces when moved at faster speeds - to the point that the metals they are made of will come apart.

For every stroke of a piston one way it must be almost instantly reversed to move back the other way - the Physics law of inertia - that a body in motion tends to stay in motion and a body at rest remains at rest. 

Heavy steel weights of of connecting rods and pistons moving back and forth instantly reversing generate tremendous forces and are liable to come apart when moved beyond certain design limits.  Engineers are usually able to calculate these forces mathmatically.  Generally smaller and lighter is better unless in doing so makes them inherently weaker.

Steam locomotive design considerations usually considered larger wheels as capable of moving faster because the moving parts moved slower.  This however effected the tractive effort that smaller wheels could generate to pull heavy loads.

The motion of connecting rods is divided in two ways.  Half the rod is rotating and half the rod is reciprocating motion.  Balance of the long or short rod reduced the mass of weight that needed to be started and stopped each stroke.  Generally passenger engines used short rods and freight long rods.  Passenger 4-8-4 would use the second drive wheel.  Freight 2-8-4 would use the third drive wheel but it was the actual length that made the difference.  Some articulated steam engines crowded the design of the chasis so that the rod would connect to a different drive wheel just to fit the cylinders.  For example the Union Pacific 4-8-4 passenger Northerns use the second set drive wheel compared to the Big Boy 4-8-8-4.

Another consideration of long vs short rods is the angularity created in their movement.  The long rod engine will cause the piston to remain at or near the cylinder head for a longer period of time than the same engine with a short rod.  This "dwell" time can be calculated by engineers in a study of "time-angle-area" computation and effects the way in which the steam exits and enters the cylinder and the expansive effect it has upon the piston.

The days of steam locomotive development in America was a time of "industrial arts" and we live today in an age of "industrial science."  By this I mean that most of the engineering of steam engines was by practical testing of operating railroads.  There were not engineering labratories that figured out design problems before the engines were constructed.  They did not test them on test stands for thousands of hours to develop and resolve problems.  No rather one steam engine grew out of the design of previous models and the practical success they had.

- Dr. D  

  • Member since
    August, 2010
  • From: Henrico, VA
  • 8,186 posts
Posted by Firelock76 on Saturday, November 25, 2017 6:48 PM

Yay!  Dr. D's back!  Cool!

  • Member since
    May, 2015
  • 21 posts
Posted by Loco2124 on Saturday, November 25, 2017 8:24 PM
WHOA!
  • Member since
    February, 2005
  • From: Vancouver Island, BC
  • 21,315 posts
Posted by selector on Saturday, November 25, 2017 11:50 PM

Dr D

...

Another consideration of long vs short rods is the angularity created in their movement.  The long rod engine will cause the piston to remain at or near the cylinder head for a longer period of time than the same engine with a short rod.  This "dwell" time can be calculated by engineers in a study of "time-angle-area" computation and effects the way in which the steam exits and enters the cylinder and the expansive effect it has upon the piston.

...

- Dr. D  

 

I don't understand this.  On both variations of rod length, the piston stroke is the same and the driver diameter is the same.  Also, the crank is the same, in both angle and length.  So, regardless of the length of the rod, as any one of the side-rod-linked drivers rotates about is axis one full revolution, so will the main driver with the main crank.  That would mean that piston dwell at either end of the piston's run, still commensurate with those same driver diameters and main crank length, would be exactly the same.

  • Member since
    December, 2005
  • 184 posts
Posted by AnthonyV on Tuesday, November 28, 2017 10:04 AM

selector

 

 
Dr D

...

Another consideration of long vs short rods is the angularity created in their movement.  The long rod engine will cause the piston to remain at or near the cylinder head for a longer period of time than the same engine with a short rod.  This "dwell" time can be calculated by engineers in a study of "time-angle-area" computation and effects the way in which the steam exits and enters the cylinder and the expansive effect it has upon the piston.

...

- Dr. D  

 

 

 

I don't understand this.  On both variations of rod length, the piston stroke is the same and the driver diameter is the same.  Also, the crank is the same, in both angle and length.  So, regardless of the length of the rod, as any one of the side-rod-linked drivers rotates about is axis one full revolution, so will the main driver with the main crank.  That would mean that piston dwell at either end of the piston's run, still commensurate with those same driver diameters and main crank length, would be exactly the same.

 

It's been a while since i have thought about this, but I pretty sure that for a given stroke, a longer rod results in a lower piston speed near TDC and BDC than for a shorter rod.  The result is that the piston spends a longer time at/near either end of the stoke.

I'll leave it up to others to discuss the practical effect this has on steam engine performance.

 

  • Member since
    September, 2014
  • 72 posts
Posted by JC UPTON on Sunday, December 17, 2017 1:51 PM

Am I wrong to think that the length of the rod has nothing to do with the piston travel inside the cylinder

Is not that piston travel the real stroke?

from the Far East of the Sunset Route

(In the shadow of the Huey P Long bridge)

  • Member since
    February, 2005
  • From: Vancouver Island, BC
  • 21,315 posts
Posted by selector on Sunday, December 17, 2017 3:12 PM

Actually, it does.  The length of the main rod must reach to both the extent of the main crankpin's travel rearwards, but also as far as it must go forwards, commensurately with the reach inside the cylinder of the piston.  In turn, the piston is connected to a piston rod which must make the crosshead move a certain distance along a crosshead guide which must be long enough to safely support and guide the main bearing for the main rod as it moves back and forth in consonance with the main crankpin's travel.  

You ask if the piston travel is the real stroke.  Yes, if all you wish to take note of is 'stroke' per se.  But the crank, rearward, also has a nominal distance back and forth fixed on a rotating axis which offers little movement except rising and falling as per the suspension.  If the main driver is forced to rotate, but not by cylinder pressure on either side, the piston will be induced to move only as far as the main crank's circle diameter makes it.

  • Member since
    December, 2005
  • 184 posts
Posted by AnthonyV on Sunday, December 17, 2017 5:15 PM

JC UPTON

Am I wrong to think that the length of the rod has nothing to do with the piston travel inside the cylinder

Is not that piston travel the real stroke?

 

The stroke is twice the crank length.

  • Member since
    September, 2003
  • 5,628 posts
Posted by Overmod on Sunday, December 17, 2017 7:01 PM

AnthonyV
JC UPTON

Am I wrong to think that the length of the rod has nothing to do with the piston travel inside the cylinder

Is not that piston travel the real stroke?

The stroke is twice the crank length.

A bit more precisely, the piston travel is constrained to be essentially the main-pin circle, for mechanical reasons, on a normal reciprocating steam locomotive in which the piston-rod axis intersects the main driver axis.  Naturally there is some longitudinal adjustment on where the piston head is located on the piston rod, but the travel of the active 'face' on a DA piston from FDC to BDC will be equal to the crank circle plus whatever slight inertial or bearing clearance change there might be.  That will be true of the front face and the back face regardless of how long the piston might be (and this accounts in part for why some uniflow cylinders have to be so long).  It is also true for 'angled' cylinders like those on some high-speed British locomotives, again so long as the piston-rod axis intersects the center of the axle.

Note however that the length of the rod is independent of the crank circle: it is determined for side rods by the distance between axle centers, and you will quickly appreciate that the stroke of, say, an ATSF 2900 class with 80" drivers is very different from that of a PRR T1 with 80" drivers, but the length of the side rods is very similar.  Likewise, the length of the main rod is much more a matter of desired angularity fixing the distance from the inside rear cylinder head face to the tangent at the main crankpin circle.

Again I differentiate main and side rod circle, because those can be easily different (as they are on the PRR T1).

  • Member since
    December, 2005
  • 184 posts
Posted by AnthonyV on Monday, December 18, 2017 8:07 AM

Overmod

 

 
AnthonyV
JC UPTON

Am I wrong to think that the length of the rod has nothing to do with the piston travel inside the cylinder

Is not that piston travel the real stroke?

The stroke is twice the crank length.

 

A bit more precisely, the piston travel is constrained to be essentially the main-pin circle, for mechanical reasons, on a normal reciprocating steam locomotive in which the piston-rod axis intersects the main driver axis.

If the crank length (or radius) is defined as the distance from the main driver center to the main pin center, and has a value of say 16", how would you calculate the stroke?

  • Member since
    September, 2003
  • 5,628 posts
Posted by Overmod on Monday, December 18, 2017 8:47 AM

AnthonyV
If the crank length (or radius) is defined as the distance from the main driver center to the main pin center, and has a value of say 16", how would you calculate the stroke?

Just as you did, but with the geometrical model I used to simplify understanding: twice the 'crank length' is the crank circle (measured at the center of the crankpin) and therefore the stroke is 32".  I think it is somewhat easier for people having trouble comprehending the geometry involved  to compare things 1:1, and it is instantly clear looking at a diagram with crank circle what the relationship to stroke will be.

Of course, your original answer did not actually provide Mr. Upton with the information he was looking for; saying the piston travel equals the crank travel is little more than a tautology, and he was looking for the correlation of rod length and stroke.  Which is most emphatically NOT related directly to piston travel or stroke length. except trivially that the permissible crank circle can't be less than that which compromises driver-center strength at the hub (for the PRR T1, 26") or greater than that which compromises driver-center strength at the rim (I have not calculated this directly as it's pointless in practical engine design, but probably somewhere around 64" maximum for an 80" driver with 72" center).

Of course, the rods can't be shorter than the corresponding crank circle diameter (or twice the crank length, or piston travel) in normal configuration.

As I noted, rod length for side rods is determined by the axle spacing, and not by any nontrivial characteristic of either piston travel or crank dimension.  If I understood correctly, that was part of the question actually asked.

 

  • Member since
    December, 2007
  • From: Georgia USA SW of Atlanta
  • 8,569 posts
Posted by blue streak 1 on Monday, December 18, 2017 7:26 PM

Will try to muddy the discussion further.   Give two identical locomotives what are the differences if the crank circle of one is larger than the other ? 

1.  Watching videos of 844 it seems crank circle is fairly small.

2.  would bigger crank circle give more tractive effort at slower speeds ?

3.  Would the longer stroke of larger crank circle remove more energy from the steam ?

4.  Bigger circles mean longer strokes .  What are the loads on the cylinders ?

5.  Would higher steam pressure on the smaller crank circle off set Q. #2 ?

6.  Is Max available speed easier to obtain with a smaller crank circle ?

7.  Any idea what the dynamic loads of the rods are with different circles ?

 

  • Member since
    September, 2003
  • 5,628 posts
Posted by Overmod on Monday, December 18, 2017 11:13 PM

blue streak 1
Will try to muddy the discussion further.   Give two identical locomotives what are the differences if the crank circle of one is larger than the other?

Does not muddy anything at all.  Larger crank circle gives greater displacement per stroke, and also increases peak velocity and accelerations in the rods.  It also has the effect of increasing effective torque for a given effective pressure at some point in the stroke (or for the range of pressure/mean effective pressure for the stroke as a whole)

1.  Watching videos of 844 it seems crank circle is fairly small.

It's actually pretty big at 32".  (The FEFs were some of the earliest explicitly high-speed 4-8-4s, using a combination of relatively small bore and longer stroke to achieve better balancing that a more 'square' engine like an ATSF 3776 or 2900 class would; the FEF-3s are 25" x 32" which is only 1/2" smaller bore than a NYC Niagara, and using higher boiler pressure)  It looks small against the 80" diameter, just as the 29.5" circle on the 3460 class Hudsons looks small against their 84" drivers.

 

2. would bigger crank circle give more tractive effort at slower speeds?

Yes, it would.  Part of that is better leverage; part is longer stroke for a given bore.

3. Would the longer stroke of larger crank circle remove more energy from the steam?

This is a bit more complex than you think.  Longer stroke makes timing and lap somewhat less complicated, but it increases swept area (and the potential for surface condensation, which is not at all the kind of 'energy removal' you want to hear about!) and it may increase the steam mass trapped during compression.

4. Bigger circles mean longer strokes.  What are the loads on the cylinders?

The real 'loads' are not on the cylinders but on the piston, piston rod, crosshead, and rods  Remember that at a given rotational speed the piston assembly is coming to a stop at both FDC and BDC, having to be first accelerated and then decelerated, and longer strokes make for greater accelerations for a given driver diameter.

5. Would higher steam pressure on the smaller crank circle offset Q.#2?

To an extent, yes.  A related problem is that the cost of traditional boiler maintenance, as well as structural weight, went up with boiler pressure out of proportion to the efficiency gains made.  Another is that higher pressure makes the torque 'peakier' and the engine is more prone to slip at points of peak leverage while still facing the stalling problems of the shorter stroke.

6.  Is Max available speed easier to obtain with a smaller crank circle?

Obviously; the inertial as well as direct augment forces are less, and the lower displacement traditionally requires lower mass flow

7. Any idea what the dynamic loads of the rods are with different circles?

They go up dramatically with increase in crank circle.  Remember that this also applies to lateral or buckling load in thin-section Timken rods. 

The PRR T1 design actually had less stroke than the Milwaukee A Atlantics; and Vauclain of Baldwin said in 1945 that the stroke would probably have been less still if strength concerns in the main driver center between axle and main-pin fits had permitted.

  • Member since
    February, 2005
  • 1,708 posts
Posted by timz on Tuesday, December 19, 2017 2:08 PM

blue streak 1
2.  would bigger crank circle give more tractive effort at slower speeds ?

Maybe you mean, if two engines have the same calculated tractive effort, and one has a longer stroke, will it pull better than the other at low speed?

No reason to think so. Le Massena used to say the 34-inch stroke on NKP/C&O 2-8-4s was great, but no one knows why.

  • Member since
    April, 2001
  • From: Roanoke, VA
  • 1,608 posts
Posted by BigJim on Wednesday, December 20, 2017 7:26 AM

timz
No reason to think so. Le Massena used to say the 34-inch stroke on NKP/C&O 2-8-4s was great, but no one knows why.

Well, Le Massena used to make up a lot of stuff. And, being "published", the Minions ate it up!

.

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