Pulling Power Test

[rickb773]

Subtitle: Who is going to put away all those boxcars!

As I was making my struggling attempt to transition to DCC, I figured with my soldering days years past, that the best way to get enough engines to re-power my layout on my limited budget was to get new Atlas (my “go to” cost effective) DC engines and plug in DCC decoders. I am not a fan of sound equipped diesels (steam yes – diesel no!) thus I could save $30-$50 per engine.

What surprised me was the pulling power (or lack thereof) of the new engines. The GP7/9s and RS-3s seemed closer in power to my VO-1000s (which struggle to pull 10-11 cars on level track) than to my reliable fleet of Atlas GP-38s. So, I figured I would run a test to see if my impressions matched reality.

This was far from a scientific, controlled test. 10 feet of box cars does not fit on a 9’ section of “sort of” level straight track, so some curve track enters the picture. And I really can’t say the northbound and southbound grades are on a consistent gradient. The DC engines sounded horrible and ran very slowly on my Digitrax EVOXD DCC system.




I dug out all my Micro-Train 40’ box cars to have a consistent load (and ended up adding stray Atlas 40 footers to the end).

Here are the somewhat surprising results:

• Go 1960s ConCor PA1 & E7 !!!
• Those E-R Sharks could really haul also

• The Fox Valley heritage models (my second favorite – RIP) edged out the Katos
• 2 older Atlas GP-38s could outpull 3 of the new Atlas Geeps
• I can no longer just double head my 25-car coal drag or tank sweeper.

The 3 new Atlas Geeps seem no better than my underpowered VO-1000s, the 5 new RS-3s are only marginally better and my new Geeps do not track well (I am in the process of shimming my Peco crossover turnouts).

[peteski]

I believe the pulling ability is directly related to the models weight. It would have been very telling if you included the model's weight in your chart.  The type of metal wheels are made of can also make a difference.

[bbunge]

If those new DCC non-sound Atlas engines have empty space where the speaker for the sound version would be, try putting some weight in that space.

[rickb773]

I believe the pulling ability is directly related to the models weight. It would have been very telling if you included the model's weight in your chart.  The type of metal wheels are made of can also make a difference.

New Atlas Geeps & RS3 2.2 oz;
old Atlas RS3 2.3 oz;
Atlas GP-38 2.6 oz;
Bachmann Geep 2.8 oz;
E-R RF16 2.9 oz;
LifeLike SD7 2.9 oz;
old Atlas RS11 3 oz;
Kato SD70ACe; 3.9 oz;
FVM ES44AC 4.1 oz;
ConCor PA1 5.5 oz;
ConCor E7 6 oz.

[randgust]

I've been one of the guys grumbling about this for about 20 years.   It first came up with me when I got a new Atlas SD24 as a replacement, managed to MU and handle the same 10-car train on a 2% grade as it's original, and then I left the room and let it run it with the train.  Came back an hour later and it was stalled on the hill it had actually made a few minutes before.   So that started another theory, it's not just weight, it's wheel material, as on inspection the Atlas wheel blackening was now worn shiny smooth.  And pulling had greatly decreased as a result.

I set up a school-grade spring scale in grams to be a crude dynamometer car to measure two things, tractive effort in grams, and also train resistance, so I could evaluate the effectiveness of wheels.   Add to that a good digital scale to weigh locomotives, and you can pretty easily calculate not just the weight, but the pulling power, and arrive at a coefficient of traction.   Prototype is often stated to be 25% steel on steel without sand; models with wheels on nickel silver rail rarely, almost never, get that.   

More testing over YEARS showed not only that locomotive/frame weight was drastically decreasing (As expected as lead was effectively outlawed, and DCC cutouts reduced that) but that wheels did NOT have the same coefficient of traction; it was remarkable.   It was pretty evident that the newer Atlas, like the SD24, suffered from 'slick wheel syndrome' as the wheels were a much harder and slicker alloy than 'old school' plated brass, and if the plating wore off down to brass, THAT is when you got 20-22%.   I have an Atlas MP15 that got 12%.  Old Trix Uboats were real hogs the older they got.   Traction tires help but don't help as much as I thought.  I presume that the 'slick wheels' help DCC pickup, and my experience with blackened Life-Like wheels on the SW's and GP20's indicates that a factor; they have rough pickup until they have polished up the treads if they have sat a while, but outpull most after that.

Now combine the two factors and you see tractive effort (TE) is plummeting, it's not your imagination.   It is a factor of weight, and add a factor of wheel adhesion (material) to the point that I've really proven that Kato wheels get a better grip than many Atlas, although some of the newest Atlas are now marginally better.   It has nothing to do with number of powered axles, poles in the motor, or gear ratios.   Weight and wheels.   Unpowered wheels are just drag, many worse than others.

I've been grumping on this for so long that when I tested the adhesion factors of some newer Atlas against older ones, and found the adhesion was marginally better, I've wondered if I made any difference.   Probably not, but that's my story and I'm stickin' to it.

The only way to add some consistency to your theory is find a good spring scale and figure how to turn it into a dynamometer car, mine is crude and laughable, but wow, it works.   I had it out the other night measuring drag on a train vs. what my Trix Decapod could do after modification, so it's still a really useful tool.

I've recorded my tests over the years by model, weight, powered axles, grams of tractive effort in an Excel spreadsheet much like yours.   My champs are:

Six-Axle
FP45 body on a Kato SD40-2 mid-production drive with a LOT of additional weight stuffed in.....
Kato SD45's on the original chassis
Worst:  Atlas RSD's, huge difference compared to Kato RSC.
Four Axle
Original Kato GP38 is almost double anything else its size (the 1982 one)
Worst, the Atlas MP15
Steam
GHQ L1 Mikado conversion on a Kato chassis, set a record for single-engine tractive effort, pewter boiler with a traction tire, drop the mike, literally.
Worst, and a 'bigger locomotive' footprint, was the original Life-Like NKP Berkshire, what a disappointment.

PM me if you want the spreadsheet

[bbussey]

Put three Life-Like/Atlas FAs/FBs on the point.  They'll blow everything else away in pulling power.

[bbussey]

Besides ... I've always found that the models pull close to their prototype counterparts.  If you need more power, add another unit.  I don't believe a single RS-3 prototypically could pull 40 cars , especially at grade.

[peteski]

New Atlas Geeps & RS3 2.2 oz;
old Atlas RS3 2.3 oz;
Atlas GP-38 2.6 oz;
Bachmann Geep 2.8 oz;
E-R RF16 2.9 oz;
LifeLike SD7 2.9 oz;
old Atlas RS11 3 oz;
Kato SD70ACe; 3.9 oz;
FVM ES44AC 4.1 oz;
ConCor PA1 5.5 oz;
ConCor E7 6 oz.

That pretty much confirms my theory with few exceptions.  Weren't RF16s a 2-unit set and I believe both are powered, so that is why they ended up on the top of your chart.

If you had the old Kato GP38-2 or GP50, those still had a lead-alloy chassis, so they would likely be good pullers.

Any chance you could redo your chart with the weights included?

[learmoia]

My best pullers were (historically) were..
Life Like Plastic Frame E8s
Early Kato F3/7s
Kato SD80/90MACs are surprisingly good pullers too.

[nickelplate759]

Put three Life-Like/Atlas FAs/FBs on the point.  They'll blow everything else away in pulling power.

Well, three Rapido FAs would pull even more I suspect, but I only have one, and it isn't nearly as good looking as the Life-Like / Atlas models.

[thomasjmdavis]

A few to add to the list, although I don't have a layout up and running, so I can't measure car pulling ability.  I do recall that once upon a time (circa 1995), my Con-Cor PA could pull every MTL car I owned at the time (I think 40) plus an Atlas caboose, on more or less level track.

My Kato F3s (revised version- enough metal cut out to allow for installation of DZ-126)- 4 oz.
Kato F3 (decoder ready with DN-163) - 3.1 oz.
Kato F3 (original version, 1987, original condition, with original mostly metal trucks)- 4.8 oz.  (the decoder ready weighs 5.1 including the box)

And, since I had to pull it out to get the original F3 out of the box....
Life Like E6 - 6.1 oz.

The upside to F units (or FAs, or Sharks) is that if you can't pull the consist with the locos on the front end, just add another B unit.  That's what a real railroad would do. 

[rickb773]

That pretty much confirms my theory with few exceptions.  Weren't RF16s a 2-unit set and I believe both are powered, so that is why they ended up on the top of your chart.

If you had the old Kato GP38-2 or GP50, those still had a lead-alloy chassis, so they would likely be good pullers.

Any chance you could redo your chart with the weights included?

The E-R Sharks were sold as A-B sets but my test only used a single A unit so that refutes your claim.
They were brutes. The A-B pairs easily pulled 40 cars up grade on my old layout (pictured earlier and below.)

[peteski]

Well then, since RF16 weights 2.9oz and it out pulled SD70ACe (3.9 oz) then the wheel adhesion is obviously the answer.

[rickb773]

Any chance you could redo your chart with the weights included?

Your every wish is my command.

[brokemoto]

My best pullers were (historically) were..
Life Like Plastic Frame E8s
Early Kato F3/7s

The LL plastic frame FA-2s were the real stump pullers.  The metal frame FA-1 and FA-2 were not far behind.  I notice that Original Poster does not mention them. Perhaps neither the Penn nor Reading Company ran FAs on the PRSL?

:

Six-Axle

Worst:  Atlas RSD's, huge difference compared to Kato RSC.

The problem with the Atlas China RSD-4 is that while the prototype was geared C-C, the model is geared B-1-1-B.  The old Atlas-Kato RSD-4 was geared like the prototpye:  C-C.   Conversely, the RSC-2 prototpyee was geared A-1-A while the model is C-C.