Trains.com

Subscriber & Member Login

Login, or register today to interact in our online community, comment on articles, receive our newsletter, manage your account online and more!

Measuring Drawbar Pull, Tractive Effort/Weight Results

1479 views
25 replies
1 rating 2 rating 3 rating 4 rating 5 rating
  • Member since
    August 2020
  • 127 posts
Measuring Drawbar Pull, Tractive Effort/Weight Results
Posted by Southgate 2 on Monday, February 22, 2021 3:47 AM

You often see members state the drawbar pull capacity of model locomotives. How do you measure it? Does someone sell a scale, or do you modify something that is readily available? 

I have an idea on how to make one,  but if there's a better way, I'd like to see your ideas.  Thanks  Dan

  • Member since
    August 2003
  • From: Collinwood, Ohio, USA
  • 12,137 posts
Posted by gmpullman on Monday, February 22, 2021 3:53 AM

 

I have an old spring Ohaus scale:

 Drawbar_pull2 by Edmund, on Flickr

Then I bought a digital one but I don't like it as much because it will lock on to what it thinks is the peak reading. Maybe I can "dampen" the display but to me It is easier to see the red needle move on the spring scale rather than try to comprehend the rapidly changing numbers.

 Drawbar_pull3 by Edmund, on Flickr

 DB_Baldwin2 by Edmund, on Flickr

Good Luck, Ed

  • Member since
    August 2006
  • From: Franconia, NH
  • 3,025 posts
Posted by dstarr on Monday, February 22, 2021 5:08 AM

Alternate to the spring scale method,  if you have a set of weights, lead the line from the locomotive over the edge of the table and hook on a weight.  Drawbar pull is the heaviest weight that the locomotive can pull up. 

  • Member since
    September 2003
  • 10,452 posts
Posted by mlehman on Monday, February 22, 2021 5:23 AM

I have a dynamometer car. It started as a Walthers dyno car in this body style.

Since I model the Rio Grande, things are a bit different. They repurposed a troop sleeper, so that's what I did.

I added some lighting so the gauge was easier to read.

The numbers on the gauge are arbitrary. I obtained an electro-mechanical strain gauge with the hopes of getting a better defined reading, but haven't gotten back to that yet.

 

Mike Lehman

Urbana, IL

  • Member since
    July 2009
  • From: somerset, nj
  • 3,488 posts
Posted by gregc on Monday, February 22, 2021 8:15 AM

its common to record the number while the wheels are slipping.    but i don't believe this is accurate.

it would be better to record the highest value just before the wheels begin to slip

greg - Philadelphia & Reading / Reading

  • Member since
    June 2020
  • 1,889 posts
Posted by Lastspikemike on Monday, February 22, 2021 8:36 AM

Weigh the locomotive. 

Rig up an adjustable ramp with powered track.

Measure the maximum angle at which the locomotive will still move up the ramp without spinning its drivers.

Do the trigonometry to convert locomotive weight to tractive effort. 

Or, do what I do. Couple cars to the locomotive until the drivers slip at the steepest grade on my layout. Then add a locomotive and keep on going.

Three Proto 1000 C Liners will easily pull my 8 car IHC passenger consist of Budd style cars, weighted and with their standard awful trucks up the steepest grade on our layout around the 24" (+/-) radius curve near the summit of a 3% (+/-) grade.  Two will struggle but still not spin the drivers.

Put another way: prototype railroaders knew their locomotives were grade limited so we can use the same methodology. 

Alyth Yard

Canada

  • Member since
    May 2010
  • 7,878 posts
Posted by mbinsewi on Monday, February 22, 2021 10:10 AM

I guess knowing draw bar pull is all well and good, if you feel you need to know that info.

I've always added cars until the wheels slipped, than removed 1 or 2 or whatever it took to keep the train moving.

I suppose draw bar pull is just one of those things, that modelers who like to document data,  feel is nesessary.

Mike.

  • Member since
    January 2009
  • 5,954 posts
Posted by RR_Mel on Monday, February 22, 2021 10:16 AM

I bought a Portable Electronic Scale off eBay and so far it has worked very good.  Most of my locomotives are heavy and have a lot of pulling power.  On my tests the max drawbar occurs at wheel slip and doesn’t change with the amount of wheel slip, probably because of the weight of the locomotives.

A typical Rivarossi Cab Forward out of the box weighs 11.8 ounces and has 2.8 ounces of drawbar.  After remotoring and adding 10 ounces of weight to the boiler the draw bar increases to 5.8 ounces.

My typical EMD E7s have an Athearn SD40-2 frame remotored with a Mabuchi FS280 and a Cary metal body weighing in at 33 ounces with 9+ ounces of drawbar.

I put a screw eye in a Kadee coupler gauge and use that to couple the locomotives to the scale.

 
    
 

Mel



 
My Model Railroad   
http://melvineperry.blogspot.com/
 
Bakersfield, California
 
I'm beginning to realize that aging is not for wimps.

  • Member since
    September 2002
  • 7,036 posts
Posted by ndbprr on Monday, February 22, 2021 10:25 AM

MR used to have a pulley mounted in a bracket at the edge of a table and used weights to make the determination. I suspect with today's quality we will run out of passing siding capacity before reaching engine capability

  • Member since
    September 2003
  • 15,935 posts
Posted by Overmod on Monday, February 22, 2021 10:36 AM

Perhaps someone will care to build something I thought about years ago.  A 'proper' analogue to a dynamometer car could serve a number of purposes that a plain old scale can't, for example finding small faults in trackwork that have the effect of increasing train resistance, varying perhaps by car type and individual truck performance.  For this a time-varying radio signal would be transmitted to the equivalent of a chart recorder on a computer (or equivalent) which could then display prototypical charts, give proper average drawbar-pull measurements without artifacts or misleading peak-hold (a la Mallard Whistling) or difficulty reading scale numbers designed to be read in stasis and hence poorly buffered.

Probably such a scale has been built for some dedicated application.  On the other hand it would be reasonably easy to take something like a Rolling Thunder daughterboard and tie it into Mel's scale's sensor output, with appropriate trim resistors etc., and receive the result in modified lab software...

  • Member since
    January 2009
  • 5,954 posts
Posted by RR_Mel on Monday, February 22, 2021 11:08 AM

Most transducers are available 4-20ma output and with the Arduino Bluetooth anything is possible.  I have a wireless 160° panable camera on a HO flatcar and a wireless speedometer in a HO baggage car so a wireless drawbar meter is probably doable too.

    
 

Mel



 
My Model Railroad   
http://melvineperry.blogspot.com/
 
Bakersfield, California
 
I'm beginning to realize that aging is not for wimps.
 

  • Member since
    September 2003
  • 10,452 posts
Posted by mlehman on Monday, February 22, 2021 11:09 AM

Overmod
A 'proper' analogue to a dynamometer car could serve a number of purposes that a plain old scale can't, for example finding small faults in trackwork that have the effect of increasing train resistance, varying perhaps by car type and individual truck performance.

While it's analogue and somewhat arbitrary, my hacked Walthers dyno essentially does this. You have to walk along and observe, and there is no recording unless you bring a camera, but it does clearly show the effects of grade, rolling resistance, and trackwork as the train encounters them.

The parts I bought to digitize things seemed designed to make things like a hanging scale, so I hoped they would be sensitive enough to give relevant readings. However, I found what I needed to know for my own pruposes with a little observation, so haven't felt the need to fix what's already working. Obviously, a digital readout and documentation would be helpful in cross-layout and multi-loco comparisons.

Mike Lehman

Urbana, IL

  • Member since
    January 2004
  • From: Canada, eh?
  • 11,574 posts
Posted by doctorwayne on Monday, February 22, 2021 1:07 PM

In general, on a well-balanced locomotive, drawbar pull is usually equal to about 25% of the weight of the loco, which is similar to the real ones.

I modified three Athearn U-Boats, installing two motors in each, and adding weight until they were each just over 33 ounces.  I had no way of measuring the drawbar pull, until a friend suggested that I take one to his model railroad club, were they did such tests to determine how many locos would be needed for trains of various lengths and weights.
The test facility was simply some track on a hinged board, and the procedure was to elevate the free end of the board until the loco's wheels slipped.  They then did some sort of calculation which used the degree to which the board had been elevated, which came out as 8.3oz., pretty-much dead-on 25% of the loco's weight.

When I returned home, I placed one of those diesels on a train of loaded hoppers and ran it up a 2.8% grade, laid-out on an S-curve.  It managed to get about 1/3 of the train onto the level area at the top of the grade, but then slipped and could move no further.
I backed the train down the hill, then coupled another of the same diesels to the train, which easily hauled the entire train up the grade,without slipping.  I would have added more loaded hoppers, but had no more.  The trailing train was 44 cars, each weighing 8oz., for a total of 22lbs. of trailing weight.
While I still have the hoppers (and quite a few more) the three locos were bought by a friend when I later backdated my layout to the late '30s era.

Some time later I came across this gadget from MicroMark (click on the photo to enlarge it, as the wire is otherwise difficult to see).....

The unit is battery-powered and can show the drawbar pull in either grams or ounces.  To use it, I bent a piece of piano wire, forming a hook on each end, with one end hooked to the reader, the other to the coupler on the loco's tender.  While holding the tester, power is applied to the tracks and the loco moves until its wheels slip, the read-out climbing until slip occurs.  That figure remains on the display, until the user re-sets it.
I have no way of verifying the accuracy, but by testing all locomotives, then running them with various trains of recorded weights, it should be reasonably easy to extrapolate how many locos, of various capabilities, would be needed to pull trains of any particular weight.
Before acquiring the tester, I did test a number of locomotives as to their pulling abilities, but because I don't necessarily adhere to NMRA weight recommendations, the results were "varied". Smile, Wink & Grin

Wayne

 

  • Member since
    June 2020
  • 1,889 posts
Posted by Lastspikemike on Monday, February 22, 2021 1:12 PM

It'll be approximately the coefficient of friction of nickel silver  x locomotive weight in whatever units suit you.

Of course the load pulled varies with grade, curvature and truck condition for each hauled car. Knowing drawbar pull is of esoteric interest.  

Alyth Yard

Canada

  • Member since
    August 2020
  • 127 posts
Posted by Southgate 2 on Monday, February 22, 2021 6:01 PM

Thank you for the replies.  I want to use it for the same reason that Mel does, measure improvement in locomotives as modifications are made and weight is added. Mainly curiosity,  which is part of the fun. Dan

  • Member since
    January 2009
  • 5,954 posts
Posted by RR_Mel on Monday, February 22, 2021 6:09 PM

I bought my scale off eBay.

https://www.ebay.com/sch/i.html?_from=R40&_trksid=p2047675.m570.l1313&_nkw=Portable+Hanging+Fishing+Scale+Digital+Pocket+Weight+Electronic+Hook+40kg%2F10g&_sacat=0

 

Mel



 
My Model Railroad   
http://melvineperry.blogspot.com/
 
Bakersfield, California
 
I'm beginning to realize that aging is not for wimps.

  • Member since
    May 2002
  • From: Massachusetts
  • 2,782 posts
Posted by Paul3 on Monday, February 22, 2021 9:14 PM

At my club, we use a spring scale mounted vertically as it was designed to be.  The scale measures from 0 oz. to 16 oz. in 1/4 oz. increments.  If you use certain kinds of spring scales in the horizontal position, it can induce drag on the internal components and affect the reading.

Our scale has a string attached that's tied to a piece of piano wire bent into a hook.  The string wraps around a teflon pulley on a teflon bearing designed so that the string is at coupler height.

Two important steps to take is to make sure both the track and wheels are clean.  Dirt acts like rubber tires and can change the reading by a significant amount.

We run the loco at full speed and check the reading, then we slow the engine down to a stop and check again.  In a perfect world, it'd be the same reading but it rarely is.  We usually go with the stopped reading.

I also purchased the Micro-Mark digital scale, and when I tested it out using proper test weights, it gave inconsistent readings.  Some weight indications were ok and some were not.  Worse, it was seemingly at random.  I used the same test weights on our spring scale and it was correct for each one which is why we still use the spring scale.

  • Member since
    August 2020
  • 127 posts
Posted by Southgate 2 on Tuesday, February 23, 2021 7:20 PM

I read and considdered all replies, thank you. The only hanging scales I could find were for weighing fish up to about 100 lbs, I doubt they'll be real accurate in the tenths of ounces.

The incline idea doesn't take into consideration tender weight, which I add plenty for current pickup.

Well, the pully and weight system sounds good, but with a twist here. This was the idea I mentioned in the original post, and seeing no easier accurate way around it, here it is:

The little Harbor freight scale I've been using for years to weigh locomotives has proven itself sensitive and accurate, so I applied the pully idea to it. I made the pully and frame from aluminum,   

   

 

and an idler truck from a delrin passenger truck with a coupler. Added them to my existing test track board; done!  

Place the 4 oz weight on the scale, push "tare" to zero it. Then let the loco pull the little super free rolling idler truck, and measure "negative weight" on the scale.   

   

The weight showing on the scale at the time I took the pic wasn't accurate, the little Pocher Inyo can pull an honest 2.6 oz. I've added considerable weight to it, the engine (no tender) weighs 8.58 oz.  Drawbar pull: 30%, due to traction tires.

I've added weight to other locos as well. One example, a Mehano 2-6-0 (yes, I like those!) engine only originally weighed 9 oz, and pulled 1.2 oz. With 2.2 hard earned oz added, it pulls 1.5 oz. about 14% it's weight. Those have very shiny slippery drivers. So just figuring about 25% loco weight=drawbar pull proves pretty loose. 

This is why I wanted an accurate drawbar pull scale, to see what improvements are being made.  The added weight of course adds to electrical pickup as well.

Thank you for the input, I do appreciate it! Dan

 

  • Member since
    January 2009
  • From: Maryland
  • 10,492 posts
Posted by ATLANTIC CENTRAL on Thursday, February 25, 2021 6:10 AM

This has been an interesting thread, something I have considered doing myself.

Up this point I have used a different approach to measure pulling power.

I have about 70 pieces of identical rolling stock, same car, same trucks, same weight. I measure pulling power by simply testing how many of these a loco can pull.

The car in question, Athearn 50' flat car with two vans, trucks are Kadee sprung metal trucks refitted with Intermountain wheel sets.

At 4.3 oz each these are one of the heaviest pieces of freight rolling stock I have, so it is also a good practical measure of motive power requirements for all trains simply knowing how many of these a loco can handle.

And, most of my freight cars have those same trucks, which are very free rolling and track very well. 

Sheldon 

    

  • Member since
    December 2016
  • 97 posts
Posted by speedybee on Thursday, February 25, 2021 8:52 AM

Nice work, Dan. If you measure more drawbar pulls, please post the results!

  • Member since
    June 2020
  • 1,889 posts
Posted by Lastspikemike on Thursday, February 25, 2021 9:24 AM

Tender weight is included in the incline method as long as you weigh only the locomotive first. Tender weight reduces net drawbar pull available to pull the cars.

Tender weight reduces net drawbar pull by the draw resistance of the tender. That will reduce maximum height of the incline proportionally at the point the drivers begin to slip.

Nobody needs to know the pulling power of a steam locomotive without its tender.

Alyth Yard

Canada

  • Member since
    September 2003
  • 10,452 posts
Posted by mlehman on Friday, February 26, 2021 2:09 PM

Southgate 2
The weight showing on the scale at the time I took the pic wasn't accurate, the little Pocher Inyo can pull an honest 2.6 oz. I've added considerable weight to it, the engine (no tender) weighs 8.58 oz. Drawbar pull: 30%, due to traction tires. I've added weight to other locos as well. One example, a Mehano 2-6-0 (yes, I like those!) engine only originally weighed 9 oz, and pulled 1.2 oz. With 2.2 hard earned oz added, it pulls 1.5 oz. about 14% it's weight. Those have very shiny slippery drivers. So just figuring about 25% loco weight=drawbar pull proves pretty loose. This is why I wanted an accurate drawbar pull scale, to see what improvements are being made. The added weight of course adds to electrical pickup as well.

Yes, now you're hitting on why just calculated by weight assessments of drawbar pull have some real limits. This number is good to know and gets you in the ballpark, but doesn't help that much is you're trying to make (and measure) gains in performance,

Traction tires, for example, can positively affect tractive effort. More often, other factors detract from it and thus you want to be able to record via some standardized method with common units the incremental change that would result in solving issues that detract from TE.

While fewer people run steam now, there were lots of things that could detract from calculated TE. Often the use of springing, etc on leading and trailing trucks could detract from weight on drivers. You often needed some way to help things track, but here it was more beneficial on TE to try to add weight to these trucks rather that using springing, wihch would have more effect on TE.

Just some examples, there's plenty more, including for diesels, that suggest the usefulness of being able to measire TE in units that can then translate between subjects for comparison purposes, etc.

Mike Lehman

Urbana, IL

  • Member since
    July 2009
  • From: somerset, nj
  • 3,488 posts
Posted by gregc on Friday, February 26, 2021 3:44 PM

mlehman
Yes, now you're hitting on why just calculated by weight assessments of drawbar pull have some real limits. This number is good to know and gets you in the ballpark, but doesn't help that much is you're trying to make (and measure) gains in performance

not sure what you're trying to say

in general, TE is 20-25% of the weight on the drivers, assuming the weight is evenly distributed.     just adding weight to the front or back may not have the expected result.   

i recently posted the thread "loco modifications affecting weight on drivers - Reading" to discuss when prototypes realized this.    weight wasn't evenly distributed on RDG I-5 2-8-0s built in 1891, but was very even with the I-10s built in 1918.

greg - Philadelphia & Reading / Reading

  • Member since
    September 2003
  • 10,452 posts
Posted by mlehman on Friday, February 26, 2021 5:59 PM

gregc
just adding weight to the front or back may not have the expected result.

Yes, among other things, weight and where it balances certainly can affect TE. That's why simply calculating it acxcording to weight provides a general value, but doesn't help much in telling you whether it's best balanced front to back. Then you need a way to measure any tiny improvements or detriments to TE as you adjust things for best balance.

Mike Lehman

Urbana, IL

  • Member since
    August 2020
  • 127 posts
Posted by Southgate 2 on Friday, February 26, 2021 7:47 PM

The following will illustrate what Mike is talking about. I took down a few locomotive TE measurements, as follows. Note the wide difference in weight to TE ratios. I'll add side notes that will also explain the inconsistancies. Recorded weight is of the loco only, not the tender. The tender's weight affecting TE would only really apply on grades, which I have none. The tender's rolling resistance will have whatever effect it will.

On all locos, I usually let the wheel slip over 1 revolution, and then took the lowest TE. On some they're pretty consistent, but traction tire ones can vary considerably. 

Loco                        Weight     TE  TE/weight ratio.  Notes

Pocher Genoa 4-4-0  4.77 oz   1.5 min  31.4%  Stock weight, stock traction tires. new motor and flywheel in tender, shafted to drivers

Pocher Inyo 4-4-0    8.58 oz   2.6 oz min     30%   Same as Genoa, but with a ton of lead weight added.

Pocher "JW Bowker 2-4-0  4.65 oz 1.6oz 34 %?  Stock weight, 2 traction tires. Early- motor in tender/steel chasis version. Very inconsistant TE, all over the place.

Bachmann 4-4-0          5.40 oz   1 oz       18.5%   Stock weight, stock tt (2)

MDC  0-6-0 switcher   8.1 oz    1.58  19.5%  Has plated Mantua drivers with one traction tire. Pretty consistant TE

Athearn SW 1500        10.58 oz   2.75  26%  NWSL plated wheels

Mantua 4-6-0 ("Dixie..." 8.88      1.03   11.5%   Shiny plated drivers.

Rivarossi 2 truck Heisler  9.04     1.9      18.3%  Stock model, no traction tires

Atlas YB Austria GP 38     14.17   2.6       18.4%    Stock model

Atlas Japan RSD 4/5        13.2     2.1     16%          Stock

MDC/Athearn RS 3          14.8     3.1      20.09%       Steel wheels, weight added

BB Athearn GP 35          11.9oz     2.8 oz     23.5%  Stock Old slugger, Steel wheels and metal sideframes. Pre can motor, has zinc flywheels

BB Athearn SD 40,         15.1 oz   3.5 oz      23.5%  Later BB, wide hood but plastic sideframes, steel wheels, can motor brass flywheels

P2K Alco S-1                   8.43     2.25    26 .7%       Stock

Mantua Mike, see notes    25 oz     3.8 oz   15.2%   Heaviest loco tested. Has "Power Drive", and a Cary Heavy USRA boiler. Ridgid chassis, No lead or trailing truck (incomplete but running kit) 

P2K GP7                        15.6       5oz!       32%     Only mods, Athearn gears.

...So, 20 to 25% is a general rule that could apply to diesels, with their swiveling trucks. But even then that proves loose, except maybe on BB Athearn.

  At over a pound and a half on just drivers, I expected the Heavy Mike to pull more, but 3.8 oz TE will pull more than it will ever be required to on my layout. later on sometime, I'll check how balanced it is and see if balancing it changes the ratio.

The P2K GP-7? Wow! I already knew this was the heaviest puller on the layout, but 5 OZ!   32% TE/W on metal wheels. It out traction ratioed models with traction tires. I had to ad an ounce to my 4+ oz scale dead weight. 

The Pocher locos above are AHM  Rivarossis with all steel chassis, pre plastic chassis version. 

I'm enjoying this gadget.  Dan

 

 

  • Member since
    September 2003
  • 10,452 posts
Posted by mlehman on Monday, March 1, 2021 3:50 AM

Dan,

Great data! Shows how having a scale is necessary to get beyond approximations. While things are flat on your layout, if they weren't you'd probably be looking at ways to improve the TE on some of those locos, where an accurate scale to measure this would be quite handy.

Mike Lehman

Urbana, IL

Subscriber & Member Login

Login, or register today to interact in our online community, comment on articles, receive our newsletter, manage your account online and more!

Search the Community

ADVERTISEMENT
ADVERTISEMENT
ADVERTISEMENT
Model Railroader Newsletter See all
Sign up for our FREE e-newsletter and get model railroad news in your inbox!