Required reading for steam fans, National historic mechanical engineering landmark: the "J"

Saw it sitting cold this past June in Spencer, NC. Beautiful piece of machinery  !!

The HP and Tractive Effort curve is incorrect, and is evidently one made by a N&W railfan who worked in the RR supply industry but never worked for N&W and was not an Engineer. The correct curve was published in the Jeffries book (p. 59) as well as the N&W "J" book by Miller. The correct curve was first published by Pond in an RME article in 1946.

Are you talking about figure 2? Could you scan and post what you think is the correct one, would be interesting to compare. 

I concur with Hudson5432 regarding the 300 psi curves.  I believe he has first hand information regarding this subject.  I can add this:  The N&W developed the correct curves during comparative tests with J 604 at 275 and 300 psi.  The curves also appear in the test report as Chart 1 (IIRC).  However, they are calculated curves based on the Baldwin method, not actual curves based on test results.  There's a note on the Chart specifying this.  The original test report is available at the N&W Historical Society archives in Roanoke.

I am attaching four different curves for the N&W J. The 1st curve appeared in Railway Mechanical Engineer, in the December, 1946 edition. The curve is contained in an article authored by Pond. I consider this a primary source document and this, and the fact that he was Asst Supt of Motive Power for N&W, I regard as factual. (I attached pages 656, 657, and 658.)

The second curve appeared in the Jeffries book on p. 59, and this curve is the same curve as that contained in the Pond article.

The 3rd performance curve appeared in the Miller book on the N&W J. The 3rd curve that compares the "J" to a PRR T-1 and a 5400 HP diesel electric has obvious errors. For example, it shows tractive effort of 25,000 lb at 75 mph for the "J". (Drawbar pull would be lower than tractive effort and, without using a dynamometer car, the effort to haul the engine and tender at 75 mph would have to be subtracted.) The formula for drawbar (at rear of tender) horsepower is drawbar pull times the speed in mph, and this  divided by 375, which is the horsepower constant in US units of measure. So... 25,000 lb TE at 75 mph is 5000 hp. The problem is that both the Pond curve and the Jeffries curve show this engine with 4200 HP at 75 mph, not 5000.

The 4th curve in the Miller book on page 69 is closer to the Pond curves but is not the result of an actual test. It is a calculation. What is interesting to me about this curve is that the actual test points from a dynamometer test are shown and NONE of these test points are at a speed greater than 45 mph.

For all those that are interested in this kind of detail and comparative source material,  please note that in the sources noted above by Hudson 5432, all four of the curves are identical.  Also, the 300 psi curves for DBPull and DBHP are different from those in the ASME booklet. This is where some sort of error crept in.  In the bottom example of Chart 1, you can see the explanatory note that I referred to in the upper right corner of the chart.  It spells out how the curves were calculated, using a Baldwin-developed method that was in widespread use in the railroad industry.   I was able to replicate the erroneous 300 psi curve by misusing the Baldwin method, and after comparing it with the N&W 300 psi curve, I considered the ASME booklet interpretation as wrong. Why this curve survived any editorial muster and was included in an ASME sanctioned publication, I'll never understand.  It just doesn't make any sense to me, for what that's worth.

I believe that Hudson 5432 and I know who each other are, so I'll leave it to this:  if you're TG and I'm DS, then we're on the same page and have been for years.  Glad to exchange information with you again on this subject.

You guys are impressive, whatever your initials! I am looking forward to digesting this and understanding these calculations. Thanks for taking the time to post details and explain.

Love this forum.

Hello there Dave!

Yes, I am "TG".

Thank you for your constructive comments, and very pleased to see someone else who worked in this industry contributing!

Re the ASME, I was also surprised that this erroneous info was forwarded, evidently without further checking, as their reputation is generally excellent. A 1550 HP error at 100 mph on an Engineering curve is not something that I would be proud of though...........

As info, a lot of the early issued performance curves and quoted power levels are wishful thinking. I recall reading where a D&RGW official remarked that his 4-6-6-4's had 50,000 lb drawbar pull at 60 mph. Well, that works out to 8000 drawbar HP! D&RGW Challengers may have been good engines, but certainly not capable of that power level.

PRR Altoona Test Plant, with a roller dyno, never checked drawbar pull below 35 mph, instead drawing/filling in the curve to zero speed. That is where MOST of the operation took place, as the PRR freight speed limit was 50 mph for general freight. I think that is another reason why PRR completely overlooked the diesel early on as any more than "just an inferior competitor" to steam.

Hokie71,

If that's Va Tech class of 1971, there may be another connection here.  I'm VT class of 1964.

You nailed me! You and hudson5432 are killing this EE who only had the Jones 5 credit thermo class! Check out my profile and send me an email.

Of course EEs today don't even have that!

Of the four curves, all of which have the same shape, the Miller curve is different. Three of the four curves have drawbar pull as the unit of measure. DB Pull is most appropriately measured using a dynamometer car. The Miller curve that includes the PRR T-1 and the EMD FT diesel has "tractive effort" as the unit of measure. You can calculate tractive effort and in fact that is the commonly used measure of performance, at least for diesels. The difference between TE and DB pull is that you have to deduct the effort necessary to move the engine and tender in order to arrive at the energy required at the tender coupler to pull the train (ie drawbar pull). For a passenger application and assuming speeds greater than approximately 40 mph, the difference between tractive effort and DB pull is significant due to increasing wind resistance above 40 mph. Also, since wind resistance increases as the square of the speed, and at  80-100 mph is high and rapidly increasing.

I've seen a lot of youtube videos showing N&W 611's drivers slipping. Can 611 really take advantage of having 75,000 to 80,000 pounds of tractive effort in less than ideal conditions while only having a factor of adhesion of approximately 3.6? NKP 765 and UP 844 both have a factor of adhesion of approximately 4.1 to 4.2 and don't seem to slip the drivers nearly as much as N&W 611.

hokie71 posted:

Of course EEs today don't even have that!

'05 EE, U of Akron, definitely took thermo - and definitely finding this thread a good read.  Thanks everyone!  Always enjoyed that ASME article, and in fact just shared it with an ME friend of mine who just discovered steam preservation and the J.  He's about to get a link to this thread to complete the picture.

Blystovski, good to hear a 21st century EE with thermo- seems the interest now is how many billion transistors can be squeezed on a chip.

to Hudson, Felton and the gurus , is there a good resource on what is inside a dyno car, how they operated, and how for instance draw bar pull was measured? Would be interesting to read details about these tests too.

I believe I saw a summary of dynamometer car design, for a specific railroad, in either Railway Age or Railway Mechanical Engineer a long time ago. I would have no idea how to locate that article. I believe there is a NKP dynamometer car at the museum in Bellevue, OH. I don't know if it has been stripped of the measuring equipment. In the old days, these cars had a fluid drawhead that was calibrated to the instruments, which were of course analog. In steam locomotive testing, an "observer" in the cab would radio specifics such as valve cutoff, boiler pressure, throttle opening, etc. to the test car crew. There was a table (in the NYC car) with recording pens that would trace the drawbar pull in lb. It was very important to know the exact speed, and I believe that there was a calibrated speedometer in the car. Mileposts were also important, and usually an observer would either call out or use a handheld device that would put a mark on the trace to show mileposts. Once this raw data was available, the track chart for that part of the railroad would show grades and curvature so the corrections could be made to determine drawbar horsepower. If the train was either accelerating or decelerating, or on any grade or in curvature, the HP measured would be incorrect. For this reason, the NYC used a "brake engine" positioned directly behind the test car and the train. The purpose of this engine was to hold the speed exactly constant, so the car could measure "pull" and speed exactly. Using this engine, numerous readings could be taken of the drawbar pull at that exact speed. Most tests precisely measured the "pull" at 37.5 mph. Since the "HP constant" is 375, you can then measure the pull in pounds and divide by ten, and know the drawbar HP exactly. For example, if a locomotive is recording 25,000 lb. of "pull" at the drawbar at 37.5 mph, it has 2500 drawbar HP at that speed. (In all of the valid tractive effort and drawbar curves you will ever see, the tractive effort line and the drawbar HP line will both cross at exactly 37.5 mph.) One "high speed" test  point is 75 mph, so you can, by observation, know the DBHP at that speed, but you have to "divide by two"!

In the diesel age, most equipment of this type is called a "test car". The instrumentation is all electronic. The last time I was on one, I was drafted to call out the mileposts. The engine trace at my seat was recording over 167,000 lb of drawbar pull, and the speed was so low that my video shows individual pieces of ballast that I could count moving under the end sheet on the #2 end of the lead locomotive. I got out of there (the #2 end walkway) once I realized that if a knuckle failed, it might not be "pretty". (I noticed the two knuckle holders, with knuckles, at about this time, on the rear platform end sheet.) I believe the speed was less than 2 mph.....

The maximum performance of modern diesel electrics is really hard to believe.

Whoops! You have to multiply by two at 75 mph!

I am slowly starting to get it. I was over complicating the process thinking about the dynamometers I had worked with on the tail end of truck and car diesel test stands. The last post made me realize what was going on.  Here is pretty good description from wikipedia on dynamometer cars.  the 75mph versus 37.5 is pretty clear in the first equation. Still can't find any pictures of the force measuring mechanisms they used back in the day (strain gauges make it easy!).

Feltonhill and Hudson 5432,

I always enjoy reading your posts about locomotive performance and analysis.  I am looking at various test data sets to gain an understanding of the Machine Resistance component of Locomotive Resistance.  By Machine Resistance, I'm referring to mechanical losses due to the pistons, the valve gear, the rods.  Johnson recommends using a factor of 20 lbs/ton of weight on drivers to account for MR.  In my opinion, this penalizes locomotives with high Factors of Adhesion and rewards low FOA because it is simply based on adhesive weight, and not on the dimensions of the cylinders, operating pressure, valvegear design, etc.  

To analyze MR, I start with IHP and DBHP data (or, even better, Cylinder Tractive Effort and DB Pull data), and then subtract the calculated contributions for axle bearing friction (roller or plain), flange friction, and air resistance (all estimated from methods described in Johnson's book).  I account for both locomotive and tender resistances in this way.

Thus far, I have Cylinder and Dynamometer test data from the NYC S1b Niagara (from Dawson's "Road Testing the Niagaras"), and three Santa Fe types (5001 class 2-10-4, 3460 class 4-6-4, and 3765 class 4-8-4) from Farrington's "Santa Fe's Big Three".  There is enough description of the test methods from those two roads that I have confidence in the data being correctly adjusted for acceleration, grades, curves, etc.  Do you know of data on other locomotive types where IHP and DBHP for the same test conditions is available?  

Your response is much appreciated.

Scott Griggs

Louisville, KY

You have a formidable assignment! One reason is that, depending on the cutoff at any point in time and of course the speed, the thermal of the cylinders changes, and therefore the "pumping losses", and including piston valve and piston friction, etc. I remember reading that Franklin Railway supply claimed that a "piston valve gear" required 80 HP per side to operate, and the Franklin poppet valve system required 3 HP per side. Engines that were completely roller bearing equipped had very low mechanical friction. (There is one photo of a new NYC Niagara being pulled by three young women on level track. However, I am sure they did not pull that engine very fast!) Engine and tender loaded weighed 891,000 lb.)

There was also a difference in friction that was dependent on the brand of antifriction bearings used. A NYC locomotive performance guy told me that when NYC ran tests to determine the speed at which their automatic train stop (ATS) would function, they tried testing with the SKF bearing equipped Hudson 5343 and could only achieve a little over 100 mph. So they called for Timken equipped 5344 and ran to 113.6 mph. There may be something buried in the complete Niagara road test report, which is based on the ASME Test Codes. The Niagara report contains 851 lines of data.....

I do not have the PRR T-1 Test report from Altoona and not sure it still exists, but have seen claims of 93-94% machine efficiency with the use of poppet valve gear. This seems a little high to me and of course does not include tender weight, friction, etc.

The locomotive resistance equations, either the Davis or the AAR, have an air resistance term and also a journal and bearing resistance term. Both of these equations are the result of extensive testing and are a little conservative.

The ATSF "Big Three" and the Niagara Test report are the only ones available with this degree of detail that I am aware. Note they are all actual tests done with dynamometer cars and not "calculated".

I wrote an article analyzing the performance of NYC&HRRR #999 for the 1Q2017 edition of the Central Headlight that includes this type of analysis. Should be out momentarily.

I want to tell a story here which shows how fragile the history of locomotive testing can be.  About 10 years ago, I located the PRR T1 Altoona test report at the RR Museum of Pa in Strasburg.  Another researcher and I wanted to get a copy of the Altoona test and the other person being from Australia (Neil Burnell), the job fell into my hands.  I contacted the RRMofPa and they agreed to let me make two copies of the T1 test (in exchange for a donation on my part) which I was very glad to do.  When I got to Strasburg, we discovered that the T1 test was  a carbon copy on onionskin, probably about a third generation copy.  I suggested that since each page would have to be artwork (single copy, backing of plain white paper to prevent print-through and highlight the already degraded type face,) that we make four copies.  That way the Museum archives would have two copies on standard weight paper that researchers could use without handling the already fragile original.  OK, that went off without a hitch.  There were four copies, one for me, one for Neil in Australia and two for the Museum archives.

About two years ago, when the T1 trust (replica 5550) came into being, they wanted to get a copy of the Altoona test for their website. I received an e-mail asking if I had a copy of the T1 test, and of course I did, good old paper copy.  Apparently, there was a problem.  The Museum couldn't find the original T-1 test or either of the two copies.  It very quickly dawned on me that I had what could be (at least at that moment) the only existing copy of the Altoona tests in the country.  I made arrangements to get the test copied at a local office supply store, within my sight (no drop-off or come-back-later), so that there would be no chance of loss.  The test was converted to a PDF, copies transmitted to the T1 Trust and the Museum, and everything settled back down again.  I did backups on three different media and stored them in three separate locations.  The whole event was a real heart stopper for me, and probably several others  as well.  Steam locomotive research doesn't often get that exciting!!

Fortunately the NYC Niagara tests are not that endangered.  I'm sure that Hudson5432 and the NYCSHS have copies, and I purchased a copy from their archives myself many years ago.  It is without a doubt, the gold standard of modern locomotive road testing.  The details are awesome, and the accuracy amazing, virtually laboratory quality

Locomotive test data is very hard to find, and its existence can be extremely tenuous, as you can see.

I wanted to put this out to reassure Hudson5432 that the Altoona T1 test report does, in fact, still exist, and it's certainly worth careful examination.  As for the other questions, I'll have to spend a bit more time on those to see if I can contribute anything of value here.

MattA

The N&W Class J's starting tractive effort (STE) is the subject of much discussion, but I prefer to look at it this way.  The J was designed with relatively large cylinders, such that its calculated STE would be  somewhat high.  Under poor rail conditions, it certainly could NOT get all that force to the rail.  However, keep in mind that the J's daily job was to accelerate trains from frequent stops  to track speed  relatively quickly, move 15-17 car trains over mountain grades and start them on these grades when necessary, and accelerate from 25 mph slow orders to 50 mph repeatedly.  An 17-car train doesn't not require 80,000 lbs STE to get underway on N&W's grades.  So the J's large cylinders came into play above 25 mph, where slipping would be much less of a problem than when starting.  They gave the locomotives almost breath-taking acceleration in the mid-speed ranges.  I witnessed this phenomenon in 611's cab (June 1986), and it's not imaginary!

As reference to the extreme low speed and of things, I recommend looking at the video on youtube showing 611 barely making it over the summit at Linden, Virginia, on the Manassas-Riverton line during an excursion in 2016.  But make it she did, at about 1 mph with one pop valve lifted, indicating that even under lousy rail conditions she could still hold on and get a 20-car train over the top of a 1.6% grade on something like a 2-4 degree curve!

Scott,

I've found that locomotive resistance (total, as described in Johnson's book) tends to be more constant than the formulas historically used.  I believe I read the comment somewhere, that locomotive resistance (the difference between cylinder tractive effort) and drawbar pull at the rear of the tender) tends to be more linear then exponential.  Further it varies more in response to how hard the locomotive is being worked rather than just pure speed.  I've been messing around with this idea for some time, but have never been able to separate the two components (cutoff/throttle and speed) primarily for lack of detailed data.  Empirically, I'm not sure it matters very much anyway.

As a rule of thumb, depending on whose thumb you're using, a modern 4-8-4 should have a low speed (0-5 mph) resistance of about 3,000 lbs. (total, engine and tender) which will increase to about 9,000-12,000 lbs at 80-90 mph.  I've usually found this by  subtracting drawbar pull from cylinder tractive effort.  Not very precise, to be sure, but it's a decent indicator.

Nonetheless, if you make any breakthrough, I'd like to know about it.  I frequent this forum several times a day  so I'll be sure to notice!  I appreciate your questions and am heartened by the interest of a younger generation.

Hokie71,

Drawings of the interior of dynamometer cars are available through various railroad historical society websites.  If you Google dynamometer car drawings, or something like that and look at Images, you can likely find several dyno car interiors.  There are some existing dyno cars.  N&W's is at the VMT at Roanoke, but most of the measurement machinery has been removed.  However the rest of the car is reasonably intact, except it's not open to the public because of floor problems and other safety considerations.  I was allowed to go inside to get info for a presentation I was doing, so I know what's there and what's not.

I believe there's a good drawing of the NKP dyno car available and you may want to search that to see what's out there.  The original is in a museum as mentioned above.

Hope this helps

feltonhill posted:

MattA

However, keep in mind that the J's daily job was to accelerate trains from frequent stops  to track speed  relatively quickly, move 15-17 car trains over mountain grades and start them on these grades when necessary, and accelerate from 25 mph slow orders to 50 mph repeatedly.  An 17-car train doesn't not require 80,000 lbs STE to get underway on N&W's grades.  So the J's large cylinders came into play above 25 mph, where slipping would be much less of a problem than when starting.  They gave the locomotives almost breath-taking acceleration in the mid-speed ranges.  I witnessed this phenomenon in 611's cab (June 1986), and it's not imaginary!

A good recording of the acceleration Dave is speaking of can be found on "The Fading Giant - Sounds Of Steam Railroading Vol. 2" (this being the revised, stand alone version CD that is 79:40 in duration). On track #2 you will hear how quickly a J can accelerate train #15 which was known to be heavy with head end cars and made frequent stops.

BTW, this CD is very much different than the original "F/G" and well worth the price of admission. Along with different tracks of trains #15 & #16, it also has an extended version of the Christmas Eve at Rural Retreat recording...though, I must confess, the lady playing the carols can "bring tears to a glass eye"!