Steam Friction Fudge Factor

LAST EDITED ON Nov-16-04 AT 09:02 AM (EST)

I have added a "fudge factor" to my steam friction calculations. This is something that Joe Realmuto's excellent calculator does not account for. According to numerous references in my collection of books pertaining to locomotive engineering, construction and operation,when calculating steam locomotive friction (resistance) an additional factor must be added to the standard Davis formula.The Davis formula for locomotives is as follows, and is the basis for Joe's calculations ( and real world tonnage ratings ):R=1.3T + 29N + 0.03TV + 0.0024CAV²Where
R=Resistance in lbs
T=locomotive weight in (US 2000lbs) tons
N=Number of axles
V=Velocity in MPH
C=Coefficient of friction ( air resistance )
A=Front area of locomotive in square feet

Now, here is the missing bit, quoting Ralph P. Johnson in "The Steam Locomotive":

"In calculating total resistance for the steam locomotive it is customary to use the same values for journal, flange and air resistance as the electric locomotive, but to add a constant amount for all speeds of 20 lbs per ton of weight on the driving wheels to cover the internal machine friction of the locomotive between the cylinders and the driving axles. The total gives the resistance at the rail." (I can't cite the exact reference right now, but in a newer book in my collection it states that it is permissible to reduce this factor by 20% for engines with lightweight roller bearing rods ( N&W J Class and the last five A Class 2-6-6-4's 1238-1242 for example) and by 50% for those equipped with poppet valves and Franklin Rotary Cam valve gear (PRR T-1 and Q-2 Duplex locomotives for example ))

So the modified Davis calculation may be expressed as follows:R=1.3T + 29N + 0.03TV + 0.0024CAV² + 20DWhere D=the weight on driving wheels in tons.Now, how do I translate this into the sim, and what is the effect of all of this stuff?I want to use Joe's calculator and save myself the aggravation of hand calculating the Friction parameters for each engine so I came up with an idea. if I take the factor 20D, and divide it by 29, I come up with a number of "phantom axles" that when added to the total number of actual axles, will cause the calculator to add the additional resistance. I can bore you with the math, if anyone wants to see, or you can take my word for it.Level track numbers for Union Pacific Big Boy where T=381 N=12 V=5 C=1 A=120 20D=5400, and the resulting Friction parameters from Joe's calculator:Without the fudge factor: 907.65 lbs resistance:MSTS Friction Value:3750.8N/m/s -0.10 1.3mph 34.957N/m/s 1.631
Rolling Stock Type: Steam Freight Locomotive
Axles:12 Mass: 345.6 tonnes Frontal area: 11.15 m² Cd: 1.00
Davis=Fcalc speed: 70 mph
RMS Error(2 mph to 80 mph): 44.6760 N
Maximum Error(5 mph to 80 mph): 92.57 N @ 80 mph
English Davis Equation(V-mph, R-lbs): R = 843.22 + 11.4287 V + 0.288042 V²
Metric Davis Equation(V-m/s, R-N): R = 3750.81 + 113.7197 V + 6.411361 V²
With the fudge factor (186.21 phantom axles): 6307.65 lbs resistance:
MSTS Friction Value:27771.0N/m/s -0.10 1.7mph 11.096N/m/s 1.950
Rolling Stock Type: Steam Freight Locomotive
Axles:198 Mass: 345.6 tonnes Frontal area: 11.15 m² Cd: 1.00
Davis=Fcalc speed: 66 mph
RMS Error(2 mph to 80 mph): 247.2583 N
Maximum Error(5 mph to 80 mph): 506.06 N @ 6 mph
English Davis Equation(V-mph, R-lbs): R = 6243.16 + 11.4287 V + 0.288042 V²
Metric Davis Equation(V-m/s, R-N): R = 27770.96 + 113.7197 V + 6.411361 V²
As you can see, the there's a dramatic increase in the Friction parameters for the second calculation. (of course, the NALW Big Boy is in 2 parts in MSTS, so you would calculate each part using 1/2 of the "real world" values. )The result is that the locomotive now pulls the "real world" rated load up the proper grade at a proper speed ( a walk instead of a trot, 8-10 MPH vs. 15-17 MPH ).I have created 2 slopes in the sim, 1.14% to represent the Wahsatch grade out of Ogden and 1.55% to represent the "old line" over Sherman Hill, Westbound from Cheyenne.The real world adjusted tonnage ratings were 4450 for the 1.14% and 3250 for the 1.55%, both with a car factor of 5.The NALW Big Boy now performs prototypically on both grades with the rated load. It also handles a 3600 ton fruit block up the 1.14% at about 15 MPH as was the case in real life.These are adjusted tonnage ratings. If anyone needs an explanation as to what are "adjusted tons", "car factors" and how to calculate, ask and ye shall receive. I can also provide some tonnage rating info for certain engines on certain grades, and am in the process of calculating ratings for various engines on various routes in the sim.If you are not as obsessed as I am with prototypical behavior in the sim, then don't read this post. 8^)

Regards,
Nick

Alert | IP

Printer-friendly page | Edit | Reply | Reply With Quote | Top

RE: Steam Friction Fudge Factor, Yankee15, Nov-16-04, 08:35 AM, (1) RE: Steam Friction Fudge Factor, scottb613, Nov-16-04, 09:25 AM, (2) RE: Steam Friction Fudge Factor, Yankee15, Nov-16-04, 10:24 AM, (3) RE: Steam Friction Fudge Factor, Bill Hobbs, Nov-16-04, 09:15 PM, (4) RE: Steam Friction Fudge Factor, Yankee15, Nov-17-04, 00:26 AM, (5) RE: Steam Friction Fudge Factor, Bill Hobbs, Nov-18-04, 08:04 PM, (11) RE: Steam Friction Fudge Factor, crick14a, Nov-17-04, 07:43 AM, (6) RE: Steam Friction Fudge Factor, supergoods, Nov-17-04, 12:16 PM, (7) RE: Steam Friction Fudge Factor, Yankee15, Nov-17-04, 05:46 PM, (8) RE: Steam Friction Fudge Factor, Yankee15, Nov-17-04, 06:23 PM, (9) RE: Steam Friction Fudge Factor, pvoxland, Nov-18-04, 11:20 AM, (10)

Conferences | Topics | Previous Topic | Next Topic

LAST EDITED ON Nov-16-04 AT 08:36 AM (EST)

Oh, and another thing.The default values for horsepower for older DC motored diesels in the sim are incorrect. The SD40-2, for example, has an advertised power of 3000 HP, but only has 2533 HP available for traction. This is due to transmission losses and parasitic loads on the prime mover. I have changed the horsepower from 2238kw to 1890kw and the engine performs prototypically now. Take the nominal HP of the locomotive and multiply by .8445 to get the rail HP. Examples:SD40-2 2238kw x .8445 = 1890kw
SD45 2686kw x .8445 = 2268kw
F7 1119kw x .8445 = 945kw This does not apply to the latest power from EMD and GE since "truth in advertising" appears to have taken hold. Thus the AC4400CW for example actually delivers 4400 HP at the rail, it's prime mover producing a few hundred additional horsepower to compensate for parasitic loads and transmission losses.

Regards,
Nick

Alert | IP

Printer-friendly page | Edit | Reply | Reply With Quote | Top

Hi Nick...I am very interested in prototypical behavior on steam locomotives... I read your posts with great interest... While many of the calculations are a bit beyond what I am capable of doing for an engine file - I would like to see your work... I expend my efforts modeling and skinning... That is the great thing about MSTS - folks specialize in all the different aspects...Anyway - if you ever get around to distributing some engine files to test - I would gladly take a look at them... Regards,
Scott

Alert | IP

Printer-friendly page | Edit | Reply | Reply With Quote | Top

I expend my efforts modeling and skinning... That is the great thing about MSTS - folks specialize in all the different aspects...As my Dad always says, it takes all kinds of people to get things done right.I'll continue this later.Right now I'm playing God, creating mountains in the Route Editor.

Regards,
Nick

Alert | IP

Printer-friendly page | Edit | Reply | Reply With Quote | Top

Nick,
I'm not sure whether you are referring to Fcalc1 or Fcalc2.
I glanced through the documentation for Fcalc2 this evening because I thought I recalled it taking the 20t into consideration. If you look in the appendices at the end of the document, it does show the 20t addition to the weight for locos with friction bearings. It is added to the first term in the equation.
I've been using Fcalc for a couple of months now and really enjoy the added resistance on starting. But I've found that on my older steamers, I've had to reduce the value of the second term (the exponent for the starting resistance) from -.25 to -.2 to minimize the oscillation of the drivers. Doing this actually decreases the rate of drop of the resistance, but it does allow the locos to be more stable when stopped.

Bill Hobbs

Alert | IP

Printer-friendly page | Edit | Reply | Reply With Quote | Top

I'm using Fcalc2, but here's the deal Bill.I had been using Steam/Freight/Roller Bearings, because all the U.P. power had rollers on all axles. ( I assume that it is asking for axle, not rod bearings, since on all the other rolling stock calculations it refers to axle bearings ).But the Steam/Freight/Roller Bearings calculation result does not differ in any way from the Locomotive/Freight/Roller Bearings option which is meant for non-steam locomotives.So now, at your suggestion, I tried the Steam/Freight/Solid Bearings and the resulting factor comes out about 37% higher than it should be for the Big Boy. I won't even bother plugging those numbers in the .eng file, because I already know that will cause the engine to underperform, and probably stall with a consist it should handle. Also, 20t cannot be a determining factor in that equation since it does not ask you for the weight on drivers, or the number of drivers or anything that would identify the value of 20t. It would return the same values for a 2-8-2 and a 4-6-2 if the engine weights were the same, since they both have 6 axles, even though the value of 20t could be quite different for the 2 engines.I'll stick with my method, since it gives me the same result as my manual calculations based on all the data I've collected on the subject.

Regards,
Nick

Alert | IP

Printer-friendly page | Edit | Reply | Reply With Quote | Top

LAST EDITED ON Nov-18-04 AT 08:05 PM (EST)

Nick,
I wasn't suggesting that you use the solid bearing calculations instead, but just pointing out that that calc did make use of the 20T term. You are correct in pointing out that the value used by Fcalc 2 is not the same as Johnson's equation (page 200 of The Steam Locomotive, 1945 edition). So there are really two different work-arounds needed.
Since I have only one loco at present that has roller bearings, my primary concern is for getting as accurate an estimate as possible for solid bearing locos. Until a modified Fcalc 2 is available, my practice will be to simply use the weight on drivers instead of the total weight. I graphed the differences in the Davis equation properly calculated and calculated using weight on drivers instead of total weight. Since total weight contributes so little to the final equation relative to the impact of weight on drivers, the differences were insignificant to me even for ratios of weight on drivers to total weight as low as 70%.
A concern for many users who do not have good reference libraries will be in finding weight on drivers. I suggest the following as a workaround. If the tractive effort in lbf is known, multiply by 4 (the "standard" factor of adhesion) and divide by 2205 (to convert to metric tons).
Also, as noted in my first posting, I've had trouble using the -.25 exponent and have settled for working with -.2 to reduce the oscillations of the loco when brakes are holding the loco stopped.
Nick, let me suggest you modify your recommendation for making the N term pick up the additional weight simply using .7 x weight on drivers + number of axles or .7 x (TE x 4 / 2205) + number of axles to compute the total number of axles needed to compensate for the missing term in the roller bearing equation.

Bill Hobbs

Alert | IP

Printer-friendly page | Edit | Reply | Reply With Quote | Top

This very same thing was being discussed a couple or so months ago on the UK forum and has been put into practice, I believe that Ian ( Supergoods ) has been in discussion about it with Joe.The idea came about to see if the new friction calculation would have an effect on water usage and it does the figures when used increase water usage, mind you the rest of the file calculations have to be correct also as other wise it is pointless.Regards, Rick

Alert | IP

Printer-friendly page | Edit | Reply | Reply With Quote | Top

I'll post something on this later, unfortunately my daughter crashed the hard drive doen to the Bios so I'm having trouble accessing data at the moment.

Supergoods

Alert | IP

Printer-friendly page | Edit | Reply | Reply With Quote | Top

LAST EDITED ON Nov-17-04 AT 05:54 PM (EST)

It's not a real-time solution for the water and coal usage issue, but I set all my tender weights to represent 2/3 loaded with fuel and water. Most steam tonnage rating calculations that I have seen seem to use this as an accepted standard, That is why on most railroad mechanical department locomotive diagrams you will see the weight of engine and tender given with the disclaimer "2/3 loaded with fuel and water". U.P diagrams show both weights as well as the empty weights of both engine and tender. Based on that I have created a "Big Boy dead-in-tow" engine that my diesels can tow from one fictitious display site to another. I disabled all the locomotive parameters and deleted the lights on both engine parts and the tender, and adjusted the weights, friction and braking effort. Looks cool sandwiched between two empty flats behind a DDA40X on Cajon pass. My next project is to learn how to upload a screenshot to the forum.(now if I can only figure out how to cut the piston rods....hmmm..............)BTW, the empty figures are 697300 for the engine and 171300 for the tender. 434.3 tons vs. 594.75 tons fully loaded.( Yup, it lugged two heavyweight pullmans worth of weight in fuel, water and sand, good for a little over two hours worth of maximum output. Amazing.)PS--if anyone wants to make a D-I-T model be sure and protect the original .eng files. I have enough trouble as it is, without new people cursing me.

Regards,
Nick

Alert | IP

Printer-friendly page | Edit | Reply | Reply With Quote | Top

OK, sorry Rick. I reread your message and realize that you were talking about water use by the locomotive, not a way to create the effect of diminishing weight in the tender as the engine consumes the coal and water supplies.There have been other posts on that topic and I got sidetracked (pardon the pun).

Regards,
Nick

Alert | IP

Printer-friendly page | Edit | Reply | Reply With Quote | Top

LAST EDITED ON Nov-18-04 AT 11:21 AM (EST)

Not too long ago I asked a few questions here about setting .eng file values for a Mikado steam engine and especially about performance of a prototypical consist, as far as tonnage, on a gradient. The answers were thoughtful and helpful.I was a little intrigued by the idea of calibrating those parameters to known tonnage ratings for a locomotive and consist and a protypical route and grade. That is, after coming up with the "best probable parameters" for engine and cars how does the performance within the simulator compare with actual operations?I also asked about starting a prototypical consist with the maximum tonnage rating on the steepest gradient of a prototypical route.With that last requirement (starting on a gradient), I found I had to increase modestly the "3rd" mystery value from the FCALC 2.0 computations. It becomes quite a struggle for the locomotive to get going ("accelerate" seems to overstate the performance), but it does happen.I appreciated every comment my questions brought out before. I still wonder whether it is possible to say (for example) that MSTS, given our best estimates of proper .eng and .wag settings) gives close performance to tonnage and gradient real world performance. Or does the simulator "overperform" or "underperform"?

--- Phil Voxland

Alert | IP

[ Home | MSTS Forum | MSTS FAQ | How To... | File Library ] Copyright © 2004 by Nels Anderson. All Rights Reserved.