Over on the "A" forum there's been the usual debate about how many cars a six-axle locomotive can pull up a given grade. I've had my own debates over the years, and it all started with a brand-new Atlas SD24 that was a complete wimp in comparison to any other six-axle unit I'd ever seen. The tools and techniques to figure that out (including my dynamometer car tool - a spring scale on trucks - evolved over time. Then I decided to retire my Kato GP38 and replace it with new Atlas and that was a dismal failure - the new GP couldn't handle the assignment, further raising my suspicions of light-frame, DCC cutout, slippery-wheeled power. So I redid all my data point samples again - and added a couple wrinkles. I weighed everything and calculated a factor of adhesion (suspecting the slippery wheels theory on new stuff). http://gustafson.home.westpa.net/N Scale tractive effort data research.pdf A couple things here really shocked me that I actually didn't know. The slippery wheels thing on new production is real, and everybody knows they are lighter frames by a lot. You may not realize how much that impacts at the drawbar though. What surprised me, and I'm still going to add more data here, is that the wheel wear and age is still a huge factor. Back in the 80's I was surprised that my old Trix U28's outpulled Kato units, but those Trix units were worn right down to the brass on the wheels - no plating left. My latest discovery is that a couple locomotives that were marginal now have crept up - my oldest Atlas SD50 mech (in my FP45) now has virtually the same TE as my good Katos, and my most-used Atlas Classic GP7 is in the same league. What's more, my 'new' old kato SD40 in the SD45-2 shell....is slippery in comparison. Close examination of the wheels would tend to indicate that this wheel wear through the original factory surface is behind this. The 'new' old Kato never got much use (got it used) and it hasn't found its feet yet. When we all start throwing out opinions and the stuff is all over the place - this is one more thing to consider here. I knew the 'new stuff' couldn't pull worth a darn, but seeing it IMPROVE sure made me go back and recheck stuff. Just adds one more factor to the 'so what will an SD-XX pull?' and you'll get way too much information here. If you look at my data (so far) the thing that 'ought' to raise your eyebrows is that the variances here between similar models in theoretical identical prototype size can be THREE TIMES the tractive effort or more; between a Kato GP38 and a new Atlas GP38, and between a worn-down Atlas Classic GP and a new, out of the box Atlas GP7. This has nothing to do with loco size, axles, anything except total weight, wheel material, and adhesion factors. I just never imagined that the spread could be this wide. I have lots more locos to load into this, more to come. I'm using a pretty high-precision scale for weight and a pretty crude scale for pull, so if you think the pull numbers are 'off' they may be, best thing I can say is that pay the most attention to the data in comparison. You may run the same power and get different pull, but the weights should be very close here.
I agree on the "broken in" part. Wanna really test this theory? Rig up a "break-in" track; tether a loco into place on a piece of sectional track, and let it run slipping. Pretty quickly, that should wear the wheels down through the plating. I'm of the mindset that more engines is always better, but then I have been known to pull three cars around with an ABBA set of FT's. So; don't listen to me! Still, it goes to show- there's merit to the thought that there shouldn't be a loco made without traction tires. I know they're not for everyone, but when done right? A little old 2-6-0 will pull the track up behind it! We don't have the benefit of HO guys with their 2lb. F7's.....
You know, on this topic I have been checking grades on my prospective layout with my Unitrack. My Spectrum USRA H5 2-6-6-2 was having trouble on the club layout pulling even 10 cars up the 2.5% grade. It was all I had with me at the time. I was concerned because I had to have a 3.6% grade on my home layout for it to work. So home I went and set up a 4% grade with a 15" radius curve. The poor H5 couldn't move 4 cars through it (time for some Bullfrog Snot and maybe some lead if I can find a place to stick it), but my Atlas/Kato Gp30 pulled 12 cars up it no problem. So did my Rivarossi 2-8-8-2. My Spectrum 2-8-4 Heavy Mountain (cast body) did fine too. I can only guess it's the weight? Though it seems like the H5 isn't any lighter than the GP30, but there are more wheels distributing weight. Maybe it is the plating? I'll have to run it for a while and see if it improves.
I don't have that much steam - but I do like to put that Kato/GHQ L1 in there just for the 'holy crap!' comparison factor. The other thing that is fun is to put my ABBA set of Kato F's together and watch them pull the scale out all the way to max... And my 'lastest version' Climax with the cast metal boiler, well, with the single traction tire it now climbs off the rail and derails as the adhesion is sufficient to convert it into vertical lift on the truck, it won't really slip, so that's as much as it pulls before it lifts the front axle off. I absolutely reject the idea that the solution is just to add more locomotives - that rewards the manufacturers of poor design. And the 'perfect storm' of light frames, DCC cutouts, and slippery new wheels is really a lot worse than what you would think. My opinion (not completely backed up by data) is that Kato is way ahead of Atlas on frame density and wheel material - the NW2 mech does comparitively well and the cab isn't even full of weight, AND it has a DCC compatible frame. Meanwhile, a right-out-of-the-box new MP15 has less than half the tractive effort for essentially the same chassis weight.
Great infor Randy. Look forward to seeing more. Does this mean I should send my Climax back to you for a tune-up? Jim:tb-biggrin:
Oh I'm not one to "throw locos at the problem" when traction is light..... i just really, really, really like locomotives! LOL
i do sand the wheels on all my locos before they hit the track the very first time. this improves tractive effort by as much as 40%. electric contact is improved as well. there's a coal train on my layout with 80 really heavy huberts coal cars. heavy means approx 44 grams each. this train is easily pulled by two kato es44ac. they both have sanded wheels. a short test showed it would use 3-4 brand new kato es44ac out of the box to pull the same train.
Since about the middle of last year I have been expirimenting with casting the boiler out of metal instead of resin. I'm finally to the point in modifying the molds and process where the metal is predictable and reliable and I offer the metal boiler instead of the resin one as standard equipment. But it's taken a long time to get there. I've gone from scrapping and remelting 6 of 7 to maybe one of four.... If you haven't built the kit, PM me and we'll talk.
Thanks for the analyses. It is very good to know and be aware of. While I love the analyses I'm a "add more locos", (hmm, or is that "ADD more loco"?), guy. My usual consist is two AC4400s or SD40-2s, (under my FP45 shells), up front and one Atlas light frame geep in the middle of about twenty 65' grain cars. If the lead stalls the geep will either nudge them along or stall itself. It also relieves tight curve close lines. The main issue is when trying to back through turnouts into a yard. That can get messy as the two, (now rear), units just keep on shoving if the geep stalls. oooops. Note: this is with DC using one power supply.
Well, I admit to doing that - on the Atlas SD24. Out of pure frustration. I literally took it out of the box, put it on the point of about 10 cars, and started the run-in exercise. It was going just fine. Left the room and came back.. and it was now stalled on the hill, spinning and couldn't climb it, where it was doing the same train a few minutes earlier. Once the 'roughened' blackening was worn smooth, it lost tractive effort. Scratched the wheel - there was brass under there, somewhere. So, under power and flipped upside down, I actually filed the wheel treads with a narrow jewelers file to try to get through that slick surface. It worked, actually, sorta, and wasn't any worse for (literally) wear. Added more weight to the area I pulled the rear headlight out of. Since then, all I have done is run it. It increased the TE by about 25% with those stunts. I'll be adding that unit to the spreadsheet.
If casting brass turns out to be too much, try something I've used in plastic-bodied locomotives--it would work for this without the epoxy or putty I've used. Use less resin and more buckshot.
How about I send you a few Kato SD70/80/90 Macs and you will get another set of 6 axle locomotives to try.....
About a year ago I set out to do some experimentation on how to add tractive effort to my locomotives. I have a helix with 17" radius and 2.5% grade and also an area of my layout with a very long climb that ends with a brief portion of 3% that is the place where my trains will most likely stall. Most of my 4 axle locomotives would pull only 5 to 6 cars up the 3% area. One evening I set an Atlas B40-8W up with 5 identical coal hoppers and ran the train up the grade. It only took one added car to stall the train. As the locomotive spun it's wheels I began carefully placing 00 buckshot pellets on the top of the loco. Two pellets got the train moving. Two more and I could add another car. So it seemed to me that this proved that more weight causes more downforce which equals more tractive effort. So how can this theory be applied in a practical way? Well, I went on-line and ordered lead sheets in .020" and .040" thicknesses and also tapped more from my shotgun reloading stock. I found that weight can be added to typical N scale engines by cutting strips of lead and either gluing or taping them to many areas inside a loco shell or onto open spaces on the frames. My first victim was the Atlas B40-8W. It's out-of-the-box weight was 84 gr. I found space to add 16 gr. for a total of 100gr. This loco can now pull 7 coal hoppers up my grade slipping slightly but not stalling out. Another loco I did was an Atlas MP15DC. It was a pathetic puller at 51gr. stock weight. I was able to creatively sneak 12 gr. into that little shell which allowed it to pull 6 cars up the grade instead of 4. Here you can see the loco on the scale dis-assembled. This is an example of some of the cavities and places in the sill or walkway where lead sheet can be cut and glued. Every little bit helps ad more weight Here weight is added to the inside of the shell and on top of the decoder. I used yellow tape as an insulator. Found more space inside the cab. Also used some lead sinkers which fit the space perfectly.
Lastly, some weight behind the pilots. Here is the assembled loco. As you can see, it has gained some weight. 12gr. to be exact. And finally, here is the loco in action plodding up the grade with 6 cars in tow. It's truly now, the little engine that could! The MP15 is the extreme example of how adding weight can help traction. Other locos such as GPs and B40s are easier to find areas to add weight. I do have more photos of a Life-Like GP60 that got the weight treatmentand benifited from it greatly. Adding weight to locos creates better adhesion which not only extends tractive effort, but also promotes better electrical contact in lighter locos such as the MP15DC.
My solution was to keep my grades down to 1%, I run 30-40 car trains with two powered locos and even my lightest locos will get the job done!
That's pretty impressive with the MP15. And at least I know our digital scales agree. That's really hard to get THAT MUCH more piled in anything. Back about....2002 or 2003 I had a tour of the Huntington, WV shops on CSX. They were actively putting big 'ol chunks of steel on the walkways of selected GE units, ballasting them up to maximum axle loadings. Pretty much the same idea. If you looked on the walkways it was visible- like about another 6" in height. One of the things I'm trying to get people to accept isn't only the weight idea, its the adhesion factor and just how wide that really is. This is the first time I've really been able to get some meat behind the numbers and make any sense out of it. While I've now seen an Atlas Classic 'reform' itself due to heavy use, I have yet to see similar progress out of anything in the 'slow speed motor' era do the same. I just don't think your average modeler imagines that wheel material can impact tractive effort this much even if the basic frame weight thing is intuitive. Or that Atlas and Kato six-axle units of similar physical size are that far apart. I'm still researching it.
When I built my layout, it was obvious that it was going to have some tuff grades. The layout was taken from a John Armstrong plan which was HO scale. I just used N scale track and reduced the grades a bit. A part of the layout I left out was a turning wye meant for helper locos. That point missed me completely until one day I came across the old John Armstrong book I got the plan from. Duh! Why not try running a helper set? So I did. Talk about realism. DCC makes it so simple and it adds a whole new aspect to getting trains around the layout. It is also a blast for my son and I to do as a father-son activity. Most 6 axle locos in pairs will do the job pulling about 16 to 20 cars. Now we can do 30 car trains with no problems and the added realism of that helper unit on the rear.
