Resistance- is futile

[Scottl]

The TLTR version is I'm a dumb-

I've been having fun running trains this past week but one train was acting up.  It is my favorite- a grain train with 27 cars of mixed origin, all weathered and pretty much prototypical for my era.  The power was a pair of Kato 6-axle units operated by TCS DCC decoders.  The problem was two-fold:

1.  A number of cars would derail, often on good track in curves.  Some would derail repeatedly on a trip around the layout.  This was all specific to this train.  The track was smooth and level and no other trains have these issues.
2.  The locomotives could barely pull the train, especially when going trough some curvature.  The lead locomotive was spinning its wheels and the entire train would stall on a flat grade.

Prior to this week, I had papered the second problem over with a third locomotive, but now that I have all my trains operating, I don't have one to spare.  Enough was enough, it was time to figure out what was going on. 

Based on the coupler slack between the two locomotives, they were obviously not working well together.  Also, the train lacked the obvious slinky effect that we know so well from MT couplers.  The solution lay in these two observations.

First, I made an effort to speed-match the locomotives. Given they were identical models and decoders, I was surprised it took quite a lot of CV tweaking to make them work well together. 

That definitely improved things, but the derailments continued with the train and the locomotives were still struggling with the load.

Next, I decided that I needed to test the rolling resistance of the train cars.  I set up 2 m of flex track with a 4 cm rise on one end.  I recorded the distance rolled by each car.  The results were pretty amazing: individual cars rolled between 38 and 183 cm.  Some even needed a little nudge to get moving downslope on the test track. 

Clearly some of the cars were dragging.  With my micrometer I measured the axle length of the worst offenders and found I had equipped many MT trucks with 0.553" axle wheel sets.  I replaced these with 0.540" wheelsets and found this decreased the rolling resistance dramatically.  I didn't have enough wheelsets to correct all of the cars, but I fixed the 6 worst offenders.  I now have a priority list to replace others as I get access to new wheel sets.

After assembling the train again, it was no surprise that the train not only travelled easily under the power of the locomotives, but also, the number of derailments dropped to only a few associated with two cars. 

Long story short, details that didn't matter on my previous layout with shorter trains matter a lot now that I am running 12' trains.  If you've read this and are thinking, "well, duh", I've had much the same line of thought as I discovered what was going on.

As a final note, I still don't have any coupler slinky in my train.  I have inadvertently introduced the equivalent of MT retarder springs into my train.  All it takes is putting the wrong wheelsets into a trailing car.

[Rasputen]

You didn't mention the brand of wheels you are using, but I remember on some NWSL wheelsets, the end of the axle point was not formed correctly.  This would cause a lot of drag too.  I had to get a 10x eye loupe to inspect each axle end.

[Maletrain]

I have been doing something similar over the last couple of weeks, building a 37 car + caboose train that will be a B&O "stone train" as soon as I can get 37 limestone loads made. 

I checked each car before adding it to the train.  They are all Atlas 55 ton double hopper cars in 1950 Western Maryland paint - per the prototype pictures I have of this train on the Pittsburgh Division.  Most of the cars are second hand, because Atlas has not been making them for a while.  Some had quite a bit of "modifications" by the time I acquired them.  Some had Micro Trains trucks substituted for the Atlas trucks, and several had incorrect axle length wheel sets - both MTL 0.540" axle wheels easily falling out of Atlas trucks (which drag when not actually out), and 0.553" axles in MTL trucks, which don't fall out but do drag. And, there were some metal wheels with flanges that would cut pizzas.  And one even had Rapido couplers.

I put the metal wheels I had available in the last 10 cars to help guard against string-lining the train.  There are 5 cars of ESM 0.553" axle wheel sets next to the caboose, with 5 cars with 0.553" Fox Valley wheel sets ahead of the ESMs.  My strategy is that the ESMs roll better than any of the others, and the Fox Valley wheelsets are marginally better than the OEM Atlas plastic wheels in these cars.  The replaced OEM wheels in the last 10 cars gave me enough spares to replace all of the other non-OEM things in the other 27 cars.  Except, I had to order some sets of Atlas 50 ton friction bearing freight trucks with knuckle couplers to replace the MTL trucks and the Atlas set with Rapido couplers.

Results were better than expected.  I had been pulling this train around a flat loop with 15" radius end curves using 4 DC ER Sharknose units that I am "running in" before conversion to DCC.  My first thinking was to make sure this 10' long train did not string-line on the 15" 180-degree curves.  I does not.

But, before I replaced the MTL trucks with the MTL couplers, I did have uncoupling by the MTL cars - which is why I replaced those trucks.  Apparently, there was enough slinky in just 1 MTL truck set to uncouple the train at that car - without the train making it around the loop even once.

After I had the train reliably circling the loop, I decided to substitute the intended loco for the finished train - a ConCor USRA Heavy "Santa Fe" 2-10-2 that needs a tender replacement and some serious boiler bashing to represent a B&O "Big Six" 2-10-2..  That pulled these 38 sets of trucks reliably, with no slippage.  I still need to think about performance on grades, so more ESM trucks are in the budget. 

It is nice to see such a long train snaking around the curves with no slinky effects.  I just wish the OEM Atlas couplers provided closer coupling. Especially with the Sharks now spaced at 30 scale inches, the roughly 60 scale inch spacing between the hopper bodies really sticks out.

Still waiting to see what Npossible couplers can do for that.

[djconway]

Rolling resistance is one of the hardest things to get to a minimum. 
Micro-Trains (MT) trucks with MT low profile (the old ones) were about the best stock trucks.
1- Look for burs on the axel points.
2- Measure all axels, a little (+/- 0.0005") seems ok, I haven't figured out how to fix long or shorts. But sort them and keep them sized.
3- Not all truck frames are square to the bolster or parallel, use the long/short axels from step 2 as needed. you are talking 1/1000" differences here.
4- Debur the bolster pin hole

Micro-Trains -- I'll ask once again Please go back to your pre 2010 Low Profile wheel configuration, the new Medium Profile wheels tend to pick switch points far more often. (the flange looks almost square, not tapered)                                               

[ednadolski]

As a final note, I still don't have any coupler slinky in my train.  I have inadvertently introduced the equivalent of MT retarder springs into my train.  All it takes is putting the wrong wheelsets into a trailing car.

Sometimes tho you need a little rolling resistance, like when spotting a car.  Low rolling resistance could allow the car to move off-spot (this can happen in any scale -- some of my P:48 trucks will roll several feet, on what looks like a flat section of track).  It can also be harder to couple, when the car wants to bounce away from the loco.

I think of excess drag as basically like trying to move cars with the brakes on.  I'd rather not have to back a long train up-grade and thru sharp curves, if it had too many cars with drag springs installed.

Ed

[Maletrain]

I agree that Micro Trains trucks and plastic wheel sets roll very well.  And, they are very consistent.

Replacement wheel sets may not be as consistent in axle length, and can also have gauge and wobble issues.  And, if they don't come with issues, you can create some yourself by the way you handle them getting them in and out of trucks. 

I always check gauge and wobble on metal wheel sets, and am careful to put side pressure on only the wheel that that does NOT have the insulator between it and the axle when I push it into the truck.

So far, I have not found any problems with ESM wheels (and am not expecting any).  They roll more freely than any others I have tried.  I think that may be due to them having a sharper angle on the conical axle ends than others I have looked at with a 10X loupe.  Not sure if that would have any realistic effect on the wear of the bearing surface in the trucks.  So far, I am not seeing any increase in rolling resistance as I drag my test train around the loop for several hours - and I am watching for it. 

[Maletrain]

Sometimes tho you need a little rolling resistance, like when spotting a car.  Low rolling resistance could allow the car to move off-spot (this can happen in any scale -- some of my P:48 trucks will roll several feet, on what looks like a flat section of track).  It can also be harder to couple, when the car wants to bounce away from the loco.

When doing ops, I tend to use the end of my uncoupling pick as a brake for coupling to single or a few cars.  I just put the end down on the roadbed on the opposite end of the car from the loco, stopping the car from rolling away when the loco hits the coupler.

Another issue with free rolling cars is that they can roll off spot due to gravity on "looks level to me" track, or even a fan breeze. I have some modified miniature clothespins that clamp on the rail to act as breaks when I need to leave a car where it wants to roll.

[ Guests cannot view attachments ]

The red one shows the necessary modification to enable gripping a rail.  Amazon has a variety of different sizes that can work with various scales.