Anybody else RUNNING new Atlas 89' flats?

[nkalanaga]

Yeah, but when I need a washer, not a common event, I need it NOW!  If I needed a bunch of them, and wasn't in a hurry, buying them would probably be the best idea, and certainly easier.

[randgust]

OK, well, it finally hit the top of the tinkering pile again.    The good news is that the big, thin, nylon washers that come with MT passenger car trucks are a perfect fit for a test-raise of these cars without a real modification to the bolsters.  And I got all the flash ground off the bottom of the car so that it doesn't grind at all anymore on the wheel flanges.

The metal wheels seem to do just fine, surprisingly.  I also tested medium-flange MT wheels, those worked just as well.  No derailments.

But holy smokes, now that I can run-test them in an actual train, those body-mounts combined with the short-shank coupler give me the worst case of stringlining I've seen yet on my 13" min radius curves on just under 2%.    Metal wheels work a little better, but it's still lifting the outside wheels right off the rails on the curve; that's on a 14-car train  with this car 3 cars deep.   I haven't had this problem with either the BLMA or the Trainworkx car, either.   I 'think' its the body-mount coupler position, it's further toward the end than anything I've seen before and has much less swing.

I really can't afford derailing this train through hidden trackage and dumping cars and trailers 52" off the floor.  So before I do any more damage to it, back in the box it goes and I'll sell this one.  Nice looking cars, but not for me.   I've tested these cars a lot before I posted this, you can debate it if you want, I'm just saying they don't work on my layout when every other 89' piggyback I've ever bought from Trix, Con-Cor, MT, Alan Curtis, TrainWorx and BLMA is operating just fine in the same train.

Well, maybe one more tinker.  I've got piles and piles of MT 1016's with a longer shank, may as well give it a shot.  I'm about as stubborn as they come.

[tehachapifan]

My experience has been hit-and-miss with body-mount couplers on just about any length car that are mixed in with those with truck mount couplers...and I have 20" min curves. I have to run all my 89' flats from various manufacturers with body mounts nearer to the rear of the train to reduce the chance of them getting pulled off the rail to the inside of the curve by the truck-mount couplers. I should mention I have fairly steep grades approaching 2 to 2 1/2% and run long trains which, combines with small flanged wheels and the mismatched coupler thing, increases the stringline forces.

[nkalanaga]

This is one place where one almost HAS to use either all body mounts or all truck mounts.  A truck mounted coupler is basically a very short car, think "ore car".  Even the prototype can't run them coupled to an 89 ft flat through most curves.

Randgust:  The long shank couplers may help.   I have almost as great a variety of cars as you, missing only the Alan Curtis and Atlas, and even my old Arnold car with 1025 couplers will run well with other body mounts.  My Con-Cor's have been lowered and have MT offset-shank couplers body mounted, the MT cars have the swinging offset-shanks trimmed and body mounted, and the BLMA/TW cars have whatever came on them (1015s?)  I have one 16 inch radius curve and they all go through that, but not coupled to truck mounted couplers on anything, or body mounts on short cars.  I try to limit adjacent cars to 50 ft or longer.

[jagged ben]

...

Well, maybe one more tinker.  I've got piles and piles of MT 1016's with a longer shank, may as well give it a shot.  I'm about as stubborn as they come.

Yeah man, do it for me!   

If the BLMA can do it with those, then I think the Atlas should be able to as well.
Glad I don't have to deal with 13" curves.

[randgust]

What's wild about watching this was that it wasn't always derailing, but it was always lifting the car right off the wheel treads on the outside rail, running it on the inside wheel treads only, then it would slowly set back down on the outside rail again...yikes....  And that's a basically heavy car, too.

[GaryHinshaw]

I don't have any Atlas flats, but I have been hit by string-lining with the BLMA flats on 2.3% grades with 18" curves.  I think I understand the physics of it now, but I'm not sure I have a helpful solution to offer. 

For a car with a wheelbase and overhang like a typical 89' flat (5" wheelbase and 1" overhang from bolster to coupler), the body-mounted car will experience a roughly 40% higher string-lining torque than the same car with truck-mounted couplers.  If the body-mounted car is string-lining, the only practical options are to increase the mass of the car, or decrease the drawbar force on the car by shortening the weight (length) of the train behind it.  (You could also try a truck with a lower bolster - see below, or convert to truck-mounts.  Of course increasing curve radii also helps, but that is usually not practical.)

-Gary

P.S. Here is a sketch of the essential geometry.  The basic issue is that the torque applied to the car by the drawbar force (red arrow for body-mounts, green arrow for truck mounts) is larger because the body-mount force is applied at a larger angle to the car's center-line than for the truck-mounted force.  For an 89' flat, with its large overhang, the ratio is very close to 40%, independent of curve radius.



Here are the terms:
τ_w - positive torque due to weight of car
w  - weight of car
g  - wheel gauge (9 mm)

τ_d - negative torque due to drawbar force
h  - height of truck bolster (point at which force is applied to the truck)
F  - drawbar force (basically the weight of the train behind the car, de-rated by the grade of the track)

I think the BLMA flats are more forgiving than other flats because of the lower truck bolster height, h, so you might also try swapping trucks to get a lower bolster.

[bbussey]

The bolster on these Atlas trucks is the same as the bolster height on the BLMA trucks, so swapping in BLMA trucks will make no difference.

Since it isn't practical to install truck mounted couplers on this model, and since the cars are diecast already, it seems the only option in this case would be to lessen the load behind them (shorter train) or move the cars toward the rear of the consist.

[GaryHinshaw]

The bolster on these Atlas trucks is the same as the bolster height on the BLMA trucks, so swapping in BLMA trucks will make no difference.

Good to know Bryan, but too bad it's not a possible solution. 

Randgust, do your cars string-line of they're loaded?  That's an easy way to increase weight.  Another option is to add pushers, to decrease F.   

[nkalanaga]

I'm with Bryan on the stringlining issue, and I've seen cars ride on the inside wheels without derailing, just as Randgust reported.  The only solutions are wider curves or less trailing tonnage.  Even the prototype has issues with this, although it's rare on mainlines, simply because the curves are much wider.  It can be a problem on sharply curved industrial tracks.  Most of our curves are sharp by industrial standards...

One possible partial solution is to check the wheels on the cars behind the problem.  It isn't the same issue, but I found that it was much easier to push MT 2-bay hoppers, with truck mounted couplers, through #4 crossovers if they had MT lo-pro wheels.  Everything else worked the same with MT or Precision Masters wheels, but the hoppers derailed being backed.  Whether it was just enough less rolling resistance, or a slicker plastic in the flanges, I don't know.  You might find that some of the cars roll better if the wheels are changed, which would reduce the load a little.  If so, the current wheels could be used on non-intermodal cars, where the drag wouldn't cause any problems, saving a little bit of money.

[bbussey]

Randgust, do your cars string-line of they're loaded?  That's an easy way to increase weight...

The only issue with adding weighted trailer loads is that you'd shift the center of gravity on the car, which could make the situation worse.

The BLMA cars come equipped with #1016 couplers.  Is it possible that extra amount of coupler pivot could make a difference on the Atlas cars?

If I was encountering this problem though, I would treat it as an operational wrinkle.  Same as not putting tankcars carrying explosive cargo directly behind the locomotives or directly in front of the caboose.  All TOFC cars have to be on the rear of a consist, and if necessary restrict how many can be in a consist.

[ljudice]

I found unweighted trailer loads on the BLMA flats improved their reliability 100% over empties.

Next layout - no grades....

[randgust]

Your geometry analysis is right on, and agrees with field results I've seen that fly against the common wisdom that body-mounts perform better than truck mounts.   Somewhere around 18" curves and 20 cars the equation flops over, as well as the cars.

I did the experiment with putting in 1016's.  It's a kludge.  While they fit the box, without an MT box around them there is no centering action for the coupler.   If I wanted to spend a couple days hacking out the box and putting in a 1016 box, it 'might' work, or I might simply ruin the car.

I did experiment with grinding out the inside of the box a bit to get more coupler swing, that made no appreciable difference.

Mind you, I'm an Atlas fan for years, but they get a D- for this one between the quality issues around the injector marks, flange scraping, the interference of the uncoupling rods with the coupler swing, poor coupler swing design, and basic car design that really makes it a 15" radius or above, so it was mis-advertised as well for what curves it can handle.   

I'm not giving these cars extraordinary treatment when any other 89' I've got is a free-runner without restrictions.  Not worth it.   $25 a car, $12 a trailer,  dumping a train on the floor from a derailment is more than I want to add up from damage when nothing else does this.   Back in the box it goes.

The only good thing about this car is that it has freed up a bunch of MT TTX flats on the market at somewhat lower prices, and those run flawlessly on my layout, replacing the surviving Con-Cor flats that are being retired.

[tehachapifan]

...Next layout - no grades....

Amen to that!