Tehachapi, BC

[davefoxx]

Of all the Tehachapi-inspired layouts out there, my gut says this is destined to become the Tehachapi layout by which all others will be measured.

That gets an enthusiastic +1 from me!

[AlkemScaleModels]

Are you including the LaMesa Model RR club version in this calculation?

[John]

Are you including the LaMesa Model RR club version in this calculation?

LOL

[GaryHinshaw]

I appreciate the vote of confidence David and Dave.  I really don't aspire to compete on the world stage of model railroading; I just want this pike to live up to my own expectations.  If it inspires some people in the process, so much the better.

It's been a bit of a dry spell due to work & travel, but I did get back in the garage this weekend and have a bit of news to report.  I now have the main line through the Loop resting in place and so far it passes at least one of my reliability tests: a lightweight BLMA 70T truck will roll down the Loop at a pretty good clip without derailing.  Next up is to wire the track and run some trains.  Photos to follow.

In the meantime I have continued to test some trains in the Cameron/Monolith area and have encountered my first problem.  I have been very concerned about trains string-lining, and have delayed gluing track down until I have enough experience to know that things will be reliable.  The curve between Cameron and Monolith is 18" on a 2.2% grade, and it currently has no super-elevation.  Here is my first encounter with string-lining in this curve:



It's a new BLMA 89' flat at the head end of a 22 car train.  FWIW, an MT 89' flat in the same position in the same train shows no tendency to string-line (same for a Red Caboose auto-rack).  Why is that??  The cars have almost identical mass, length, wheel-base, etc.  As much as I hate to admit it, my theory is body-mount couplers.

Let's think about the physics for a minute: string-lining arises when the torque about the axis that intercepts the inner wheels is too large for the weight of the car to overcome.  In this case, the relevant torque from the drawbar force is proportional to h sin_theta, where h is the height above the rails at which the drawbar force is imparted to the body of the car, and theta is the angle between the midpoint of the car and the point that the drawbar force is exerted (as measured from the center of the curve).  The sin_theta factor is similar for the two flats, so that mostly drops out of the comparison.  But the height h is different: in the body mount case, the drawbar force is exerted on the body by the coupler mounting screw at the height the screw meets the body.  In the truck-mounted case, the force is exerted through the bolster pin at the height the pin meets the body. 

The height h, and thus the string-lining torque, can be as much as 50% larger in the body-mount case.  As an illustration, just compare the bolster height on the truck-mounted spine car (where the coupler box attaches) to the top of the coupler shaft on the BLMA flat:



Of course this only matters if there is enough torque to overcome the car mass, but in my case there is, much to my consternation.

Now, I'm not trying to bash body-mounts or BLMA here, I'm just noting my experience and pointing out that in some cases, the physics of truck-mounted couplers put less stress on the equipment than body mounts do.  I'm not sure where this is headed, but that's another topic.

Cheers,
Gary

[ednadolski]

It looks like you've got a car with a pretty long overhang coupled to a relatively short one.  The autorack has lot more swing in the coupler, so that may 'absorb' the effect of the overhang.

So it's more of a geometry matter than body vs. truck per se.  Looking from the top, how much does the coupler on the BLMA flat want to swing out over the outer rail?

Ed

[Leggy]

Lets also not forget that body mounted couplers are generally regarded as the better option as they put less force on the bogies and work better with longer trains.

[Sokramiketes]

Do you have any other cars that length with body mounts that arn't BLMA?  Club members have had that same problem with the BLMA spines and Trinity reefers.  I suspect there's something with the articulation (or lack thereof) with the trucks causing an issue as with other BLMA cars. 

Usually the MTL 89' flats are derailment prone too, because the funky-sorta-talgo coupler bar keeps their trucks from articulating well also.  Nate took all those arms off his Circus train flats and body mounted the couplers and they run well.  So I don't think it's just a body mount issue. 

[davefoxx]

The curve between Cameron and Monolith is 18" on a 2.2% grade, and it currently has no super-elevation.  Here is my first encounter with string-lining in this curve:

You know, if you DID have superelevation on this curve, the BLMA flat would then have all four wheels on the rails.  There.  Problem solved.   

DFF

[GaryHinshaw]

I do need to do some more testing, but I can make a few comments about the physics:

1. It is absolutely clear that in order to string-line, you need sufficient torque about the inner wheel contact points to overcome the weight of the car.  This comes from the transverse force applied by the drawbar pull on a curve.

2. There are a few factors that enter into the transverse force (such as car and curve geometry and the overall drawbar force), but the torque due to this force is directly proportional to the height above the rails at which it is applied to the body of the car; i.e., the higher you push on something, the easier it is to tip it over.

(One clarification to my previous post: if the body-mount coupler post is rigid, the effective lever arm is the center of the coupler shaft not the top of it.  But it's still higher than the truck-mounted lever arm.)

In the comparison between the MT and BLMA flats, it's likely that the transverse forces on the drawbars are different between the two due to the angle and overhang of the coupler (i.e. higher on the body-mount drawbar), but that is a consequence of body mounting.  It's also true that a lower lever arm gives you more margin to sustain a higher transverse force without string-lining.

I don't think I have any other 89' cars with body-mounts at this time, but I think I'll make some and see what happens.  Please note that I crave the realism of body mounts as much as the next guy (perhaps more so), and I love the detail on these flats!  But I crave reliability even more, and if truck-mounted equipment does better in this setting, then that's what I'll go with.  Can't wait to see how long it takes me to switch back to pizza cutters.   

-gfh

[Scottl]

Interesting to see, I'm with you Gary- reliable operation is a higher priority for me, especially with hidden track.

[wm3798]

I think the key to solving this riddle is to have a line of cars that are similarly equipped.  Having a truck mount coupled to a body mount, particularly one with a long overhang, is no doubt compounding the problem.  I'm no physicist, so I can't address the torque and all that bother, but I do know that a body mounted coupler wants to maintain the centerline of the car, while the truck mount wants to maintain the centerline of the track.

The BLMA car is trying to find the centerline of the MT car, but the MT truck doesn't care.  The result is pressure on the MT truck to pull to the outside of the curve, and conversely, the end of the BLMA car may want to be pulled to the inside of the curve.  The car that is less flexible or even the tiniest bit lighter will lose that battle, which is what your photo seems to imply.

Dave F's suggestion of adding some superelevation may help, because it will provide the effect of broadening the radius a bit.  However, it may not make a difference, because both cars will be experiencing the same curve, just as they are in the picture, and the slope of the grade may exacerbate the problem on the BLMA car, because the center of gravity is now shifted to the inside of the curve...

Dear god.  What the hell did I just say!!

Lee

[3DTrains]

Gary,

Do you think the above might also be influenced by the car's wheels and the track traveled over? There's a slight profile difference between available wheelsets, as well as a notable difference between ME and Atlas track profiles, and I'm wondering if this would be a factor or would be negligible? Obviously you're layout is limited to using ME concrete, but what about the wheels (BLMA, FVM, Atlas, IMR, etc.)? What happens to the above cars using low-profile plastic wheels? Would the friction from plastic wheels increase the likelihood of string-lining?

Cheers!
Marc

[Sokramiketes]

Let's think about the physics for a minute: string-lining arises when the torque about the axis that intercepts the inner wheels is too large for the weight of the car to overcome.  In this case, the relevant torque from the drawbar force is proportional to h sin_theta, where h is the height above the rails at which the drawbar force is imparted to the body of the car, and theta is the angle between the midpoint of the car and the point that the drawbar force is exerted (as measured from the center of the curve).  The sin_theta factor is similar for the two flats, so that mostly drops out of the comparison.  But the height h is different: in the body mount case, the drawbar force is exerted on the body by the coupler mounting screw at the height the screw meets the body.  In the truck-mounted case, the force is exerted through the bolster pin at the height the pin meets the body. 

The height h, and thus the string-lining torque, can be as much as 50% larger in the body-mount case.  As an illustration, just compare the bolster height on the truck-mounted spine car (where the coupler box attaches) to the top of the coupler shaft on the BLMA flat:

I don't think the torque calculation changes based on where the coupler box attaches.  The centerline of the coupler shaft should be at the same height above your axis (through the inner wheel) in both instances.  It doesn't really matter how it is transferred to the wheel, the torque would be the same because the measurements in space are the same.  The "lever" is the same size.

Now, on body mounts the body will move first.  If there is enough articulation in the bolster, it won't lift the truck.  But if that joint is tight, or the bolster pin is pushed in too much, then the body will lift the truck with it.  (Though in an extreme case like your showing, the truck will be lifted no matter what, but I'm still not seeing any droop between the angle of the body and the angle of the truck.  Is the other truck allowed to articulate side to side on this BLMA flat?  Or did both trucks lift up?  (On should have side to side play, the other not... three point style).

[GaryHinshaw]

Some new test results:

1. I swapped in some BLMA and ExactRail reefers into the front of this train (both longish cars with overhanging body-mounts) and both performed well with no inclination to string-line.

2. Putting a small weight (maybe 1 oz?) on the BLMA flat stopped it from string-lining under otherwise identical circumstances.

3. I haven't made another 89' car with body-mounts yet, but I did have a second BLMA flat, so I decided to turn it into a Frankenflat by putting MT trucks, FVM wheels, and 1019 couplers under it, with a washer for clearance.  This beast looks like cr*p, and rides high and wobbly:



but it does not string-line:



In this case I had both flats at the head of the train.  I also tried this with only the Frankenflat, coupled to the spine car, and it was still fine.  The articulation in the original trucks on both flats were fine.  The bolster pins were not too tight, and the range of motion was what I would consider normal.  This string-lining is the entire body rotating, with both trucks lifting simultaneously.  I'm just going to keep these cars in restricted service. 

Now back to track-laying!



I have about 24' of mainline through the Loop shaped and set in place.  A solo BLMA truck passes the roll test from top to bottom.  Wiring and trains to follow.

Cheers,
Gary

P.S.  Mike, I agree with you about the lever arm: if the coupler post is rigidly attached to the body (as it is with the BLMA flat), the effective lever arm is the center of the coupler shaft.  I corrected myself in my second post above.  In the case of a truck mounted coupler, the situation is more complicated because the system is not a rigid body. It is evidently more stable in pulling mode; pushing mode, or downhill running, is another story.  The big test to come is how a full train of auto racks do up and down this hill.

[Philip H]

Gary, if a 1oz weight keeps it from becoming a Frankencar the all you need to do is have one of those cast resin containers that show up in batches on ebay mounted to the deck permanently.  Paint one side for one company, the other for a different company, and move out.

FWIW that track work is stunning, but if there's no video of the truck rolling down it, I call foul ball . . .