Tehachapi, BC

[Scottl]

Looks great, but still no sign of local power.  NS today?  LOL.

The trick with the rail joining is very clever.  I immediately liked it until I read his blog and realized you have to attach feeders to every piece of track.  Even on my medium sized layout (10' X 7'), I have probably 40 pieces of flex track!  Some people hate ballasting, but I hate soldering feeders more! 

[GaryHinshaw]

That's funny, of the ~45 locos in my fleet only 5 are foreign road power.  There must be some hidden significance to my choice of photos...  (Actually, one of the biggest gaps in my roster now is a small stable of NS Dash-9s and GEVOs for run throughs.)

I too hated soldering feeders -- until last night!  Jamie's method of exposing the rail base is really simple, as long as you haven't already installed your track.  And you have to admit that feeders are more reliable than joiners...   Now I'm curious to see how the joinerless method works on an 18" curve.  But with stiff track, and rail clips the whole way through, I think it will be fine.

-gfh

[MichaelWinicki]

Hey Gary... Looks great!

What's the benefit of not using railjoiners?

[DKS]

Hey Gary... Looks great!

What's the benefit of not using railjoiners?

I used the same method on my WR&N IV, and would do it that way again on any future layout. There are many benefits. First, attaching feeders to every length of track ensures reliable delivery of current; you're not relying on any rail joiners to carry current. It also does away with physical rail joiners, which have always been something of an eyesore to me. And with no physical connection between sections of track, expansion and contraction of subroadbed vs. track is never an issue. It's not that onerous to attach feeders to every length of track, since that amounts to feeders every three feet or so.

[GaryHinshaw]

I agree with everything David said.  Michael's question got me to wondering how this technique would work with a truly flexible track like Atlas, and it seems to be fine.  Here's a quick pair of shots from a 5-minute test in the garage:




The method, in brief:

* Cut 2-4 ties off one piece of track,
* Cut the fixed rail of the other piece short by the same 2-4 ties, but be careful not to cut the tie strips.
* Slide track together and stagger the joint on the flexible rail by a few ties.

In the shots above the curve radius is 18" and I literally just set the track down and held the curve with the straight pins in the picture.  The joint is virtually invisible and there is only the slightest of kinks at the joint.  The clear mechanical advantage here is that you're holding the rail curve and track gauge with clips at every tie, instead of cantilevering the rail over open space and relying on fairly weak joiners to do so.  Of course, thousands of modelers use joiners with no trouble, this is just another option.

Cheers,
Gary

[M.C. Fujiwara]

When building curved turnouts and trying to get the curve through the frog, I've found that I had to bend the rail and then cut back 1/2" or more from the end: it's pretty hard (nay, impossible!) to get the last few mm of rail to curve without kink.
(darn physics).

So perhaps you could pre-bend the rails on the flex, and then trim back 1/4" -1/2" to maintain the smooth curve.

[ednadolski]

Hm, that last shot looks to my eye like a bit of a kink on the inner rail, and a bit of a spread on the outer one.   Hard to tell if there is anything there that may affect performance or not.   My only concern would be if there could ever be enough expansion/creep to push the two rail ends against each other and cause the alignment to slip.

FWIW, one option to help ensure rail alignment might be some photo-etched joiners for C55 rail like these.    I've never tried them myself, but they seem like they ought to work OK.

Ed

[GaryHinshaw]

Haven't made too much progress recently due to work, but I did finally managed to tack down and wire up enough track to run a full length train and do a bit of performance testing.   My default train length will be about 11' which, in the case of the grain train below amounts to 24 hoppers and 3 Dash-9s:



Happily, the 3 locos were able to handle this train up the ~2.3% grade in this section without too much sweat, and the train doesn't overwhelm the space.  I'm guessing that a 13' train with 4 locos should work out ok as a maximum train length (just long enough to crossover itself on the Loop).  Next up is to see how this train does through the 18" curve at the top of the hill.  I never thought 18" would look like a sharp curve, but it sure does in this context.

-gfh

P.S. Those Kato Dash-9s look so naked without detailing, but they sure are silky smooth runners.  After so many years without a pike, it's awesome to watch this snake slither up the hill!

[wm3798]

Your benchwork is museum quality...  Good thing your scenery work is so good!  I'd miss that view otherwise!
L

[Coxy]

Hey Gary,

Post more shots if you got 'em. Eye candy really!!

Seems that ~12' train length, give or take a foot or so, is a good length for an N scale layout. Longer that that doesn't really contribute much more to the sense of running a long train and can introduce train performance issues.

It does sound odd to cast 18" radius as sharp. Did you settle on superelevation for the loop or decide to leave it flat? On "tighter" curves, (is 18" really tight?), I would expect N scale superelevation could create a few problems like the mess here at Caliente:







Cheers,
Steven
 

[GaryHinshaw]

What benchwork Lee?  All I see is the Mojave slope...

Wow, amazing photos Steven!  Did you take them?  Of course string-lining is exactly what I'm worried about, which is why track laying is proceeding so deliberately.  There is .020" in the two broad curves under this train, and I hope to get the 18" curve in place today so I can do some stress testing.  Then I'll return to the Loop shelf.

Cheers,
Gary

[C855B]

Those are amazing pix. In a perverse way, it's reassuring to know this happens to the prototype, too.

In a way, I can surmise the problem may have been caused by a DPU not hearing the "go!" command from the head end, which has been a problem in the Tehachapis since RCL was invented. In the early '70s, SP installed a very elaborate repeater system between Cable and Ilmon to relay control signals between the head end and the remotes. There was, as I understand, heated discussions between SP and ATSF about radio helpers, and that SP was not going to let Santa Fe run remotes that weren't setup for the repeater frequencies. So they didn't.

[Coxy]

Gary, I happened on this scene on my way back to the Bay Area from a family vacation to Disneyland last July (the kids get Disneyland, I get Tehachapi!!  ). I almost didn't notice the derailment at first. Other than the cars on the ground, there were no MOW vehicles.

There's more shots of the scene on my flicker site if you're interested. If you start with this shot (http://www.flickr.com/photos/coxsj/6489495727/in/photostream/), and click "Newer" at the top right, you'll see shots of the cars on the ground and shots of a WB BNSF Z-train passing the carnage taken from the signal just west of Tunnel 1. (BTW has that signal always had two heads? I notice it showing Y-over-Y for the siding at Caliente in one of my shots.). There's also (hundreds) more shots from the top of the hill to the bottom depending on how much of a fix you're looking for! Glad I've taken plenty of pics and video over the years, access to good photography locations on the line seemed more noticeably curtailed.

C855B, not sure what happened to cause the wreck. It's interesting to speculate.  Your suggestion sounds very plausible for what pulled 'em off. The track seems to have been well cleared, so the cars on the ground could have been down for days or weeks. The first priority in a wreck seems to be getting the line open again and the fallen cars are dealt with at a later date. You guys may also remember the stringline accident at tunnel 10 that happened right in front of the Trainorders webcam (early 2000's?). I'm pretty sure there was a couple of BN SD's on the ground for a few weeks after the line was reopened.

The centerbeams on the ground at Caliente were empty. If they were empty in their train, I would have picked the train as a westbound (downhill), headed for the pacific northwest. However stringlining would be more of an issue for an uphill train. The loads could have been removed during the initial cleanup, so this could have been an uphill train. If that was the case, robot helpers not kicking in as the head end throttled up could easily have caused the stringline. Of course if the centerbeams were empty as well, they would have been prime candidates to succumb to tensile forces on the curve.

Gary, thinking 1:160 for a minute, I guess the key issue is how close to stringling will your trains be on that 18" curve + local grades when laid without SE? If it is already marginal, or marginal for some train profiles - say lots of light cars at the head end, or stalled tail end help as C855B suggests - then super-elevation could, literally, be the tipping point and lead to cars on the scenery. On the other hand, if it's not marginal at that location, then superelevation goes back to being a cool cosmetic track treatment. Hopefully that 18" curve is broad enough. In any case, is there much superelevation on the prototype at Walong?

That curve at the extreme end of Caliente horseshoe has always amazed me how tight it is. The grade steepens there as the track heads up hill and the curve itself is very tight. Quite frankly, it is miraculous to me that trains stay on the track at that point , especially heavier trains.

I visited California from Australia in the early '90s, when the now-infamous US housing boom was gathering steam. I have fond memories at Caliente watching ginormous SP lumber trains with countless scarlet and grey SD's on the point and at mid-train and heaving on the tail end, just hammering out of Caliente and swinging around that curve as they made their run at the hill. I just marveled at the flange squeal, the rail singing, the roar of the EMD prime movers and the sheer brute force of it all lugging all that dead weight up the hill, with the whole train folding back on itself around that horseshoe curve, in the desert heat.  Just an awesome experience!  Visceral stuff!  Sights and sounds like that sealed it for me as far as southwest railroading was concerned.

Steven

[GaryHinshaw]

Thanks for the follow-up Steven.  That must have been a pretty disconcerting site to see so many cars on the ground like that.  I hope when (not if) that happens on my pike, that the cars just tip over, without falling hundreds of scale feet to the concrete floor below...

The remainder of this post is squarely in the Engineering Report category.  I did manage to get a few more sections of flex in place over the weekend - enough to start pulling trains around the 18" curve, on a 2.3% grade, at Monolith.  Here is a shot of the head-end of a 30 car test train with 3 locos (~13' long overall) navigating the curve:



The zoom in this shot emphasizes the sharpness of the curve, even though it is quite broad by most standards.  Happily, I have not had any troubles negotiating this stretch, even with a tall, light Auto-Max at the head end (with crappy plastic wheels). There is no super-elevation at this point, but - so far at least - there is no sign that the head-end cars want to string-line.

There is a slightly kinked track joint right in the middle of the curve above (the track is not glued down yet - it's just sitting in place for now). This has yet to cause a problem, but I don't intend to leave it that way.  On the side, I've been practicing some track laying methods and have this test drying overnight:



There is a joint in the middle of the test curve that I'm holding with a Fast Tracks Sweep Stick while the adhesive dries.  If it stays put, this joint will be very smooth.  The adhesive here is Titebond "Green" construction adhesive, and for my money, it seems pretty good for track laying.  It has the appearance and consistency of room temp. peanut butter and it grips like an SOB, yet it dries slowly enough that you can fine tune the track while the stuff sets. And when it does set, it is solid, yet slightly flexible.

Finally, I got about 30' of bus wire installed:

 

Once I glue the track in, I'll clean up the feeders and solder them to the bus.

Thanks for looking,
Gary

[nscalemike]

Looking great Gary, as always!  It's been said before, I really like your benchwork/subroadbed.  My helix track was about a 19" radius on the inside and about a 2.3% grade and never had a stringline problem, so I'm guessing you'll be fine.

Keep it up!

Mike