Tehachapi, BC

[GaryHinshaw]

Very interesting feedback jb!  If I may paraphrase your comments: each end of the yard has only a single track ladder and therefore can only handle one train at a time.  Since these two yards funnel down to a double-track main, I didn't see that as a big limitation, but your suggestions add some interesting options at pretty low cost.  Let me make some sketches and revisit this a bit.

As far as capacity: Mojave staging can handle 6 NB trains and Bakersfield can handle 6 SB trains comfortably, so they are pretty balanced.  I also have the storage yard which I plan to have staged with additional trains, if needed.  This latter yard has short connections to both north and south staging, and the assumption is that trains will be transferred during a typical session.  I still don't know how many trains I expect to push across the line in a typical session; I need to get the lower deck finished to find out!

-gfh

[GaryHinshaw]

Yikes! I certainly know this particular pain. I'm still running Yosemite and have been avoiding upgrading... does it hose BootCamp somehow? 10.12 (Sierra) is around the corner, too.

Sigh.  I had some more forced laptop down time since the last post here.  Long story short: I had a hard drive cable failure for the second time in 5 years.  In the end, I replaced the cable, swapped my drive for a nice new 500 GB SSD, and did a fresh install of OSX 10.11.5, and now the machine is better than new!  But that upgrade, along with some work travel, has precluded much progress on TBC.

I did get back to the Bakersfield plan to work on jb's suggestion about adding more leads to the yard.  First, I went back to the prototype and sketched out the following simplified and compressed schematic:



Northbound is to the right, and in reality, there is about 2 miles between Kern Junction and the BNSF yard, while the UP yard straddles Kern junction (not all tracks shown). 

Here is my latest iteration of Bakersfield:



There are now 2 leads at either end which divide the yard into arrival and departure yards, 4 tracks each.  The outer half is UP arr./BNSF dep. and the inner half is UP dep./BNSF arr.   With 4 leads and 2 mains, there are 8 possible routes one might wish to follow.  All but 2 of them are accessible in this configuration: track 1<-->UP arr. and track 2<-->BNSF dep., neither of which would be required for standard right-hand running.   I could open up the latter route by moving the crossover to the left of the BNSF dep. lead, but: 1) that is not prototypical, and 2) it would preclude simultaneous arrivals and departures from the BNSF side.  (The current arrangement supports simultaneous arr./dep. from either RR's yard.) There are both left and right crossovers further up the main to enable left-hand running, if desired.

In normal ops, I would imagine staging ~3 trains in each departure yard, and as one train departs from a given RR, it opens up a track in the other RR's arrival yard, so it should stay fairly fluid during a session.  All in all, this is a much better arrangement and I really appreciate the feedback.  Further comments always welcome.

[jagged ben]

Glad I could be helpful!   

[ednadolski]

It's a little hard to judge the precise geometry from the sketch,  but I'm still left with the impression that a double-slip might be worth considering for that place near the 5' mark where the two turnouts meet almost point-to-point.  If those points are too close, then a train on Track #2 for the BNSF arr. yard could experience a potentially troublesome S-curve.  (It's probably not prototypical, OTOH the prototype would not have as much compression between the two turnouts.)

If you will have #10 turnouts, that might mitigate the S-effect at the expense of some space.

Ed

[GaryHinshaw]

Interesting question Ed.  If I understand you correctly, you are suggesting I consider an arrangement like this:



as an alternative to the current plan:



correct?  I can see the merits.  Happily, I expect to begin benchwork construction for this segment starting this Saturday, and geometry testing will soon follow, so I will keep this under advisement.   

I have to admit that I kind of like the look of the S curve in the original plan, but reliability trumps all.  Alas, I don't have any auto racks with body mount couplers to really put this to the test, so I'll have to use some judgement.

[ednadolski]

Yes, that's what I had in mind.  As I recall the 'rule of thumb' to avoid S-curve issues with adjacent, facing-point turnouts is to have a straight track between the points that is at least the length of the longest car that will traverse that route. If space it really tight, then you might be able to get away with using the wheelbase length, but overhang and the angle of the turnouts could become factors. Alternately, the double-slip will, of course, eliminate the S-curve entirely.

When testing, it's probably good to use combinations of cars/locos of various weights and short/long overhang, and run both pulling/tension and pushing/compression.  (Do you still have that 'Frankencar'?)

The issue with close facing-point geometries is that it forces the couplers to swing against each other, and the heavier car (or loco) will "win".   I can't seem to find my copy of the John Armstrong track planning book, but that's where I remember seeing some coverage about this topic.

Ed

[GaryHinshaw]

Roger that Ed. Actually, it occurs to me that I have a similar geometry already in service in the Vortex^®.  Here is a diagrammatic comparison:



The red track is the Kern Junction plan, flipped and rotated to align with an interlocking in the Vortex where the helix squeezes from 3 tracks to 2.  The S curves in that section are tighter than in Kern, but they have posed no problem at all for any of the trains that have been through there (which is quite a few now).  So I think I'm good.  In any event, I do have enough room to push the crossover to the left (in the original plan) quite a bit if I think it's necessary.  Thanks for the input!

[ednadolski]

Roger that Ed. Actually, it occurs to me that I have a similar geometry already in service in the Vortex^®.  Here is a diagrammatic comparison:



The red track is the Kern Junction plan, flipped and rotated to align with an interlocking in the Vortex where the helix squeezes from 3 tracks to 2.  The S curves in that section are tighter than in Kern, but they have posed no problem at all for any of the trains that have been through there (which is quite a few now).  So I think I'm good.  In any event, I do have enough room to push the crossover to the left (in the original plan) quite a bit if I think it's necessary.  Thanks for the input!

Gary I may be overlooking something but I don't think I see the same S-curve situation in the Vortex track in your last pic. There does not seem to be any point-to-point turnouts of the same handedness in there.   The concern I has was over this route:



Which looks to me like with long-overhang cars it could generate a situation like this:  http://familygardentrains.com/primer/reliable/s-curve1.gif

(Note, that pic is rather exaggerated to illustrate the point.)

Ed

[davefoxx]

@GaryHinshaw,

If you put the back-to-back points as close together as you can get them, you will likely have no problem at all, especially if you're using large turnouts.

DFF

[C855B]

^^^^^ - Astute observation. That creates a geometry very closely resembling a double-slip without actually going there. I should think about this in places in my plan.

[Missaberoad]

Also depending on the size of the turnouts you are using this may not be an issue at all.
The rule is to avoid tight S curves.

What size turnouts are you planning to use in your yard?

[GaryHinshaw]

New plan (unrelated to potential S-curve issues).  There were two recent changes to the plan that I did not fully think through: 1) splitting the yard into arrival and departure yards and doubling the number of leads, and 2) moving the loco service outside the balloon tracks, rather than inside.  The plan above allows pretty good flow between the 4 leads and the 2 mains, but I did not think through loco moves or transfer jobs to/from the storage yard, both of which have issues now.  Hence I'm thinking about breaking with the prototype and including both a left and right crossover as shown:



The plan lets all 4 staging leads access both mains.  More importantly, it lets the top staging lead access the storage yard lead off track 2, and it streamlines the movement of locos in and out of the yard, especially to/from the top lead.

I'll mock up the S curve shortly and report back, but I think it will be fine.  The baseline plan is large-radius Peco turnouts with close points.

[jagged ben]

Well, I'm a little sad to see the prototypical Kern Junction track arrangement fall to the wayside, but that storage yard is certainly an important consideration.

With respect to S curves, my understanding is that you either want the points right up next to each other or you want a full car length (at least) in between them.  It's in between distances that cause the most problems.

[ednadolski]

Looks to me like you could eliminate the S-curve problem by simply swapping turnout "A" and crossover "B":




(And that would save some testing work)


Ed

[svedblen]

Looks to me like you could eliminate the S-curve problem by simply swapping turnout "A" and crossover "B":

Cutting the Gordian knot! Nice thinking Ed.