Tehachapi Loop II

[mark dance]

Or use Peco as it is hidden staging.

Md

[ednadolski]

Mark, I've never used the Peco flex before, tho it seems that most folks like it.    Does it have the loose/sliding rail?   Is the gauge accurate, or does it run wide, or tight?   I should probably pick up some to try out.

@GaryHinshaw I don't mind the look of nails tho I would be concerned if I ever had to make future adjustments, since there will be several levels of helix above this.  I use fine wheel flanges so I'm not too worried that they would hit the nail heads.  Likely it would take 3-4 nails on either side of each gap to force the rail back into the desired curvature.   @pdx1955, a similar thought, that when/if I ever need to repair a solder joint it would be hard to get at with the upper levels in place.

@wcfn100 I'm rather hesitant about the idea of gauge widening.   The prototype doesn't do this (afaik) and the only reason it is done for modeling is as a compromise to allow rigid, long wheelbase equipment to run on very sharp model curves without binding on the flanges.   My understanding is that mechanically, it is preferable to have as little slop as possible, to keep wheelsets from 'wandering' and developing excess friction.  This is not to say it cannot work, just that I am not sure of any benefit in my use case (as I have no steamers and all my curves are all +24" radius anyways).   I'm thinking that one thing I should do before trying anything else is to cut a few test gaps in the outer rails, to see if the Atlas track will pop out of curvature on that side too. (If it does, then I will be reaching for the scraper...).

Cheers,
Ed

[GaryHinshaw]

As you probably know, I use Peco flex and turnouts in all of my staging, and I'm very happy with it.  It does not have one loose rail like Atlas, but it does have tie gaps preferentially under one rail, and that rail should typically be on the inside of the curve.  It has silky smooth flexibility (unlike ME), but it also holds its shape (like ME), so it's really the best of both worlds.  I'd use it everywhere if it looked more prototypical.

I'd still give nails a try, like you describe, since you have nothing to lose.  I think it's very unlikely that you'd have to adjust it in the future.  But by all means try Peco if you'd prefer.

[svedblen]

Ed, I do not think it is wise trying to fix this with spikes. It would be like doing a flex track joint in a curve with spikes only, and no rail joiners. You would not try that, would you?

[wcfn100]

@wcfn100 I'm rather hesitant about the idea of gauge widening.   The prototype doesn't do this (afaik)

Actually, I only mentioned it because I think they do (or did if just for steam).  I'll see if I can find whatever my brain is referencing.

Jason

[wcfn100]

Here it is, Army Technical Manual 5-628 (1991)



Don't know how this relates to all railroads or eras, but I knew I had read it.

Jason

[nuno81291]

Ednadolski your work is far too good for a compromise such as nails... multiple layers of helix over it? fugetaboutit' something like the peco flex or going back in with ME should be much more bullet proof, and I wouldn't want my staging let alone helix to be anything less than bullet proof. I have done the nail thing with atlas55, it works, but I did have to go back in to adjust over the brief life of that module.

[GaryHinshaw]

I think the "bullet-proofness" of nails depends on the substrate.  With bare plywood there, it is going to be very solid.  To be clear, I would place them right against the rail base (one for each rail) in tight-fitting pre-drilled holes.  If the results seem at all suspect, then switch track.

[Missaberoad]

A 12 degree curve is pretty sharp, for example Tehachapi loop Is mostly 10 degree curvature.

I think, but I stand to be corrected that you wouldn't find mainline track as sharp as that chart indicates.

All that being said in N scale a 12 degree degree curve is a ~34" radius so 

[John]

replace a couple of the plastic ties on either side with PC board ties .. solder to guage and move one to the next one ..

[railnerd]

American Tie and Timber Gapmasters are awesome for avoiding this…

http://www.americantieandtimber.com

[Santa Fe Guy]

American Tie and Timber Gapmasters are awesome for avoiding this…

http://www.americantieandtimber.com

They do look like they could do the job very nicely. I am considering using these on my new switching layout (well the old SFRSD got pulled down and sold when we sold the house) in lieu of any rail joiners at all.
Rod.

[lajmdlr]

A 12 degree curve is pretty sharp, for example Tehachapi loop Is mostly 10 degree curvature.

I think, but I stand to be corrected that you wouldn't find mainline track as sharp as that chart indicates.

All that being said in N scale a 12 degree degree curve is a ~34" radius so 

Uintah Ry had 65 degree curves on 7.5% grade & Los Angeles Junction has one place w/ an 85 degree curve & a #3 switch. Here's a table showing HO, S & O gauge degree of curve in inches: http://www.railwayeng.com/Track/Curves.htm

[Missaberoad]

Uintah Ry had 65 degree curves on 7.5% grade & Los Angeles Junction has one place w/ an 85 degree curve & a #3 switch. Here's a table showing HO, S & O gauge degree of curve in inches: http://www.railwayeng.com/Track/Curves.htm

Awesome examples and wow that's sharp (Moro castle looks too sharp to be a model railroad) 

http://ngdiscussion.net/phorum/file.php?1,file=14682,in_body_attachment=1

And 85 degree curve in N would be a 5.5" radius!

There's also this example in Portland OR which is ~70 degree curve

http://3.bp.blogspot.com/_sL3WQYmDW_A/TKUfsY_CWyI/AAAAAAAAApg/06HgpcmRoPo/s1600/Tight+curve.jpg

But Since the Unitah was a turn of century narrow gauge line and the LA Jct is industrial trackage I stand by my original statement that you wouldn't find curvature sharper then 12 degrees on mainline trackage...

Edit: D&RGW in Byers canyon is the sharpest I can find at ~15 to 16 degrees.

BC Rail also had a 16 degree curve on the Squamish sub
(and a lot more apparently 
North Vancouver to Kelly Lake, 189 miles,
108 curves of 12+
19 curves of 13+
2 curves of 14+
1 curve of 16)

Cantara Loop on the former SP is supposed to be 14 degrees.

Even the Spiral tunnels on CP are 10 degree curves, there might be some sharper ones North of lake superior, would need to find a track profile.

The BN Sumas sub had a 15 degree curve that was restricted to 4 axle power and 5mph...

But anyways back to Tehachapi... My vote is for PCB ties... 

[ednadolski]

I'm not too sure that it will be possible to rescue/repair these joints.   Despite its flexibility, the Atlas rail has no desire to keep to any curvature unless it is forced at multiple points.   More so, it is very tricky to re-align the railhead once it has moved out of position.   Here is my "Frankenjoint" attempt using a couple of track nails:


IMG_1777


The plywood is very solid and takes/holds the nails well.  The nails can also be adjusted a bit with a few (gentle!) taps of a light hammer and nail set.  However the margin of error in the alignment of the rail ends is very small -- and even after lining them up they still can flex a bit.  I think @svedblen said it best: it is very much like trying to make a rail joint on a curve without using any joiner at all.   (It can be done, as @GaryHinshaw has shown, but only with track like Micro Engineering that has virtually no room for the rail to move out of alignment. The sliding rail on the Atlas track is way too loose for that.)

The only way to avoid avoid the alignment problems is to secure the rail in place with PCB ties,  or @railnerd's Gapmasters, before making the Dremel cut.  On curves I think there should be no less than 3 PCB ties on either side of the cut (geometrically, two points define a straight line, and 3 are needed to define a curve).

I am reluctant to replace the track completely, and perhaps I can avoid that if I can manage to slide out each inside rail and slip an uncut length of rail in its place. That said, I am also now having some doubts about keeping with the Atlas track (altho it worked fine in my other helix, that did not have isolation gaps), and the Peco track is sounding like a better choice.   If Peco has good insulated rail joiners, then maybe it would be better just to go with those and pass on the whole notion of PCB ties and trying to force the Atlas rail into something that it does not want to do.

Ed