Definitely not on the curve!

If you are thinking gaps due to temperature changes, I would think it would be best on the closest straight section to the curve.  Use code 70 rail joiners as slip joints.  I would myself place the slip joint for each rail on the opposite end.  

My 2 Cents!

Nelson,The end of the

Nelson,

The end of the soldered tracks is open/not attached

so a gap where the flex meets should be fine?

I'll be dealing with temps that range from 18C to 26C.

middle

I have an HO 18" radius turnaround, three sections of Atlas code 83 soldered at each joint.

It recently buckled, ever so slightly and caused derailment of a smaller 2-8-0 old time loco.

I pulled up around 10" to let it settle down (latex caulk and sharp putty knife) and plan to cut gap and install jumper wires, in the center or the curve..

Oh I see!

Well then there should be no need for gaps.  If the end rails are left alone to expand and contract you should be fine.  

Expansion

A large part of the "rail" expansion is actually in the roadbed.  A lot of it depends on how "stiff" your tracks are.  

Switches are generally rigidly locked into position, they won't move.  You can either cut gaps so that the rail expands away from the switches or gaps just short of/at the switch and let the expansion go towards the switch.

Another thing that "locks" track are electrical feeders and PC board ties.  PC board ties allow for no expansion or contraction.  I soldered code 55 rail to PC ties in a yard and when the roadbed finally settled in it buckled in every track, busting the rail loose from teh ties in several places.  I unsolderd the track and used ME Micro spikes and haven't had an issue since, because the spikes have enough give to let the track shift.

Dave, good info to have. I'll

Dave, good info to have. I'll keep that in mind for the future for the fourth module, which will have turnouts and some ME track probably.

Thanks Nelson. 

Thanks Nelson. 

Dennis, any pics of it?

Dennis, any pics of it?

Whatever the reason for track

Whatever the reason, track expansion and contraction can occur.  It happened to me on my first layout.  Since then I have decided to make provision for it on my last and current layout.  I have soldered rails on curves to avoid kinks but left ability to expand and contract on straight sections.

I decided to solder the feeder wires to the rail joiners so the rails could slide a bit in them to "breath" for any expansion and contraction, which hopefully will be minimal in my basement where a dehumidifier to keep the humidity from rising about about 50% or so.  Because of that, I have frequent drops.

I know some may lecture us that we must solder all rail joints because you can't rely on mechanical joints like rail joiners to reliably pass electricity down the line.  But OTOH, if you solder all your rail joints, there could be expansion/contraction issues as I learned.  I soldered all rail joints on my first layout and had some pretty major expansion and contractions problems.  In the winter the solder joints pulled apart and broke and in the summer the rails kinked when they expanded.  The environment of that layout was worse than my present one but it was a lesson learned.  I always plan for at least a bit of expansion and contraction since then which means I don't solder all the rail joints.

I did see mechanical rail joint power failure first hand when a friend of mine back in high school built a fairly elaborate 5x10 layout in California and had some dead spots due to this.  But I don't think he had frequent drops like at least every other rail joint.

we must solder all rail joints ?

Hi Jim, I tthink a good compromise is to solder the joints of about 3 sections of track ( say 9 feet) then leave an unsoldered joiner , then solder 9 more feet, etc.  Three sections soldered together will have at least three feeders so should alway maintain electrical power ,and the unsoldered joint will still conduct most of the time but doesn't matte rif it does not. As you mentioned I solder the curves and put the sliding joints on the straight aways. It's rare to have a curve exceeding 9 feet ....DaveB

The way I do

I model in N scale and yes N scale track has also expansion capabilities

To avoid any troubles, I didn't solder piece of track together and some join rail have a very small gap between them

I used first code 55 Peco track and now exclusively ME code 55 weathered track.

I didn't solder pieces of flextrack track together except in curves to avoid kink or miss aligned track join

In fact I only solder rail join, where it's impossible to avoid a kink due of tension in the track, the rest I just put joiners between the piece of track;electrical gaps if necessary are cut after when the track is full laid.

This let place for expansion of the rail and avoid any surprise

I put feeders on each separate piece of track, even if join are soldered between two pieces, this give a bullet proof electrical continuity which is really a must in N scale and like in  all of the others scale of course.

Environment

One thing to take into consideration is your layout environment.  I solder my rail joints and have never had track problems related to expansion and contraction.

Does that mean physics stops working in my layout space?  No, of course not. 

All my layouts including my current layout have been in a main room of the house that has standard insulation and allows for 24-hour temperature control.  Normal expansion and contraction are subtle enough where it does not create an issue.  I would also imagine, although slight, there is enough give where the track can adjust to these slight changes.  

This is much different than say a layout in an unfinished space and/or space where temperature and moisture levels are not regulated constantly.  
 

Joe

I use Peco code 83 track

I use Peco code 83 track because both rails are removable. I run them through a fasttracks roller to give then a natural curve and return them to the tie strip. Now I can cut a gap anywhere in the curve and not worry about kinks in the rail.

Steemtrayn, great way to do

Steemtrayn, great way to do it.

I have removed the rails easily on Peco code 83 in the past for my inglenook, but was surprised howvery difficult it was to on Peco code 70 so I didn't bother. 

I have a fast tracks roller and forgot about having it when I laid the rail in my turntable pit.

Be careful to allow sufficient expansion gaps

In my expansion computations for nickel silver rail for my Run like a Dream TRACKWORK book, I determined the expansion gap size for every three feet of rail needs to be 10 thousandths of an inch. I typically solder two pieces of rail together at the joints and leave a 20 thousandths inch gap every six feet.

For nine feet, you will need to leave a 30 thousandths gap, which is right on the edge of having a gap wide enough to cause a noticeable bump when something crosses it, so be careful.

  I determined the expansion

Hi Joe, What temperature range did you use for the calcs?  The range in my layout room is from about 30 degrees to 100 degrees  but I only work and run trains when the shop is between about 50 and 70. So I'm not concerned if the gaps get larger at very cold temps only that they don't get too tight and warp something at very high temps. I gap the rails about .015" per nine feet if the room is 50 degrees and a bit less if it's 70 degrees and haven't had any problems. ....DaveB

Seems to me just looking at

Seems to me just looking at the rail metal expansion factor and temperature is ignoring most of the problem. In the end, it's a dynamic system and the net of all the moving parts is what counts. The rest of the system isn't going to hold still while just the rail expands or contracts.

You have to look at humidity and how it affects the benchwork in combination with temperature and rail expansion. And it's not the absolute change in rail length but the change in rail length relative to the benchwork - which is itself changing. So its the delta resulting from both changes that matters. 

It's going to depend on what material you use for benchwork, as plywood and dimensional lumber have different properties. And whether you do spline or not. 

If your temperature range is

If your temperature range is only 18 - 26C I doubt you have much to worry about. Also I agree with Marc - solder only on curves and have feeders to every piece of track.

Tony in Gisborne, Australia

It’s both temp and humidity

Foam?

May I ask. Does either temp or humidity affect foam? 2" pink/blue or posterboard or any such layer used in roadbed or even track ties? So, for example: the only thing affected is the rail/nickel silver and everything else is a plastic product - is there any worry about expansion/contraction on the layout?

Foam coefficient of expansion

Wood (cork is a wood product) gets mostly affected by humidity, the effect of temperature differences is negligible.

Foam, on the other hand, is most affected by temperature shifts and humidity changes are negligible. With foam, the amount of change for a 20 degree temperature shift is equivalent to how much wood changes size with a 40% humidity change.

So in short, you can just assume wood and foam are the same as to size changes -- with wood it's humidity and with foam it's temperature.

You can paint wood with a primer sealer and protect it against the effects of humidity, but there's nothing you can do to protect foam from temperature changes except tightly controlling the temperature -- and pray the power never goes out or the AC never breaks down during a heat wave.

I cover all this in some depth (along with many other considerations) in my book Run like a Dream TRACKWORK.

P.S. Foam may also continue to shrink from outgassing as it ages. Ken Patterson has had some older parts of his layout shrink by 1/4" or more due to foam outgas shrinkage from age. When both pieces shrink by 1/4" that can result in a gap as wide as 1/2" based on Ken's experience.

Also, foam is not tightly controlled when extruded as to dimension because it's intended to go inside walls as insulation so who cares if one piece is off by 1/8" from corner to corner. I've taken calipers to foam sheets in Home Depot and Lowes, and from corner to corner, almost no sheets are consistent. Also when I sight down an extruded sheet, I can see waves in the surface -- the stuff is rarely perfectly flat.

Personally, foam is great for scenery, but not the best choice for roadbed, IMO.

Morgan,So far my two modules

Morgan,

So far my two modules that have extruded foam glued to hollow core doors have held up well without notable shrinkage. The modules have been exposed to high humidity and heat like 100F in storage two times and once very dry cool storage for over a year. No effects noticed on the homasote road bed glued to it nor the track; and I'm overly finicky when it comes to track. Perhaps sealing the foam on all sidesand ends with latex paint limits its shrinkage. The only shringage that is notable on my modules is in thickness of the slabs. 

I no longer use Celluclay on small pieces of extruded foam, as it can warp it.

Thanks

To both JoeF and Dee. As you both know, I am fanatic about painting stuff. My 2" foam was cut to the 12" by 8' and then painted top, bottom, ends and sides. And based upon what you both have mentioned, seems like I'm fairly safe. I live in a mobile home and power outages happen. But I do know that the temp holds pretty steady. But humidity is always a "crap shoot". So, now that I know to watch out for the humidity, at least I feel better knowing that I've painted my foam and there's no pink exposed to the air/or the humidity. I honestly don't remember seeing any waves on the foam when laying roadbed & track. But at least I know what to watch for. So a huge thanks for this information.

I think you're safe

I think you're safe Morgan. 

Also my modules spent 9 days in a UHaul truck at temps well above 90F.