Joe, what about Bill Darnaby?

Bill Darnaby's Maumee Route is constructed entirely atop extruded foamboard with no supporting plywood beneath. It's going on 20 years old now. Any reports from him about shrinkage? 

I have been using Foam Board

For over 20 years, and have yet to have shrinkage issues.

Ken L.

It appears to be a YMMV issue

One other note

One other note on foam - the manufacturers also say they keep changing the formula and manufacturing process, hoping to get a more stable product. Now stop and think - how stable is a material going to be if you keep changing how you make it over and over?

One of my principles in my Run like a Dream series is to find layout materials that are highly predictable from batch to batch and go with them. Foam fails that guideline big time - and the manufacturers admit it by publishing up to a 2% dimensional shift over time as "normal".

Old foam maybe different

Agree with Joe, maybe 20 year old foam is better.

My major experience with foam was in the mid-70's with foam taken out of an old chicken coop.

It was blue. That foam hadn't lost any dimensions.

Harold

Polyisocyanurate

I really enjoy these Reverse Running articles! One type of foam not mentioned in the article is the one I happen to use: polyisocyanurate ("polyiso"), sold under the brand, RMax Thermasheath-3, at Home Depot. I chose this for my sub-roadbed because XPS (pink or blue stuff) isn't typically stocked at Southern California home centers. The industry-standard set by the association that sets these sorts of things for rigid-foam insulation was 2%, but has since been revised to 1%. These percentages indicate the maximum a manufacturer's product is allowed to shrink, and not necessarily how much they'll actually shrink.

According to one scientific study, the typical amount of shrinkage of 4' x 4' test pieces of polyisocyanurate was only about 3mm, with a nominal amount of only 2mm (that's about 0.0016%). The report stated that this shrinkage was mainly due to out-gassing and was independent of temperature-change. Impressive, since temperature-swings were significant in these tests, since these were roof-insulation studies. According to the six-year study, the polyiso tended to stabilize after two years (i.e., it stopped shrinking). The cool thing about polyiso is that you can spray-paint it since it's chemically resistant. It's also compatible with common adhesives. Plus, it's easily cut with just a hobby knife (no hot-wire needed!), and just as easily shaped using a 3M sanding sponge. The bad thing about is that it comes with a thin sheet of paper-backed aluminum foil on both sides. Other than that, it's great stuff!

Well, maybe not

In 1999 I spoke with several contractors before deciding to use the foam. From what I was told and read, shrinkage would only occur in severe heat... not cold. (I thoroughly understand and agree with ymmv). I saw the same rumors floating around back in 2008... I think model railroader forum had a thing on it.  All that taken into consideration.. maybe the heat outside could be the perfect suspect for Kens issue with shrinkage. oh yeah.. wood also shrinks, expands, and warps too ( just fun thoughts here). Then there are other things to factor in.. what is the foam covered with, actual roadbed for track, etc.etc.. In summary I just want to point out that there are enough examples of no shrinkage and other info provided by the makers and users to justify continued use. Just as YMMV, personal preference also varies. So no matter what any one chooses.. do your homework and dig in

It is The Devil's Material!

A while back I experimented with foam roadbed and was not impressed. You can't nail anything to it (I prefer to use track nails to secure my track to the layout). It has a "sounding board" effect, which I was never able to solve. Someone suggested using plywood under the foam - if I'm going to have to buy plywood anyhow, why bother with the foam? 

For scenery, it is diffucult and messy to make landforms with. That is, unless you use a hot knife which produces toxic fumes!

And now we find out that, as George Kostanza might say, there is significant shrinkage...

No thanks!

George J

I like plywood better. If I

I like plywood better. If I place a hand on it and lean on it plywood does not dent or break.

Cork

My layout is built on the second floor of my home. It is more or less a finished attic; carpet, dry wall, central air vents. However, the central air doesn't quite keep up with the upstairs temperature changes. While the average home layout in a bedroom or a basement may experience a 20 degree range, my layout probably experiences closer to a 30 degree range. 

Summer time in WV can be nearly tropical, just a couple weeks ago temperatures were in the mid 90's with humidity in the 70's. Yet by January, it can be 10 degrees outside with a stiff breeze and a foot of snow. This also affects the interior temperature of the upstairs. 

I was dissatisfied with the humps in the foam board. It may appear to be smooth , but the humps show when you place roadbed and tracks. So, I'm using the old tried and true method of cork roadbed on 3/4" plywood and risers.

My last layout . . .

. . . well, last permanent layout, was built atop hollow-core bi-fold doors capped with one-inch pink extruded foam. And that upstairs train room got HOT in the summer. 

I noticed no shrinkage, warping or other defects related to the foam in the eight years that layout existed. It was stable. 

So, yes, YMMV, but so far the anecdotal evidence is running counter to your thesis, Joe. However, you provided good food for thought.

To what degree does having a foam/plywood sandwich reduce

(or eliminate) the effects of shrinkage?  Would using luan plywood glued to foam (perhaps using Great Stuff) be stable enough over time?

@MikeM

Wondering that myself. Does laminating the foam board to something more dimensionally stable, such as a hollow-core door or cabinet-grade plywood, in turn reduce or eliminate whatever tendency for the foam to shrink?

Sandwich?

I shouldn't think so. If the foam wants to shrink, it will shrink, probably leaving you with something trapezoidal, in the worst case maybe warping the plywood. I don't buy the 2% number, I think we would have noticed 2 inch holes opening up by now. But a gap of a quarter of an inch seems quite likely.

However--- none of our materials are dimensionally stable. We have just developed techniques to deal with the problems. We arrange wood beams so warping stresses tend to be in compression, we support long pieces with braces at a right angle, we cover cork roadbed with rocks in a matrix of soft glue so any gaps that open up will be under the ballast, and we gap our rail so it won't kink.

I'm not sure what the solution will be for foam. I built my layout out of 24" pieces tiled and stacked, and I can slide an index card into joints that were tight three years ago. Those will eventually be covered with plaster, so no biggie. The track is on cork/rubber composite, which has enough give and slip that the foam could dance all over without disturbing the rails. Maybe using small lengths of foam is all we need to do, perhaps more sophisticated techniques will turn up.

pqe

Yes and no

Yes and no. In my Run like a Dream: TRACKWORK book, I publish my research on what products are the most dimensionally stable and also which products have the tightest manufacturing tolerances. So some products are better than others, although none is perfect.

In order of most dimensionally stable to least dimensionally stable (and also factoring in tighter manufacturing tolerances), the order is ...

• Drywall (most stable, thickness manufacturing tolerance +/- 1/64")
• Masonite 
• Plywood
• Dimensional lumber
• Cork
• Homasote
• Foamboard (least stable, thickness manufacturing tolerance +/- 1/16")

​As for dimensional changes, I assumed a 20 degree temp change and a 40 percent humidity change. This works in a climate controlled indoor location, but for places like an uninsulated space or modules that get transported to shows, the changes could be outside this range.

With dimensional lumber, cork, and homasote, the dimensional changes are mostly from humidity and a couple good coats of shellac primer to seal the material can help greatly. 

Drywall, masonite, and plywood are slightly affected by both humidity and temperature changes, but because of the way they're made and the materials that are used, they change only slightly (on the order of 1/32") compared to the rest of the list. Plywood is the most interesting because the laminations are cross-grain and pressure glued - making any dimensional changes in the laminations self-correcting - the different laminations actually pull against each other keeping the total board surprisingly stable dimensionally.

With foam the culprit is temperature (humidity changes have little effect), and you can't coat foam with anything to insulate it from temperature changes. Foam is the most unpredictable - one piece can change very little but the next piece can change more. I have one piece of foam I bought a few months back that has warped upward in the middle while laying flat, so the warp went *against* gravity. The next piece I buy may stay flat as a pancake. You just never know with foam, and that's the trouble.

MDF

Joe,

Do you have any data on MDF (Medium Density Fiberboard). I've been using it for years and haven't had an noticeable problems with it. I use 1/2" thick pieces for my (sub)roadbed and it seems to work well. 

George J

MDF ...

" most dimensionally stable to least dimensionally stable"

 "most dimensionally stable to least dimensionally stable"

     Those only tell the theoretical stability, the working characteristics are a lot different. Sheetrock is physically very unstable, if hit with a hammer it crushes, or if a nail is driven in it there is no holding power so it's not as good for roadbed as the lower rated homasote which might rate low on stability but stands up well to hammer blows and takes nails quite nicely. I've found cut offs of homasote in my yard that are 20-30 years old and still solid so would not worry about it's stability.  The kind of foam we use for model railroading is not a very strong material so it's dimensional  stability should be of less concern than it's strength and resistance to crushing and even breaking under load. Foam is a good insulating and core material ( think ice chests and surfboards )but not the best thing to make utilitarian loaded structures like benchwork from. Along with stability one should also be thinking about things like how to mount switch motors, fascias, legs, etc. when deciding on the benchwork and roadbed materials. Weight is a minor issue for most layouts, even modular portable layouts need some weight to stand up to operators bumps and leans......DaveB 

I agree

Shellac primer

Are you using an actual shellac based product ("extract" from the Lac beetle) or is it something synthetic? Real shellac is often used as a finish on violin family instruments. It's not particular tough or water resistant and at least originally, was a finish developed hundreds of years ago based on it's appearance and sonic properties. Was this a water based product? Shellac (at least instrument grade) is usually a bag of flakes (it's literally lac beetle poop) mixed with alcohol.

I'm curious about the best way to make masonite stable since usually when you paint it, it warps like crazy until it dries. Suggestions?

Bob

Foam or plywood: Flexibility vs. ease of wiring the layout.

I built my layout on top of some existing MDF workbenches. On top of these, I laid a thin layer of a polyethylene floor underlayment as an attempt at some measure of sound-isolation. I then laid 3/4" melamine which serves as the layout's lowest-grade. On top of that is the 2" polyisocyanurate ("polyiso") I mentioned earlier which elevates most of the layout. I'm not worried about the polyiso, since it's highly dimensionally stable, shrinking only about 0.0016% (primarily due to out-gassing, not temperature- or humidity-change). I imagine the melamine laminate is pretty stable since it's a thermally fused plastic of some kind. The melamine is laminated to particle board, which has lower density and strength than MDF, but I would guess that the melamine laminate likely increases its dimensional stability, possibly surpassing that of MDF's.

I'm not married to the idea of a foam-based sub-roadbed, although the polyiso sure is easy to work with (the layout is still in its early stages and can be easily re-worked). Plus, I'm not crazy about how crushable foam is. In fact, I may just switch it out for another material just to make wiring easier; e.g., something like two layers of 3/8"-plywood, separated by a matrix of  2" x 4" blocks of dimensional lumber. This would make for a mostly empty "middle," where wires may go where they please. If you know where your depressions are going to be, then a foam sub-roadbed really isn't necessary (though, I'm not a huge fan of jigsaws, since I'd have to cut-out some irregularly shaped pond depressions). But, looking at Joe's list, I see that plywood is more dimensionally stable than I thought, so that's probably a good choice.

Joe: What is the purpose of the drywall sandwich?

Is this to keep the plywood from warping?

Tie width for your upcoming e-book

Hey Joe,

I took a look at your sneak peak of your upcoming e-book.

I have a request... Could you add tie width to your table of manufacturer tie dimensions? Actually, length would also be great.  Heck, tie spacing would also be nice

I've noticed different manufacturer tie widths vary.  I've noticed this while experimenting with handlaid track techniques.

With a mixture of wood ties, PC ties from different sources, ME flex track, CV tie strips - it quickly becomes evident that "7in x 9in" means different things to different manufacturers

Drywall-plywood sandwich