Modelling what might have been in southeast BC and northwest Wasington

Trying to get the #$@&%*! plywood flat

After all of the work that went in to installing the joists and risers so that the section would be level, the stars conspired against us in the form of warped plywood. This is the tale of that pursuit.

But first, a bit of background on my usual construction methods. All of the sub-roadbed so far on the layout is composed on two layers of 3/8” fir plywood, laminated together from narrow arcs and straight bits that fit under the tracks. With this approach there are no splice plates per se, as the plywood pieces form a continuous two layer sandwich all around the layout. This has proven to be a very good approach as can be attested to by the lack of running issues on the rest of the layout. One problem with it, however, is that the areas between the tracks are wide open, and more bits have to be pieced together to fill in for buildings, etc. This was quite a job under Grand Forks, as it seemed that I was always adding another bit to support something or another.

And then the bright idea hit me! For Curlew, as it was all going to be completely flat so as to facilitate switching and not having cars roll away from the operators, I decided to try laminating two layers of large sheets of the same 3/8” plywood, in the hope that I could get a nice flat area for both the tracks and the town, all at once. All of the plywood these days seems to be warped in one way or another, and the best that I could find locally was no different. I have always glued the layers back to back, so that any warps should cancel out. This has worked very well on the thin strips used before, so I assumed that it would be similar for the larger sheets. Well, it seems that I was wrong. Once everything was glued and dried, I noticed that there were in fact a few places where it went up and down, and not just a tiny bit. My guess is that the dynamics of the wood are different along the edges than in the middle of the sheet as the worst spots are where there is a joint on one side or the other. The three big main pieces were all offset from one another to avoid having adjacent joints on both layers. While this works well with the thin strips, it seems that it doesn’t for large sheets. At one spot the two top sheets curl up a bit making a cusp, even though they were screwed and clamped together as flat as I could make it. Earlier posts show the lengths I went to to clamp them flat.

Ok, I thought, no real problem, as we can sort it all out when the plywood gets screwed down to the risers. They can be adjusted upwards and downwards to fix bumps or dips. John’s pictures show the effort that went into getting the many risers attached at just the right elevation, by using the laser level (thank you very much!). Again, it seemed like a good idea to all of us at the time…

I quickly realized that I needed some specialized tools to see just how warped the plywood was. I devised some targets for the laser level that show if any point on the plywood is at the correct height or not. I realized that using the laser on the edge of the plywood is not good enough when it is wide and may be cupped. I came up with a two step block idea where the goal is to split the beam on the edge, with the upper half of the beam hitting the second block. That one is set back a bit, so it is easy to see if the beam is only on the lower block, (too high), only on the upper block (too low), or split between the two (just right). I will make up some from high grade birch plywood so that they are all the exact same thickness. I had to hunt through many pieces of my 1×4 supply to find a bunch that were the same thickness, so it’s time for a special tool.

Movable laser leveling blocks

I found that as I slid these blocks around on the plywood surface, the laser beam was all over the place! The next picture shows three blocks, with the middle one indicating that the plywood humps up and back down over a short distance. Not great for standing cars on a siding!

A high spot in the warped plywood.

Well, I spent most of two days trying to get the whole mess reasonably level, and will have to rely on using a wood plane a lot in some spots to shave off the bumps. OK, so what about using the risers to compensate? Great idea in principle, and also in practice on narrow bits, but it fails miserably when dealing with such large pieces. The strength of the laminated plywood is impressive, so much so that it resists bending almost as much as the L-girders themselves. At one spot, I was trying to force it up, and all I was doing was bending the L-girder down. At that point, I realized that while some correction of the warps was possible, and I managed to do just that, it is not as flat as I would like it, and I (we?) will have to live with it. As I said, the wood plane should fix the worst spots. I did use my test truck and piece of track to see if anything would roll when parked, and in most places it doesn’t, so I think it will be OK. Oh, and the risers that the gang installed so carefully on Saturday? Almost all of them were moved in this process, I am afraid. Heavy sigh.

Getting there…

The next issue was the joints in the plywood. I had tried to arrange the pieces to minimize where the tracks would cross, but some were unavoidable. The edges of the plywood tended to curl a bit, making for some quite uneven seams. Out with my trusty plane and a lot of “elbow grease”, checking frequently.

Starting to flatten the joints.

In some places the entire top ply was removed, uncovering voids in the plywood that had to be filled.

Planing and filling the joints to make them flat.

In the end, it turned out pretty flat, but there are still a couple of places that may be a challenge to switch and not have the cars roll away. We shall see in due course.