By Craig Bisgeier
As of November 2003, the Do-It-Yourself I-beams (Click here for the original in-depth article) I built are well over three years old, and have been intentionally subjected for the last two years to a stress test. The two 16' long beams, supported only at their ends, were subjected to a load of over 1,000 lbs. stacked on or near the center with no additional support. That's a lot more weight than an evenly dispersed section of model railroad layout would have. I think the fact that the beam bridge is still standing after that says a lot of good things about the design. That said, there were a few minor design issues which can be easily fixed. Here's a list of problems that occurred, and their solutions.
The most worrisome problem I found after removing the weight was that one of the beams had started to twist about 4 feet from one end. Looking back, it isn't surprising this happened - it occurred where there were joints between all four of the 1x2 flange boards, and the area was also between cross-braces that stabilized the two beams against one another. This combination of flaws allowed the beam to flex at the weak point where it was unsupported. Fortunately the 3/8" plywood held and bent a bit but did not break, showing it was strong enough for the application even when the engineering failed.
Two things will fix this. First, it seems obvious now that more cross-bracing is needed, especially where there are weak points in the I-beams. From here on, all my beam bridges will get a rectangular cross brace every 4 feet instead of every 8 feet, which will place them at every potential weak point (breaks in either the flange boards or the web) along a beam. This will better support the beam and keep it from twisting at the joints. Even if you use longer flange boards and there is no flange joint to protect, I recommend placing cross-braces every 4 feet to keep the structure rigid.
Second, if I make any more beams using the 1x2 flange board method, I will stagger the breaks in the flange by a foot or so. This will mean every break in the flange on one side will be supported by a continuous flange on the other side, instead of allowing a weak point where two joints are right across the web from one another. This should minimize the flexing the beam can experience, and coupled with the extra cross-bracing, this flexing problem should no longer occur.
Another problem turned out to be the long unsupported span of some 15' between the legs. As initially built there was minimal sag on the beams even when a 300 lb. weight was placed in the center. But over time and with considerable weight concentrated at the center, the beams did take on a permanent sag of about 1.25". Under the +1000 lb. weight the sag was more than 2" before the load was removed.
With a smaller, better distributed load across the whole span, I doubt the sag would have been so pronounced (if it would have occurred at all), but just the same for strength I think it is probably a better idea to place a set of legs every 7-10 feet along a 'standard' do-it-yourself I-beam.
If I had a place where I had to have an unsupported span greater than 12 feet, I would probably consider making a beam with a thicker and deeper web (probably ½" plywood or greater, cut in widths 10" or greater) and heavier flange boards (1x3) and in greater lengths than 8 feet if possible. Of course, now you are getting close in size to the commercially made continuous I-beams, and if you can get them into your space that might be the better and simpler way to go.
The final problem was with sway. Under the stress load, when the bridge was bumped into from the side, the center of the structure would sway side to side up to 2" or more from center. With a full grid of benchwork fixed to the tops of the bridge, I think the beams would have been far less subject to these forces, but no one wants to take a chance on their scenery cracking or crumbling as the benchwork flexes over time.
The reason for this motion turned out to be inadequate diagonal bracing between the cross-braces between the beams and the sides of the I-beams. Actually there was very little diagonal bracing which was a mistake on my part. This will be corrected by doubling the bracing in each 'well' opening, creating an X-pattern made from 1x2's jammed tightly into the corners that will brace the bridge from flexing forces in either direction.
Finally, the 1x2 flange board method has worked well enough, but if I build any more I-beams I will use the previously discussed method of slotted 2x2's as flanges. I have had the lumber cut for this new type of I-beam for well over a year now, I just haven't gotten around to building any yet because the major part of the layout is on hold for now. But I believe it will be a better, stronger solution if you have the tools to cut a slot in 8 foot or longer 2x2's. (I made my 2x2's by cutting 2x4 studs down the center, yielding two 1.5" x 1.75" boards.) If and when I get around to making up some of these beams, I will create a new web page describing their construction and use.
So with a few small design corrections and some additional bracing and support, I think the do-it-yourself I-beam method has been a qualified success, and will continue to work well into the future of my railroad, and I hope yours too.