Does anyone know of an analysis of deflection under load for L-girders and joists made of plywood or metal studs comparable to what Linn Westcott did for dimensional lumber in How to Build Model Railroad Benchwork Second Edition?
I can’t directly answer your question, sorry, but I thought I would point out that gluing/screwing/both elements of a plywood L-girder might be iffy due to the layered nature of the edge of the vertical component in plywood…and I say this not really knowing. I have assumed and read others stating that screwing into the edges of plywood is not a good idea. Screws should go into the side across the layers, not between them.
However, that vertical component, especially if 1/2" or thicker, should sag very little except across ridiculous lengths never contemplated by knowledgeable layout frame builders IMO. I could see thin metal house framing materials sagging a bit, but also on great lengths.
I used spruce 1X4 for my last layout’s I beams, and they are rock solid. The longest is 8’ long.
Crandell
Not to worry, it works great in practice. Adhesives are amazing, both those used to make the plywood and the wood glue used to put pieces of it together into L-girders.
As with nearly everything related to benchwork, much more time is spent in handwringing over variations, options and details than it would take to just build the dang thing.
As far as deflection goes on a 2" x 4" steel stud I don’t know if there is any. I built a 12’ x 22’ steel stud roof covered with coroplast for our dog runs. It is incredibly light but strong. I built it on the ground and three of us threw it up on top of the run after. I throw a sheet of plywood up there once a year to get up and wash it and it is solid. I have a similar roof on another dog run made out of 2" x 4"s and 1" x 4"s and it has a lot mushier feel to it when I am up on it.
I have used metal studs to build some unique projects with great success. They are cheap, very light weight, a hack saw goes through them like butter, they screw together easily with self tapping screws and are perfect for many applications in our hobby. However, they need to be braced properly or they twist, bend and crumple. For this reason I don’t think it would be a good choice for L girder construction. I would stick to wood.
A steel stud grid bench with foam on top goes a long way to removing heat and humidity issues for some people. Depending where one lives on our planet, it can be a good route to go.
I bet you can find the info to your question if you google it. I’m sorry I can only offer my experience using metal studs. It has been good. I used one as fire proof roadbed across the fireplace.
Brent[C):-)]
Brent
Tried that, thought it would be faster to ask here as the first 3 pages of results were no help.
Alan
I would expect plywood to be stronger than dimensional lumber for the same thickness. (Plywood joists for floors are thinner than dimensional lumber used for joists).
But not all plywood is equal either. Most of he stuff sold in big box stores is not that high a quality. Still, if you use 3/4" plywood it should be stronger. But I wouldn’t rely just on screws for the L girders, use glue too. The cross pieces which rest on the L girder will probably be okay with just screws.
Personally, I would use 13 ply Birch plywood, but you’ll probably need to find a lumberyard other than the big box stores for that. At least the big box stores in my area don’t have it. I have some 4 ft long bookshelves using it with no intermediate supports. They have never sagged in over 30 years no matter how many books and magazines I put on them.
Enjoy
Paul
Hey Doc
I have used both dimensional and plywood and since I have a table saw my preference is the plywood.
I was able to buy a good grade of 3/4" plywood and rip it for all the pieces I needed .
I used nails and glue to make the “L” beams ( in my case I made “C” beams) The legs were two pieces of the 3/4" to make a 3" X 3" leg.The joists were 3/4" X 3" with the band board at the end screwed on with 2 1/2" screws. I did predrill for the first inch so as not split the joist. I was pleased as to how little it cost me to do the extension this way. For levelers I shimmed the bench level, then dropped a short piece of 2"X2" inside the legs and ran four screws into the 2"X2". No metal levelers needed. Also notice leg to “C” beam joint.
A poor man’s way. [:-^]
Lee
![http://cs.trains.com/TRCCS/themes/trc/utility/[IMGhttp://i260.photobucket.com/albums/ii34/Yankeeflyer/test022.jpg[/IMG]:550:0](http://cs.trains.com/TRCCS/themes/trc/utility/%5BIMGhttp://i260.photobucket.com/albums/ii34/Yankeeflyer/test022.jpg%5B/IMG%5D:550:0){kind=link}
![http://cs.trains.com/TRCCS/themes/trc/utility/[IMGhttp://i260.photobucket.com/albums/ii34/Yankeeflyer/IMG_0509.jpg[/IMG]:550:0](http://cs.trains.com/TRCCS/themes/trc/utility/%5BIMGhttp://i260.photobucket.com/albums/ii34/Yankeeflyer/IMG_0509.jpg%5B/IMG%5D:550:0){kind=link}
Hey Doc
I have used both dimensional and plywood and since I have a table saw my preference is the plywood.
I was able to buy a good grade of 3/4" plywood and rip it for all the pieces I needed .
I used nails and glue to make the “L” beams ( in my case I made “C” beams) The legs were two pieces of the 3/4" to make a 3" X 3" leg.The joists were 3/4" X 3" with the band board at the end screwed on with 2 1/2" screws. I did predrill for the first inch so as not split the joist. I was pleased as to how little it cost me to do the extension this way. For levelers I shimmed the bench level, then dropped a short piece of 2"X2" inside the legs and ran four screws into the 2"X2". No metal levelers needed. Also notice leg to “C” beam joint.
A poor man’s way. [:-^]
Lee
I have used 3/4" ‘common birch’ plywood($25-33/sheet) for building dimensional lumber. I have the local building supply store cut the sheets into 1x2 and 1x4 pieces(about 25 cents/cut). I have built ‘L-Girders’ using the above material and it is a dream to work with. I glue up the edges of the material and use #8 wood screws(you really need to drill a pilot hole with a ‘screw-mate’). I have seen no problems with running screws into the 3/4" edge of the plywood lumber. Remember, it really is the glue that is holding the ‘L-Girder’ together - the wood screw is just a ‘clamp’ to hold everything together while the glue dries. I you are really cheap, I suppose you could back out the screws. I have spliced ‘L-Girders’ over 20’ long on my layout. They are supported by two ‘L’ brackets. I can climb onto the layout with no problems! I have not tried to measure ‘deflection’, but with plaster wrap hills 10-16" hign and over 20’ of plaster rock castings on a bluff, the layout does not appear to be suffering!
Jim
This isn’t directly on the original topic of deflection, but I thought it is a feature that differs between wood and steel worth mentioning as its closely related. When using wood for the joists, they can project a fair distance past the L-girder or other supporting system and still be strong and stiff. That’s because wood is very resistant to twisting forces.
If you have a steel joist end project out, it’s susceptible to twisting. Until you tie the end of the steel joist together with others via your fascia or other means, it’s going to be springier than the equivalent wood would be. My guess is that most people would use wood for the joists, but if you do happen to use steel, you’d want to take this into account in your design.
Around here (CT) the big box stores charge around $45 a sheet; I would rather have my carpenter rip the plywood rather than the guy at the big box store; but if the resulting “dimensional lumber” you ended up with is true, why not.
It’s not a question of being cheap, rather of not having hidden metal getting in the way of future construction. The glue should hold the joint by itself.
I wouldn’t get too hung up on the math as in part Westcott was simply trying to convey to a potentially (at the time) skeptical audience that if you use dimensional lumber in a conventional way you have to plan ahead for how it behaves under stress, and his L girder system takes advantage of the fact that wood likes to bend but not to twist, as an above poster points out., so making an L girder means it resists both because bending becomes twisting
That is why if you take a length of 1x4 lumber and rip it in half and make an L girder out of it, it will have an entirely different degree of strength and rigidity – same amount of wood, totally different properties. Westcott went into this in depth first because that was the kind of guy he was but also because he sensed, correctly I think, that the L girder system had not caught on as much as he intended it to.
As to plywood, I happened upon a large supply of 5/8" plywood when a friend of mine did a major renovation on his home and i got some of the old floor boards. I did not want it for the tops of the benchwork so it sat for a while, but then my wife wanted a series of tall tables for her art workshop. I followed an idea Jim Hediger published years ago about building an entire “train table” out of ripped plywood, with the legs being an L shaped piece.
I was hampered in that I did not have the best possible types of saws to make the straightest possible cuts, but even with my inadequate tools I remain impressed by the strength of the tables and legs, and screws into the sides of the plywood – what some guys regard as simply a no-n0 — for the L shaped legs and where sides meet ends, have held up just fine, with minimal glue by the way. To get to the point, as Jim B says, plywood can indeed be an alternative to what ordinarily most people would regard as a job for dimensional lumber.
Somewhat OT by the way, but as bad as some plywood is getting at the lumber yards, man the d
Even the plywood is of questionable quality. I usually reject 50% of the kiln-dried studs when buying. (All the new lumber going into the renovations is kiln-dried fir.)
The house I recently bought was built in 1938. The dimensional lumber (slightly bigger than today’s) is in great shape; relatively straight and true with negligible twisting.
I believe that he was also trying to make the point that benchwork tends to be way over built for what it has to do. And having a table of useful lengths for structural support makes planning benchwork more productive.
Still, I think it would be instructive for someone to actually assess how strong our current construction methods are in comparison.
Miehman is right on the spot here. I tried a “hybrid” metal and wood bench work, and found the bounce or spring to much less to my liking than similar wood construction. As I’ve noted elsewhere, that can remedied by a more permanent fixture.
If you’re interested, there’s a thread here with good info.
and I have some additional photos on my website. Just go to tips and tricks, and than bench work from the side menu.
Funny, I’ve been using steel studs everywhere that Linn Westcott used wood, have flying joist ends' as long as 10 inches and have never noticed any springiness.’ They WILL twist, but that’s easily detected and corrected when adding risers and subgrade.
Of course, I’m modeling in 1:80 scale. If you’re running 1:12 scale live steam your results may vary.
Chuck (Modeling Central Japan in September, 1964 - on steel stud benchwork)
Pure semantics, Chuck. “Springiness” was probably the wrong word to use in association with a discussion of deflection anyway. My point is they can be twisted, at least the steel studs I’ve seen, until the ends are tied down. You can’t do that with a stick of wood.
But if you say they can hang out and not run into issues, I’ll take your word for it. Someone else had results along the lines I suggested in my caution. Not saying it won’t work, just that it may require some additional thought in how to do it. I can’t say I have direct experience with steel benchwork, but my limited experience with steel studs tells me they’re not necessarily a direct substitute for wood in every application.
You might share some of your experience with steel stud benchwork, as it’s something that few have experience with and there’s little documentation in the hobby press about it.
From my experience, “springiness” was exactly the right word. As I’ve mentioned, I think it’s remediable, but there was a lot of bounce from my experiment.
Chuck was very helpful when I set up my test section, so having his input is very valuable in this case