I was reading another thread on cantilevered shelves which reminded me that I wanted to explore building a large (12′ x 32″) workbench/saw stand. I’m cramped for space and I want to be able to push my cabinet saw and 8″ jointer underneath the bench when they’re not in use.
I’ve been thinking about doing it as a torsion box skinned with G1S plywood. However, I have not been able to come up with a good way to attach it to the wall. French cleats might work if the box were narrow, but if the shelf is 32″ wide AND supporting a mortiser, drill press and my 10″ slider a cleat just won’t cut it. The only option I’ve come up with that keeps the underside clear is to use an end bracket that goes UP from the ends of the bench at a 45-degree angle and which is anchored to the studs. Even this method doesn’t deal with the inevitable sag I’ll get in the middle of a long span.
In the unlikely event that I’ve been clear in describing my problem, please feel free to bombard me with other suggestions and comments.
If I’m not clear, feel free to respond with a simple “HUH?”
Replies
You're asking a lot to cantilever a workbench. For a given load, the deflection goes as the square of the length of the cantilever, so a 32" bench will deflect seven times as much as a 12" shelf for the same load.
How about some swing-out legs? They would be hinged to the wall, and you'd swing them out to support the bench, then swing them back to make room for the machines. Of course, you'd have to remember not to put anything too heavy on the bench top while the legs were folded.
-Steve
12 feet long? I think you have room for some legs that will still allow a TS and Jointer to fit underneath. With some legs and either a thick wood or torsion box style top the spans should not be a problem. Remember that your jointer (at 8" I assume the bed is like 6' or so) may well be able to go in and out of its stall on a diagonal to get part of the bed behind a set of legs - i.e., it does not have to be pushed stratight in and out at its max width.
Edited 12/4/2007 2:21 pm ET by Samson
How about up? Can you attach a wire from the outside corners to the rafters or to the top corner of the wall? Depending on your width, you could repeat the wire every 4 feet or so or combine with one leg in the middle.
Edited 12/4/2007 2:47 pm by GettinTher
I've just checked the wall space where I plan to put the bench and I have 15' to work with. My Unisaw is roughly 7' in width, and my jointer is also 7'. This means, in theory, that I can use corner posts and a center post and still park both items. However, I'm still thinking about a cantilevered design, if only because it would be an engineering challenge and cool to do!
How about the following: Pull off the drywall, sister the studs at 32" intervals, then drill a 1" horizontal hole through the sistered 2x4. Put a matching hole through a beam built into the torsion box and then insert (epoxy?) a 1" steel bar 30" long into the stud and slide the entire workbench onto the steel bars? Assuming I can get all the holes aligned this appears to offer the necessary strength without support from underneath.
If you make the sister studs out of hard maple or something similar, you might have a chance. If they're softwood, you'll simply crush the wood fibers and force the holes out of round when you put a load on the bench. The compressive strength of ordinary stud material is only a few hundred psi. Given your 32" lever arm, it would be easy to reach that limit with only a moderate amount of load.
-Steve
Dang! It would be nice to suspend the laws of physics just once! It has also occurred to me that if I have a bench at 36", then mount my 10" SCMS the boards I'm cutting will be at about 40"...a mite too high, I think. Back to the drawing boards!I enjoy this part of woodworking! Solving problems is half the fun.
Who says all your bench surfaces along its length must be the same height? Most jointers are significantly lower than 36"; is yours? The SCMS could be in a dip that made its cutting table even with the height of the bench over the TS. But you're crazy not to use legs. Unnecessarily complicated engineering violates law of parsimony, not to mention tempts the law of gravity. ;-)
Edited 12/4/2007 7:58 pm ET by Samson
I can do parsimony...just ask my kids! I know that, in the end, I'll do something logical and simple but since I haven't started ripping any walls apart or cutting any wood it's an interesting design exercise. Also, this is more fun than firing up the snowblower and dealing with the 15" of snow we've had dumped on us in the last 24 hours!
Ron,
What is at the ends of your 15' run...can you sister in a parallel beam ?
Ron,
What's on the other side of the wall? Is it possible to add down bracing on the other side of the wall? I am thinking you could extend the ends of the planks for the torsion top back thru the wall then add diagonal braces down to the ends on the other side of the wall. The wall becomes a fulcrum.
Just a thought.
Regards,
Bob @ Kidderville Acres
A Woodworkers mind should be the sharpest tool in the shop!
Edited 12/5/2007 10:07 am ET by KiddervilleAcres
I've seen something similar done with a free floating buffet counter, except it used simple 1/2" pipe into the studs and backside of the buffet. You could walk on that sucker without it flexing (there was a pipe in every stud).
Bob
One way to achieve solid support with minimal space interference is to use 1/2" pipe stanchions from the hardware store. They are easily adjustable for height and can be relocated if the size of the units stored underneath changes.
Frosty
"I sometimes think we consider the good fortune of the early bird and overlook the bad fortune of the early worm." FDR - 1922
OK, here is a trick from my other hobby (model railroads). if you want to minimize the legs/supports you can first instead of using legs, put an angle support from the front of the table (underside) to the bottom of the wall. This keeps the legs out of the way. You still have supports but no legs. No kicking legs, no hurt toes. Also this is more stable (the whole triangle thing)
Next trick. Take some 1x and make an "L" or even a "T" shape out of it. If you have a longer span then you have 1x you can stager the joint in the two parts, but remember that the part that is on the bottom leg (you use these inverted) need to be reinforced at the joint as that part is under tension, while the top does not as it is in compression. You will be amazed at the strength this give you. I did this and with two of these spanned about 12 feet of my railroad and they were strong enough that I could sit in the middle of the layout with minimal deflection and I weighed in at about 220 at the time. So if you did this (I used 1x4, screwed and glued) and only spanned say 6' or so (enough to get each tool under the bench) I would expect that you could make a strong bench. It may not hold a car engine block, but it should work for wood working. I would suggest extra support if you know you will be beating on something with a hammer. (locate that if possible over a leg)
Doug
Actually this is areally easy problem to solve.
Since you don't mind ripping off the drywall take a peice of 1x2 and 1/8" thick angle iron. In your sister stud, cut a 1" deep kerf, extend the angle iron from the kerf to the end of your work bench (or get a 36" piece of angle iron, 32 for bench + 1/2 for drywall + 3 1/2 for stud) put these on each stud. Build a torsion box out of 1/2 ply on the bottom and 1/2" mdf on top. Use honeycomb cardboard for the filler, amazingly strong stuff.
You'll be spreading your load across a bunch of supports and each by itself will have plenty of holding power.
I build my lumber racks this way and only have 4 supports per 8'and have no problems with deflection.
A few prblems though. Fixing the drywall will be a a PITA, as you mentioned the height, and the price of the angle iron, your looking at a few hundred!
"I build my lumber racks this way and only have 4 supports per 8' and have no problems with deflection."
This isn't a good comparison, for a number of reasons:
1) You can tolerate much more deflection in a lumber rack than in a workbench. No one would notice 1/8" or even 1/4" in a lumber rack, but it would be a significant problem at the front edge of a workbench.
2) The arms on your lumber rack presumably aren't 32" long. The lever arm is proportional to the length of the arm. So, for example, given the same supporting structure, a 32" arm would deflect more than 2-1/2 times as much as a 12" arm. Likewise, the torque exerted by the arm is directly proportional to the length. A 50lb. load at the end of a 32" arm puts more than 2-1/2 times as much stress on the support than a 50lb. load at the end of a 12" arm.
3) Deflection within the arm itself goes as the square of the length. So, a 32" arm deflects seven times as much as a 12" arm for the same load. This is in addition to the deflection in (2).
It's an engineering truism that cantilevers are hard. And they get much, much harder as they get longer, so an engineering solution that works for a short cantilever is very likely to be impractical for one that's even just twice as long.
-Steve
Okay, I will agree that there is some deflection in the steel. And being 32" long increases the loading But that does not mean it can't support the weight. I am not suggesting an old piece of bed frame I am suggesting a pretty strong piece of steel.
Using a piece of angle iron the size and in the way I suggest will basically be using it as an I beam. Of course we will have to ignore stresses across the face and side to side since this is not an I beam, but being fixed in a bench we're close enough. But for the purposes of vertical loading all of the math is the same. We are interested in achiving a surface that is flat with only a minimal amount of deflection. let's say 1/16 of an inch max at the 32" point. This of course assumes all of the load will be distributed across the surface of the bench, now since we don't know exactly what will be on the bench lets make some more assumptions.
The weight of the bench itself will be 200#, the various tools mounted to the bench, 1000#, random wood etc. on the bench 1000#. total weight 2200#
So before we do the math we know the following:ldepth of bench 32", width of bench 15', size of angle iron support 1x2x1/8" thick, min of 11 supports spaced 16", total weight on bench #2200.
to make this easier let's assume the weight is evenly distributed across the bench so: 2200#/11 supports = 200# per support (boy that worked out well :)
our beam is size :2, weight perfoot: 2.3, depth of section: 2, flange width: 1, made of steel who's Modulus of elasticity = 30,000,000
The properties that determine the strength of a beam are its section modulus and moment of inertia. For our beam (2WF x2.3#) those properties are:
S = 1.3332 in.^3 I = 2.6664 in.^4
So the bending moment in the beam is:
M = 200# x 33 in = 6600 in.#
The 33 comes from adding the 32" to the distance from the flange (top mounting edge) to the tip of the beams center line.
The stress in the steel is then found:
S = 6600 in.# / 1.3332 in.^3 = 4950.5 psi
The yield strength of structural steel is normally 36,000 psi.Comparing 4950.5 to 36,000 you can see that there is no danger in doing any permanent damage to our support.
As we mentioned the beam will deflect some when the bench is loaded, let's find out how much. The formula for calculating the deflection is:
deflection = (weight x (distance from mounting point ^ 3)/8(Modulus of elasticity x moment of inertia) or:
d = (W x l^3)/8EI) So:
d = (200 x (32 ^3)/8(30,000,000 x 2.664) = .01025
This is less than 1/64 of an inch.
so if our allowable deflection is 1/16 we solve for W
W = d8EI/l^3 = 1180#
or for all 11 supports we can have a total of 12980# total weight for the bench and stuff on it.
Now this also assumes everything is going to be a dead load, and since he will be working on this bench etc, I would consider this more of a live load and try to stay at the 200# limit. But I am going to have to go with my original statement that this would meet his goal, of having a usable cantalevered bench.
I also admit I thought the weight it could support would be higher. You definetly could not make these supports out of wood.
And although the deflection is 7 times greater, than a 12" arm I would have to say the total amount of deflection is insignifigant.
I still think the big problems are the height of this bench, and the total cost. I am 6'5" and this bench would be to high for me, and if built to my design would run over $500.
Also all of the formulas and constants I used can be found in engineering books. If anyone is so inclined feel free to check my math. I am a software engineer and not a mechanical one, so if I did make a mistake that's my excuse!
1/16th inch spring at the outer edge??? Acceptable on a workbench??? Not for me.
How much deflection would the bench have with (the much simpler, cheaper, and practical alternative) legs? Something approaching the compression limits of the bench material and floor, I'd guess.
Just to clarify the calculation showed:
d = (200 x (32 ^3)/8(30,000,000 x 2.664) = .01025
This is less than 1/64 of an inch. but it is greater than 1/128.
to get the 1/16 he would have to park 2 cars on the bench, and really that's not bad for a cantalevered bench with 2 cars on it.
Also If you have a hardwood bench it probably moves from time to time and I don't think 1/16 across 32 inchs is alot. But that's just my personal opion. but much more than that and I would flatten it again. But then again some of my friend are using old work benchs that not only have a small spring but have dips, rolls, and lots of wear marks, and the furniture they produce is amazing!
I agree that the steel itself is of lesser concern here. Given steel's much higher strength, it's not too hard to construct a steel beam that won't deflect under the expected loads. (However, I still say that statements such as "It worked for my 12" lumber rack" are insufficient evidence to suggest that it will work in this case.)
But the cantilevered beam itself is only part of the problem. The mounting of the beam to the wall is more likely to be the weak link, primarily because of wood's rather low compressive strength. (I think I said this already, but I may not have made it clear.)
If you insert a steel pipe into a drilled hole, or bolt a steel angle to the side of a stud, or capture the flange in a saw kerf in the stud, you're concentrating all of the stress onto a relatively small cross-sectional area of the wood. And this is where the differences between a 12" shelf and a 32" workbench are really going to come into play. Anything you can do to spread that stress over a larger area will increase your chances of success. That's why I (and others) suggested the use of an L-shaped steel bracket, where the vertical portion of the L is attached to the exposed edge of the stud. This also appears to be what Ikea uses in their design, mentioned previously in this thread (or maybe it was the other cantilevered shelf thread--I can't remember which is which anymore).
-Steve
Your right I missed your first reply the wood would be the limiting factor. Once bolted to the wall it probably would crush the fibers and the whole bench would sag considerably. see above post for some alternatives.
WOW!! Thanks for the analysis! I'm also a software engineer by training (admittedly 40 years ago when computers were fixable and programs were written with either an assembler or (believe it!)machine language. I still remember my first prof starting his Day 1 lecture with the statement "As is obvious to the most casual observer" and then losing me entirely for at least 6 weeks.I should point out that this will be a secondary workbench and the most critical tool will be my 1) SCMS and 2) my mortiser. I built my primary bench from scratch based upon plans from Lee Valley, and it is my day-to-day workhorse. I think I can live with a 1/16" deflection since the bench won't be used for assembly.If I were to make my sistered studs a minimum of 16" in length, and weld a 12" piece of 1 x 2 angle right angles to the 32" angle iron, and centered on the back of the stud, would this be sufficient to offset the torgue exerted at the mounting point to the wall? I'm also intrigued by the idea of having the bench height at two different heights i.e., left side at 36" (over the tablesaw) and right side at 30" over the jointer. If I use the entire 15' of available space this should still leave room to mount the SCMS near the center of the bench. I could then build a simple box extension for the right side to act as a support for cutoffs.This is fun! You guys have given me some great ideas!!
Edited 12/6/2007 11:52 am ET by RonInOttawa
I think for a dead load all of this would work. I know the calulations are a lot harder for a live load. When you attach it to a wall I would worry about the bolts ripping through the studs bringing the shelf down. Also there will probably be compression of the wood and over time the shelf would sag. You would have to trap the studs in angle iron somehow. Again this would drive the cost up quite abit.
Would it be possible to use some overhead supports as well as the walls and leg? I.e if you had supports from the top in the "Dead" Area behind your SCMS and a leg in the middle.
My calculations proved the angle iron could hold it, but that doesn't mean it's a good idea! I proved a lot of bad ideas will work in my lifetime.
I added a quick pic of how I would build it. The angled braces are attached to the wall and could even be worked into dustcollection.
And you should add more legs and braces if possible
This forum post is now archived. Commenting has been disabled