LumberJocks

All Replies on Torsion Box Workbench base--a new (I think) idea. Opinions please.

  • Advertise with us
View Walker's profile

Torsion Box Workbench base--a new (I think) idea. Opinions please.

by Walker
posted 02-09-2017 12:57 AM


14 replies so far

View clin's profile

clin

1070 posts in 1531 days


#1 posted 02-09-2017 03:56 AM

What you are doing is similar to a recent build I made. Here’s a link to that project.

Click for details

In my case I built the base first, then used an approach similar to the Wood Whisperer’s to create a dead flat build surface. Just using my base rather than saw horses.

Making a bunch of straight 2×4’s etc is a lot of work itself. More than building the torsion box itself. So I understand your interest in trying to build the base flat and use that as the reference surface to build the torsion box. However, I don’t think this is as good as a temporary surface, made with many supports. The reason is that your base won’t have enough supports to properly support the layer of MDF that is the initial build surface. Also, I assume your top will overhang the base. If this overhang is significant, again, the reference surface for the torsion box won’t be supported as well.

Of course, not being supported “as well’, doesn’t mean it isn’t supported well enough. You could always build your base, lay the MDF on it and lay a straight edge and see what you have. But I think you’ll be surprised at just how much it can sag even across a relatively short span.

As I see it, the whole point of building a torsion box for a worktop is to create a very flat top. Torsion boxes are also relatively light weight, but typically we want workbenches as heavy as possible. Point is, if you aren’t trying to build the top flat to within, say <20>ll never get the hardboard top and side boards to be perfectly aligned. So you end up with a little surface lip, here and there, that can make it hard to slide things off the top. At some point, I plan to trim my side board flush to the MDF top of the torsion box and cut a new hardboard top that goes all the way to the edges.

-- Clin

View Dave G's profile

Dave G

335 posts in 2583 days


#2 posted 02-09-2017 10:15 AM

When a flat reference is needed I usually build upside down on my cast iron table saw top.

Eventually you may consider building a Roubo bench.

-- Dave, New England - “We are made to persist. that's how we find out who we are.” ― Tobias Wolff

View Robert Stockwell's profile

Robert Stockwell

8 posts in 1092 days


#3 posted 02-09-2017 11:01 AM

Thanks Walker, It’s a great post

View HokieKen's profile (online now)

HokieKen

11288 posts in 1674 days


#4 posted 02-09-2017 01:22 PM

I’d say that very few of us have flat floors in our shops. Most of us are working in our garages or basements which can’t really have flat floors for drainage purposes. So, you’re not alone there! This is why we build our benches upside down.

I think what your proposing will work if done right. First thing would be to throw the idea of softwood aprons out. Construction lumber will move within minutes of milling. MDF is a good choice IMO. I have used it for a lot of tool tables and it’s my favorite for a stable, reliable, flat surface. The key is to keep it dry. Liquid water will kill it quick and the core material is porous so in highly humid environments it may distort. IME though, even in VA where we have pretty big swings in humidity from summer to winter, it’s stable. To ensure stability though, you can seal it up so all sides have equivalent moisture barriers. Many simply use paint for this. I personally leave the faces like they are and laminate formica or hardwood to the cut edges.

I would skin the bottom of your rails. With the span you have, you’re going to have sag. The stability of a torsion box comes from the top and bottom skins preventing that sag. Your torsion box top will act as your top skin but without the bottom skin, you’ll have a harder time getting it flat to begin with.

If this is going to serve as a workbench for any hand tool work (and I assume it will with dog holes and holdfasts) you’re probably going to want to beef your stretchers up. I’d go with the 4×4s you’re using for the legs and pinned mortise and tenon joints. I think a lot of people underestimate the racking forces that happen when you’re working wood with a handplane or saw or pounding away with mortise chisels. Half-laps would probably work as well.

My only other suggestion would be on your pipe supports. I like the idea and your friend is right, they pipes will support the weight just fine. The only issue I see is if you use MDF. That core is soft and easily deformed. Eventually, the MDF will give way and wallow itself out on the pipes. Maybe once you get the rails where you want them, add some cleats on the legs underneath to support them. That way the weight is not all resting on the tangency to the pipes but over a larger area.

Good luck with your build and welcome to lumberjocks!

-- Kenny, SW VA, Go Hokies!!!

View Walker's profile

Walker

160 posts in 1007 days


#5 posted 02-09-2017 02:23 PM

Thanks for the replies so far!

@Clin: I see that you combined both ideas. Building the base first, but then still doing the 2×4’s like in the fine woodworking article. Were you not confident your base was flat or was it mostly the overhang referenced causing initial sag? My table will be sort of an odd size because of where I need to put it. It will be the full 96in long but only 32 wide. The rails on the base would be about 80×24. That leaves 8 in overhang on each end the long way, and 4in overhang each side the short way. Like you said though, I could put my sheet of mdf down and if it does sag, then I could proceed the way you did. I do want it to be flat as possible given its double duty. It will be used for gluing edge joints.

Also noted about the floating hardboard top not working so well. I think since I want the dog holes to stay aligned, I have to countersink at least a couple of screw in it anyway. Maybe cut the sideboard a small bit short of the top (1/8th?) so the inevitable ‘lip’ is the other way? That would allow things to slide off and not catch, but aesthetic OCD might kick in.

@Dave G: The Roubo bench looks interesting. Perhaps if future circumstances allow space for an assembly table and a workbench, I’ll make the Roubo. I am interested in the cross-brace scissor deal going on for the leg vise. I was toying around with a similar idea for bracing my leg posts. After all, they’ll only be about 15” apart (measured inside of each 4×4). If I get to it later, I’ll make a sketchup pic to explain better what I’m thinking.

@HokieKen: You make good points about the rail material, but they only enhance my dilemma. : ) If sealed well, the MDF will move less. (so don’t use softwood) However, over time it may wallow into the pipes (so softwood would be better?). A nice hardwood is the answer, but I just can’t spend that much money. I was also considering the compressive strength of the rail. Yellow Pine has a much higher strength that the MDF. Your idea for cleats is a perfect compromise though! I could use hardwood for those, or even some steel plates if I could salvage some. For the record I’m in Rochester, NY and the humidity in my basement seems relatively stable.

I also like the idea of skinning the bottom of the rail. Now that you mention it, the doubled up longer rail is like a mini torsion box in itself. The torsion box itself, will have a top and bottom skin. Planning on 1/2” MDF for both skins and the half lapped webbing (accept where the dog holes will be, those will be wood backed).

The stretchers remain a point of concern for me. In the past, I’ve not been very successful making anything straight, square, or flush. Thus the reason I’m trying to upgrade to more precise tools and methods, the bench being a great place to start. I’m not very good at mortise and tenons, I’m worried messing up the stretchers will throw the whole thing off square. Plus there is the need to move the workbench and then re-level it. I struggling to find the right words, but would a 4×4 mortised stretcher be too solid? Meaning could each leg still adjust height independently?

This has been very helpful so far. Thanks for brainstorming with me!

-- ~Walker

View clin's profile

clin

1070 posts in 1531 days


#6 posted 02-09-2017 05:02 PM



Thanks for the replies so far!

@Clin: I see that you combined both ideas. Building the base first, but then still doing the 2×4 s like in the fine woodworking article. Were you not confident your base was flat or was it mostly the overhang referenced causing initial sag? My table will be sort of an odd size because of where I need to put it. It will be the full 96in long but only 32 wide. The rails on the base would be about 80×24. That leaves 8 in overhang on each end the long way, and 4in overhang each side the short way. Like you said though, I could put my sheet of mdf down and if it does sag, then I could proceed the way you did. I do want it to be flat as possible given its double duty. It will be used for gluing edge joints.

Also noted about the floating hardboard top not working so well. I think since I want the dog holes to stay aligned, I have to countersink at least a couple of screw in it anyway. Maybe cut the sideboard a small bit short of the top (1/8th?) so the inevitable lip is the other way? That would allow things to slide off and not catch, but aesthetic OCD might kick in.

- Walker

My base came out very square. But again, there’s enough unsupported space, front to back, between your rails that your first piece of MDF you lay on it will sag a bit. How much is too much? Depends on on flay you want it. As mentioned, no reason not to just do that and check it. If the sag is acceptable, go ahead with the torsion box build.

To some degree any sag in this initial build surface is reduced by the webs. Since the webs will not follow every dip. Though if you assemble with a bunch of short webs, like the Wood Whisperer’s technique, the webs will follow the build surface more.

In my case, I assembled the two back legs, and back rails/apron. These are M&T joints. I did the same for the front two legs. Though there is not a bottom rail on my bench, at least not attached to the front legs.

Once I had the back and front I stood these up and clamped side panels to them. Using the leg levelers, I adjusted until the front and back aprons were in the same plane. Used a long level and winding sticks to get this true. At that point I screwed the side panels to the legs. I later added doublers/stretchers to the panel for added stiffness and to create a lip to support the lower shelf and lower rail.

I still had to screw down the hardboard here and there. That stuff isn’t necessarily as flat as you think. I realize it looks nice recessed inside the sideboards. And as you said, you could leave the sideboards a little shy or low of the hardboard top. But you could still catch this sharp lip. While if run all the way to the edge, you would typically ease the edge.

Again. I see no reason to build a complex rail that is itself a torsion box. Or, to even expect it to carry any significant weight. Two or more bolts at each corner is a much better joint than a single 1” pipe. You need resistance to racking, the pipe gives you no resistance to racking. Anything you do that is good for racking is going to be plenty strong for any weight the rail might transfer to the leg. And the torsion box is more than capable of carry all the weight and transferring it directly to the legs.

If the top were thin, like a table top, then sure the aprons support the top. But that is just not necessary with a torsion box top. It of course doesn’t hurt to allow the apron to support the top. If you expect to set really heavy weights (like stacks of anvils) in the center of your 8’ long bench, then the top will sag enough to transfer some weight to the aprons. But even then, I’m not sure it maters. I.E., does it matter if the top sags 1/16” when you have 800 lbs of anvils stacked in the center of it? And is it that much better if the aprons reduce this sag by 1/64”.

-- Clin

View Woodknack's profile

Woodknack

12924 posts in 2915 days


#7 posted 02-09-2017 06:26 PM

I didn’t read the walls of text but you have a design problem in that your aprons appear to join the legs with metal fasteners and that isn’t enough to prevent racking. You need a shoulder to make it really rigid, like a mortise and tenon, or in your case I would use lap joints. If you forgo the aprons then the top needs the same kind of connection that won’t allow the legs to to wiggle. Parden if this was already covered.

-- Rick M, http://thewoodknack.blogspot.com/

View dgapilot's profile

dgapilot

4 posts in 1007 days


#8 posted 02-09-2017 06:26 PM

Great idea. One thing I see right away, is how are you going to keep the entire base from racking over? If you have pivots with the pipe, and you are working with 2 parallelograms forming a cube, the frame can move in either of 2 directions. Triangles are your friend! I would suggest building assemblies that are rectangles, check for square using trammel points and secure them together with either a web or diagonal rails. Once you have 2 perfectly square rectangular frames, then join them with the cross pieces. If the frames are square. you can level them individually, then it’s just a case of leveling the cross braces.

View Walker's profile

Walker

160 posts in 1007 days


#9 posted 02-09-2017 11:41 PM

wait, so I can’t store my stacks of anvils on this? Then what am I going to do with these?

-- ~Walker

View Walker's profile

Walker

160 posts in 1007 days


#10 posted 02-10-2017 12:55 AM

The whole theory I’m going for here is to intentionally allow racking, at least until things are level. The pipes would purposely allow independent movement of all four posts and rails, and once they are level lock things in. I just need to find a clever, reversible process to lock things down. My initial thought was end caps on the pipes (perhaps with washers) could be tightened to stiffen up the connection (I know that’s not really illustrated in my pics) I don’t know how tight I could get these though. Time to buy a few test pieces and do some experiments. HokieKen’s idea for cleats I think would work well. I could also just screw the rails in also, once they’re set. I had an idea for an adjustable X-bracing, illustrated (extremely poorly) below. Attached with bolts/washers/wingnuts through slots in the posts.

Though now I’m leaning toward something closer to Clin’s build. Constructing the leg assemblies for each end, and get those matched first. Again, I’m just not confident in my ability to make a mortise and tenon that stays true. It sounds like you were able to adjust the feet well enough, even with your apron/stretchers clamped on.

I realize I’m on the precipice of over-engineering. As I stated in the first post, I am trying to reinvent the wheel; proven methods already exist. But my mom says I’m special and it’s ok to be different! Ok, all joking aside, help me reign this in to something practical.

-- ~Walker

View clin's profile

clin

1070 posts in 1531 days


#11 posted 02-10-2017 03:10 AM

Walker,

I think you are creating a very complex design just to address squaring up the base at the time of assembly. Regardless of the exact design, just clamp things as you square things up. Since you are putting on leg levelers, you just lightly clamp, adjust leveler, loosen or tighten clamp. Check for square/level, rinse and repeat.

Once built, you levelers will account for variations in the floor.

Also, many of the joints can be firmly fitted, before final assembly. You do not need all the joints to be moveable. As I mentioned before, I assembled the back legs to their rails, and the front to the front rail. Just glued it as square as I could. You can’t really screw that up as the only line that matters is the rail. And that is straight no matter which way the legs point.

Then when you join the front to the back, you just need to get the front and back rails in the same plane. You can even firmly attached these at one end. And then you only need to adjust the other end. I highly recommend just using plywood panels as the ends. These make racking of the ends almost impossible. And once secured, you can still double them or add whatever additional structure you want.

As far as making M&T joints. A workbench is a good thing to practice on. You could have the mortise be a thru-mortise, and be exposed at the top of the legs. That way the mortise is just a slot in the top end of the leg. You can cut the sides of the mortise with a hand saw and then chip out the material.

Many of ways to cut a large tenon. Numerous crosscuts on a table saw and then some chiseling and parring is pretty easy. Or just cut them old school by hand.

Even if you get a bit sloppy, they will still be quite strong. The real strength of the M&T is where the shoulder hits the leg. A great way to create a very tight M&T is to do a drawbore. Lot’s of examples on line. This is a form of putting dowels or pegs through the joint to hold it. But, you offset the holes in the tenon about 1/16” so the peg pulls the tenon in extremely tightly. You don’t even need glue, though why not. This technique is holding up on structures made hundreds if not thousands of years ago.

As mentioned in my project post, I made my base by laminating plywood to form the legs and frame members. Mortises are easily created by leaving appropriate gaps in the wood. And similarly tenons by offsetting the wood properly. Gluing all that up is a lot of work. But you get some nice fitting joints and the final structure is spooky rigid.

My table weighs ~300 lbs. I use a jack to take the weight off the legs to adjust levelers. If I raise one leg 1/16” the other leg, at the same end goes up the same amount. No detectable flex in the structure at all.

-- Clin

View HokieKen's profile (online now)

HokieKen

11288 posts in 1674 days


#12 posted 02-10-2017 01:02 PM

I think you’re going for too much adjustability. It’ll bite you in the ass down the line. Like Clin said^clamps are your friend. Once you have everything squared and flat and where you want it, then go back and start replacing clamps with glue/screws.

X-braces are fine but if you use good joinery for the stretchers, you don’t need them. If M&T is too much to bite off, then lap joints will work. Like Clin said though, this is a good place to practice M&T and if they’re a little sloppy, you can wedge them or use some dowels to cross-pin them and still have a solid joint.

I’m an engineer so I feel where you’re coming from :-) But, you’re creating a lot of complexity that will, most likely, end up being a detriment. You want this bench solid when you’re done. The more “floating” joints you have, the less rigidity you have.

-- Kenny, SW VA, Go Hokies!!!

View Woodknack's profile

Woodknack

12924 posts in 2915 days


#13 posted 02-10-2017 06:29 PM


I realize I m on the precipice of over-engineering.
- Walker

No, you’re over the edge. :) My suggestion to new woodworkers is always, “Don’t try to innovate until you understand why things are built as they are now. Furniture design evolved over centuries and joinery has been tested and perfected, it works.” I think your idea for a torsion box benchtop is fine but your base is broken. It’s causing problems not solving them. And cross bracing on a bench or table screams “non-woodworker”. X-bracing is for metalwork, the only reason to use it on a workbench is to fix a poorly built one.


...I m just not confident in my ability to make a mortise and tenon that stays true.
- Walker

You could easily use a lap joint which is just as strong and will prevent racking. But the fear of making a M&T will always be there until you make one. And woodworking is forgiving if you use established joinery.


...purposely allow independent movement of all four posts and rails, and once they are level lock things in.
- Walker

To help break you into the lingo, carpenters and masons build level, woodworkers build square. (Notice I didn’t say carpenters build square, as anyone home knows, lol) <kidding> You want to build the bench square and then level it using feet or shims.

-- Rick M, http://thewoodknack.blogspot.com/

View Robert's profile

Robert

3554 posts in 2016 days


#14 posted 02-10-2017 10:53 PM

I built the top first, then the base.
Walker,

I’m sure you could build the base first, you would just have to be sure it is perfectly level, square, flat, planar, etc. which might be a bit of a hassle.

I actually built the top first by set up two sawhorses with 3 2×4’s screwed down and then levelled and used winding sticks to eliminate twist.

Once this is set, then lay the top sheet down first and cover with plastic to prevent glue stick. Use this surface to arrange and assemble the torsion components.

After the torsion assembly is dry, lay the bottom ply on and fasten with glue and screws. Then flip the structure over and repeat for the top. The WoodWhisperer has a pretty good video on this.

Hope this helps.

-- Everything is a prototype thats why its one of a kind!!

Have your say...

You must be signed in to reply.

DISCLAIMER: Any posts on LJ are posted by individuals acting in their own right and do not necessarily reflect the views of LJ. LJ will not be held liable for the actions of any user.

Latest Projects | Latest Blog Entries | Latest Forum Topics

HomeRefurbers.com