Need input on creating this laminated coffee table base

[Chris Merriam]

My next project is going to be a coffee table. We've settled on this design from Restoration Hardware. The trick is their table is actually made out of metal: an iron base and a sheet metal top. I want to make the whole thing out of wood (currently thinking walnut).

I've never done bent lamination before, been reading plenty about it, but thought I'd throw this out for any opinions before I seriously get started. You can see that each leg of the base is basically two U-shapes. Well, one's a "u" and one's an "n" I guess.

I was thinking of laminating each piece like this: /````\ there'd be a couple inches completely flat, then the curves would kick in. Then I would glue up two of the u-shaped pieces (along the flat edge) to form one leg of the table.

Do you think the curves are doable via lamination? Do you think the leg will be strong enough?

In the pic below I imported the design into sketchup and did my best to get valid measurements of key dimensions (like the height and radius of the curves).

Thanks for any input!249766

[Jerry Miner]

Hi Chris---

I've done a fair amount of bent lamination and I think this project looks quite do-able. The radii look fairly tight, so you will want pretty thin layers. Prep a piece or two and see how easily it conforms to your form. Err on the side of thinner pieces. And don't forget about springback.

Strength won't be an issue at all, as I see it. You will need to create some rigidity between the leg sections.
Overall, it looks like a good project for an introduction to bent lamination. Have fun and good luck.

[Chris Merriam]

Thanks Jerry, glad to hear I have a green light to proceed! Good thing you mentioned rigidity between the leg sections, I hadn't even put any thought into that yet. I was only thinking how easy it would be to duplicate what they did with the iron rod stretcher. But since I'm working with wood, and not having the steel top welded to the base, this thing won't have much racking resistance. I might need to come up with some kind of low profile apron system. Hmm, maybe two long 1x1 rails right beneath the top, connecting the tops of the opposing leg assemblies. That would give me a total of three rails between the legs.

That might work out better because that will give me a better way to attach the table top. Right now I was thinking of screwing down through elongated holes in the top and into the four points of the legs. The top will be 1.5in thick, and I was going to recess the screw heads, then cover them with some large contrasting wood caps, lightly pillowed like Greene&Greene, but much larger, maybe two or 3 inches square to look proportional to the top.

[Jerry Miner]

249876249877I think you should think a little harder about racking strength. A 1 x 1 apron won't offer much. I don't want to highjack your design, but you might consider an angle brace, or even an arched brace----something along the lines of the attached sketches

[Chris Merriam]

Ooh I like that second one!! That'll be a challenge to make. couple questions though:

Does the stretcher attach to the bottom side of the top? By what means, and how will it allow for wood movement?

how much wiggle room will the stretcher allow across the curve? Seems like a pretty tall order to glue up something that long and get it to precisely meet in three points: the two legs and the top.

thanks for the help on this Jerry.

[Sam Murdoch]

I like the 2nd one too. I don't think you would need to attach the stretcher to the top. Still, the top of the stretcher could meet the underside of the top. Make the beam long of course and cut it to fit. If the beam is as thick as your legs I think flexing would be negligible, leaning to none. Attach to the legs with tenons or dowels even, though I'm inclined to the tenon for best gluing and rack resistance. And as Jerry cautioned - don't forget about springback - especially for the leg pieces. Great project!

[Chris Merriam]

I'm not used to thinking in three dimensions with these curves. That makes perfect sense to build the beam extra long then just trim to fit, plus as long as I make the base first then there is plenty of wiggle room to determine the final length of the top. I was shooting for 67in long but who cares if it ends up +\- an inch or two.

Santa brought me a Domino for Christmas, so I was already thinking I would use that to join the stretcher to the legs. I'll just need to step up to the biggest tenon size they have and probably double up the tenons on each end.

[Dave Richards]

FWIW, if the stretcher is attached in the middle of the top, that's not going to be an issue for wood movement of the top. The place you need to deal with that is out at the tops of the legs.

FWIW, I had a little play with your images in SketchUp, Chris. You probably want to rework the dimensions from what you got because some of them are not right. You could get closer with PhotoMatch although the image has been modified after it was shot so things don't line up perfectly. Still, you can get closer than what you've got.

[Chris Merriam]

Thanks Dave, I figured sketchup would have some kind of dimensioning utility like that, but I googled it for only about one minute and didn't see anything so I just winged it. I knew the height width and depth for sure so I just scaled the images to match.

I'll have another go at it using photo match, thanks for the tip.

[Jamie Buxton]

You can certainly use bent lamination to make those big C shapes. However, when you attempt to cut that flat on the C, things may get a little wonky on you.
Here's the deal.... In general, the more laminates you have in a bent lamination, the less springback. Conversely, fewer laminates results in more springback. When you cut the flat into the C, you're reducing the number of laminates in that area, so there will be more springback there. That is, the C will try to straighten out.

[Chris Merriam]

I was thinking best case I would laminate them in that shape (with the flats transitioning to curves). If that isn't possible then I'd do the C shape and cut the flat as you mentioned, but I didn't realize that it would springback even after being laminated and cured. I'll have to cross my fingers that option 1 will work.

[Jerry Miner]

Does the stretcher attach to the bottom side of the top? By what means, and how will it allow for wood movement?

Yes, I would attach it: dowel, screw, loose tenon (Domino), .... Wood movement here is not a problem---it attaches at only one place to the top so does not restrict wood movement (as Dave said in post #8)


how much wiggle room will the stretcher allow across the curve? Seems like a pretty tall order to glue up something that long and get it to precisely meet in three points: the two legs and the top.

I think you get it now that the arch is built long and trimmed to size after.

Jamie is right about the C changing shape after the flat at the joint is cut. The answer, IMO is to make the C's overbent and over-long, trim the meeting flat first, then determine the cuts at the ends of the "arms" of the C last.

This will be a great learning experience in bent lamination, and part of the beauty is the dimensions are not written in stone, so you can be a little flexible with radius and overall width of the leg assemblies. Keep track of how it all goes, and when you do your next bent lamination project, you will have significant experience!

Have fun, and let us know how it goes

[Jerry Miner]

I was thinking best case I would laminate them in that shape (with the flats transitioning to curves)

Chris--- I think you'd be better off bending the ends in a continuous curve and cutting the flats after. Design-wise, you want the inside of the C to be a continuous curve. Making the form transition from curve to flat on the outside and keep a continuous curve on the inside could be problematic. I would keep this project as simple as possible: simple, continuous curves for the laminations, with flats and joints cut after.

[Chris Merriam]

Lol I completely missed the obvious there, thanks for setting me straight! I certainly want a nice curve on the inside.

I'm going to show your sketch to the wife tomorrow for approval then I'll hop into sketchup and get this thing drawn out for real.

What do you recommend for the form material on this: mdf, plywood?

I understand some forms can be crushed, not sure if my case is that extreme or not.

[Jerry Miner]

I usually use mdf because it's relatively cheap, stable, and easy to work with. But you can use almost anything. Particleboard is a little cheaper, and fine for this. Plywood is good, too. I usually use a liner of something smooth---could be masonite, an extra laminate, Formica, acrylic, ....

Provide a place for clamping when you make the form----expect to use LOTS of clamps.

[Dave Richards]

As far as those flats on the arches are concerned, the top arch could have a cleat attached to the ends after they are trimmed. This will give you a way to attach the top and if you install it before cutting the flat, it'll prevent the spring back. You could attach a temporary cleat to the bottom arch, cut the flat attach the two arches and then remove the cleat. Everything should be kept under control.

http://farm9.staticflickr.com/8498/8337778934_ffe3cf7388_z.jpg (http://www.flickr.com/photos/8294157@N08/8337778934/)

I didn't spend any time on this but hopefully it gets the idea across.

I was thinking on my way to work that maybe it would make sense to add some screws in addition to the dominos. You could bore in from below and add plugs made of matching wood to cover the holes. Something like this:

http://farm9.staticflickr.com/8079/8337961268_e16f7f58ba_z.jpg (http://www.flickr.com/photos/8294157@N08/8337961268/)

[Mel Fulks]

I would use solid wood ,not laminated. Each end of the table would be made of four pcs. There are old ones made that way. If you do laminate it ,spring back will be slight,since so many layers will be needed. That style is a Roman thing,can't think of the name right now.

[Chris Merriam]

Thanks for the additional input. Dave, some good thoughts on reinforcing with screws, or maybe even some dowels? The cleat idea would simplify things greatly, it kind of takes away from the design of the original a bit, but then again, nobody will ever really see it. I'll have to ponder on that on for a while. It sure feels a lot more secure than just attaching the top at the four corners alone though.

Mel, I started down the solid wood route first, but then read that in these cases a lamination will actually be stronger than a solid piece. It seems like a solid piece would always have the grain in the wrong direction at some point.

[Dave Richards]

Chris, how were you planning to attach the top? It seems to me that the cleat would make it easier to do because the cleat can be attached with a couple of screws at each end into the upper arch and then the top attached to that. The cleat won't be visible if you make the edges of the top thicker as I indicated in my sketches. The top os supposed to be shown in section so there'd be the same sort of lip on the end.

As to the solid wood route, I expect Mel is thinking of something laminated from wideer boards. Maybe you'd make it three layers thick. I did that for the keel on the first sailboat I built.

http://img.photobucket.com/albums/v506/weekender410/Weekender/51702c.jpg

Lousy picture but the best I can come up with. I don't know if you can see it but the pieces overlap at the "corners" so there's no real short grain anywhere and it is very strong.

http://farm4.staticflickr.com/3323/3516738758_fb2a46b516.jpg (http://www.flickr.com/photos/8294157@N08/3516738758/)
The felloes on my steering wheel are essentially done the same way with the joints staggered.

[Jamie Buxton]

If you attempt to reproduce the original design exactly, you'd need to mill flats on the bent-laminated C's. That might give you trouble. Instead, you could change the design a bit. Bent-laminate the C's, position them back-to-back, and bandsaw a piece from solid lumber to fit between them.

[Mel Fulks]

As beefy as your design is I don't think that's a problem,most of the old ones are slimmer.Certainly it can be done either way.

[Chris Merriam]

I was going to make the top out of solid 8/4, planed down to 1.5in or so, so a lip wouldn't work unless I went with a 3/4 top. That's certainly a possibility, I was just trying to get away from 3/4 dimensions on everything, and I want this to have a really solid feel.

This is still a mental work in progress, but I was thinking of attaching the top by putting threaded inserts in each of the four legs, then running a hefty lag bolt down through a counterbored elongated hole in the top and into the insert. At each of the insertion points I was thinking of making a large (2 or 3 inches square) cap to cover the connection, like a greene and green plug, but not so pointy and tall, maybe just rounded over a bit. That would give the top a little bit of decoration. But certainly a cleat would be much easier and secure than doing all that.

As far as solid wood, I understand laminating a wide board to make it thicker, but with a u shape you'd end up with the grain going in the wrong direction along some part of the u. If the grain is horizontal then the base of the U is nice and strong, but the legs are weaker, and vice versa. Or am I over-thinking it?

That is one nice looking steering wheel!

[Prashun Patel]

Chris, I'm with Mel. Having done bent lam on an oval coffee table that required flats on the ends, I can tell you that it's not very fun. You need some good jigs, and honestly, unless it fits perfectly, you risk some tension at the joints. Mine actually cracked. I went with a solid wood configuration as a fix.

If you lay out your circle like the 'steering wheel' it'll be plenty strong - especially with the dominos. However, here's another option:

'bricklay' the arches. If your legs are say 1.5" thick, use 2 layers of 3/4" 'steering wheel' glue-ups with the joints staggered. This will be the strongest. If you are persnickety, it might require some grain matching of your pieces, but it's totally doable. The nice thing about this kind of configuration is that you'll basically be working with octagons, which will make squaring your ends up for joinery a snap. Only after all joints are mortised should you bandsaw it to shape.

If you like the arched stretcher (I do) I would also bricklay or 'steering wheel' that.

I wouldn't be overly concerned about racking or grain direction. Any of the designs you've posted should (in my experience) be plenty stable.

I think bricklay will give you the most stable and easiest to work with solution.

[Dave Richards]

Chris, the thing I was trying to describe is something like this:
http://farm9.staticflickr.com/8083/8338640024_b5559b6183.jpg (http://www.flickr.com/photos/8294157@N08/8338640024/)

There's no real short grain or weak points in a glue up like this. It's is very strong and it would be easier to build than the bent laminations. And there'd be no spring back to worry about.

If you do make the top as I showed, I would make the top thicker anyway. Maybe an inch or inch and a quarter. I would do the Marc Singer trick for hiding the end grain, too. If you make the top full thickness all the way across, you could excavate for the cleats so they don't show.

To me, with the style of the table, having plugs over the bolts would seem distracting. Just my thought.

[Chris Merriam]

Dave, you are a sketchup wizard, that is very clear now and I was thinking of the completely wrong thing in my head earlier. I think you've also reflected the bricklaying that Prashun recommends. That definitely seems to be the way to go on this. If I make the diagonal parts a bit wider it will also give me room to carve out the foot at the end of each leg rather than milling and attaching it separately.

Then according to Prashun's recommendation I'll do the stretcher the same way, just with a different layout/set of angles.

As far as the top, I was thinking the same thing on embedding the cleat a bit, that might be the best way to go as well. Just need to leave a little room for cleat seasonal movement.

Not familiar with Marc's end grain trick, I'll have to look that one up later today.

Thanks so much for everyone's help on this, I feel a lot better about this design now!

[Prashun Patel]

You are a Sketchup master. I wish I could sketch like that!!! Hats off, Dave.

[Dave Richards]

Thank you, Gentlemen.:o

Chris,

Here's an example of the "Marc Singer" method I was referring to. Sorry about the disgraceful end grain material. I don't have anything good in the way of end grains on this computer.

So the Marc Singer method involves making the top over length and over width. You then rip the sides off with miters and cut the end the same way. Then you glue the pieces back on as in the lower version and presto! No visible end grain and no issues with seasonal movement. You also get the appearance of a thicker top without the cost of thicker wood.

http://farm9.staticflickr.com/8499/8338930266_29d99bc97b.jpg (http://www.flickr.com/photos/8294157@N08/8338930266/)

Hopefully it makes sense from the image that the top is shown upside down.

I tried to find a photo or two showing this where I thought they might be but came up empty handed so I doodled this out. I saw in person a coffee table Marc did this way and it looked amazing.

[Chris Merriam]

Thanks for the drawing, very helpful. I never would have thought of that, my brain just doesn't work that way. Using Marc's way would also hide the cleat. That might just be the best way to go and I'll forego having a true 1.5in thick top.

I was also toying with the idea of doing a little inlay (my first time at that too), just something simple like a little black epoxy border on the top. Maybe after the miters dry I can just drop the inlay right on top of that glue line to hide any imperfections.

[Dave Richards]

That inlay thing could be cool. You could cut a rabbet on the edges and fill them with epoxy. Then, once the epoxy has cured, run a router around with a chamfer bit and put a bevel on it.

[Prashun Patel]

I do a lot of gluing up of 3/4 stock to make wider things like legs. I will tell you that on a table top, it's a little tricky. You will spend (or should spend) good time flattening and jointing the panels so they glue up well (not to mention the bevel trick below). After a couple years, I find that the effort to do all this is rarely worth the savings of the extra price of the 5/4 or 8/4 stock to begin with. On the top, I'd just go thick to begin with.

[Chris Merriam]

Two quick follow up questions guys:

When laminating the boards together, is it necessary to use 3 layers or would 2 be ok? I'm probably going to use 8/4 stock so ill have plenty of thickness.

And, what's the official name of this bricklaying technique? All my googling returns bent lamination. I'd like to learn more about this style of work.

Thanks!

[Dave Richards]

You could probably get away with 2 layers just fine. There's a plan for a smaller boat similar to the sailoat I built with the keel made that way. It only uses two layers. Three just seems more "right" to me. Especially so if the grain will show. The center layer could even be a little thinner than the other two if you wanted.

Official name? Good question. I've seen the method used for making arches and wagon wheels as well as curved stair rails and other things. A client of mine built a trestle table I designed for him and the legs are built up that way. We've got another one that will include benches with matching legs and he'll build all the legs that way again. In the case of the three layers it's sort of a built up mitered bridle joint. If two, it would be half lap joints.

By the way, when I made the keel for my boat, I shaped one outside layer to the desired shape. Then I glued the other layers to it and used a router with a pattern trimming bit (bearing on the top) to cut as deep as I Could. I flipped it over and used a flush trim bit (bearing on the end) to finish it up. In your case, since you'll need to make several matching ones, you'll probably want to start with a patter but you might find it useful to clean them up the same way.

[Prashun Patel]

I used an article on FWW (requires subscription, unfortunately). Just search 'Federal Card Table'

It's typically used curved table aprons, which can then be veneered to cover the seams to give the appearance of something that's steam bent.

http://www.sawmillcreek.org/showthread.php?191800-Elliptical-Ash-Coffee-Table

[Chris Merriam]

Thanks for the replies. There does seem to be something magical about using 3 plys, it would just mean extra resawing, but we're not talking about any significant lengths, so I'll probably do it.

Prashun that's a great table, I read the original post a few months back. I'm really trying to get out of the "everything square" casework look and try the curvy stuff that you do so well.

[Dave Richards]

Chris, it sounds to me as if you're ready to build a boat.:D

[Prashun Patel]

Chris, thanks. It's easier than it looks. Curved work allows me to mask my joinery imperfections.

One piece of advice: think twice about resawing the 8/4. If you can cheaply get extra 4/4, I'd do it out of that instead. Resawing 8/4 for aesthetic reasons is one thing, but to break it down to make 4/4 is IMVHO a waste of good 8/4, and possibly unnecessary effort.

[Chris Merriam]

Lol, no need for a boat yet! I'd have nothing to do with it other than to mount it on the wall and stare at it.

I started to do some math to figure out my wood needs (I don't keep any on-hand for the most part). The top alone is going to cost me $150, ouch! The 8/4 comes in 8ft lengths, so if I use 63 inches for the top, that leaves maybe 25in or so left over. It's going to take 4 or 5 boards, so I'll have a good bit left over, so that's what I was planning on using to laminate the legs. Honestly, with the small local mill I just found, the difference in cost for 8/4 vs 4/4 walnut is not too much, $3.75 vs $3.25, which seems to be a pretty darn good price. The bigger local place I normally go to wants $5.75 for 4/4 walnut. I will however pick up a couple 4/4 boards to do the arched stretcher. I guess I could save the cutoff 8/4 for some later project, but honestly I just don't like dealing with tons of scrap and oddly sized pieces. I don't ever really build anything small (maybe I should do some boxes) so it just sits in the corner of the shop and stares at me.

[Dave Richards]

Make an accurate SketchUp model of the table and work out the parts you'll need to make the legs. It might be that you could get away with a number of shorter pieces for the inner layer if you make it out of three layers. This would reduce the required width of the stock you'll need for the middle layer. I agree with Prashun about not resawing the 8/4. Save it for something like normal table legs. It's a shame to waste more walnut than you have to.

[Chris Merriam]

Well it's been a few weeks so I thought I'd post a status update and an additional question for the experts.

I pretty much did everything we discussed. Got some 8/4 walnut for the top, and 4/4 for all the other parts. The legs and stretcher ended up being two plies of 4/4, because the rough stock was generously cut, it was really 1.25in thick, and after jointing I got it down to about 7/8 or so. So two layers gave a nice fat leg. I spent a long time in Sketchup and at the workbench trying to figure out how to best glue up the leg stock to provide a smooth grain flow. My end result is just "ok", nothing like Dave's fancy boat steering wheel though. I think the main issue in the end was the available raw walnut I had to work with, and the fact that I didn't buy any extra. I just couldn't afford to be super picky about the stock I bought. Here's some pics of the legs:
254115254116

I was pretty leery about building the stretcher correctly, but I modeled it in Sketchup and printed out a full size template. Taped all the sheets of paper together and just traced it right on to my blank. I probably should have cut the domino mortises before I finished roughing out the blank, but I managed to make do after the fact. The stretcher-to-leg angle is 60 degrees (+/- a degree), set the domino fence to 60 and went at it. I used two dominoes in the stretcher, an 8x50 and a 6x40. The 6x40 is only in the stretcher for 12mm, the other 28 is in the leg. I cut those last night, crossed my fingers and did a dry fit, came out perfect. Here are the final pics of the stretcher and base:
254117254118254119

So that's where I am today. It's time to build a cleat, attach the legs to the cleat, then attach the cleat to the tabletop. That's where I'm stuck. Can't wrap my head around the best way to attach the legs to the cleat to allow for a strong anti-racking connection. Any advice? Couple points:

1. The tops of the legs are about 2x3 inches square, and they are end grain (ugh)

2. I'm shooting on a max cleat thickness of no more than one inch

3. Current options I'm considering:
a. Dominoes (I can fit two of the largest (8x50) in each leg
b. Hanger bolt bored into the leg and then attached via some method into the cleat. I found a product at tablelegs.com that might do the trick:254120

4. I was basically counting on the stretcher not providing all the racking resistance. But maybe it's stronger than I think?

Thanks for any advice guys! And thanks for getting me this far!

[Prashun Patel]

Wow, looks great!!!
If the cleat is 1" thick, you'll have no problem. If it's a little wider than the leg base, then screw it into the tops of the legs, then screw the bottom of the cleat into the table at it's widest part. Just elongate the holes, and then snug it down.

[Dave Richards]

It looks really good Chris. thanks for the update. I agree with Prashun about the cleats.

[Chris Merriam]

Thanks guys. I guess my worst nightmare is someone bumping/kicking the end of the table and it folds over and collapses, so maybe I was trying to over-engineer it a bit. Any recommendations for leg-to-cleat screw size, length, and count? I can probably fit 2-4 screws per leg, maybe some 3in deck screws? Or maybe a nice thick #10 screw, not sure how long those go though.

[Prashun Patel]

Here's the thing to remember: No matter how beefy you make your base, a wide table top will exert a good amount of torque on your base. While you may get the cleat to stay perfectly snug to the top, the legs may end up rocking a bit on an 'apronless' table like this. To really resist that, you need to cross up the legs with some kind of bracing.

That being said, it has not been my experience that a coffee table requires a deep connection. I've made tables with cleats thinner than 1" and relatively short screws.

I would use 3 screws on each cleat: One in the center of each cleat, screwed tight and two on the ends, tightened through elongated holes. Watch the table for a year and note the movement. If it's severe, shim between the cleat and the top. If you are using furniture pads on the feet, you can do it there too.

If you are feeling ambitious, you can recess the cleat in an elongated mortise in the top. This way, any shims will be invisible.

If you are feeling even more ambitious and your design likes it, you can make your cleat a sliding dovetail in the top, and use a single screw in the center of each cleat.

[Chris Merriam]

Interesting. I was worried about everything below the cleat, while you are concerned with what's above it, lol! Never thought about torque before, and this will be a heavy 63x23 1.5 inch thick top. In addition to the cleat helping keep things flat, I'm also seriously leaning towards breadboard ends as well. Kind of redundant, but I wanted to hide the end grain and add a little detail to the top.

A sliding dovetail cleat, very interesting. Would like to try it but never done a sliding dovetail yet. I'm too scared to try it on this project!

[Jerry Miner]

Chris--- That table is really looking good. I don't think you need to overthink the cleat-to-leg connection. A couple #10 screws would be plenty (hanger bolt won't really do any more for you, IMHO)---1 1/2" penetration into the leg should be enough. Dominoes would work, too, but if it were me, I'd keep it simple with a couple of screws.

Will you be attaching the stretcher to the underside of the top? It will give you lots of racking resistance. I predict a very stout and stable table. Good work!

[Chris Merriam]

Thanks Jerry, I agree on the #10 screws, that'll be the way I go. When I mentioned earlier that I kind of winged it on the stretcher I really meant it! I mocked it up in Sketchup and built it exactly to spec. But apparently because of my poor Sketchup skills I made the arc too tall! So, I NEED a thick cleat to give me more room for the stretcher, and I'm still over by about 1/4 inch. Rather than take stock away from the center of the stretcher I was thinking of boring out a little rectangular trench in the top. So as it works out, I can indeed have the stretcher mount to the top. I can probably run a couple screws from under the stretcher right into the top. They won't be seen, but maybe I'll go ahead and fill them with matching plugs too.

Thanks for the input. I wish I could get back to the exciting construction details, but right now I'm in the agonizing throes of sanding all those curves. My poor job of smoothing my leg template is really coming back to bite me. But in my defense, it's really hard to see the lumpy curve in MDF. Once I cut into that dark walnut though, I have bands of alternating light and dark colors that tell me exactly where all my dips are. I've spent the last two nights on just one of the legs, ugh!