How do I do this the correct way?

[Alan Lightstone]

I can tell everyone how to do this wrong. When I first made this cover for the center of a fire pit table, I made a rookie mistake. After steam bending a 1.5" tall curved circular piece that is the base for the cypress circle cover, I glued the two pieces together along the entire circumference. :eek: Somehow forgot that wood moves. It's been my mantra since.

352390

No great surprise, in Florida with 90+ degree heat and 200% humidity, it wasn't too long until the cover turned into a warped pizza.

Now, three years later, parts of this have developed rot, and I've taken the piece apart, removed the rot, and am refinishing it now.

My question is, what is the correct way to connect the bottom rim to the round top. Only on two edges? I also now have the rim in 4 pieces (roughly 90 degrees of the circumference each), so they will need to be attached separated.

I can't think how this can be done with something like a breadboard, so what are my options so it doesn't just warp badly again?

[Jamie Buxton]

I would not use those carefully-bent pieces. Instead I'd saw the rim from solid wood. Imagine you are looking at the big slab just like you cut the top from. Saw a ring from it that is shaped like the rim pieces you bent up. Glue that ring to the top, observing the grain orientation of the ring and the top -- that is, the grain direction of the ring is the same as the grain direction as the top. Because the two grain directions are the same, when the wood wants to expand and contract, the ring and the top will not be fighting with each other.

[Andy Giddings]

Alan, on top of Jamie's comment, would it work just making it out of thicker material and milling a rebate around the circumference? Does the inside need to be hollow for some reason? Is the underneath just used for locating the top over the pit? If it is, why not use 3 or 4 dowel rod stubs for location purposes?

[Sam Murdoch]

OR - think of the rim as a table apron and the cover as the table top. Connect the two together in such a way as to allow the top to move (this you know - the reason for this post). Perhaps bend some stainless clips or best of all have a metal ring made with tabs to secure to the rim pieces. Fasten the clips (or the tabs on the ring) securely to the rim sections but through elongated holes where attached to the cover.

The elongated holes just need to be oriented perpendicular to the grain of the top so will be different from one rim piece to the other.

Hard to design a good system after the fact. Jamie's idea has merit but the same consideration applies - you need to consider grain orientation and the grain of the circular rim will not always be parallel to the top. I'd avoid glue but use well considered elongated holes - maybe through bolted using small carriage bolts. The heads of the bronze or stainless bolts become a detail. Would only need four fasteners if the rim is one solid piece.

Certainly worth saving - looks very nice.

[Bryan Lisowski]

I would laminate the pieces together and then cut the circle. Then again living most of my life in places that gets cold and snow, if I was in Florida, not sure I would have or need a fire pit.

[Alan Lightstone]

Alan, on top of Jamie's comment, would it work just making it out of thicker material and milling a rebate around the circumference? Does the inside need to be hollow for some reason? Is the underneath just used for locating the top over the pit? If it is, why not use 3 or 4 dowel rod stubs for location purposes?

It does need to be hollow, to clear the center of the fire pit table that has the fire glass and is a little elevated.

[Alan Lightstone]

Would multiple full-size pieces of, say, 3/4" or 1/2" wood, glued together to make a thicker piece expand and contract at the same rate, or would they split? I'm assuming that grain is oriented parallel, as well as possible. I could rout the last piece to remove a large circle from the middle, leaving, essentially a thicker edge, and thinner center.

[Prashun Patel]

why do you need the rim at all?

How about three cross-grain, sliding dovetailed or screwed in battens? They could still serve the purpose of registering the lid over the opening.

[Sam Murdoch]

If you are willing to glue up another rim - here is another thought. Glue it up with overlapping sections sawn into radii but make them the full height on the outside. Then rabbet the inner edge of the rim section so that the top simply sits in the center. Leave an 1/8" clearance all around the top to the rim. Set the top into the rim using an exterior caulk - will secure the top to the rim, keep water out and still be flexible enough to allow the top some freedom to move.

[Alan Lightstone]

If you are willing to glue up another rim - here is another thought. Glue it up with overlapping sections sawn into radii but make them the full height on the outside. Then rabbet the inner edge of the rim section so that the top simply sits in the center. Leave an 1/8" clearance all around the top to the rim. Set the top into the rim using an exterior caulk - will secure the top to the rim, keep water out and still be flexible enough to allow the top some freedom to move.

That's a very interesting thought, Sam. Hmmmm...

[Alan Lightstone]

why do you need the rim at all?

How about three cross-grain, sliding dovetailed or screwed in battens? They could still serve the purpose of registering the lid over the opening.
Not sure I understand what you mean by that, Prashun. Could you make a quick and dirty drawing of that?

[Prashun Patel]

Here's what I mean. Quick and dirty. Sorry. Each batten would go as far to the edge as your pit allows. Also, the thicker you can make them, the better. I would probably make them tapered if you're constrained around the edges.

[Alan Lightstone]

Oh, I get it now.

Interesting, but I think that would constantly kick up the sharp glass rocks when taken off the table, so I think it's a non-starter.

I thought of making many (maybe 30) small 1.5" tall battens after you mentioned them, to go just around the edge, pointing towards the center like rays eminating from the center point, but only for the last outside 1" or so of the disc, but that's also a ton of work. Shouldn't cause warping with wood movement, though.

[Alan Lightstone]

OK, tried a new approach (as well as fixing the old one for a temporary fix).

First, I used up all the teak I had lying around. Not a cheap solution, but at least it was stock I already had, didn't have to buy any, and is certainly good wood for an outdoor, wet, humid environment.

I used a modification of Prashun's approach, gluing on 36 small battens on the edge, each 10 degrees apart. They are short (1.5"), and only glued on with a small amount of epoxy each. I don't think there will be enough wood movement to cause the wood to snap off the battens, or crack, over such a short distance.

I tried a test piece with penetrating epoxy as an undercoat, but was underwhelmed at the absorption, so I abandoned that approach.

Spraying on System 3 Marine spar vanish now. Here's a few pictures of the work in progress.
354292354293354294

A few more coats of spar varnish to go, satin now, and it will be done. Time will tell if it will turn into a potato chip to. Pretty sure it won't rot, though.

[Jerry Miner]

Thanks for the update. Looks great. Should be safe movement-wise.

Are you sure that's teak? Looks more like walnut to me.

[Nick Decker]

Very nice. Looks like the solution is a cool design element as well.

[Alan Lightstone]

Didn't color correct the picture. Fluorescent lights screwed up the picture.

It's teak, for sure.

[Sam Murdoch]

I hope to be proven wrong Alan but potato chip it will be. Pretty though.
Sorry to be negative.

[Alan Lightstone]

Sam:

You may be right. But I'm getting better at building these. Time will tell.

[Alan Lightstone]

OK. The replacement has been built, and placed outdoors in my Florida torture chamber. Time will tell if Sam's potato chip prediction is true. I'm hoping he's wrong.

The teak I used this time is far, far heavier than the cypress it replaced. Makes taking it on and off a little more of a pain, but hopefully teak will be a more weather friendly wood for the constant sun and absurd humidity. I'm hoping the System Three spar varnish holds up to the conditions. It has thus far on the Titanic deck chair I built last year.

Here's the finished product. Flat, for now:
355660

[Alan Lightstone]

I hope to be proven wrong Alan but potato chip it will be. Pretty though.
Sorry to be negative.And the winner is:

Sam.

Potato chip it is. As you say, a pretty one, but quickly warped. Going to leave it (I'm sick of making those), if for no other reason as I'll be able to see how long the finish (System Three Spar Varnish) lasts outdoors.

[Prashun Patel]

The battens need to go across the whole width to resist the warp.

[Alan Lightstone]

The battens need to go across the whole width to resist the warp.Unfortunately, not possible due to the fire pit table's design.

Oh well...

[Prashun Patel]

Alas I recall you said that. Unfortunately unless the battens ar straight grained, thick, and traverse The grain of the top, they will just be along for the ride.

I have to say selfishly, though, this was educational for me. While I understand and abide by the science of wood movement, i don't have enough practical experience to know how much of the time it is critical. like everyone I get lucky sometimes and wonder if that's the norm or the exception.

I thought of a solution for you: remove those battens and convert them to tapered shims. Let the top warp away and just shim it where you need it !!!!!! ;)

[Alan Lightstone]

Funny. Practical but funny.

I had seriously thought of buying more teak, orienting and gluing it to the top making the piece thicker, then routing out the center circle.

That became a non starter as it would just amount to hundreds of dollars of teak just thrown away. Should have worked, though, regarding the warp.

[Sam Murdoch]

This was one I didn't care to "win". You've worked so hard on this one. I'm confident some version will succeed.
Steady on Alan :)

[Alan Lightstone]

What's really odd, is that this teak version warped far more substantially than the cypress one, and far quicker. No clue what to make of that.