Construction problem.

[Steve Maskery]

Ladies and Gentlemain, as Frankie Howerd might have said, no, look, listen, here.

I've designed this 'ere chair:

The way I have drawn it, the back is only 12mm thick. I could beef up the lower part so that it is 20mm thick or so at floor level, but that still means 16mm max at seat level. My problem is obvious, how on earth do I attach the rails to the back?

This is the joint that gets most strain, but I'm not at all convinced that a tenon that is only 12-16mm long would be sufficient.

This is a bedroom chair, not a dining chair so the back won't be seen like a dining chair back would, so if I have to use brass caps it won't be the end of the world, but I'd prefer not to have to use them, I'd like to do some reasonably trad joinery here if possible.

These are the sort of things I mean from Woodfit:

I have seen chairs built with laminated backs like this before, but I've no idea what joinery the makers used. Anyone have any expertise in this area? It will be a lot of work to get wrong!

Cheers
Steve

 

[OLD]

How about fixing a block between the rails where they meet the back you then have a much larger surface area at the interface between back and combined rails so lots of ways to make the connection including your tenons as there could be other hidden fixings and large glue area

 

[Steve Maskery]

Hmm, good idea Old, hadn't thought of that. I could have two through tenons and another couple of blind tenons in between. Or indeed, all through tenons, if I reckon I could get them all identical.

I'll think on.

Thanks very much.

Steve

 

[DaveL]

Steve,

I think you need a set of nice metal brackets and a big slice of perspex. :twisted: :wink:

 

[-David-]

Steve

I am no cabinet or furnitur maker, just a DIY'er. So with that in mind my notion might be 'silly'.
I would not rely on the rails, you have a good hidden void between the rails, and sufficient depth to fix a wide batten to the back and then fix the seat to that. I am sure that would take care of the downward forces. As far as the twisting forces I would still use the rails. I would dowell from the back to the ends of the rails and the batten, making a feature of the dowell ends.
Oh no!!
Just seen that OLD has got there before me.


David

 

[Steve Maskery]

Steve,

I think you need a set of nice metal brackets and a big slice of perspex. :twisted: :wink:

Absolutely!

Yes, I thought the back was lovely, shame about the rest, really. I didn't like the way he'd mounted it into a cylinder at the bottom either, I reckon it would be very uncomfortable for the back, let alone the bum.

I've seen several fan-shaped chairs and alwways liked them. I found a nice one in a hotel room in Sweden about 20 years ago and always fancied making that. This is a hybrid of several really.

Well spotted!

Cheers
Steve

PS If anyone is wonderering what on earth we are on about, we saw a chair last weekend made from some nice wood and some horrible Perspex.

 

[Chris Knight]

Steve,

That is a tricky one. I reckon that it would might be strong enough with just the rails tenoned into mortices in the back but OLD has made a good suggestion to enhance the bearing/gluing area.

I think the problem of potential failure is likely to be compounded by the back flexing in the joint region which could effectively multiply the leverage stresses at the joint very considerably.

I suggest you make quick full scale mock-up and test it to destruction. You could use a flat piece of ply for the back to simplify the mock-up's construction.

 

[jasonB]

You could use some form of metal fixings like the ones you have shown. Then hide the heads by routing out a 1mm deep rectangular recess and glueing in some end grain veneer, I won't tell anyone they are not through tennons, you could even do some fancy diagonal wedging

I once used this method on a table that had 50+ through tennons, eight of them took a lot of load so I used square metal bar epoxied in with a 3mm bit of endgrain to cover the hole and look like a tennon.

Jason

 

[Steve Maskery]

What if Iadopted an invisible mechanical approach? I could run 2 channels down the full length of each rail, bury 4 nuts in the back (before veneering, of course), and run 4 lengths of studding through the whole lot, tightening them up from the front of the cross-rail. If I made the seat so that it rolled down over the front rail, the tightening nuts wouldn't be seen.

?

Cheers
Steve

 

[Bean]

That would work Steve

 

[cambournepete]

What about using wedged tenons or wedged double tenons - would they be strong enough?

You could glue a block to the back and then joint into that - not sure if that suits your design though - which I rather like btw.

 

[engineer one]

my little engineering brain reminds me of one inevitable thing
Steve, people ALWAYS lean back on chairs, and by that i mean
go onto the back two legs and lean further back. in your case,
the very shape of the back, and its thickness will add to the
bending moment.

Firstly as Old has said, i am sure that a cross bar under the seat
is vital, but i also wonder whether you should consider an
angled bar from near the floor up to the front of the seat.
This would stop the seat bending at the front, which i think it would
with the front legs not further supported. I know it takes away
some of the initial beauty, but in these circumstances, i always think that
referring back to the mackintosh style chairs with high backs
is a useful lesson.

Your mechanical supports have some value, but i wonder whether the bending moment will not cause the ply to shatter because it of the
localised stress when tightened.

The design is great, but why do conventional chairs have lower
stretchers????? :-k :-k

I know you have more experience than me in actual construction,
but i think the bending moment is greater than you think it may be.

good luck
paul :wink:

 

[Scrit]

Steve

Have you considered bonding-in something like some BigHead fasteners in the back and using these with your threaded stay rods? These fasteners were actually designed to be embedded into fibreglass or plastics on boats, etc and would probably work quite well in your situation as the point load on the joints would be spread onto the full area of the plate (and they make some fairly large plates). If you then end the threaded rods in 40mm diameter holes on the inside of the rails it would be possible to use one end of a kitchen worktop connector (in fact just the long hex nut and half-moon bushing) to provide a completely hidden joining system with a lot of strength

Scrit

 

[Steve Maskery]

Wow, Scrit that is EXCELLENT! I didn't know such things existed!

There is just the thing in the photo, but I can't find it in their catalogue. Methinks a little phone call is in order.

Thanks very much indeed.

Cheers
Steve

 

[Colin C]

Hi Steve
I think you want tot look at an antique hall chair and see how they are fixed.
What about fixing a rail between the back rails ( dovetailing it to both rails and you can still have through tenons )
you could also put a surporting peice at the back to help strengthen the joint and also make a feature of it.
If that is clear :?
What was wrong with that perspex chair :^o 8-[
You are right it did look unfinished

 

[engineer one]

steve,
i am still concerned about the back breaking around the seat fixing, and no matter what you do with the fixing, it will still be very weak there.

so i have another thought to make it easier, i think.

can you not also bend the back a little in the horizontal plane, so that it is part of a cylinder in both planes. in that way you will truly strengthen the whole back, and you could use 12 mm all the way.

as for the tenons, how about through ones, and tusks???

i do not think the fixing is the problem, i am convinced it is the
bending moment.

paul :wink:

 

[OLD]

I think that engineer one may have a point about stress areas however we all tend to over engineer things, but the answers may come from research into furniture of the 1960 where bent ply and minimal lines were common i believe, although i don't have any information to hand.

 

[engineer one]

i agree that woodworkers tend to "over-engineer",
however ,in this litigious society,
we have to be careful because the client NEVER tells the
truth, by that i mean they will tell you they will use the
item in the way designed, but they never do.

i only caution that you look carefully at the relevant areas,
and old is right to suggest looking at the plywood furniture of
the 50's and 60's

paul :wink:

 

[Nick W]

I think I'd join a solid piece of timber between the rear of the two rails, probably tenoned into them. Shape that piece carefully and then glue and screw through that into the back. It should give quite a large area of long-grain to long-grain join. And yes, taper the back so that it is 16mm at the join area. Maybe even have the back an even 20mm up to the join area and taper it from there up?

 

[Steve Maskery]

Thank you all for your excellent suggestions.

I'd like to keep it as clean as possible, so I thnk tusk tenons are probably not what I want - but thanks for the idea.

Looking at the furniture of the 50s and 60s is a good one, I'll have to do some research.

I've just ordered the catalogue from BigHead. I like this idea a lot. The biggest problem with it will be mounting it so that it is in line with the rails, rather than perpendicular to the face. But maybe some of the ones on a strap will be bendable enough to be moulded to the shape of the back.

I agree making the back cupped would increase its stiffness. Good idea. But it does make the joinery even more difficult. Hmm, tricky one.

Anyway lots to think about, thank you all very much indeed.

Cheers
Steve

PS Neil, you've let me down. I had a bet with myself that you'd suggest Miller Dowels! :wink: