How should I construct this shelf?

[el_Pedr0]

Hi all.

The drawer unit I'm designing has a shelf with an inlaid glass panel. I'm planning to make this cabinet out of walnut veneered mdf for the panels and solid walnut for various parts where the edge will be visible.

My provisional plan for this shelf is to make it out of four lengths of solid walnut with a 45 degree joint at each corner. Is that the most logical way to build it?
If so, what should the joints be? Some kind of mortise and tenon joint? - I'd prefer not to see the joint on the front edge of the shelf or from the top.

Proposed dimensions:
Thickness of wood: 19mm
Width of each length: 56mm plus 10mm lip to hold glass
glass: 725 x 377 x 6mm toughened
overall shelf dims: 837 x 489

 

[recipio]

The glass will determine the overall dimensions as obviously it can't be cut. You seem to be making a glass topped drawer - or is it a shelf ? Its hard to see how you can avoid seeing the joint - either a simple butt joint with dowels will work or a mitred joint if you can insert a small biscuit to reinforce it. Correct me if I am wrong. !

 

[Droogs]

I would make a M/T frame and have a groove with entry from the back in the carcass

 

[TheTiddles]

Mitre with a spline and drop the glass in a rebate on the top, basically a picture frame horizontal

 

[CaptainBudget]

I'd go with something like Droog's suggestion, mitred joints will always be weak even with a spline and I don't know what you intend to put on these shelves...

Below sketch is how I'd do it, the short edges are covered by the case sides so no visible end grain. The small squares are where I would put Mortice and Tenon joints. This frame will not fail without a fight provided the joints are cut accurately.

A Housing joint in each side of the carcass will locate this frame and be more than strong enough even in MDF

For the glass itself I would go with one of two options:

1. Rebate a groove for the glass panel and gule the frame up around the panel, issue is if the glass gets damaged/broken you can't replace it

2. Cut an open rabbet and drop the glass panel in from above after glue-up, either leaving as-is (so it can be pushed out from underneath if required as per Tiddle's suggestion) or nailing a bead on top to fix in place.

Hope this helps...

 

[thetyreman]

I'd be tempted to make something like this but plan for having a rebate for the glass.

 

[el_Pedr0]

OK. Great. So mortice and tenon or mitre and spline. Sounds like M/T has the advantage of strength - even though this is just for handbags and such rather than my power tools. But would one suit a contemporary design more than the other?

I'd be tempted to make something like this but plan for having a rebate for the glass.

I do like the look of this joint. But the fact that you can see its construction from the front edge is not in keeping with the rest of the wardrobe. No other part of the wardrobe would have that kind of detail.

have a groove with entry from the back in the carcass

Interesting. I had assumed that I'd be doing something like cam and dowl to attach it to the carcass. Or screwing it in from the outside of the carcass, because the sides will be hidden when the cupboard is fitted. But these kinds of decisions shamefully reveal that I'm more familiar with the construction of flat pack furniture than bespoke cabinetry

The glass will determine the overall dimensions as obviously it can't be cut.

I'll have the glass cut to size by the glass supplier.

You seem to be making a glass topped drawer - or is it a shelf ?

This bit is a shelf with inlaid glass. When you look down through the glass, you see the contents of the top drawer. But the shelf itself is fixed - it doesn't move with the drawer.

Its hard to see how you can avoid seeing the joint - either a simple butt joint with dowels will work or a mitred joint if you can insert a small biscuit to reinforce it.

Sorry - first post was a bit misleading. I don't mind the line between the two components - but I want to avoid seeing the construction of the joint.

 

[MARK.B.]

Could you just make it slightly under size and add a lipping/edge of solid walnut

 

[el_Pedr0]

Two other things that might have a bearing here.
1) the front member of this shelf has two notches (19mm x 37mm) cut out to allow for the vertical 'fascias' on the unit. Does this create a weak point if it is made out of 19mm thick solid walnut? - i.e. might that front strip snap off?

2) I am also planning on making a couple of glass-fronted drawers below this shelf. It seems like a good idea from an aesthetics perspective to use similar type of construction on the drawer fronts as for this shelf. Here's the drawer fronts:

This is inspired by this unit:

 

[el_Pedr0]

Some follow up questions for @Droogs or @CaptainBudget, if i may.
How deep should I cut the channel in the 19mm MDF carcass sides for the housing joint?

And with this housing joint, would you make the shelf exactly to the width between the carcass sides plus the depth of the channels? Or is it wise to make the shelf slightly narrower for tolerance/expansion, and if so how many mm smaller on each side?

 

[Droogs]

middle 1/3 for the tenon and mortice, called the rule of thirds funny enough. Doesn't have to be exact so say middle 6mm or 1/4". If you run grooves for sliding the shelf into the case then obvs. as thick as the shelf and about half thickness of the vertical panel up to 10mm, thats all you need. The shelf you make a couple of mm narrower than the gap + the 2 grooves depth (to ease fitment). The notch at the front wont make a difference to the strength of the joint or the carcass once put together. The shelf is also held in place by a back panel or a couple of squint nails at the back.

hth

 

[CaptainBudget]

Nothing I can really contribute that Droogs hasn't said already at this stage beyond the below:

I don't think the housing joint needs to be deeper than 1/4" (~6mm), even this shallow it will take a LOT of force to make the sides fail and drop the shelf (I'd expect the shelf itself to fail first) and it stops you weakening the sides too much, especially if there is a housing joint on the opposite face for another shelf. Obviously the width should match the thickness of the shelf as tightly as practical, which then makes the shelf structural and increases rigidity in the carcase.

I would make the shelf as close to the overall width as practical any gap between the edge of the shelf and the bottom of the housing joint is purely to ease fitment. This also allows you to maximise the strength of the joint; worst case you may just have to plane a couple of shavings off an edge to get it to fit. This also allows you to use the shelf during glue-up as a spacer to help force the assembly square.

 

[Droogs]

yeah I should have mentioned about having a shelf on the other side of the vertical panel oops my bad, strangely I did think about 6mm being the norm when I wrote the above but didn't type it for some reason, ah well

 

[Just4Fun]

I am a bit late to this party but I'll comment anyway. I was pondering on the problem because I don't know what to call the joint I had in mind. If indeed it has a name. Anyway I just cut a sample, on the basis that a picture is worth 1000 words.

Here is one piece, which has a mortice in the middle and a sort of partial lap on the underside.

Here is the other piece which has a tenon in the middle and ...

a housing for the lap on the underside.

When assembled it looks like a normal mitre from the top

and a sort of stopped lap joint from underneath.

This is just a test/demo piece so it is a bit gappy but with a bit more care it could look OK I think - if nobody gets down on their hands an knees to look at the under side. If I were doing it though I would probably just go for a mitre with a hidden M&T like in the sample but without the lap element. The lap makes it quite a bit more fiddly to cut and looks bad I think, although it does add more glue area.

 

[el_Pedr0]

Thank you - not too late to the party. I'm still making a seemingly infinite number of tweaks in the sketch up model so haven't even ordered any timber yet.

I was wondering why no one had suggested an M&T mitred joint (though the paul sellers clip above is close). I was starting to assume that such a joint must have some limitations to rule it out.

I must admit that I like the look of the mitre - to my ignorant eye theres something about it that suggests finer craftsmanship. However, there are other sections in my build where two members meet at right angles and putting a mitre on them could be a right pain. So I'm starting to think that perhaps I ought to go for 'normal' (i.e. square not mitred) M&T joints thoughout.

 

[Droogs]

If you have an Android phone get J-Joinery and install it from the play store. It's free and you will find it informative and interesting

https://play.google.com/store/apps/details?id=jp.dreamo.japanese.joinery&hl=en_GB&gl=US

 

[Just4Fun]

I was wondering why no one had suggested an M&T mitred joint (though the paul sellers clip above is close). I was starting to assume that such a joint must have some limitations to rule it out.

I have used joints like that in the past and have had no problems with them.

I must admit that I like the look of the mitre - to my ignorant eye theres something about it that suggests finer craftsmanship.

I certainly find it surprisingly difficult to get tight mitre joints that are perfectly square. Either one of those two is easy enough but getting both on the same joint can be a frustrating exercise, for me at least. Nevertheless I continue to use mitres and it is very satisfying to eventually get them right.

BTW the sample I cut was tricky because you can't mark the lap housing from the lap. You could rely on accurate measurement or do what I did and make a paper pattern from the lap to mark the lap housing.

 

[CaptainBudget]

I was wondering why no one had suggested an M&T mitred joint (though the paul sellers clip above is close). I was starting to assume that such a joint must have some limitations to rule it out.

It is not as strong as a normal M&T because the glue surfaces are smaller, but I would imagine is strong enough.

The biggest problem is you have to be 100% accurate in your marking and cutting or they end up a gappy mess and out of square. It's also impossible to plane down a perfectly cut frame that's fractionally too wide without ruining it. Definitely a "show" joint that may not be worth the faff in this application as the nature of the case means your eye is not drawn to it. The cutouts in the front of your shelves you want for the case will also significantly obscure these from view in the finished piece.

That's not a call for us to make (you've got to decide if the detail justifies the extra labour/faff), but personally I'd just stick with a normal M&T in this one.

Either way if you're going down this route I would make a practice piece in scrap first.

 

[el_Pedr0]

Really pragmatic advice. I had been feeling like I was getting lost in the detail so had been anyway moving towards normal M&T. These last couple of comments have sealed it for me.

This piece of furniture will be built into a house that has been fully renovated. So hopefully it'll be a permanent showpiece of a much larger project. So it's incredibly tempting to refuse to compromise on any detail. But all the fretting would be academic if I made it so intricate that I never actually get round to completing it.

Thanks all. In due course I'll post my design and hope that you'll all critique it so I can make improvements.

 

[Just4Fun]

It is not as strong as a normal M&T because the glue surfaces are smaller, but I would imagine is strong enough.

Yes, I agree with that. I have not tested the strength of any of thee joints but if I had to rank them my guess would be (from weakest to strongest:
Simple mitre
Mitre with splines
Mitre with hidden M&T
Conventional 90-degree M&T

Either way if you're going down this route I would make a practice piece in scrap first.

Good advice when tackling any new joint. I always do this.