Looking for an elegant joint solution

[ade99]

Hi everyone,
I’m newish here. This is my first thread. I have posted in other threads but this is my first question!…
I was wondering if anyone has any ideas for an elegant solution to joining the legs of a shoe rack frame to a top shelf to allow for the wood movement in the top shelf. It’s all going to be in oak. I’ve milled and dimensioned the wood for the frame but haven’t started the joinery yet (or anything of the top shelf). I need to decide on the joint between the legs and the top first. The top will possibly/probably be a live edge from one of the two pieces shown in the photo below. I’ve also included a photo of the milled frame pieces and also a couple of renders from a Fusion 360 model (where the top is only illustrative of the top shelf without a live edge for showing my wife)!
As a solution, I had thought of sliding dovetails but then wondered about leaving an open dovetail socket at the back of the piece. Not the end of the world but the back of the piece will sort of be visible through the window of the outside (unused) door where the unit is going to be placed. I then thought about making shorter dovetails at the top of each leg (approximately a quarter of the longest dimension of the leg - front to back) and make access to sockets in the underside of the tops by using short mortices, just behind each dovetail socket (offsetting the rear leg dovetails further back towards the middle of the leg top, to allow for both forward and backward movement in those rear sockets (with no glue in those joints).
Other points for consideration: I’d like invisible joints (like a mortise and tenon for all of the other joints in the frame). I had also hoped to not have to use stretchers between the tops of the legs for aesthetic reasons (even spacing between all of the shelves) and didn’t want the unit to be any higher than is already planned due to visibility from the outside. My wife wants to put a vase or similar on the shelf so having the shelf a touch lower than the door’s window is preferable. The unit will also be set away from the door, 100mm or so, as there’s a draught excluder (sausage) at the bottom of that door.
Any thoughts and ideas would be much appreciated. Many thanks in advance

 

[Jacob]

Some over-thinking going on?
I'd lay the table top on to the ends of the 4 stiles as in your 3rd pic, plus a little housing for each of the 4 ends to keep the thing in place but a loose fit.
But your equal spacing of shelf tops might look odd as the opening below would then be different from the lower openings. Another rail like the others would look better. On the other hand who is going to notice, either way?
With a top rail all round you could sit the top on it and join it with buttons. One each end just to hold it in situ.
Your drawing shows dowels (?) going lengthways. Flimsy - they'd all bend. Better across the width.
There is no bracing shown, to resist racking. Brackets under the bottom rail might do it?

 

[ade99]

Some over-thinking going on?
I'd lay the table top on to the ends of the 4 stiles as in your 3rd pic, plus a little housing for each of the 4 ends to keep the thing in place but a loose fit.
But your equal spacing of shelf tops might look odd as the opening below would then be different from the lower openings. Another rail like the others would look better. On the other hand who is going to notice, either way?
With a top rail all round you could sit the top on it and join it with buttons. One each end just to hold it in situ.
Your drawing shows dowels (?) going lengthways. Flimsy - they'd all bend. Better across the width.
There is no bracing shown, to resist racking. Brackets under the bottom rail might do it?

Thanks for the reply.
Some over-thinking going on… absolutely definitely! Lol!
I’m not in the trade and am always looking to further my skills… Each project I do is more complex than the last.
I wanted to see if it was possible to achieve an invisible joint (but also allow for any wood movement). The oak slabs in the photo have been air dried for a few years but I’m sure they’ll still move a fare bit. I also wanted to keep the aesthetics of having evenly spaced and sized stretchers carried on up to the top-shelf.
The pieces for the frame haven’t been cut to length yet, only their cross-sectional dimensions have been finalised, so there is plenty of material available for a tenon or a dovetail etc, at the top of the stiles. The stretcher thickness (vertical height in the drawing) is 42mm after milling and the top shelf will probably be around 50-60mm in thickness after milling. I’d prefer not to have a top stretcher purely for a balanced aesthetic (as the combined height of that and the top shelf would be in the region of 100mm)….
I’ve run a test with the ”dowels” which are actually 12mm square cross-sectional pieces (with a 3mm round-over) and will also have a central perpendicular support. The test ran fine so I’m happy with that part. I was worried at first too, hence having run a test. I didn’t include a photo of those pieces because I wasn’t worried about them.
And cross-bracing to prevent racking… Mmm, I hadn’t thought about that though. The whole thing is going to be mortise & tenoned. My thinking was that that sort of joinery should be strong enough to hold it square. I was going to glue it all up in sections too, making sure everything was square at each stage… both ends first and separately the long stretchers glued up with their centrally placed cross supports (for the “dowels”). The tricky bit will then be getting both ends glued up with the central section & “dowels”.
I wanted to get the joinery cut at the tops of the stiles before any gluing up started though. Once the frame is finished my thinking was to then cut the housings in the underside of the (dimensioned) top, using the finished frame against the underside of the top for accuracy of the housings.

 

[niall Y]

@Jacob 's suggestion of a rail / apron all round, with the top fastened to these with buttons or expansion plates is also my initial thought. You might also try jointing in rails at the ends only. A long rail can then be tenoned in to these at their centre, forming a wide H shape. The top can then be screwed to these with metal expansion plates.

Either way, you are going to compromise on the look, or the space available under the top. You could, however, ease the problem of your height restriction by lowering the height under the bottom shelf

You have obviously come to to the realisation, that if you just tenon the legs into the top, then you are going to be at the mercy of the top shrinking or expanding in width, However, even allowing for expansion and contraction with such a relatively chunky top, in relation to the rest of the stricture, you still might run into problems if the top develops a twist.

 

[ade99]

@Jacob 's suggestion of a rail / apron all round, with the top fastened to these with buttons or expansion plates is also my initial thought. You might also try jointing in rails at the ends only. A long rail can then be tenoned in to these at their centre, forming a wide H shape. The top can then be screwed to these with metal expansion plates.

Either way, you are going to compromise on the look, or the space available under the top. You could, however, ease the problem of your height restriction by lowering the height under the bottom shelf

You have obviously come to to the realisation, that if you just tenon the legs into the top, then you are going to be at the mercy of the top shrinking or expanding in width, However, even allowing for expansion and contraction with such a relatively chunky top, in relation to the rest of the stricture, you still might run into problems if the top develops a twist.

Thank you. Good thoughts and suggestions.
I had thought about adding stretchers at the side, just under the top but not the long “H” rail between them. Interesting thought. I did want to keep the bottom space under the bottom shelf larger than the rest though, to allow for boots so don’t really want to reduce that height either!
The top twisting was another thing I hadn’t really considered. Thanks for that thought. The slabs have been drying for a few years. I was going to mill them in stages to allow as many stresses as possible to be released before milling them to a final dimension (whatever that might be).

 

[Inspector]

If the moisture content in the slab is close to other wood you have in the house then attach the legs to the top with mortices, tenons (cut with a tenon cutter in a drill/drill press or doweled. The small amount of movement the top may go through will be "absorbed" by the top of the legs flexing.

You can input the moisture values, dimension and species to find out how much the top will move. If too much then use the skirt at the top and buttons as suggested.

https://woodbin.com/calcs/shrinkulator/

Since you show the rack sitting in a sunny window you should put the slab in front of it for a while to let the heat from the sun dry it out. Flip it over often so it drys a little more evenly. That should get most of the warping, along with any surface checking and cracking out of the way.

Pete

 

[johnnyb]

I wouldn't hesitate to tenon that to the top. If its dry it'll be fine. As suggested those legs will take up any movement.

 

[ade99]

If the moisture content in the slab is close to other wood you have in the house then attach the legs to the top with mortices, tenons (cut with a tenon cutter in a drill/drill press or doweled. The small amount of movement the top may go through will be "absorbed" by the top of the legs flexing.

You can input the moisture values, dimension and species to find out how much the top will move. If too much then use the skirt at the top and buttons as suggested.

https://woodbin.com/calcs/shrinkulator/

Since you show the rack sitting in a sunny window you should put the slab in front of it for a while to let the heat from the sun dry it out. Flip it over often so it drys a little more evenly. That should get most of the warping, along with any surface checking and cracking out of the way.

Pete

Thank you.
Really good thoughts... so I've just checked the moisture values for wood in the house (on finished pieces - a couple of coffee tables with slab tops) and also the wood in the workshop. There only seems to be a few percent difference which is interesting. I was expecting more. (I have a good meter - Wagner Orion 950).
Thanks for the shrinkulator link too... that computed ~2.5mm movement. I also tried the KatzMoses calculator which gave a higher figure of ~5mm. I'll include those in my calculations. Thank you.
And your last point... brilliant. As the frame is the first bit to be made, I'll make that and take a longer term view with the top, letting it sit on top of the frame in situ, and let it dry (and twist) whilst flipping it and then attach it at a later date once its settled down.

 

[ade99]

I wouldn't hesitate to tenon that to the top. If its dry it'll be fine. As suggested those legs will take up any movement.

Thank you. Using the worst case calcs from the shrinkage calculators above, 5mm max averaged between the two legs would be a max flex of 2.5mm perpendicular to the 150mm (leg length down to the next joint) which doesn't seem too bad! Or does it? I'm not sure. That would be the simplest solution. It might put excess stress on those mortise and tenon joints though.

 

[HOJ]

Got to be honest, will that not look odd just parked in front of door? presume you wont need to get out of it in case of an emergency.

All the "shoe racks" I've made have had doors on them and solid shelves helps to stop the carp & dust from dropping on to the ones below.

 

[ade99]

Got to be honest, will that not look odd just parked in front of door? presume you wont need to get out of it in case of an emergency.

All the "shoe racks" I've made have had doors on them and solid shelves helps to stop the carp & dust from dropping on to the ones below.

Point taken.
To be honest, that door has only been opened once in the three years that we’ve been here though. That was when we needed to drain the heating system down and the shortest route to a drain outside is via that door. It’s an old house with an extension and the extension has two newer exits making this door pretty redundant. But in the case of an emergency, as it’s not a built-in, the shoe rack should be pretty easy to move out of the way.
As far as solid shelves and doors: I do hear you on this but this rack is more for non-muddy shoes, trainers and slippers etc taken off before heading upstairs. We tend to leave the muddy stuff by the back door (where I’ll be also be making a more solid seating area with space for boots and other muddy bits underneath).

 

[Sam Kilby]

If you really want to go belt and braces, and use it as an excercise (maybe practice for larger projects like a dining table etc)
The best, joinery based solution in my opinion would be breadboard ends. These could have the legs tenoned into them so they are fixed, then the boarded up top could be on a sliding dovetail. You could either taper this, or leave it standard. You could glue the rear 1/4 of it to allow the rest to move.
This would also act as a cleat to help arrest any drying movement of the top.
If you don’t fancy the d/t you could use mortice and tenons with haunches for the top to breadboard, but will have to think about leaving space in the mortices for movement, and I would be looking at pegging some of them somehow..


It’s a lot of extra work, and may not be necessary on such a small top, but yeah it’s definitely the belt and braces way. Solid timber panelling or cabinet doors all had a floating panel, and that wasn’t very large across its width either !

 

[ade99]

If you really want to go belt and braces, and use it as an excercise (maybe practice for larger projects like a dining table etc)
The best, joinery based solution in my opinion would be breadboard ends. These could have the legs tenoned into them so they are fixed, then the boarded up top could be on a sliding dovetail. You could either taper this, or leave it standard. You could glue the rear 1/4 of it to allow the rest to move.
This would also act as a cleat to help arrest any drying movement of the top.
If you don’t fancy the d/t you could use mortice and tenons with haunches for the top to breadboard, but will have to think about leaving space in the mortices for movement, and I would be looking at pegging some of them somehow..


It’s a lot of extra work, and may not be necessary on such a small top, but yeah it’s definitely the belt and braces way. Solid timber panelling or cabinet doors all had a floating panel, and that wasn’t very large across its width either !

Thanks Sam,
I do definitely want to make a large farmhouse-style dining table in the future. I'll need to move my workshop first though. It's way too small for anything that big at the moment... but when I do, I'll def be looking at breadboard ends and following your methodology for that. Shouldn't be too long hopefully.
For this one, though, I'm looking at using a pretty small single slab so not sure the breadboard ends is applicable here.

As an exercise, I'm going to model the short dovetail sockets (in front of mortices for access) to see if that's a viable thing or not

 

[Jacob]

If you really want to go belt and braces, and use it as an excercise (maybe practice for larger projects like a dining table etc)
The best, joinery based solution in my opinion would be breadboard ends. These could have the legs tenoned into them so they are fixed, then the boarded up top could be on a sliding dovetail. You could either taper this, or leave it standard. You could glue the rear 1/4 of it to allow the rest to move.
This would also act as a cleat to help arrest any drying movement of the top.
If you don’t fancy the d/t you could use mortice and tenons with haunches for the top to breadboard, but will have to think about leaving space in the mortices for movement, and I would be looking at pegging some of them somehow..


It’s a lot of extra work, and may not be necessary on such a small top, but yeah it’s definitely the belt and braces way. Solid timber panelling or cabinet doors all had a floating panel, and that wasn’t very large across its width either !

A bit complicated! You should have a look at some actual tables. Breadboard ends aren't that necessary - it's a bit specialised. Table tops generally are fixed with "buttons" and similar devices to allow movement.
No need to reinvent the wheel - keep looking at actual furniture as primary source of practical design details, you don't have to work them up from first principles!

 

[niall Y]

I would be a bit wary of just using tenons in this instance. Not that it isn't a good enough joint, but that the mortices will present very close to the edges of the top.

Windsor chairs share this form of construction but their legs are free to pivot on a single flexible stretcher to accommodate any shrinkage or expansion in the seat.

Bearing in mind the OP's design has multiple " stretchers" / rails, holding the sides rigid, then the legs will have to bend at the top to cope with any movement.

It might be a lot more likely that to cope with any shrinkage of the top, that a split will open up in the end grain sides of the mortices

 

[johnnyb]

If I gave anything so much thought I'd struggle to make anything! That little top won't move much. If I'd made that shoe rack I'd just screw the top down and pellet it. Or better screw it from underneath. Stuff that small and delicate hasn't really got any doff to break or crack anything. Ps put another stretcher at the top and screw through that. Or buttons into the missing top stretcher. If I made it I'd extend the legs at the top(to stop the mortices splitting) and take the corners out of the top.Float the top with little spacers. And screw through the top rails.just oversize the hole half a mm.

 

[johnnyb]

Ps the whole assembly is not unlike clothes drying racks I make(and sell)
Mortice and tenons whilst neat and strong on small frames are really tedious to get really perfect even with machines. Hence my desire to make everything identical as little variations can really slow things down and scupper accuracy.

 

[johnnyb]

I would also consider how thin the legs are OK for single blind mortices but to thin for intersecting corner mortices. So basically stagger the mortices or beef up the leg.

 

[ade99]

I would be a bit wary of just using tenons in this instance. Not that it isn't a good enough joint, but that the mortices will present very close to the edges of the top.

Windsor chairs share this form of construction but their legs are free to pivot on a single flexible stretcher to accommodate any shrinkage or expansion in the seat.

Bearing in mind the OP's design has multiple " stretchers" / rails, holding the sides rigid, then the legs will have to bend at the top to cope with any movement.

It might be a lot more likely that to cope with any shrinkage of the top, that a split will open up in the end grain sides of the mortices

Hi and thanks.
As you say...
"I would be a bit wary of just using tenons in this instance"... that was my worry too!
"then the legs will have to bend at the top to cope with any movement" - that was my fear and hence the rabbit hole!
"a split will open up in the end grain sides of the mortices" - again, another similar thought I feared!
_

I did have another play in Fusion and came up with something I think I'm going to try (unless you all have better reasons for me not to)...
NB:
1. Front dovetails would be glued in place (to keep the front of the unit's overhang constant). The rear dovetails, which are set 5mm further back would be left loose (maybe even waxed as well).
2. The rear sockets are 5mm longer to allow for movement (~5mm from the earlier "shrinkulator" calcs)
3. It's a bit hard to see in the model but the rear section of each housing has straight mortices and the forward sections are dovetail sockets.
4. There's a minimum 3mm contact region all the way around the outside of the housing (underside of the top shelf) to meet the shoulder of the stiles/legs...but making that any larger and my fear is that the thinnest part of the dovetail becomes too thin/weak.

 

[Spectric]

For me the legs are to skinny in comparison to the rails, also with a shoe rack the shelves are often angled downwards at the front. To get ideas take a look at furniture in secondhand places to see how it is assembled, often good simple ideas that you can take forward to your own projects.