Some text in the Modal..

Ant Smith
Creativity is an action not an attribute... Photographic Skill: The Book - now available CLICK HERE


Designing a curio cabinet

This project looks at designing a woodwork project and explored:

  • Having and forming ideas

  • 3D modelling of conceptual designs

  • 3D modelling for construction

Where do ideas come from? To be honest, it hardly matters, except to say they're not so much invented as discovered. There's really no such thing as “genius” – some exclusive band of people churning out ideas for society; most genius types I've met have really been little more than over possessive knowledge-thiefs. No, ideas are just there, hanging about waiting to be discovered through the act of discourse; simply by interacting with our world and with others in it. There's quite a nice little article on this over on lifehack.org. We can all of us have ideas, and if we trust and challenge them we can turn them into fact.

My woodworking articles are all about this belief that any of us can do anything any other human can do given the will, tenacity and rightful approach. I just don't believe that we're limited by birth – death is the limiting thing in our lives, not birth! I guess that learning to 'do woodwork', at my age, has somewhat moved me... BUT OF COURSE, it's way more complex than this given society's oppressive forces; what I mean is there's far too many external barriers already in place to allow our own internal barriers any place in our lives. Perhaps I'll write more on this, but not here.

So for this project I wanted to start from absolute zero and share the whole journey from idea to finished object. Therefore, this article journals the process I went through to design a curio cabinet. I'll follow-up later, once I've actually built it.

Step 0: 'Having' the Idea

I keep lists, oh so very many lists, of things I could do; in case I ever find I'm not doing anything. Trouble is, a lot of things on the lists are on a list because at the time they came to me I wanted to do other things. Which means they're less than perfect ideas because they didn't immediately impel me to get out of my bath and crack straight on with them. That's probably not totally fair, but my point is the very best ideas drip serendipity.

When critiquing photography my first concern is always 'why just then, at that very moment in that very place'. If there's no clear answer to that then I find the image, actually boring.

The best ideas are of their time, of their moment.

So when Chris (my wife) mentioned she'd like some way of displaying the accoutrements of our life together, I 'had the idea' of making a display cabinet.

Chris gifts me a lot of ideas about things I could make out of wood; it's one of the reasons I love her so very much.

Step 1: Formation

My next step is to try and understand this idea.

In the realm of woodworking, at an early stage, that means understanding more precisely what this thing will be for and what it will be like.

a few of our curios

To get a better handle on 'what the cabinet is for' I took a look at some of the curios we have. You can see a few of them here. That little plastic Dalmatian at the front was stolen from a baby by our beloved cat Fanny, most of the rest of our curios were legally acquired and all of them play a part in the stories of our life together.

Which is interesting. I started thinking of the cabinet as something with shelves or compartments that we can put stuff on or in. But in contemplating the actual items, thinking about the real-world 'use-cases' of the thing I'm making I suddenly realise that actually I'm making a set of stories, not just a bunch of shelves. It's less a matter of creating a space for these items to sit and more a matter of making a bunch of mini-sets for each grouping of the curios. Now, that I'm sure will impact the design – see how it is the interaction with things that creates, draws out, the idea?

Now I know I'm making a cabinet which will happen to hold things, but more precisely will provide a set of story-telling spaces.

I'll come back to these objects and use them to further refine the design later – but now I have a slightly clearer idea of where I'm heading I want to improve my vision of what the thing will be like.

To do that I'm going to look at what other people think display cabinets look like.

So I do an internet image search for 'display cabinets'. Which turns out to be a really unhelpful thing to do. It turns out most people think display cabinets are big things, almost like cupboards. So I search for 'small display cabinets'. Which is almost as disappointing as now all I'm seeing is smaller cupboards; nothing that really chimes with the vague notions I have in my head.

This is a semiotic problem; a disconnect between my internal understanding of the term and society's understanding of it – I'm sure Noamh Chomsky or Derrida, would have something to say at this point. But I don't have time for philosophy right now. So instead I think about what's wrong with the results I'm getting and I realise that actually I don't want a big floor standing cabinet, I want something that goes on the wall.

A search for which shows me a lot of slimline wall mounted... cupboards. Still not really the kind of thing I'm looking to make.

Talking it over with Christine and she says 'Oh, you mean like a Gay Box?' - which isn't a term I'd come across before and curiously isn't the best thing to look for in a search engine when you want just a few furniture design ideas. My, what a lot of box sets on offer...

It did though help me to start thinking of the different terms we use for cabinets and I realised what I'm looking for isn't a display cabinet or a gay box – but simply a curio cabinet! Hey, it may be obvious now but it wasn't when I started out knowing nothing about this project. My image search (curio wall cabinets wood) turned up a rich set of results that I could mull over.

It was a similar story when I made The Bench - I had a vague idea that I wanted to make a bench so I went through a number of iterations of refining my search so I could look at benches that were progressively more and more like the vague notion I had in my head.

“Chinese curio cabinets“

example of a rosewood chinese curio cabinet

I was most struck by the internal layouts of these, like this one from some site calling themselves houzz.com at a very fancy $464.31 (sold out):

It's way too fanciful of course, but I like the layout concept.

It turns out there are almost endless different kinds of curio cabinet. Great, lots of inspiration. While looking through them I started to take note as to why some weren't at all the kind of thing I would make:

  • No door, I need a door on the cabinet so I don't have to be dusting our curios from time to time.

  • Open sides, same problem as above.

  • Far too intricate for my skills, and anyway I want the curios to be the thing to look at not the scroll work on the cabinet!

  • Layout too limited, I need compartments not shelves because I want to present little stories not just stick stuff on shelves

  • Layout too regular, My curios are all different sizes; aren't everybody's? Who lives a life where all of the curios they want to display are exactly the size of, I dunno, say matchbox cars?

All of which really helps to clarify and focus the kind of thing I want to make.

Step 3: Shaping the design

At this point I stop looking at other people's work. It's a good idea not to completely re-invent the wheel, and looking at how others have answered a problem is super important. But I don't want to be like them, other people, they're disappointing; I want to be my own person, making my own cabinets – not making a slightly shonky version of something you could buy from a shop.

So I stopped looking at other things now my head is awash with ideas and turned to making some rough sketches – not even wireframes at this point, just something to give me an idea of layout and scale.

There are 14 curios currently on my desk that will find a home in the cabinet. Of course we have more packed away in boxes that will end up there as well. And it really isn't beyond possibility that we will yet acquire more; so I can't let the current crop of 14 overly constrain the design – we need some headroom just in case life remains interesting:

Curio Display Requirement = Current Known Crop + Current Unknown Crop + Future acquisitions.

Let's assume those currently on my desk represent half of what we own and want to display. These have been acquired over 20+ years, and I would optimistically expect another good 10 years of living an interesting life that attracts curios before we find ourselves constrained by things like care homes and graves; so perhaps we're going to need to put 35 things in this cabinet. Which seems a lot. Perhaps I should make a cabinet for 20 things and then make another when we need it? Something to ponder.

The 14 curios I have grouped together into 4 clear stories, with a couple of items left over that might make more sense once we unearth the Current Unknown Crop:

  • Wedding items

  • Grotesques

  • Good luck charms

  • Holiday mementoes

All of which means, allowing for the curios I don't yet know about, the cabinet should provide for:

  • 5 'staging areas' where groups of about 4 curios can be arranged to tell a story

  • The remainder comprising mini compartments for stand-out one-offs, in a range of sizes or shapes

And looking at how the curios I have arrange into stories I can see that I'll need a couple of tall-thin staging areas, and perhaps 3 short-wide staging areas. Something like:

How the cabinet might be laid-out

You can see I've arranged this layout in a 5x4 grid, which will now help me determine scale; I am suspecting that the scale consideration will cause me to re-think the layout but then design is an inherently iterative process.

I've got the set of Current Known Curios sprawled out across my A1 self-healing cutting mat, looking for all the world like a Toys in Peril encounter. The healing mat has a 2cm grid on it. I arrange the curios in to their groups against the grid, adjusting the above layout for width based on what I am seeing. I calculate that each horizontal unit in the initial layout would be 8cm.

Then I lay the curios on their faces and check the heights of the groupings; further adjusting the initial layout.

I end-up with a layout that is 5 units across by 7 units down, safe in the knowledge that if I make 1 unit equal 8cm every curio will find a home:

scaled layout

In this layout I've added an upward rhythm with a 1:2:4 ratio of the uprights on the far left-hand side. In fact I can read the vertical rhythms of the piece left-to-right at each unit (8cm) interval:

(click or tap any table row to enlarge)

8cm Rhythm

What do these rhythms of transitions tell me about the geometry? You know, I'm just not sure. Perhaps one day everything will seem clearer. In the meantime I'll just keep noting such things, until it all makes sense.

Anyway, I feel like I'm quite liking this layout, but is it buildable?

Currently the lines in the layout simply demarcate the notional areas, or stages, inside the cabinet; they don't necessarily represent where the shelves and walls are. As I'm contemplating the layout it occurs that I can construct the interior not from a lattice of interlocked walls, but from 6 small open ended boxes either sitting on or under one of two full width shelves:

simplified layout

I could take a more mixed approach and use 2 walls at the bottom instead of the three boxes, which would simplify the build and reduce the material costs. But I like the box idea because it makes the internal space much more flexible, for if we should need or want to radically adjust the content. But perhaps 3 boxes is enough flexibility? So the whole design becomes 4 fixed cavities at the bottom with 4 fluid staging areas at the top provided from 3 movable boxes...

I think it's time to see what it might look like in 3D.

Step 4: Aesthetic Modelling

When I started out in woodworking (back in 2019) I made 1:10 scale models in cardboard before progressing so that I had a good appreciation of what the thing would look like; to ensure I'd be happy with the proportions. They were fun, but required a lot of pencil sketches first to work out all the pieces needed. So I've switched to using 3D modelling software on Orac-1.

I'm currently using Autodesk's 123D Design which is super simple and gives really smooth fly-arounds of even highly complex objects (well it does on Orac-1). Perhaps its fast because it's so very simple and doesn't have to do anything other than crunch vectors; but whatever Autodesk don't supply it anymore. You can probably still find downloads of it (it was last licensed as freeware) but you'll have to do your own searching. Or use something else, 3D modelling is enough fun for it to be worth learning some software or other!

Plan to throw one away; you will, anyhow.

The Mythical Man-Month, Fred Brooks

3D modelling is gaining popularity as PCs become powerful enough to support it and as 3D printers become cheaper. Which means there's loads of people out there with piles of 3D printed scrap who'll tell you how tough it is to line things up precisely in 3D. And yup, it really is. But since we're only making design drawings from it we can avoid most of the heartache that comes with poorly aligned components in a 3D model. Nevertheless a degree of accuracy is needed in the final model and you will find yourself throwing models away in order to recreate them because its way easier to line up primitive blocks than it is to line-up shaped pieces.

All of which means, I make my first model quick-&-doirty. I don't create the joinery of the pieces and I don't care if things overlap. I just get the shape made. There's no use slaving over it, it's doomed to be born a wrackling and to die very, very soon. So I really just bash it out.

Look, here's one for the curio cabinet, as we've discussed it so far:

The initial 3D model

Now, what's interesting is, even though this is the simplest model arising from our considerations so far, it still took some thinking about. My earlier (2D) layout simply didn't take into account the width of the construction material. Since I'd decided that 8cm was the minimum viable width for the compartments I've had to make the boxes with larger external dimensions to ensure we attain that width.

The upshot is that my first box, on the left, is very much slimmer than the layout design. In fact the internal dimension is half that which I had initially envisioned.

I can live with that. But also, note how we end up with double walls and double floors where the boxes abut with each other or the carcass. This has the effect of throwing the upper box-internal staging areas out of alignment with the lower fixed compartment staging areas.

Nothing is balancing quite how I want it to.

Thinking about this model I can see that the only way to be able to control the alignment of the upper and lower staging-areas is to use the same 'inner box' method for both. But I just don't want to add that extra complexity into the design.

The alignment problem would be reduced (not eliminated) if the interior boxes were made from thinner wood stock. I actually think the design looks too 'chunky' when you see a double wall or floor anyway. Also if the fixed cabinet was a totally different wood (or treatment) to the inner box I feel something may be added to the aesthetic of the piece. Lets take a look...

The second draft 3D model

In this model I've slimmed down the inner-box walls, moved the first box to the far left, and shown different materials for the frame and the inner boxes.

It's starting to feel good, reminiscent at least of what I think might have been in my head. The inner boxes have introduced the kind of 'haphazardness' in the levels that I've been longing for!

Despite what it means in terms of extra build complexity I want to look at using boxes in the lower compartment also, to see the impact. I said earlier that boxes also need more material (than shelves with walls), which means more cost – but since I've slimmed the box walls down it might turn out to be cheaper to use boxes anyway, especially if I use a premium wood for the cabinet, and a pine for the internals.

The first 2 drafts of the 3D model took no more than 20 minutes a piece. This next draft took me well over an hour. Now that I have a model, I can tweak it and try alternatives just as much as I want; much more so that making cardboard models anyway: So here's draft 3:

The third draft 3D model

I've adjusted the proportions so that the inner boxes can be arranged in either orientation, tessellating well with others. The two inner regions of the frame are 24x48cm and 36x48cm. The two box sizes are 12x16cm and 12x24cm; in fact it's difficult to create an ugly arrangement of the boxes! They are intended to be moved about (so the display can be kept fresh, if we so desire) but the arrangement I show here is just about my favourite, I particularly like the upward sweep of the lower 3 boxes.

Note I've ended up with 4 larger and 2 smaller boxes. I initially created 3 of each but because the upper part of the cabinet is larger than the lower part it actually made more sense to work with proportionally more larger boxes.

By the way, the vertical rhythms now are:

8cm Rhythm
1.5 :0.5:1:2

It still doesn't tell me anything, but I like the sound of it when I sing that rhythm in my head (we have neighbours you know).

I haven't as yet considered the back, the door, the crown nor the base – but these don't require the same degree of weighing up as the cabinet proportions and internals; in this project anyway. So I will look at those after I've refined this carcass in terms of its construction requirements; which may well end up modifying this basic aesthetic design (I don't really think it will in this case, but often does in more complex designs).

Construction Modelling

This design phase is about marking up the pieces with respect to their required joinery. This ought to be relatively simple in this case but sometimes while modelling I find the number of joints a piece needs to fit the model compromises the structural integrity, so much so that on occasion you find they've been chopped completely in half!

Since the construction model is all about directing the build it's best to work to true scale. Before proceeding we need to select the materials, to determine the actual thickness of the stock.

Obviously cost is important so I need an estimate of the stock required.

The carcass is 52x66cm externally with an 11cm depth and with one full width shelf. This makes the carcass stock length (66cm X 2) + (52cm X 3) = 2.88m.

The large boxes (4 of) are 24 X 12cm with a 10cm depth, requiring 4 X ( (24cm X 2) + (12cm X 2)) = 2.88m

The small boxes (2 of) are 16 X 12cm also with a 10cm depth, requiring 2 X ( (16cm X 2) + (12cm X 2)) = 1.12m

So that about 3m of carcass wood and 4m of inner box wood.

With those requirements as a guide I've chosen 20mm thick Black Walnut with a depth of 119mm for the carcass; and 12mm thick American White Oak with a depth of 94mm for the boxes. And I'm anticipating using the same white oak at a depth of 20mm to create the door frame (and I don't think I'm going to want anything special for the crown and base). This is just over £100 in stock costs for the cabinet. Plus bloody delivery charges of course which are expensive. As is the VAT.

The trouble with buying wood is you need to check the lengths of the pieces you want to cut versus the lengths of the stock that will be delivered. You can order just the right length and find you don't have enough because each piece of stock doesn't provide a whole number of the pieces needed – you end up with a waste off-cut.

Luckily in my last project I created a Cut Planner. I'm going to use this before creating my order so I can be sure to buy just enough, because walnut and oak are damned expensive.

The overall cost is knocking £200 (including glassware). So I make a new model based on these sizes to see if it looks like I want to spend that much!!

The for construction 3D model

Here you can see that I've added the frames for the doors – and yes, I am very happy with this design, I think I'm resolved to spending the money! In this model I've also adjusted all of the components to match the above materials; and at this point, instead of just allowing pieces to overlap in the model (which obviously they can't in real life), I add in all of the joinery cuts I will need to make, thinking about how I will make them, so I know that the joinery is achievable. This exploded view shows all of the joinery:

The for construction 3D model - exploded view

The cabinet will have a mirrored back and glass doors. I was initially thinking about cutting grooves for these to slide into – but two issues occurred to me while I was visualising the construction: If the glassware slides into grooves, it will need to be in-place while I complete the glue-up and there's a fair chance I might end up smashing it; If the glassware is fixed inside grooves then should it break at some time in the cabinet's lifetime it could only be replaced by completely taking the cabinet apart. So instead of cutting grooves I've created simple rebates at the back of the cabinet and the doors. This will allow me to add the glassware at the very last, using picture frame turn buttons to retain them; which will allow for easy replacement.

I'll cut the rebates with my circular saw, which means they will run end-to-end on any rebated pieces. This can lead to little notches, such as where the rebated shelf sits in a dado in the side walls:

Effect of end-to-end rebating

In this case it isn't a problem because it's at the back. The rebates cut into the door frames don't cause a problem either because the door frame pieces are all end-lapped at the same depth as the rebate, so the lapping piece's lap covers the rebate up. There is one place where this rebating creates the most complex concern, in the dovetailing of the main carcass:

Efect of end-to-end rebating

I guess this wouldn't be an issue if I didn't cut the rebates end-to-end (if I were to cut blind rebates) but I think the right tool for that would be a router which I don't have. Perhaps a skilled application of the circular saw (using a plunge cut) could avoid the issue, and I'll probably try that but I have no idea if I'll be successful. The other option would be to hand-cut blind rebates – but honestly I want to make the joinery as easy as possible since I'm working with a new, unfamiliar and expensive material. Cutting these notches during the dovetailing will be fiddly, but much easier than cutting a 46cm blind rebate by hand.

I could have had the notches either at the top and bottom or else on the left and right hand sides. I've chosen to place them top and bottom so that they are less noticeable, should the fiddly work be somehow executed less than perfectly... It is the placement of the pins and tails of the joint that determines where this notch appears. My 3D model has allowed me to be aware of it before cutting any wood; and so has allowed me to be proactive regards where it is best placed.

At this point my design is all done and I'm ready to order the wood. Well, intellectually ready, I'm not actually wholly reconciled emotionally to the cost of it yet – but if and when I do make this cabinet I'll follow-up with another post, so that you will come to know if the design is any good in practice!

The Design Process in Summary

So that's it, that's how I design my wood working projects – which in summary is:

  • Have the idea, however it is one has ideas.

  • Clarify the functional needs.

  • Clarify the aesthetic look.

  • Specify the (precise) functional needs.

  • Sketch the aesthetic response to the functional needs.

  • 3D Model Draft 1 – How the 2D sketch will look in 3D.

  • 3D Model Draft 2 and beyond – Balance the 3D compositional aesthetic.

  • Specify the for-construction materials.

  • 3D Model For Construction – recreate the 3D model in perfect scale wrt the construction materials; ensuring the joinery required is physically achievable.

If you want a copy of the 3D model Drop Me A Line and I'll send you the file.

More in Woodworking