Fitting drawers - wider at the back?

[AndyT]

Over in the projects section, I'm documenting what is my first attempt at building a 'proper' small chest of drawers; all solid wood construction and trad techniques as far as I understand them. It's wandered off a bit into what I think is an interesting question about how to fit drawers and rather than further derail my own thread, I thought I'd start a fresh discussion here.

To summarise very briefly; CheshireChappie mentioned an old technique whereby the space into which a drawer slides should fit tightly at the front and have a little more clearance at the back. I decided to adopt the technique. Jacob said he thought it was a myth and was unnecessary. Then n0legs found a video of a contemporary maker shimming the fronts of his drawer boxes to give the same effect.

I've been doing some more thinking, and I've also gone and looked at some evidence from old work, which is something that Jacob and I agree on; if you want to see how people who made their living from hand woodwork went about it, there are lots of useful clues visible in their work.

We don't have a lot of antiques in our house but we do have a combination wardrobe and also a dressing chest, both I think from around 1900, which I judge to be well-made where it matters but with no work wasted where it does not. (So, for example, the inside of the back is sawn not planed where it is hidden behind drawers; the undersides of drawer bottoms are not smoothed; dust panels are left short rather than fussily trimmed to size. Lapped dovetails are sometimes cut slightly beyond the baseline but mostly not.)

I looked at the wardrobe to see how the drawers were sized. This shows the general arrangement, inside doors, under what is now hanging space:

Here's the inside of where the left hand drawer goes:

and here I am measuring the width at the back. I'm holding a piece of square edged white plastic where the side of the drawer would go, up against the central runner, so as to show the full width on the folding rule.

It's 21 3/8".

Here is the same measurement at the front:

It's 21¼". That's an eighth of an inch - about 3mm - less.

Is the drawer wedge shaped?

Here's the back - 21 1/4"

and the front - also 21 1/4"

The drawer sides are only 5/16" thick and show no signs of having to be planed down locally to fit. In practice, they slide very sweetly.

Thinking about it some more, it does make sense - the side to side clearance, as the drawer is pulled forward, stays constant; the constraint is the slightly narrower opening at the front, where the total clearance is too small for me to measure like this. This also means that the drawer fronts visually fill the openings when the drawers are closed. But most of the side of the drawer has extra clearance - up to a maximum of 1/8". You can demonstrate this another way by grasping both handles and twisting the drawer a little - the back of it can waggle from side to side slightly, though the front can only pivot.

If the drawer had been planed wedge shaped so it fitted like a plug into a matching wedge shaped opening, the amount of side to side movement would increase as the drawer was pulled forward, making the drawer get looser and looser - which I suggest would be a defect to be avoided.

I found the same thing on the dressing chest.

Here is a general view

The light was against me here, so I measured the openings with my low-tech gauge - an offcut of mini trunking and its snap on lid. They were tight side to side at the back, but brought to the front, they overlap the opening on the left hand drawer

and on the right

The drawers are square. I wondered if the taper was made on the central runner - but as far as I can see it's not.

I also measured the width across one of the full width drawer openings and locked the tape measure when it was tight across the back of the opening, then brought it to the front; the size at the back is greater by about 1/8".

Again, the drawers are square:

So, I suggest that it's not a myth, but a way of making sure that drawers open and close well without jamming, and without the faff of having to make them anything other than perfectly square and true.

 

[Jacob]

You could be right!
NB I don't think drawer sides would usually be planed down to fit as it's no prob making them fit in the first place. It'd be a solution of last resort if everything had gone wrong.
So basically what you are showing are tapered "guides" (which isn't quite the same as making the whole structure tapered - which I was casting doubt on in the first place).
Makes sense - and it means the front of the drawer will align with the stops on the front rail even if the rest of the structure is out of alignment.
So the question could be - were they tapered at the beginning or just as part of the fitting process? Or even trimmed later when a drawer jammed? The guides are usually fitted last, sometimes loose or just lightly pinned/glued.
Then what about solid sided carcasses without guides, would they be tapered? I guess not i.e. easier to taper the drawers.
Then what about the other fable about drawers being made so that they tighten as they are pulled out? I think that one is certainly dodgy!

It's much more interesting looking at stuff rather than relying on books/mags! And "ordinary" stuff is rewarding as they were under pressure to optimise their efforts - so you can see the short cuts and dodges (e.g. over cut DTs).

PS I just had a quick measure of several suitable offerings here. There seemed to be so many discrepancies that it'd be impossible extract a general rule. In fact some drawers seemed to tighten at the back as they went in.
The only critical thing seems to be that the fronts should align with the stops on the front rails and the drawers should be lose enough to allow this.

 

[Sheffield Tony]

I'm sure I saw this given as a tip in a recent Fine Woodworking. There it was deliberate design, not a fitting adjustment. The idea being that outwardly, the sign of your extra clearance is never noticed whether the drawer is open or closed, unlike if you had tapered the drawer.

I have some sympathy with the argument I think Jacob was making in Andy's other thread - that if you can make a nicely parallel sided drawer, and a front opening that is a good fit, why should you need the extra clearance at the back ?

 

[ColeyS1]

perhaps it being wider at the back is to allow for racking on wider drawers perhaps :?

 

[Jacob]

I'm sure I saw this given as a tip in a recent Fine Woodworking. There it was deliberate design, not a fitting adjustment. The idea being that outwardly, the sign of your extra clearance is never noticed whether the drawer is open or closed, unlike if you had tapered the drawer.

I have some sympathy with the argument I think Jacob was making in Andy's other thread - that if you can make a nicely parallel sided drawer, and a front opening that is a good fit, why should you need the extra clearance at the back ?

I guess in the end you simply choose which "rule" to follow (or non at all) but in the end it doesn't make a lot of difference as long as the thing works as it is supposed to - and if it didn't there'd be some last minute fitting adjustments to confuse the issue.

 

[AndyT]

A quick point of clarification - as far as I can see, the central guide on the two smaller drawers is straight and not tapered, with the extra width being got not by planing down guides but by adjusting the whole carcase. It's hard to be definite on the wardrobe, where the construction is complicated, but the chest is simpler.

That has solid ends, so the two full length drawers are running between smooth solid sides with no guide pieces at the sides at all. Measuring on the outside of the whole freestanding piece, I find that it is 3/16" wider at the back than at the front. As the sides are a nice consistent 7/8" thickness, that same taper is present on the inside of the box.

This sort of deliberate inaccuracy is little enough that it can be accommodated within the 'give' of the joints at the front - indeed, it would serve to close them up nice and tight where they are visible.

 

[Jacob]

Quick belt around the house!
My poshest piece - a very nicely made Georgian Davenport - shows a consistent 2mm wider at the back of the drawers. :shock: It must be deliberate as it is very well made.
Everything else - inconclusive - lots of small variations top to bottom.
So I was wrong - it is a rule which was used and it makes sense.
But it doesn't seem to matter much if you don't do it. And it seems to be the opposite of what our American friend was doing.
I imagine it'd also be handy at the cheaper end as a way of getting drawers to fit and accommodating errors.

 

[Sgian Dubh]

I remember learning about this way back in the 1970s when I was doing my training, and the reasoning then was much the same: a slightly wider carcase opening, or at least the runners, kickers and their guides at the sides element (depending on the carcase structure) at the back ensured you could create a snug fit of the drawer at the front, even if the back end was a hint sloppy. I've seen it done in new work, spotted it from time to time in some really good antiques, and made a very limited number of cabinets that way myself. To be honest I've never really found it a valuable technique. I don't think this method of assembly was ever one that was universally considered as the ultimate in craftsmanship, rather it was a method preferred by some. I strongly suspect through examining (working on really) many cabinets over the decades (antique, old and new) that the greatest majority (almost all) aimed for square openings and square drawer boxes to fit the openings.

An alternative explained to me at about the same time of my training, which essentially does the same thing, is to make the carcase square, but then make the drawer itself slightly narrower at the back than the front, which is easily done by cutting the back a millimetre or three narrower than the front prior to doing the dovetailing.

I've also heard a case made a few times for making the drawer box a hint wider at the back than at the front to prevent it slopping around in the opening, or to make it more difficult to simply yank the drawer out with the added risk then of it falling on the floor and breaking, but I've never bought into that one, and always thought the arguments for the technique a bit contrived, but I could be wrong I suppose, ha, ha. Slainte.

 

[custard]

I was taught,

1. Make the drawer cavity as square as possible.
2. Make up your drawer sides from quarter sawn stock and leave in stick under your bench for several weeks at least. Lay out your dovetail pattern on a piece of MDF for transferring to the drawer sides.
3. Fit the drawer back to the front of the drawer cavity opening, make it a push fit. Ditto the drawer front.
4. Make the drawer sides so that they just slide into the drawer, each on their respective sides, lay out the sides so the outside face planes sweetly from front to back.
5. Dovetail the drawer, with your marking gauge set so that the tails are about 0.5mm proud when assembled. Two reasons for this, first it means you don't need a shaped cramping block when you glue up, and second when you've planed the slightly protruding tails down flush the drawer box should still be a perfect fit in the drawer cavity.
6. Take the drawer apart, work the groove in the drawer front for the drawer bottom, trim the drawer back about 6mm down from the top and high enough from the base to accommodate the drawer slips and the drawer bottom.
7. Glue up the drawer, remove cramps almost immediately after checking for square and leave the drawer overnight.
8. Test fit the drawer.
9. Make and fit the slips, muntin if needed, and bottom (the drawer bottom is generally made from Cedar of Lebanon with the grain running from side to side, if there are two or more drawers positioned side by side the grain in the drawer bottoms should flow unbroken from one drawer to the next).
10. Test fit the drawer again.

I know this procedure works, in that it reliably produces beautiful drawers that operate smoothly. The problem is it can easily take me 8 or 10 hours to make and fit one drawer, quite a bit longer if I hit any snags. Okay, I'm slow, but I'm not that slow. So using this process means you need customers who are willing to pay!

 

[Jacob]

Interesting.
What I was taught was to make everything strictly to the rod, in any order; drawers first if you felt like it. If you've done it properly everything fits. If not, a bit of fiddling about is called for.
You'd pay just as much attention to grain etc as appropriate for the job. All the components would be made up to finished size taken from the rod - no fitting of one against the other but checked against the rod.
DTs I'd rather leave the pins proud and plane them off down to the sides, removing glue in the process and showing a nice clean end grain. No planing of the drawer sides except to clean up and/or fit if they don't fit too well.

This procedure works too.

 

[Cheshirechappie]

I think the 'rod' method makes a lot of sense for a 'standard' design, perhaps the Victorian equivalent of modern Ikea made in quantity, fast, for as low a price as can be managed. What might be termed 'ordinary' furniture (bearing in mind that most 19th century 'ordinary' furniture was far better made than most modern 'ordinary' stuff.) It would be sensible to keep the rod hung on the workshop wall, and work to it when an order came in, perhaps batch-building several items. The downside of the rod method - at least as far as drawers go - is that some clearance must be allowed for on the rod between drawer box and opening, thus producing a slightly sloppy fit; for a mass-market piece like a kitchen table made to a price, that's not necessarily a show-stopper.

For better class, higher-priced work, the extra effort of making the casework very slightly tapered is worth it to get a drawer that fits nicely at the front. Such pieces were usually made one at a time by an individual craftsman, not going ****-for-leather but not wasteing time either, so there was was scope for dodges that saved time on fitting work, but didn't adversely affect final appearance of the finished piece.

Moving on up the price scale, when the finest work is called for, money is much less an object and there's time for careful fitting, the method outlined by Custard can be justified.

Horses for courses. Interesting stuff...

 

[Matt@]

I'm slightly confused - are we saying the drawer aperture is out of square of is the drawer out of square? I can see no reason for the drawer being anything other than square not least because this is the easiest way to make a drawer! ie components cut to same length (backs and fronts) I would guess any drawer apertures being out of square is due to clumsy positioning of runners due to the fact that as long as the aperture does not narrow as it goes to the back then the drawer will not bind. At the end of the day, its the drawer stops that create the final position of the drawer relative to the carcase.

 

[bugbear]

I'm sure Hooper and Wells (God rest their souls) feel vindicated.

How could we ever have doubted the Cabinet Makers' bible?

BugBear

 

[Jacob]

.....The downside of the rod method - at least as far as drawers go - is that some clearance must be allowed for on the rod between drawer box and opening, thus producing a slightly sloppy fit; ........

Not so.
Whatever the performance criteria required for "better class, higher-priced work" can be incorporated into a rod e.g including tapered fitting, tight fit with any clearances or other details you'd choose to use, even with brass knobs on.
It wouldn't suit the arts n crafts brigade as they work to made up plodding romantic mythical rules and procedures of their own. But it'd certainly suit say Chippendale and the Georgians - who churned out high (and low) quality stuff as fast and efficiently as they could.
The "wider back to the chest of drawer" issue is actually something you'd expect from a batch producer, as it accommodates errors neatly - particularly racking of either or both drawers and carcase. A bit like shadow lines, quirks etc. - a convenient dodge. In fact I've changed my mind and I'm all for it!

I can't think of any downside to the rod - it's a brilliantly effective way of getting stuff done which is why the procedure is so widely used, throughout many other trades as well. It's very odd that many woodworkers don't seem to know about it.

 

[bugbear]

... which is why the procedure is so widely used, throughout many other trades as well.

Could you give examples please?

BugBear

 

[Jacob]

... which is why the procedure is so widely used, throughout many other trades as well.

Could you give examples please?

BugBear

You've read it all before BB it's in the archive! Could you save me the bother and dig out a few quotes?

Steel yards, boat yards, sail makers, for starters. Somewhat replaced by computerisation but the full size layout drawing is still key to many fabrication processes.

 

[Jacob]

I think the 'rod' method makes a lot of sense for a 'standard' design,.......

Also makes sense for a one off. Extremely useful in fact. It's the final step of the design process - committed to paper or board. Can of course be altered (rubber + pencil) as work progresses, if say the stock turns out to be undersized or something.
The alternative of fitting each component piece by piece like a 3D jig saw, is extremely difficult and inefficient.

 

[custard]

I think the 'rod' method makes a lot of sense for a 'standard' design,.......

Also makes sense for a one off. Extremely useful in fact. It's the final step of the design process - committed to paper or board. Can of course be altered (rubber + pencil) as work progresses, if say the stock turns out to be undersized or something.
The alternative of fitting each component piece by piece like a 3D jig saw, is extremely difficult and inefficient.

Rod and fitting aren't mutually exclusive. I start with a rod, build the carcass as square as I possibly can, then fit subsequent components like doors and drawers to the carcass. It's unlikely that the carcass will be perfectly true, so when it's there in the workshop in front of me it's the carcass, rather than the rod, that determines subsequent measurements.

 

[woodbrains]

I've also heard a case made a few times for making the drawer box a hint wider at the back than at the front to prevent it slopping around in the opening, or to make it more difficult to simply yank the drawer out with the added risk then of it falling on the floor and breaking, but I've never bought into that one, and always thought the arguments for the technique a bit contrived, but I could be wrong I suppose, ha, ha. Slainte.

Hello,

I was taught to make the drawer pocket slightly wider at the back and the drawer too. The aim was to have a force fit of the drawer into the pocket which gradually 'lets go' about 1/4 the way in, the rest of the travel is smooth and free. When the drawer is used, it can be pulled out about 2/3-3/4 the way out and then stops. This might be contrived or elegant, depending on your point of view! :roll: it might be unusual, but not unheard of.

Mike.

 

[Sgian Dubh]

I was taught to make the drawer pocket slightly wider at the back and the drawer too. The aim was to have a force fit of the drawer into the pocket which gradually 'lets go' about 1/4 the way in, the rest of the travel is smooth and free. When the drawer is used, it can be pulled out about 2/3-3/4 the way out and then stops. This might be contrived or elegant, depending on your point of view! it might be unusual, but not unheard of. Mike.

Mike, I'd forgotten all about that configuration you've outlined. Now that you've reminded me of it I can't recall where, when, and how I came across a similar or matching description. I don't think I've ever attempted to make drawer openings and drawers like that, although it's possible at some point in my career I've inadvertently ended up with the layout because of what are sometimes euphemistically known as 'calculatory anomalies', ha, ha. Slainte.