Tuesday, May 13, 2008

Mind the Arc

I built this cabinet in 2005, and last year I noticed that the gap in between the two doors was no longer as even as they had been (they were never perfect, but close). The top of the gap was a little thinner than the bottom. Or so it seemed. I ignored it at first, chalking it up to sloppy construction. But it got worse. After a few days, the gap at the top was less than a 1/8" and the gap at the bottom was 3/8".

(insert little girl scream)

Here's what happened. When I built the carcase, I didn't pay attention to the arc of the endgrain. The second photo, taken from beneath the cabinet, shows the endgrain (on the tails). You can see that the crown of the arc is arching away from the carcase . What happens with a flatsawn board like this is the grain essentially wants to straighten itself out. So, the last tail in between the carcase and the door started to pull away from the last pin, effectively taking the bottom of the door with it. What I thought was an indestructible joint was failing...badly.

So I added glue, clamped it up, and drove a screw through the pin.

A better way to build a carcase or box is to orient the crown of the arcs toward the box. So, when the grain tries to straighten itself out, the pressure will be into rather than away from the joints.

I haven't had the nerve to check the other side of the carcase, but if it does happen to the other board, I'll know not to hyperventilate.


Nick Brygidyr said...

please remember i say this in the nicest way possible...but it looks like you also made that pin a lil too big...*runs away*

Kari Hultman said...

Good thing you're fast! heh heh

Unknown said...
This comment has been removed by the author.
Unknown said...

I really have no idea what you are saying here. I need arrows.

Anonymous said...

Hi Kari,

The proper orientation for boards in panels is first for appearance. The alternate crowns and what have you is not really going to keep wood from moving. There has been plenty of talk about it, and decidedly, just make them look good is best.

Flat sawn boards are what they are, and the wider they get the more we have to accept movement. That said, here are my observations. I don't think the crown orientation did this on it's own.

Your dovetails are properly oriented to support the bottom of the cabinet, though perhaps there could have been more of them and smaller ones, similar to that which you did in the doors, to help control movements. More Dovetails and smaller ones may have better locked the board.

Further, the movement you had to correct for I would bet is more from the tool storage weight in the door being transferred to the piano hinge and onto the carcase. That is why that lower corner was trying to slide out of it's dovetail. That is a heavy load, localized on the hinge area, actually

Had it been a common cupboard door, this would have never happened. There is not enough weight in the door stressing the cabinet side.

I bet a miller dowel would do as well as the steel fastener as well. It is an upgrade that would probably be a good idea on both sides of the piano hinge top and bottom due to the door loadings transferred to the carcase.

This was a great article and a heads up for people concerned with tool cabinet door loadings. Thanks!


Kari Hultman said...

Wyld, I'll try to email you an illustration.

Rob, thanks for the comments. I agree that boards should be pretty-side up. You can see that I ignored that, too, given all the sapwood--ha!

More dovetails might have helped. That's a good thought. Regarding the load from tool storage, however--there are only 8 chisels hanging inside both doors and they don't weigh much. If it were a case of load, then both doors would have experienced the same problem. In fact, both sides of the cabinet have the same number and size dovetails, so the other side of the cabinet should have pulled away as well (it hasn't). Hmmmm

I just went out to the shop and checked the end grain on the other side of the cabinet--it's riftsawn. I'm going to stick to my guns with my first idea—-that the end grain arc orientation played a key role, especially since that board is rather wide. But I do like your idea of more dovetails. I'll probably just throw this tool cabinet out and start over. ; )

Shazza said...


Anonymous said...

Heck Kari,

I bet I know people who would love having that cab if it is scrap to you! Ha!

I'd just run at least three more screws into the cabinet dovetails as you have to shore up the loadings, and maybe do a corresponding treatment on on the doors. The long grain in a miller dowel would work well and fit tight too.

Flatsawn boards... the little buggers. You just can't trust 'em! :-D


Woodfired! said...

At the end of the day it's still a fine looking cabinet and life's journey is about learning stuff and arguing about what we've learnt.

Great post Kari.

PS If you'd made the gap between the doors an inch wide, no-one would ever have noticed!

Kari Hultman said...

That's one of the best things about woodworking--you can always learn something new.

Great idea about the inch wide gap--thanks! ; )

Geemoney said...


Please don't laugh if this is too silly a question, but in these kinds of cases, especially with some of what Rob has said, would it be better to go with mortice and tenon joinery? I know that dovetails are the most common choice in this situation. However, would a different joint have helped in this application? I know, hindsight is 20/20, but I just wonder whether or not M/T joinery would be an advantage here. I could see it withstanding the wood movement a little bit more (which might ultimately be a bad thing, granted), but also potentially helping with the loads created by the hanging doors.

Admittedly, I have M/T and bridal joints on the brain lately, but I would love to hear your take on it for this application.

Beautiful dovetails, however.

(More background about where my question comes: This question also comes up because of an old(er) article I read recently on Fine Woodworking, where the strength of various joints was tested. As I remember, the M/T came in on top, with dovetails following. In the article, the point was made that the number and angle of the dovetails was a relative non-issue, but there was nothing mentioned about these joints with respect to wood movement. The number I have in my mind is about 2000psi to move a joint to failure at the strong end of the spectrum, whether dovetail or M/T. As the article pointed out, this is far beyond most applications.)

Kari Hultman said...

Geemoney, that's an excellent question! I put a lot of thought into the cabinet's construction and my main goal was to make sure it didn't fall apart under the weight of the tools. I figured that the top two corners of the case would carry the most weight and that dovetails, since they lock together, would carry the weight best of all joints.

That being said, there are ALWAYS other ways to do things. I guess you are thinking about tenons along the top and bottom and mortises on the sides, right? Like how some mission bookcases are built? I don't see why that wouldn't work. In fact, that would look cool. Also dadoing the top and bottom into the sides (also like a bookcase) would work, especially if the back were rabbeted in. The other reason I chose dovetails was for appearance. I just like them.

Once all of my tools are loaded in [someday in the far off future], the dovetails will really get a workout and hopefully I won't be posting a photo with the cabinet and contents crashed on the floor.

I'll post about the door design someday and would love to hear how others would have handled the construction.

Thanks for commenting!

Geemoney said...

Due to various circumstances, my work with what I will call "real" wood is currently limited to pine. Otherwise, I am limited to beech planks which have been manufactured from multiple thin boards. Also, only one thickness, 18mm.

Because of that, I have been spared the necessity of having to deal with various issues, one of the major ones being exactly what you talk about in this post - planning for wood movement based on the growth rings and grain direction. I still get some wicked movement, however, when cutting things to size on manufactured wood, which always surprises me.

As I get better I hope also to be able to use more dovetails. Currently, however, I have been trying to stick with a)easier joints and b)joints that hide my lack of experience/skill.

Two things led me to this site, the work you've done with New Mexican design (my home state, so it's near and dear to me), and links from other woodworking sites. I have to admit to some serious ignorance about Mission bookcases, though what you describe is dead on with what I was envisioning. I have some research to do, apparently.

And if the cabinet fails, please don't show any pictures. It looks too nice, and I would rather remember it as it currently is.

Kari Hultman said...


1) Nothing wrong with working with pine; I love it--the smell, the look, the ease of cutting...

2) We're all in the same boat: trying to improve our skills. I'm an expert on this many ww topics----zero.

3) New Mexico is a gorgeous state! I've been there several times.

4) Congratulations on your baby boy!
; )

Unknown said...

Wow, and thank you for the picture!

LadyBurg said...

hehehee - little girls scream. Made me giggle. I can picture the scream. And love the new pic! Very very nice.

LadyBurg said...

Ok, just read all the comments. What are you all talking about? :)

Kari Hultman said...

Ladyburg, come over anytime you want to hear more than you ever wanted to know about woodworking.

Woodfired! said...

A few thoughts on joint selection:

1. A woodworker's tool cabinet is usually their pride and joy. An opportunity to practise skills and show off those skills to clients and friends. So rather than use joints that we "can get away with", we want something exciting and challenging. That said we still need to choose joints that are appropriate for the job and contribute to the piece's overall integrity. And, I hope, show some respect for the tradition that the joint comes from.

2. Modern glues on the whole perform better than the glues that were used when these joints became traditions. (Ducks quickly to avoid flying gluepots!) This gives us a somewhat freer hand in choosing joints and the proportions of them.

3. Modern glues not withstanding, some joints hold better than others when glues finally fail. That's why dovetails are more commonly used than mortise and tenon for this purpose. It depends on how long you expect your piece to last. I'm sure Kari's cabinet will be being used by someone in 100+ years - even if it's not for tools. That user might be grateful for the choice of dovetail joints.

4. The choice of joint is more important if the base of the cabinet is expected to carry significant weight. I would have chosen dovetails in this case (and indeed I did on my own tool cabinet). Another alternative though would be box or finger joints. These are usually done with small spacing giving lots of fingers and therefore a massive glue area. Still dependent on the glue lasting but immensely strong in the meantime. The mechanical fit of the fingers would also mitigate against the cupped board being able to pull away (which I guess is what Rob was getting at in suggesting more tails). However we all know that wood will have it's way eventually.

5. I'm inclined to agree with Kari that the board's orientation contributed to the movement. Where strength is important (and this partly depends on your assessment of how much the board will cup - which depends on the tightness of the arc as well as how well seasoned it is) and if you can predict the direction of cupping then placing the convex on the outside means that the middle of the board is trying to pull away and this will be resisted by the joint far more than an edge.

Some of these views might seem contradictory but I'm just throwing out some considerations. We have to make our own minds up given the requirements and circumstances.

Sorry this was so long but it's an important subject to some of us and there is much to discuss.


Kari Hultman said...

Mark, thank you for your insight. You point out another exciting thing about woodworking and that is the brainpower that goes into designing and building a piece. Considering how and where a project will be used, wood movement, grain direction, joinery, size, shape, thickness & species of wood, and personal preference all play a role in how the piece finally turns out.

All of the variables are important and you are correct that there is much to discuss. (Maybe a future post on your site???)

Woodfired! said...

Hint, hint! I wish I had your stamina for regular posting. I'm sure it's not just me that appreciates your efforts. I do have some thoughts, techniques and images building up - I just need the time to write them up.
The school where I learnt my woodwork skills had great emphasis on the design process. In critiques you had to justify every design decision you had made and explain how each one contributes to the narrative that the piece is telling. An approach that sounds esoteric but is surprisingly useful when faced with design decisions.
I might just write a post on it soon. Thanks for the prod.

Kari Hultman said...

Mark, it's not necessarily a good thing that I have so much time to write this blog. Just means I don't have much of a social life! haha

I would LOVE to read more about the design process. Please blog about that.

Anonymous said...

Hi Guys Mark from Oz
All very good comments above, you guys have some great discussions. I think you were right with your reading of the joint Kari, the only thing I would do, is to replace the screw with a nice contrasting color dowel in each corner, to lock it in and keep up appearances. Great looking cabinet.

Of course, thats if you don't throw it out:)


Woodfired! said...

Another Mark from Oz!!

Mark said...

Oops didn't mean to step on your toes. Where from? I'm in Newcastle. I'll Go MarkA if you like. Good blog you guys have here.

Woodfired! said...

No defence of territory intended Mark.

I'm in Canberra.

No need to change your handle. I've been going to change mine to Woodfired! anyway - that's the name of my blog. (Attempts to capture my passion for wood with my love of good pizza!)

Woodfired! said...

In fact I have!

(aka Markew)

Kari Hultman said...

Newcastle Mark, I can't explain it, but I sort of like that screw driven into the dovetail joint. I like the incongruity of an ordinary utility woodscrew drilled into a project that I spent so much time on and one in which I was so careful with the joinery. (I know that's weird!)

Canberra Mark/Markew/Woodfired....you need to write more blog posts, my friend! ; )

Mark said...

You like the wabi sabi of the screw showing it was made and adjusted by a real person, a bit of whimsy in the order and strength maybe. With everything I make I like to leave something a chisel mark etc showing it wasn't made by a machine or perfect production line.