Magnetic Wood and Magnet Selection


We find hanging utensils in the kitchen to be the best storage method; most of the vertical space in the kitchen is covered with kitchen equipment.  However, the sides of the cabinets above the sink had nothing hanging on them! This problem demanded a solution. Last time I did a project like this I made pegboard, but pegboard is inflexible both in spacing and in the type of hooks available. This time: magnets.

I started with 1/16″ thick sheets of steel, which I covered with oak veneer. Technical support at 3M advised me to use spray adhesive, so I adhered the veneer with Super 77. The trim on the cabinets conceals the side edges, but not the bottom edges, so I epoxied small oak strips along the top and bottom to hide those edges and blend into the cabinetry.

Once I had the veneered steel screwed to the cabinets I realized that I couldn’t find magnetic hooks suitable for hanging kitchen tools.  I could hang a few things with Super Fridge Magnets and some things with the Standoff Magnetic Tool Holder, but most of the things I wanted to hang would require a different solution.

These  magnetic bases have a protruding threaded rod. To transform them into hanging rods I attached standoffs, metal tubes that are threaded on the inside. At McMaster I could buy standoffs in a variety of lengths, though for some reason the stainless steel ones were twice the price of aluminum.  To get a finished look I found two options. McMaster sells elegant end caps for the stand offs, but they are expensive—more than the shorter length standoffs. You can see them below in stainless steel (which is the same price as aluminum). I also bought a box of stainless fillister head machine screws which make reasonable looking caps for the standoffs at just pennies a piece. This scheme enabled me to make hooks as long or as short as needed. I worried that the absence of a hooked end might allow tools to fall off the wall and plunge into the sink, but that hasn’t been a problem for most things.

The one exception is the red funnel which slips right off of a rod. A simple solution is to insert a screw that is longer than the threaded rod so the screw head acts as a stop.


Hanging this funnel required a trick to keep it from sliding off the rod.


We also have some spatulas whose hang holes do not admit a 1/4″ shaft, and the threaded rod does not come in a smaller size that is compatible with the magnetic screw bases. I attached a 1/2″ diameter magnet to a small piece of wood and inserted a 3/16″ diameter stainless rod to create a suitably sized hanger. It didn’t work: I planned to hang four spatulas from this rod, but the hook would slide down and fall off the wall if I loaded it with more than two of the utensils.

Next I tried polymagnets. I recently learned about these magnets where the magnetic field varies spatially across the magnet to produce different effects. They make several types of magnet pairs, including the spring, latch, and centering magnets. For this task I was testing the “attach” type magnets, which supposedly concentrate the magnetic field closer to the magnet so that the magnet grips more strongly. I had on hand some 1″ diameter, 1/32″ thick “attach” magnets as well as two 1″ diameter, 1/8″ thick “attach” magnets (one with a mounting hole and one without). The 1/32″ thick magnets have the same volume of magnetic material as the 1/2″ diameter, 1/8″ thick magnets I used to make my unsuccessful hooks. Another observation is that two of the 1/32″ magnets got stuck to each other at one point and they were extremely difficult to separate. It was impossible to slide them apart like one usually does to separate rare earth magnets. I had to separate them using a wedge, something I’ve never had to do before, even with larger 3/4″ diameter, 1/8″ thick magnets. These magnets are not weak.

I glued a 1″ diameter 1/32″ thick “attach” magnet to a small piece of wood and again inserted a length of steel rod.   This new hook was very weak and slide right down the wall.  I glued on a second “attach” magnet, and now it could hold the weight of 3 spatulas, but not 4.  When attached to the refrigerator, however, the hook would hold the necessary weight—the four spatulas I wanted to hang weighed about a pound in total. Note also that the rolling pin weighed more than a pound and yet was easily held to the veneered steel by one hook with a 1/2″ magnet.  The hook with two polymagnets has double the amount of magnetic material,  but it fails to do the job.

I abandoned that design.  The hanging rod in my original design was below the magnet instead of centered over it.  The wood was 1/4″ thick.  The magnets are 1/8″ thick.  I didn’t think the remaining 1/8″ of wood sufficed to support the hanging rod.  The off-center hanging rod seems to make the shear hold weaker because the load on the rod can tilt the magnet away from the surface.  I made the centered rod design (right) by using a magnet with a screw mounting hole in the center. The rod goes through the screw mounting hole, and I filled the space with epoxy to ensure that the rod would be well supported.  This new hook holds all four spatulas without any difficulty.


Spatula hooks. The bloodwood hook on the left with a conventional half inch magnet has the hang bar off center and can hold only two spatulas. The chechen wood hook on the right, with a one inch polymagnet and centered bar easily holds four spatulas.


Measuring magnets

So what was going on with my magnets? I tried to measure the capacity by filling a bag with weights (dried beans) and determining the point at which the hook began to slip. This was kind of messy and I didn’t feel very confident about my accuracy, but I did find that the hook with the two polymagnets held 4.25 lbs on the refrigerator and only 1.1 pounds on the wood covered steel. One important factor in this sort of holding power is friction. Is it possible that my finish on the wood was too slippery? Or did the magnets simply hold with less strength on the wood.

forcesheer_smTo get to the bottom of this I purchased a Chatillon force gauge cheaply on eBay. This device measures force in pounds and enabled me to measure the pull force to detach magnets from surfaces as well as the shear force (the hanging load that would cause the magnet to begin sliding). Measuring shear force is relatively easy: I attached a loop of string to the hook and pull downward with the gauge.

Measuring pull force is a bit more difficult. I used a strip of Tyvek and placed the magnet over it. This means that the magnetic force is decreased by the extra distance created by the layer of Tyvek. It also is difficult to be sure that the pull is applied directly perpendicular to the plane of the magnet, and I suspect that an angled pull may lift the magnet more easily.


Measurement of the magnets resulted in a complicated set of data with some inconsistencies. I tested magnets on two surfaces, the refrigerator and the wood veneered steel. If the pull force of the magnet is F then the shear force is supposed to be equal to σF where the coefficient of friction, σ, depends just on the surfaces. Lee Valley sells two self-adhesive high friction materials. One is their High Friction Disks, specifically marketed for use with rare earth magnets. The other is the Super High Friction Tape. When I ordered the polymagnets I also included some Traction Tape to use with them. My hook that wouldn’t support a pound of spatulas gave the following test results

Two thin Polymagnets glued to wood
Pull Force (lb) Shear Force (lb) σ
Fridge Wood Fridge Wood Fridge Wood
Bare magnets 5 1 1.5 0.25 0.3 0.25
Friction disks 3 0.5 3.75 1 1.25 2
Friction tape 1.75 0.5 5.5 2 3.1 2
Traction tape 4.25 1.1

When placed on the refrigerator the pair of 1/32″ thick 1″ diameter “attach” polymagnets gave a respectable pull force, five pounds with nothing but a sheet of Tyvek separating them from the refrigerator. The friction disks are only 0.022″ thick and the friction tape is 0.016″ thick. But this small extra separation from the metal created by the friction disk or friction tape reduced the pull force considerably to 1/3–1/2 of the original amount. Despite this decrease in pull force, the shear force increased. Without a friction tape of some kind the coefficient of friction was around 0.25, but with friction tape it rose as high as 3.1. Note that for this hook the high friction tape gave the best result: with two pounds of holding power even on the wood it would have sufficed to hold the 1 pound load of spatulas.

Why is the magnetic force of attraction to the refrigerator so different from the force of attraction to the veneered steel?  And why can’t two magnets hold on as strongly as a single magnet hook? Is the high friction tape really so much better than the other tapes?  I made the same measurements of the screw bases:

Small (16mm) screw base magnet
Pull Force (lb) Shear Force (lb) σ
Fridge Wood Fridge Wood Fridge Wood
Bare magnet 4.5 3.5 2.75 1 0.61 0.28
Friction disk 2.5 2.75 2 2 0.8 0.73
Friction tape 2.5 1.5 2 2.5 0.8 1.7

The strength of attraction to the wood and the fridge are closer together in this case, and the shear forces are very similar. In some cases the attraction to the wood is higher than the attraction to the fridge. This result is quite different from the previous one. One anomaly stands out: the shear force for the bare magnet on the fridge is very high. I noticed that when I dragged this magnet on the fridge it was scraping little white bits of the vinyl off, so I think that rough edges of the magnet were digging in and providing a huge boost to friction. I tried sanding the base but it didn’t significantly change the results, though. (Maybe I didn’t go far enough.)  Another anomaly is the high σ value with friction tape on wood; I have no explanation for this one.  These screw bases feature a 1/2″ diameter magnet that I assumed was 1/8″ thick, which would mean they have half the magnetic material of the two polymagnets I used above, and yet the holding power on the wood is higher. Moving on to the larger screw base:

Large (20mm) screw base magnet
Pull Force (lb) Shear Force (lb) σ
Fridge Wood Fridge Wood Fridge Wood
Bare magnet 4.75 6.25 1.5 1.25 0.3 0.2
Friction disk 3 4 3.25 4 1.1 1
Friction tape 3 4 3.5 4.5 1.2 1.1

This larger base clings to the fridge with greater strength than the smaller base, and on the wood it is even more powerful. The shear force strength is also somewhat stronger on the wood. So a bigger magnet is gripping more tightly and giving more holding power, right?

Wrong. Both screw bases have the same sized magnet.  The larger base has a plastic ring encircling the magnet.


I disassembled one of each hook and found that the magnets were slightly bigger than 1/2″ in diameter (0.54″) and at 0.15″ thick, they were thicker than expected.  And they were indeed the same.  These neodymium magnets come in different grades with different strengths, but I tested the pull force and found no difference: both held to the fridge with 2 lbs and the wood with 2.5 lbs.

So apparently the difference in magnetic forces depends on the way the magnet is enclosed. I measured the pull force for a bare 1/2″ magnet and got 1.5 pounds for both the fridge and the wood. When I put the magnet into a steel cup the force rose to 3 pounds.  The observations above suggest that I might get even more force if I had an oversized cup surrounding my 1/2″ magnet.  So I tried inserting my 1/2″ magnet into a base designed for a 5/8″ magnet.  The force rose slightly to 3.25 lb on either the wood or the fridge.  I didn’t see the striking increase I got with the magnet base on the wood.   I tried a larger 3/4″ magnet cup and the force decreased slightly.  The only curiosity was that when I inverted the magnet cup, so the magnet was backed with steel but not covered on the sides, I saw an increase in the attraction to the wood to 3.5 lb, while the attraction to the fridge declined to 2.5 lb.

What conclusions can we draw from these measurements? Clearly backing the magnet with steel makes a big difference.  It seems like it’s hard to predict the attractive force of different configurations, so measurement may be the only way to know what will happen. The thin polymagnets did not adhere well to the wood. Perhaps their magnetic force is shorter range than the normal magnets, and the veneer is thicker than the vinyl that covers the refrigerator. This behavior doesn’t offer any obvious advantages, so I don’t plan to use those magnets again.

The coefficient of friction for the bare magnets came out consistently around 0.2–0.3, and inserting friction tape of some kind improves this to about 1. The small gap created by the friction tape significantly reduces magnetic pull, but you still get more shear force. Don’t skip the friction tape if you care about shear loading! But I didn’t get consistent friction coefficients as claimed by the physics books. My measurements were all over the place, ranging from 0.73 up to 3.1. Why the big spread? The friction coefficient is supposed to be independent of contact area, but I’m observing higher values when I have more contact area and the lowest values when contact area is the least. The high friction tape seemed to perform better with the polymagnets, but not at any other time, so I can’t identify a winner among the friction tapes.

Tilting can play a role in limiting the load a magnet can hold on a vertical surface. These magnetic hooks fail  not by sliding down the fridge but by tipping forward: the failure isn’t in shear.  I tested my wooden hooks (pictured above) on the veneered steel surface.  The bloodwood hook could support 1.75 lbs of force at the base of its 1.5″ rod but only 1 lb of force at the tip.  The chechen hook has a four times bigger magnet and could support 5 lbs at its base, but again only 1 lb at the tip of its longer 2.125″ rod.  When the load was at the tip it  usually failed by tilting, or it vibrated as it moved.  Perhaps even when magnets are sliding down the fridge they are tipped forward, compressing the friction tape at the bottom and moving the magnet away from the fridge at the top.

I’ll finish with some measurements of a few other polymagnets. I have a 1″ diameter, 1/8″ thick “attach” magnet. It has a marked side that I thought was supposed to be the active side. Pull force on the fridge was 7 lbs on the active side and 13.5 lbs on the reverse side. On the wood I got 3 lbs on the active side and 5.5 lbs on the reverse side. I also have a polymagnet with ring shaped magnetic regions designed so that a pair of them will center on each other. This magnet is 3/4″ in diameter and 1/8″ thick. With the marked side to the fridge it had a 9 lb pull force and with the reverse side only 4 lbs. On the wood, this magnet gave a pathetic 2 lb pull on the active side. A normal magnet of the same dimensions clung to the fridge with 2 lbs of force, and the wood with 2.5 lbs. The 1/2″ magnets I tested above are only 44% as big and yet 75% as powerful in this situation, so for hanging stuff on the fridge, at least, save money and use the smaller magnets—or spring for that centering polymagnet.


Draining Rack


We have various soaps next to the kitchen sink, and I noticed that as a result, the sink area was always wet, which led to mold growth. To address this problem, I tried building a drainage rack out of wood, which I hoped would allow the water to dry and prevent the growth of mold. I’m not sure whether it was helping with the mold, but it turned out to have another problem: the wood itself produced brown stains on the counter.


Underneath the tray it looks worse.

It seemed that for this project, I needed a different material.  I envisioned something with holes in the top and a sloping drainage tray to send water back to the sink.  Shapeways offered a 3D printed ceramic material that is dishwasher safe and food safe.  It seemed ideal.   So I started designing in SketchUp.  SketchUp is easy to use for lots of things, but I found it tricky for this task.  Arranging the holes in the top proved tricky, tedious, and hard to change once it was done.  The printed object will be nicer with rounded corners and edges, but SketchUp is not very good with rounding over corners. But eventually I got a model done and submitted it to Shapeways.

It was rejected because the walls of the design were too thin somewhere.  After resolving that problem my design got rejected because it wasn’t adequately supported.  But the exact nature of the support I needed remained elusive as I tried to redesign the model to comply with their vague requirements.  Then the ceramic material got discontinued entirely.

Months later I was admitted to the Shapeways pilot project for a new porcelain material. I had to reduce the size of my model to make it printable, but the only hitch I had was failing to meet a minimum size requirement.  So finally, I was able to make the print.

My original design was a two piece ceramic structure with a drainage tray in the base and a top that lift off for cleaning.  This was quite expensive, so in a quest to lower the price, I devised a design with a single ceramic part and a slit to hold a drainage tray.  I ordered a piece of white plastic from McMaster and cut it to size to fit into the slot.

The result is functional and reasonable looking.  I was a little disappointed by the way the glaze looked:  I was expecting a more uniform deep green color.  The long unsupported top span sagged a little bit during fabrication, but this doesn’t affect its function.  The rack has stood up to a couple months of use and has definitely improved the sink area.




Charging Station Prototype

We have enough rechargeable devices that we needed a way to charge them that was less chaotic, that didn’t leave devices on the floor.  I looked at commercially available options, but nothing seemed to do what I want.  So I started constructing my own wall mounting approach.

My idea was to have a shallow shelf that devices rest on, with some mechanism to hold and organize the cables and a box to hold the charger hub.   I thought initially that cables could rest in slots that were narrow enough so that the plug wouldn’t fall through:


Cables can slip into the skinny slot to hold them in place. The wider part of the slot permits the cable end to stick down when a device is charging.

This approach didn’t work because the cables sometimes pull forward out of the slots.  The next idea was to use rubber to hold the cables in place.  But which type of rubber?  McMaster has an overwhelming number of options, both types of rubber, thicknesses, and hardnesses.   I tested several options:

dsc09092It seemed like the 1/32″ EPDM with a Shore hardness of 40A was promising.  So the prototype looks like this:

dsc00043I realized that velcro alone did not suffice to manage the cables, so I added the cable wrapping posts.  Also it seemed like the EPDM rubber was not holding cables as well as I expected, especially with thicker cables.  So I tried a change in the design with a thicker piece of rubber and a hole to hold the cable.  This seems to work better at holding cables, though it is more difficult to get them in and out.


The black rubber has a hole punched in it to hold the cable. This does seem to be more secure, though choosing the right rubber is still an issue.

We’ve had this in use for a while now, but one problem I’ve noticed is that the users don’t like to leave the cables secured in the rubber slots.  They prefer to just leave them dangling.  Does this mean the rubber slots are unnecessary?

How can this design be improved?

Couch Cabinet

dsc00014After finishing the Utilitarian Cabinet I said I was going to lay off the plywood for a long while. Events conspired against me: we needed a small table or cabinet to fill the narrow space beside the couch and the remaining walnut plywood was just right for the job.

For this cabinet I borrowed a Festool Domino to make the joints for the case and drawers.  The Domino is a lot nicer than dowels. It can make one pair of tight fitting mortises for alignment and loose fitting holes elsewhere, which makes the joint much easier to assemble and disassemble than dowels.  This cabinet went together without the struggles that we had with the dowel-joined Utilitarian Cabinet—I was able to do it without a helper. The drawers went together very easily as well, and the dominos helped in assembling the miter joints in the pedestal and in affixing the pedestal to the case.

I also experimented with the undermounting Blum Tandem-Plus Blumotion drawer slides, which waste less space at the sides of the drawers.   With the drawer cavity only seven inches wide I wanted to make the widest drawers I could.  Understanding the requirements for these slides proved to be rather difficult. Blum is very bad about posting detailed information, and they also sell different products in Canada than they do in the USA, so Lee Valley has products that are not otherwise available in the USA. It wasn’t immediately obvious to me—though it should have been—that the undermounting slides would have a minimum drawer width. It is not easy to figure out what this minimum width is.  Lee Valley didn’t know, and when I asked Blum USA about the Canadian slides they couldn’t find the answer!  I eventually learned that the Tandem-Plus would work on my narrow drawers as long as I used special “locking devices”.  The undermounting slides waste less horizontal space but they consume over an inch of vertical space.   This led to a problem when I positioned the handle without checking the position carefully and the mounting screw hole hit the drawer bottom.


See the recess I carved into the drawer bottom to allow me to attach the drawer pull?

When I went to fit the drawers into the cabinet they worked right the first time. This is a huge contrast to my experience with the side mounting slides I used in the Utilitarian Cabinet where I had to spend hours on drawer fitting.  I had a brief problem with it when I did the final test fit and the drawer hit the case.  It turned out I had accidentally pressed the levers that raised the drawer.  Lowering the drawer back to the correct position was very easy.  The finished drawers open and close very smoothly and much more easily than the drawers in the Utilitarian Cabinet, though we’re not sure we like the Blumotion drawer closing action.  Because the drawers are tall and narrow I used the wood with the grain running vertically and I mounted the handles vertically.


Note the continuous grain on the mahogany drawer fronts.

The corner in which the cabinet sits is not square.  It was necessary to curve the edge considerably to get a reasonable fit.  dsc00021

Kitchen Organization

This post is a round up of small kitchen organization projects, inspired by a few recently completed ones, but including some projects from years ago.

Organize a Bottom Freezer

This one isn’t very interesting as a woodworking project, but might be interesting to people looking for a solution to the problem of organizing a bottom freezer. Our old refrigerator had a top freezer, and I had containers in the door, labeled on the side, and shelves in the freezer space. The new freezer has a big open cavity and a sliding top drawer. How could I use this space effectively? When I looked around online for ideas on freezer organization, I didn’t see anything that looked both effective and efficient.

For the sliding top drawer, my solution is to use these air-tight containers that can be turned sideways and labeled on their sides.
For labeling I find that the Pentel Wet Erase Chalk Marker works well. It resists water, but can be washed off of smooth surfaces easily when desired (though it mysteriously does not come off in the dishwasher). The only problem is that these particular containers have a rough area on the side, from which the pen does not clean off.

For organizing the bottom some people put things in containers, but this wastes scarce freezer space. I made wooden dividers to partition the big, open space into sections sized to quart freezer bags.


The left side is divided into six spaces. The right side is divided into four larger spaces.


I cut these dividers from quarter inch plywood using my band saw, but a jigsaw would work too. There is no need to finish them.  Observe how the dividers fit into the gap in the freezer frame at the far right and left sides.


These are the dividers for the right side. The top one is cut to match the slope of the back send of the freezer. The lower piece, which runs left to right when assembled, has a small notch to fit onto the plastic divider that came with the freezer.


Measuring Cups & Spoons

We have a lot of measuring cups and spoons. Some of them hang on hooks from holes in the handles, but we had a plastic set with no hang holes that I hung using magnets. The nice thing about the magnets was that it was very easy to hang them or grab them. The troublesome thing was that adhering magnets to the polyethylene measuring cups required an expensive adhesive and they would eventually fall off. I got new stainless steel measuring cups and their different shape and increased weight made the magnets a more troublesome solution. So I came up with this approach that exploits the little tabs on the ends of the cups. (However the one cup measure has no tab and requires a little support bar underneath.)


This measuring cup rack is angled 45 degrees to the wall. It might be slightly easier to use if it was a bit closer to horizontal.

I attempted to drill the large holes using an expansive auger bit. These tools seem to be plentiful on ebay but they  don’t seem to work. (Why did they make so many if the tools didn’t work?)  In each case the center hole cut by the lead screw would grow until the bit could wobble around in the hole with nothing guiding the cut. The deepest I could cut was about 3/16″ before it became impossible to make further progress. I had to cut out the centers with a coping saw and finish the edges with a drawknife. With a drill press and a circle cutter this job would be easy, in theory, but when I equipped my drill press with this unbalanced cutter the lateral forces caused the chuck to fall off.  The chuck continued to fall off periodically after that, until I learned the trick of freezing the arbor.

I marked the sizes using punches and then inking them with pens.  Curiously some pens, like the Pigma Micron, seem to be ruined by writing on wood.  They never started writing again after I used them for this job.

For years we hung the other measuring cups on a store bought rack, which really didn’t work well. The hooks would fall off, and it wasn’t well organized. So to go with the rack above I made a rack with designated hooks:


For the measuring spoons I made this rack by attaching veneer to a square foot sheet of plywood:

Pen and Paper on the Fridge


We need pen and paper available at the fridge. To make a pen holder I started with a small scrap of cherry.  I used the router to cut out a recess in the block of wood and I installed some magnets on the back.  If I was going to make one of these today, I’d probably glue together two layers to avoid using the router.

To hold paper we used a plastic rack for many years, but its weak magnets resulted in frequent trips to the floor, and it eventually cracked beyond repair.  I made a much nicer wooden replacement from quarter sawn cherry I had left over from the file cabinet.

Now we just need to figure out how to keep people from walking off with the pens.


Room for four pens.


Five quarter inch rare earth magnets keep this firmly in place on the refrigerator.


Look at the fantastic quarter sawn fleck figure.


This needed to be thicker than the pen holder. I made it from several pieces rather than hollowing it out by router. It is also much newer than the pen holder. Is the lighter color due to fewer years of darkening? Or is it a lighter colored tree?


Next time I think I would use a smaller number of larger magnets. But this holds securely even on the slightly curved surface of our new refrigerator.

Hanging Utensils on the Wall

We bought a rack for hanging utensils on the wall, but then we ran out of room. So I made a pegboard out of oak. I built it using frame and panel construction with oak plywood for the panel. I used a piece of pegboard to guide the drilling of the holes in the oak plywood.


The Talon pegboard hooks hold utensils well, though I did have to clip off some of them with wire cutters to get a good fit for some items.


From the front the oak frame blends in with the cabinet. The non-woodworkers don’t notice the change in grain.


Cooling Rack

For mounting cooling racks I made magnetic bars by setting 1/8″ thick half-inch rare earth magnets into 1/4″ thick wooden bars. The magnets hold the bar onto the fridge and also hold the steel cooling racks. They worked well—until we switched to a stainless steel rack that was only barely magnetic. Each of the three bars could hold its own steel cooling rack, but all three bars together barely keep the stainless one on the fridge.  A better design would probably make use of hooks instead of magnets to hold the stainless steel racks.


Three bars with magnets embedded in them.


Barely holding up a stainless cooling rack.

Draining rack by the sink—fail

We have the dish soap and hand soap by the sink and that area stays constantly wet, leading to mold. I tried building a draining rack to improve this situation. I used the router to cut grooves in a piece of ipe.  (I think I killed the router bit.) Underneath the groves are perpendicular so water drains all the way through.  The first problem was that the wood stayed wet underneath.  I discovered that tiny insects had taken up residence.   To correct this problem I raised the wood up off the counter on rubber feet.  This got rid of the bugs, but I now have brown staining coming from the wood.  It seems that wood is not the right material for this application.


Draining soap rack made from ipe.



Rack in use.


Utilitarian Cabinet


This project is a utility cabinet to hold paper and related tools, with a space for the rotary trimmer on top. It seemed like a good opportunity to use up 1.5 sheets of walnut plywood that I bought years ago for a project that got canceled. I also had a walnut board I had bought to use for edging, and some iron-on walnut veneer tape. Of course, the first thing I had to do was hit the lumberyard for three more sheets of plywood and another chunk of walnut. But the plywood I needed to buy was Baltic birch for interior structure.dsc07887_sm

In addition to storing standard sheets of letter size paper, I wanted storage for stacks of over-sized paper, large sheets of Japanese paper that are rolled onto tubes, and assorted bottles of glue and paint.  To do this I decided to put racks on the door. I threw in some drawers because I figured they would be useful.  One odd design element was two skinny cubbies on the left, which are there so that you can open the drawers without having to open the door more than 90 degrees.


The drawers can open when the doors are only open ninety degrees. A friend gave me some leftover quarter sawn sycamore for the left hand rack. I used the some maple scraps for the right hand rack.

For hardware I wanted to give knife hinges a try because I’ve always liked their elegant, minimalist look.  I decided to use rare earth magnets to hold the door closed, with a magnet on the bottom edge of the door and another one in the case.  This would be elegant and unobtrusive.  However, experimentation with this idea suggested that the magnets weren’t strong enough, so I did a direct and inelegant magnet-to-metal closure. Even this seems surprisingly weak: it holds the door closed but it won’t pull it closed. It seems like my kitchen cabinets close more strongly with weaker magnets.  Note also the layout of the drawer pulls:  I placed them all the same distance from the bottom edge of the drawer, which looks a bit odd if you view the cabinet from directly in front, but it looks right when you see the cabinet from a normal viewing angle, where the top of the top drawer is hidden by the cabinet’s case.


This door rack was meant to hold all the bottles. I had a lot of trouble figuring out the right dimensions. What can I put in the tiny shelf in the middle that is empty right now? Another observation is that the rack is full already.


This rack holds tubes with oversized sheets of paper. Plenty of room remains for expansion on this side.


The top provides plenty of space for using the paper trimmer.


The back of the cabinet is finished. This required a lot of extra work. Did it make sense, given that I’m planning to put the cabinet against a wall?

Construction Notes

It’s been a while since I made a piece of furniture out of plywood. I think it will be a long while before I do it again. Plywood and solid wood each call for different tools and impose different design constraints.  I’m not set up for plywood.  I cut my parts with a circular saw and some of the cuts weren’t perfect. Truing up edges of plywood with a hand plane is not a pleasant task. I realized one of the reasons for frame and panel construction beyond dealing with wood movement: all four sides of the panel are (mostly) side grain, which is easily planed.

I decided to join the case together using dowel joinery, so I made a jig by drilling holes in a scrap of wood with the drill press and tried to fit a test joint together. It required huge clamping forces to close the joint and it would pop back open the moment I removed the clamps. I figured the holes must be slightly crooked due to slop from using a wooden jig, so I bought metal bushings and tried again, but I couldn’t get that to work either. I gave up and bought an expensive commercial jig, which worked fine; the need for that extra jig was kind of annoying because I don’t anticipate doing a lot of dowel joinery in the future. (What could I have done instead?)

Another problem I didn’t notice soon enough is that all of the plywood was warped. Once the joints were cut and I fit the case together I noticed that the sides bowed out and the bottom bowed up. I tried to force the bottom flat by gluing a stiffening piece of solid wood on the underside, but my piece of wood was only 1.5″ thick, which was evidently too wimpy to help. When I glued in the dividers I used clamps to force the sides flat. This process was successful. However, I neglected to pay attention to the top, which had previously been the only flat part of the case. Forcing the sides flat caused the top to bow upwards, so the top is not flat.  If I’d been paying attention I could have corrected this before the glue dried.  With solid wood I could plane the top flat but not with plywood. So I’m stuck. It seems like the solution to this problem is to assume that every piece of plywood will be warped and to build into the design some sort of straightening element.

I chose a mitered edging for the doors and I realized a disadvantage of mitered edging: the nice corners of the miter get messed up if you have to plane the door to fit. Of course, the doors turned out to be warped as well. I hoped that attaching the racks would help hold them flat, and I think it did help, but they remain slightly curved, which complicated the fitting of the knife hinges.  I tried to tune the fit of the doors in the opening by shimming the hinges, but this shifted the hinges out of alignment with each other.  They got harder to turn and I found that brass was wearing off into the hinge joint.  I’m hoping that brass will wear until the hinges work smoothly. The warped plywood also created mismatches where the doors met each other. At the bottom the right door sticks out from the case. I tried to conceal this by rounding off the edge of the door. At the top the doors don’t meet. I couldn’t think of anything extra to do that would actually make this look better. At least I was able to get a small, uniform gap between the two doors.


Note the mismatch of the doors at the top. I think this occurred because of the warped case. Efforts to shim the hinges didn’t fix it.


A modest roundover and slight shaping actually hides this problem rather well.


The right hand door sticks out from the case it looks pretty bad.


I worked the edge extensively with a plane and spokeshave to make the doors meet, and the result is surprisingly good. A casual observer doesn’t notice anything amiss.

I made the space for the drawers 18″ deep, but I realized belatedly that once the drawer front was taken into account, a full 18″ long drawer slide might not fit. Fortunately some manufacturers’ slides are 450 mm (17.7″) instead, but this limited my options. I wanted overtravel because the drawers would be underneath a 4.5 inch overhang. Accuride tech support insisted that for twenty inch wide drawers I needed to use their premium slide (7434) because of concerns about racking even though the weights would be very small. The cheaper one is explicitly listed only for drawers less than 16″ wide. I wonder if this is really necessary? Luck was with me and I was able to find exactly the slides I needed on ebay for half price.


The drawer slides cover the dowel pins, but one is visible on the top drawer.

Because the cabinet is utilitarian, and because the drawers are hidden inside, I made the drawers entirely out of Baltic birch plywood. I cut the drawer fronts so that the grain runs continuously across all four drawers, but the grain of the Baltic birch is so mild it’s hard to notice. I made the drawers with rabbets in the front, which I cut on the router table, and I pinned all four joints with quarter inch dowels. Even so, the drawers did not come out perfectly square and true. The process of fitting the drawer slides was long and tedious. I discovered that at least in one case, the reveals changed when the drawer was loaded. It seems like fitting solid wood drawers is a simpler and less frustrating process. (When I installed metal drawer slides in the file cabinet I also found it difficult and time consuming.) I had thought about using an applied front, but wasn’t sure how I could hold them in place accurately enough and long enough to screw them on. I think the applied front is the way to go when dealing with these slides. Maybe three minute epoxy is the answer.


Note the gap between the cabinet wall and the top and bottom shelf. The center fixed shelf has no gap.

I bought some inexpensive shelf support sleeves and pins, thinking that the pins would slip neatly and snugly into the sleeves. It seems that the hardware makers have a different concept: the pins sit loosely in the sleeves. This is not satisfying. Brusso makes expensive solid brass pins and sleeves that I might try in the future, but they seemed too expensive for this project, especially considering that I wanted to have shelf holes every 3/4 inch for fine adjustment of the shelves. I selected some L-shaped brackets. They fit well in their 1/4″ holes, but I don’t really like these either because they force you to leave a larger gap at the end of the shelf. The handles I selected taught me another hardware lesson.  They look good, but they are hard to use because you have to bend down to reach the graspable part of the handle. Even my short kids have complained about this!

I edged all of the shelves with the iron-on walnut veneer edging. This stuff is easy to apply, but trying to get a nice finished result seems difficult unless you like sharp corners on the edge of your work. As I sanded I would start to sand away the edge of the walnut and expose the edge of the top layer of veneer underneath. It’s a good thing the Baltic birch has very thick veneer layers, or I probably would have exposed more than one. I don’t think I’ll use this type of edging material again because of this. Another sanding difficulty was that I found it remarkably easy to sand through the walnut veneer on the plywood, and difficult to notice that I had done so until much later. Advice on fixing this problem involves coloring the wood with pencils or something like that. I must not have the right pencils, because when I tried that it just looked worse. After I realized the dangers I felt very nervous about sanding. Should I just leave that defect in the wood because it’ll look even worse if I sand through the veneer?

I  learned many things with this project, and despite the various problem I have identified above, the cabinet looks decent in the same room with the walnut coffee table, and it meets the goal of being utilitarian—it works.


Roorkhee Chair v. 2


My original Roorkhee chair was comfortable, but it was very low to the ground and deeply reclined. I prefer a more erect seating style, and thought I might raise the chair and make it less tilted. This proved to be considerably more difficult than I anticipated. It was easy enough to test the effect of taller legs by putting the first one up on blocks, but what happens if I decrease the seat tilt? I made some new back legs in pine with holes in different spots for testing and I found that raising up the back to make the seat more level wasn’t enough by itself. It wasn’t comfortable like that. It needed some other changes like a decrease in seat depth. I didn’t have a good way to prototype that. Decreasing the seat depth also leads to decreasing the width (if you want to keep the frame square to avoid complicating the assembly) and it’s not clear that would be a good thing.  In addition it seemed like the location of the pivot point for the back rest needed to change.  What seemed like a small change to the design turned out instead to requires global alterations.  So I gave up and made a chair whose seat is farther from the ground, but with no other changes.


Another change with version two was the choice of leather. For the first one I used discount leather. That leather was too stretchy, especially for the arms. So this time I took a look at leather from Wickett & Craig, which Schwarz had used in some of his later chairs. Schwarz used the Oiled Latigo, but when I looked at the samples, all of the oiled Latigo leathers were almost black. I ended up choosing the English Bridle leather which is apparently the same leather, not oiled. The Latigo comes in a 6-8 oz thickness but the English Bridle is not split, meaning the leather comes off the cow the thickness it is. I could have paid an extra $1.10 per square foot to have it split and re-dyed, but decided to go ahead and use it as it came in a 8-10 oz weight.

One problem with the original Roorkhee was the narrow width of the arms. I noticed that I would try to rest my arms on the arm rests and they would fall off, so I corrected that in this version with arms that are wider.


The wider arm rest is definitely an improvement.

For the first chair I cut the leather and put it on the chair. But I subsequently learned about finishing the edges, which definitely produces a much nicer look. To do this I first beveled the edges with a leather edge beveling tool. Then I applied burnishing compound and burnished the leather vigorously with a leather burnishing tool. Finally I applied edge finish to the burnished edges.


The top edge is rounded and smooth after burnishing and edge painting.


The roundover on the front edges was intended to make the backrest more comfortable, though the leather is so stiff that it may not matter in this case.

For this chair I made the legs from canarywood scraps I had on hand—the only 8/4 material I had enough of. The wood is pretty, but I think it is harder than the ideal for hand powered reaming. (It is a good bit harder than hard maple.) I had more trouble cutting the mortises than on the first chair, and the finished joints were worse. I used maple dowels, and I ended up making the back supports out of some bland apple sapwood. From sitting in the original chair I had noticed that my shoulders and back would rub uncomfortably on the edges of the backrest supports, so for this version I was careful to give them a big roundover, though in the end it may not have mattered so much because of the heavy weight of the leather.

Here you can see the process of folding over the leather of the back rest. On the first chair, this part of the project was a little irregular, so I took more care this time, making measurements, keeping everything parallel, and clamping the pieces together before making the holes for the rivets. As it turns out, this particular measurement produced a very snug fit with the stiff leather I was using.


I determined the fold for the seat back by clamping the rail flush with the edge and folding the leather over for a snug fit. I marked the fold, removed the rail, and then clamped the leather down to punch the rivet holes.

The original design calls for square legs with a turned recess near the top and a turned taper towards the feet.  Schwarz implies that the turned recess at the top is important because it gives you a place to grip the chair.  My first one didn’t have this.  For the second one I cut an octagonal recess.  In order to do this I made a series of parallel crosscuts with the bandsaw and then cracked out the waste.   This produces a rough square recessed area.  Then I smoothed it out with the router plane, chair2_sm and then I marked out the proper edges and used rasps to remove the corners to transform the square into an octagon. The result looks pretty good, but in the end, the only function it serves is that it gives a space for the leather that is wrapped across the backrest to go when the backrest tilts.  I never grab the chair by these “handholds” because the frame just twists.  If I do another chair I think I might just make a single shallow cutout on the inside of the back legs for the leather and skip the rest of them.

Unlike the recess in the middle of the leg, the taper to the feet is simple to do without a lathe. I started by marking the octagon on the bottom of the leg and choosing the length of the taper to be 10.5 inches. Hand planes quickly removed the corners of the legs up to the octagon’s edges, leaving the shape shown below.
legtaper1Next I measured from one triangular facet to the opposite facet to find the spot where the thickness was equal to the 1.75 inch thickness of my leg stock. This is the right spot to start tapering the legs to get a regular octagonal cross section. I marked this starting line and then tapered down to the octagon’s boundary marked on the foot, finally obtaining the completed tapered octagon:

legtaper2When I picked 10.5 inches I figured vaguely that the length of the taper on the square sides would be a bit smaller, perhaps by a factor like the square root of two. But working out the correct formula would have revealed that if I tapered the corners 10.5 inches I would only get a taper on the faces that went up about 4.25 inches. If I do this again I’ll run the taper up much higher. The length of the taper on the flat side is (ST)d / (S2T) where S is the thickness of the full leg, T is the desired thickness at the tapered end, and d is the length of the taper on the corner of the leg stock.

chair3_smSo how does this chair work? Alas, it is not as comfortable as the first chair. One observation I made after sitting in the first chair for a while was that the way I constructed the seat, which seemed initially like an unfortunate mistake, was actually a fortuitous move. The seat came out considerably looser than I had intended, and this made the seat more comfortable by giving it mobility. I was careful to model this second chair after than one in that respect.  (I marked a line eleven inches from each edge and folded the leather to that line.)  However, the stiffer leather makes the back of the chair less comfortable. With the softer leather, the bottom part of the back rest conforms to my lower back and seems to give a remarkably comfortable support to the lower back. With this stiffer leather, the bottom of the backrest is simply too close to the seat and it pokes me on the butt without conforming at all.  This isn’t so comfortable. After giving the chair more use I may try trimming off a few inches from the bottom of the seat.

chair5_smBut if I do that I’ll need to solve another problem first. Often when one leaves the chair the back ends up in a strange position. I have tried various schemes to stop this from happening. A perhaps related problem is the propensity of the bolts holding the back on to loosen themselves. I tried to solve these problems by inserting a lock nut under the wing nut, by using a wing nut and a thumbscrew, by inserting leather washers to increase friction, and by inserting Belleville spring washers to try to create tension that keeps everything together. None of these things worked—it always unscrews itself with use. If I remove leather from the bottom of the seat I’ll be taking more weight from the bottom and I’ll create an imbalance that will be even more troublesome. The only solution I can see is to add weights to the bottom of the seat supports, though that still doesn’t address the problem of the chair unscrewing itself.


The backrest connected using a wingnut and a brass thumbscrew in an unsuccessful attempt to create a locknut combination that would keep the chair from unscrewing itself.