• View
• Edit
• History
• Print

Wood Coasters

Why Coasters?

I think that a common project for woodworking beginners is a cutting board. You find some interesting species of wood, cut pieces that are straight, flat and square, glue them together and then make the resulting board flat and smooth. The end result is usually attractive as long as you have a) the right woodworking equipment and b) some aesthetic sense to select the wood and the pattern. If something goes wrong, you can figure out what it is and what you need to do to make it go right in the future. So why, instead, would I make a coaster? Well, 1) it seems that fewer other people are doing it, 2) it's less wood to throw way if it goes wrong, 3) it's a smaller object to have around if everything goes right, 4) if you give a coaster to someone, as opposed to a cutting board, they are less likely to think "Good grief, what am I going to do with this?", 5) to finish a cutting board, after gluing the pieces together, you really need a thicknesser of some sort, whereas a coaster, being smaller, can, in a pinch, be finished by hand or by a hand-held power sander, 6) a coaster can be given a non-food-safe protective coating, such as urethane, which should increase the time for which the woods retain their fresh-cut appearance.

The First Coaster

My objective at first was to find out whether or not I could make an attractive coaster by gluing up thin (¼") wood strips, cut from boards readily available from a local store. I decided to use the domestic wood species maple (curly), cherry and walnut (the ice cream flavors). For some reason, that now escapes me, the cherry was to be a narrow band around the edge. The strongly-contrasting maple and walnut would be in strips of various widths in the "core". What I expected to go wrong was a) my crummy table saw would not produce strips with parallel sides and/or edges properly square to the sides and b) when I squeezed the strips together after applying the glue, they would bow into a curve or even fly up in the air. I made the strips longer than necessary, expecting problems at the ends. Laid out dry, they seemed OK. To clamp them while the glue dried, I made a jig from the cast iron bed of my old, defunct, table saw, by jamming scrap wood strips in the miter slots and packing more scrap between them and the coaster core. To try to hold it flat, I put a foam block on top and piled on top of that everything heavy that I could find. It worked amazingly well. So, I trimmed the core to its final size, about 3 5/8" square, and set about making the band. This was really tedious and time consuming. I have a 45° miter jig for my cross-cut sled, so cutting the angles wasn't a problem. The problem was the length. I had to shave a tiny bit off at a time and test fit the result, all of this four times. I resolved to either not use an edge band again or to make a jig specifically for cutting the edge pieces to the correct length. For the glue-up, I had to get even more inventive to get pressure on all four sides, but in the end the result was good. Obviously, if I was going to do this again, I would need to design and build a glue-up jig. There was easily noticeable unevenness on both sides, strips being higher or lower for no reason that I could figure out. I corrected this by holding the coaster to the bottom of my bench-mounted belt sander and then hand sanding to remove the scratch marks. I have no idea how flat and true the result is, but it looks OK, and that was good enough at the time. I applied three coats of General Finishes Enduro-Var water-based urethane. Then I cut a piece of self-adhesive cork sheet to 1/16" smaller than the coaster all around and applied it to the bottom. The end result was what you see in the photograph above-right. Aesthetically, this was a big success. I took the finished coaster along to a family meal to show around on a "this is the kind of thing I get up to in retirement" basis. Everyone oohed and aahed in what appeared to me to be genuine admiration. My sister-in-law asked outright if she could keep it. I said no, but that I would make her two new ones, and so my future as a coaster maker was cemented.

Yellowheart and Padauk Coasters

Who knows where inspiration comes from? I certainly don't. On one of my trips to my local woodworking store, I picked up a small yellowheart turning blank and a 24" by 3" by ¼" piece of African padauk. Neither species looked promising, but I felt the need to branch out beyond the ice cream flavor woods (maple, walnut and cherry). The yellowheart was especially problematic as even when cut into ¼" thick strips the color and grain were both uninspiring. At some point, my brain decided that if the colors weren't going to carry the day, it would have to be geometry, and a design with diagonals popped into my head. So, I calculated some dimensions to make reasonably efficient use of the wood, made the cuts and glued up what I thought would be the core (note that the angled cuts required the use of clamps on my cross-cut sled, and, in this manner, were perfectly safe). At that point, thinking about which wood to use for the edge band, I decided that an edge band would just not look right. Unfortunately, in anticipation of the band, the side-to-side dimension was about 3 5/8", rather on the small side, but I just went ahead anyway, leaving the other dimension closer to 4". Using my router and a chamfer follower bit, I put a very small bevel on the top edge all around. After sanding on the belt sander and by hand, I applied the first of three coats of urethane. Much to my surprise, the colors of both woods brightened and grain details that I had not noticed before popped out. My wife loved them. A few days later, a neighbor came over for afternoon tea and declared them "spectacular". Of course, I offered them to her (knowing, now, that I could make more), but she declined, saying that she had no use for coasters. I pointed out that she was not interacting with them as coasters, but as objets d'art. So, I made two more (not making the mistake of sizing them as a core for an edge band) and took the originals over to her (the size mistake won't matter if she doesn't use them as coasters). She accepted at once. Now I've used up all my yellowheart. Disclosure : the photograph is of the second pair.

Here (mainly for my own records) is the "recipe" for these coasters.

Live-Edge Curly Maple Coaster

Picking through the curly maple at my local woodworking store, I was usually disappointed by the small degree of figuring in the wood. It looked nothing like the picture on the Web site. A store employee suggested that I rummage among the turning blanks and find a piece with good figure, which I could then resaw myself into ¼" strips. Having no idea whether this would work, but wanting to find out, I selected the most strongly figured piece that I could find (which was not easy to determine as the pieces were rough sawn). Well, it did work, and I used the resulting wood unless I needed a piece wider than the turning blank (1½"). One corner of the blank intersected the natural edge of the tree from which it was cut. I did not use the pieces cut at this edge until I hit upon the idea of making a coaster to feature the "live edge". I chose a simple, mirror-symmetric design with walnut in the center flanked by narrower strips of padauk. The glue-up was simple, there being no band, but it took a lot of work on the belt sander to get rid of unevenness in the strips. The finished coaster (with, again, three coats of urethane and self-adhesive cork sheet) is slightly dished, which I think is visible in the photograph at above / right. Oh, well.

Coasters for the Mother-In-Law

As the years pass, my mother-in-law gets more and more generous with Christmas and birthday gifts. It's always money. I try to spend it on something specific that I can thank her for. This year, I decided to spend it on woodworking supplies. So, she is partly responsible for me suddenly taking up woodworking! My wife and I are going down to Florida to be with her for her 97th birthday (update : that didn't happen due to COVID-19). It seems appropriate to take as a gift some product of my new leisure occupation. I don't think she will be very critical, so I came up with a simple, mirror-symmetric design of curly maple and bocote with an African mahogany edge band. Bocote was a new discovery, having suddenly showed up in my local store. I also wanted to revisit the difficult process of cutting the pieces for the band by designing and using a jig that would produce a strip with mitered ends of exactly the right length without "sneaking up". This worked very well. These were also the first coasters that benefited from my newly-constructed drum sander.

Here (mainly for my own records) is the "recipe" for these coasters.

Déjà Vu All Over Again

Look familiar? Remember my sister-in-law wanted to keep the first coaster I made? And I said no, but I would make her two new ones? Well, here they are.

Here (mainly for my own records) is the "recipe" for these coasters.

Bolivian Rosewood

The woodworking store that is most convenient for me seems to want to sell machinery and gadgets more than wood. In theory, it has a good variety of domestic and exotic woods in a wide range of sizes, but, when I go there, the wood stock seems picked over and neglected. Many species aren't in stock or are not in stock in a size that I can use. What is in stock is often horribly warped, flawed or has an uninteresting grain pattern. Prior to the trip I am about to describe, my most interesting finds were a small yellowheart turning blank, an African padauk ¼" panel and a badly warped bocote ¼" panel. To get curly maple with a decent figure, I'd had to buy a large turning blank and saw it down.

But on my fourth (or was it the fifth) visit, it appeared that I had arrived soon after the wood selection had been re-stocked. I came away with nicer and larger pieces of African padauk and bocote plus purpleheart, bubinga, zebrawood, leopardwood and Bolivian rosewood. And $200 poorer. The last one, the rosewood, really caught my eye and I wanted to do something with it immediately. But I didn't have any ideas as to exactly what. Obviously one or more coasters, as the panel was only ¼" thick. But I had no thoughts on a pattern that would show off the wood's grain. The first 7½" of one of the 24" long panels had a bad flaw at one end and a hole at the other. I decided to just cut this piece off and do whatever came into my head. The un-flawed part could wait until I came up with an idea.

I cut off the flawed part, which was about 4" long. It fell into two pieces. (For what happened to the 3½" piece with the hole, see the next section.) I removed the flawed edge from both pieces with angled cuts. I found a piece of curly maple about 1" wide that, when placed between the two pieces, showed a grain figure that ran directly across the width, following the angle of the cut that removed the flaw. Measuring the total width, I needed about another ½" to get close to 4", so I picked out two more narrow strips of curly maple. I glued all this up, leveled both sides, trimmed to 3¾" square and routed a slight 45° bevel around the top edge. Then I applied the urethane finish and cork sheet on the bottom. The result was very pleasing, especially considering how spontaneous the design was.

Here (mainly for my own records) is the "recipe" for this coaster.

Bookmatching Experiment

As a result of circumstances described in the previous section, I found myself with a piece of Bolivian rosewood ¼" thick, 3" wide and 3½" long, marred by a small (natural) hole surrounded by very fragile wood. The interesting grain pattern inspired me to try the technique of bookmatching. I started by trying to saw the piece in two, on the table saw, through the thickness. This wasn't easy and didn't work out well. I started the cut at about ½" depth and used a piece of ¼" thick scrap to adjust the fence for an equal amount of wood (about 1/16") on either side of the kerf (about 3/32"). I made the cut on both sides of the rosewood and then increased the depth and cut again. And again. I held the wood against the fence with a push stick, being careful not to press it against the blade, while I fed it into the saw with my fingers right at the top and my hand wrapped around the fence for protection in case anything went wrong. At this point, I had the two thin sheets I was looking for, but joined by a thin rib along the middle that needed to be sawed out. I should probably have moved away from the table saw at this point and used a hand saw, sanding away any residual ridge. But, I tried to make one more cut to remove the rib. Nothing really bad happened, but without any way to hold the sheets in place, they wandered into the blade and came out irregular in thickness.

OK, so now the cut surfaces are a mess. But, having removed about 3/32" of wood, I can see that, in any case, the grain patterns are not at all obvious mirror images. The pattern shifted too much in the thickness of the kerf. Perhaps I have learned that bookmatching only really works with veneer, when the sheets are separated by a knife rather than a saw. It turned out that the outside faces of the thin sheets matched about as well as the damaged inside faces. So I decided to carry on anyway.

I needed a 3½" square "core", around which I would put a ¼" wide edge band to hide the fact that the core is only 1/16" thick. Working around the hole, I was able to cut almost 1¾" from each piece for this purpose. The rest went to make a species sample to add to my growing collection. To get up to the 3½" width, I added a very thin strip of maple. This also helps to disguise the failure to properly match the grain. I glued the thin sheet to some 3/16" MDF which resulted in the ¼" thickness of my coasters and covered up the mess I'd made sawing the original piece in two. After the edge band was added to get a 4" square coaster and the cork sheet was stuck to the bottom, the MDF was completely hidden. Although my attempt at bookmatching was basically a failure, I learned a lot and ended up with an attractive coaster anyway.

Here (mainly for my own records) is the "recipe" for these coasters.

Some Woods I've Been Ignoring

Having worked off my obsession with rosewood, I decided to make some simple designs using woods that I've bought but so far haven't done anything with. I'm remembering yellowheart and padauk, which looked much better in a finished coaster than they did when I picked them up in the store. Maybe I have another sleeper in my wood box. The results are shown in the photograph on the right.

The leftmost coaster (of which I made a pair) is mostly leopardwood with stripes of wenge and maple. The point of this was to showcase the leopardwood, the other species being just for contrast (and to pad the 3" width of my leopardwood stock to the 4" that I like for coasters). I noticed that the flecking of the leopardwood changed from broad to narrow about two thirds of the way across, so I cut along this line to insert the accent pieces. I thought the effect was quite subtle, but my wife noticed and appreciated it at once.

The middle coaster (of which I also made a pair) is wenge with a narrow strip of maple and then an unidentified wood that I call "any fruit tree". The wenge has a very interesting grain pattern, but, unfortunately, it is very hard to see because of the dark color of the wood. Unfinished or finished, it's the same. You have to hold it at the right angle to the light to see the pattern. I think I will in future use this wood only as an accent. The story of "any fruit tree" is that I was picking through turning blanks in the store looking for woods that weren't available as sheets. These are marked with a very terse two- or three-letter code on the end. I found sycamore ("SYC"), jatoba ("JAT") and yellowheart ("YH"), but the chart nearby did not translate the code "AFT". I consulted with an assistant, who agreed that it was a mystery but thought perhaps it was mango. I said that it looked interesting and he began a sentence "Oh, yes, any fruit tree ...", whereupon I interrupted to say that "AFT" obviously stood for "any fruit tree". He laughed and I tossed the blank into my basket. I'm not sure, in the end, what I will do with it, but now I've used up half of what I bought.

The coaster on the right (of which I made only one) is bocote, jatoba and maple. I have only a small amount of jatoba, cut from a small turning blank. Like padauk, the wood seemed to offer little promise until I applied a urethane finish and then it came alive. I'm please with the design, which is an adaptation of the design of the yellowheart and padauk coasters, above. The bocote on the left and right sides has a fairly straight grain pattern so as not to confuse the diagonal element. I also experimented with the maple (I'm running out of maple with a good curly figure) by turning the ¼" sheet edgeways to get a quarter sawn view of the growth rings (this does not show in the photograph).

Here (mainly for my own records) are the "recipes" for these coasters. Also on this page are larger photographs of each design of coaster.

Walnut and Maple Chevron Coaster

Although I've decided not to make cutting boards (explained at the start of this page), I see no harm in watching YouTube videos of people who do if they have an idea that I can use in a coaster. In the video Chevron Cutting Board Tutorial, "Around the House" makes a board out of walnut and maple that appears three-dimensional when seen from the correct angle. It is more complicated than cutting strips and gluing them side-by-side, but no more complicated (I think) than my edge bands. What I suspect will give me trouble is cutting into strips the result of the first glue-up on a 45° diagonal. But, nothing ventured, nothing gained, so I set about making a small, thin, square replica of the cutting board in the video. I have the same number of zig-alternating-with-zag strips (seven), but fewer zig-zags in each strip. As I expected, cutting the strips did not go perfectly and the joints are not as tight as I would like them to be. The three-dimensional effect is not as strong as with the (larger) cutting board. Perhaps I will try again, perhaps with different woods.

Here (mainly for my own records) is the "recipe" for this coaster. Also on this page is a picture of a coaster made largely from the sections that would otherwise be wasted after the first glue-up (thumbnail image at right, right-click to view).

Purple Lover's Coaster

To be clear, this is a coaster for someone who loves the color purple. I am not in a romantic relationship with Barney the dinosaur. This person loves purple more than I've ever known anyone to love any color, so it stands to reason that I need to make her purpleheart coasters. Since the color is the whole point, I won't mess about with complicated patterns; I'll just take a piece of purpleheart, cut it unequally in two and stitch in some accent pieces, padding the width up to the 3 7/8" that I've settled on for most of my coasters. After holding all my wood samples next to the purpleheart, I decided that anything yellow, gold, red or brown would not look good. Pale maple looked OK, but not too much of it. Dark wenge looked OK, but didn't produce enough contrast. So, I decided to add both woods and use the maple to delineate the wenge. Just a single stripe seemed too plain, so my final design uses two very narrow strips on either side of the main wenge insert with a narrow strip of wenge separating them. I'm satisfied with the result. I don't personally think it's one of my best coasters, but then I'm rather indifferent to the color purple. I will probably use the narrow strips idea with other woods in the future.

Here (mainly for my own records) is the "recipe" for this coaster.

Pictorial Coaster (of a Fish)

Casting around for ideas for coaster designs and, at the same time, people to give them to, I thought of Cyndi "The Fish Lady". We order fish from her every week and pick it up at Brookline Farmer's Market in the summer and at When Pigs Fly Bakery out of season. (If you haven't cooked and eaten fish pulled from the water that very morning, you should try something like this. Especially for bluefish, which deteriorates rapidly after being caught.) Anyway, I decided to make a coaster to give to her depicting a fish. This would give me a chance to try a pictorial design. I figured that I could make a convincing fish with a small number of straight-edged shapes set in vertical strips of the background wood. I would have contrasting narrow strips representing reeds in the water. Also, on my list of things to try was drilling holes. This would be a way to represent the fish's eye and bubbles rising from its mouth. My wife gave me the idea of using sycamore for the water, exploiting its ripply grain pattern. I don't have (or even know of, at present) a greenish wood, so the reeds are yellowheart. For the fish, I chose a jatoba because its a plausible color without any distracting grain and I had some small scraps to use up. I felt that this would be a painstaking and time consuming project and I was not wrong. I'm pleased with the results, but I don't think I will come up with any more pictorial designs this complicated. I like the holes-as-bubbles and I think the eye came out well.

Here (mainly for my own records) is the "recipe" for this coaster. It is long, complicated and not really complete, so you probably won't want to read it. Also on this page is a photograph from a different angle (thumbnail image at right, right-click to view) that better shows the holes-as-bubbles.

Did anyone get the pun right at the beginning of this section? "Casting around ..."? Casting? Fish? No? Oh, come on!

Coaster Set for a RISD Alumna, Designed Using SketchUp

After the pictorial fish coaster, I felt a bit burned out and lacking in inspiration. I had, however, committed (at least in my own mind) to making a set of four coasters for a friend who has impeccable taste in interior design. She attended RISD a few years behind Tina Weymouth. She suggested (some time before) that I make a set of coasters all using the same wood species but with different geometrical designs. I asked how many should be in a set and she said a minimum of four. Now, she wasn't thinking that these coasters would be for her — that was my idea.

But, how to come up with the designs? I don't want to make too many ho-hum coasters and four would be too many right off the bat. I was thinking of drafting the design in a simple image editor, such as MSPaint. That would give me a view of the colors of the wood species. Then it somehow came to my notice that SketchUp was no longer a Google product. So after a bit more research, I installed SketchUp Make 2017 and set about learning it. This was not too hard as I had experience with 2D drawing programs and a 3D game level editor. Creating the basic coaster shape and dividing it into strips for the various wood species was easy. It took a little time to figure out how to turn a photograph of a piece of wood into a SketchUp "material" so that it would render in a convincing manner. Once it seemed that everything was coming together, I installed the Twilight Render Hobby plugin. The image above / right was rendered with the '07. High+' preset. This is more realistic than I anticipated and would help me design coasters going forward.

My first design decision was that if I was going to make four coasters I would not give them an edge band, so as to avoid the extra work. My second decision was that to be a "set", the coasters needed a unifying theme stronger than just using the same wood species. If one was all straight lines, one had the simple diagonals of the yellowheart and padauk coasters, another was chevrons, etc., then they would look like an assortment rather than a cohesive set. My third decision would be the wood species. This was limited by what I had available in sufficient quantity, as this all happened during the COVID-19 lockdown. The prospective recipient of the coasters had given me a very thin piece of strongly-figured tiger maple. I had a good-sized piece of ¼" thick walnut which would contrast nicely with the maple but had a muted grain pattern so as not to steal the show. Wenge is very dark, almost black, and thin strips would sharpen the divide between these two woods. So, I could make four coasters by varying the widths and positions of one or more maple strips, bordered by wenge, in a walnut background. But, again to reduce the amount of work, I wanted to make two coasters at a time by gluing up double-length wood strips and then cutting the result in half. So, my fourth decision was to add strips going across the coaster. This was something I had not tried before - but I would be able to see what it was going to look like using SketchUp! For this, I made the bold choice of African padauk, which is orange / red, but, again, if this didn't look right after mocking it all up in SketchUp, I would be able to change my mind before cutting wood.

Now that I had a clear design concept, I needed to refine it by choosing actual dimensions. One coaster pair would have a narrow strip of maple, but still wide enough to show off the grain figure. This would be slightly to the left of center. The other coaster pair would have a wider strip, further off-center. The wenge strips on either side of the maple would all be the same 1/16" width. To differentiate all four coasters, one of each pair would have a single padauk strip across it and the other would have two narrower ones, close together and in a different position. This looked good to me in SketchUp as I first drew it, but I could have easily fiddled with the positions of all the strips until it looked right. I created "materials" to apply to the faces in the model from photographs of the wood and adjusted the colors so that what was on the screen matched the wood I was holding in my hand. Then I arranged the four coasters in a way that would make a good photograph and rendered the scene. The result is attached here, above / right.

Actually making the coasters was not too hard. There was no edge band and there were no angled cuts. The very thin wenge was a small problem as the strips tended to vibrate and wander into the table saw blade, but I solved that by cutting them a little thicker and then running them through the drum sander. I glued the wenge onto both maple strips first, stacking them side-by-side in the glue-up jig. Next I glued on the walnut for one coaster pair and then for the other. (The glue-up jig holds two coasters at a time, either separately or as a double-length piece that will later be cut in two.) For each step, I allowed the glue at least an hour to set up. While this was happening, I worked on cutting the next pieces of wood that were needed. First squaring off the ends, I cut two sections off each coaster pair glue-up for the individual coasters with the single cross-ways padauk strip. While this was in the jig, I cut the three sections needed for each of the coasters with two cross-ways strips and discarded the minuscule scrap left over. After the final glue-up of these pieces, I trimmed the coasters to their final dimensions (3 7/8" square, which I have adopted as a kind-of standard), beveled the top edge and sanded and finished them (which is a lot more work than it sounds). The next day, I arranged the coasters in the same way as I had in SketchUp and photographed them under a cloudy sky. This photograph is also attached here, to the right.

Comparing the SketchUp rendering (above) and the photograph, two things are immediately apparent. First, the appearance of the geometry of the wood pieces is captured exactly, validating the use of SketchUp to come up with pleasing shape combinations. Second, the color is completely different. I don't see this as a huge problem, but I could make some guesses as to how it happened. I did not use the exact same pieces of wood to set the color of the SketchUp materials and to make the coaster. The dominant factor is probably that when adjusting the materials I was working on the computer indoors, probably under artificial light, and when taking the photograph I was outdoors.

As I sit here now, indoors in the day time, in a room with pale apricot walls, and compare an actual coaster with the two images, its color is in between the two. If I turn on the table lamp beside me, it looks more like the rendering than the photograph. So, I'm back in SketchUp adjusting the material colors to more closely match the photograph while also looking at an actual coaster. The result in the new rendering at right. I think what I have learned is not to adjust the colors of materials in the evening under artificial light, but to do it in the daytime when there is better light. Ideally, when the weather is better, I would take the computer and the wood sample outside.

Here (mainly for my own records) is the "recipe" for these coasters and their holder, including notes on what went well and what went poorly and could be improved.

The Coaster Finish (Urethane) on Bird's Eye Maple

After the set of four coasters with a holder, all designed with help from SketchUp, I tried my hand at making small boxes. As a "keepsake" box won't, in normal use, come into contact with water, this gave me a chance to try a finish other than urethane. Teak oil was first, as we had some around the house. The resulting piece felt more like wood than the coasters (if you don't know what I mean, you could read my discussion of this topic). However, the finish was very matte and I prefer a bit of a sheen so I bought some Danish oil and applied this as a second coat. This was a small improvement. Overall, I like the feel much better. Urethane gives the impression of a layer of plastic between you and the wood. Basically, that's exactly what it is. But, the oil seemed to have given a slight amber color to the very pale bird's eye maple that I used for the box lids and also didn't accentuate the figure of the wood as much as I expected. I had cut the lids from a ¾" thick board so that there remained ¼" thick pieces of exactly the same wood. So, I turned them into coasters with a urethane finish so that I could make a direct comparison. I didn't spend too much time on the design. I added to each side first a strip of bloodwood, my first time using this species, and then bocote, which has become one of my favorites. As I expected the, maple did not take on the amber tint. I do still need to figure out whether that was due to the first application of teak oil. Maybe if I only used Danish oil I could avoid the tint. (No, the teak oil has nothing to do with it, I get the amber tint with just Danish oil.) The urethane does also seem to give the bird's eye figure a little extra sparkle as compared to the oil finishes. It's hard to see - you certainly won't be able to see it in the photographs - but my wife and I agree that it's there. So, as with all things, it's a trade off. I will probably switch to an oil finish for things that aren't in danger of getting wet because of its tactile qualities (and ease of application) and stick with urethane for coasters. According to multiple Web sources, urethane can be applied over Danish oil if the latter is allowed to dry thoroughly, estimates for which vary from days to weeks. I don't know why I would do this, though.