A couple of weeks ago I brought in a whole mess of rough sawn walnut from our lumber piles outside and stood it against a wall inside the shop. I let it stand there for a while to acclimate to the different conditions inside the shop. If a board is going to want to twist, warp or cup as it changes moisture levels I want to know it before I cut it into parts. I can still use such a board for smaller parts, but the longer pieces need to come from the flattest, straightest boards I can find. I move the surface planer around beside the workbench where I have enough room on the in-feed and out-feed sides to maneuver 8 to 10 foot long boards. I feed them in from where they are standing and take them off over by the band saw, where I can then slide them onto the workbench next to the planer, stacking them up as they come through. What I’m doing here is to remove the rough layer of wood making it easier to see the coloring and grain pattern of the wood. I’m removing as little wood as possible right now to achieve this clarity. When I’m done I have a pile of various sized planks stacked on the bench. Now I move the planer out of the way (it weighs almost 600 pounds so it’s a good think it has wheels!) and sort through the planks looking for the pieces I want for the first parts to be made. I stack the rejected candidates in this first round on the floor. I always start parts-making by selecting wood for the largest parts first, then working down toward the smallest. Often I can make the smallest parts from cut-offs produced in making the larger parts, but more importantly is that business about needing the flattest, straightest pieces for the largest parts. It would be heart-breaking to find that I’d cut up a nice flat piece into small parts then find that the rest are not as well suited for a large part. Once I have found the pieces I want, I match them up for color and grain pattern. The idea here is to minimize the variance where one piece of wood joins another. I’ll be making 2 panels 13-15 inches wide and almost 7 feet long. Each will require 3 pieces of wood to attain that width, and I want to hide those edge joints as much as possible. After the panel is glued up it will be planed to finished thickness and teh ugly spots will be removed. When I have achieved the best matches I can – for perfection is difficult on parts as long as these. That would require boards that show perfectly straight grain for their entire length and width. Something you don’t see much anymore because all the old growth timber has been cut and the smaller trees they’re cutting now rarely yield such perfect boards – I mark them with a big 'V' across the back face of all three boards to keep them properly oriented during the next step, which is to joint them. Jointing is done on a Jointer. (Duh!) It is called this because one of its primary functions is making very straight, square edges that can be glued together to join multiple small boards into wide ones. When we bought this machine we got one with a bed that is 8” wide and a little over 6 feet long because we sometimes need to joint long boards like these. A smaller model with a shorter bed would make it difficult to straighten a long board because a long, sweeping curve would just ride on the bed, making the board narrower with each pass, but no straighter. There are larger models too, some industrial models look like aircraft carriers, but this was the largest tool we had room for in this shop. And because when you move into the industrial grade, prices quadruple! As the boards are jointed they are laid back on the work bench so I can check to see that the joints between boards mate up well with no gaps. If I find gaps, I joint those two surfaces again. When I’m happy with them they go into the assembly room where I have two tables set up and a battery of clamps at hand, the edges get a coat of Type III polyurethane glue, I feel to make sure the jointed faces, which are down against the clamps, are mated up well to reduce ridges that will have to be removed later. I use poly glue for panelizing because this is not a joint that will need to be taken apart someday for maintenance, as most joinery will. Hopefully not for a hundred years or so, but eventually wood movement and stresses will cause the piece to need re-gluing. Where that is a possibility I use standard (aliphatic) wood glue so these joints will be repairable when the time comes. I used to use hide glue for all my joinery, but this requires mixing glue crystals with water and keeping it at 170° all the time it’s being used. And left-over glue in the pot tends to stink like something dead – because it’s made from animal hide. The extra hassles make it unsuitable for new furniture construction where traditionalism is not required and time is money. It’s still the best choice for antique restoration though. I’ll let this assembly sit in clamps overnight so the glue can develop a good hold. Poly glue takes 5 days to develop its maximum holding power as it cures but will provide a good “grip” in a few hours, especially if the joints do not have stress in them from forcing poorly fitted boards together under clamping pressure to close up gaps. The rest of the afternoon will be spent jointing the other set of boards for the second side panel so I can glue it up first thing in the morning. Then I’ll move onto making rail and stile stock. Why didn’t I make the sides in frame and panel construction? Oh, well, that’s because of that hexagonal “belly” the clock has. The pieces that will slope out from the sides will be too small to make as frame and panel, so I’d have had to switch from frame and panel to solid wood then back to frame and panel. That, I think, would look weird so I elected to make the sides of solid wood. The back, the front doors, and the base front will be frame & panel using a fancy fluted stock with a rope trim insert. That ought to be fun!
Wednesday, February 24
Friday, February 19
Today I will start chronicling our latest project – a Grandfather Clock for Harold in Chattanooga TN. But first I will mention that I have FINALLY gotten one of the Steamer Trunks completed and ready to post as available on the web site. This one is the oak with walnut trim: the all walnut trunk is still in progress. In this photo the box and lid have just come out of the finishing room after receiving three coats of semi-gloss lacquer and are about to have the hardware reinstalled. Harold’s clock started with an image he cut out of a magazine. He has an antique grandfather clock that is made to hang on a wall, with just the clock works in a fancy metal case and the pendulum and weights hanging in the open below. He wants a case for that clock to make it more traditional. Harold liked the looks of the clock in the photo, but a discussion of the cold-molding process that would be required to make the “belly” of that clock, and the time and expense involved, led him to go with the octagonal swelling instead. The joinery will still be tricky, but no mold needs to be built first. So the first step is to create a template of a regular hexagon in the size I think we will need. My high school geometry class comes in handy once again as I construct the hexagon and lay out the side walls. This is a starting point, from here I tweak it as needed to keep the “belly” in proportion to the rest of the clock. My little pencil sketch (above) was not to scale, and attempting to create a scale drawing in Sketch-up led me to believe I’d be better off (time-wise) to just work it out as I build the mock-up. It took me two days to build the mock-up out of ½” crate board (triple wall corrugated cardboard). I will not build the entire clock, just those elements that will be necessary to give Harold a look at the design and proportion and to see if his clock works fit inside as they should. I took detailed measurements while at his home, but there were a couple of things I wasn’t able to properly quantify – like what the actual swing of the pendulum would be while it is in operation, which it was not at the time. Now that I have the mock-up done I need to get it to Harold. He’s a busy fellow so that may take some time. While we work on getting that arranged, I will begin the lumber prep steps.
Tuesday, February 16
The final step in producing a set of TV Tray Tables is to package them so they will arrive at their new home in the same pristine condition as when they left here. This is not as small a matter as you may think. I often get supplies that were simply tossed in a box, some air pillows tossed in to fill the empty space and they send it off… and they usually arrive just fine. But these tables are a little more delicate. So, I start my wrapping padding the frame of the stand with Styrofoam and bubble wrap. Then I line the bottom and ends of a box with Styrofoam sheet. We have these boxes made up for us specifically for shipping these table sets. They are sized so they just BARELY fit into the Tier 2 UPS shipping schedule. Any bigger and the fees go way up. Then I begin bundling the tables into pairs. In this set we have only one pair, in a full set we would have 4 tables, two pairs. I use blocks of Styrofoam to space the tables so they don’t rub on one another during shipment. I’ll lay the other table on top of this one and use cellophane stretch wrap to bind them tightly together. Then set the bound pair of tables on the stand and bind them together. Laying that bundle into the box, I fill the voids along the sides with shredded paper and Styrofoam spacer blocks to hold them in position. Once I’ve sealed up the box, weighed it and processed a shipping label this one is ready to go out on the dock to await the FedEx truck. Time to clean up the mess and get ready for our next project.
Monday, February 15
A trio of small blocks are all that are left to build and we are ready for final assembly of these tables. I’ve made up stock for the two small leg-mount blocks (foreground) and one wide latch block for each table. Shaping these blocks is done on the router, but holding such small parts as the leg blocks with my hands while rounding over the edges is extraordinarily dangerous and a great way to loose a fingertip or two. (The router is NOT running in this photo) So instead I use a small hand screw clamp to hold the pieces; it takes a little longer to grab and release the parts with this than it would with my fingers, but it’s much safer. After shaping I sand the parts using a mop sander. I could use the hand screw clamp here too, but about the worst this device will do is to reshape my fingernails, so I just do it by hand – I’m not proud of my nails anyway. I do make sure I sand on the side that if the mop grabs the block and flings it, it will fling it away from me, not at me. Then I drill pilot holes, pre-finish the inside-the-joint faces and attach them to the tops of the legs. The latch blocks are made in a similar fashion except I also cut a dado in the center that will house the latch tab, which works along with the hole in the upper spreader to snap the legs into the open position so the table can not collapse while in use or being moved around. Exactly how we make that part is a closely guarded secret, if I show you how it’s done I’d have to kill you. We don’t want that! [5011-R] Then it’s just a matter of aligning everything just right and gluing and clamping the blocks to the underside of the tray. We can’t use screws here, so glue joints have to be done well or the thing will fall apart. We don’t want that either! Finish sanding follows, then vacuuming and tack ragging and moving the tables and stand into the finishing room for two full coats of lacquer each. After that is hard, I scuff sand them tack them off again and shoot the entire set with a third coat of lacquer. When that coat is set up hard I check the tray panels and scuff sand them again if needed then apply a skim coat of high quality tung-oil based polyurethane to just the ribbon panels. This step is necessary to protect the tables better against modern living than the lacquer alone will. Sweaty drink glasses or warm plates placed on lacquer can cause white spots or rings to form if the water is allowed to absorb into the finish. Even people who have an elegant dining table and care for it properly may forget and abuse these tables, so we add this extra measure of protection to help avoid problems. They’re done and ready to package up. We’ll look at that process in the next episode.
Wednesday, February 10
The starting steps in building the stand are like those for all other parts; cutting the parts blanks a little oversize from rough lumber, jointing one face and one edge, thickness planning the rough face to finished thickness and ripping the rough edge to finished width. Then we lay-out the shape of the parts on the blanks using the templates. I use a Forstner bit to provide a smooth transition for inside curves by boring a carefully placed hole then band sawing the straight lines and the outside curves. These cuts are all smoothed up on the stationary belt sander. Then I lay-out and counter-bore screw holes. Where the uprights meet the feet I use dowels to reinforce the joint, these holes are laid out and bored with the help of a doweling jig that insures the holes are centered in each piece. Once all the parts are made I assemble them with glue and screws then plug the screw holes. After the glue dries I trim and sand the plugs flush and finish sand the stand. This part is now ready for finishing and will get set aside until the lacquer flies. I should note that this is a *custom* stand made to hold two tables, Wanda did not need a full set of four tables. All that’s left to make are the little blocks that hold everything together.
Tuesday, February 9
We begin the process of making and installing the tray rails by setting up the big band saw for resawing. The resaw band has wicked, large teeth that will tear the flesh from my hands if I’m not careful, leather gloves are a necessity. Then comes marking and resawing the billets that were made up earlier. It is vital that I mark the billets so that after they’ve been cut I can keep them paired up and oriented for proper grain flow. More about that in a bit. As the pieces come off the band saw I set them aside in pairs, when all are done I surface plane the strips to smooth the band sawn faces and reduce the strips to the finished thickness. I also check them to see if there is any bow in the vertical plane, if so I joint that out before trimming them to finished height on the table saw. Then I change the table saw over to use a dado head and set it up to plow the groove on the inside face of each rail stock strip. This groove will house the panel assembly. I use a set-up block from the template set to get the dado head height and rip fence position right. When everything is in position and double checked, I plow the panel groove in all the strips, paying close attention to inside/outside and upper/lower edges. I finish up the rail stock preparation by rounding over the upper-inside edge of the strips. This edge will be very difficult to work after the tray is assembled, so I take care of this one now. The others will get rounded after assembly. Now comes the tricky part… unwinding each pair of rail strips so they wrap around the tray with an unbroken grain pattern. To do this I use an old box-makers trick that starts with resawing a solid piece of wood, then cutting the two pieces so each piece cut mates up with the next. I start at one end of a pair, lay out the left side rail, then the back rail. Come up to the other strip and lay out the right side rail then the front rail. With very little excess length in each strip, where the two raw ends meet at the left-rear and right-front corners, the grain pattern will still flow nicely around the corner. The other two corners are cut from the middles of the strips and matching those is easy. I set up the miter saw to cut the 45 degree miters at each corner and cut all the parts using stop blocks to insure that all longs rails are the same length and all short rails are the same length. The first of each are the guinea pigs, these I cut just a little long and sneak up on a perfect fit. Once the precise length is determined I cut the rest to the same length. When all are cut it’s time for a dry fit to be sure all went well. No glue here, just friction. When I’m sure all is right I take the rails off again and sand the inside faces. Again, these faces will be difficult to work properly after assembly, so sanding them now is a great labor saver. Then glue is applied very carefully to the joints where the rails meet the backer panels. The rails are not glued to the ribbon panels. Clamps are place to apply gentle pressure to hold the corners closed until the glue dries. No screws or nails can be used in these joints because they are so slim and delicate, proper glue joints are mandatory. I’ll let these assemblies set up overnight, then pull the clamps, route the outer edges and sand the outer surfaces of the rails. And that will complete the tray assemblies.
Sunday, February 7
I am actually almost done with these tables, but have been focusing on getting the woodworking done and not on blogging about it. I’ll get that caught up as soon as I am able. Sorry for the delays, but this order is over-due so I've been putting 12 hour days trying to get it caught up This step will be to make the ribbon panels. I call them ribbon panels because they’re made up of thin strips (or ribbons) of solid wood that are precision edge glued together to form, a wide panel for the tray top. There are several advantages to doing it this way, including the fact that if the table top should be dinged or scratched, the damage can be sanded out and refinished without destroying the table. Using veneered plywood here would mean that repairing such damage would be very difficult if even possible because the veneers used are so thin (mere thousandths of an inch) and will sand through very easily. The ribbon panels also look much prettier. The down side is that building these panels takes a good deal of time and effort, where simply cutting a piece of plywood and popping it in the slot is much faster. But we at Smoky Mountain Woodworks are not so much about speed or cheap as we are about quality. We want you to have furniture that you will be proud to own – and proud to hand down to the next generation. We start with pieces of wood that have been jointed and surface planed so that the faces are absolutely flat and smooth. These faces will become our glue joints, so there cam be no defects or snipe. Bear in mind that we are making a full set of trays and stand, two all walnut tables and a custom stand for Wanda, the other two tables will be our Signature Series made of walnut and oak and will be used for photographic purposes, then will probably head off to our living room where we will conduct extensive product testing . As I cut the strips I lay them on the extension table in the same order I cut them from the billet, this preserves the natural flow of the grain pattern and makes it easier to get a good looking panel. I like to use lumber that is as close to flat sawn for these ribbon panels because when I re-saw them the strips end up being quarter sawn and displaying some strikingly beautiful grain, especially in woods like oak. When I have all the strips I will need cut (and a few extras, just in case) I begin to lay them out and check the patterns. The ideal method here is to cut all the strips for one panel from one wide board, so that the pattern flows naturally from one edge to the other. But, these days most of the old growth timber has been cut and getting wide boards that are predominantly flat sawn is rare. So I will use two or three boards per panel, and by careful attention to color and form, match them up for a good looking panel. I use a couple of strips of wide masking tape pulled tightly across the panel to secure it once I’m happy with the lay-out. If any of the strips have developed a little spring to them un cutting, I’ll use a bar clamp to pull the strips snug while I tape. Fold the assemblage in half and take it into the assembly room for glue-up. I built this lop-sided “A” shaped jig to help me apply glue between the strips quickly and efficiently. The tape acts like hinges between strips when placed the panel is placed on the jig up side (tape side) down. I use a small brush to apply Type III polyurethane glue evenly between the first strip, move the panel up, glue the next joint, move it up, and keep going until I get to the last strip. When all joints are glued I move the panel over to a set of clamps that were placed at hand, 3 underneath two on top. By alternating the clamps I can insure that I don’t mold a bow into the panel due to the clamping pressure, and I am assured that the panel pieces won’t go “SPROING” and fly all over the place (giggling incessantly) when I apply pressure with the clamps. While clamping I pay attention to the seams, getting them all as flat and even as possible – there are a lot of them to monitor as I snug up the clamps! I’ll leave this stage sit overnight so the glue will develop a good hold before I take it out and machine it further. It takes 5 days for this poly glue to develop its maximum hold (by then these tables will be on their way to Wanda) but this glue will grab well in a few hours. Giving it overnight to set up makes sure none of the joints will pop open during machining, the surface planer especially will induce a good amount of vibration in the panel as the knives strike the surface repeatedly. The next morning I pull off the clamps, scrape the surface to remove glue squeeze-out, trim the panel to rough width (so it fits through the 15” wide surface planer) and surface plane it to remove most of the roughness. Quarter sawn wood has a tendency to chip out in the surface planer, so I switch to the wide drum sander (photo) as soon as feasible for final smoothing. This tool takes of miniscule amounts of wood with each pass, so reducing the thickness very much will take a long time, but results in a superior surface in the end. Wha-La! Two perfect panels! Next I trim these panels to exact width, which is 1/8” narrower than the backer panels, and glue the ribbon panels to the backer panels, leaving a 1/16” overhang of backer on each side. This does two things: first it allows some room for the panels to expand in humid weather and it prevents the ribbon panel from getting glued to the rails during final assembly of the tray. Because these panels are solid, natural wood, they will expand and contract across the grain as the wood fibers swell and contract with the changing humidity. No finish, no matter how good will prevent that –it’s just a fact of life, we furniture makers have to learn to deal with it so parts don’t crack and joints don’t fall apart. When the glue is set-up I take the panels to the table saw equipped with a large miter sled and trim the ribbon panels to be even with the ends of the backer panels. Why don’t I leave a gap here like I did on the long sides? Because wood dos not expand and contract along its length, only across it’s width. These edges can be flush, and in fact making them flush will help to strengthen the tray against racking. Now I finish sand the panel assemblies, front and back. Finish sanding them well after the edge rails have been added will be far more trouble than doing it now. When I’ve gone through all the grits I wipe the panel down with clean mineral spirits to simulate a finish, this is how it will look when I spray on the lacquer – are there and glue spots, or chips, or crazy grain that need more work? Now would be the time to fix them, not after the first coat of lacquer is on. Once the mineral spirits has evaporated I pre-finish the long edges of the ribbon panels. This is so that when the panels contact a narrow ribbon or raw wood does not peek out from under the rail edge. Yes, it is possible that capillary action will pull some finish into the joint between rail and panel, but I don’t want to count on that. Yes it’s possible that even if a sliver of raw wood exposes itself that no one will even notice, but I don’t want to count on that either. I’d rather do it his way and be sure. This completes construction of the panel assemblies. Next up, tray rails.