Back/Bowl Page1 · Page 2 · Page 3 · Page4 · Pegbox · Neck · Soundboard · Soundboard page 2
Braces · Attaching the soundboard · Fingerboard · Finishing · Gallery of the completed oud
I attached the profile pattern to the back of the bender in order for it to be readily accessible while bending the ribs. I attached it with an angled block so it tilts rearward, making it easier to check the rib bending progress.
The ribs cut to general shape. For this oud, there will be 13 ribs. The neck end tapers to about 1cm, the tail end stops at a point. The ribs will shaped to their final dimensions after bending.
The ribs soak in water in preparation for bending. 15-20 minutes seems good for this walnut.
How the ribs are bent. The technique is not as simple as it seems. It is all too easy to over bend the curve or end up with some flat spots. One can flatten the rib if it gets too tight. The curve should be smooth. I can see now why an elliptical bending iron is better--the curve is shallower, thus lessening the chance of flat spots. I found that the first rib I tried (rib #1) took about 30 minutes to get right. The next two, half as long. I learned to start at the tail end and do a small section at a time in order to prevent over bending. The most difficult part is the tight curve at the deepest part of the back. Once it gets past that and towards the neck end, it goes very quickly. I considered mounting the bender vertically, and may still try this, but this method seems to work fine.
The first three ribs bent to shape. After a few minutes of drying, they opened up a bit. A little dry bending over the iron will take them back.
After the ribs are bent, they are planed flat. I was surprised at how flat the ribs ended up after bending. One would think from looking at the flat rib cut to shape, that it would not be so. A few passes over an inverted sharp jointer plane ensures the angle of the edge is not lost. The angle is automatically planed onto the edge by the way the rib sits on the plane because of the rib's taper. Neat!
To check if the rib edges are flat, place the rib on a very flat surface. In this case, a piece of 3/4" Baltic Birch plywood. I considered using my jointer table, but this seems accurate enough. Hankey's book calls for sanding the edge flat on the same type of surface. I got pretty close with just my jointer plane, and it produces the best glue surface anyway. Nevertheless, I will probably try the sandpaper method as well to see if it is more accurate.
I decided to try the sanding technique for the edges after planing. The planing was a bit tedious. The sanding however registers the entire rib edge at once and produced an almost perfectly flat edge. When I tested it on the flat surface there was only about a 1/64" gap which flattened out with almost no pressure.
The 1st rib glued to the neck and end blocks. I began at the neck end.
I spread a thin layer of glue on both block and rib. One wants only little tiny beads of glue as squeeze-out. Any more and it becomes a mess, too little and it is starved. Finding the right amount is the tricky part. The rib is held in position with thumb tacks. I think a little roughening of the pin of the tack would help it hold better, perhaps with a file or some coarse sandpaper. I overlapped the end of the block a little to allow for planing the neck surface flat.
I glued up the end block the same way, instead using c-clamps and plastic cauls to protect the rib. I let it set up about 20 minutes only to find that both ends had slipped off center about 2mm! The rib was crooked. Luckily the glue was not totally cured. I rubbed each end against the bending iron and slipped an old Japanese Dozuki saw blade (very, very thin) between the rib and block. I did one at a time and reglued them each in their proper position. An hour later I removed the clamps and cleaned up the excess glue. The back had begun to emerge!
The second rib gets glued to the first. I jointed the edge as before, then tweaked the profile of the rib to match as close as I could the previous rib. The closer it matches the less work it is to glue it up, and the less internal stress is introduced, not to mention the back keeps a better shape as a result. After it is ready to glue, I joint the other edge so the next rib can be fitted to it. I clamped again at the neck end with the tacks, and with c-clamps at the tail end. The rest gets squeezed together with hand pressure and taped. I made sure the ribs lined up along their thickness as close as I could. This leaves the most material for smoothing. The rest of the ribs (except the last two) are fitted and glued this way.
The two ribs with clamps and tape removed.
Here one can see a section where the ribs are a bit offset and not totally flush. This is not a gap in the glue joint. The offset is only about .3mm so it will sand out easily.
A shot of the neck end. The glue joint is satisfactory.
The tail end. This was very tricky to glue up. The glue joint is visible--but only about the width of a hair or two. It is very tight though right where the end block tapers off. The joint went together fine on a dry run. There is simply no way to draw the two ribs together very tightly without some sort of special clamp or caul. Masking tape will hold a good joint while the glue dries, but in this application it is not the best. I tried applying a c-clamp at an angle to perhaps draw the joint together, but this did not work. I will try to make the rest better, but I suppose this being my first oud, I can live with small discrepancies such as this. I wouldn't think it would affect the tone. Also, I plan to cover the convergence of the points with a circle inlaid into the end block, not with a large piece of veneer as the Turkish do. This latter method would cover my mistake, but it is also a bit of a compromise in my opinion, the round inlay being a more tasteful way of covering the points.
After I got the mediocre results on the first tail end rib joint I came up with this technique to get a better joint. After I have completely bent the rib to match the previous one and jointed the edge, I press the new rib to its mate with a good amount of hand pressure and tape it in place. I then continue to push as I clamp an offcut of rib stock right up against rib. This way, when I glue it up, the joint is very tight and cannot move. Also, because the offcut is the same thickness as the rib, it does not interfere with clamping the new rib to the end block. I also made sure to rub some paraffin wax all over the little offcut so it does not get glued to the end block or the edge of the rib.
Here you can see the results of the improved gluing method. The joint is much better (almost invisible) than the first one below it. Clarification: I realized that the above text may be confusing after a discussion with Richard Hankey about the small offcut clamping technique described above. In order for the back to be built up successfully and retain its proper shape, each rib must meet flat against it's mating rib and match as close as possible its curve. If the rib needs any pressure at the ends to close the joint, it will force the center portion of the rib to flatten out. It is very important that each rib not be forced out of shape, otherwise the back of the oud will end up flattened and the shape at the front will not be round. I devised the small offcut in order to prevent the rib from moving away from the previous rib while gluing, not to force the new rib against the previous rib. This is easy to do, and if done just a little each time will magnify the mistake at each rib. I in fact made this mistake on the third rib, but corrected it with the straightening jig seen below.
The back with three ribs completed.

I let about a week pass before I worked on my oud again, and when I went back to it, I found that the third rib did not quite match the profile pattern. It was a bit flatter than it should about 1/2 way between the apex and the neck. About 5mm. I followed Hankey's advise about checking each rib against the profile pattern after each glue-up and correcting it immediately. To do so, I made a jig that is simply two blocks which are shaped for each side of the rib profile. I wet the misshaped rib and clamped it in the jig overnight. I was a bit skeptical that it would work, but sure enough, in the morning when I removed the clamps the rib was almost right at the proper shape. However, later in the day it had lost a bit of its shape. I will probably repeat this, but instead will glue the next rib to it as soon as possible to retain the shape. I anticipate having to make a few of these corrections as I near the full shape the of the back. I figure that by the time I am at the top two ribs, the shape will be just about right. I don't think that it is extremely important that the shape of the face of the oud be exactly the same as the profile pattern. I will be satisfied with a discrepancy of a few millimeters here or there. I see now why so many oud makers use a mold to make the back. There is no chance of going away from the shape. I still think Hankey's open form method may be superior, but requires more skill to execute.