Sunday, July 22, 2012

Convex Double Bevel Sharpening

The sharpening setup I used, a variation of my portable sharpening station. I made the holder from shop scraps in half an hour; the oversize base provides a clamping edge for clamping front or back. From left to right, DMT extra-extra coarse diamond plate, Norton coarse, medium, and fine India stones, leather strop block, and Norton translucent Arkansas stone. I'm currently using Flexcut gold compoud on the strop.

For the past couple of years I've been using two somewhat different freehand sharpening methods, double bevel and convex bevel.  Both seek to sharpen an edge to the same 30 degrees. However, where double bevel explicitly tries to achieve it by very precise control, often using a sharpening guide rather than freehand, convex bevel is a much freer method, where the angle back behind the edge is irrelevant.

For some time I've been thinking it might be useful to combine the two, and the idea finally converged when I thought of them in terms of tolerances. Double bevel is a matter of tight tolerances, even on the 25 degree primary bevel. Convex bevel is much more a matter of loose tolerances back behind the edge.

I like to apply tolerances at their appropriate scale wherever applicable. Some things require tighter tolerances, and some things can tolerate looser ones. Looser tolerances generally require less work and are more forgiving, so when appropriate, they can actually be better. Save the tight tolerances for when they're actually needed.

Sharpening is a contentious topic. Some might see this as a worthwhile compromise, while others might view it as a horrifying abomination that satisfies no one, like Republocrats and Demublicans. Heresy or revelation, I find it combines the best attributes of both methods while remaining fast, simple, and effective. If you're just learning to sharpen or are having problems with your current method, give this a try and see if it helps.

Think of the sharp end of a chisel or plane iron as having two regions: the region behind the edge, referred to as the primary bevel when using the double bevel method, is at 25 degrees; the region at the edge, referred to as the secondary bevel, is at 30 degrees. The convex bevel method forms them  as one continuous curve, but still produces essentially the same two regions.

These regions have different requirements. The primary bevel doesn't need to be precisely angled. It just needs to be strong to support the edge. Because it's thicker, it can be ground at the narrower angle.

The edge would cut more readily at that narrower angle, but it would break down too easily. Because it's the vanishing line where two planes meet, it's thin. In order to provide usable durability, it has to be ground at the higher angle. Careless sharpening can easily ruin a good edge, so more precision is called for.

So let's combine the carefree convex bevel approach behind the edge with the controlled double bevel approach at the edge.

The video below demonstrates the method, first on an old Stanley socket chisel, presumably O1 steel, then on a wide IBC Pinnacle A2 steel plane iron. A2 has a reputation for being harder to sharpen, but other than needing a few more strokes on each stone, I didn't find it any more difficult with this method.

One thing I forgot to mention in the video, I don't put oil on the Arkansas stone because there's usually enough on the tool even after a quick wipe with the paper towel. The Arkansas stone doesn't take much before the tool is just skimming across on a film of oil instead of honing.

I start with a primary convex bevel coming up to 25 degrees at the edge; it probably falls off by 5 degrees as it curves back. You can certainly hold it to a closer tolerance if you feel the need to do so. Then I follow up with a much more careful flat secondary bevel.

I do heavy, rough shaping of the convex bevel on a DMT extra-extra coarse diamond plate. This removes material very fast, and in fact you need to pay close attention with chisels so that you don't take the end out of square. Once I've established the desired shape, I don't need to use the plate again. Thereafter I maintain the convex bevel on a Norton coarse Index stone, requiring just a few strokes each sharpening. I leave the primary bevel at this coarse surface.

For the secondary bevel, I start with coarse India stone, then progress through medium and fine India stones, Norton translucent Arkansas stone, and leather strop with compound. The coarse stone nips off the end of the 25 degree convex primary bevel quickly, leaving the flat 30 degree edge. The remaining stones and strop then polish that edge.

I maintain control of the secondary bevel by using simple pull strokes. Body geometry implies that pulling the tool back over a distance will alter the angle a bit, so the result is another convex bevel, not a true flat, but at a much finer scale. So this method actually forms a double convex bevel, primary bevel at loose tolerances, and secondary bevel at fine tolerances.

For both angles, I use simple angled wooden blocks cut to the desired angles as sight guides. I set them next to the stones, and angle the tool up or down until it appears to be parallel with the top of the guide. This is simple and provides adequate precision. As you practice with the setup, you can elminate the sight guides.

The sight blocks, 25 and 30 degrees. Easily made from shop scraps ripped at the desired angle.

Depending on the abrasive and the metal in the tool, you only need between 5 and 10 strokes on each stone. This quickly raises a burr. You can raise a larger burr with more strokes, but that just produces a larger flap of metal that will take more work to polish off. Might as well stop as soon as you've gone far enough.

Keeping to a minimum of strokes also makes it fast. The video draws the process out, but I was able to do the entire sequence on the chisel in 58 seconds without feeling rushed. So say you only need at most 90 seconds for any given tool. You can further save seconds during a couple of touchups by only going back as far as the fine India stone. That might mean you need one or two extra strokes on the coarse stone when reforming the convex primary bevel.

Note that this all assumes you've already ground flat and polished the flat back of the blade, as least the inch at the end. Without a flat back, all your work to form the flat plane of the bevel will be meeting up with a rough surface, so will not produce a good edge.

I used oilstones, but this works with any abrasive. The focus here is not on any particular abrasive, use whatever type you prefer, but on the technique.

The stone holder with a sanding block and a Norton 8000 grit water stone replacing the DMT extra-extra coarse plate and the Norton translucent Arkansas stone.

The final word I'd pass on is not to view this as wholly prescriptive. Look at it as a starting point and adjust as you see fit, varying the motion, the number of strokes, whether you add forward strokes to the pull strokes, etc. Again, focus on the technique, finding what works for you. Experimentation is good, even if it doesn't work out, because it shows the limits and the things to avoid.

(Continue to part 2)


  1. What do you do with your stone holder when you are not using it? I've been thinking of making one, but wonder if I need some kind of cover to keep dust off the stones when not using it.

  2. You demonstrate very thoroughly your method. Excellent video.

    Where do you feel that tight tolerances would be important in honing?

  3. Cliff, I'd like to find a dedicated spot in my shop to leave it rather than pull it out and clamp it down when I need it, since in reality I always need it. I may build a simple stand for it, like Jim Kingshott's tool stands. I would also build a cover to keep the sawdust from mixing with the oil.

    Thanks, Tico! I think the tight tolerances are important when honing the secondary bevel, the edge. You need to be consistent with the angle. Too shallow, and you're only honing behind the edge, not reaching it. Too high, and you're honing across the edge, actually taking it off.

    So where the primary bevel could vary by several degrees, here you want to hold it steady within a tenth of a degree. The very slight convex motion of the secondary bevel here means that you should just be coming to the edge without flipping over it.

  4. Regarding my last comment, the other interesting thing to note is that once the shape is established on the coarse stone, because the successive stones are finer, they remove smaller and smaller amounts of material, so they are less able to affect the shape.

    At this point I'm just speculating, I would love to see this at microscopic scale, but I would assume that these finer abrasives would therefore do progressively less damage if you happened to go over the edge. Meanwhile, by polishing the surface, they remove the microscopic projections that would increase friction during the cut.

  5. Great instruction! I am glad to see that I am not the only one using the "old fashioned" oil stones. I know that your video work is time consuming, but keep it up.

  6. Oh Boy !!! you have opened a can of worms now> :) the Paul Sellers haters will be out in force. This method makes an awful lot of sense to me. I must break down and try it. I now use tormek and then polish with oil stones. Very much enjoy your blog.

  7. Great post on sharpening, I hope that you do more videos in the future. I know they help me out in renforcing the text part of a lesson or point that someone is trying to get across. Keep up the great work!

  8. Forget micro bevels and guides and get on to real woodworking. They only needed to sharpen and hone a convex bevel. Japanese sword polishers have been putting convex double bevels on knives. thanks