Convex edges and the Wicked Edge

[Gaston444]

Good post, and while I have no 'side', (I just posted the diagram for reference), I gotta ask... from your summary, wouldn't the opposite also be true? Could you not say, "At the same final edge angles, the convex is sharper (since geometry behind the edge also matters), and at the same shoulder, the convex is stronger"? If no, why not? Seems like all your saying is: thinner is sharper but weaker, thicker is stronger but less sharp.

Yes you could say that, and that is why many convex edged factory knives are made with very thin edges, and as a result they really do cut better than most factory V-edges...: Bark River and some Blackjacks comes to mind. (The Cold Steel San Mai III Trailmaster I had was much too thickly "swelled" to have any cutting advantage. The blade not being particularly wide worsened this poor geometry: I wonder if the cheaper true V-Edge Trailmasters are not sharper)

So then, why not a thinner convex? The main issue is that, for the same edge thickness at the point of measurement, which is critical to lateral strength, you can get better sharpness out of a V-edge, and this for several reasons:

-Absolute lateral edge strength is not dependant on edge angle: It is false to say that a 10 degree per side edge angle is inherently weaker than a 15 or 20 degree per side edge angle, even on the same knife... It is therefore a mistake to say that you want to compare at equal edge angles.

In fact I have observed the exact opposite: Lower edge angles are more resistant to damage while chopping in Maple: Edges that produced micro-bends at 15 degrees per side suddenly appeared "tougher" when sharpened thinner to 12 or even 10 degrees per side...

The reason for that is that the deceleration in the wood is more gradual at more closed angles, which then makes this deceleration more directionally "stable", and more stability means less yawing when the blade sinks into the wood: Yawing and abrupt deceleration is what disrupts the edge apex cohesion.

There is also another reason why "lower edge angles are stronger": Because the edge bites more deeply at a lower edge angles, this means the deceleration is spread over more depth of the blade: This deeper depth of bite means there is more wood "trapping" the steel in a more secure and forceful "grip": Thus the very wood that is cut is also acting to "preserve" the metal that is cutting it...: This probably accounts for a lot of this counter-intuitive lower edge angle "strength effect".

It could also be that the sharp shoulder of a V-edge, occurring earlier than on a Convex edge of similar edge angle, enhances this protective "wood-pinching" effect.

That edge apex needs protection, since at lower edge angles (below 15 per side) I can create a micro-folds in the edge apex by rubbing hard with my nail... Yet, amazingly enough, this weak-seeming metal (if of good quality: Not chipping or curling from hitting wood) will suffer less edge apex damage in wood than if you open up the angle... This even when poor/missed hits created large laterally displaced chunks from the blade glancing and leaning... The wood trapped around the edge retained its protective effect even when hits were not perfectly square, and when the blade glanced...

So then to compare at the same edge angle is a mistake: You want the lowest possible edge angle that will still give a given edge thickness at a given point from the apex, not the lowest edge thickness that will give a similar edge angle: The edge thickness is the basic value you start from.

The optimal edge shoulder thickness is to me a pretty immoveable value: For big choppers it is 0.020", or 0.5mm, no more than around 2 or maybe 3 mm from the edge apex: Anything above that in lateral thickness is going to get increasingly hard to sharpen to a low edge angle, the breaking point being around 0.030", and even 0.030" will require taller and taller edge bevels as the knife wears, meaning a greater and greater surface to abrade (unless on a tall hollow grind, which for some height remains constant in thickness, a huge durability advantage over flat grinds after decades of use, which is why Randalls last so long)... Anything below 0,020" is more suitable for a folder, and will tend to chip out or bend too easily while chopping...

For a 0.020" shoulder thickness at a given distance from the edge apex, the sharpest edge will always be a V-edge.

Then there is the natural tendency of sharpening to be too "rounded": Rounded is out of ideal, since the flatter the sides of the V is, the sharper and more aggressive the actual apex will be, and the more protected it will be by sinking deeper into the wood...

The edge thickness comes first, the edge angle second, and that is exactly the mistake convex edge proponents make: They reverse that order. They sometimes bring in convexed axe heads as an argument, but axe heads have so much momentum they work also by laterally displacing the wood, so there is no comparison to how a knife blade works.

Similarly, on Japanese sword the convex edge was related to their manufacture, and on 1970's custom knives it was largely related to cutting free-hanging rope...

Worse of all, convex edges encourage a rocking motion while sharpening, which is a very bad sharpening habit: You want to remove any rocking motion, not encourage it! Rocking motion is not something you can humanely do in a consistent way: This is why there is this emphasis on "soft" sharpening mediums and high levels of polishing: Unfortunately edges whose cutting ability depend on their polish will also lose that polish very fast, while rougher edge surfaces that are flatter will cut longer...

Keeping things straight and flat is within human perception (contrary to what most claim), if the strokes are kept as longitudinal as possible: A very shallow diagonal motion works best. Never stroke at 90 degrees to the edge, except for very minor finishing touches, as you cannot control the rocking over repeated motions.

Gaston

 

[HeavyHanded]

Personally I have noticed no difference in lateral stability or resistance to mineral damage based on grind type.

A Scandi at 24 has about the same lateral stability as a convex at 24 as a V bevel at 24. The only difference between them is how far up the blade any resulting damage might go, and this is all about primary grind thickness, also not dependent on grind type.

The only real difference is how the wedging forces are distributed, and convex spreads them out over the largest surface area. Personally I have zero doubt the convex profile drives an edge deeper with same terminal apex and same thickness behind the edge at 2-3mm than any other grind type. Once it begins to cut, the contact surface separates the wood with the least wedging - there is no shoulder to act as a pinch point. Also, lacking a shoulder there is no tendency to steer to one side of the bevel.

Actual apex angle is going to be decided by the bulk of intended work, same as with a felling axe tuned for pine or hardwood. Once you start talking about a wide variety of materials to be cut, I'd imagine different grinds will outperform others, nothing is 100% across the board. For ordinary cutting chores I don't believe most folk will see any advantage to either grind type.


For sharpening, convex on a hard surface is also about the easiest grind type to work on. Grind a bit behind the edge and then advance toward the apex. A new cutting bevel can be ground with just a few passes across the apex. This can be as precise as one cares to make it, or one can be a bit lazy - it will still fall in a fairly narrow range and certainly not inherently rounded at the edge any more than any other grind. One maintains the same overall geometry compared to V bevels that are generally neglected across their primary grind until performance drops off.

 

[Gaston444]

Personally I have zero doubt the convex profile drives an edge deeper with same terminal apex and same thickness behind the edge at 2-3mm than any other grind type.

You're off to a bad start when you describe something that is geometrically absolutely impossible...

With the same terminal apex angle, the thickness of a convex edge cannot be the same at the same 2 or 3 mm from the edge. It has to be thinner.

If it is the same edge thickness at 2 or 3 mm from the apex, and has the same terminal apex angle, then it has to have a straight line in there somewhere, which means it is no longer a true convex, but a combination of convex and V-edge...: Mostly a V-Edge with slightly rounded shoulders, and even then, the terminal apex angle would have to be greater, since the straight portion by definition starts lower because of the rounded shoulders...

I don't know how many times this basic geometric notion has to be said...: At the same shoulder thickness and distance from the apex, a convex has to have a more open terminal apex angle... This is exactly why I am against them...

A convex edge IS a duller edge angle at the apex: There is no way around it... What will it take for people to understand this I don't know... This is the very basic geometric notion all convex edge proponents have trouble coming to grasp with...

I'll leave aside concave ground edges. What you are saying amounts to saying that a bulged curved line is straighter than a straight line...

Sorry, but nothing is straighter than a straight line, which is why sharpening should aim for the flattest edge sides as possible...

Gaston

P.S. As an aside for the edge fragility issue: I test for micro-folds by running my nail away from the edge, where an invisible sliver of bent edge will "grab" whitish nail material and show the micro-fold that would otherwise be completely invisible.

In my experience, most people do not rigorously check for this damage in this way, which is why they assume the edge has no micro-fold damage. I've noted 12 degree per side has less tendency to do micro-folds while chopping (than 20 degrees) in this way.

As an interesting side note, I've noted that my Randall Model 12 showed no detectable micro-folds in hundreds of full-force chops in Maple at barely 10 degrees per side, yet this same knife developed micro-folds in just a few batoning hits, splitting four 2.5" diameter, perfectly straight, Maple branches.

So in 15-20 batoning hits, the Model 12's edge had more observable damage than in hundreds of chops, this in the same wood from the same tree...

It could be the hollow ground geometry is at play here, but my preliminary recommendation is... Avoid batoning like the plague...

G.

 

[HeavyHanded]

You're off to a bad start when you describe something that is geometrically absolutely impossible...

With the same terminal apex angle, the thickness of a convex edge cannot be the same at the same 2 or 3 mm from the edge. It has to be thinner.

Right you are!
I was thinking of my machetes when writing, but in retrospect am describing something more in line with 7mm back from the edge - and yes, the convex region would be above the V bevel at that same distance even though edge angle is the same. At the same distance the V bevel would be into the primary, convex still arching back.

A convex edge IS a duller edge angle at the apex: There is no way around it... What will it take for people to understand this I don't know... This is the very basic geometric notion all convex edge proponents have trouble coming to grasp with...

I do not agree at all with this. As mentioned above, I shape all my convex on a hard surface and as such the edge does not round down into oblivion anymore than any other freehand edge. I shape them on my wet wheel as well. You may claim that makes them a V bevel with rounded shoulders but to me that still qualifies as a convex.

P.S. As an aside for the edge fragility issue: I test for micro-folds by running my nail away from the edge, where an invisible sliver of bent edge will "grab" whitish nail material and show the micro-fold that would otherwise be completely invisible.

In my experience, most people do not rigorously check for this damage in this way, which is why they assume the edge has no micro-fold damage. I've noted 12 degree per side has less tendency to do micro-folds while chopping (than 20 degrees) in this way.

As an interesting side note, I've noted that my Randall Model 12 showed no detectable micro-folds in hundreds of full-force chops in Maple at barely 10 degrees per side, yet this same knife developed micro-folds in just a few batoning hits, splitting four 2.5" diameter, perfectly straight, Maple branches.

So in 15-20 batoning hits, the Model 12's edge had more observable damage than in hundreds of chops, this in the same wood from the same tree...

It could be the hollow ground geometry is at play here, but my preliminary recommendation is... Avoid batoning like the plague...

G.

If I'm inspecting my edge for damage it is the same as inspecting for burrs or incomplete grinds - very little is going to escape. In all honesty I seldom find edge chipping or rolling of any note on my machete, axe or hatchet edges. They get set at 30 inclusive and maybe drift over time between 32 and 28, but overall I just don't have problems with them - primarily seasoned American Beech and pine if camping, ash and maple around the house.

Though I agree, batoning is something to do in a SHTF situation and should be avoided if you don't have a spare cutting tool - I've done it on occasion with machetes but only a few times with my belt knife and just to experiment. The amount of pressure back into the edge is several times greater than the hardest manual chop, and I don't like breaking my stuff for no good reason.


Back to the OP, I'd still think having the clamp open slightly and pivoting the blade in the clamp as you go will give the best results. You could also lay it out in three facets for a more precise outcome but more time to execute as well - grind the edge, grind off the shoulder, grind off the transition between the two - but takes advantage of the precision afforded by the equipment.