Convexing on a belt grinder?

[kennyj]

Something I've been pondering for a while. Convex edges seem to be quite popular nowadays. I keep seeing lots of noise about how to create/maintain a part convex edge, most of them very cheap (and with a coincidental disadvantage in terms of accuracy, ease of use, etc.) Everything I've read about convexing in detail goes over the fact that to do it "right" you need a belt grinder, but this is quickly brushed aside as being an overly-expensive endeavor.

And yet, belt grinders can be had for as little as $100 (not counting any new belts that might be necessary,) while there are numerous sharpening systems that are more expensive (though some of them inarguably worth the money) and many of us have individual knives worth many times more.

Now, I'm not saying that people should start buying belt grinders just for knife sharpening (it's overkill, and you can't carry one with you on a camping trip) but for those of us that do have occasional need for a grinder of some sort, or occasional access to an actual belt grinder, it only seems logical that this avenue be explored.

Is there any information on how to best set up and use a belt grinder to profile and sharpen a blade for the mere mortals among us?

 

[Cliff Stamp]

Everything I've read about convexing in detail goes over the fact that to do it "right" you need a belt grinder ...

A lot of people propogate this misinformation, tapered grinds are very old and existed long before belt sanders, you can do it with a file readily.

I'm not saying that people should start buying belt grinders just for knife sharpening ...

For sharpening no, however if you are interested in reprofiling they are very nice to have. Even the cheap one inch ones are more than enough to adjust edge angles almost instantly on small knives and even just minutes on large chopping blades. An angle grinder is nice to do similar for axes.

Is there any information on how to best set up and use a belt grinder to profile and sharpen a blade for the mere mortals among us?

If you press a blade into the slack belt region above the platen it will give and thus produce a convex edge. How much curvature is induced depend on how hard you press and the angle you hold relative to the belt. I just use them for shaping so generally only have x-coarse belts. Cool the blade often to avoid overheating, working barehanded helps to avoid overheating the metal, however the very edge can be burnt long before the body of the blade becomes uncomfortable to touch.

-Cliff

 

[kennyj]

Would it be safe to utilize some sort of cold water bath to facilitate faster / easier cooling? I'm thinking: stroke, dip, dry, stroke, dip, dry.

 

[Cliff Stamp]

If the blade is fairly large and especially if the edge is thick and you are not grinding too acute, the massive amount of metal is often enough to serve as a heat sink. On smaller blades and especially when the edges get very thin and acute you often have to cool on every pass to prevent them from overheating. When you go to extremes (edges under 0.010" thick and angles of <10 degrees) you also have to press really light or you will burn the edge immediately even on one pass. Note the type of steel matters of course, common tempers are 300-1000F. It is pretty hard to burn a HSS for example but trivial to do so on full hard 1095.

-Cliff

 

[cbwx34]

Cliff,

Do you have any documentation to back this up? I've read other posts from knife makers that differs from what you've posted, and would like to see if any actual testing has been done that shows who is correct.

cbw

 

[Cliff Stamp]

Which points.

-Cliff

 

[thombrogan]

Can you document that 1095 tempers at a lower temperature than M2? Can you document that heat dissipates faster in larger masses than in smaller masses?

"Metal expands from heat? Who's talking mumbo-jumbo now, Lisa?" ~Bart Simpson

 

[cbwx34]

How about any of it? Let's start with this... "When you go to extremes (edges under 0.010" thick and angles of <10 degrees) you also have to press really light or you will burn the edge immediately even on one pass". That seems pretty specific....

 

[thombrogan]

That particular one is a statement you can easily disprove or verify. A superficial study of thermodynamics would make the statement able to be taken for granted without the bringing in any modifying conditions (such as the presence of large amounts of tungsten and molybdenum in M2).

 

[thombrogan]

Curtis,

Are you saying you know of no cases in which exerting more pressure against a moving object causes more friction and that you also don't know of any cases of increased levels of friction causing increased levels of heat?

 

[kennyj]

Can you document that heat dissipates faster in larger masses than in smaller masses?

http://en.wikipedia.org/wiki/Heat_conduction covers the basics of how heat moves through solid objects. High school stuff, the main points are nothing we haven't seen before (at least, I hope.)

Also, take note of how your computer's CPU is cooled. The heatsink is essentially a chunk of aluminum with fins to increase surface area. The surface area is the most important part of the equation, but as you can easily observe, it's a pretty chunky little thing. It could easily be much, much smaller if surface area were the only consideration.

All other things being equal, a larger piece of metal means that a given amount of heat has more room to move into. There's more surface area for air cooling as well. Because the heat being generated has that much more material to heat up before the heated surface reaches a given temperature, as said heat is constantly migrating to wherever it can, it can withstand heat exposure of increased intensity or time without the exposed surface becoming any hotter.

The degree to which this improvement is made depends on quite a few things and is subtle, but it's easy to observe in extreme cases. Turn your oven on at its lowest point (usually about 150 F, the same as a warm beverage) and take two pots or pans, one large with a thick bottom, the other small with a thin bottom. Place both in the warm oven for the a short period of time, say one minute, remove both simultaneously and see which one is hotter.

Of course, this works in reverse as well - more material means more heat retention - as this experiment will also demonstrate. It'll take longer to heat a thicker chunk of metal, but it'll also take longer to cool. However, when working with smaller amounts of heat, it becomes evident that larger pieces will radiate heat more efficiently due to their increased surface area, and when combined with the increased time taken to raise temperature due to increased mass, a larger peice can be considered easier to keep cool than a smaller piece.

When you're talking about tasks which take short periods of time, this becomes quite relevant. The thicker, chunkier piece of metal will absorb more heat before it gets too hot to work.

Of course, this entire discussion is running under the assumption that all other things besides mass are equal. This is not the case in the real world. Many other things will come into play, like the thermal conductivity of the material in question, its shape, where heat is being applied, how the material is being cooled, etc.

I believe Cliff's statements regarding the varieties of steel and edge thicknesses were intended more as hypothetical examples than hard evidence.

 

[Cliff Stamp]

I believe Cliff's statements regarding the varieties of steel and edge thicknesses were intended more as hypothetical examples than hard evidence.

Generally I don't get that specific on a hypothetical. About the lowest you can functionally go on the slack belt one inch sanders if you press very lightly is a bevel which runs from about 8 at the shoulder to 12 at the very edge and it takes a *really* light touch to do this and you have to lay the whole knife on the belt (the exact angles produced will depends on the width of the blade and the nature of the existing grind). My Battle Mistress has this grind as evidenced by the big strip of metal removed from the spine, and it took a long time to apply because the pressure was so light. You are looking on the order of an hour to reprofile at that angle due to the very low pressure. Going from 20 to 15 is pretty much instant, going from 15 to ten takes forever in comparison. I have done it *many* times, hundreds to date as I reprofile all of my personal knives and all the knives I sharpen for freinds. The lowest I have done on a slack sander was about five degrees per side. It took me many sessions of about half an hour on a D2 blade from Mel Sorg, because you barely have to touch the blade or the curvature induced will steeped and thus so little metal gets abraded. You can actually do it faster by hand which I figured out later.

I also worked with the platen as it allows more acute bevels at higher forces so I could grind off the convexity in the edge because it was lowering cutting performance. I didn't want that curvature in the edge, I wanted the entire bevel at the shoulder angle. As I have noted elsewhere, the main reason people see improved performance when they reprofile edges by applying convex bevels is that they added a relief, not that the bevel went from flat to convex. Anyway, platen grinding generates heat so much faster that on edges of the class as noted in the above I found it very difficult to even do one pass without them discoloring (unless they are high temper steels of course). Heat transfer through a plate is inversely proportional to the thickness, one of the fundamental laws of thermodynamics and the heat necessary to bring a volume of material up to temperature is proportional to the volume (another basic fundamental thermodynamical principle). Thus very thin edges will get overheated in very little time because it takes very little heat to do so and it will move very rapidly from one side of the edge to the other, *far* faster than it will move back through the blade towards the spine.

You can also convex grind on a platen by rocking the blade, it is way harder than doing it on the slack region, but it is possible. Fikes demonstrates the back principles in his video where he rocks a blade while honing it with a large benchstone. Mel Sorg was the first maker I saw advocate sharpening convex ground blades (spine to edge was one flowing bevel) on flat stones. Alvin Johnston had been describing how to grind convex edges on flat stones for much longer. Recently he has ground convex knives (primary grind) with such tools as an angle grinder and wheel.

-Cliff

 

[cbwx34]

Are you saying you know of no cases in which exerting more pressure against a moving object causes more friction and that you also don't know of any cases of increased levels of friction causing increased levels of heat?

No. My question was whether or not anyone has actually done any testing as to what point an edge will actually 'burn' and affect the temper, when sharpening on a grinder. I've seen this subject bounced around before.... I thought maybe Cliff had done some actual testing in this area... or maybe knew of some... since there seemed to be some specific info in the post.

cbw

 

[Cliff Stamp]

Edges will "burn" when the tempering temperature has been exceeded. This varies from about 300F to 1000F for most cutlery steels. You can also effect the temper by holding them at elevated, but lower temperatures, but it takes a much longer time than you would spend grinding. How fast this temperature sets in depends on the cross section of steel mainly and how hard you are pushing, as well as the condition of the abrasive, worn belts overheat very fast compared to new ones. It is very difficult to burn very heavy khukuris for example or large bowies on a new belt but trivial to overheat a small paring knife on a worn one. On large knifes, especially machetes, you can often completely regrind the edge and it never gets beyond hot to touch. I have actually made blades much hotter just cutting things. Which brings up the point raised on rec.knives years ago about can you effect the temper of the very edge on some knives if you cut fast enough. I have got them that hot that they will burn you readily if you touch them just by cutting cardboard rapidly.

-Cliff

 

[cbwx34]

Thanks. Still doesn't answer my original question of whether or not anyone's ever done any testing. I'll assume at this point, the answer is no. I'm not implying that you don't have a pretty good idea of when a knife edge is burned... I just thought that with all the testing you appear to have done, maybe at some point you or someone else had pointed a temperature sensor at a knife being sharpened, and went... hey... that thing is XXX deg. hot. (Maybe they're not that accurate... I don't know). That's all. No implication that your original post wasn't true... no implication that friction doesn't cause heat, etc. Just thought along the way you might have done or read something along those lines.

cbw

 

[Cliff Stamp]

I have burned edges, intentionally and otherwise as noted. They will change color as you overheat them, from straw to full black in severe cases.

-Cliff