How to study tool dulling

Re: How to study tool dulling

Warren in Lancaster, PA

>One of the big variables with testing different sharpening methods is the skill of the person using the method. Whether we are talking about oilstones, or waterstone, or diamonds, or sandpaper, or wheels or strops, a person who has used the method for years will often get better results. As Thom suggested a few days ago, a person who knows his steel will also get better results. It would not be surprising if a person testing got the best results using the method or abrasives or steel that they habitually use.

As just one example, a person who has used the same stone for years will get a feel for the optimum pressure to put on the steel while sharpening. This type of variable makes a difference in speed of sharpening, condition of the stone, formation of the burr, etc. It would be difficult to account for this feel in testing conditions.

a little free thought and some previous musings...

Dave Thompson

>I had imagined a system that measures the load required to push a blade through a uniform medium. It would be sort of like a Brinell/Rockwell/Janka test configured like a guilloteen. A load is applied to the back of a blade as it cuts through a 1cm x 1cm substance ( pick one HMWP, balsa, etc.) The measurement of dullness/sharpness comes from either how much weight is applied, or perhaps, how deep the penetration with a set weight.

Ofcourse, comparisons of sharpness must assume the same bevel angle, and now that I think about it... thickness of blade would need to be uniform between comparison study....which may limit it's usefulness.

Dave

Re: How to study tool dulling

thomd

>A metal lathe is designed to cut with what we normally call a scraping action: The top of the tool is set even with the centerline of the lathe spindle, and in a horizontal position. a tool could de designed that would simply change the angle to be a "cut", one would have to determine what kind of cut, and develop a feed mechanism, like a little plane so that the angle remained consistant as the workpiece diameter decreased.

Metal lathes do have a device for advancing the tool post, but it is not timed to follow the declining diameter of any workpiece. A routine could be devised for a CNC lathe.

A simpler machine might be a super surfacer. A mechanical plane. It does operate at higher speeds than a regular plane, but not terribly so.

At the end of the day it doesn't really mater what the tests tell us if as Warren says we can't seem to reproduce the results. I think a similar sort of thing is waxing skis, and different people will

Instron, perhaps

jim reed @ tallahassee

>My only lab experience measuring force is using an Instron. Here it could be used to measure penetration into a reference solid. In my opinion wood is too variable and some sort of plastic might be the best reference material. The first set of tests would be to measure initial penetrations to get an average. Then you do something to dull the blade. Then you measure terminal penetrations and average those results. These measurements are very precise and in order for the results to mean anything, someone must write a pretty detailed testing protocol. There is alot of work here because of the variables. I think that to develop the testing protocol and perform some repeatable experiments, this is a 2000 man hour project for a talented lab technician.

Perspective *Graphs*

woodburnbob

>Bill, I'm sure you know all this has been done in great volume in the first half of the twentieth century regarding tooling for cutting metal. For others reading through, any of the old used books on tool design you might find in the machine shop section of your favorite used book store or library will contain reams of this material.

I scanned 3 representative graphs from a book I had handy, simply to illustrate the sort of things you'll see in such books. It's from this sort of data that the all the feeds/speeds/depth of cut recommendations come...whether single point tools or multi-point tools like milling cutters (like router bits). Precisely, how the data were gathered, i.e., design of mechanisms, endpoints and so forth I don't see. The Machinery Handbook also has all this stuff.

What I haven't seen is analogous stuff for woodworking cutters. But I suspect it exists. Factory owners want to know how to set up machine to get the greatest output per horsepower/time and least tool wear. IIRC, the chip types in Leonard Lees book on sharpening are in fact drawn not from hand planes, but from jointer-planer-shaper cutting heads. That is, high speed rotating cutters run from motors. Important information, I presume, for sawmills, millwork houses, furniture companies and the like. But a hand plane or chisel aren't exactly cutting heads.

One of the most robust finding about tool life is that the more polished an edge the longer it will take to fail (dull). It seems a very solid law of metal tooling, but whether it has relevance for wood cutting by hand is as I say another matter.

The time/effort/cost of simply sharpening a plane iron or chisel by a craftsman or tool connoisseur is relatively insignificant compared with shutting down an assembly line, tearing apart a complex machine to get the cutters out, sending them to the tool sharpening shop, waiting and so forth.

Of course, our interest in "how sharp" and "how durable" is largely about wonder about the validity of marketing hype and about compulsive attention to comparative shopping, that and a burning desire just to KNOW, wouldn't you say?

Re: How to study tool dulling

Donald Pierson

>In 1998/99 I wrote some software to control a factory automation system that was part of Intel's production line. The line manufactured cards used in most of the laptop computers. The system included a robot that had an end-effector that measured forces and torques in all three directions. If the cutting tool of your machine/tester was mounted on such and end-effector software could read the forces/torques dynamically. As the tool dulls these forces/torques would increase. As I recall the company that made the end-effector was JR3 and their website is www.jr3.com

Re: How to study tool dulling

Chris Scholz

>Bill and Phil,

In addition to your proposed setup and of course Brent Beach's and Steve Elliott's work, another setup that I encountered is in the context of super surfacers. Difficulty there was to accurately determine the friction between wood and "sole". They used a custom jig with various strain gauges to solve that problem.

For the actual balde wear measurement researchers seem to use what I belive are surface profilometers.

Attached results show the measurement setup and measured values of a SV' parameter (that is related to blade tip wear as defined in the image). The measurements were taken using a super surfacer plaing oak with the following geometric parameters: cutting speed: 50m/min, blade protrusion below sole 0.06mm (0.0002in here in the US), mouth width 0.4mm (0.016in), cutting angle beta 30.8 degree, bedding angel alpha+beta 35 degree, distance capiron to cutting edge 0.25mm (0.001in), blade material: HSS. How relevant do you think this is for hand planes?

(Source: R. Fischer, Improvement of the surface quality of wood through linear planing action. Dresden Institute of Technology, Institute for Wood and Paper Technology, Final Report AiF-Project-Nr. 10811 B/1, pp. 79f, my own tranlations, no affiliation).

Chris

Re: How to study tool dulling

woodburnbob

>Chris, your graph of "sv" (presumably an index of edge radius) versus linear feet is quite analogous to Steve Elliot's data plotting a functional measure of sharpness versus linear feet.

Correlating image measures, edge radius measures and functional cutting measures is excellent "basic science", tending to validate all three...if in the future someone brings together these three sets of data.

Steve fit his data to an arbitrary best-fit spline curve rather than a straight line, I beleive. Fischer, for the bar chart you show, could have done the same.

If you do a simple linear regression on the two sets of data, I bet the slope of the line is statistically identical.

But my guess is that the best fit wouldn't be a straight line, nor necessarily an Excel spreadsheet spline algorithm, but rather some sort of polynomial or exponential function, such as: the first 10 feet deducts 10% from sharpness, the next 10 feet deducts 10% from what's left, the next 10 feet deducts 10% from what's left, so on and so forth...not unlike paying off a mortgage, or watching your fortune grow under compounding interest.

Then what would we argue about ?

Moses Yoder in White Pigeon, MI

>

I'm Keeping an Eye on You Moses. -)

Jim Shaver Oakville, Ont.

>

How bout the effects of BU/BD on dulling?

John in New Mexico

>