The Big Plane Experiment
Brent Beach and Steve Elliot have done some very interesting experiments. I am thinking about ways to organize collected data so more comparisons can easily be made and data can be added by multiple people over time. Also with all the data organized together it is easier to look for interaction affects of more than one parameter. I do not remember much but I took a statistics course on experimental design in university. The important part is every possible combination of treatments in the domain need not be performed to get good information. Analysis of variance, etc.
Organizing Data
Domain
- blade manufacturer
- blade metal
- blade hardness
- blade thickness
- bevel angle
- back bevel angle
- back bevel width
- honing abrasive
- wood species
- angle to growth rings ( 0 = face, 90 = edge grain)
- angle to tree axis (-90 = end grain, 0 = long grain, 90 = end grain)
- cross grain angle (0 is in grain direction, 45 is like a scrub plane)
- plane screw angle
- length planed (0 ft is a freshly sharpened blade)
- plane manufacturer
- plane model
- bevel down or up
- bed angle
- shaving thickness
- chip breaker manufacturer
- chip breaker model
- chip breaker bevel angle
- chip breaker distance from cutting edge when sharp
- mouth opening (feeler gauge between mouth and blade)
- experimentor
- experiment date
To insure fair comparisons I think the bevel and possibly a back bevel should both be single, flat bevels.
Range
- a description of the resulting wood surface
- a picture of the wood surface
- subjective rating of the wood surface: tear out [0,10] (good measure when planing against the grain)
- subjective rating of the wood surface: chipped blade traces (good measure when planing with the grain)
- subjective rating of ease of honing
- subjective rating of ease of planing
- picture of shaving
- thread cutting test results
- initial sharpness results from the Catra blade tester ($91 CAN/sample for regular testing including initial sharpness and edge durability)
- the force required to push the plane through wood (how to measure? some set up like the Catra blade tester with a 3 dimension force measuring vise)
- a microscope picture of the cutting edge (>= 700x magnification)
- a profile of the edge made by the Catra blade profile microscope (190x magnification, can see radius greater than 20 microns, $15 CAN/sample) or microsection (not sure what this is but they told me the blade profile microscope would not see wear on slightly dulled plane blades) by the Catra folks (resolution <1 micron, $35 CAN/sample)
After all that I've read I would be most interested in the following test domain
- blade = {Hock high carbon, Hock A2, Finck A2, Veritas A2, Lie-Nielson A2, Clifton, Academy Saws M2} and later {Knight high carbon, Knight Japanese, other Japanese blades, other tool steels}
- bevel angle = {25, 27.5, 30, 32.5, 35 degrees}
- back bevel angle = 0 degrees
- honing abrasive = {0.5, 0.25 micron diamond paste on a cast iron plate; Shapton professional 8000x, 15000x, 30000x; 0.5, 0.3 micron 3M abrasive paper}
- wood species = {pine, hard maple, curly hard maple, cherry, oak}
- length planed = {0, 10, 50, 100, 200, 350, 500, 700, 1000 ft}
- grain angle = ? against the grain by varying degrees
After all this testing compare the blade rankings to the rankings produced by the Catra blade tester for edge durability starting with freshly sharpened blades.
Workshop Index
Feedback? Questions?
