Michael Bulatowicz
01-28-2020, 4:16 PM
I spent some time thinking about the topic of edge retention in chopping while in the midst of chopping out the waste in a couple of Roy Underhill style bench hooks I made in oak this past weekend.
I realize that there is a perception that the topic of steel as it relates to edge retention has been beaten to death here and elsewhere. I disagree, particularly as it relates to edge retention in chopping. Various people have gone back and forth regarding hardness, brittleness, carbides of various compositions, toughness, sharpening to the finest possible edge in order to improve edge retention, abrasion resistance, and so on.
It occurs to me that there are a few important points that I have not seen raised on this or any other woodworking forum (perhaps I’ve simply missed those posts/threads).
In spite of my tendency to get overly verbose, I’ll try to keep the introduction here comparatively brief and details comparatively sparse; for the curious, we can add details in the discussion.
Point one: hardness as measured by an indenter (such as the Rockwell C hardness test) is a measure of the tendency of a material to resist “cold flow” and is not a measure of the hardness as one would measure on a basis of, for example, ability to resist being scratched (such as Mohs hardness). The Rockwell C number that one ends up measuring arises from a complicated interaction of the steel grain size; the size, distribution, concentration, and composition of carbides; steel crystalline structure distribution (how much martensite, austenite, etc. is in the steel); the alloy content, distribution, and structure; impurities; carbon content; and so on. It is, therefore, related to but not descriptive of the microstructure of the steel. It is in no way a measure of the brittleness of the material, though as has been hashed out many times it does correlate to brittleness. As pointed out here and elsewhere, one cannot, therefore, simply compare two steels of different composition or even the same composition but different processing details on the basis of a hardness spec alone.
Point two: metal fatigue. Bend a paper clip back and forth enough times and it’ll snap. If you haven't tried it before, it's a cheap and easy demonstration of metal fatigue. At no point are you applying a large enough load to break the paper clip all at once--you're not shearing through the steel or pulling it apart, but it breaks nonetheless due to the cyclic loading. The same basic idea appliees to a chisel edge in chopping; whether you’re using a hammer, mallet, or anything else to apply an impulse to the chisel edge (by way of hitting the handle, hopefully), you’re applying a cyclic load to the very thin cutting edge, which will eventually start to fracture and fail. Brittle failure, on the other hand, will happen upon a single application of a load that exceeds the fracture stress of the material (itself a complicated topic, but let’s leave that alone in the introduction). So, if you hit the chisel with a consistent blow any true brittle failure should happen very quickly—upon the first or second strike, if you’re consistent enough with applying the same force each time. An edge that, in straight-grained wood clear of knots, doesn't immediately chip but instead develops chipping only after many strikes has either been impacting relatively large imbedded hard particles in the wood (much larger than the typical imbedded silica one finds with many woods) or is instead degrading due to metal fatigue (often as exacerbated by abrasion). Fatigue failure starts with a small crack (or cracks) that develop and grow with repeated loading until the material breaks; without knowledge of the load history of a material and/or having a microscope to examine the fractured areas, fatigue failure is often easy to mistake for brittle failure after the fact, but is in truth a fundamentally different mechanism.
Many factors affect fatigue life. For maximum fatigue life, the list would include:
1. Fine grain structure—requires powdered metal and/or careful control of the mechanical deformation and temperature profile during processing.
2. Fine, well-distributed carbides that are well-bonded to the bulk steel (poorly bonded carbides worsen fatigue life even though they still increase hardness and abrasion resistance). So, impact forging (hammer or drop forging are examples) to shatter the carbides and mechanical deformation to distribute these broken-down carbides. Or, powdered metal that starts with inherently small carbides in the small powder particle size. Iron carbide exhibits the best bond strength to iron of any carbide of which I am aware. Stronger than chromium carbide or vanadium carbide, for example.
3. Minimum impurities. During processing and heat treatment, impurities tend to gather at grain boundaries, weakening the grain-to-grain bonds. Impurities in general and sulfur (sulphur) in particular tend to make steels brittle and fragile by weakening the grain-to-grain bonds.
4. Properly oriented grain structure (think “long grain” along the chisel’s long axis, with the chisel’s bevel being like a miter on the end of a board—a rather poor metaphor, but applicable nonetheless). The advantages here should be obvious to any woodworker. This is achieved by mechanical deformation, for example by hammering or pressing the steel at temperatures too low for full recrystallization.
5. Hard steel—this includes hardening by mechanical means such as burnishing (for example, stropping on a plain strop). All else being equal, a harder steel will have a longer fatigue life under identical cyclic loading.
6. High carbon content. All else being equal, a steel with more carbon will have a longer fatigue life (as long as the carbon content remains low enough to actually still call it steel). Edges that have experienced decarburization will fail much faster, for example (as already pointed out by a number of people).
7. Very smooth steel—keeping the scratches as small as possible at the sharp edge will maximize the fatigue life.
8. Oriented scratches: side-to-side sharpening can be expected to result in an edge that fails earlier due to fatigue compared to forward-backward sharpening because of the direction of the applied cyclic loading.
These points serve to help explain the observations of a number of people regarding the edge retention of various steels in chopping. If fatigue failure is a dominant or at least major edge failure mechanism in chopping the following should exhibit maximum edge life, when done “right” for edge retention:
1. Impact-forged high carbon steel (including but not necessarily limited to hand-forged high-carbon steel)
2. Powdered metal compositions
Okay, so it appears I’ve once again failed to overcome my tendency towards verbosity. Or, have I? There’s lots more where that came from.
Questions? Disagreements? Insults for having dared to re-open such a topic? I welcome them all.
Best regards,
Michael Bulatowicz
I realize that there is a perception that the topic of steel as it relates to edge retention has been beaten to death here and elsewhere. I disagree, particularly as it relates to edge retention in chopping. Various people have gone back and forth regarding hardness, brittleness, carbides of various compositions, toughness, sharpening to the finest possible edge in order to improve edge retention, abrasion resistance, and so on.
It occurs to me that there are a few important points that I have not seen raised on this or any other woodworking forum (perhaps I’ve simply missed those posts/threads).
In spite of my tendency to get overly verbose, I’ll try to keep the introduction here comparatively brief and details comparatively sparse; for the curious, we can add details in the discussion.
Point one: hardness as measured by an indenter (such as the Rockwell C hardness test) is a measure of the tendency of a material to resist “cold flow” and is not a measure of the hardness as one would measure on a basis of, for example, ability to resist being scratched (such as Mohs hardness). The Rockwell C number that one ends up measuring arises from a complicated interaction of the steel grain size; the size, distribution, concentration, and composition of carbides; steel crystalline structure distribution (how much martensite, austenite, etc. is in the steel); the alloy content, distribution, and structure; impurities; carbon content; and so on. It is, therefore, related to but not descriptive of the microstructure of the steel. It is in no way a measure of the brittleness of the material, though as has been hashed out many times it does correlate to brittleness. As pointed out here and elsewhere, one cannot, therefore, simply compare two steels of different composition or even the same composition but different processing details on the basis of a hardness spec alone.
Point two: metal fatigue. Bend a paper clip back and forth enough times and it’ll snap. If you haven't tried it before, it's a cheap and easy demonstration of metal fatigue. At no point are you applying a large enough load to break the paper clip all at once--you're not shearing through the steel or pulling it apart, but it breaks nonetheless due to the cyclic loading. The same basic idea appliees to a chisel edge in chopping; whether you’re using a hammer, mallet, or anything else to apply an impulse to the chisel edge (by way of hitting the handle, hopefully), you’re applying a cyclic load to the very thin cutting edge, which will eventually start to fracture and fail. Brittle failure, on the other hand, will happen upon a single application of a load that exceeds the fracture stress of the material (itself a complicated topic, but let’s leave that alone in the introduction). So, if you hit the chisel with a consistent blow any true brittle failure should happen very quickly—upon the first or second strike, if you’re consistent enough with applying the same force each time. An edge that, in straight-grained wood clear of knots, doesn't immediately chip but instead develops chipping only after many strikes has either been impacting relatively large imbedded hard particles in the wood (much larger than the typical imbedded silica one finds with many woods) or is instead degrading due to metal fatigue (often as exacerbated by abrasion). Fatigue failure starts with a small crack (or cracks) that develop and grow with repeated loading until the material breaks; without knowledge of the load history of a material and/or having a microscope to examine the fractured areas, fatigue failure is often easy to mistake for brittle failure after the fact, but is in truth a fundamentally different mechanism.
Many factors affect fatigue life. For maximum fatigue life, the list would include:
1. Fine grain structure—requires powdered metal and/or careful control of the mechanical deformation and temperature profile during processing.
2. Fine, well-distributed carbides that are well-bonded to the bulk steel (poorly bonded carbides worsen fatigue life even though they still increase hardness and abrasion resistance). So, impact forging (hammer or drop forging are examples) to shatter the carbides and mechanical deformation to distribute these broken-down carbides. Or, powdered metal that starts with inherently small carbides in the small powder particle size. Iron carbide exhibits the best bond strength to iron of any carbide of which I am aware. Stronger than chromium carbide or vanadium carbide, for example.
3. Minimum impurities. During processing and heat treatment, impurities tend to gather at grain boundaries, weakening the grain-to-grain bonds. Impurities in general and sulfur (sulphur) in particular tend to make steels brittle and fragile by weakening the grain-to-grain bonds.
4. Properly oriented grain structure (think “long grain” along the chisel’s long axis, with the chisel’s bevel being like a miter on the end of a board—a rather poor metaphor, but applicable nonetheless). The advantages here should be obvious to any woodworker. This is achieved by mechanical deformation, for example by hammering or pressing the steel at temperatures too low for full recrystallization.
5. Hard steel—this includes hardening by mechanical means such as burnishing (for example, stropping on a plain strop). All else being equal, a harder steel will have a longer fatigue life under identical cyclic loading.
6. High carbon content. All else being equal, a steel with more carbon will have a longer fatigue life (as long as the carbon content remains low enough to actually still call it steel). Edges that have experienced decarburization will fail much faster, for example (as already pointed out by a number of people).
7. Very smooth steel—keeping the scratches as small as possible at the sharp edge will maximize the fatigue life.
8. Oriented scratches: side-to-side sharpening can be expected to result in an edge that fails earlier due to fatigue compared to forward-backward sharpening because of the direction of the applied cyclic loading.
These points serve to help explain the observations of a number of people regarding the edge retention of various steels in chopping. If fatigue failure is a dominant or at least major edge failure mechanism in chopping the following should exhibit maximum edge life, when done “right” for edge retention:
1. Impact-forged high carbon steel (including but not necessarily limited to hand-forged high-carbon steel)
2. Powdered metal compositions
Okay, so it appears I’ve once again failed to overcome my tendency towards verbosity. Or, have I? There’s lots more where that came from.
Questions? Disagreements? Insults for having dared to re-open such a topic? I welcome them all.
Best regards,
Michael Bulatowicz