Abrasion Rate vs “Grit”

It is a common assumption that coarse hones (lower grit rating, larger grit dimension) remove metal faster than fine hones.  However….

In the Swarf! article, it was observed that coarser hones remove larger pieces of metal.   At the same time, we should expect that a finer hone will contain a higher density of abrasive particles.  The overall rate of metal removal will depend on both the size and number of metal swarf particles removed over a given distance traversed by the blade on the surface of the hone.

In this conceptual drawing, the red circles represent 5 micron particles and the blue circles represent 10 micron particles. There can be four times as many 5 micron particles as 10 micron particles in the same area.

Obviously there are other factors to consider, the pressure applied, the shape of the grit particles (and how they wear with use) and the mechanical properties of the steel, to name a few.  The obvious approach is to simply measure the rate of metal removal on a variety of hones to see what happens…

In this experiment, a small piece of hardened carbon steel was used as a workpiece to compare the abrasion rate for a variety of hones.  The sample was cut from the end of one of the vintage straight razors used in other experiments for this blog.  This is not intended to be a systematic study with precise control of the pressure and stroke, but rather to simply to determine ballpark numbers for the metal removal rate for a few hones over a range of rated grits.


It should be expected that the rate of material removal will increase with the downward pressure applied.  In this experiment, the workpiece was “scrubbed” back-and-forth by hand with approximately 500g equivalent force applied.   The force and surface area were chosen to be in the typical range used for honing a straight razor.  The angle of the workpiece was varied relative to the direction of motion every ten strokes to minimize abrading parallel to existing scratches.  At varying intervals, the workpiece was weighed with a 1 milligram precision scale to determine the change in mass due to metal removal.

The pressure was chosen to be typical of that used with straight razors, however it is certainly less than would be applied to remove substantial amounts of steel when free-hand sharpening a very dull knife.  Since pressure is inversely proportional to surface area, it should be expected that a similar pressure to that used here will be achieved when flattening the back of a broad chisel or plane blade.


Each of the waterstones (Sigma, Shapton, Chosera, King) were lapped with an Atoma 400 (or 140 in the case of the 240 and 320) diamond plate to produce a fresh surface, and each stone was lapped again after every 500 strokes of the workpiece.  All of the diamond plates have all been used sufficiently to be broken-in, but not nearly worn-out.

The results for the Shapton Glass 16k hone are shown in the plot below. The average rate of removal was determined by fitting the data to a straight line.  Steel is removed at a rate of 2.7 micro-grams per stroke, or 2.7 milligrams per thousand strokes.  Based on the dimensions of the workpiece, each milligram decrease in mass corresponds to 2 microns of steel abraded.  One micron of steel is removed every 200 strokes.


The measurements were performed on a variety of hones and grits, and the results are tabulated below.  The rate is given in micrograms per stroke (1 microgram corresponds to 2 nanometers) and in microns/ 100 strokes.


For the waterstones, there is essentially no measurable difference (within experimental uncertainty) for the abrasion rate of the Shapton 320, Chosera 1k,and the Shapton 2k and 4k hones.  These four hones removed metal at the same rate.  The 8k and 16k Shaptons removed metal at about 1/3 the rate of the coarser hones.  The removal ratio between the King 1k and 6k hones is much higher than observed with the various grits of Shapton hones, however the King stones are rated for their actual grit size rather than their “effective” grit size.

Only two of the waterstones tested “became muddy” during use, the Sigma 240 and the the King 1k.  These two hones were also among the fastest metal removers.

The results shown in the Swarf! article demonstrated that the dimensions of the removed metal particles do increase with the rated grit of the hone.   However, the rate of metal removal will depend on both the size AND quantity of the metal debris.  The results here corroborate the idea that the finer hones remove smaller particles, but that they can remove proportionally more of those particles, for an overall similar removal rate.

The Atoma diamond plates produced the same removal rate with 140 and 1200 grit with the 400 grit performing at half that speed.  Having fewer contact points than the DMT plates produces a higher pressure at each diamond for a given force.

The four DMT plates were the same ones used in the Diamond Plate Progression where we observed the seemingly anomalous inverse relationship between grit and keenness.  The measured removal rate displays the same inverse relationship with the “finest” hones being fastest and the “coarsest” hones being slowest.  Astonishingly, the DMT Coarse (325) removed steel at the same rate as the Shapton 16k glass stone.  Among the fastest hones were the DMT EEF (8k) and DMT EF (1200), and this is almost certainly due to the presence of oversized diamonds in those hones.  The slow removal rate for the Coarse and Fine plates indicates that there is not sufficient pressure for the full diamond particles to abrade the surface, rather that abrasion is dominated by small asperities on the surface of those larger diamonds.

The experiment was repeated for a few of the hones, but with an estimated 5 times the force applied.  At approximately 2.5 kg force equivalent, this is the limit of what I could apply with two fingers and still keep the workpiece moving  smoothly across the hones.


At this higher pressure, we begin to see the coarse hones outpace the finer hones.  Even still, there is little advantage in the 320 and 1k grit stones over the 4k, when the greater apex damage caused by those coarse hones is considered.

There is a simple explanation for these results: coarse stones are only faster if sufficient pressure is used.   In the study and literature on the topic of grinding and wear (tribology) the terms “sliding, ploughing and cutting” are commonly used.  There is a critical pressure required for an abrasive to cut into the workpiece, otherwise it will simply slide without removing metal.  Efficient grinding occurs when the particles are cutting.   If the exposed abrasive becomes worn, the contact area increases and the effective pressure decreases (for a given applied force) and we will move from cutting to sliding.   In performing these measurements, I did observe an increased abrasion rate on the coarsest hones immediately after lapping, but this was typically short lived (less than 50 laps) and had minimal influence on the measurements.

As usual, we should be cautious in generalizing these results.   It is not my intention to “rate” various brands and types of hones.  However, this is not a difficult experiment for others to perform themselves.

30 responses to “Abrasion Rate vs “Grit””

  1. Thank you again. I hope you didn’t wear out your fingers rubbing that little piece of steel on those rocks. Weighing the differential is a delicate procedure that is waaaay beyond most of us.

    This information gives more support to your article on “too big of a jump”. Progressing from say a 1k stone to an 8000 DMT might make quick grinding, although the damage from rogue diamonds on the 8000 plate might make this a poor choice. However, progressing from a 1k bevel setter to an 8k finisher, like your Shapton, without any intermediary stone in the progression seems like it would take longer to achieve a finished edge than using the 4k in the progression.

    I have found that using a slurry on my synthetic stones makes honing much quicker. I have been experimenting with a 4000 Gouken Hayabusa and 8000 Gouken Fuji, mid priced, matched hones. I rub the two stones together to make a slurry on the 4000 which cuts quickly and the slurry keeps the stone surfaces clean. I progress to the 8000 hone with slurry which I slowly thin out to a clean surface (no slurry) for the last few laps, and edge trailing strokes. I am pleased with both the performance of the hones and the results on my bevel and edge. I think the slurry makes a slightly micro-convex edge. The burr is minimal, and requires only a few strokes on the loaded denim strop for removal.

    If the edge is chipped, I find that 800 SiC (corundum) paper, laid wet on glass, removes metal quicker than anything I have — cheap, flat and fast.


      • Thanks for the correction. You did not go on to say that SiC, or silicon carbide, is also known as carborundum which was spell checked to corundum — a different compound as you pointed out. Aluminum oxide “sandpaper” is often made to look black, like carborundum “sandpaper”, but the performance is different on steel. I much prefer carborundum paper for sharpening purposes because the grit is very sharp and it seems to be friable, breaking down to a finer grit during use. I like it for sharpening both razors and knives. I had a ¼” glass plate cut to 4 ½” x 11″, with polished edges, so that I can use half a sheet of 8 ½ x11 inch paper. I also use the paper dry with masking tape over the ends to hold the paper on the glass. A rubber eraser cleans out the swarf.


  2. Thank You for this valuable report!! When they say one have to get to know his stones…
    I have quite a little bunch on stones and i can say that there are finer grit ones that cut faster than the rougher; Now you have given sound numbers to this kind of observations!!!


  3. Very interesting conclusions. So if the goal was quickest metal removal to take out damage on larger blades what would the best choice be? Am I gathering that a coarse diamond stone would be a poor choice since you need to use high force which greatly decreases the life?


    • I have no use for coarse diamond stones. However, the DMT EF (1200) is probably my most used hone for fast removal in low-pressure situations; straight razors, chisels and plane blades. For water stones, I wouldn’t go below 1k.

      The simplest (and most common) way to increase pressure is to decrease the surface area by convexing the bevel. This allows you to access the higher removal rate of the coarser stones. This occurs naturally when you freehand sharpen.


  4. I would love to see some pictures and tests done with a Washita stone since they cut so differently then other stones. More like a cheese grater than an abrasive particle although I believe that is also part of the process. It has been explained in old literature that a Washita’s cutting power comes from the craters or depressions with particles tightly packed around their rims. Since they didn’t have the technology that we do now, back in the 1800’s, it would be interesting to see how close they were. In experience it seems to play out the way they were thinking.

    Liked by 1 person

    • I would speculate that loose grains of silica play an important role in the performance of these, but I don’t know how to design a suitable experiment to demonstrate this. I have a couple of images from an old Washita pocket-knife stone.



      Liked by 1 person

  5. Thank you for the images. To someone like me that is amazed by and uses these stones it is very interesting. From the images alone I would say that the loose grains do play a role and also that the old literature was surprisingly accurate in that there are grains packed around the cavities. They are very surprising stones. Thanks again.


  6. Your research is immensely helpful and engrossing. Ive learned a lot for sure. Any chance of you doing something on the black arkansas stones? And perhaps also shallow/deep scratches and whether various hones may work better freshly lapped or in a glazed condition? Especially regarding the difference in a razor edge honed on a burnished surgical black arkansas and another honed on a matte(freshly lapped) arkansas? Preferably the same stone?
    Dunno why there is so much mote info o , say, coticules than the arkansas stones. Why shouldnt every honer be an expert on the locally available natural stones? I think it is a serious deficiency to ignore them. I hone on anything i can afford. Cant afford a coticule yet. But i want experience with every stone I can get just in case the future sees a shortage of any particular foreign sourced hone.


  7. I am just piling on to thank you, Todd, for all the illuminating and incredibly useful results you’ve shared here. What a treasure trove of result sets you have produced. Your refreshingly objective presentation of each topic area where you let the picture walk the talk is a most admirable approach to an emotional and subjective and too often contentious topic.

    I cannot help notice the slow-down in articles production during the last 12-18 months. I do hope that all is well with you and yours. I cannot imagine you’ve run out of article topics to expose. Then again this kind of thing can be exhausting!
    Anyway, you’ve created an incredibly legacy here for generations to come.

    My only regret is that I didn’t stumble into your site sooner if for no other reason than it would have saved me lots of money wasted on unnecessary stuff!

    I would love to know if there is a single greatest result surprise you uncovered?

    Finally, I’m sure you’ve heard this many times already, but I think many of these images would be awesome blown-up as posters that could be framed. Perhaps even supporting the habit so to speak?

    Thanks again,



  8. This is amazing stuff. I have to say though it does not follow my experience. I find my king 300 significantly faster than my king 1k. And my course crystolon is much faster than the 300.
    Also many people consider the coarse DMT to be a very fast grinder of metal but you show it barely does anything.
    I’m curious why your scientific evidence doesn’t match the evidence of the use of the abrasives


    • I went to great effort to explain that PRESSURE is the critical parameter here. Coarse grits are ONLY effective if we surpass the critical pressure for grinding. Pressure is affected more by contact area than applied force – for example if you vary the angle slightly, you will make contact on corners instead of flats and dramatically decrease contact area and increasing pressure.

      Try flattening the back of a chisel or plane iron with a DMT Coarse.


  9. read this after the king 1k/6k article. Just made me think that in production grinding(not necessarily on fine edges) the feed rate is limited by swarf loading during a single scratch, like if a wood saw with small teeth takes a heavy bite(per tooth) and must travel through a thick board the chips will pack the gullet solid creating several problems with friction and feed pressure. They also use high pressure coolant to blast swarf out of the grain structure. So large grains can enable faster feed on long scratch tracks.
    Also bond strength is matched to the material’s tendency to grab grit at the intended feed speeds and pressures; bond too weak and the grit is wasted (extreme it can’t develop the pressure to cut), too strong of a bond for the conditions and material and the grit goes smooth without exposing a fresh sharp layer(glazing) and creates excess heat.


  10. Great work here. As someone who started producing natural whetstones, I wrote an article and proposed a ‘cut rating’ should replace ‘grit rating’ as a more meaningful measure of what you’re calling abrasion rate – I think you’re the guy to float a new ISO standard!

    “Perhaps a whetstone boffin could create something like a natural whetstone ‘cut rating’ (to supersede ‘grit rating’), which would refer to the effect a stone has (which is what we’re ultimately concerned with) on a standard grade of steel after a standard number of hours, to allow for particle and bond hardness, as well as the size and shape of the particle, to have a role in determining the end result.”


    Liked by 1 person

  11. So in practice, is it possible to just use a single relatively high-grit stone, say the Shapton Glass 4000, and achieve the same result as progressively increasing the grit within the same amount of time?


    • Not in general, but for straight razors it seems to be the case. The combination of low pressure, high steel hardness and low sharpening angle makes coarser grits no more effective or even less effective than a stone like the SG 4k.


  12. Interesting, and it confirms my experience using a KME sharpening system with their diamond plates. Despite everything I’ve read (until now), it was clear to me that more metal was being removed with the 600 and 1500 grit plates than with the 140 grit. I was following KME’s recommendation to apply light pressure with all the plates, but I believe that more pressure is needed with the courser grits (and those plates are probably less prone to wear with the added pressure).


  13. Todd, I’ve just read your thread on Bladeforums “(Why) Are Coarse hones faster than Fine hones?”
    Not sure if I can post links here… let me try: https://www.bladeforums.com/threads/why-are-coarse-hones-faster-than-fine-hones.1449226/
    Very interesting thread. No one offered you a satisfactory answer (within your model). I’ve been thinking about it, and here’s my answer. Better late than never, right?
    1) Let’s consider the situation of static pressure; force straight down into the hone, but without any horizontal movement (for now). A coarse grit particle will penetrate twice as deep, and will make a dent that is twice as long and twice as wide, than a fine grit particle. Therefore, the total volume of material, displaced by a coarse grit will be 8x of that displaced by a fine grit. Thus, one coarse grit will displace 2x the amount of material displaced by four fine ones.
    2) Now let’s add horizontal movement; say, left-to-right (referenced your pictures in the original post of the Bladeforums thread). When one grit makes a scratch that is equal to the diameter of that grit, it enters the trough left by the preceding grit. Notice this is also true for the very first (in the direction of movement) grit – the one that initially was outside of the abraded material and has just entered – it’s in a trough made by its own scratch as well by the previously first grit, which is now second. This is false only for the last grit, which has just left the material.
    3) Now all grits are “in the air”; each of them is in a trough. The downward force lowers the material onto the grits, and the cycle repeats from step 1.
    4) You stated “for the same distance traveled, the length of those swarf particles would be the same for both hones”. However, as demonstrated, coarse and fine grits don’t travel the same distance. The length of the scratches they produce equals to their diameter (within your model). Each swarf particle from a coarse hone is twice as thick, twice as wide, twice as long and eight times bigger than the one from a fine hone.
    5) So, a hone that is twice as coarse (and all other things being equal) should indeed work twice as fast.
    6) I don’t know why your DMT 325 is so blunt. Probably DMT’s fault.


    • Some more remarks:
      1) As the coarsness of a hone increases, particle density decreases quadratically, while volume of displaced material increases cubically (for a given level of force). That’s why coarse hones tend to be faster than fine ones.
      2) Imagine we removed every second particle from our hone (say, from the coarse one). The remaining grits will sink deeper into material; the total volume of displaced material will increase by sqrt(2)^3 (that is, by 1.4142… in the power of 3), which is approximately 2.8 (if the grits are tall enough for that). So there is only half of particles remaining, but each one is more than twice as effective – the overall effectivness of a hone is increased by about 40%. More precisely, by sqrt(2).
      This is presumably why your fine diamond plates are so effective. Of corse, I’m cheating here, because you already pointed that out several times 🙂
      3) The fastest hone has just one huge abrasive particle. It might be a bit difficult to use, though…
      4) I think sideways force (parallel to the surface of the hone) shouldn’t change irregardless of coarsness? But thinking is hard and I’m already tired of it… I’ll leave it as an exercise for the readers.


      • And one more remark:
        There is no magic. With a coarser hone we achieve deeper penetration, more distance so more work (in downward direction) is performed. Recall that “work” is “force * distance”. If we maintain the same – presumably low – level of force as with a fine hone, then we’ll need to maintain that force for longer time, since penetration cannot be instant.
        In order to achieve the same speed of penetration with the coarse hone, we’ll need to… indeed, press harder.
        And if we keep the low level of force, then in the same time our coarse hone will penetrate just as deeply as a fine hone (!); it will work then the same. Thus, there is really no magic.
        Therefore, if the force you can apply is limited (e.g., you are grinding a straight razor and it flexes too much if push too hard), a fine hone should indeed work as fast as for the amount of material removed.
        Removing more metal requires more time grinding, or more effort grinding. There is no way around it. Considering that, my assertion 5 – “a hone that is twice as coarse (and all other things being equal) should indeed work twice as fast” – is not really correct.
        Damn, it was a long time since I last thought about physics. Anyway, hopefully the authour and some others will find this interesting.


        • > Considering that, my assertion 5 – “a hone that is twice as coarse (and all other things being equal) should indeed work twice as fast” – is not really correct.

          And, to be honest, it’s simply incorrect, because “fast” implies time, and the unstated assumption (in the three-step grinding model above) that steps 1 and 2 take the same time for both coarse and fine hones, is wrong in real world. Doubled penetration depth should take twice as much time, and so should sideways movement that is twice as long! Or, alternatively, 2x force in the same time.


          • Now that I rested and thought some more about how time works in Todd’s model…
            The speed of penetration must depend on the rate of material displacement. Obviously, material cannot move to a new place instantaneously, it must take some time. It seems reasonable that the amount of material displaced per unit of time is directly proportional to pressure. At least, seems reasonable for steel to me… Coarse grits exert the same pressure (overall) as fine grits (since their surface area increases quadratically with coarsness, while density drops also quadratically) but displace twice the amount of material, therefore they must work for twice as long.


Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.

%d bloggers like this: