Measuring the maximum temperature of an edge during belt sanding: initial results

[REK Knives]

How much did we know before starting?

Ok, here is another link I've found thanks to your quoting it before hypefreeblades forums • View topic - Unwanted Tempering Whilst Sharpening
with bibliography

Thats about burning a few microns in hand grinding dry. Since edge is only about a micron, and damage is measurable ... my takeaway use water (at least) in hand grinding as well


That is kind of a different question ... if the question is important maybe a simple test of edge retention ?

But I don't think the question is too important, unless coolant consumption is eating into your profits ... then it might be a good idea to pay for some micrographs ...

Roman Landes seems open to questions/discussion on hypefreeblades, so might be a good idea to ask directly for details.

Edge retention tests, when done properly are anything but simple from what I've seen. I've been wanting to work on this with cliff and he's willing but I haven't had time nor money to go out and purchase some cheap test knives (he's in Canada and we would need several so multiple runs could be made).

At the risk of derailing this thread 😀 if you scroll to the bottom of this page and the first post at the top of the next I asked almost these exact same questions to which Roman danced around the answer and did not answer any of them directly. I am not saying he did this on purpose, he may have been busy and not really paying attention, but they went unanswered nonetheless.

 

[HeavyHanded]

Edge retention tests, when done properly are anything but simple from what I've seen. I've been wanting to work on this with cliff and he's willing but I haven't had time nor money to go out and purchase some cheap test knives (he's in Canada and we would need several so multiple runs could be made).

At the risk of derailing this thread 😀 if you scroll to the bottom of this page and the first post at the top of the next I asked almost these exact same questions to which Roman danced around the answer and did not answer any of them directly. I am not saying he did this on purpose, he may have been busy and not really paying attention, but they went unanswered nonetheless.

Thanks for the link. This topic always fascinates me not the least because there never seems to be a complete accounting of methods/time/belt speed/comparative temps (there are no control images of a wet ground tool edge to accompany the heat-damaged one) etc etc, and in general no controls are used or mentioned in these discussions. Also there is never mention of heat generated during various common usage that surely approaches some grinding/polishing operations. I also have yet to find thermocouples that operate at high enough speed at sub micron scale under the conditions needed to study realtime interaction between steel and an abrasive. Not that they don't exist, but high speed/resolution thermal study at that scale is usually done with optics.

Add to that different abrasives all have different thermal characteristics - diamond for example conducts heat much more readily than SiC or AlumOx, different abrasives will have different percentages of rake angles, different characteristics of chip creation, plowing, rubbing action etc etc that all effect heat generated from grinding, and you have a real circus.

I certainly have noticed that many factory edges do not hold up well, but then most are overly obtuse to begin with and that kills edge retention right off the top. No argument that dry grinding has all the potential to harm HT and will if care is not exercised, but in use I have seen absolutely zero evidence that at slow/hand speeds and with the abrasive surface in good condition there is any difference at all.

As for the chip burning an eye from a distance coming off a ceramic rod, sounds like I should just bring a crock stick when camping and sharpen my knife over my tinder. If a micro-spec of steel can carry a thermal load across open air, and still have enough energy to burn the surface of a lubricated eyeball, it should work wonders starting a camp fire. There must be another explanation for that story.

 

[chiral.grolim]

... there never seems to be a complete accounting of methods/time/belt speed/comparative temps (there are no control images of a wet ground tool edge to accompany the heat-damaged one) etc etc, and in general no controls are used or mentioned in these discussions. Also there is never mention of heat generated during various common usage that surely approaches some grinding/polishing operations. I also have yet to find thermocouples that operate at high enough speed at sub micron scale under the conditions needed to study realtime interaction between steel and an abrasive. Not that they don't exist, but high speed/resolution thermal study at that scale is usually done with optics.

Add to that different abrasives all have different thermal characteristics - diamond for example conducts heat much more readily than SiC or AlumOx, different abrasives will have different percentages of rake angles, different characteristics of chip creation, plowing, rubbing action etc etc that all effect heat generated from grinding, and you have a real circus.

I certainly have noticed that many factory edges do not hold up well, but then most are overly obtuse to begin with and that kills edge retention right off the top. No argument that dry grinding has all the potential to harm HT and will if care is not exercised, but in use I have seen absolutely zero evidence that at slow/hand speeds and with the abrasive surface in good condition there is any difference at all.

As for the chip burning an eye from a distance coming off a ceramic rod, sounds like I should just bring a crock stick when camping and sharpen my knife over my tinder. If a micro-spec of steel can carry a thermal load across open air, and still have enough energy to burn the surface of a lubricated eyeball, it should work wonders starting a camp fire. There must be another explanation for that story.

:thumbup::thumbup:

This is why I appreciate what Cyrano is doing here - clearly presenting ALL of his methods/materials along with the results for criticism, etc. which he takes back for further testing. Regarding common usage that heats an edge, it would be interesting to see what his method evinces on a blade used for cutting cardboard or chopping wood - I have personally felt how uncomfortably hot the bevel of my knife or axe can get after doing a lot of cutting. I guess I should be using coolant whenever I'm cutting dry material with my knife? When we talk about knives having more/less abrasion-resistance via carbides, do we think the CATRA tests do not generate much heat at the apex of the knives? Or do we think the heat generated is insufficient to significantly impact performance?

The fact of the matter is that, just as Cyrano indicated in his first post, there is STILL NO DEFINITIVE ANSWER regarding damage to apices through over-heating during grinding - we know that it CAN happen, we certainly know HOW to cause it if desired (Cyrano shows us how in this very thread), but it remains to be shown that it MUST happen if specific coolant is not used - steel type, grit, pressure, time, other circumstances.
Landes offered limited evidence of a specific tool made from steel that very likely has low heat-resistance (since we don't know the steel, no conclusion can really be drawn), ground in an unspecified manner (he gives the grit, but no time or pressure which are key components without which no conclusions can be drawn), and the hardness tester used finds a change in hardness within the first 100 microns (NOT the first 300) but we do not have the control measurements of the hardness at those points prior to sharpening. *shrug* So what he offers is a warning that power-grinding dry might damage the first 100 microns of your edge. :thumbup: And stay away from crok-stiks?

Question: why do "high speed steels" exist? What are their properties? What attributes contribute to "abrasion resistance"?

 

[ToddS]

The two measurements that are below the average hardness are made very close to the edge of the blade cross-section. These type of measurements must be made at least 3 indent diameters from the edge - the image is of too low a quality to determine whether those last two data points are far enough from the edge. Also, this type of polishing typically rounds off the specimen near the potting compound interface- so the surface may not be normal to the indenter at those points, which again influences the result.

I used this type of thermal indicator (from Omega) about 25 years ago while working on my PhD thesis and found it to be very unreliable away from thermal equilibrium. Typically the phase change occurs at much lower temperature than the actual surface temperature. A large part of my thesis work involved measurement of temperature in the outer hundreds of nanometers from the surface.

Several years ago I did an experiment to prove that stropping does not involve "heating" the apex. By placing a blade (a straight razor in this case) apex-down on a hot plate (with a thin sheet of aluminum foil to ensure good thermal contact) then measuring the spine temperature to determine whether there is any merit to the old saw about heat flow being slow through the small cross-section of the bevel (16 degree inclusive in this case). The spine temperature did in fact rise rapidly. The key point that is usually missed is that although the cross-section is small, so is the contact area.

I would speculate that any damage may be mechanical, and the rise in temperature occurs as a result, not as a cause.

 

[Joe_Lane]

I knew nutnfancy was onto something with that wd40 when battoning

 

[Gaston444]

It's constructive in that he is asking what the controls were and implying that without knowing the variables we do not really know much more than when we started.

For instance. One can (and people do, based off of landes research) argue that wet grinding (which I do, I have converter my 2x72 belt girder to be cooled with a coolant during grinding) will not damage the edge at all but nowhere that I've seen, and I've looked, has this topic been researched or discussed (i.e. Wet grinding vs dry grinding, with the same fresh grit belt and speed, etc with micro thermocouples at the apex and what difference is truly made with wet grinding vs dry) . Who is to say that wet grinding at x speed won't damage the temper and if so what pressures were used, how long did the knife stay on one spot on the belt, how dull were the abrasives, etc etc.

Great post which shows the huge amount of variables involved... You'd have to define a controlled motion robot arm to even begin to equalize all the variables...

The two main variables in outcome I see is the amount of apex softening and/or brittleness, and the depth under the apex of the softening and/or brittleness...

I find the story of the particle being "burned" into the eye tissue, from hand sharpening, hard to believe: I think it was not burned into the tissue, but that it embedded itself through velocity, not heat..

I think this because there is a great distance to travel for a proportionately very small object... The smaller an object is, the bigger its surface compared to its internal volume: This means the smaller the object is, the quicker it cools down... This is why artic animals tend to be bigger... Travelling from the sharpening surface to the eye, even if very fast, means travelling a distance several thousands of times its own size, and it certainly was not glowing hot to begin with, yet it supposedly showed burn marks in tissue that is essentially mostly liquid... Does a meteorite leave burn marks in a swamp?

I think the particle was embedded and the doctor assumed it was heat and velocity combined, but it was really just velocity. From the embedding alone I would have assumed a power tool just like him...

Gaston

 

[Bill DeShivs]

While I appreciate all the scientific testing, I can tell you this:
There are way too many variables involved to come to more than a scientific "maybe."
Experience tells me that it is absolutely possible to overheat an edge when using powered equipment-maybe even possible at a microscopic level when hand sharpening. This is why experience is important when sharpening. It can tell you what not to do.

Much ado is made of the "burr" when sharpening. It's entirely possible that the burr is either annealed or work-hardened metal. Simple stropping works it back and forth and causes the burr to break off.

 

[bucketstove]

The fact of the matter is that, just as Cyrano indicated in his first post, there is STILL NO DEFINITIVE ANSWER regarding damage to apices through over-heating during grinding - we know that it CAN happen, we certainly know HOW to cause it if desired (Cyrano shows us how in this very thread), but it remains to be shown that it MUST happen if specific coolant is not used - steel type, grit, pressure, time, other circumstances.
Landes offered limited evidence of a specific tool made from steel that very likely has low heat-resistance (since we don't know the steel, no conclusion can really be drawn), ground in an unspecified manner (he gives the grit, but no time or pressure which are key components without which no conclusions can be drawn), and the hardness tester used finds a change in hardness within the first 100 microns (NOT the first 300) but we do not have the control measurements of the hardness at those points prior to sharpening. *shrug* So what he offers is a warning that power-grinding dry might damage the first 100 microns of your edge. :thumbup: And stay away from crok-stiks?

Um, wow
Stay way from crok-stiks is what you get out of that discussion?
Not use lube even with crock sticks?


Maybe its just me, but if I were to try to pick apart what Roman Landes says, I might read his published/peer reviewed papers/book ... but I don't read much german, and not too great at reading charts either

 

[chiral.grolim]

Um, wow
Stay way from crok-stiks is what you get out of that discussion?
Not use lube even with crock sticks?

Maybe its just me, but if I were to try to pick apart what Roman Landes says, I might read his published/peer reviewed papers/book ... but I don't read much german, and not too great at reading charts either

I am very good at reading both peer-reviewed papers and charts, it is part of my "trade". But like you, I don't read much German. I do however read English and have not found much to corroborate Landes' assertions (most of which he simply asserts and then references German papers that we cannot review to critique methods/materials) among English/American papers on the same topics. What IS presented for possible review/critique are charts like the above, which I and others have critiqued. If that chart represents the "best" demonstration of the point, then it does little more than what Cyrano did here in this very thread - present possible concerns/considerations for power-grinding steels - except that the chart does not come with all of the important details of steel-type, grinding pressure, & grinding time. Landes' own remarks in response to these very questions in the threads you linked indicate that he considers those other factors irrelevant, even to the point that he asserts that dry crok-stik sharpening will burn your edge, supported by his little anecdote. If you accept that assertion without repeatable demonstration... ?

I point out the crok-stik remark as the "last straw" because it is beyond the pale (to make an Irish reference).

Again, there is very little evidence presented to suggest that power-sharpening creates a noticeably negative impact on edge-stability unless a deliberate effort is made to induce damage. Not so? If this is incorrect, please enlighten us. It is also the case for dry hand-sharpening, and most importantly for dry USE - the most widely accepted test of edge-performance is CATRA which is not done using coolant or lubricant, furthermore nearly every domestic and industrial use of knife edges is done dry (you might except kitchen processing of cold wet materials).

It might be noted that we have developed "high speed steels" that are less sensitive to heat generated during use precisely because of heat-damage to the apex, but steel-type has not been brought into the discussion of burning edges... why not? Again, those alterations in steel formulation are irrelevant? They were undertaken without rigorous testing and investigation? We make costly complex steel formulations for no reason?

Furthermore, as has been pointed out in those linked threads, no evidence has been presented that the application of "coolant" actually prevents the supposed heat-damage to the very apex. We know heat is bad, but we don't know that the amount we are generating by our various methods, or the efforts we are taking to reduce it, are having an appreciable impact. I think we ALL would appreciate seeing that, but it seems that running an appropriately controlled experiment to discern that is too complicated and all differences really are just lost in the "noise" of more important factors like impact-damage from abrasive particles against the apex or variations in micro-geometry.

Nevertheless, i appreciate that Cyrano is trying to find a way to roughly guide users as to how one CAN generate sufficient heat to damage an apex and perhaps how one can work to avoid it.

 

[HeavyHanded]

I have one last thought on this aside from commenting on any additional material Cyrano provides. Prefaced by the statement that I do not know one way or the other, but as many have noted it may not make a difference compared to other factors of steel type and geometry. I have no standing to dismiss or verify any of Landes findings, and at some intersection of conditions I'm 100% certain the HT will be compromised with dry powered grinding or in extreme conditions even in the presence of a bit of coolant.

I am left with one puzzling thought - if it is that easy to generate extremely high damaging temps on HT'd steel from dry abrasion, what of the common file? Surely it would loose its temper along the cutting edges since files are almost universally used dry. Unless the abrasion process is dramatically different from a shaped steel cutter to a mineral abrasive, the file would be ruined almost immediately, as well as the temper of any cutting edge being shaped with same.

Too many questions, I'll keep wetting my belts with a sponge when power grinding, but for hand sharpening or slow speed powered work will go with my gut and observations.

 

[chiral.grolim]

Too many questions, I'll keep wetting my belts with a sponge when power grinding, but for hand sharpening or slow speed powered work will go with my gut and observations.

I do wonder how often the use of coolant is advisable to protect, as you mention, the "file" i.e. the abrasion/cutting tool itself whose teeth or carbides might more easily be damaged/dulled/dislodged by excessive heat in the surrounding material... use a coolant to reduce wear on the tool doing the cutting as opposed to damage to the surface being cut.

 

[chiral.grolim]

So I found a paper from UMass published 1995 in the JOURNAL OF ENGINEERING FOR INDUSTRY that may be applicable:
http://manufacturingscience.asmedigitalcollection.asme.org/article.aspx?articleid=1447963

The authors embedded thermocouples into steel (including AISI 1020 at 40 Rc and also 52100 and O1 steels hardened to 62 Rc, none "high speed" or stainless alloys) and ground the surfaces repeatedly using abrasive wheels running 30 m/s (for reference, the HF 1x30 runs ~15 m/s) cutting to a depth of 25 microns on each pass, with the steel plate feeding at ~13 cm/s - I am not skilled enough to translate this to applied pressure of the grinder against the work-surface, it may be that the pressure is lower than hand-sharpening but I doubt it...

They found that the total energy generated by the grinding-action could be converted to a maximum heat <850'C for 4 milliseconds.
They found that 60-75% of this energy is transferred to the workpiece as heat when using Aluminum-oxide wheels under these conditions, resulting in heating the first 10-20 microns of the surface to ~500'C for 4 ms.
They found that only ~20% of the energy was transferred to the workpiece as heat when using CBN wheels under these conditions, resulting in heating the first 10-20 microns of the surface to ~120'C for 4 ms.
They found that you could not use the CBN wheels under these parameters even with lubrication to achieve good cutting of the 1020 plate because it is too soft and glazes readily, the harder steel plates cut better.
The CBN abrasive has higher thermal conductivity and retains its sharpness longer (so they suppose) such that significantly more heat is conducted away from the work-surface...


And another article from the journal Materials Research 2002: http://www.scielo.br/scielo.php?script=sci_arttext&pid=S1516-14392002000200016
Among the conclusions is this:

The cooling properties of the cutting fluids can be neglected and don't cause any improvement in the reduction of the grinding zone temperature and in prevailing compressive residual stresses.

Now both of these papers are not specifically related to knife edges, and both DO recommend the use of lubricant while power-grinding, but they are also discussing grinding at much higher speeds and cutting depths, and the lubricant is recommended primarily to preserve the life of the abrasive which tends to be more heat sensitive than the work surface. *shrug*

Here is another paper, 2008: http://wumrc.engin.umich.edu/wp-con...ocouple_fixation_method_for_grinding_temp.pdf
This one is trying to develop a more accurate method of measuring total grinding temperatures (not just the amount absorbed by the work-surface) via thermocouples.
They were grinding (dry compared to wet) with an alumina-oxide wheel into The workpiece material was Dura-Bar 100-70-02 (carbon content of 3.5–3.9%, Rockwell hardness HRC 50, thermal conductivity of 63 W/m K, and thermal diffusivity of 1.63).
They measured a maximum temperature rise at the very last microns of contact to be 500'C with ~85% transfer to the work-surface (425'C) correlating closely with the first paper posted above. Once again, this was noted:

Grinding experiments and heat transfer analyses showed that grinding fluids provide negligible cooling within the grinding zone.

Are these data points relevant to the discussion?
52100 is tempered at ~200'C for two hours, what is the impact of heating even the last 10 microns to 200'C for 4 ms?

It is mentioned in these and other papers I've found that water or majority-water lubricants largely retain the thermal properties of water and so vaporize before they reach the point of contact, such that their impact is not felt by the contact-area itself which responds as if it were being cut dry. The primary reason for using these liquids at all is their lubricity (reducing friction) and to wash away swarf that can clog the abrasive and cause irregular grinding.

Again, none of this means that you cannot over-heat an edge through power-grinding or maybe even hand-grinding, thereby ruining the temper at the very apex. I think we know that it CAN happen, bluntcut has put forward (on various occasions) the factors involved in getting it done, but I challenge the assertion that it MUST happen without use of a coolant/lubricant or that the use of these matters at all for that last few microns of apex.

 

[Twindog]

The technical aspects of this discussion are beyond my ken. But anecdotally, power sharpening can produce awesome edges, at least when used to do the heavy initial hogging of metal.

Following a discussion on chipping of 3V, Jerry Holsom sharpened my 12-inch chopper in that steel on a belt, and that amazing edge has held up for a long, long time. If the edge hardness of that edge had truly been knocked down from 60 Rc to 55 Rc, I'd know it immediately.

I've been experimenting with a mix of power and hand sharpening systems -- a 1x42 belt sander, Wicked Edge, Sharpmaker and hand strop. The belt sander is so nice for reprofiling an edge. I can reprofile a high-hardness steel edge in minutes on the belt grinder, compared to hours on the Wicked Edge. Then I follow up the power reprofile by putting a fine edge on the blade with a Wicked Edge. Then I finish with a strop. I can't tell any difference in performance (how long the edge lasts) between using the power system for the initial grind and a complete hand reprofile and sharpening.

The other big advantage of the belt grinder is touching up a slightly worn edge with a leather strop and green compound, followed by a gentle hand stropping with the same compound. Fantastic and easy edge. For a minute or two of my time, I get an edge that wants to grab everything it touches. I think the hand stropping gets rid of the residual burr that it hard to remove with a power belt.

 

[BluntCut MetalWorks]

It's well known surface grinding generate a lot of heat, where temperature could exceed steel melting point. When grind depth is fixed between 2 fixed surfaces (grinding wheel and work piece), the ratio between fracture and smear-off depend on abrasive strength and sharpness. Since grinding pressure essentially infinite (for non-chattering operation), and cut depth is fixed, therefore the remaining percent of surface-to-be-remove not done by fracture&plow, will be done the smear/friction/pile-up. This smear force is beyond work-piece material UTS, hence damaging heat happened in very short amount of time. CBN wheel is stronger and perhaps stay sharp longer, so fracture/plow % is higher than AlO wheel, in turn less smear = less heat.

While sharpen knife with belt, where belt & blade are not fixed - i.e. easy flex and or move away depend on amt of downward pressure. Swarf trap/pile-up doesn't change to total pressure involves, unlike force is cummulative for the case (above) of 2 fixed surfaces at X feed rate. So, with a low SFPM(says ~200 SFM) and sharp belt, it's possible to sharpen w/o damaging the top 1um of the blade apex. Repeat, sharp abrasive is very important - avoid burnishing/non-abrading-impact/smear/rubbing as much as possible.

So I found a paper from UMass published 1995 in the JOURNAL OF ENGINEERING FOR INDUSTRY that may be applicable:
http://manufacturingscience.asmedigitalcollection.asme.org/article.aspx?articleid=1447963

The authors embedded thermocouples into steel (including AISI 1020 at 40 Rc and also 52100 and O1 steels hardened to 62 Rc, none "high speed" or stainless alloys) and ground the surfaces repeatedly using abrasive wheels running 30 m/s (for reference, the HF 1x30 runs ~15 m/s) cutting to a depth of 25 microns on each pass, with the steel plate feeding at ~13 cm/s - I am not skilled enough to translate this to applied pressure of the grinder against the work-surface, it may be that the pressure is lower than hand-sharpening but I doubt it...

They found that the total energy generated by the grinding-action could be converted to a maximum heat <850'C for 4 milliseconds.
They found that 60-75% of this energy is transferred to the workpiece as heat when using Aluminum-oxide wheels under these conditions, resulting in heating the first 10-20 microns of the surface to ~500'C for 4 ms.
They found that only ~20% of the energy was transferred to the workpiece as heat when using CBN wheels under these conditions, resulting in heating the first 10-20 microns of the surface to ~120'C for 4 ms.
They found that you could not use the CBN wheels under these parameters even with lubrication to achieve good cutting of the 1020 plate because it is too soft and glazes readily, the harder steel plates cut better.
The CBN abrasive has higher thermal conductivity and retains its sharpness longer (so they suppose) such that significantly more heat is conducted away from the work-surface...


And another article from the journal Materials Research 2002: http://www.scielo.br/scielo.php?script=sci_arttext&pid=S1516-14392002000200016
Among the conclusions is this:


Now both of these papers are not specifically related to knife edges, and both DO recommend the use of lubricant while power-grinding, but they are also discussing grinding at much higher speeds and cutting depths, and the lubricant is recommended primarily to preserve the life of the abrasive which tends to be more heat sensitive than the work surface. *shrug*

Here is another paper, 2008: http://wumrc.engin.umich.edu/wp-con...ocouple_fixation_method_for_grinding_temp.pdf
This one is trying to develop a more accurate method of measuring total grinding temperatures (not just the amount absorbed by the work-surface) via thermocouples.
They were grinding (dry compared to wet) with an alumina-oxide wheel into The workpiece material was Dura-Bar 100-70-02 (carbon content of 3.5–3.9%, Rockwell hardness HRC 50, thermal conductivity of 63 W/m K, and thermal diffusivity of 1.63).
They measured a maximum temperature rise at the very last microns of contact to be 500'C with ~85% transfer to the work-surface (425'C) correlating closely with the first paper posted above. Once again, this was noted:



Are these data points relevant to the discussion?
52100 is tempered at ~200'C for two hours, what is the impact of heating even the last 10 microns to 200'C for 4 ms?

It is mentioned in these and other papers I've found that water or majority-water lubricants largely retain the thermal properties of water and so vaporize before they reach the point of contact, such that their impact is not felt by the contact-area itself which responds as if it were being cut dry. The primary reason for using these liquids at all is their lubricity (reducing friction) and to wash away swarf that can clog the abrasive and cause irregular grinding.

Again, none of this means that you cannot over-heat an edge through power-grinding or maybe even hand-grinding, thereby ruining the temper at the very apex. I think we know that it CAN happen, bluntcut has put forward (on various occasions) the factors involved in getting it done, but I challenge the assertion that it MUST happen without use of a coolant/lubricant or that the use of these matters at all for that last few microns of apex.