Thread: A question about quenching media, and the science behind quenching

But you are taking observations and learning from them. I am fascinated by the fact that so many "good" blades have been made without the makers having the benefits of the education available to us. Humans are keen observers and learn from their mistakes.

Right, and your mention of ritual is/was a part of this human learning. Trying to repeat what has worked before as accurately as possible. An audience member at a talk, that I attended, about Japanese swords suggested that a lot of the ritual and prayer during the sword making was related to the timing of different operations during the forging.

To my understanding, aside from inclusions and cracks in steel, the reason blades get damaged on quenching isn't the reduction on size of the martensite crystals, while a millimeter next to it a different form of iron-carbon combination exists, like ferrite, that doesn't change shape? Steel gets literary stretched or pressed, until it snaps. Is there some other serious reason a blade breaks on quenching?
What I'm purposing is to have the whole blade, or cubic centimeter/inch to drop its temperature uniformly whether it's the edge or center (of course, the center will be a bit harder since it's further away). The liquid alloy will get hotter as you dip the blade, ending up a few degrees colder than the steel in an instant.
This liquid alloy won't work like water, a polymer or salt or NaOH and any other combination but like the outer layer of a few molecules from a piece of steel that is being quenched since the heat transfer will be similar. Theoretically, the temperature difference between the hot blade and finished-tempered blade will not matter because of that heat transfer.

The same way heat travels inside the hot piece of steel that gets colder will also be transferred out of it is what I'm saying reducing the chance for failure.
So, a fast cooling rate between Ms and Mf won't matter since it will be happening to the whole blade, and, from the center of it, to the outer layer, there will be no deformation, aside from the one because it got cold. After that, increasing the volume of the alloy, the surface of the pot with the other pot that contains water or leaving it to cool down by itself as simple as posible is up to the blacksmith-scientist.

@ Vasilis:

“If the piece of steel cools uniformly, no matter its speed, wouldn't that reduce greatly the formation of these cracks?”

“To my understanding, aside from inclusions and cracks in steel, the reason blades get damaged on quenching isn't the reduction on size of the martensite crystals, while a millimeter next to it a different form of iron-carbon combination exists, like ferrite, that doesn't change shape? Steel gets literary stretched or pressed, until it snaps."

You are right, the cracks form due to the stress of the surface of the piece cooling faster than the centre of the piece, not due to (in most part) there being slightly different rates of cooling at different parts of the surface (due to the formation of vapor jacket, which with good technique can be considered negligible anyway).

There is no way possible (when quenching austenite straight to martensite) of dropping the temperature at the surface and in the centre of the piece at the same rate (the physics of heat transfer in steel does not allow for that). Different cooling rates means different structures present in the steel (which have different shapes/densities) which results in stress. Faster cooling rates (i.e. faster quench mediums) exacerbates the differences and leads to increased failure.

Marquenching is the solution you are looking for, as it insures that the entire piece is at the same temperature before it is cooled (relatively slowly) through the martensite transition zone. Therefore giving you your desired “uniform cooling”. Obviously there will still be a temperature gradient across the piece as it cools through the martensite transition zone, but as the cooling rate is so slow (i.e. in ambient air) the gradient can be considered negligible


“I would like to find a way to achieve hardness, for low alloy steel with carbon content close to 1%, of above 65 RC (like some Japanese razors that have showed a hardness rating of 67RC, how could they do that in an age where knowledge and technology was behind) and am searching for ways to do that. If I could reduce the failure rate, that would be a big bonus since grinding a blank, at least for me, takes days.”

Try marquenching, alternatively:
More uniform grinding and the elimination of stress risers in the blank will also help, the former takes years of practice and the latter is a matter of simple design (and may be helped by doing more of the grinding post heat treatment so that the quenched blank is thicker and less intricate).
Quenching spine first may also help

However, blade failure is an occupational hazard for smiths using water as a quenching medium, and water is necessary (when quenching straight from austenite to martensite) to get low alloy/plain carbon steel razors to their maximum hardness (when using forms relevant to blades/razors). I recall seeing a YouTube video of the Iwasaki workshop that showed a pile of razor blanks that had failed during heat treatment.

@ Mike Blue:

“I will just stick this out there...I do not quench in brine. I see no reason to do so using a well-prepared known alloy of steel and do not recommend it unless you like to gamble on how many blades you will get to survive this violent process. It's just not needed but persists in the Mythology of How Things Should Really Be Done. I can identify a very limited set of conditions that might require brine. Mostly they are unknown steels from the junkyard that won't harden using normal techniques.”

I agree (and I imagine anyone who understands metallurgy would do likewise) that brine is unnecessary for any steel when used in forms appropriate to razor/knife blade. However, Tim Zowadas post (http://straightrazorpalace.com/forge...elf-today.html) beautifully exemplifies a further set of (very limited) conditions where brine is appropriate.

“And, Rob Gunther's Superquench is a soap solution that addresses the surface tension. It will leave a hardened thin skin on even low carbon steels. The first honing will remove anything good about using this on a cheap steel to achieve a quick result. It's best to use known materials of good quality and good heat treatment rituals.”

I found Superquech an interesting phenomenon that allows some hardening of “unhardenable” steel (an alternative to case hardening if you will), but you are wise in advising that it cannot in any way replace a good heat treatment of a “hardenable” steel”.

@ Vasilis:

“If I could reduce the failure rate, that would be a big bonus since grinding a blank, at least for me, takes days.”

The laminar construction and/or differential hardening techniques used (initially) by Japanese smiths may also help. By hardening only the cutting edge, the rest of the blank works as a stable support matrix during the quench to prevent (or at less lessen) the stress of hardening cracking the blade.

Originally Posted by bluesman7

I've been interrupting my quench at somewhere around 450 F (a SWAG and a 480 degree temp stick) The last time I did this I was thinking that the steel, in theory, should still be austenite at this point. So directly out of the quench I checked the steel with a magnet and low and behold it was non-magnetic despite being almost a 1000 degrees below the curie temperature. Ain't science great!

Just an interesting side note, mildly related to the topic.

While normalizing my last batch of 1095 blades, I did what you are talking about: Played with temps and magnet, cool stuff!

On my phone so short answer.
Laminating with lower carbon steel makes it harder not easier to keep things straight because both sides will not be identical. Additionally the piece may just tear each other in half lengthwise if the core hardens and then the sided contract strongly.

Also uniform cooling is all but impossible on a razor with less than 1/16 edge and 1/4 spine

Great discussion btw.
My experience is that warp is caused most commonly by not enough normalisation or not holding the blade straight during the quench so one side cools faster.

Also gently quenching instead of ramming it in as fast as possible seems to work for me

Gentlemen, you are right.
I was trying to figure out a way to quench a piece while trying to achieve same thermal conductivity of that liquid alloy with the piece of steel being quenched, but I forgot that you have to "force" martensite out of austenite. Without martempering, that's a paradox.
I thought that thermal conductivity was a key for that. Only, you have to go beyond that and drop temperature more violently than steel's own thermal conductivity so that "carbon atoms do not have time to diffuse out of the crystal structure in large enough quantities" for the formation of martensite. It sounds so obvious now that it was way too simple to "achieve maximum hardness with close to zero shock on the piece of steel while losing energy" to be true, laws of phisics don't allow it, martempering aside, again. But playing with the rate the temperature drops could help achieving something that again belongs to the category of martempering.

My new question; can we achieve maximum hardness with martempering? To my understanding, we can achieve high hardness while the piece of steel suffers the effects of crystalline changes to a lesser degree, so, good hardness, less damaged blades. But can we go above the, say 62 RC with that method? Again, for a simple steel, low alloy and close to 1% carbon (I'm using O2 steel, it's the cheapest available high quality blanks I can locally find. I'm not asking for this steel specifically but generally).
Is it possible to achieve higher hardness than the 60-62? Submerging blades in liquid Nitrogen and other fancy ways to reach subzero temperatures before or between cycles of tempering sounds bothersome for half a degree RC to do them.
(I remember reading an old chemistry book 20 years ago, elementary school probably, 1960-70 university book, where it was written that with ice and enough CaCl2 you can drop the temperature to below 60 degrees C).

Originally Posted by Vasilis

Gentlemen, you are right.
I was trying to figure out a way to quench a piece while trying to achieve same thermal conductivity of that liquid alloy with the piece of steel being quenched, but I forgot that you have to "force" martensite out of austenite. Without martempering, that's a paradox.
I thought that thermal conductivity was a key for that. Only, you have to go beyond that and drop temperature more violently than steel's own thermal conductivity so that "carbon atoms do not have time to diffuse out of the crystal structure in large enough quantities" for the formation of martensite. It sounds so obvious now that it was way too simple to "achieve maximum hardness with close to zero shock on the piece of steel while losing energy" to be true, laws of phisics don't allow it, martempering aside, again. But playing with the rate the temperature drops could help achieving something that again belongs to the category of martempering.

I would encourage you to continue to think about the problem. As often happens, breakthroughs occur because the status quo argues that nothing new is left to be discovered. On the other hand, you want to get shop work done and the methods that are available do have a track record of some success. And sometimes you will run the experiment knowing it's not likely to succeed because you're bored or you just have to pee on the electric fence for yourself.

My new question; can we achieve maximum hardness with martempering? To my understanding, we can achieve high hardness while the piece of steel suffers the effects of crystalline changes to a lesser degree, so, good hardness, less damaged blades. But can we go above the, say 62 RC with that method? Again, for a simple steel, low alloy and close to 1% carbon (I'm using O2 steel, it's the cheapest available high quality blanks I can locally find. I'm not asking for this steel specifically but generally).
Is it possible to achieve higher hardness than the 60-62? Submerging blades in liquid Nitrogen and other fancy ways to reach subzero temperatures before or between cycles of tempering sounds bothersome for half a degree RC to do them.
(I remember reading an old chemistry book 20 years ago, elementary school probably, 1960-70 university book, where it was written that with ice and enough CaCl2 you can drop the temperature to below 60 degrees C).

I am going to suggest another course of reading. Find the isothermal transformation diagrams about the particular steel (they each have one if they are common steels or the steel supplier should have one for each that they sell) and read, study, then test the variables. Each of these diagrams has large sample size so the numbers are really pretty valid and reliable. The curves shown will give you predictable information about the correct temperatures and times necessary to achieve the hardness value you desire. If that hardness number is not on the diagram, it's probably not possible for that steel, at least under the conditions specified. Example: 1030 (a simple carbon steel) will not achieve greater than Rc45 under normal conditions of heat treatment. That is as-quenched hardness. Tempering it will reduce the max Rc. You could use a Superquench to achieve a harder "skin" but it's not very deep and the parent material is not Rc45. Such quenchants were not tested developing the diagrams. Most of that benefit would be abraded away honing anyway.

About liquid nitrogen: The people selling the equipment to process steels with this stuff are the ones who produce the research into it's benefits. I see that as a conflict of interest. There are steels that benefit, notably high-alloy steels, e.g. stain resistant types, that often produce retained austenite (RA) when HT'd. Ordinary carbon steels, if HT'd correctly, do not produce enough RA to benefit practically from such treatments. There may be a statistical proof of change (see conflict problem above) but that does not translate into an economic benefit to the smith/maker or the customer (your point about bothersome). John Verhoeven's book specifically mentions LN treatments in the chapters on stainless steels. No where in the chapters on carbon steels do you read about LN or cold treatments during HT processes.

Quenching a hot billet into a liquid nitrogen bath could be very exciting for the operator. There was a How It's Made show that was produced about razors that showed very briefly a thin strip of steel for multiblade plastic razors emerging from the induction heater to austenitize the steel and being immediately quenched between two LN frozen metal dies. That's the way to do it correctly. The currently sold products for home shops require heat then quench then a lower temperature soak/exposure rather than an immediate quench to low temperature. It takes more time added to the product and additional expense in equipment. LN evaporates and must be replaced more frequently than oil or water. Quenching in oil might flare up a bit, water hisses, neither is nearly as exciting as LN expanding trying to absorb all that heat. I also think the idea of a vapor jacket would be even more dramatic a problem with condensed gasses like LN. The thermal shock of that large a temperature drop given all that has been said before, might leave you with a handful of fractured pieces of steel. Your original concern was to quench without so much shock.

It's a good thought problem.