PDA

View Full Version : "Couldn't leave well enough alone"--- my brine quench experiment



jkf96a
10-13-2010, 09:39 PM
I've been quenching in vet grade mineral oil for nearly two years, with generally acceptable results. I don't have good metallurgical testing equipment, but I do use the blades I have kept, and use them hard. I mostly use my kiln for the heating, except on blades that are too long to fit, in which case I use my charcoal forge. I use mostly 1080, with an occasional file thrown in for fun. I generally do full quench, and I also use satanite for hamons once in a while.

I made a knife out of a file, probably the 25th file knife I've done. I clay coated, heated to 1475 with a soak of 5 min or so, and quenched in the mineral oil warmed to about 100 degrees. The blade did not get hard, and the hamon met the edge only about an inch from the tip. To me this indicated that the quench speed was too slow considering the unknown (but apparently higher) carbon content. I made some brine, using a gallon of water and a cup and a half of rock salt. Did the same heat treating procedure (clay, 1475, soak), then quenched in the brine, with the brine at about 100 degrees. I agitated during the quench, and kept the blade in until it was below the boiling point of the brine. I had predicted to my wife that if I tried this, 25% of the blades would crack. Let's just say 100% of the first one cracked, near the end of the cooling time, before I pulled it from the liquid. I have quenched (and cracked) in water before, and I could definitely tell a difference between the brine and the plain water. The bubbles were much smaller and much more uniform with the brine.

I had a 1080 blade sitting nearby that I had inadvertently "blued" the edge while grinding off the heat treat scale. It needed to be reheat treated, and the oven was hot, so I did the same heat treat as the previous knife (clay, 1475, soak, brine). This time, since the ping came late in the process on the previous blade, I did a three second in w/ agitation, then out interrupted quench. Guess what happened? Ping!

Since I had these two cracked blades, one known 1080 and the other a file with higher but unknown carbon content, I decided to break them into tiny bits to look at the grain. On the file, from tip to ricasso, the grain at the spine (under the clay) was coarse, and the grain below the clay was fine and gray. This is what I expected to see. On the 1080, the grain was not as fine as the file, but was uniform throughout. This was not what I expected to see.

Not many conclusions to be made here, other than I now have 100% failure rate on a brine quench. It is noticeably smoother than water, and noticeably faster than mineral oil. I mostly intended just to relay the experience. Anyone have suggestions for next time?

Kevin R. Cashen
10-14-2010, 08:23 AM
You are in a tough area between oil and water, there are only a couple of oils made that are fast enough to fill that gap properly. I just shot some micrographs last night that would be great for this thread but they are for an article a friend is writing so he has to get first dibs. The images are of a roughly eutectoid (around .8% C) bloomery steel that was quenched in various mediums, two of which are a very good fast oil and a 9% brine solution. If ever there was a “water quenching” steel it would be bloomery material, but a good number of the images of the brine quenched piece were of micro-fracturing and one serious macro crack. My friend tried just plain water and had macro-fracturing. The success came with a very fast quench oil in thickness under 1/8”, - good hard steel, no cracking. At ” no oil was up to the task of getting the steel out of the mid 40’s HRC, but the brine did reach 63-65 HRC at the cost of at least micro-fracturing.

I have seen many recommended methods of quenching modern steel with water and they all seem to have the same points in common which revolve around under-austentizing to avoid catastrophe. Which in light of the bloomery products behavior makes total sense, if bloomery steel can’t completely handle the stress of water in blade cross sections, modern steel with significantly more alloying (especially Mn everywhere) is really not going to like it.

The frightening part for me was not the crack I could see with the naked eye, it was the fact that the micro-fractures were only around .005” in length but were numerous; this would have left me with the impression of a very good solid blade loaded with many opportunities for surprises later on. My samples were also soaked at 1475F for 4 minutes, indicating that the window to leave enough out of solution to be safe must be even lower yet.