? For Kevin Cashen

Mark Williams

Mark Williams

Moderator
Joined
Nov 28, 2000
Messages
4,760
I cant remember the thread right off hand, but I distictly remember you stating something to the effect that grain size being too small is not a good thing. I cant remember why that is. Could you elaborate on this please?
 
In simple steels, that do not have the benefit of alloying elements which can promote deeper hardening, the smaller the grain size the more shallow hardening the steel may become. I am not sure if you want the full technical explanation as to why this is or not, so I will give you this simple version for now, so as not to bore folks ;)
 
Bore us a little Kevin. I'm unclear as to what grain we are looking for in.....let's say the King of steel (5160) :confused:
 
And when we get to the next level down ,how we measure the size of the crystals , it's done with X-ray diffraction techniques . No I'm not going to explain further because I don't remember !....Many things involve the grain boundaries ,that's where diffusion changes occur first . The more grain boundaries the more places to start transformations .And Kevin can put that all into understandable english .
 
Oh Boy! Kevin's gettin ready to explain some techno steel stuff! I'm getting some popcorn ready. :)
 
Diffusion based processes (those resulting from the movement of carbon atoms) are driven by rates of nucleation (big word simple meaning- a place for the process to get a toe-hold and start) the more points of nucleation that greater the rate of diffusion, or at least the greater the amount of it going on. Pearlite is a product of diffusion. It needs points of nucleation to get a start. As mete mentioned (and then left me holding the bag ;)) the grain boundaries are points of high energy where nucleation can readily occur. Finer grains mean more grains, more grains mean more grain boundaries to separate them, this all equates to more areas for pearlite to grow. More areas for pearlite to nucleate will naturally push the nose on the TTT curve to the left, decreasing the time required to cool the steel below the point at which pearlite forms, lessening the chances of reaching maximum hardness.

Try this sometime, get a piece of 10xx (1080, 1084, 1095) that has good bevel ground on it and quench (full quench) it in a good oil, and then polish to see the temper line and where it is at. Normalize (or cycle) the blade a bit and then quench and notice where the temper (harden) line is now. If you are really good at temp. control and keep the heats in the proper range to refine keep refining those grains, you will be able to watch that martensite line move closer and closer to the edge, as the grain gets finer.

Fine grain is good for toughness to be sure, but everything in this business seems to be a compromise and a balance. Large grains promote deeper hardening but are not good for toughness for a couple of reasons. First, and simplest, fractures really like to travel along grain boundaries, the larger the grain, the straighter the course for the crack to take. Then you have embrittling precipitates in the grain boundaries that tend to be bolder in course grained structures. Finally, if you have large grains above .60%C the resulting martensite plates will be huge and will run into each other at high angles (like big ice sheets slamming into each other), creating really ugly micro fracturing that doesn't do a thing for the toughness of the steel :( I know folks who have been involved with research involving the cutting ability if extremely fine grains (beyond what is typically achieved) and their results were disappointing. It would appear that edges could need a balance as well, although I would like to do some studies myself to see what could be going on there.

As for the desired grain size, that is hard to nail dead on without the equipment. I hope to have some appropriate microscope eyepieces for such work soon, but even then I won’t be doing too much since it is a bit of work to prepare the specimens to properly rate grain size. One needs to prep and polish the specimen and then etch it correctly to get what you need under the microscope, and then you need to count how many grains occupy a given space under magnification. I believe it is around the average number of grains per square inch at 100X magnification. The ASTM chart typically ranges from 1 to 10 or so, in grain sizes with the higher number being finer grains. Mete will have to tell you anything about x-ray defraction since that is well beyond my means.

For most smiths the most common option is to break a treated piece and look at the fractured grain size. The rule of thumb for this could be – if you can make out the individual grains well with the naked eye, they are probably too big :eek: Fractured grain specimens are all but useless under a good microscope since the uneven surfaces make focusing nearly impossible.

Peter, I have limited experienced based information to offer on 5160, in my shop it is more like duke or earl, with other steels firmly reigning as kings or emperors, and the last lone bar wandered out of my shop about 5 years ago. It is, however no mystery that the steel is as popular with smiths as it is; it is very hard to make a bad blade out of it. Although the .6%C (often actually .55%) will limit the maximum hardness, due to the leftover ferrite, it will also keep the martensite entirely of the lathe variety (instead of plate). This makes the steel tougher by nature, and will even allow a few “oops” in the overheating and still not be as brittle as expected with a larger grain structure. The chromium allows it to harden deeper than one could expect from other steels of the same carbon level, like 1050 or 1060. All of this adds up to a steel that is very easy to forge, and somewhat self correcting as far as grain refinement (due to the chromium), it will harden up nicely with less chance of distortion from radical quenches, and will be pretty tough, even if the temps weren’t dead on. I have mostly fond memories of working the steel, (before I started mixing 2 steels in one blade), until the current reigning monarch of large blade steel in my shop, deposed it, and banished it from the kingdom. ;)
 
Excellant Kevin. Very understandable. The experiment you mensioned explains what I have see in some of my multiple quenched 1084 blades.
 
that does exsplain a few things :D I got to poke that in knife site somewhere so you won't have to wright it all out again when the thread gets lost :D ,
 
I'm back, got distracted by Simona !! Kevin you did fine . There is a direct relationship between microscopic grain size and fracture grain size.But perhaps the best thing to do with grain size is to have a way to compare your heat treating techniques and resulting grain size . Do some experimenting ,learn how to relate heat treatment and grain size and corelate with other properties .
 
That was very understandable Kevin. :cool: Thanks !
If you keep working at it, you may even get the hang of 5160 and quit playing with those fishing lure blanks :footinmou :footinmou :footinmou :footinmou :footinmou :footinmou :footinmou ;)
 
peter nap said:
If you keep working at it, you may even get the hang of 5160 and quit playing with those fishing lure blanks :footinmou :footinmou :footinmou ;)
Click to expand...

Whaddya mean, Don? Kevin forging stainless or something??? ;) ;) :footinmou :footinmou :footinmou
 
peter nap said:
That was very understandable Kevin. :cool: Thanks !
If you keep working at it, you may even get the hang of 5160 and quit playing with those fishing lure blanks :footinmou :footinmou :footinmou :footinmou :footinmou :footinmou :footinmou ;)
Click to expand...


Heheh :D What are fishing lures made of anyhow? I think some of the ones I have seen are some sort of zink or die cast metal.
 
Keven or anyone with a answer. While we are talking about grain size, I have a question on normalizing times. Other makers have mentioned that thy get better performance by normalizing 24hr apart that means that it takes 3 days to normalize a blade 3 times. Is there anything to this? the steel in question is 5160. Also how cool dose a blade need to get to before reheating. Gib
 
Let me see if I have this right. Some steels will not actually benefit from multiple heating and quench cycles, due to the grain size being reduced too far/small??

:eek:
 
You guys went above and beyond me from the get go but I do know a little bit about X-ray diffraction from a couple classes at school. so I'll try to explain how it works and maybe that will help your discussion? Or at least prove that I stayed awake for one chemistry class and afew physics lectures ;)

Everyone seems to understand the idea of the grain stucture. It has a basic shape, sort of a lattice or matrix. Depending on the material the individual parts are hooked together different ways, some are face centered, some are edge centered (there are more but I don't remember them). X-ray diffraction allows you to study this matrix,the size of its individual peices and how they fit together.
In redneck terms, its like spraying a garden hose at a porch screen and studying the pattern of the water that makes it through.

In more detail, when your dealing with atomic structure, you have very dense areas(nucleus) and relatively open areas around them. If you have a thin enough sample of your material, you can shoot an X-ray at the sample and it will be able to pass through the open areas and will be deflected by the dense areas. There is a very fine relationship between the wavelength of the X-ray and the material it can pass through.
The open areas are very small and act like thin slits, which cause diffraction. Basically the slit becomes the center of a new wave coming out the other side. Because you have a bunch of these little slits together you get double slit interference.As the waves come out the other side, they can have positive and negative intereference on each other. A wave has a high spot and a low spot, if 2 high spots meet, the new wave is stronger, if a high and low spot meet, they cancel out, etc. The pattern of intereference leaves a very distinct pattern of light (x-ray) left as there will be very intense spots from the positive intereference and dark spots from the negative interference. Smart folks well trained in physics, and metallurgy can use the laws of diffraction to study this pattern and use it to calculate the size and spacing of the slits/open spaces which then tells them how the matrix is put together and how big its peices are.
This takes a lot of physics as well as chemistry/metallurgy.
I'm sure I've made a few mistakes here, but its been over a year since I took any chemistry or physics classes, and I really didn't like the stuff that well to begin with :D Hopefully you get an idea of how the process works now though.
 
Matt, I took an x-ray crystallography course as an undergraduate - the course was a graduate physics course ! Well anyway for practical use we can for example see exactly what happens when we temper steel. The martensite crystal is an elongated cube with carbon atoms stuffed into the edge of the crystal . As we temper at higher and higher temperatures ,more of the carbon atoms come out of the crystal and the length of the crystal gets shorter and shorter , finally becoming a cube ,ferrite,without any carbon......Gib, since you are normalizing ,that is air cooling,it doesn't take much time for the blade to reach room temperature. Nothing else will happen to that steel in a hour or a day. The only change will occur when you reheat it .In fact most of the changes in normalizing occur as it goes from austenite to ferrite above 1000F. Three days ? More myth.
 
Gib Guignard said:
Keven or anyone with a answer. While we are talking about grain size, I have a question on normalizing times. Other makers have mentioned that thy get better performance by normalizing 24hr apart that means that it takes 3 days to normalize a blade 3 times. Is there anything to this? the steel in question is 5160. Also how cool dose a blade need to get to before reheating. Gib
Click to expand...

Mete has addressed this quite well so all I can say will only add and reinforce his post. The main idea around normalizing is to equalize and refine austenite grains, with perhaps some evening of the carbide dispersement. You heat to a given temperature above Ac1, the point at which new grains will begin to form (industry often goes above Acm/A3, which is even higher), and then allow the steel to cool as evenly as possible (air) below the point at which austenite will transform. If you wish to reduce the grain size, subsequent cycles will use the framework of the previously formed grains to increase the points of nucleation (there's that word again ;)) for the new grains. Just as with pearlite, increased points of nucleation equals more newly formed grains and more grains in the same area=smaller grains.

Once again we have to have tradeoffs and compromises, however. The finer those grains become, the quicker they will start to grow, so you will have to pay attention to your subsequent temperatures and back off a bit more every time in order to get the next smaller grain size, eventually you will reach a point where further normalizing will just be a waste of time. Of course, with all of this you will also find that you may have to move to lighter oil or perhaps water based quenchants to get the same depth of hardening.

Now where was I going with all of this? Oh yeah! 24 hours between normalizations would do nothing but waste a lot of time, but then if you price your knives by the hour :rolleyes: . All you have to do is transform to austenite and then let it transform back again, this should take no more than five minutes. The boundaries of the austenite grains are set as soon as the parent austenite transforms into a more stable structure which may have its own grain structure. If that structure is pearlite, it will be done before the steel reaches 800F. and it will be about as stable as it gets. If somebody feels they are getting better grain refinement by waiting longer than what it takes to get to room temperature this is one of those claims that just defies all of the principles of ferrous metallurgy, and I think they would need to explain it quite a bit better than a hunch or a belief.
 
ragnoor said:
Let me see if I have this right. Some steels will not actually benefit from multiple heating and quench cycles, due to the grain size being reduced too far/small??

:eek:
Click to expand...

For political correctness, we should probably stick to normalizing. Almost any steel that has undergone uncertain heating operations, such as forging, will benefit from normalizing. Regardless of grain size, grain uniformity is also something to go for. Small grains are great, but there is a balance in all things and we must always remember that the universe insists on checks and balances. There will be a point at which it will be increasingly difficult to get the grain finer, and that finer grain will need a little more cooling power to harden. Of course this does indeed change from steel to steel and many substituational alloy atoms will help to easily overcome this. Those same elements can retard grain growth by sitting in the grain boundaries and stabalizing them. So, while I would say that any steel could benefit from proper normalizing, some will respond to it more profoundly than others and some will need some new considerations in the heat treatment.
 
You must log in or register to reply here.
Back
Top