Reducing grain size in pattern welded damascus steel

So with that, does that mean you recommend doing all your nomalizing cycles in a salt pot? The 7 cycles is more than double what most people recommend, not that I want to play with a salt pot right now anyway but it is something to think about.
 
Thank you Stacy

@tattooedfreak,

About the seven cycles, I wanted to see what the limits of my setup were, would the refining stop at a specific grain size or would it just keep on refining. It turns out the cooling and heating speeds are key to finer grains. Salt pots are real good at heating and cooling fast. They are not necessary, but if you want to go for ultra fine grains, I think it is the way to go. I haven't done it yet, first I need to build myself two saltpots ;-)
 
Thank you for this. This is my favourite part of knife making.
 
I agree that a high temp and low tem salt pot setup would be superb for maximum blade quality. It will not make a poor blade good, but it might make a good blade better. How much better is more of a laboratory thing than a practical thing, but I am sure you are right that it would improve the overall quality and uniformity of HT.
 
I agree on the saltpot, though i have a question about the cooling performed during the experiment; instead of the air tubing/nozzles method, simply extracting the rack from the oven for a still air cooling seems to me a quicker way and maybe would have yielded a higher astm number.
I don't know, probably has been done that way due to the tiny volume/section of the samples, in order to avoid martensite formation, i wonder.
I appreciated a lot reading, thank you! :thumbup:

Stefano
 
You are right, it was to avoid bainite or martensite formation. By keeping it in the oven I could leave it there at 550 degrees Celsius, until pearlite finish.
 
Van Zanten said:
You are right, it was to avoid bainite or martensite formation. By keeping it in the oven I could leave it there at 550 degrees Celsius, until pearlite finish.
Click to expand...

thank you,
did you try "classical" air cooling with bigger samples, and compare the grain size to the experiment's samples?
Are you sure about the actual temperature involved, and accuracy? I noticed the samples were exposed and close to the bare coils.
I have no direct references on astm numbers but visual "feel" and comparison with the ol' snapped file ;) in your opinion did the samples "look" fine and did you visually appreciated the differences between as forged and normalized?
 
Thanks. I enjoyed the paper. I have a couple of questions. How many forging cycles did your billet go through? For someone forging a knife by hand the blade would undergo multiple cycles. How did the grain size vs normalization cycles compare to each other?
 
@stezann,

thank you,
did you try "classical" air cooling with bigger samples, and compare the grain size to the experiment's samples?
Click to expand...

No I haven't done anything with classic air cooling. The whole scope of this research was about the AC1 - AC3 cycling and how to perform that.

Are you sure about the actual temperature involved, and accuracy? I noticed the samples were exposed and close to the bare coils.
Click to expand...

That is a really interesting thing to think about. The way I see it is; the size of the thermocouple is about the same size of the samples. Also, the distance from the coils to the thermocouple is almost equal. So I don't think there has been a temperature overshoot. I'm not 100% sure though.

I have no direct references on astm numbers but visual "feel" and comparison with the ol' snapped file in your opinion did the samples "look" fine and did you visually appreciated the differences between as forged and normalized?
Click to expand...

Well I did some breaking tests, but it is really hard to tell something accurate from the grain you can see. I can see the difference in ASTM grain size 1 and 8, but from 8 on and higher I don't think one can do it. I did test my setup by putting a piece of O2 for 10 minutes at welding heat in a coke fire. I was very very coarse and the specimen broke very easily by touching it with a hammer in a vise. I normalized this piece with my setup and tried to break it again, the second time I had to beat like crazy with a big hammer, it bended 90 degrees and finally broke. That was a really special thing to do and convinced me the setup was working. Pictures about the coarse and the finer grain are here (both the same sample):

p6pl.jpg


3mlv.jpg


@ Bo T:

Thank you. The steel used was from this ladder patterned knife :

koksmes_aug13_a.jpg


So there was a lot of heating and cooling. 240 layers. It took me a real long time to get all the groves in the ladder pattern. The cooling in the press might also have refined the grain. ASTM grain size 12,5 is pretty high..

What do you mean by a grain size cycle and a normalization cycle? I refer to normalizing as heating and cooling through the transformation temperatures.
 
Normalizing is often referring to the 1650f heat to get everything into solution, and the appropriate cooling rate for the steel. Grain refinement cycles refer to the heating above critical to form new grain boundaries, such as 1550f, cool to magnetic, then 1450f and cool to magnetic. Many do a 1200f brief anneal at the end of this. Speaking with Kevin Cashen this past weekend, he suggested not going to below 1450f, but the cycle started with a 1200f subcritical anneal. I have a few questions in an email regarding cycling to him and will open this for discussion when I hear back.

BTW, great knife!
 
You had several samples. Sample 1 went through 1 normalization, sample 2 went through 2 normalizations, etc. Did you measure the grain size for samples 1 - 4? Also, You reduced the thickness of your billet by using a press(?) so as you noted the reduced grain size of 12.5 might be due to dynamic crystallization?
 
I knew something like this was coming from you after our Elmax discussion. Very refreshing work. I have a couple questions, but they'll have to wait 'til after work.
 
@Willie71,

First I have to convert this to Celsius, I'm very bad at imperial, inches and Fahrenheit ;-): (I really appreciate your input, but if I may, I would like to encourage you to speak in terms of pearlite, austenite etc instead of temperatures, the normalizing mechanisms etc are pretty equal to most steels.)

Normalizing is often referring to the 898 C heat to get everything into solution, and the appropriate cooling rate for the steel. Grain refinement cycles refer to the heating above critical to form new grain boundaries, such as 843 C, cool to magnetic, then 787 C and cool to magnetic. Many do a 648 C brief anneal at the end of this. Speaking with Kevin Cashen this past weekend, he suggested not going to below 787 C , but the cycle started with a 648 C subcritical anneal. I have a few questions in an email regarding cycling to him and will open this for discussion when I hear back.

BTW, great knife!
Click to expand...

I understand what you mean. But to give you a better answer I have to know what steels we are talking about, and whether it is a monosteel or damascus. For instance, an eutectic steel (0,77%C) does not have a state in which both pearlite and austenite exist (two phases mixed), upon crossing AC1 a whole new set of grains is formed, not two sets like hypo- or hypereutectic steels do when crossing first AC1 and later AC3.

@Bo T,

I have made pictures of all the samples. I have only counted them on sample 5 and 6; ..> 7 cycles and 'as forged'. I did 3 countings per layer, so 6 countings per sample, and each counting took me 30 minutes to make the image, so I decided to skip the others. Sample 1 to 4 showed results that were in between the others. (linear growth, as seen by eye).

I wrote about the reduction of the press and dynamic recrystallization in chapter II,E.

@Me2,

Hey, it is nice to hear from you again! Glad to hear you like it! By the way, here is a picture I made from Elmax! Magnified 200x.

g_sampleE_200x.jpg
 
Last edited:
Ok, here are my questions. Keep in mind I read the paper while on lunch break on my phone and have not done the calculations or looked very closely at the diagrams/micrographs.

You have counted the grains per area before normalizing and after 7 normalizing cycles. The before normalizing counts were lower than the after 7 normalizing counts. This indicates more grains per unit area, or smaller grains. However, your calculations show a larger grain size after normalizing, size 9 after and size 12.5 or 10.5 before normalizing. You used a conversion due to difficulty in getting a full area in the micrograph. Are these differences due to the conversion, or am I missing something else?

Second, you use a system to air cool, and use some impressive measurements to give temperature ranges and cooling curves. Have you considered plate quenching either steel instead of air cooling? This would be easier than using the system and less dangerous and complicated than salt pots.
 
beautiful images!!! and amazing knife!!!
That is how i picture the process in my mind: I think normalizing refers to go to ac3 for hypereutectoids... you get all carbon into solution in austenite reaching the right temp to accomplish the task (for eutectoids is the lower possible for steel), and allow soak for even diffusion of all that carbonwhich equalizes within the matrix: the grain grows high, but they levels to the same average size and the steel is now normalized. The further cycling are done at progressively lower temperature, closer to ac1, and are meant to refine the average size of "grains"...but they are not normalization beacuse we don't touch all the carbon (except for eutectoids) but leave alone the carbide phase in the previously well distributed state, not reaching anymore the higher temperatures required to drown it into austenite solution, but "texturing" the matrix with finer crystals, thus optimizing the ratio between grain boundaries and carbides.
I'm really enjoing this discussion, it is clear that this is one of the most fascinating side of knifemaking, at least for me :)
I'm with you on the TC readings...you are right IMHO.
I'll be following with interest this great thread.
 
me2 said:
Second, you use a system to air cool, and use some impressive measurements to give temperature ranges and cooling curves. Have you considered plate quenching either steel instead of air cooling? This would be easier than using the system and less dangerous and complicated than salt pots.
Click to expand...

If he plate quenches he gets martensite....the samples are tiny
 
Pink steel ?? we don't make pink steel in the USA. More snide remarks later.

Have you made measurement of the carbon content as you go from one steel to the other ? It would be nice to see the gradient .

Normally with steel having carbides in addition to the martensite matrix you would HT to deal with carbide size first , then deal with grain size. Comment for your steels ?
 
Last edited:
You must log in or register to reply here.
Back
Top