Reducing grain size in pattern welded damascus steel

me2 said:
Maybe, but that allows an even finer structure to start from.
Click to expand...

indeed, but i think this study is very specific about what happens with cycling and martensite and other structures has been carefully taken out of the equation.
it was a very interesting reading, and gives all the elements to explain what happened and why...and to discuss together
thank you VanZanten for sharing

cheers

Stefano
 
@Me2,

The count values are directly related to the surface area and magnification. So a higher count value does not automatically give a higher ASTM grain size number, the conversion factor has to be taken into account. It is best to look at the count values at the sample's photograph, the magnification and conversion factor are given there in the table. The ASTM grain size number was higher on the "as forged" samples, the normalizing lowered it.

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...

My teacher told me the best practice was to go for pearlite formation only. That's why I didn't want any air or plate cooling, I needed accurate temperature control with the tooling I had. So I decided to do it all inside the oven where a TC was mounted. My teacher told me it was very important to enter the pearlite nose as early in time and as low in temperature as possible. In other words, enter the pearlite nose left-under. I think a salt pot might go more in that direction. It is important to go for an isothermal situation. So I kept the samples at 550 degrees Celsius for a few minutes (can be seen in table 2; Additional pearlite formation time) to make sure the pearlite formation finishes. After that I would climp up again and make a new cycle.

Maybe, but that allows an even finer structure to start from.
Click to expand...

I didn't want to form bainite or martensite. I think forming martensite might give cracking problems. But if you can tell me advantages of going for these phases, please let me know.

@stezann,

Thank you!
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.
Click to expand...

Interesting. I suppose with carbides you are only talking about Fe3C? So you deliberately let the grains grow for a while pretty high in the austinization temp to go for a more even grain distribution? Do you perhaps have some information about this online? I think it is really interesting what you write about the ratio between grain boundaries and carbides but I would like to read some more detailed information about it.

@ mete,

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 ?
Click to expand...

I believe dealing with carbides in a martensite matrix is done by tempering? At these low temperatures I think you can't really make any changes to the grain size any more. What do you mean?

I haven't made measurements of the carbon content between the layers. It would be interesting though, but I'm pretty sure it has leveled out evenly at these thin layers, and taking all the forging time into account.
 
Last edited:
Yes the plate quench suggestion was more of a "where to go from here" suggestion, given that he found the limit with this method was a 9. Perhaps a faster cooling rate would allow finer starting structures, and induce finer followup structures.

Cracking is certainly an issue. The hardenability of these 2 steels is pretty different, with O2 being higher, so you might get a significant portion of martensite with a plate quench in the O2. I doubt you'd get very much in the other steel. It's mainly a suggestion as to where to go from here.

The advantage is an extrapolation of the statement that finer original pearlite grains produce finer austenite grains upon transformation. Both bainite and martensite can be finer still, giving an advantage in starting point. Martensite does have that nasty quench cracking issue though.

A couple more questions, and these are more far reaching than before, and can be considered for entertainment purposes only. I don't know how much more research you intend to do along the lines of this paper.

It is interesting that the forging produced a finer grain size, and is sufficient to reduce the grain growth from welding. It would be interesting to know how consistent this is from one billet to the next. Given your use of a forging press, I'm speculating you could be quite consistent.

Given a good level of consistency, could the forging process be refined to give smaller grains size rather than relying on normalizing? The problem with these 2 questions is you have to make quite a few billets, which is very labor intensive and expensive.

The normalized grain size is a little larger, but consistent from steel to steel, and still quite fine. One question that comes to mind: does having a more consistent grain size between the 2 steels offer an advantage over the finer, but inconsistent, size found as forged? This could be anything from more consistent hardenability to better edge holding.

The next question is a general question for all knives. The point of this paper was to pursue a finer grain size in the quest for better edge holding. A finer grains size after normalizing was not found, but that does not change the original intent of the paper. Does the difference in grain size from very fine to ultra-fine show measurable difference in edge retention? I have used knives with grain sizes off the ASTM chart (much smaller than a ASTM size of 15). These did not really seem to be any better than knives with more conventional grain sizes, though my use of them was very limited and other factors were in the way.
 
Last edited:
Van Zanten said:
@Me2,

The count values are directly related to the surface area and magnification. So a higher count value does not automatically give a higher ASTM grain size number, the conversion factor has to be taken into account. It is best to look at the count values at the sample's photograph, the magnification and conversion factor are given there in the table. The ASTM grain size number was higher on the "as forged" samples, the normalizing lowered it.



My teacher told me the best practice was to go for pearlite formation only. That's why I didn't want any air or plate cooling, I needed accurate temperature control with the tooling I had. So I decided to do it all inside the oven where a TC was mounted. My teacher told me it was very important to enter the pearlite nose as early in time and as low in temperature as possible. In other words, enter the pearlite nose left-under. I think a salt pot might go more in that direction. It is important to go for an isothermal situation. So I kept the samples at 550 degrees Celsius for a few minutes (can be seen in table 2; Additional pearlite formation time) to make sure the pearlite formation finishes. After that I would climp up again and make a new cycle.



I didn't want to form bainite or martensite. I think forming martensite might give cracking problems. But if you can tell me advantages of going for these phases, please let me know.

@stezann,

Thank you!


Interesting. I suppose with carbides you are only talking about Fe3C? So you deliberately let the grains grow for a while pretty high in the austinization temp to go for a more even grain distribution? Do you perhaps have some information about this online? I think it is really interesting what you write about the ratio between grain boundaries and carbides but I would like to read some more detailed information about it.

@ mete,



I believe dealing with carbides in a martensite matrix is done by tempering? At these low temperatures I think you can't really make any changes to the grain size any more. What do you mean?

I haven't made measurements of the carbon content between the layers. It would be interesting though, but I'm pretty sure it has leveled out evenly at these thin layers, and taking all the forging time into account.
Click to expand...

This is what I was referring to. My discussions were specific to 52100, and other hypereuctoid steels. The idea is to get even carbon and carbide distribution, even though it results in an increased grain size. The two decreasing temp cycles grow new grains within the existing grain boundaries. The final austentizing temp requires some experimenting, as you only want to get 0.8 to 0.85% carbon into solution to prevent retained austentite. The remaining carbon is free to form carbides. You should get a mix of plate and lathe martensite and minimal micro fracturing in the martensite.
 
Let me start again and clarify things.
First it's best for damascus to have two steels that have very similar HT characteristics. Our popular 1084/15N20 is a good choice.
For best properties we want fine grain and fine carbides . To get that we deal with carbides first. Get high enough austentizing temperature totally dissolve carbides. [Stong carbide formers like Vanadium in high amounts present some problems here. ] Quench this to provide the many nucleations points of martensite and keep the hypereutectoid carbides small and away from grain boundaries. Always temper this martensite .This will add more fine carbides .
Now that the carbides are formed we can proceed to grain size management. The large austenite grain boundaries will still influence things and provide nucleation points as will the martensite . The repeated normalizing , austenite to below transformation temp.This done usually three times at a lower temperature each time.
The results then in small carbide size and small grain size --are just what we want .
That should be a little more clear.
 
mete said:
Let me start again and clarify things.
First it's best for damascus to have two steels that have very similar HT characteristics. Our popular 1084/15N20 is a good choice.
For best properties we want fine grain and fine carbides . To get that we deal with carbides first. Get high enough austentizing temperature totally dissolve carbides. [Stong carbide formers like Vanadium in high amounts present some problems here. ] Quench this to provide the many nucleations points of martensite and keep the hypereutectoid carbides small and away from grain boundaries. Always temper this martensite .This will add more fine carbides .
Now that the carbides are formed we can proceed to grain size management. The large austenite grain boundaries will still influence things and provide nucleation points as will the martensite . The repeated normalizing , austenite to below transformation temp.This done usually three times at a lower temperature each time.
The results then in small carbide size and small grain size --are just what we want .
That should be a little more clear.
Click to expand...

Especially with hypereutectoids, i don't think quenching directly from the high austenitizing temperature necessary to dissolve all the carbides would be a good idea, because it likely leads to micro and macro cracking with big plates of martensite impinging each others and a whole lot of retained austenite; though quenching from a lower heat in the last cycling steps surely would provide a lot of nucleation points and finer structure, desired in prevision of the final austenitizing before final quenching.
 
OT, Stefano in the WRC Rally Sardegna do cars drive across water ? Very interesting quenching process going on. You can vary speed through the water ,original temperature of the disc and those determine whether or not you crack the discs ! Pazzo LOL.
 
I was thinking more of quenching in use .If you're not a rally person perhaps you haven't seen it. I don't have photos of that. Maybe I should give lessons on HT to the drivers .
I do radio work for rallys here. That document is funny -I noticed three different attempts to write the greek letter delta with a standard typewriter !! It'll buff out , rally on !!
 
i have seen discs glowing in racing application, point is the alloy used in those applications has very low carbon, if they are not ceramic/carbon disc nevertheless.
In normal use we don't austenitize disc during breaking. ;)
i try to avoid quenching a blade from 1600 °F when normalizing :) I have never seen suggested to do it.
On the contrary it is a common pratice to throw a quench in beetween the grain refining steps at lower austenitizing heats, to further refine the crystal structure and prepare for the subcritical annealing. Am i missing something?
Cheers

Stefano
 
@me2,

Given a good level of consistency, could the forging process be refined to give smaller grains size rather than relying on normalizing? The problem with these 2 questions is you have to make quite a few billets, which is very labor intensive and expensive.
Click to expand...

Well I think you could rely on the reduction of the press alone to get very fine grains. But, with damascus, there is a lot of "staring to the steel, thinking what to do" etc. At least in my case. I always try to make a new pattern. If you would just make the same random billet over and over again, the forging vs welding ratio would be better. Please bare in mind I always take to welding heats. For simpler patterns one should be sufficient and that would also contribute to finer grains.

The normalized grain size is a little larger, but consistent from steel to steel, and still quite fine. One question that comes to mind: does having a more consistent grain size between the 2 steels offer an advantage over the finer, but inconsistent, size found as forged? This could be anything from more consistent hardenability to better edge holding.
Click to expand...

I wonder... I think it is better to have one finer layer combined with one less fine layer. It might play a role for the micro serration at the edge.

Does the difference in grain size from very fine to ultra-fine show measurable difference in edge retention?
Click to expand...
Well it should. An edge is less prone to plastic deformation or chipping when it has higher toughness. (Hall-Petch strengthening)

Where did you get that extremely fine grained knife from?

On the other side, I would still put a big amount of work in a knife to make a very little technical advantage. All the little things together create something very special.
 
The steel was friction forged D2 from Diamond Blade knives. There is a point of diminishing returns with grain refinement, but most knife makers never get close. Finer is better, but beyond a certain point, it seems like bragging rights rather than real improvement.
 
Very astute point.
Similarly, with retained austenite and multiple tempering cycles, there is a point where it is only discernible in a laboratory. The real world stops seeing any change after two tempers.
Grain refinement has theoretical limits that are far smaller than practical knife use needs.
 
also, when making knives, it should be considered that reducing too much the grain would provide a whole lot of grain boundaries and nodes which act as footholds for the pearlite reaction to start; That is the reason above the reduction of the hardenability related to the super fine grain size.
It would so shift the "nose" to the left and increase the risk of getting a mixed structure in the end product. That would provide more toughness at the expense of hardness and strenght
 
also, when making knives, it should be considered that reducing too much the grain would provide a whole lot of grain boundaries and nodes which act as footholds for the pearlite reaction to start; That is the reason above the reduction of the hardenability related to the super fine grain size.
It would so shift the "nose" to the left and increase the risk of getting a mixed structure in the end product. That would provide more toughness at the expense of hardness and strenght
Click to expand...

+1

I wonder at what ASTM grain size number the steel should be that it won't hit 64HRc anymore after quenching. (O2 for example)
 
Verhoeven addresses this in his book you reference in the paper. It is highly alloy dependent. The effect of grain size was still noticable in 5160, reduced from somewhere in the fine to very fine down to the ultra fine range. I don't remember the numbers exactly. The effect was GREATLY reduced, but was still measurable. In something like O2, it should be even less noticable, maybe to the point of not being a practical concern. Then again, if you can consistently get down to the 14 to 15 range of grain size, you might have to change to faster quenchants. Its an interesting question, and I think there is research on it out there, but the question becomes does it exist for your steel, is your steel thick enough for it to be an issue, and are you processing it to get the size that small? If I remember correctly the 1080ish steel Verhoeven used had the hardenability reduced so much that there was a fully pearlitic spine and a martensitic edge on a 1 inch wide, ¼ inch thick full flat ground blade, divided nearly down the center.
 
aha.. interesting. Fortunately my blades are very thin. We'll see how it turns out after the salt pots have been built.
 
Just wanted to say, thanks for going the distance guys- you've provided some very interesting reading material here, a cut above the usual everyday shoptalk fluff. My choice to go the salt pot route with my next shop improvements has only been reinforced.
 
You must log in or register to reply here.
Back
Top