- Joined
- Sep 9, 2003
- Messages
- 2,361
**Notice: this thread is an analysis of a test piece that was taken to the extreme, it is not a recommendation or endorsement for soaking any longer than the standard recommended times**
Phillip Patton was good enough to send his pieces of O-1, which he soaked for 5 hours to see if Kevin Cashen was just messing with folks or not
(which I really enjoyed and commended him for). I must stress once again that I am not in any way advocating soaking O1 for this outrageous amount of time but it was great to prove the point that grain growth is only an issue if you lack control of the temperature, but with that control, time becomes our friend. Since I have not microscopically examined any steel that has been cooked for this long, I jumped at the chance to compare and gather information for my research. My metallurgical polishing equipment is still limited and I have repolished the pieces a few times and have decided that we will just have to live with the scratches.
Rockwell hardness of the steel was an average of 64.5 to 65 HRC, I rebroke a piece to get some fresh ends to look at and with a notch it still took some effort with a 3lb cross peen to break it, no less than the pieces I often break after a standard soak. I have included an image of that fractured end to give an idea of the grain. It is hard to tell from a photo but this steel was very fine grained that any knifemaker should be happy to have in his blades.
Cross sections of the piece where mounted and polished beyond a 2000X finish by hand ending at 8000X. and then etched in a 5% nital solution.
I did several etches to bring out various features. A quicker and lighter etch would highlight the martensite much better and reveal superfine black structures I believe could be some sort of carbide, but it would also show the larger residual carbides with less contrast. Etching can cause large enough differences to cause one to misread the results. I had originally questioned whether the steel had fell short of complete hardness or had been tempered because of my initial hasty polishing and etching jobs. It made the martensite look more granular and globby instead of plate or needle like.
This martensite is actually very solid and plate like. Here it is at 400X after a light etch:
Here it is after etching to bring out a stronger contrast and show the residual carbide at 1000X:
And here is a comparison of Mr. Patton’s 5 hour O-1 and some of my 10 minute soaked O-1:
When one takes into consideration that any differences may be from polishing and etching as well as my messing with the scale a bit when I crop the images, the similarities become more striking. If the carbide size in the 5 hour is not from any snafu in my cropping, it could be the condition of the prior heat treatment before the soak. While there were still loads of carbides left there (making grain growth very unlikely) they were overall lower in number and more widely spaced compared to the 10 minute soak. Also if one looks at the general coarseness/fineness of the martensite it is plain to see that there is little difference between the two.
I have no way of knowing what the grain size was before the soak so to say that zero grain growth occurred would be impossible, but when one considers that these images are 4 times that magnification that grains a traditionally measured at, and still look very fine, the outcome is pretty darned good. If there was grain growth then Mr. Patton had the initial grain size incredibly small to have this be larger in comparison.
Although I , or Mr. Patton I am sure, would never soak a blade this long, I see no problems from carbide issues or grain size with this piece, but…
Now for the down side of soaking too long… DECARB. Mr. Patton said that he enclosed this piece in foil with a bit of tissue to avoid these problems and he did forge on it before hand, so I would be willing to say that much of the decarburized metal occurred in forging (definitely the complete layer of oxide did), but I have done metallography for folks who have been very careful to pack the steel in foil with carbon bearing materials, only to see those white colored flecks appear at the edges anyhow.
Here are some images of the edges and corners.
The very outer edge has heavily defined and dark boundaries between the white ferritic grains, this in my opinion is heavy oxidation, the same thing happens when steel is burned (uncontrolled, runaway oxidation) the process attacks the grain boundaries more heavily and eats them away until there is weak patchwork left with not much holding it together. As we look deeper, the white grains become more a solid field of ferrite until dissipating into the martensitic body of the steel.
Notice how the corner has a deeper zone of decarb? At a corner the penetration can combine from the two directions to accumulate, now apply this to a knife edge and ask yourself how much material you would like to remove from a forging to get to that good martensite! Most folks think 2 dimensionally about this and polish down either edge bevel and call it good, a wise smith would also leave meat to remove on the very edge even after heat treating unless he really did have a way to totally eliminate this problem.
In conclusion, I must predictably say that I found no real surprises here. It is much less of an “I told you so” than a comforting affirmation that the rules and principles of metallurgy do indeed still work and can allow us to predict outcomes in a very reliable fashion. Re-inventing the wheel can be fun but it is not very productive. Folks reading this that have tight control over their temperatures can soak away at reasonable times with little worry that they will get anything but great results from it. Folks who have less accurate equipment for heating will indeed need to watch the temperature so closely that time will make things quite difficult. With soaks less than 6-8 minutes I have seldom achieved complete martensite hardness with O-1.
Phillip Patton was good enough to send his pieces of O-1, which he soaked for 5 hours to see if Kevin Cashen was just messing with folks or not
(which I really enjoyed and commended him for). I must stress once again that I am not in any way advocating soaking O1 for this outrageous amount of time but it was great to prove the point that grain growth is only an issue if you lack control of the temperature, but with that control, time becomes our friend. Since I have not microscopically examined any steel that has been cooked for this long, I jumped at the chance to compare and gather information for my research. My metallurgical polishing equipment is still limited and I have repolished the pieces a few times and have decided that we will just have to live with the scratches.Rockwell hardness of the steel was an average of 64.5 to 65 HRC, I rebroke a piece to get some fresh ends to look at and with a notch it still took some effort with a 3lb cross peen to break it, no less than the pieces I often break after a standard soak. I have included an image of that fractured end to give an idea of the grain. It is hard to tell from a photo but this steel was very fine grained that any knifemaker should be happy to have in his blades.
Cross sections of the piece where mounted and polished beyond a 2000X finish by hand ending at 8000X. and then etched in a 5% nital solution.
I did several etches to bring out various features. A quicker and lighter etch would highlight the martensite much better and reveal superfine black structures I believe could be some sort of carbide, but it would also show the larger residual carbides with less contrast. Etching can cause large enough differences to cause one to misread the results. I had originally questioned whether the steel had fell short of complete hardness or had been tempered because of my initial hasty polishing and etching jobs. It made the martensite look more granular and globby instead of plate or needle like.
This martensite is actually very solid and plate like. Here it is at 400X after a light etch:
Here it is after etching to bring out a stronger contrast and show the residual carbide at 1000X:
And here is a comparison of Mr. Patton’s 5 hour O-1 and some of my 10 minute soaked O-1:
When one takes into consideration that any differences may be from polishing and etching as well as my messing with the scale a bit when I crop the images, the similarities become more striking. If the carbide size in the 5 hour is not from any snafu in my cropping, it could be the condition of the prior heat treatment before the soak. While there were still loads of carbides left there (making grain growth very unlikely) they were overall lower in number and more widely spaced compared to the 10 minute soak. Also if one looks at the general coarseness/fineness of the martensite it is plain to see that there is little difference between the two.
I have no way of knowing what the grain size was before the soak so to say that zero grain growth occurred would be impossible, but when one considers that these images are 4 times that magnification that grains a traditionally measured at, and still look very fine, the outcome is pretty darned good. If there was grain growth then Mr. Patton had the initial grain size incredibly small to have this be larger in comparison.
Although I , or Mr. Patton I am sure, would never soak a blade this long, I see no problems from carbide issues or grain size with this piece, but…
Now for the down side of soaking too long… DECARB. Mr. Patton said that he enclosed this piece in foil with a bit of tissue to avoid these problems and he did forge on it before hand, so I would be willing to say that much of the decarburized metal occurred in forging (definitely the complete layer of oxide did), but I have done metallography for folks who have been very careful to pack the steel in foil with carbon bearing materials, only to see those white colored flecks appear at the edges anyhow.
Here are some images of the edges and corners.
The very outer edge has heavily defined and dark boundaries between the white ferritic grains, this in my opinion is heavy oxidation, the same thing happens when steel is burned (uncontrolled, runaway oxidation) the process attacks the grain boundaries more heavily and eats them away until there is weak patchwork left with not much holding it together. As we look deeper, the white grains become more a solid field of ferrite until dissipating into the martensitic body of the steel.
Notice how the corner has a deeper zone of decarb? At a corner the penetration can combine from the two directions to accumulate, now apply this to a knife edge and ask yourself how much material you would like to remove from a forging to get to that good martensite! Most folks think 2 dimensionally about this and polish down either edge bevel and call it good, a wise smith would also leave meat to remove on the very edge even after heat treating unless he really did have a way to totally eliminate this problem.
In conclusion, I must predictably say that I found no real surprises here. It is much less of an “I told you so” than a comforting affirmation that the rules and principles of metallurgy do indeed still work and can allow us to predict outcomes in a very reliable fashion. Re-inventing the wheel can be fun but it is not very productive. Folks reading this that have tight control over their temperatures can soak away at reasonable times with little worry that they will get anything but great results from it. Folks who have less accurate equipment for heating will indeed need to watch the temperature so closely that time will make things quite difficult. With soaks less than 6-8 minutes I have seldom achieved complete martensite hardness with O-1.
Fortunately knives are very simple shapes and this becomes less critical.
Folks ask why I go so fine before heat treating, my answer- there is a huge difference in hand sanding O1 at 25-30HRC from doing it when it is 62HRC!
