O1 Sword - Heat Treating Questions

D

duurza

Joined
Jul 4, 2015
Messages
107
So, I finished profiling my O1 sword in March and haven't touched it until yesterday. I've done all of my simulations and am ready to do the heat treatment. Basically, I will be going to a blacksmith here in Toronto, Canada. I will only have access to a propane forge for heating and a metal tube (internally 3.5 inches x 1.5 inches) for quenching.

The overall blade length is 28.25 inches. The tang is 9.75 inches. Original piece was 1.5 inches x 36 inches x 3/16 inch thickness. I profiled it to a diamond cross section. The center line is still 3/16 inch thick. Current edge thickness is 1/64 inch. I am not absolutely sure but I believe the steel comes from Precision Marshall.

http://www.pmsteel.com/pdf/PRESCO-2015.pdf

My plan for the heat treatment will be a bit different. I know that O1 is an oil hardening piece; however, I do not wish to purchase that much oil nor am I willing to risk a fire indoors. I've written several simulation programs that estimate the heat transfer during the quench process and I need somewhere around 500 [W/m^2 K] in order to just pass the pearlite nose. I am using this diagram for reference.

http://www.cashenblades.com/steel/o1.html

Based on my calculations, by quenching in boiling water (100 C), the radiative heat loss through the vapor jacket will be roughly 700 but decreases as the steel cools. I believe that that works for this purpose as I only need to bypass the pearlite nose; afterwards, I want it to cool down slowly. I am going to give this a try. The known only issue at this moment is that the heat transfer coefficient spikes to 60000 when the temperature difference between the steel and water (at 100 C) is 30 C. This is a little bit frightening as that is really high but intuition tells me that this spike should not be an issue because it goes through it so fast and immediately after, there is the protective vapor jacket (water hardening steels survive non boiling water quenches). I am referencing

https://en.wikipedia.org/wiki/Nucleate_boiling#/media/File:Boiling_Curve.jpg

Basically, the point on the line divided by the temperature difference on the X-Axis will yield the heat transfer coefficient.

I will first fire up the forge without the piece inside. As I see that the lining comes to the color of around 1500 F, I will turn off the flames and place the piece in there. As the sides cool down I will turn the heat back up and move the piece in and out to ensure as even a heat as possible. When it is around the desired color, I will hold for around 5 minutes. I will quench into boiling water for around 10 seconds at which point, the temperature of the entire piece should have evenly dropped to around 400 C. From that point, I will remove the piece from the water and hold it horizontally for 60 seconds. Radiation and natural convection should reduce the temperature down to 280 C. At that point, I can do one of two things:
A) hold it vertically until I can touch it with my fingers (vertical orientation reduces heat transfer). This would effectively be like a marquench.
B) wrap it with ceramic wool and wait until the entire piece cools. I should get a mixed structure with hopefully some bainite.

I am thinking of tempering this on my stove top element until I get a blue color or even higher. I'm less worried about overall hardness as this is more of a project and wall hanger.

My questions right now are these:

1) The profiling of the blade I did by hand using a file. The tang I had to forge out because it wasn't long enough. Do I need to normalize the piece? As I understand it, normalizing is only required if the piece is forged so that all of the crystals can be similar in size and ideally smaller than when they grew during the forging process.

2) The blade of the sword has not been exposed to heat. It is in the same condition as when I purchased it and it was really soft to work with. Does that mean that the carbides are not properly distributed and that I need to do something to evenly spread it out? If so, what are the steps?

3) As stated above, the edge thickness right now is 1/64 inches (0.015625 inches). Is this too thin already? If it is too thick, what should I aim for?

Here are some pictures of it. I will probably do this earliest at the end of this month or closer to Christmas. This is more of a science project so I want to experience the process myself. It will also not entirely be a functional piece so even if I can't get everything correct, I am ok with that.

Comments and suggestions would be nice.









 
Just to quickly give some answers to your questions:

1. No need to normalize O1 if all you did was stock removal. It comes fine spheroidized, ready to machine, ready to harden.
2. Carbides are already distributed evenly when the O1 was rolled out. Nothing to worry about here either.
3. 0.015" is about the minimum, and I mean minimum, thickness for an oil quench. Prefer 0.020" and even better 0.030". However, the stock is thick and isn't tall, being a sword, so you probably will get by OK without worries.

Also, I strongly suggest you ditch the H2O and oil quench O1. Use canola oil warmed to 130°F if you don't have access to commercial quenchant.

Good luck!
 
Thanks. What about the handle though? About an inch from the shoulder to the pommel end, it was lengthened by forging. Are large grains detrimental only in situations where impact will be encountered? Also, I'm pretty sure I need to hot punch some holes in the handle. After forging, it air hardened to the point where my drill bits just won't penetrate and I need those holes to secure the tang, the pommel and also give myself some way of easily holding the blade with tongs during the quench.
 
I don't think having large grain form the forging in your handle will be detrimental. Maybe someone else with more experience needs to chime in on that, tho. I wouldn't think that would be an issue, but not sure.

For holes in the handle that is somewhat hard, you can use solid carbide bits, you can use the cheapo carbide tipped masonry bit (will be slightly oversized, go full throttle, as much heat as possible, no coolant, idea is to spot anneal), or you can cycle the handle at a dull red heat a few times (around 1300f) to try to soften it up.
 
I will need to see if I can get access to such equipment. I definitely won't be able to heat to a dull red and be able to anneal it like that. I only have access to a propane forge. There's no temperature sensor on that. The blacksmith is an artisan blacksmith so he doesn't really have a requirement for hardening steels.
 
duurza said:
I will need to see if I can get access to such equipment. I definitely won't be able to heat to a dull red and be able to anneal it like that. I only have access to a propane forge. There's no temperature sensor on that. The blacksmith is an artisan blacksmith so he doesn't really have a requirement for hardening steels.
Click to expand...

In your other thread, I posted a video of decalescence. Heat it to dull red, but don't go hot enough for decalescence.
 
So you have the ability to heat it to dull red to check deca/reca....but you can't heat it to dull red to anneal the tang?
 
Your forge is only 5" deep?
If so I wish you good luck heat treating a blade that long. Last time I used a forge to heat a sword (32" arming sword) It took awhile and my forge is like 16" long. I clamped the forge in a post vise and ran the sword verticle in and out. It was a lot of work chasing an even 1500° color. This heat treat was one of the big driving factors in building my 32" long heat treat oven. I put to much time into my knives to wory about a forge heat treat. I'm not saying a forge is not good for heat treat as I have done TONS of knives and heat treated in the forge, but there is ALLWAYS that what if in the back of my mind.
 
JTknives said:
Your forge is only 5" deep?
If so I wish you good luck heat treating a blade that long. Last time I used a forge to heat a sword (32" arming sword) It took awhile and my forge is like 16" long. I clamped the forge in a post vise and ran the sword verticle in and out. It was a lot of work chasing an even 1500° color. This heat treat was one of the big driving factors in building my 32" long heat treat oven. I put to much time into my knives to wory about a forge heat treat. I'm not saying a forge is not good for heat treat as I have done TONS of knives and heat treated in the forge, but there is ALLWAYS that what if in the back of my mind.
Click to expand...

I'm not using that. I am going to a blacksmith who has a 48 inch forge. I'll just practice watching the colors with scrap steel with my homemade one.
 
samuraistuart said:
Move the tang in and out of the forge as it heats up.
Click to expand...

I can't. It's physically like a bucket.... There's no opening on the other end. I made it to heat treat smaller knives and at the time, I never thought of tackling such a large project.
 
So does anyone have any opinion on whether or not I should normalize the tang?

The blade portion was done via stock removal but the tang was forged to lengthen it. I also won't have access to any good drilling equipment so the blacksmith and I will be hot punching some holes (good skill to learn I guess). I will probably need to fix up the tang just a bit while I am there so it will be heated up to forging temperatures.

So... should I normalize the tang or is that not needed?
 
I don't believe normalizing the tang is necessary either.

You don't have to fully anneal your tang to drill it. Just getting it to turn blue will allow you to drill it with HSS or Cobalt.
 
Ok thanks.

Also, I won't be able to drill it since I won't have access to a drill in time. Hot punching will have to do.
 
On a sword tang I soften it after the quench and temper. Once the hardening is done, heat the end half of the tang to red and watch the colors run down the tang. When the yellow/bronze color hits the ricasso, take the heat away and cool the ricasso with a wet rag. Let the tang cool off slowly and it should be easy to drill, tap, and shape. You can do this with your little forge, because you only need to heat the end 4-5".
 
So, I have a couple of days before I quench my sword and I thought that it might be interesting to share my simulation results in a more graphical manner. For the simulation program, I've simplified the geometry of my blade to a 28" long x 1.5" wide by 1/10" thick.

I've taken a few time-temperature points for the O1 TTT from (http://www.cashenblades.com/steel/o1.html) and overlayed the data over it.

The first image below will show three things that I've looked into:
1) Air cooling (should be more appropriately called Radiation Cooling since radiation is never weaker than natural convection at any stage of cooling)
2) Boiling Water Quenching (water is pushed to 100 °C and the steam jacket is used to provide just enough of a cooling medium)
values taken from https://en.wikipedia.org/wiki/Nucleate_boiling#/media/File:Boiling_Curve.jpg)
3) Forced air cooling with radiation

Ambient temperatures are all 0 °C. In fact, it didn't really make a difference as long as the ambient temperature was between -20 °C and 20 °C (until the temperature was below 200 °C at which transformation to Martensite would have already begun).



The above picture shows that if left uninsulated, O1 will air harden to an extent (long blue line that passes through the core of the TTT). It also shows that forced air (wind tunnel) could be a viable method to harden O1 as long as the speed is above 200 m/s. I've stopped near 320 m/s because that approaches the speed of sound.

Since I like to explore things, I am going to quench my sword with boiling water. I will take advantage of the vapor jacket as it acts as a strong enough heat transfer medium to avoid the pearlite nose but does not draw the heat out too fast that there would be a temperature gradient across the piece. I just have to ensure that I do not quench so long that the vapor jacket collapses. See the image below.



As can be seen in the image above, as long as the temperature coming out of the boiling water quench is around 425 °C or below, I should be able to avoid the curve. This would theoretically allow me to harden and minimize the stress caused by the cooling itself.

If I get my hands on some thermal insulation at the blacksmith's, I might time the air cooling phase after the boiling water quench and further slow down the cooling once the temperature is around 225 °C.

Thoughts or comments welcomed!


EDIT: The simulation is based on the millisecond. That means that it takes the current temperature, calculates all of the other variables that are temperature dependent and then calculates the next temperature after 1 ms of cooling. It then repeats this 1000 times per second of simulation. I could push this work on the microsecond but that would take nearly 4 minutes per 250 second simulation which was a little too long for my liking. Also, the temperatures weren't significantly different (no difference greater than 2 °C).
 
Last edited:
I like your Idea of quenching in steam.

Have you figured how to avoid getting in contact with the hot steam?
 
You must log in or register to reply here.
Back
Top