Jump to content
I Forge Iron

Hardening and tempering in one step - water tub and torch method?

Recommended Posts

53 minutes ago, anvil said:

I was going to post just what Steve said. At best you are normalizing the blade. And using a torch, the blade will not have a consistent heat along it's edge.

One thing I can assure you is that it's not simply normalized. I've done that plenty of times and this is a definite hardened edge. You are correct - it isn't consistent in this case, but the examples I was shown were thicker blades and it seemed to be very even. I'm curious what would happen if the edge was tested for hardness in many places, like every 1/2 inch or so. I wonder if it would come up softer in some spots (not on my sword, it's obviously not even; on his knives).

Just to make my experience with this clear - though my post count is low, I've been treating and tempering steel for seven years using the proper methods. This is something new to me and I'm as skeptical as everyone else. All I can say is that it sure seems to have worked. I'm going to try it out on something I don't care about and do some tests to destruction and see how it compares. I'll knock up a few crummy knives and heat treat half of them traditionally and half with this method and see what happens. I'll definitely post the results of my tests.

1 hour ago, JHCC said:

So just to repeat: he’s got the bulk of the blade submerged in water except the edge, he heats the edge until it’s glowing, keeps it in the water until the edge stops glowing, and then he lets it air-cool?

I’m no metallurgist, but this seems problematic for a couple of reasons. The torch heating might well get the edge above transformation temperature (to convert the ferrite into austenite), but is the water bath enough of a heat sink to cool the edge fast enough to convert the austenite to martensite (especially in a non-air-hardening steel like 5160)? Even if so, how does that martensite then get tempered to remove its brittleness, when there’s no more heat being applied?

You are correct - that is what he's doing.

I agree with everything you just said. It seems like it shouldn't harden it, and then afterwards, there is no tempering process except what the water imparts. But that's what I'm seeing - the edge is file-hard and isn't chipping when I strike hardwood, which I have seen happen with one of my blades that I didn't temper properly.

1 hour ago, Steve Sells said:

without the quench its mostly just normalizing which leaves a tough blade,  but as for an edge, well  I can put an edge on an Aluminum can quite easily but it wont last either

I've put an edge on aluminum! Back when I was a kid I had a mail-order "ninja" sword (those are sarcastic air quotes). I couldn't understand why it dulled the instant I tried to cut with it. And why it bent sideways when I hit a post with it. Ah, memories. :D

What I'm seeing here looks and feels like a hardened and treated edge. It doesn't add up, I agree. Like I said in my other reply, I'm going to do some testing and see what happens.


Thanks for your feedback on this, everyone. I really appreciate the knowledge you guys have on these matters, and you could have just as easily ignored my original post as a "newb" question. 

Link to comment
Share on other sites

Like everyone else I'm skeptical, but I try not to be dismissive.   Since you're willing to spend the time and energy with further testing I for one am very interested in your results in a side by side comparison of the same steel using this method compared to the prescribed heat treating process for the steel.


Link to comment
Share on other sites

  • 3 weeks later...

Greetings again, everyone. I'm happy to report that I finished forging and tempering my test blades. I'm sharpening now and will start to conduct my tests this week!

Initial results seem promising. Both sets of knives are properly hard and durable, but the testing will yield real data. I'll have more pictures and also some video of the vise bending test. Should be interesting.

Link to comment
Share on other sites

Interested in seeing your results.  Assume you only oven tempered the blades that were heat treated using the accepted, conventional heat treatment method, not the water tank technique you described above.  Please be clear on what specific steel was used (and how you know it is actually that steel type), thickness at the spine and edge during heat treatment, any normalizing or annealing done to blades, and whether they were forged or just stock removal (hopefully the latter, as that would put less variables into the test).

How are you planning on testing the hardness?  Do you have calibrated hardness files, chisels or a Rockwell tester?  Note that I believe that the last, and most accurate, test method needs to be done on a flat, even thickness, section of the blade (ricasso) to be truly accurate. 

How will you test the durability?  Will you do a brass rod test, chopping test into hardwood, extended cardboard or rope slicing test, or just a break test to examine the grain?

Are you going to be rigorous about blind testing the samples (hiding identifying markers for which batch is which during testing)?

Even if this technique for hardening and tempering works, I can see it being difficult to use for any blade shapes other than those with full flat edges.   How would you torch the front end of a blade with a straight back, spear point or trailing point?  Are you lifting the blank out of the water to torch the front sections of those blades?  What does that do to the rest of the edge treatment?

Link to comment
Share on other sites

I'll fill you all in on the details a little bit later today - it's definitely 5160 as I bought it from a reputable metal supplier and it was cut from the same piece of stock, so each piece is identical to the next. 

Unfortunately, I can't blind test as I made them and I'll definitely know which is which. I suppose if I had gone in with that plan in the first place, that would have been doable. Even so, I'm on the fence about this technique, so I'll be as impartial as possible. I'll also post pictures on imgur so people can see the results themselves.

You are right about the shape making it difficult. My knife-shaped-objects are slightly curved, and it made the treatment difficult. I had someone help me, though, so it worked out.

As for the test, I'm going for chopping, checking the edge for sharpness and deformation, and then I'm sticking them in a vise and bending until 90 degrees. I don't think I'll be able to get there, so I'll probably use a tool for leverage.

Hopefully, I'll have some results later today or over the break!

Link to comment
Share on other sites

  • 3 weeks later...
On 11/27/2019 at 3:35 PM, doggonemess said:

Hopefully, I'll have some results later today or over the break!

I apologize for the delay in posting the results; holidays happened and I got really distracted. 

The link below goes to an imgur page of the test results. I haven't posted any captions yet, but I'll describe what I did and it's presented chronologically. The blades are identifiable by the letter stamped on the handle as indicated below.

As you can see, I made four KSO (Knife Shaped Object) blades of almost exactly the same dimensions. They're pretty ugly, but I'm not trying to win a beauty contest, just test their properties. Two came out larger than the others, so I used one large and one small for both methods, to even it out.

The steel is cut from the exact same bar of 5160, which I bought from Alabama Damascus http://www.alabamadamascussteel.com/. I chose their company because they were the best price for a small quantity. Most of the other companies I checked gouged on shipping or required a large purchase. 

During forging, I marked each blade with "T" or "S". The "S" indicates the standard treatment, and the "T" is the water trough. The standard treatment involved hardening in oil and tempering in a 400 F oven on top of a stone for two hours, then allowed to cool in the oven overnight. The trough method I've described above and uploaded videos of a bladesmith doing it.

I decided on three tests, each inspired by the ABS. Now, some of the results would be difficult to compare to other knives but the point of this exercise was to compare these identical knives to each other.

1. Edge hardness 
To perform this test, I clamped a piece of 2 x 4 to my deck. Because the blades were so short, I didn't want to try cutting all the way through, so I decided on an arbitrary number of strikes and then compare the edges. I started with 40 wacks, since it worked so well for Lizzie. 

At the end, none of the blades showed any sign of wear and were still shaving-sharp. So I upped the difficulty and got a piece of ipe. Anyone who knows wood knows that this is one of the hardest commonly available woods. I gave that one 40 wacks with all the blades and none showed any deformation or chipping, and all were still just as sharp.

2. Bending
For this test, I stuck each blade handle-up in a post vise and bent to almost 90 degrees. I say almost since I didn't set up a gauge behind the vise and, I'll admit it, I got really nervous when they were bending past 45. I took the video and corrected the vise angle in Photoshop to get some rough angles.
I'm wearing a welding mask because I couldn't find my faceshield. I'm sure I looked silly to my neighbors.
I recorded these tests and posted them to youtube as I knew that it would be very easy to make false claims about this sort of thing, and wanted to be certain that the results were believable.
The first test was the longer of the two trough hardened blades. It bent almost to 75 degrees without cracking and then back to about 45 degrees of set. I was just as surprised as everyone else will be.
The second test was the longer of the oil-hardened blades. I was feeling confident at this point so I had no worries. It broke violently at about 60 degrees.
The third test was the shorter of the trough blades, resulting in almost the same outcome, bent to a max of 65 degrees. 
The fourth test was the shorter oil-hardened blade; it broke even sooner at around 40 degrees.

3. Edge toughness
Since all the blades were ruined at this point, I figured I might as well see what other kinds of punishment they could take. For this test, I dropped a 1 Kg wrought iron rod from a distance of 30 cm. This is obviously not a standardized test; my goal was to compare the damage to each of the knives as see how they compared to each other.
You can see the tests here, but to sum up, I placed a blade in the vise edge-up, held the chunk of metal against a mark on the backdrop and dropped it. They all performed about the same, resulting in a chip about the size of a fingernail clipping. I was very surprised, I thought the trough knives would roll the edge, even though they performed the same during the chopping test.
T1: https://www.youtube.com/watch?v=EZD3LRdhWjk
S1: https://www.youtube.com/watch?v=BcjWfrzAvjc
T2: https://www.youtube.com/watch?v=hGYGvP3H6XU
S2: https://www.youtube.com/watch?v=3VkbRQMNaKA
Fun fact - the wrought iron bar is a piece of the USS Constellation, given to me by the guy in charge of restoration. It was some bonus material left over from making hooks for the ship.


The results of this fairly informal test are pretty clear to me. With 5160, this method seems to produce a hard edge with a very flexible spine that performed just as well as traditionally hardened blades. The trough treated knives were very flexible and durable. I'm going to make some more blades using the trough method and see how something a little more substantial does in the tests.

If there is anything I would take away from this personally, it's that I obviously don't temper my blades long enough the standard way. I've got to work on that, clearly.

Watch the videos, look at the pictures, let me know what you guys think. I'm open to any feedback, and happy to receive it. 


Link to comment
Share on other sites

For the next round of testing, do a double-blind setup. Put a piece of tape on each finished knife (remembering to include a non-hardened control) and number them 1, 2, 3, etc. Keep a record of which number goes with which blade and which blade has what heat treatment. Give them to someone else, who will add another piece of tape, write on a new set of numbers, make a list of which new number is paired with which old number, remove the old numbers, and return the blades to you. You will do the tests, record the results, and then use the list of number pairs to determine which blade is which.

This way, neither you nor the person recording the numbers used in testing will have any knowledge of which knife has which heat treatment (especially if your blades are as identical as possible), and thus the problem of observer bias will be if not eliminated, at least minimized.

Link to comment
Share on other sites

Okay, true enough. There is very little quality control in the torch method, but I've never done differential hardening, so the first several times would yield unpredictable results. With the torch, the process is easy to monitor - when a section turns bright orange, you move a little further down the blade. It would be much better if there was a way to guide the heat down the blade mechanically. 

Link to comment
Share on other sites

I can say for sure that a mild steel knife wouldn't survive the chopping test and remain sharp. I've made a few knives from inferior steel and they couldn't keep an edge at all. I would expect them to also fold the edge rather than crack when abused. But I should have more data shortly - I made the knives out of 5160 and left them untreated. So that will give us the baseline performance for that piece of metal.

Link to comment
Share on other sites

  • 3 months later...

Greetings, all. After a few months of general nuttiness, I finally got around to doing the untreated test.

I got two pieces of 5160 from the original stock that I made the previous four knives with and made two knife shaped objects. They're ugly, but it's not a beauty contest.

I was lazy about taking pictures and video this time, so there isn't as much detail in those. You can see the results of the chopping, bending, and heavy thing drop tests in this imgur album. 


In the close ups of the blade, you can see the section in front of my finger where the edge has flattened or rolled after the chopping test. It was fairly sharp before that.

For the bend, I did take video of both in case something interesting happened.  Nothing did - they just bent a whole lot.


I then bent them back, which you can see in the album.

After that, I dropped the wrought iron bar on the edges and it caused less damage. This is visible in the second to last image. It just slightly deformed the edges instead of breaking a chip off.

The end result shows that these normalized pieces behaved very differently from the hardened steel tests. 

While none of this is conclusive, it suggests that there is something more to this method than it appears. I don't know what the mechanism is, but I have made a few blades this way and they perform as well as my other (properly) heat treated blades. If anyone has access to an electron microscope, I'd love to send some samples. Barring that, I'm going with my gut on this one and using this technique to treat longer edges, as it's much easier than quenching and tempering.

Link to comment
Share on other sites

  • 2 years later...

It's pretty evident that everyone has an opinion.  Lots of good information in the string though.  I've actually done something similar to this technique to put an edge on short run blanking dies using 4140 steel.  It works but as has been said, not on all allows of steel.  Since you are a hobbyist,  try it a few times and see how it works for you.   Keep track of what you use for stock.  If it's an old file and it worked, buy the same brand file....LOL.  


Link to comment
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.

Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

  • Create New...