Hi,

I'm not entirely sure this is the right place for this kind of discussion, but I'll post anyway. Please relocate if necessary, then forgive me ;). Oh, wait, let's first introduce myself a bit ! I'm Stan, DIY enthusiast and woodworker from Belgium. I like most things involving creating beautiful objects and experimenting with new techniques. 

Last Sunday, I was invited by a friend to try and forge a knife in his workshop. Exactly the kind of things I love. What could possibly go wrong? Well, I had a wonderful day, but ended up with a bit of a disappointing result (to say the least). To be honest, failure didn't really come as a surprise (when you forge your first knife, it's pretty rare to end up with something incredible...) but I wasn't expecting what I got: A blade full of cracks.

Since I'd like to give it another go but without repeating the mistake(s) I made at the first attempt, I come here looking for an explanation about what actually went wrong...

I'll detail the process hereunder: 

1. I started with a cylindrical piece of metal that used to be part of a stone chisel, quite hard.

2. I forged it (gas forge - part drop forge hammer, part hammer) until I ended up with a flat piece of metal about 4-5mm (0.2 - 0.15 inches) thick. 

3. I used a angle grinder to cut the shape of the blade (13cm/5 inches long + 13cm/5 inches handle), then a backstand to grind it until it completely flat on both sides, and I started to give it a very light bevel, which left me with 2.8mm/0.11 inches at the spine, 2.1 mm/0.08 inches at the - future - cutting edge).

4. Hoping not to end up with a banana blade, I put it back in the forge and quenched only the cutting edge (dropping about half the width of the blade held horizontally in water).

5. I tempered it in a kitchen oven (205°C - °F) for 1,5 hours

6. I started grinding again (being very careful not to overheat the blade this time)... until I realized the blade was completely cracked.

7. I cried

8. (Now) I'm looking for answers

I hope one of you (after having had a good laugh at the idiot belgian who tought he would get away with this nonsense) will be able to give me the reason why things went that way and what I should do to avoid such catastrophic failure in my next attempt.

Thanks in advance,

Stan

Edited December 2, 2025Dec 2 by Mod34
Moved to proper section.

Good morning, LeStan, welcome to IFI.
One major piece of information we'd need to answer the question properly.  What was the alloy of steel?  
Every tool steel has a specific 'recipe' one needs to follow for a successful heat treat.  Specifically, how hot the steel should be when quenching and what quench medium to use.  There are air hardening steels. oil hardening steels and water hardening steels.  To complicate matters even further, there are fast oils and slow oils.

Without this information, knife makers are making a guess as to the proper heat treating procedure.
What you can do next time is repeat your attempt but use a slower quenching medium.  Most bladesmiths would say use a known steel (not a mystery steel) to prevent this very thing. 

There are other smiths who would take a small piece of the unknown steel, and heat treat it using a variety of quench mediums to see which one performs best.  You can start with the most aggressive (water) and work your way back until the sample hardens without cracking, or you can start with the least aggressive (air) and quench in progressively faster mediums until the blank hardens.   

What Billy said. When using mystery steel I always quench in oil first.

Nice shape on that blade. Keep at it, I have a pile of cracked blades and failed projects.

Have fun :-)

As you guessed, it was a "mystery steel"... Probably not the best idea, but I used what was at hand. Next time I'll use a leaf spring from a car suspension... So i won't be sure about the exact alloy either, but it looks like a very good advice to use offcuts as samples to test heat treatment for that specific alloy AND to go for less aggressive quenching mediums. Especially since I'm pretty sure my first successful attempt won't end up being the most used knife around the house, so it doesn't need to be insanely hard...

@Rojo Pedro: I also think the shape was pretty cool and since I only noticed the cracks after starting to grind again, I really thought I was going to end up with a splendid knife (and it was even straight !), but it turns out beginner's luck only brings you so far... 

Thank you very much for this precious advice,... I'll make sure I share my next failure in this topic (if it's not closed by then)

Stan

Also, while I'm  no knifemaker, I usually hear "banana blade" used to refer to the curving effect that occurs when you forge the taper from the thick back to the thin edge.  It looks like you successfully countered that in the forge. 

Quenching half of the blade is something of an advanced technique, and can potentially contribute to cracking.  It's probably better to start out quenching the whole blade, and then tempering the back softer (you can put the whole thing in the oven, and then temper the soft part further with a propane torch or over a hot piece of steel).  My recollection is that steel expands slightly when it goes from austenite to martensite, so quenching only the edge would actually make a banana effect slightly worse.

Welcome aboard Stan, glad to have you. If you put your general location in the header you'll have a much better chance of meeting up with members living within visiting distance.

I don't have much to add to Billy's advice but there is an old saying, If a good blade you will win forge it thick and grinding it thin. Next time leave more thickness before heat treating it and heat treat BEFORE you finish grind it.

Frosty The Lucky.

@Mike BR the idea of quenching only the cutting edge (be it good or bad...) was to put less thermal stress on the upper part of the blade and hope that - being straight - it would keep the whole thing to go banana. In a way, it seems to have worked (the blade is straight after all...). I obviously don't know if quenching the whole blade would have made it curl or not. 

@Frosty I did heat treat before finishing grinding, but maybe I should have left more material. I had a look at what the "ideal thickness" is, but found a lot of different answers to that question... You don't want to start grinding (without overheating) a huge hardened chunk of metal either, so I still don't know how thick "forging it thick" actually is...

I forgot to mention when I greeted you that this is a forum, the @ tag messes up the OS requiring them to have to clean it out and will only get you a message from the moderators.

Please just use my login or call me Frosty.

Frosty The Lucky.

 

Ouch, sorry, didn't know about that... Most forums allow that, but I'll avoid it in the future

No problem Stan, Iforgeiron has a learning curve and takes a little time to get used to. Tagging is the norm many other places.

It's good, making slips happens to everybody, heck even mods make mistakes.

Frosty The Lucky.

I will repeat everything already said and add this: 

Try and shatter a piece. 

What I mean is, get it hot, quench it too quickly, and hit it with a hammer. (of course, wrap it in a cloth and especially wear eye protection and face shield). When you see the breaks, you will notice it is a loose grained metal, like kosher salt. You want a fine texture, like powdered sugar or thereabouts. I did this with a student recently and it was a beautiful learning opportunity. 

Cracks can form at any stage from any number of variables. Part of getting consistent results is to control these to the degree possible. 

Temperature, time heated, time cooling, hammer blows.... all can affect the finished product. 

You may have been forging a piece of metal with cracks in it. Breaking it at the crack can also tell you when the cracks formed. Oxidized (meaning dark) happened while forging, generally, and silver/light grey occur during or post quench. Finally, make sure you temper soon after quenching. 

Thermal stress occurs whenever a piece changes temperature unevenly.  So quenching only part of a blade increases stress.  Of course, it's impossible to get a perfectly even quench -- the outside of a piece cools faster than the inside, and thinner sections cool faster than thick ones.  That's why it's important to use the slowest quench that will harden the piece (and not quench air hardening steel in oil, or oil hardening steel in water, for example).  With that said, there are specific reasons to quench only the edge of a blade, but doing so is riskier than using a full quench.  

AT (!) Ridgeway Forge Studio, thanks for your input! Didn't know at all that it was possible to figure out when the cracks occured...  I'll first copy the design (that kind of came by accident but that I really like), then (try to?) break the blade and see what's inside

Tempering started about 1/2h after quenching, so I hope the timing is OK

Mike: Next time I'll quit messing around and quench it fully,... Didn't think that as "advanced", otherwise I wouldn't have done it, since I'm definitely not an "advanced" knife maker. Quite the opposite 

 

The only other thing worth mentioning is that normalizing the material a few times before taking a quench would be beneficial, depending on the steel.

All the above is good advice/information.   I'll add one more thing.  If you continue to try to forge when the steel has cooled too much, it's easy to introduce cracks.  The next question is, of course, what is too cool?  The answer is it depends on the alloy.  Generally speaking though, you want to quit hitting with blows that change the shape before it drops into the dull red range.  You can still do light blows for straightening safely (usually) at that heat, but for moving metal generally you want to be in the orange range. 

Again, this is alloy dependent.  For mystery metal you will need to do some experimentation as mentioned above.

And here I've been needing to make stone chisels - that could have been part of the problem too, a lot of the stone chisels now have carbide tips. I would tend to guess quenched in the wrong medium, ground it too thin first, or worked the blade too cold at some point.  It's got a really nice "sheep's foot" profile, especially for a first one and grinding with an angle grinder.

There's a process for heat treat that's a little different for mystery steel; it's often best to take a little and make test coupons to see which medium works best. My normal technique is normalize (heat to non-magnetic and let cool slowly), two or three times (or one if I'm being lazy), heat just a little past non-magnetic, then quench in a vigorous up and down motion with the whole blade. Temper immediately - for a blade, the easiest way is in an oven or toaster oven, set about 400-450 for two hours. Wait a day and do it again.

You can do this going from the least aggressive quenches to the harshest (oil, water, brine), stopping if/when it hardens (a file skates across and makes a glassy noise), but the more times you heat treat a thin blade, the more chances of failure. It's a little simplistic but will get you there on most steels, hence the test coupons.