Terrified of quenching O2 knife

[Gustav]

Hey.

I'm a new bladesmith (with a couple of years experience of traditional blacksmithing) and I've made a couple of knives out of O2 steel. The first one turned out great, but the second one had cracks all over the edge and spine. I've done a little bit of research and people say: the edge hardens first, so when the thicker spine hardens the edge can't flex => cracks. What I think caused it:

• I had problems normalizing it before hardening.
• The oil (canola oil) was probably not hot enough.
• I used a coke forge and struggled to reach an even temperature.

Read a bit online and found out O2 isn't for inexperienced bladesmiths. I didn't know this when I began forging my next knife, a kitchen knife. It's not hardened yet, and I'm terrified of doing it. My question is: how do I make sure I don't get any cracks in my new knife? I've bought some vermiculite to normalize in, but someone said online that you shouldn't normalize O2 in vermiculite. He said you should do a "sandwich" with hot pieces of thick metal under and over the knife. Does vermiculite work or was he right? Is there anything else you recommend me doing to increase my chances of a successful quench with O2? I've ordered 80CrV2 steel (which supposedly is a more beginner-friendly steel) so next time will be easier.

Very thankful for all answers!

// Gustav

[anvil]

First off you are mixing up normalizing with annealing.  You just air cool from critical to normalize. You cool slowly in vermiculite or lime to anneal.

There is a free app for android called heat treaters guide companion that will give you all the info to heat treat O2.

Dont be afraid of O2, its actually pretty user friendly.

You just need practice.

[Gustav]

anvil - Ok if I want to shape the knife after forging so a file can remove material, is it air cooling (normalizing) or vermiculite (annealing)? I've always thought of annealing as the process after quenching with the tempering colors and normalizing as the "remove all hardness and stress before shaping and quenching".

Before the quench there's the normalizing cycles so it's critical temp then air cool regardless of how thin it is? It feels like it cools down too quickly when air cooled so it's not very soft.

[anvil]

Heres the process:

1:Normalize

2:anneal

3:cold work

4:harden

5: temper

You need to know what steel you are using because modern steels are temp specific.

As a general rule. 

"Critical temp" = cherry red or about when your steel loses its magnetism,  actually just a bit hotter.

Normalize by reaching critical temp and air cooling. This refines the grain.

Anneal by reaching critical temp and cooling slow in lime or vermiculite. This brings your steel to dead soft. Now you can do your filing.

Harden by reaching critical temp and quenching. Heres where you need to know your steels. Some quench in water, brine, oil, or air. This step makes your steel brittle hard. If you drop it, it may shatter. If you leave it sitting out over night, it may crack or break by morning.

Tempering is when you draw back the hardness in order to have a working tool that wont break, shatter, chip, or dull too fast. These are the colors you see that run from straw thru blue to black. And the blue you see when you grind a drill bit or a knife is in this range, and in this case means you screwed up! Tempering is how you can get a "hard" cutting edge, a spring, or a tough hammer face.

Again, this general, but not specific for a particular steel. Get the apk i mentioned. It is avalable for android and iphone. There is a sticky on it in the heat treating section.

And the "O" in O2 means it is an oil quench steel.

Hope this helps.

 

[Gustav]

6 minutes ago, anvil said:

1:Normalize

2:anneal

[...]

Now that's something I didn't know. Normalize and then anneal! One question however: after grinding etc, do I need to normalize and anneal again just before quenching? For instance, if the steel gets pretty hot while grinding (maybe not glowing, but hot), is it necessary to redo steps 1 and 2 before hardening?

6 minutes ago, anvil said:

Hope this helps.

It definitely did!

[anvil]

If you feel heat when you are grinding or see a y of the temp colors (straw thru blue to black) you blew it!. Dont let it get that hot. 

If you do, then the safest way is to go back to step 1 and start again.Especially when at the bottom of the learning curve. 

 

[Gustav]

5 hours ago, anvil said:

 

I'll make sure to regularly dip it in water.

Thanks!

[anvil]

Have fun!

One more thing. Cracking can be caused by forging either too hot or too cold. This can happen whilst forging, or show up when you heat treat. 

[Steve Sells]

there is an entire section devoted to knife making here in the form, I suggest you stop what you are doing and read some of it before moving on. I will move this post there

[Buzzkill]

 

 

On 5/5/2019 at 6:33 AM, anvil said:

1:Normalize

2:anneal

3:cold work

4:harden

5: temper

Hmm.

That's different than the sequence with which I am familiar.

As I understand it the purpose of normalizing is to relax the steel and define/refine grain boundaries.  The purpose of annealing is to grow the grain size and put the steel in its most machinable state.  In other words annealing undoes normalizing (or hardening).  If I have this wrong I'd like to know.

My usual routine is:

1. Forge to shape

2. Anneal

3. Most of the stock removal/cold work

4. Normalize (x3)

5. Quench (harden)

6. Temper

7. Remainder of cold work (minor stock removal, sanding, polishing, etc.)

This is for simple steel alloys and may not apply to some complex alloys. It's possible I have this wrong, but with what I think I know I don't see the point of normalizing before annealing.

 

[anvil]

 

i always follow the specs given for the steel. A good reference is " The heat treaters guide companion". Their is a post on this in, i think, the heat treat section.  Also, of course, i should have listed "forge" as step one.     My go to sources are the above and my copy of Carpenter Tech heat treat specs for many of their steels.

I did a "quick" temp check once upon a time and it seems that as a general rule, the temp to normalize is a bit higher than that for annealing. Thus, normalizing second, before hardening, means your steel is not fully annealed when you harden. since grain growth is refined at normalizing temp, if you do this, neither step counters the other.

Ill give one example:

W1: normalize (1600-1700f)  anneal (1375-1400f)

So basically its how i learned to HT, It is how steels were heat treated historically, and it is the process recommended by contemporary companies that make tool steels today.

Thats good enough for me. 

[Latticino]

In this case I believe you are disputing terminology. This is one of the common problems these days with discussing heat treatment. The annealing that Buzz is , hopefully, referring to is spheriodal  annealing, the one that anvil is referencing is subcritical  annealing. The former involves entering the austinizing range and leaves the steel at its relatively softest state, for maximum cold workability. The latter is the final step in stress relief, used just before heating and quenching, to harden the blade while minimizing warping.  This last is often labeled part of the three stage normalizing process, where each cycle is air cooled from descending temperatures starting above critical.

Note that the former annealing will not work for most knife steels unless cooled slower than just holding in still air.  Typically using either a programmable oven, or encasing is a very good insulator immediately after heating (like vermiculite or wood ash) is the technique. I sometimes just leave my blade overnight in my forge when it gets shut down. Close the doors and it has enough thermal mass to cool slowly.

Couple of other minor points:

1. The heat treaters app/book is intended for more industrial crossections than are typically used for blades. Follow it precisely without allowing for this at your own risk.
2. I have found Kevin Kashen's site to be a true wealth of information on how modern bladesmiths heat treat current common knife steels.  Strongly recommend that you look him up.

 

Just looked back thru this thread and see that anvil made some great suggestions and clarifications, going to let this post stand anyway as it seems there are still questions.   

[anvil]

To clarify, the process for annealing is to reach annealing temp, then cool slowly in lime. Normalize is air cooled.

 

The sequence for all is as i listed, which match both sources and all the spec sheets ive collected. He used to have a page detaiing how to heat treat to pass the ABS journeyman test. The process he laid out is the same as mine for all my heat treating.

My personal belief is that, all things considered, you cant go wrong, following the specs for any given steel, industrial or otherwise, assuming the specs are from the manufacturer, not the distributor. And when in doubt, call their tech service. Personally I just dont accept that there is a major difference between industrial, blacksmithing, or knife smithing when heat treating. There are certainly nuances and a bit of extrapolation is needed, but the process, temps, etc work well for all three. 

I also recommend Verhoeven. Ive skimmed his book a few times, read it thru a couple of times and taken profuse notes o ce or twice as well. Ive not found where he nor Cashen actually vary from the spec sheets again, all things considered. 

Verhoeven spends far more time dealing with normalizing and barely touches on annealing. My gut feeling tells me that this is the reason that knife makers downplay, or dont anneal. I believe he does that because normalizing and grain growth are "healed" by normalizing and perhaps this is very critical for knifesmiths to be aware of. But i do not believe he imies that anneing is any less important to the process. However, ive found no statement or suggestion in his book that indicates not to anneal. If anyone can point out different, id love to have a page reference.

Finally, my approach is more conventional or spec oriented and works for my ht needs than what we see for knifesmiths. This covers our blacksmith tools of all sorts, as well as stone and wood tools, dirt bars tempered for different types of soils/rocks and the few knives ive made.

I make no claim that this is "the way", nor that normalizing 3 times, or any of the other ht variations do or dont work.

Because This topic seems to often become quite heated, and i refuse to debate "better or worse",  suffices to say, it works for me.  

[Cedar Crest Forge]

1 hour ago, anvil said:

Normalize is air cooled.

Nice post. I think you meant just the 02 for "normalize = air cooled", as some tool steels like A2 do not normalize that way.

I am not the expert on tool steel; just clarifying the topic.

https://www.engineersedge.com/materials/normalizing_annealing_tool_steels_13253.htm

Gustav, as for the knife, if there is terror in your process, then may I suggest working with some smaller test pieces until you get your process refined. Take some strips of steel ,bevel them, and process.. Break a few at the end of the process, to view your grain structure, etc.

[anvil]

Yup, its pretty hard to list all steels in an article here.   Thats why i stated Follow the Specs. As for A2, according to the specs,,, Do Not Normalize. So, there is really no comparison between normalizing technique and Do Not Normalize. 

For me, rule # 1 is Follow the Specs.

When in doubt, or got a problem? Follow the Specs. 

If you are still having problems,,, refer to rule # 1.  

Genrally you will find that most steels follow the process above that I posted above,, IE as stated in the specs!

A few steels specifically state In The Specs,,, do not normalize.

Even fewer(that ive found) state(in the specs) to anneal twice. Once as i listed, and a second at a lower temp after heavy machining as this can add stress.

Ive never found a steel that lists anneal before normalize. Exceptions welcome. 

Ive never found a steel that states not to anneal. Correct me if im wrong as there are a "ton" of steels out there!

As far as normalize=air cooled. To clarify, bring your steel to normalizing temp then cool in open air where there are no drafts. Again, specific steels may vary, follow the specs. 

Lol, i hope ive made my ht routine clear. Just in case you miss the message,, Follow The Specs.

Works for me

 

[Gustav]

I normalized in air and then annealed in vermiculite yesterday and took the knife out today. It's soft! One question though: what's the difference if you compare dead soft carbon steel and just regular old mild steel? The knife is soft, but mild steel is softer. This might be a no-brainer, but not for me.

[JHCC]

Very generally, metals can deform when their crystal lattices don't have any interruptions that stop the molecules from shifting around against each other. The more highly alloyed a steel is (remembering that steel itself is an alloy of iron and carbon), the more "stuff" there is to disrupt the lattice.

(This, by the way, is why cast iron is both hard and brittle: the excess carbon precipitates out of solution into graphite flakes that not only resist deformation, but also allow cracks to propagate easily.)

[Buzzkill]

On 5/9/2019 at 6:03 PM, Latticino said:

In this case I believe you are disputing terminology. This is one of the common problems these days with discussing heat treatment. The annealing that Buzz is , hopefully, referring to is spheriodal  annealing, the one that anvil is referencing is subcritical  annealing.

I think you are correct here. Like you, I tend to anneal by leaving the piece in the forge when I'm done and then closing off the ends.  That tends to be above critical temperature rather than below.  It appears as though I need to do a little more studying about what's actually happening to the structure of the steel during these different phases of heat treatment.  I don't seem to quite have my arms around it completely at this point.  As long as I'm going through all the effort I'd prefer to get the most out of a steel that it has to offer.

[Latticino]

1 hour ago, Gustav said:

One question though: what's the difference if you compare dead soft carbon steel and just regular old mild steel?

A lot depends on what you mean by "soft".  If from a deformation standpoint, John's (JHCC) explanation of sliding of the layers of the metallic lattice certainly follows what I remember from my Material Science classes lo those many years past. 

If, however, you mean machinability (or essentially resistance to abrasion) I think you have to take into account the shape of the iron, and other, carbides that can form in various types of steel.  Depending on the heat treatment process these can form in larger or smaller aggregations, and are most likely more evident in high carbon steels than in mild steel.  In fact certain processes can result in flat lamellar plates of these carbides which can be quite resistant to drilling or saw cutting, while the rest of the steel matrix is very soft and fully annealed.

[JHCC]

Good point.

[Gustav]

I've been very busy and finally found time to get going on this project. If anyone's still following this I have some results:

Forged

Normalized (is it really necessary before annealing?)

Annealed in vermiculite

Ground the knife

Normalized x 3

Quenched in well heated canola oil

Tempered in oven at 200 °C two times, 1 h each.

 

I got an even heat by placing a square pipe in the coke forge, it was supricingly effective. It pretty much created a propane forge lol. The quenching went good, even though I got a slight warp that I wasn't able to correct. It's probably something I'll have to grind away if I have enough material. However, there were no cracks. Also, the file-test was just wonderful. The high pitch just echoed in the shop and I let out a long sigh of relief.

The picture is the knife before the tempering cycles. It is now at a straw/light bronze color with some patches of purple (not on the edge).

[Buzzkill]

Success is one of the sweetest things in life.  Congrats!   I'm looking forward to seeing the finished knife.

[Gustav]

I do however have something I'm worried about. When I ground away the surface after HT, i really wanted the edge to be clean. I then dipped the knife in water and quickly ground away some material throughout 70% of the edge with the knife perpendicular to the sander. Later, I tried filing and noticed that the tip (the part I didn't grind) might be a little bit harder than the other part of the edge. I am however uncertain and could just be thinking this, but I'm thinking about how to fix it if 70% of the edge actually got some of its hardness "removed". I did not see any tempering colors whatsoever during grinding, however.

One thing I thought about is to file away a bit of the edge to hopefully "get to" harder steel a tiny bit closer to the spine. How should I do this? Use an ordinary file until I can't remove more material, or sand it away? What grits?

I'd also like to have those rockwell files that test hardness. What are they called, how many should one have and which HRC should they range from?

Or if you have any tips on how to quickly check if the hardness of the edge is "good" and not ruined. HT went well, the problem (if there is any) does not have to do with quenching or tempering. Like if a file shouldn't even bite into the steel at all after HT or if it's fine if it bites a little bit.

[BIGGUNDOCTOR]

One thing I found while making items out of O-1 in the machine shop was that thin sections of O-1 will harden in air,  I had sections around 1/8" and under do it, so be careful when working with it.

[ThomasPowers]

As the heat treating books base their recommendations on tests done with 1" thick pieces; knifemakers have learned that often in knife thick pieces you go one quench gentler; so oil => air, water => oil,   best to learn how your alloy works on a test piece than something with a due date and money riding on it.