File knife brittle even after extreme temper. What gives?

[harry_r]

Hey all, my sister was in town and I found the motivation to fire up the coal forge to try and make some quick blades out of old files.

 

First, the setting:

 

take to non-magnetic (brick forge with charcoal), Oil quench, oven temper at 400 for 1.5Hrs.

 

One had a slight warp. We wrapped the blade in damp towel and gently heated the area with a torch to ease thing, and then applied a little force.

 

Snap. The grain looked like a normal file. Very tight/powder.

 

The question:

 

This seemed a temper failure, and as the blade was trash anyway I decided to test by reheating with the torch. Each time I was able to snap the blade with very little hammer force. Towards the end I heated the remaining piece to dull red (it was noon/bright but held the torch for long time), and it still snapped rather then bend.

 

I could see grain size differences in the various parts (have a pic), but what confuses me is why it remained so brittle. After force heat I confirmed a file would bite, yet still brittle.

 

Why did the temper, even extreme to the point of allowing file to bite, not work?

 

I might think heat treat went wrong, for example over temp, but the first snap had very small/uniform grain so I doubt this hypothesis.

 

Mystery metal? These were old files I got from my father's estate. I believe US made and no newer than 1980s vintage.

 

Just the nature of a file based knife? If so, who wants one......

 

 

Thanks in advance for any thoughts.

 

Regards

 

Edit to add best pic I have of the grain at first snap.

 

 

 

Edited March 31, 2015 by harry_r

[Alan Longmire]

That's actually really big grain, plus it looks a bit discolored. I would guess overheating and pre-existing cracks. Did you grind the teeth off before forging? That would help you see any cracks, if present.

[grpaavola]

This may be a stupid question... but, did you normalize before and after forging?

 

-Gabriel

[Miles Hebbard]

As Alan said...looks big to me, nothing to refer to, but having snapped many a blade...that grain doesn't look even...and as Gabriel asked, did you? A good check before working old steel is to grind it smooth, then etch, (vinegar will do,) rub down with water paper on a hard backing, and cracks will show up...as a 'rust line' then you can expect the steel to break! I've learned to test every piece I make before all that hard work, if a flaw shows, put it aside 'till you make a P/weld billet, it should weld it's self together again during this process! But that does look as if you've over heated the steel...

[Miles Hebbard]

Oh, by the way,I temper old files/saw blades etc, at 150 C, two cycles of 1hr each...then test, (on an oil stone) to see if it is too hard/soft, experiance of sharpening many knives tells me...

[harry_r]

Thanks all for the feedback. A few answers:

 

1. We did normalize, but just once after forging. Things were rushed so cannot say they cooled all the way to ambient. This was mostly to clear any warps.

 

2. We left some teeth to show it was a file. Did not check carefully for cracks before trying to straighten, but as this was near the tang, and all portion broke with similar ease (or I should say with no bending first), I don't think it was quench cracks. The files were in good shape prior to all this.

Having said this, I do recall the blade in question was dropped a few time per-temper. I suppose that could have started some bad things.

 

Am surprised to hear the feedback on the grain. To be clear, that is a mag shot. To my eye with reading glasses it looked a lot like any broken file. Just gray powder.

 

Thanks and regards.

[grpaavola]

That is blown out!

 

Next time normalize 3x before and 3x after!

 

trust me

-G

 

OK... I feel like I need to clarify. Normalizing is not going to solve all problems one may encounter with an old file, but it will add some insurance. Reworking steel is not for the faint of heart! That is why many professional bladesmiths don't recycle. Just don't give up! The 11th Commandment (I believe) was something like... Yea, before one shall smite thine steel thou shall normalize thrice. Neigh, once. For that is to little. Nor, twice but thrice. Five is just right out!

 

I think that was the nutshell though...

Edited April 1, 2015 by grpaavola

[Wes Detrick]

This pictures makes the rounds every so often, but I am going to post anyhow. It shows grain comparisons. As everyone said, that picture above is huge grain. I would say that the grain you have is either the top one or the next one down. More normalization is needed for sure( as everyone has already said . )

 

Edited April 1, 2015 by Wes Detrick

[M. Cochran]

Thanks for thawing that, Wes, I think I'm gonna print that out to hang on my shop wall.

[harry_r]

Yes, thanks again all.

 

Given I have crude tools and not much skill, here is the rub.

 

1. Overheat can cause large grains.

2. To remedy, we normalize 3x.

3. But each normalize is another chance to overheat.

 

As such it seems to get a good HT one has to be able to take to the right temperature 4 times, not just once. Knowing my luck on the 3rd cycle I overheat negating the first two and ending up in same place.

 

This blade stuff is hard.

 

Best regards, and that pic is great.

[Alan Longmire]

It can be done with crude tools (all I have!). The best thing is to do it in near-darkness so you can really see the colors come up. Throw away your magnet, it just tells you when you've hit 1425 degrees. Watch for the shadows in the steel. When the last swirling shadow disappears you have acheived full transformation, pull the blade immediately and let cool to black. Repeat as required. It's even easier if you put a piece of pipe with one end closed off in the fire and use that as your heat-treating chamber. A little charcoal or even wood in the end will take care of oxidation, and you can watch the whole blade as it comes up to heat.

[harry_r]

Thanks Alan. I do need to try that pipe trick. I'd considered this before as a way to prevent clay from being knocked off, but can see the general utility.

 

It was getting pretty dark when we did the HT, and I also tried a wak that had a 26" blade. To preach to the choir, I can tell you getting an even heat of any level on the longer blade was a trick; To think I could spot shadows along its whole length seems a pipe dream, to use a pun. Once buried in the coals I find I cannot see what is happening, and in trying I get night blindness from staring into the coals, which then hampers ability to see low level colors when I remove it. Having it in a dark-ish pipe would help with this I believe.

 

If you think the grain on that file is big..... I clearly cooked the 01 tool steel wak. It too failed. I believe it actually cracked in the quench, but not overtly. After temper I was toying with it over my knee and it simply snapped. I don't see any other cracks and was thinking I can perhaps make a tanto of it after some normalization. But, given one crack is rare, I should probably remove oxides and etch as suggested to check for any others. I have so much work into this already I need to cut losses if its toast.

 

Edit for spelling and:

 

I will post some pics of the wak failure here rather than a new thread. I have one just before it went into the quench that I believe shows a real hot spot in the area of the break. This one is hard to get over. I did it with hacksaw, and files and then on water/diamond stones, and and still have the arthritis to show for it. Way too much work pre-heat treat for sure. Makes me want to harden the bar and just tough out the stock removal on hardened steel given so much stands to be lost in the quench.

 

 

 

Edited April 1, 2015 by harry_r

[Alan Longmire]

Yep, that got too hot! You do know you can't get a hamon on O-1, right? WAY too deep-hardening an alloy for differential.heat treat no matter how much clay you put on it. I feel your pain on all the handwork, that is depressing.

[harry_r]

Agree on lack of hamon with O1. Sorry if I was not clear. There was no clay or hamon attempt here. I have tried for hamon in past with some 1080 from aldo in tanto form. The hamon attempts were hot water into oil while these last failed file knife/O1 wack attempts were straight oil (and no clay). I had posted in the past about issues with clay falling off during the constant in/out of the coals to check if at the right temp, which is when the first mention of the iron pipe was made.

 

I've yet to get the pipe, so instead I used some rutlands furnace cement (from advice here) which seemed to stick better than my home made clay and I had some success. This is to say no cracks and something that could be called a hamon. The hamon itself has all kinds of mune yaki I did not plan on which I suppose indicates poor temp control (again) or issues with clay coming off in the quench. As such, to think of being able to make a chrysanthemum hamon seems pretty darn tricky. Got to hand it to those old time smiths.

 

 

Best regards

[GEzell]

My first 20 or so knives were made from old files, and I still occasionally will make a blade from a file... One thing I've found is that most files have a higher carbon content than you'd expect, somewhere in the range of 1.1-1.5%... with this high a carbon content, 400 degrees is not hot enough for tempering a blade. In my experience, you want to go 475-500 degrees at least, twice, for at least one hour each time, to get a blade that will not be brittle.

[harry_r]

@ George. Thanks. I do believe in this case the brittleness was due to grain size and not just insufficient temper, though that likely also contributed. After the first break I tempered to the extreme with a torch, to the point another file would bite, and the pieces still snapped rather than bend.

 

 

As for the pipe, I suppose 3" might accommodate most long blades depending on their level curvature. A 4" piece 3 feet long is about 50 bucks. I would think the smallest diameter you can get away with is best.

 

What is the consensus here as far as capping one end? It would seem leaving open at both ends would cause a draw.

 

Regards

[Alan Longmire]

I ise a sligjlhtly flattened 2" pipe with an end cap welded on. An open pipe does indeed induce a draft that prevents even heating. By flattening the pipe a bit you get greater width plus a step at the end that lets you get a grip with your tongs.

 

Remember, no galvanized pipe in the forge! Even if you have the ventilation to deal with the zinc fume you will get a yellow-plated blade...

[harry_r]

Thanks again, Alan. The flattening idea is brilliant. Might relieve some stress also. I believe I noted in another thread that you also cover the top of the pipe in coals. That would help even out the heat as well.

 

BTW, some great info from you here watching the shadows and how to do proper descending heat normalize with lack of temp. control. I would love to see those shadows:

 

http://www.bladesmithsforum.com/index.php?showtopic=25957&p=244915

 

Very kind of you to help us newbs out. I just returned from chopping up some new charcoal, having decided I will try this pipe thing. I like the idea of it helping to protect clay when that is again applicable, as lessens the need for the chopping part as well.

 

Cheers

[deker]

A suggestion for you as you work on dialing in your HT in a coal forge. Take one of your files (preferably a Nicholson....I can tell you that they are at least consistent, and IIRC they are W-1) and break a piece off to get a measure of the factory grain. Then, purposely overheat it and quench it and break off another piece to compare. Then, normalize 2-3x, quench, and break off another piece. Lather, rinse and repeat until you have more control of what's going on and a better eye for color and decalescense/recalescence. The key is to have the factory HT sample available for grain comparison.

 

I do this as a demonstration in every knife making class I teach to show the transformations. Every time it's in a coal forge with a pipe shoved down in the fire. When I'm not rushed I can usually get very close to the factory grain on the return trip.

[harry_r]

Deker, that is a good idea. Why try and learn on something that matters? I do have plenty of files.

 

Best regards all.

Edited April 3, 2015 by harry_r

[Brian Myers]

A bit slow getting on this topic, but having worked with a lot of files, I can tell you that normalizing and tempering can help make sure you have a smooth and stress-free grain. I'll normalize at least twice to a good bright red, letting it air cool in between, then one more good heat and I'll shove it into ashes or sand overnight for a slow cool-down. When it comes to tempering, I like two 450 to 500 degree heats, 2 hours apiece with a cool down in between. I'll polish off the temper colors in between to look for any cracking.

 

Remember, files are high carbon and have been put through alot in their lives, they have it in them to be nice knives, but they need a lot of stress reduction!

[harry_r]

Thanks Brian, rather late then never, eh?

 

I have been busy with work so have not tried to practice any normalization. The issue I have is a "good bright red' may be too hot and lead to grain growth. I want to see those shadows. I have the black pipe, just need the time.

 

And to reiterate, after the blade snapped on the file knife I really heated it, to the extent it should have lost all hardness, yet it was still brittle in that it broke rather than bent when hit with a hammer. I believe that was due to grain growth/lack of proper normalize to begin with. IOW, not a temper fail so much as cooked the steel during forging/heat treat.

 

 

Regards

[harry_r]

An update to this old thread.

 

I finally got around to testing the pipe and doing the normalize/HT at night.

 

The pipe does help a lot. And I can see where doing this in any type of light (late evening) can easily result in overheat. The scale is so much less as well.

 

I am not able to see the shadows form while in the pipe. But am able to see them on cool down.

 

I normalized both parts of my o1 wak, then heat treated the tip which will become a kridashi or sorts. I had a 1070 tanto that was already clayed, but I assume also overheated when I normalized my old way, so I scraped it off and normalized it again as well.

 

Thanks for the advice.

 

Regards

 

Edit to add. I tested with an old file first. Did 3 normalization cycles, and I found the file showed the shadows the best. The 01 being the least easy to see. I then quenched the file into cold water and did a snap test. Grain seems reasonable small to me.

 

 

Edited November 29, 2015 by harry_r