Scrap Steel Testing

Old Metal Working threads with reference value, but not necessarily tutorials.

Scrap Steel Testing

Orien M
Registered User
Orien M
Registered User
Joined: January 25th, 2007, 8:48 am

October 30th, 2010, 5:19 pm #1

I've just realized that while I am always directing beginner bladesmiths to test out scrap steel, I've never posted anything about the actual testing methods! So, I'll try to list them out here...if you have any methods to add, questions, etc, then please post them up.

I'm going to divide this into "pre-HT" and "post-HT" tests. The pre-HT tests are done on the raw material...the post-HT ones are done on the finished blade.

PRE-HT:
Spark testing - This will give you a (rough) estimate of carbon content. Apply steel to a grinder, and look at the sparks...high-carbon steels make short, "fuzzy" sparks that burn up quickly. Lower carbon steels make progressively longer, less active sparks. Wrought iron makes very long sparks that barely burn at all. (There are better ways to ID wrought, though. Hacksaw halfway through a piece and bend at the cut, wrought has wood-like fibers where mild steel just looks homogeneous.)

Quench testing - this is where you heat some of your steel to critical, quench, and then try to break off some of the hardened bit with a hammer (wear goggles!). If it snaps, it's hard; if it bends, it's not. You can also try to file it, and see if the file 'skates'. I tend to use a series of quenchants: hot oil is the gentlest, then warm water, and if it won't harden in that, you can try soapy brine (superquench). This will tell you A: that this steel will harden, and B: What type of quenchant it "prefers".

Grain size testing - Kind of a sub-process of quench testing. Once I know a given steel will harden, I'll do some normalizing, re-harden, and hope that when broken, the grain will be nice and fine. If it's still coarse, I know I have to adjust the temps downward, and re-do the test until I get a fine grain.

Temper testing - I don't have a heat-controlled tempering oven, so I use the color of the oxides on the heated steel to judge temper. The colors are not very accurate by themselves, so I'll temper the piece to what looks good (medium straw, usually), grind a sharp edge on, and try to chip or bend the edge. If it won't chip or bend, great...if it does, I know I have to adjust the temper one way or the other.

"Quenchline" testing - you can get a sense of a given steel's ability to form quenchlines or hamon. Take a little strip of your steel, quickly heat one end up to critical (similar to heating just the edge of a finished blade...if you look closely you should be able to see the 'shadow' of decalescence creeping up the steel) and quench the whole thing. Then sand a little and etch...if you're lucky, and have found a shallow-hardening steel, there will be a visible line after etching.

Beyond these, it's usually a good idea to check all scrap over for physical flaws...cracks, bad rust pits, and so on.

POST-HT:
File testing - I use this one when the blade first comes out of the quench. The whole edge should be hard, and 'skate' a file evenly (some small amount of decarb may have occurred, so if you do find a soft spot, file a little more and see if it's not hard underneath). The file test works while tempering too...you can actually feel the file start to bite as the hardness drops.

Cutting tests - important! Done after the edge is ground and sharpened. I typically start by whittling some soft wood, then try hardwood, if it does well with that then on to antler or bone. Make sure to try out various parts of the edge. This is also when I do the brass rod test (pressing the edge against a rod of brass and attempting to chip or bend it), and I sometimes will try to whittle a piece of aluminum or copper a little, too. A good HT should survive all these tests, although edge geometry also has a lot to do with it.

Impact testing - when all else looks good, I often give the flat of the blade a good, hard smack against my anvil stump a time or two. It should NOT break in half from this test...

No doubt I'm forgetting some possibilities, but this should give a good basis for doing some scrap testing. Hope it will be useful to someone...

-Orien
Last edited by Orien M on October 30th, 2010, 7:11 pm, edited 1 time in total.
My handmade knives and tools on Etsy... http://www.etsy.com/shop/OldSchoolTools
Quote
Like
Share

Joined: April 1st, 2006, 10:11 pm

October 31st, 2010, 12:54 pm #2

Thanks a lot for posting this, Orien. I know it was some work to type this all up.
Matt
Quote
Like
Share

jamiemackie
Registered User
jamiemackie
Registered User
Joined: April 14th, 2008, 10:27 pm

October 31st, 2010, 7:29 pm #3

One thing I will add. When working with EN45 (a comon spring steel in the UK) and other steels with very high austenising temps you will obviously need to go higher and as a side effect you can get a thickish layer of decarb that will make you think youve got a soft blade. If it doesnt harden and you know it was once a spring the try a higher temp (900c for en45) and file a bit deeper before chucking it. I forged a boowie from en45 and when I checked with a file it was soft, I went and spoke to Owen bascially saying 'Owen you son of a.. youve given me mild steel and ive spent hours making this knife with it' , he explained and gave it a good filing and it had hardend but had a nice thick layer of decarb.
Quote
Like
Share

Knifesmith
Registered User
Knifesmith
Registered User
Joined: January 25th, 2008, 9:09 pm

November 1st, 2010, 9:28 am #4

Here's a really crappy drawing I found in an old welding text that has at least a basic illustration of sparks that will perhaps help the more graphically inclined:

I feel like I've gotten fairly comfortable i.d-ing at least the steels I tend to use, but I do keep small scraps of a number of steels to use as comparison "sparkers" if I forget to mark a particular steel or have a "mystery steel".

Also, it's probably important to note that alloying agents do weird things to the spark test.  High speed steels tend to throw really long sparks with barely any bursts that look almost like wrought iron sparks, but are more reddish near the wheel.  Some stainlesses toss sparks that are about as long as the pain HC steels, but less active on the "burst".  And some of the really weird alloys like Stellite barely toss any sparks at all.

--------------
Eric Dobratz
Scarab Forge
Quote
Like
Share

Orien M
Registered User
Orien M
Registered User
Joined: January 25th, 2007, 8:48 am

November 1st, 2010, 4:03 pm #5

That's a great illustration; thanks for posting that Eric! I use spark testing for a rough carbon content only...the alloys do throw it off somewhat, although if you had labeled samples to compare you could probably get more precise results.

(I have this one really weird center punch that is crazy hard steel, but barely sparks at all...keep wondering what THAT might be. Stellite? )
My handmade knives and tools on Etsy... http://www.etsy.com/shop/OldSchoolTools
Quote
Like
Share

Joined: April 1st, 2006, 10:11 pm

November 1st, 2010, 5:49 pm #6

Tungsten carbide, maybe? Most grinding media should barely touch it, so I'd expect minimal sparks. (Seems like an awfully expensive center punch, but I suppose it's possible...)
Matt
Quote
Like
Share

Orien M
Registered User
Orien M
Registered User
Joined: January 25th, 2007, 8:48 am

November 1st, 2010, 6:33 pm #7

Well, the weird thing is that it's obviously hammer forged, and I didn't think tungsten carbide could be worked that way. I got it for $2 at a pawn shop, thinking it might be good blade stock, but thought better of it after I saw what the sparks looked like. (It's a sweet center punch, though .)

Changing topics, I had to do (and got pics of) a bit of testing yesterday. This is a pair of old wagon springs, cut in half, ground and etched. If you look real close (in the bottom left especially), there's a bit of layered pattern visible...these springs are shear steel (yay ). During grinding, the sparks looked nicely fuzzy, so I think they have a respectable amount of carbon as well (maybe 1080-ish).



Next step in testing these will be to harden some, and experiment with normalizing and temper temps (might be worth a little hamon testing too, this stuff is typically low-Mn and shallow hardening). After that I can forge some blades! This one pack of springs contains enough material for three big and two small blades. I love working with this stuff, and it looks really cool when finished, almost wood-grain like.
Last edited by Orien M on November 1st, 2010, 7:43 pm, edited 3 times in total.
My handmade knives and tools on Etsy... http://www.etsy.com/shop/OldSchoolTools
Quote
Like
Share

Joined: April 1st, 2006, 10:11 pm

November 1st, 2010, 10:02 pm #8

That's awesome material, Orien. I'd love to stumble across some shear steel.

I also doubt that WC can be forged, though I don't actually know much about it.
Matt
Quote
Like
Share

Knifesmith
Registered User
Knifesmith
Registered User
Joined: January 25th, 2008, 9:09 pm

November 1st, 2010, 10:41 pm #9

Yeah, the tungsten carbides should barely throw off any sparks, and the ones they do are really short.

--------------
Eric Dobratz
Scarab Forge
Quote
Like
Share

Orien M
Registered User
Orien M
Registered User
Joined: January 25th, 2007, 8:48 am

November 2nd, 2010, 1:03 am #10

That's what it looked like...very few, very short dull sparks. It reminds me of grinding ceramic tile, almost...weird. It's a big punch, 1/2" diameter by maybe 3" long..seems like a huge piece of carbide, if indeed it is that. Roswell is nearby...space punch?

More fun with shear steel testing...this afternoon I forged out a little bit of my wagon spring and did some hardening tests. I normalized a little and quenched in oil. Results were not great...as you can see, one side got hard, but the other just bent. I repeated the oil quench test again to make sure, and got even more bendable edges on the 2nd try. Seems I was off base with my carbon estimate, if it won't harden well in oil it's more like 1050-60 than 1080.

(Note, that's not a crack on the edge, but a cold shut...it was there before hardening. The spring had a rivet hole in it, which was forged closed).



Next I re-normalized and tried quenching in water. Much better hardness...some quench cracks, too! But when I repeated the water quench on the other end of the sample at a lower temp, the edge hardened without cracking (the steel had a finer grain size on that end, too).



So for this set of springs, I'm leaning toward a water or brine quench, but I'll really have to watch my temperatures. I used the last water-quenching to try to form a quenchline...I have the sample etching now, so we'll see soon what kind of line, if any, I got.
Last edited by Orien M on November 2nd, 2010, 1:21 am, edited 2 times in total.
My handmade knives and tools on Etsy... http://www.etsy.com/shop/OldSchoolTools
Quote
Like
Share

Orien M
Registered User
Orien M
Registered User
Joined: January 25th, 2007, 8:48 am

November 2nd, 2010, 1:57 am #11

Oh yeah: I should mention also that this is just the first round of tests. As you can see, the grain in all the breaks is looking pretty big & nasty...I only normalized 1x before each quench. Now that I know what to quench in and roughly how hot, I will prepare another sample, normalize it better to get the grain finer, and a do a "proper" knife HT, edge testing and all before trying a real blade. This material is relatively rare, and I don't want to mess it up!

(With a more common scrap type (a file, say), I might not get quite this in depth...in my experience most files can be treated "like a file", and only rarely do you find an exception that doesn't act "file-like" in HT. So often one set of quench tests will be enough to get it figured out.)
Last edited by Orien M on November 2nd, 2010, 2:10 am, edited 2 times in total.
My handmade knives and tools on Etsy... http://www.etsy.com/shop/OldSchoolTools
Quote
Like
Share

Orien M
Registered User
Orien M
Registered User
Joined: January 25th, 2007, 8:48 am

November 2nd, 2010, 4:42 pm #12

Pulled this out of the vinegar etch this morning. The piece (edge heated, quenched in water) did take a visible temper line...looks like, as I was hoping, this steel is shallow hardening.



With the next sample I need to change the orientation of the piece...this stuff, like wrought iron, has a pretty defined fibrous structure. With modern, homogenous steels it doesn't really matter which end you use, but by putting my bevels on the 'end grain' of the piece like this, I suspect I made the edges more prone to crack than they would otherwise have been. Sample #2 will have the fibers running the other way, parallel to the edge.
My handmade knives and tools on Etsy... http://www.etsy.com/shop/OldSchoolTools
Quote
Like
Share

Orien M
Registered User
Orien M
Registered User
Joined: January 25th, 2007, 8:48 am

November 2nd, 2010, 5:34 pm #13

Changing topics again here, I wanted to add something I forgot to mention above. There is loads of material out there that is NOT suited to the task of becoming a knife blade; when should a piece of steel be rejected as blade stock? In my opinion, there are several indicators...

-First, it may not harden at all! Might just be plain old mild steel...also, some items you might think would be suitable (big hex/allen keys and "HC" railroad spikes come to mind here) actually are made to be tough, not hard. If you try quenching something and it won't harden in water, it's probably too low-C to bother with. To be fair, superquench will harden some of these, but there are much better steels to fool around with. (I do still collect old allen keys, since sometimes "tough but not hard" is a useful quality.)

-Highly alloyed steels can be a real pain, too. If you don't soak these at temp (or don't have the capacity for long, steady heats at all), then you can't get the alloying elements into solution. There's a tendency to harden irregularly, and end up with soft or extra-hard spots. Even leaf springs (often 5160, I'm told ) often show this sort of behavior. Also some alloy steels (like HSS, cobalt, etc) don't forge well...others can't be annealed, hardened or tempered by normal means at all, like air-hardening steels. (My dislike of alloys is one reason I have been using antique stock so much, since it seldom contains anything but iron and carbon.)

-Most stainless steels are next to impossible to HT by home shop methods, I avoid these like the plague also. I'm told that working stainless on equipment that has been used for regular steel will give the stainless rust problems, in any case.

-I'll also reject anything that just "doesn't behave right" in tests, for whatever reason...case hardened files, pieces that warp over and over despite normalizing (they're out there), and other unexpected behaviors.

So to sum up: be merciless in rejecting a piece of scrap steel if something repeatedly happens in testing that you can't explain or control! It will save headaches later, trust me....
Last edited by Orien M on November 2nd, 2010, 5:49 pm, edited 1 time in total.
My handmade knives and tools on Etsy... http://www.etsy.com/shop/OldSchoolTools
Quote
Like
Share

Orien M
Registered User
Orien M
Registered User
Joined: January 25th, 2007, 8:48 am

November 7th, 2010, 12:00 am #14

Back to testing my shear steel samples...this is the remains of sample #2 (a tiny knife blade shape, this time) after 3X normalizing and quenching in warm water. I held the blade in the quench a second too long and cracked it (whoops!), so I busted it into a couple pieces to look at the grain. It's a lot finer now...I overheated the edge a little, as you can tell by the slightly coarser grain there. The spine did not harden, despite fully submerging the piece in the quench...another indicator that this steel is shallow-hardening.



Water-quenching is stressful, both for the steel and the smith! But, I'm getting closer to an effective heat-treatment.

At this point, I decided to just go ahead and try a knife....



...and ended up with this little guy. This is definitely a shallow hardening steel! I clayed the blade before quenching, and got this nice hamon. Unfortunately, the blade also developed a small quench crack about 1" behind the tip, but the unhardened spine held it together. I put a handle on it anyway...I'll try it in the kitchen, and use it for edge testing before attempting another.
My handmade knives and tools on Etsy... http://www.etsy.com/shop/OldSchoolTools
Quote
Like
Share

Joined: April 1st, 2006, 10:11 pm

November 8th, 2010, 10:23 pm #15

Great stuff, Orien. Thanks for sharing, as they say.
Matt
Quote
Like
Share

woodsroamer
Registered User
woodsroamer
Registered User
Joined: October 4th, 2008, 4:43 am

May 5th, 2011, 7:16 pm #16

Really appreciate this thread. Thank you for taking the time to contribute another informative piece. 
Quote
Like
Share

downunderrunner
Registered User
downunderrunner
Registered User
Joined: April 24th, 2012, 8:47 am

October 22nd, 2012, 10:13 pm #17

actually sounds like your "center punch" could be the broken off end of a DU long rod penetrator, from an APDS anti tank round, about the right size for it to be from a 105mm APDS round from the late 70's early 80's
<br>
Quote
Like
Share

Knifesmith
Registered User
Knifesmith
Registered User
Joined: January 25th, 2008, 9:09 pm

October 22nd, 2012, 11:42 pm #18

eeps! Time to take the punch over to Los Alamos for radioactivity testing Orien!

--------------
Eric Dobratz
Scarab Forge
Quote
Like
Share