How to identify steel

Joined
Jul 3, 2007
Messages
392
I'm guessing that this question has been asked and answered, but I can't find it... so sorry for duplication of effort, if such is the case.

I was fortunate enough to have a friend with a ware house that was being emptied of aircraft parts. Found a plethora of steel.
Problem: how do identify what my obviously high quality steel, high carbon and some stainless, REALLY is so I can forge some knives and heat treat. Lots of bearings and such. Some I have been able to find out about by checking parts manuals and such.

Is there a readily accessable method to determine exactly what I have? Is there some widget that, say a steel salvage outfit might have, I can hook up a piece and have it tell me "this is O-1, this is 1084, this is..."?

I'm open to suggestions.
 
There is a point-and-shoot device called a "Handheld X-Ray Fluorescence Spectrometer" that is used by some metal sorters nowadays. Here is one example. They will get you in the ballpark of differentiating alloy steel from plain carbon steel, etc. How accurate they are at subtle differences I am not sure.

Of course, the best way to get accurate identification is to send samples to a testing lab. That, however, can be quite expensive.
 
I can help you with the bearings. Just email or PM me the part number and name of the manufacturer and I should be able to identify most or all of them. Most likely the bearings will be a carburizing grade and won't be any good for knife material, but it's at least worth having me check.
 
How large are the roller bearings?
I have found that the larger bearing the more likely it will be 52100 or the like.
I have a hundred and fifty 1 1/4 x 3 inch rollers that I salvaged.

I was told to test them by, forging a tang shape, three or four inches long, from one of the bearings, heat and quench, let it cool then place it in a vice and tap it with a small hammer; if its 52100 it will snap like glass.
Many times you don't need the hammer, it will snap with thumb pressure.

Fred
 
How large are the roller bearings?
I have found that the larger bearing the more likely it will be 52100 or the like.
I have a hundred and fifty 1 1/4 x 3 inch rollers that I salvaged.

I was told to test them by, forging a tang shape, three or four inches long, from one of the bearings, heat and quench, let it cool then place it in a vice and tap it with a small hammer; if its 52100 it will snap like glass.
Many times you don't need the hammer, it will snap with thumb pressure.

Fred

Fred,
In the future, just PM me with the part numbers and I'll be able to tell you without all the testing. The size of the bearing has nothing to do with the material it is made of. It's the application that it is going to be used in that determines what it will be made of. For example, a bearing going into an application where there is shock loading (rock crushers, coal crushers, rolling mills, etc.) will be a carburized grade, while a bearing that needs to carry heavy loads, but without the shock (swing shaft bearings on draglines and shovels in mining, or large pillow blocks on fans and conveyors) will be a through hardened grade.

In addition, Tapered shaped rollers are almost always case carburized no matter what the size. Actually the few through hardened tapered shaped rollers are actually small rollers that are usually not large enough to get a knife out of.

Spherical shaped rollers (shaped like a barrel), could be through hardened or case carburized. The chances are that the smaller ones are through hardened and the larger ones will be split about 50/50 between through hardened and case carburized.

Cylindrical shaped rollers are most likely going to be through hardened, but there are some case carburized bearings out there, depending on the application that the bearing was used in.

Ball shaped rollers are almost always through hardened.

To give you an idea of how size doesn't matter (with bearings that is) a rolling mill bearing can be anywhere from 19" shaft size to 84" shaft size and all are a carburizing grade and not good for knife material. Large pillow blocks and many spherical bearings can be any size and be either carburizing or through hardened.

I misidentified a bearing for Nick Wheeler a while back, which caused me to do some research with our company metallurgist into what steels we use for what bearings. Although the bearing was through hardened as I had told him, it wasn't 52100, but was instead an ASTM485-1 steel. Very similar to 52100 and would be heat treated the same as 52100 for a knife application, but never the less, it wasn't 52100. Based on what happened with the bearing identification for Nick, I now have a great contact to verify what the steel is, and a nice little chart that I can refer to for Timken/Torrington/Fafnir product. It's not a chart that I can give out, since it's company confidential, but I can identify bearing steels on part numbers to assist my fellow forgers.

For Timken/Torrington/Fafnir bearings, I'll always be able to tell you exactly what steel you have based on the part number. For other manufacturers, 99% of the time I'll be able to tell you whether it's case carburized or through hardened, but won't be able to give you the exact metallurgy of the bearing.

When a bearing is through hardened, it can be heat treated identical to 52100, whether it's 52100 or some other version of through hardened steel.

I hope this helps.
 
Last edited:
Thanks that's good to know.

I get them from the big mills down here on the river.

I have been doing this so long that the operators will call me when they are replacing the bearings on a given machine.



Personally, I stick with steel that has papers. That is the easiest way to know what you are working with.

Fred
 
Most of mine don't have parts numbers. The ones that I do have, I found to be 52100.

I have about 50 cylindrical ones that show some signs of oxydation (not stainless), they are about 2 " diameter and 5 inches long. Heavy as hell. A recycling guy client of mine looked at them and told me that he believed they were thoughroghly hardened but he didn't have the equipment to tell me anything.

Out of curiousity, what would it cost to send to a lab? I may be able to get 100 more of these for free, so it might be worth the while for this newbie maker to have some cheap stuff to practice with.
 
Last edited:
I think that a while ago Sam Salvati said that Fastenal will do an analysis for around $50.00. I found a local steel manufacturing plant here locally that will also test for $50.00.
 
Back
Top