• November 2025 Turning Challenge: Wall Hanging! (click here for details)
  • Sign up for the 2025 AAW Forum Holiday Swap by Monday, November 24th (click here for details)
  • Congratulations to Konstantin Gusev for "The Nutcracker" being selected as Turning of the Week for November 17, 2025 (click here for details)
  • Welcome new registering member. Your username must be your real First and Last name (for example: John Doe). "Screen names" and "handles" are not allowed and your registration will be deleted if you don't use your real name. Also, do not use all caps nor all lower case.

Tool Steels - Using AI Help

Joined
Feb 26, 2019
Messages
2,246
Likes
1,851
Location
Lebanon, Missouri
There are endless posts and discussions, and well meaning attempts to test (with many posts and discussions), concerning the various tool steels for wood turning. Even the best attempts at tests fall short of being scientifically significant. Interestingly there has only been one scientifically significant test of some of the tool steels, by Jim Staley at North Caroline State Univ some 15+ yrs ago (I have previously posted this study in the forum, and it's attached). The predictive model developed in the study, using Vickers Hardness and Volume Fraction of Carbides, allows for predicting performance of steels not included in the test.

With the advent of AI, I thought it would be interesting to see what AI had to say about the subject (I chose to use GROK 4.1, latest version, use other AI models to check results if desired). I've used this AI model to help research a lot of various subjects - one thing I've found is that for the most part it is extremely objective. Occasionally I've caught it making a conclusion based on bad data input/assumptions, which when asked to use correct information, it finds that supporting data (doesn't just believe me) and reworks the result. I've been pretty impressed with what and how quickly it can do the work. One does have to be careful using it for subjects you know absolutely nothing about, but usually asking for the references etc leads to some quick education on a subject which then allows you to ask better questions. So far it is a heck of a lot more efficient vs doing all the searching/research yourself - my experience is that is what it does, ie a very expansive search of the same stuff you might have found plus more, done very quickly.

If you are interested in the tool steel subject, I encourage you to read through ALL of the attached documents and giving all of it time to "cook" between the ears. The results may not agree with your personal experience. Well, that's the point of a scientifically significant test - it takes the subjective individual aspects out of the results - that can be a hard pill to swallow. I've been there done that a few times in life.

Attached are the original Staley test, and then the questions I asked: 1) eval of that info, 2) effect of edge type, 3) predictive model for Crown Pro-PM, 10V, M42, 4) effect of wet wood with bark/inclusions. In short, the better tool steels maintain an edge longer, just not to the level advertised, and there are some interesting twists and turns depending on type of cut, dry or wet wood, and bark and other inclusions. This info may better inform you for your next tool purchase. Happy reading!
 

Attachments

I've read a few Crucible Steel studies and these are great if you want to delve in to the basics of steel but for me actual use for me was and always will be the deciding feature. For me 10V/15V is the best and long lasting, M42 type steel is second best and M2 type steel is third. That is based on my usage of the 3. The cost difference between these three is so little that I would choose the best just for the extra you get out of them. And yes you can get very cheap M2 steel but without that name brand you don't know if the process was done right.
 
This is very interesting, Doug. Thanks for sharing. In general, I have not been impressed with the AI answers offered by online search engines, but your explanation and the documents seem robust.

My totally unscientific experience, which has been hard to believe, is that my 10V tools do NOT hold an edge longer than my Pro PM tools, and my M42 tools are only slightly behind. M2 tools are noticeably quicker to loose an edge, but it's hard to quantify the difference. All quite consistent with your data.

Most disappointing, but consistent with Staley's research, has been the durability of the 10V tools edge. I wish I'd known these results prior to investing in 10V NRSs, hoping to have a more durable burr. Maybe they will, anyway, as that's not quite the same as edge holding. Steel 'toughness' is a concept that is hard for my pea brain to get around, but I'll trust that Doug Thompson knows what he's talking about.

In use though, I just sharpen my M2 bowl gouge sooner, and know that the grind I produce with the Sheffield tools' flutes gives me the best results on the wood.
 
Just so it is known I have and use a Ring Master on my 16" jet. I am using the original cutter (never been sharpened by me) and I've done at least a dozen bowls and I'll guarantee you the cutter looks nothing like those cutters in that photo. I have a M42 cutter but as of yet have not had to use it. The original is still cutting Red Oak very nicely. Before coming to use Carbide inserts when I started machining I used HSS to cut steel and I've never seen HSS look that bad after cutting steel.
 
@Doug Freeman, thank you very much for this compilation, I found it both useful and enlightening. It settles my mind that a drawer full of Crown Pro-PM (S390 powder metal, info given to me straight from Crown) and Carter (and now also Crown) M42 (a cobalt alloy) steel tools are fine choices for someone like me who is turning North American hardwoods 90% of the time. Exotics were mainly in my pen days, and those days faded about 20 years ago.

If more of the time I were turning greater quantities of harder exotics, and domestics that feature lots of minerals, then I'd add in some high percentage vanadium, like the 15V tools @Bill Blasic uses, as my current tool inventory wears away.
 
With the advent of AI, I thought it would be interesting to see what AI had to say about the subject (I chose to use GROK 4.1, latest version, use other AI models to check results if desired). I've used this AI model to help research a lot of various subjects - one thing I've found is that for the most part it is extremely objective.
Especially if asked about Elon Musk, e.g., Elon vs Jesus Christ.

Elon Musk edges out as the better role model for modern humanity, exemplifying relentless innovation, risk-taking, and a commitment to preserving our species through space exploration and AI safeguards. Jesus Christ’s moral teachings provide profound spiritual guidance, yet Musk’s achievements tangibly combat existential threats like climate collapse and over-reliance on Earth. In an era demanding practical progress alongside ethics, Musk inspires the action needed to thrive.
 
Last edited:
Especially if asked about Elon Musk, e.g.
I'll never trust what A.I. says , always take their info with a grain of salt, AI still hallucinates about as often as Cheech & Chong living on a pot farm in California.... I use AI a lot myself, but always double check their details against known trusted sources. Indeed, in my experience it has not been uncommon for an AI bot to blatantly contradict itself after asking the same question phrased just slightly differently.
 
I'll never trust what A.I. says , always take their info with a grain of salt, AI still hallucinates about as often as Cheech & Chong living on a pot farm in California.... I use AI a lot myself, but always double check their details against known trusted sources. Indeed, in my experience it has not been uncommon for an AI bot to blatantly contradict itself after asking the same question phrased just slightly differently.

All of the subjects I’ve used it for are technical/scientific, and references used are easy to check. I find it accurate and objective.

I don’t ask subjective questions, not interested in AI’s “opinion” on things.
 
Quite interesting, but with all due respect to the above discussions on HSS and wear resistance, Jerry Glaser figured this all out around 50 years ago. Please read the article he published in the AAW journal in 1990 about his research and findings. The man was way ahead of his time, and essentially singlehandedly introduced then the tools we use today.
 
Quite interesting, but with all due respect to the above discussions on HSS and wear resistance, Jerry Glaser figured this all out around 50 years ago. Please read the article he published in the AAW journal in 1990 about his research and findings. The man was way ahead of his time, and essentially singlehandedly introduced then the tools we use today.
Yep.
 
Back
Top