Author Message


Posts: 79

Location: United States Conyers, Georgia
Occupation: computer programmer
Age: 64
#1110   2018-01-21 07:25          
I got a lot out of this meeting's presentation by Mark Hopper! His demos are always exceptional and this time was no exception. It was his take on what often is a dry academic topic: heat treating.

First Mark went over the theory by drawing and reviewing a steel state chart (see below). I've seen these charts before, but they are usually temperature vs. carbon content. This one is temperature vs. time, showing how one typical kind of steel reacts.

Here's a description of this chart from

Most of this discussion focused on grain size, how that impacts steel toughness, and how to control it. The short answer is normalize it. Refer to the particular steel's manufacturer's data sheet to see specific recommendations for normalizing temperature and number of times to normalize it.

Normalizing is heating to critical temperature, then air cooling to below 300 degrees. As opposed to annealing, which is heating to critical temperature then cooling in something like vermiculite or wood ash for a slow cooling over night

Forging and machining steel impacts grain structure and creates internal stresses in steel. So work procedure should always be:

  1. forge
  2. normalize or anneal
  3. machine
  4. normalize
  5. quenching
  6. tempering

Testing Normalization
Next was the fun part: exploring grain size base on how steel is quenched and how it is normalized. For this series of tests, Mark used O-1 steel soaked at critical temperature to grow grain size for this test. The same steel was normalized multiple times, and broken each time to expose the internal grain and gauge the difficulty of braking the steel. The details are summarized in the chart below, with pictures of the resulting grain size for each test.

1quenched in water (no normalizing)easy to shatter
2quenched in oil (no normalizing)easy to break
3quenched in oil (one normalizing cycle)harder to break
4quenched in oil (two normalizing cycles)even more harder to break
5quenched in oil (three normalizing cycles)hardest to break

Note the grain size exposed in each broken piece:

At the end of testing, the O-1 steel was tested with Rockwell Hardness testing files and determined to be harder than 65c Rockwell (theoretical limit for O-1 is 68c Rockwell).

Key point: normalizing increases toughness without reducing hardness.

The demo continued with a quick tempering exercise. Since the O-1 sample retained hardness throughout the 3 normalizing cycles, it was not necessary to raise it to critical temperature again. The process used was:
  1. polish the surface to be able to see the tempering colors
  2. heat with a propane torch, lightly going back and forth on one end until the tempering colors begin running
  3. quench in water once a full range of colors are visible

Testing showed a total 10 points Rockwell C hardness change from no color (65c) to softest side (55c).

Key points: once normalizing is complete, the tempering process can continue without rehardening, and quenching in the tempering process can occur in water, since the metal was not heated to critical.

Useful tools
Mark recommended downloading a heat treater's guide app to refer to specific heat treating details by type of steel. This may be found at:
for Android
for iPhone and iPad

Added 4 minutes later:

In the forging process:
  1. forge
  2. normalize or anneal
  3. machine
  4. normalize
  5. quenching
  6. tempering

I was thinking step 5 "quenching" was going through a full hardening process (heating to critical, then quenching. But that's not what Mark did. This must mean plain old cooling it down. Especially since one of the key points he made of normalizing is that hardness is not affected by normalizing.

Does that make sense?

This post was edited by jguy (2018-01-21 13:32, 978 days ago)
Jim Guy
Page created in 0.032 seconds