HELP! please critique my heat treatment

[knifenewb]

Hey guys,
New bladesmith here. I am making some knives out of files so this is my planned heat treatment. Please critique.

1. anneal files
2. shape blade, bevel, and handle
3. re-anneal finish, formed blade (to relieve stresses formed during shaping and ginding)
4. triple normalize blade checking for flaws between each cycle. if flaws are present they will be fixed and the triple normalizing process repeated until no flaws are present through all three cycles.
5. heat to slightly over non-magnetic and quench in oil
6. temper 350 degrees f for 2 hours, 2 times with one day between cycles

One other question -- if and when grain growth occurs is it reversed by annealing? what methods do you use other than preventing overheating do you use to minimize grain growth?

Thanks for the help guys.

[NRunals]

Looks pretty good to me. As far as grain growth control make sure that if you get it hot you hit it. in other words, don't heat up too large of a section, only heat what you need to work on.

[ThomasPowers]

If anything annealing promotes grain growth it's normalization that draws grain size down. Now I personally would not do step 3 an anneal right before a normalize???

I would also not continuously restart the triple normalization count. Perhaps if the first one came out helical I'd restart the count but after that the tweaks should be quite minor.

Also the industrial standard is to draw temper 1 hour per inch of thickness, 2 hours for something less than 1/4" is a bit much. In fact it's generally the cycle that does the work so cycling it up and down another time would probably do more than to let it sit at temp longer. I draw temper 3 times without much of a sitting period.

What testing will you do? After all it doesn't matter how you do heat treat as long as you get the RESULTS you want.

[MattBower]

I agree with Thomas that Step 3 is unnecessary and maybe counterproductive, and over-normalizing may be counterproductive as well.

I would add that your three normalizing steps should involve descending heats. E.g., normalize at 1600+ (exact temp isn't very significant as long as it's uniform, and well above "critical"), then at 1550, then at 1500 or even 1450. This is probably the most commonly recommended method for dealing with grain growth, and I can tell you from personal observation that it really works.

What sort of oil are you planning to quench in?

Unfortunately file steel is very far from uniform, but the three files I've had analyzed turned out to be quite high in carbon -- 1.2%-1.3%. That presents some problems in achieving a really good heat treat without great temperature control, but I don't want to overwhelm you with info.