Confessions of an Over-Heater: advise needed

[P.Brewster]

I finally built a PID-controlled oven after using a forge for HT'g. I work with 1075 and W-1; I quench in Parks #50.

The forge was too small and too un-even to be ideal, but I took my time and made it work.

I was aware that my grain size was on the large size (from seeing broken throwers), and I'm aware that the main cause of this is over-heating. However, I thought my heat-treat process was under control for the most part.

The first time I used the oven, I was thrown off base.

I initially placed the thermocouple relatively close to the heating elements. Most of the weird things I observed (such as 20° swings in temperature, and blades that were still magnetic after 20 mins at "1480°") ceased after I finally moved the thermocouple to the door, away from the elements.

Besides that, I observed that the steel was what I would call a subdued red, and barely non-magnetic (after 20 mins in a pre-heated 1480° oven). With the forge, I quenched when the steel was a self-luminous bright red; completely non-magnetic.

I also noticed that the quenching was much less violent with the oven than the forge.

Now, I've pretty much concluded that I was simply over-heating the steel in the forge. However, at the time, I suspected that the PID or something else was faulty, so I utilized the two basic tests that were immediately available.

I HT'd some bad blanks and bar (no tempering), broke them to examine the grain, and did the dubious file-skating test.

My standard practice is to remove the steel from the quenchant at 10 to 12 seconds to straighten. Then I let them air cool. I use the flash-point of the oil to estimate 400° or so... my infrared thermometer said it was closer to 300° but it is not a reliable thermometer.

So I did the same thing with the test pieces. I'm not sure if this has a significant tempering effect; nor am I sure if the technique has a "name".

I observed that the W-1 would skate the file 100%, and it would snap cleanly when clamped down and smacked with a hammer.

However, I observed that the 1075 would skate a file to some extent, but not 100%. Also, it would not snap cleanly. It would bend approx 10°, then snap. I confirmed this with several blanks, stopping when my 3-lb hammer blows ripped out the bolts that mount my vise to the wooden workbench.

I really thought that the 1075, despite the lower carbon-content, would snap cleanly and skate a file, without a temper.

The grain looked desirable (as far as I can tell). I am aware that the file test simply tells me if the steel's surface is harder/softer than the file itself.

The pic shows broken oven items on the left; a forge item on the right.

So my questions are:

Does it sound like my HT process is decent? In other words, do my observations seem normal, or do they seem like something is still not right?

thanks!

 

[butcher_block]

milky looking grain is what you would be after
you should see CPM steel when its jsut right

if you have an HSS mil bits that are shot brake one of them and see how it looks

 

[JasonSCarter]

Patriqq I use W1 alot and I have a back up low tech thermometer to tell me when it's at quench temp...Table salt!... it melts at 1475ºF so what I do is put a line of it on my work and when it starts melting like snow it's ready for quenching. I dont do an interupted quench either, though, I put mine in the oil at about a 30º down angle, perfectly straight. you wont get any warpage doing it this way. I hope this helps

Jason

 

[LRB]

Patrick, just a little tidbit, FYI. The file test does NOT tell you that the steel is harder, or as hard as a file. A common mill file will not cut steel over 58/59 Rc. It skates beyond that hardness. If you get the opportunity, try the file test on a blade that you know is around 60/61Rc , and you will find this true. Most of us do not have any better method to test the hardness out of quench, but be aware that is as accurate as a common file gets.

 

[GHEzell]

Thanks for that Jason, I'll have to try that....

 

[Don Hanson III]

Pat, you're on the right track!

Was a big eye opener years ago when I discovered I was overheating blades in the forge.

 

[birdog4]

Patriqq, this is called "experience"

 

[MSCantrell]

Patriqq I use W1 alot and I have a back up low tech thermometer to tell me when it's at quench temp...Table salt!... it melts at 1475ºF so what I do is put a line of it on my work and when it starts melting like snow it's ready for quenching.

WHAT?!?! That's so elegant I can't believe we don't all do it.

What's the catch, where's the downside?

 

[69_knives]

This is taken from Kevin Cashen's website:

Anybody who gets really obsessive/compulsive about this whole metallurgy thing will eventually need to have a look inside of their steel and see the microstructures that they are always reading about and pursuing. Martensite is hard and will skate a file, but the same is true of very fine pearlite, or Bainite, loaded with carbides, the Hardness tester can’t tell you the grain size that is just as important for performance.

There's also the infamous decarb that won't skate a file but you may still have a hard blade underneath.

 

[james7729]

Thanks patrigg, I'm just getting ready to use my sugar creek oven the first time this weekend. Helps to hear about your experience.

 

[fitzo]

Patriqq I use W1 alot and I have a back up low tech thermometer to tell me when it's at quench temp...Table salt!... it melts at 1475ºF so what I do is put a line of it on my work and when it starts melting like snow it's ready for quenching. Jason

WHAT?!?! That's so elegant I can't believe we don't all do it.

What's the catch, where's the downside?

The melting point of 1475F is for very pure sodium chloride.
Any time a compound has impurities present, the melting point is lowered.
Usually, common table salt isn't pure sodium chloride, but is often iodized and/or has other constituents added to make it flow properly in a shaker.
Any adulterants like these will lower the melting point. Just how much is difficult to predict offhand.
Thus, if one wants to use salt melting temperature as an accurate indicator, one needs chemically pure sodium chloride to get the best accuracy.

 

[James_Terrio]

Thus, if one wants to use salt melting temperature as an accurate indicator, one needs chemically pure sodium chloride to get the best accuracy.

There's probably a website out there somewhere that lists melting points of various common compounds. Then google for a chemical supplier in your area. I used to work for one and we sold small amounts (like 1/2# or 1# plastic jars) of various things for dirt cheap. For instance, oxalic acid crystals to make a rust removing solution, pure sodium hydroxide (lye) for drain cleaner, etc. A good supplier will be able to tell how pure the compound is. You may have to ask to speak to the house chemist or warehouse foreman, not just the minimum wage person at the front desk. Ask for a copy of the MSDS and if they feel the compound is safe to use in such a manner.

 

[P.Brewster]

thanks guys

 

[Carcara]

Tempil stick crayons are made that melt at 1450 and 1500 degrees, They are
rated at + or- 1 percent accurate to the target temp. Perhaps a better low tech solution than buying reagent grade sodium chloride.

 

[JasonSCarter]

The melting point of 1475F is for very pure sodium chloride.
Any time a compound has impurities present, the melting point is lowered.
Usually, common table salt isn't pure sodium chloride, but is often iodized and/or has other constituents added to make it flow properly in a shaker.
Any adulterants like these will lower the melting point. Just how much is difficult to predict offhand.
Thus, if one wants to use salt melting temperature as an accurate indicator, one needs chemically pure sodium chloride to get the best accuracy.

Thanx for the chemistry lesson, Fitzo always good to have sound science up front... so I guess this is the downside MSCantrell, But your right it is an elegant solution. I asked the Lord for a low tech hi temp thermometer solution and thats what he gave me. I'll check the temp with my multimeter and make sure it's right...but I havent had it fail me yet. and the blades I've used as sacrificial learning curves all had that milky fine grain to them. I wonder if the time difference of when it just starts and when it starts to melt FAST, which is when I start pulling it out of the coal fire the temp doesnt get to where it should be...

Jason

 

[fitzo]

Jason, when I saw you write in that other thread, "...when it starts to melt FAST...", I thought "excellent".

If you wait until it is at the end of melting, it should be very close to that 1475. Melting point is usually expressed as a range of where it starts and where it ends. The purer the compound, the narrower that range, and the closer to that number you'll find in the literature.

And, I agree with Matt, it's a good lo-tech solution. I have long suggested to people starting out pattern welding to use the vigorous bubbling of borax flux as a good approximation of 1900F, the lower end of the welding range. Pure chemicals have pretty reliable physical constants.

I simply wanted folks to understand that some salt will be better than others for accuracy, and once you have excellent results, hang onto that pound of Morton or whatever. It sounds like you have worked this out for your needs quite thoroughly, and know what works in your situation. I applaud your ingenuity.:thumbup:

 

[JasonSCarter]

Thanks Mike, I really appreciate your comments and thoughts on this. Your warning that all salt is not made equally is most helpful and something that should be taken into account for.

Jason