Do I need to Normalize

[kevLaur]

I've been working on my first knife made from a old file for a while now and have got a bit confused reading from many different sources. I annealed the file overnight in a very hot fire, then cut,ground and shaped my blade. I belive it is ready for heat treating which I will do in a coal forge but need to know weather it needs normalizing first. Isn't that what the orginal annealing did?

[Steve Sells]

maybe.

[bigfootnampa]

Actually the annealing might lead to enlarged grain size which is where the normalizing could be useful. It will not hurt to normalize once or twice before hardening as a safety. Annealing softens the steel in part by the growth of larger grain which also weakens the steel... so you don't really want that in a finished knife. Hardening might correct it adequately... but why not be sure? Steve is really FAR better qualified to answer this than I am but he wants you to read his heat treat stickies FIRST! Not a bad idea to do that BTW.

[Steve Sells]

Thank you bigfootnampa.

[kevLaur]

I did read the sticky, but it takes longer to wrap my noggin around things then it used to. I do not know why I have been resisting the normalizing since it will be in the fire anyway I will give it two or three heats before quench . THANKS for pushing me along!

[bigfootnampa]

YW Steve. Well now Kevlaur; I hope that you will be quenching the steel as part of each normalizing cycle as well as for the hardening quench? The only real difference is that you don't take the metal to as high a temp before quenching. If you do NOT quench when normalizing you could still end up with larger than ideal grain... unless you get enough air quench to do the job... possible with some steels and smaller, thinner blades. But quenching is a more reliable way to control the cooling so that you resize the grain of the metal to get a stronger blade. I hope it turns out well... let us know, if you please.

[ThomasPowers]

And it depends on the alloy too; for S-1 the ASM handbook does NOT suggest normalizing. For A steels it's decidedly difficult to "normalize"

[Frank Turley]

For plain carbon steel, "Quenchcrack" a metallurgist for Houston, says that when the steel is annealed, it retains "large carbides" which are undesirable. Therefore, he suggests normalizing which is a better preparation for austenitizing.

[Woody]

Here we go again, when did quenching become part of the normalizing process? I just picked this up off a google search.


Normalizing: Carbon steel is heated to approximately 55 °C above Ac₃ or Ac_m for 1 hour; this assures the steel completely transforms to austenite. The steel is then air-cooled, which is a cooling rate of approximately 38 °C (100.4 °F) per minute. This results in a fine pearlitic structure, and a more-uniform structure. Normalized steel has a higher strength than annealed steel; it has a relatively high strength and ductility.^[13]

Quenching: Carbon steel with at least 0.4 wt% C is heated to normalizing temperatures and then rapidly cooled (quenched) in water, brine, or oil to the critical temperature. The critical temperature is dependent on the carbon content, but as a general rule is lower as the carbon content increases. This results in a martensitic structure; a form of steel that possesses a super-saturated carbon content in a deformed body-centered cubic (BCC) crystalline structure, properly termed body-centered tetragonal (BCT), with much internal stress. Thus quenched steel is extremely hard but brittle, usually too brittle for practical purposes. These internal stresses cause stress cracks on the surface. Quenched steel is approximately three to four (with more carbon) fold harder than normalized steel.<sup>[

Normalizing


1 Heat the furnace or kiln to 1,500 degrees Fahrenheit and put on your protective eyewear and clothing. (See Reference 2)</sup>

• ^{2 Calculate the amount of time you will need to leave the steel in the furnace. You can do this by allowing an hour for every cubic inch of steel you are normalizing. (See reference 2)}
• ^{3 Place the steel in the furnace and leave it for the calculated time.}
• ^{4 Remove the steel from the furnace once the time is up using the heat-resistant tongs and leave it to cool in the open air.}
  

^{Normalizing is the process of raising the temperature to over 60 º C (108 ºF), above line A₃ or line A_CM fully into the Austenite range. It is held at this temperature to fully convert the structure into Austenite, and then removed form the furnace and cooled at room temperature under natural convection. This results in a grain structure of fine Pearlite with excess of Ferrite or Cementite. The resulting material is soft; the degree of softness depends on the actual ambient conditions of cooling. This process is considerably cheaper than full annealing since there is not the added cost of controlled furnace cooling.}

The main difference between full annealing and normalizing is that fully annealed parts are uniform in softness (and machinablilty) throughout the entire part; since the entire part is exposed to the controlled furnace cooling. In the case of the normalized part, depending on the part geometry, the cooling is non-uniform resulting in non-uniform material properties across the part. This may not be desirable if further machining is desired, since it makes the machining job somewhat unpredictable. In such a case it is better to do full annealing

[bigfootnampa]

Woody you are correct... BUT, normalizing by oil quench is considered a way to get faster and more thorough results. So by oil quenching after heating to normalizing temp you would get more refined (smaller) grain than by air cooling. This is a bit more advanced technique... but still easy to do. If you google a bit more you'll find some references to this also. Of course somewhat dependent on specific alloys, but useful for most low alloy or plain carbon steels.

[Steve Sells]

thermal cycling for grain refinment is good, but it is not the same as normalizing.

[Woody]

Like I said, "here we go again" seems we went through this about last December or so.. I am well aware of "more advanced techniques" however trying to use the same name to identify two completely different processes only creates confussion. If you quench the steel you end up with a comletely different grain structure than if you air cool to ambient temperature. Perhaps what you are alluding to is defined below:

Metallurgy of Steel for Bladesmiths & Others who Heat Treat and Forge Steel
John D. Verhoeven Emeritus Professor Iowa State University

New Grains formed by Phase Transformation
For simplicity consider first a pearlitic steel. At room temperature there will be some average pearlite grain size. When this steel is heated above its Ac1 temperature
austenite grains will begin to form. The new austenite grains start to form on the old pearlite grain boundaries as shown schematically in Fig. 8.4.
After a short time all the old pearlite grains are replaced with a completely new set of austenite grains. The new austenite grains have their smallest
size immediately after the pearlite is consumed, before significant grain growth occurs as the temperature rises and time proceeds. Following are
two factors that enhance the formation of the smallest possible initial austenite grain size.

(1) Faster heating rates cause the austenite grains to nucleate closer together and enhance small grain size.

(2) Smaller original pearlite grains produce smaller austenite grains.

When the austenite is cooled back down below the Ar1 temperature a whole new set of pearlite grains is formed and the same two factors: rate of transformation and size of the prior grains, control the size of the new grains. Hence, on simply heating and cooling through the transformation temperature three different sets of grains are involved. When dealing with hypoeutectoid steels the same ideas apply only now one must heat above the Ac3 temperature before 100 % austenite is formed. If one wants to reduce the grain size of a ferrite/pearlite steel, a simple technique is to heat into the austenite region holding the maximum temperature as low as possible and the time as short as possible and then rapidly cooling back to room temperature as fast as possible without forming bainite, (see my note below) assuming it is not desired. Since smaller initial grain size promotes smaller final grain size, repeating this cycles several times will enhance fine grain size. When heat treating to form martensite, toughness is also enhanced by fine grained austenite because it results in a finer lath or plate size in the martensite. Again the same
ideas apply. Rapid heating and repeated cycling produce smaller martensite microstructures. Grange [8.3] has presented a study showing the beneficial effect of
small austenite grain size on the mechanical properties of 8640 steel. He achieved grain sizes in the ultrafine range of ASTM No. 13 to 15 by a 4 cycle process where the steel was austenitized in molten lead for around 10 s, cooled to room temperature, cold worked and then cycled again. A series of similar experiments was performed here on 3 steels to examine the effectiveness of thermal cycling alone, no cold working was employed. The steels were heated by immersion in a salt pot. Initially the steels were austenitized for 15 min. at 1650 oF and oil quenched in rapidly stirred oil. Then the steels were given 3 thermal cycles consisting of a 4 minute austenitization in 1450 oF salt and a quench in rapidly stirred oil. The grain sizes were measured with the same technique described by Grange [8.3] and the ASTM numbers before and after the 3 cycle treatment are given in Table 8.2. It is seen that ultrafine grain sizes were obtained. Figure 8.5 presents photomicrographs of the martensite structures found in the 1086 steel before and after the cycling. The composition of this steel is in the range where we expect the martensite to
be a mixture of lath and plate morphologies, and in the uncycled coarser grained sample Fig. 8.5 (A), one can see dark plates in a matrix of
the lath structure. However, in the finer grained austenite produced by thermal cycling, Fig. 7(, the martensite structure is clearly finer and the plates are not easily identified. Table 8.2 ASTM grain size no. of austenite before and after the 3 cycle treatment done here.

Steel type Grain Size Before Grain Size After

1045 9 14

1086 11 15

5150 8.5 14
Note: I have looked it up and to cool without forming bainite requires cooling to less than 200 C in less than 10 seconds.

[bigfootnampa]

Well sorry if I have confused anyone. I tried to be clear. I am following (recommending) procedures that have worked for me and that were referred to as "Nomalizing by oil quench" by ABS master bladesmiths Howard Clark and Hank Knickmeyer. http://forums.dfoggknives.com/indexphp?showtopic=21043 < Link here I guess we could come up with a different name for the process... it is the only way that I normally use to normalize so it seems routine to me.

[Woody]

I just love those master bladesmiths, Ed Fowler was at the ABANA Conference here, he is still advocating edge packing although he didn't call it that, just that you never hammer on the spine of the blade when forging only on the edge to break up the the grain structure and make it smaller and a new one on me he called "flash annealing" where you hold the hot iron in a darkened area until it goes back to black and then flashes, do this twice and then edge quench. I think it is sort of like the thermal cycling Steve mentioned, this is after the blade has been differientally quenched 3 times. He also advocated forging 52100 at a temperature so low it scared me. He forges it just above a black heat to avoid growing the grain. To my way of thinking that is dancing around in stress crack city but it seems to work for him. He never did demo forging a blade, just his method of heat treating blades. His quench tank was an electric frying pan with about a half pint of oil which he said must be maintained at 160 degrees F. the blade was triple quenched, then "flash annealed" twice and quenched again. Unfortunately I missed Resroat's demos because of conflicts with doctors appoinments (I should have skipped the doctor's appointments, the cortisone injections did nothing) and helping in the Art Gallery.. Dr West from the School of Mines did an excellent 2 hour lecture on myth busting some of "ideas" in blacksmithing and knife making. He thoroughly explained the process of heat treating and grain growth. I wish I could remember it all, fortunately I have access to him on a regular basis. His lecture was one of the best things at the conference.

[thingmaker3]

thermal cycling for grain refinment is good, but it is not the same as normalizing.

Can you expand on that, Steve? I'd like to read your thoughts on any advantages provided by normalizing but not by thermal cycling with quenching.

[pkrankow]

...where you hold the hot iron in a darkened area until it goes back to black and then flashes

recalescence.

Strange, but cool, effect. Recalescence is cited as one of several means of determining the correct temperature to quench from for simple high carbon steel in old textbooks. It is caused by the phase change of the steel causing a release of energy so the temperature slightly increases.

Phil

[Steve Sells]

Can you expand on that, Steve? I'd like to read your thoughts on any advantages provided by normalizing but not by thermal cycling with quenching.

I would be happy to, tho I am not sure of what your question is. normalizing is to relax the stess from our steel. Thermal cycling is when I heat to AC3 then air cool till I see the "Flash" spoken of earlier, indicating the transformation temp. then reheatingt again, They both serve different purposes.

If we dont see the shift in color, it thats ok, the point is to cycle between above AC3 and to below transformation re defining the grain size, 3 or 4 times sems to be plenty in most cases to get the grain to a reasonable small size.

As for the ABS MS ... I have known a few that were very good smiths, Sadly most are so full of themselves they cant see straight. I have one tell me fist thing I need to do is go their beginner course for a week of training, becaue obvously I didnt learn correctly, when asked to explain he replied that only they know the secrets of proper forging... even tho this so called Master never even looked at any of my blades., they seem only interested in the money for that basic beginner class.

[bigfootnampa]

The few ABS MS that I know are pretty cool guys and seriously skilled smiths! Recalescense is new to me! Thanks Phil! Somewhere I saw a neat demo where blades were broken and micro-photographed to show the grain sizes as they were cycled through three heats... I thought it was here but can't find it now. Anyway that was a pretty convincing demo! I do notice that the ABS testing procedure is very specialized toward the making of patterned steel and heat treating it to obtain very specific performances... interesting tests but not fully relevant to good blade performance in most cases (IMO).

Steve, I think what thingmaker3 wants to know is if he is going to cycle a blade through 3 heats to refine the grain... will there be any reasons to also normalize that blade either before or after the thermal cycling treatments? It seems like a good question and I also would like to know what your thinking is on that. Personally I figure that the thermal cycling alone will be adequate... but what do I know?

[Steve Sells]

I normalize most everything I forge, soon as I finish forging. to avoid warpage, and stress fractures. Just how I was taught, and isnt hard, or take too much time as I was just forging anyway, so one more heat is easy and while I am normalizing things it is a great time to correct any twisting.

I do the thermal cycling after normalizing and before hardening, if I did an aneal after that it allows grain growth that would defeat the cycling I just pereformed. adn anneal does not allow as much grain growth, but I do not believe it is needed again just after an cycling session.

[basher]

I am confused, lots of different similar versions of the same thing.
so speaking from personal experience (just to add more confusion to the mix!)....

I would only ever anneal, Ie slow cool in an insulated box from critical temp (otherwise known as austenising temp, around 800C for most simple carbon steels) If I wanted a material in a state that It could be machined easily.
My usual preference in such cases would be to quench the material from critical and then give it a sub critical anneal, which is another name for a very high temper where heat the steel to around 750C or just below its austenising temperature. It is often easier to achieve a machinable steel this way without the complicated ramp down Ht controls needed to do a full anneal.
I tend to normalise to reduce grain size.....Heat material to just above critical and allow to air cool I do this 3 times normally. I will often normalise twice before grinding and leave one normalisation for before hardening . I work big blades and the last normalisation is a good chance to straighten and helps reduce stress build up from the grinding.
When I teach I always heat a file in the fire , quench and break it and then normalise the same file and quench and break it again. Grain reduction from the first normalisation is massive and easy to see (sugary grain to silk) the second and third normalisations are a lot more subtle visually, to the naked eye.
The reduction returns with each normalisation are decreasing and that is why 3 normalisations has become a modern standard for bladesmiths. there are variations on this like descending normalisations (which I use for blades that get a clay quench) and multiple normalisations for reducing hardenability, but we wont go into them.

I have in the past quenched from critical to below martensite start point to obtain grain reduction . I no longer do this as I do not honestly see any advantage in the simple steels I use. there would be advantages in some steels 52100 , L6 possably 5160 where the un advantageous formation of large carbides during an air cooled normalisation are avoided by fast quenching to martensite.
Quenching is a stressful thing to put steel through in the thin sections used by a bladesmith and I have a reluctance to do it 4 times to one blade.

I have seen no evidence that quenching to martensite gives finer grain reduction than normalising in simple carbon steels. I would be interested if there is such?

Heat treating can be confusing especially as the terms and accepted practice changes from person to person and over time. It is worth noting that the changes are for the better. it just gets a little confusing if you find outdated books , information or smiths/bladesmiths before you find modern good information.

I have learned at least 4 versions of good heat treating practice over my 18 years each one improving on the last.

It seems strange in hind sight that blacksmiths and bladesmiths have such problems with information that has been standard in industry for so long.

[Rich Hale]

• Basher than you so much for your thoughts: Each time I read this thread i shake my head and wonder just how messed up the head of a person reading this can get by the things entered. i stick with simple steels and use the methods I have fouind that worked. i know this as I have also broken files and looked to see wot i have done and can do with grain structure. i aslo do the same with the blade steels I work with, it is a good investment of blade steel and time to take a piece of the same size as the finished blade and open it up after each process and see wot has been done internally. I also feel that a blade should be tested for flex, cutting ability and how its edge holds up with a lot of use. If a blade maker cannot test for these items they are just guessing as to wot they are really making. A log book to track all of this informaion is vital Just as important to know wot has worked so you can repeat it years latere if you use steel that youi have not used for a long time.
  

[WmHorus]

Long conversations about certain topics make my head hurt....Dont get me wrong they are full of great info but it still gives me a migraine.

[thingmaker3]

I would be happy to, tho I am not sure of what your question is.

I apoplogize for my ambiguity. To rephrase: why does recrystalization relive stress when done only once but not when done multiple times? How do the subsequent recrystalizations manage to re-stack the dislocations eliminated by the first recrystalization?

[Steve Sells]

I apoplogize for my ambiguity. To rephrase: why does recrystalization relive stress when done only once but not when done multiple times? How do the subsequent recrystalizations manage to re-stack the dislocations eliminated by the first recrystalization?

I never said it didnt, I mearly posted that thermal cycling is not the same as just anealing for stress relief. Of cource the cycling relieves stress. I was pointing out I felt that allowing a slower cooling also means longer time at the high tems and can allow some grain growth, since my intend in cycling in grain reduction I dont let it sit too long before heating again when I am cycleing, I didnt intend to imply anything other than that.

[thingmaker3]

Thank you, Steve. That makes sense.