1. Welcome to the New & Improved BladeForums. New software info here. Please report problems in Tech Support, and read existing threads before posting! - Spark
  2. I've changed the default forum style to Flat Awesome based on feedback. Don't like it? Click here to change how the forums look Feedback on this is welcome here.

normalizing vs annealing

Discussion in 'Shop Talk - BladeSmith Questions and Answers' started by smorgan345, Oct 5, 2010.

  1. smorgan345


    Mar 8, 2007
    I'm working my way through the "working the three steel types" sticky, very informative. I'm wondering though, what is the difference between annealing and normalizing and when exactly is each used? I will only be doing stock removal at this time. Also, what is spheroidizing? Sorry for any spelling errors. Thx
  2. Remyrw


    Jun 17, 2010
    Annealing is a slow process used to soften a steel to an "easy" to work condition. Often the steel we buy is already annealed.

    Normalizing is when you heat it up to let the steel relax and release any stress that built up from getting shipped all over the place. It's not easy being steel ya know. Rough life with all that pounding and pouring... Seriously though, normalizing is a stress relieving process that also helps make sure you don't get cracks or warping during HT. I normalize after profiling and mostly flat grinding the blank. Then I finish flattening it and grind the blade. I normalize again prior to HT, basically as part of the process, checking before doing the actual HT that it's still looking right.

    I don't think I really need to do both, probably normalizing before HT would be sufficient, but I haven't done enough to test that theory.

    Spheroidizing is annealing on steroids and makes for even softer and more ductile material. In practice, when tempering we are doing a very limited form of this to reduce the hardness of the steel to a practical level and relieve stresses that built up during quench.

    If I've got any of this wrong I'm sure one of the more experienced folks will come educate both of us. :)
  3. mete

    mete Gold Member Gold Member

    Jun 10, 2003
    Normalizing does relieve stress but it also makes the steel 'normal' in that it goes into the austenite range and in the process makes the structure more uniform throughout the piece. What you need to do after forging !
    Stress relieving also can be done by heating to 1200 F for a couple of hours .This does not form scale as normalizing will. With longer times the martensitic structures will turn into spheroidized carbides --for a 'sub-critical anneal.
  4. smorgan345


    Mar 8, 2007
    Thanks guys. If I understand right, in both normalizing and annealing, you heat the steel to just below critical and let it cool slowly? I guess specifically Im wondering what the theoretical differences are between the two activities. Is the temp or hold times or cooling rates typically higher for one than the other? Normalizing is just stress relief and annealing is actually changing the molecular structure? I'm swinging at the air with my terminology, bear with me.

    I'm using stock removal and not forging so is it still necessary to normalize before performing my heat treat?

    One more thing while I have your ear, I am plannnig on getting a heat treat oven to do my heat treats. It's important to me to get the optimum performance from the steel and I've decided this will be the best way to do that for me. But with an oven, the entire blade gets the same heat treat. Is that good? There is no way to differentially heat treat with an oven I suppose. Wouldn't I want the different parts of the blade to at least temper differently? ie. tang and spine vs cutting edge? Thanks!
  5. Leatherface


    Dec 3, 2005
    use a limiter plate if you want to differential the HT...but I would just heat treat the blade as a whole and keep the handle and tang out of the oil
  6. Remyrw


    Jun 17, 2010
    Normalizing is pretty much up to temp, let it cool back down in open air. Most folks do it in a set of two or three with decreasing temps each time. Up to (just picking a number) 1400, cool off. Up to 1300, cool, 1200, cool. You don't bring it down slowly like with annealing, you just don't quench it. This gets tricky with air cooling steels, so you have to read up for each steel and find out what the exact rate is. I use 1080 so I literally heat the piece up and take it out of the heat and stand there with it for a couple minutes then repeat at a slightly lower temp, then again. I also do stock removal, and normalizing is still worthwhile. It's not AS critical, but it's worth doing.

    Normalizing is changing the molecular structure, but not completely. What Mete was saying is that it's evening it out so you have a uniform structure internally to work with when you do the HT. Annealing is generally a much slower cool down and is specifically meant to give an even and soft structure to the steel. Normalizing is not meant to change the hardness, though you might see some change. Also, with normalizing, at least the first time in a set you take it past critical, not far, just barely past. Otherwise you didn't accomplish the normalizing and just got the stress relief portion of the process.

    As for differential HT, I'm not sure but my understanding is that you can still do differential HT's. Keep in mind, the HEAT, Quench and Temper stages are three steps of a process, how you do the heat doesn't have to change how you do the other two. Think about it this way. If you uniformly heat the knife then only quench the blade then the tang just got a normalization run of whatever cycle you ran for the HT. It will not be hardened because it didn't go through the quench process. Of course, steels with a very slow quench might not work so hot if you don't want a hardened tang.

    In most cases a hardened tang isn't a big deal, this is how most production knives are done.
  7. me2


    Oct 11, 2003
    Normalizing is always done above the critical temperature. Different types of annealing are done both above and below critical temperature. The textbook definition of normalizing is above the upper transformation temperatures, eliminating all the carbides or ferrite, depending on the steel type. Annealing typically has a slower cooling rate. Normalizing uses an air cool. Normalized steel typically has a higher strength and hardness than annealed steel.

    Annealing is used to relieve stress, recrystallize cold worked steel, spheroidize carbides, and get the steel into the softest and most easily worked structure. Normalizing is used to get a more uniform structure, reduce segregation of elements in the steel, and/or strengthen the steel slightly.

    How you harden/temper the blade has more to do with what you want the knife to do. If you want a knife to bend instead of break, you can use an oven and use one of a few methods to get a soft spine and hard edge. If you want a knife that resists bending, but will break if pushed past its limits, high though they may be, you can harden the entire blade. If neither one matters and you just want a knife that cuts, like a fillet knife or kitchen knife, you can harden the whole blade. If you are using an air hardening steel, you will probably have to settle for a fully hardened blade, tang, spine and all.

    Whether you normalize or anneal when doing stock removal depends on the form of steel you get. If you get annealed steel, I'd say annealing again wouldn't be needed, though a stress relief treatment may help if you do heavy grinding or machining. If the steel comes hot rolled or cold rolled you may have to anneal it to make it easier to work. Some experimentation will be needed on your part, since what you do depends on the steel you use, what form it comes in, where you get it, and how you shape it. If you start with Cold Rolled Annealed (CRA) 1095 and just use a hacksaw, files, and sandpaper to shape, you could go straight to hardening. If you start with hot rolled 52100 and use a milling machine or high power grinder to do heavy stock removal, a preparatory annealing treatment may make life much easier.

    **All these statements are general and apply to steels normally used to make blades. Changing steel types will drastically change how the heat treatments of annealing and normalizing are done. For example, heating and quenching 316 stainless will anneal it.;) Also, keep reading Kevin's stickies. They take more than once to fully grasp.**
  8. smorgan345


    Mar 8, 2007
    Thanks guys, I appreciate it :D
  9. Kevin R. Cashen

    Kevin R. Cashen

    Sep 9, 2003
    I will have to eventually find a way to take care of that problem:D;)

    For the sake of technical accuracy it is better to drop the use of the word “molecular” from your memory when discussing metals. Annealing is perhaps the oldest of heat treatments and was born of the world of cold worked metals, non-ferrous first and then ferrous. Its main purpose was to remove the strain hardening effects of deformation; it accomplishes this by first erasing the stored strain energy in a lower temperature are of the process known as “recovery”. Next it erases the other effects at the recrystallization temperature by making all new grains. And then via slow cooling it allows all the carbon to separate out so that the steel stays in this new and soft condition. But carbon the carbon that is put into solution above the recystallization temperature will separate out into banded sheets when slow cooled. This is not a problem with lower carbon steels, but high carbon stuff will form larger sheets that will be difficult to machine, so…

    …another type of softening operation known as “spheroidizing” is used on higher carbon steel. This operation still goes through the recovery process and eliminates the strain energy, but it does not go high enough in temperature for recrystallization so the established grain size is totally unaffected. But most importantly it does not put eh carbon into solution but instead separates it out into countless tiny spheres that easily move out of the way of machining operations.

    Because of these differences some sources do not use “spheroidizing” and “annealing” interchangeably, but if we wish to treat them as the same then the more generally accurate description for annealing would be for the purpose of softening the steel.

    Normalizing may or may not result in s totally stress free or softer steel depending on the alloy and its main function is to bring the internal structure to normal and even conditions. Carbide, grain size, various segregation and even strain effects are all equalized by normalizing. It is not necessarily a lower temperature thing and is actually described as a fairly high temperature (1700F) operation in industry. Bladesmiths however have the freedom work with a variety of ranges and often use cycles at much lower temperatures and call it normalizing. But the ideal would be to start high with one cycle to equalize everything and then follow it up with lower temperature cycles to refine the homogenized structures.

    Then, as mentioned, there are pure stress relieving operations done only in the recovery range (around 1200F) which will remove the strain effects in the steel and nothing more.

    For the most part- no, but this could depend on the steel you are working with. If it is really good stuff, the chances of you detracting from its good condition get better with every unnecessary cycle. However if you got some steel that is particularly prone to nasty segregation these days (1095, 5160, cheap O1 etc…) then a few high temp cycles couldn’t hurt.

    Not necessarily, it could take up an entirely new thread but soft spines are not necessarily the best option for strength or overall blade toughness and present some issues of its own, despite the prevailing beliefs, but the facts of how strengths and toughness in steel works are at definite odds with those beliefs. If you are really attracted to differential heat treatments for performance and not aesthetics then a differential temper instead of a differential hardening is well within your means by simply reheating the spine while protecting the edge after the normal full heat treatments.

    I hope this helps clear things up.
  10. Remyrw


    Jun 17, 2010
    It does for me, Kevin. I can't remember how many times I went through the sticky, as well as your other posts, but I learn something new every time. Or maybe it just sinks in a bit more.
  11. tryppyr

    tryppyr KnifeMaker / Craftsman / Service Provider Knifemaker / Craftsman / Service Provider

    Feb 5, 2010
    Very interesting topic. Thanks for starting it and thanks for the replies.

    Based on the information I've seen here, I'll walk away with the assumption I don't need to worry overmuch about the stresses my stock removal techniques impart to the steel, especially since most of those techniques don't involve any significant form of heat.

    - Greg
  12. me2


    Oct 11, 2003
  13. smorgan345


    Mar 8, 2007
    It does, thank you. So basically difference between annealing and spheroidizing is recrystalization which is determined by temperature? And if I'm using a high carbon steel, I'm better off spheroidizing than annealing to avoid the possibility of creating banded sheets of carbon making machining more difficult?

    lets say I'm using 1084. Can you describe the basic chronological steps in normalizing it, vs the basic steps of annealing it? Thank you.
  14. mete

    mete Gold Member Gold Member

    Jun 10, 2003
    The problem with high carbon steel is the formation of grain boundary carbides when annealing.I assume by 'banded sheets of carbide' you mean pearlite !!

    Annealing would be to heat above the critical [75-100F above critical] and slow cool.Which would give spheroidized carbide but also possible grain boundary carbides with things like 1095.
    Normalizing ,75-100F above critical and air cool. This would give you pearlite .If it is then heated to 1200 F for a few hours it would produce spheroidized carbide.
  15. Stacy E. Apelt - Bladesmith

    Stacy E. Apelt - Bladesmith ilmarinen - MODERATOR Moderator Knifemaker / Craftsman / Service Provider

    Aug 20, 2004
    Kevin tends to be too scientific in his posts, I will try and put it into the terms that most understand;

    When you read Kevin's articles, replace the word carbon with Good Faeries. Few actually know that ferrite is named such because of its high faerie content. The more good faeries the better the steel...up to the point when they get into groups and start making trouble. Then you have to deal with the Good faerie vs Bad faerie take-over plots ( large carbides and grain size).

    In layman's terms:
    When Faeries are resting they all sit on the branches in nice neat rows (pearlite). This is a good situation ( annealed). It is really good when you get them to form faerie rings and sit in circles along the branch ( spheroidal anneal). The members of each circle are Red faeries ( carbon) and Blue faeries ( ferrite). They sit red-blue-red-blue ( face-centered-cubic). When the wind blows hard (metal is heated up) they all start flying around. This evenly distributes the different color faeries ( elements) and when they calm down (cool off) they sit down in a rainbow pattern ( evenly distributed carbon and alloy content). If they get scared by lightening ,they sit down fast ( quench). The bigger faeries sit down first and thus force the others to sit between them in a different pattern than usual (body-centered-cubic martensite). This may not the best arrangement, however, and squabbles for branch space may cause the branch to break due to the strife ( brittle martensite). We deal with this by slightly agitating the faeries and allowing them to nudge each other until they are sitting more comfortably ( tempering). It usually takes a couple of tries to get them to settle down and play nice ( double temper cycles).

    Now, doesn't that make it easier to understand.
    Glad I could help.
  16. Kevin R. Cashen

    Kevin R. Cashen

    Sep 9, 2003

    Are you kidding? I intitially wrote pearlite lamellae, and thought of the number of complaints of how over-technical my posts can be so I decided to simplify it to "banded carbide sheets" for easier reading... and I still get grief:D;).
  17. Kevin R. Cashen

    Kevin R. Cashen

    Sep 9, 2003
    In a perfect world with a pure iron/carbon system that temperature would be around 1335F, but…
    Almost every steel we work with has some sort of elements other than carbon which moves that temperature around, so each steel will have its own unique temperature that will fall near that range. In an average shop without equipment able to read exact temps for the given steel the magnet will suffice for ballpark figures. If the magnet quits sticking you have definitely began making new crystals, but not necessarily moving the carbon around as you may need for other heat treating operations. The new crystalline phase opens the pathways for the carbon to move it then requires time or temperature for that to happen, but in stress relieving and annealing this is not as critical.

    Simple spheroidization involves keeping things below recrystallization so the magnet should not stop sticking during the process. It may be easier just to think of this type of spheroidizing as very high temperature tempering, and is effective enough for very simple steels to make very fine spheroidal carbides that will allow grinding and some machining.

    Steel above .8% carbon will have excess carbide that will get into mischief on you if you allow it to with a slow cool from solution. As mete pointed out it will naturally want to go for the grain boundaries where it will not only burn up your mills and drills, but it will also cause embrittlement issues if not removed from those grain boundaries. But even without grain boundary cementite (I threw that big word in for mete :D, for the other folks – iron carbide), coarse pearlite (banded iron carbide) will offer a tougher machining material than spheroidized (think of shoveling a pile of shingles versus a pile of marbles).

    1084 is right on the borderline to where you would really need spheroidizing and if the most you may do is drill it, then the extra effort of spheroidizing may not be necessary. The old standard anneal among bladesmiths and blacksmiths is to heat to nonmagnetic and then stuff it into wood ash or vermiculate for a long slow cool, this is known as a “full anneal” or more precisely as a “lamellar anneal” due to the lamellar pearlite it forms.

    Spheroidizing on the other hand does not rely on the cooling but instead on the temperature at which it is held (red but still magnetic) to do all of the work.

    After forging reheat the steel as evenly as possible (evenly is the key word in normalizing, evenly heated and evenly cooled) to a nice orange heat, (1600- 1700F). This will, evenly redistribute the carbon and equalize the grain size (uneven grain size is actually worse and a little larger grain size) and undo much of the havoc you inflicted in the forging process (hand hammer faces have a hard time moving steel perfectly even). Allow to air cool evenly (put the tang in a vice don’t’ lay the blade flat on anything). Follow this up with a heat just beyond magnetic to refine the even structures you now have. This can also be followed by another heat a hair lower or even another heat to nonmagnetic and a quench to room temp to really refine things and set it up for the last normalizing heat which can actually double as spheroidizing as long as you stay below nonmagnetic, and if attempting to spheroidize you should try to hover in that magnetic red range for a little while by cycling it in and out of the forge.

    The full anneal would then be as described above with heating to above nonmagnetic (kind of undoing any grain refinement accomplished in normalizing) and burying in an insulating medium. But if you follow the above normalizing steps you should be able to move on to grinding without the full anneal with 1084.
  18. Erin Burke

    Erin Burke KnifeMaker...ish Knifemaker / Craftsman / Service Provider

    May 19, 2003
    :thumbup:Thanks Kevin... I dig this analogy. ;)
  19. smorgan345


    Mar 8, 2007
    You guys are great, thanks for your help.

Share This Page