flaming anvil trademark logo copyright (c) 1998 Patrick J. Dempsey
     HOME!  |  STORE  |  Getting Started in Blacksmithing  
   Guru's Den   
   Slack-Tub Pub II   
   Tailgate Sales   
   iForge How-To    
   Health and Safety   
   Book Reviews    
   eBooks On-line   
   Anvil Gallery   
  Calendar of Events  
   Story Page   
   AnvilCAM - II   
  Touchmark Reg.  
   Power Hammers   
   What's New   
Comic of the Week
   Daily Comics   
Daily Metalworking Comics!
   Webring Nexus   
   Our Sponsors General Site
Welcome visitor from
United States Flag
United States
Country Counter

Tell them you found it on!

Heat Treating :

Hardening and Tempering

How can you tell if a metal is hard? Can you scratch it with carbide? Can you bang it with a hammer or some other metal you know is hard?

It depends on the type of metal. "Hard" aluminium is about as hard as soft steel. There are numerous tests. Scratch tests as you mentioned are not very specific, however, you can purchase sets of graduated files that are different hardnesses. These are good for testing common tool steels but not refractory metals such as carbide or softer non-ferrous metals.

Most hardness testers either make a dent in the sample then measure the size of the dent OR bounce an object of the surface and measure the bounce. There are other tests for strength that are often more important.

Can it scratch other materials such as glass, stainless steel, carbide?

Yes, but again these are not very specific tests. Glass is actually a soft brittle material compared to most hardened steels. Carbide is second only to diamond in hardness. Most common stainlesses except those use to make knives is a soft gummy material.

What makes this metal hard? iron, steel, carbon, heat, tempering?

What makes materials hard is a complicated subject that has to do with internal crystal structures. These in turn are effected by heat treating or mechanical working. Most non-ferrous metals can only be hardened by "work hardening", hammering, rolling, bending. But they can be softened by "annealing" which is heating to a red heat and then cooling quickly (opposite to steel, see below).

The hardness of steel is determined by carbon content. No carbon and it can't be hardened other than by work hardening. Increasing the carbon content from 0.01% to .10% increases the hardenability and the strength. This is then modified by the addition of alloying metals as well as the alloy metals having their own properties.

To harden most steel it is heated to a medium red or slightly above the point where it becomes non-magnetic. It is then quenched in water, oil or air depending on the type of steel. The steel is now at its maximum hardness but is very brittle. To reduce the brittleness the metal is tempered by heating it to some where between 350°F and 1350°F. This reduced the hardness a little and the brittelness a lot. Most steels need to be tempered at about 450°F for maximum usable hardness but every steel is slightly different.

To soften steel so that it can be cold worked and machined is called annealing. To anneal steel is is heated to slightly above the hardening temperature and then cooled as slow as possible. Cooling is done in an insulating medium such as dry powdered lime or in vermiculite. High carbon and many alloy steels can only be cooled slow enough in a temperatue controlled furnace since the cooling rate must be only 20 degrees F per hour for several hours.

The set of processes, annealing, hardening and tempering are collectively known as "heat treating". For details see below.

I need to make a tooling Die for my Hammer. Would you suggest 4140 for this? Then if possible could you explain the difference between tempering and heat treating, and which of these would be used for the dies of a power hammer?

Hammer Dies: Several manufacturers use SAE 4140. The industrial guys use a variety of steels including SAE 4150, Bull hammers uses H13, Big BLU uses S7. Plain carbon steels such as SAE 1075 or SAE 1095 have also been used but require more careful tempering. Modern steels often recommended are the H series, O1, A2 and D2.

heattreating: Tempering is one stage of heattreating. The sequence for most steels is:

  • Normalize (or anneal depending on the steel)
  • Harden (heat to the A3 point and quench)
  • Temper (heat to lower brittleness and reduce stresses)
Normalizing is like annealing except it does not require as long a cooling period. It is no longer recommended for many alloy steels. Hardening methods vary mostly in the quenchant used depending on the type of steel and the section of the part (how heavy). Heating to non-magnetic works but is not always the recommended hardening temperature. Tempering most steels requires temperatures higher than a kitchen oven's MAX temperature rating. 350°F-450°F is the lowest tempering temperature for many steels. This means the part will have the maximum hardness. This is not usually desired in hammer dies. Double tempering (recommended for dies) is simply going through the tempering process a second time. Tempering should be done as soon after hardening as is convenient. Do not quench after tempering.

Harden 4140 at 1550-1600°F Oil quench
Harden 4150 at 1500-1600°F Oil quench
Harden 4340 at 1475-1525°F Oil quench

Temper to 440 to 480 Bhn, 45-50 Rc. For the above steels requires 500-600°F
Temper to 341 to 375 Bhn, 37-40 Rc. For the above steels requires 800-900°F

See Heat Treating 4140 Hammer Dies Includes temper table.

Annealing is the softening of metal by heat treatment.

Ferrous metals are annealed by heating to just above the A3 point (a point above non-magnetic that varies with the carbon content), and then cooling slowly. For common carbon steels the cooling can be done in dry ashes, lime powder or vermiculite. For high carbon and alloy steels annealing requires cooling in a furnace that has temperature controls so that the rate of cooling is no more than ~20°F/hr.

Non-ferrous metals such as aluminium, brass, copper and silver are annealed by heating to a low red and quenching in water (the opposite of steel).

- guru - Thursday, 08/02/01 19:55:04 GMT

Annealing Air Hardening Steel: The cooling rate is very critical. The book says 40°F/h from 1600-1650. That's 20-24 hours at a steady rate to still be 200°F. The non-magnetic point being 1425°F appears to be too low to anneal. . .

Since the critical time is the first 8-10 hours it probably needs to be brought down in a furnace or salt pot.

Lets put it this way, If spit doesn't sizzle a day and a half later it probably cooled too fast. I've had the best luck with quick lime but never tried to anneal air hardening.

Grandpa may have some trick for this.

- guru - Thursday, 09/28/00 03:35:12 GMT

Guru speaks the truth. To get D2 soft, first soak at the critical temperature for at least 30 minutes, then cool very slowly down to 1300°F. The temperature slide from critical to 1300°F needs to take 10 hours, in order to convert all of the austenite to pearlite.

grandpa (Daryl Meier) - Thursday, 09/28/00 04:47:51 GMT

Annealing 0-1: To anneal heat to 1400-1450°F (760-790°C), cool no faster than 40°F (22°C) per hour. This rate needs to be maintained for 4 to 5 hours. The temperature varies but the cooling rate is the same for alloy tool steels. On a small part it is easy to lose heat at too fast a rate and end up with a hard part without quenching. A thin section of water hardening steel will air harden if you are not careful.

To test the above cooling rate, heat your part to above non-magnetic and put into your annealing medium (lime or vermiculite). Come back four hours later and remove the part and observe it in low light. The part should still be a low red but hotter than purple/red. If it has cooled to a purple/red or black heat then it has cooled too fast.

To anneal a small piece of tool steel you may need to bury it with a larger piece of steel heated much hotter (an orange). Bury the two pieces next to each other but not quite touching. Test as above. Remember, the 40°F (22°C) per hour is a maximum rate, the slower the anneal the softer the steel (to a point).

- guru - Saturday, 10/28/00 01:22:40 GMT

Annealing Non-Ferrous Material: Almost all non-ferrous metals are annealed by heating to somewhere just below the melting point and then cooling in air or by quenching in water. Quenching in water is a convienience.

Alpha brasses (64-99% copper) are annealed by heating to 700 to 1400°F (the hotter the softer) and can then be be quenched.

Alpha-beta brasses (55 to 64% copper) are annealed at the same temperature and can hardened slightly by quenching from the annealing temperature.

The key word above is slightly. Cold working produces a much greater degree of hardness. The amount of hardening is so low my copper alloys book does not give specific data. If quenched from the low end of the annealing temperature there would be no disceernable difference.

Common brazing alloy is:

Cu 56 - 60%
Sn 0.8 - 1.0
Fe .25 - 1.20
Al, Si, Mg, Pb trace (no greater than 0.1% each)
Zn balance

That makes it an alpha-beta alloy. - guru - Monday, 12/11/00 15:12:49 GMT

Hardening: The ability to be hardened varies with the carbon and alloy content of steel. The higher the carbon content the harder the steel can become. Low carbon steel has very low hardenability and wrought iron which has no carbon is unhardenable.

To harden steel it is heated above the "transformation point", a low red or just above where the steel becomes non-magnetic. Then it is quenched in brine, water, oil or even air. Afterwards it is tempered by reheating. This reduces the brittleness of the steel a lot and the hardness just a little. Temper temperatures range from as low as 350°F to as high as 1400°F depending on the steel.

The quenchant depends on the type of steel. In general quenching in a more sever quenchant than necessary can cause cracks in the steel. Overheating prior to the quench can do the same.

In general hard parts are always more brittle than soft parts. Using parts that are too hard can be dangerous. On machines this can mean parts that may explode or shatter.

I left a bunch of variables open above. This is the nature of the game. The starting place is to know what kind of steel you are working with. Then go to a reference like MACHINERY'S HANDBOOK and look up the correct heat treating parameters. IF you don't know what kind of steel you are using then you have to become your own metallurgist and do some detective work. This requires lots of trial and error and attention to detail, plus a lot of knowledge.

There is no simple formula or magic bullet. Start with a book like Jack Andrew's NEW Edge of the Anvil and a copy of MACHINERY'S HANDBOOK. If you start working with a variety of steels you will also need the ASM Metal Reference Book as it has more complete listings of numerous alloys.

- guru - Friday, 06/16/00 20:21:32 GMT

Hardening /annealing silver

I don't know much about sterling silver, but I looked it up in ASM Metals Handbook vol 1 8th ed. Sterling silver is age hardening, but the solution temperature(1300-1350°F) is close to the liquidus temperature(1435°F). The precipitation of the copper rich phase is done by aging at 535°F for 2 hrs or 575°F for 1 hr. Normal annealing as done by jewelers --- heat to very dull red (about 1200°F) in a darkened area then quench in pickeling solution.

grandpa (Daryl Meier) - Wednesday, 10/25/00 04:12:46 GMT

While working silver I bring the piece to a dull orange (1100°F) and quench in water making the silver malleable until my pounding/shaping work hardens the material. You can hear the difference in sound as the piece becomes work hardened and needs to be heated again. To harden an item after all work is done I place the piece in a kiln and bring it up to Temp app 650°F and let sit for 6-7 hrs and cool down. The item now is hardened and would need to be brought back up to the 1100°F and quick quenched to be worked on again.

Silversmith Saturday, 10/28/00 00:11:51 GMT

The quenchant may be brine, water, oil or air depending on the type of steel. Low carbon steel will harden slightly but not to the degree of spring or tool steels. The parameters of the heattreating sequence is determined by the type of steel. Once hardened, the part must be tempered. Tempering is the reheating of the part to a temperature well below the hardening temperature to reduce the hardness and increase the toughness. This may range from 350°F to as high as 1350°F depending on the steel and the hardness desired. On very hard critical parts double tempering (doing more than once) is recommended. Tempering helps reduce hardening stresses and double tempering is cheap insurance.
Quenching Oil: Mineral oil is the least toxic. For something the size of a power hammer die you will want several gallons. Oil is lower density than water, has lower thermal conductivity and flashes rather than evaporate if over heated. Therefore it takes quite a bit more oil to quench a part than water.

If you quench with too little water it just boils off. If you have too little oil it goes up in explosive smoke that is often ignited by the hot steel. If you must use automotive oils use ATF. It has less (possibly toxic) additives than regular oils. - guru - Monday, 06/19/00 04:48:38 GMT

The splendid smith Burnham-Kidwell pointed out that when he changed from automotive drain oil (the old standard low-rent quenchant) to used deep-fry oil his shop went from smelling like a lousy auto repair shop to a cheap deli...a considerable improvement. Deep fry oil ( often peanut oil) is selected for it's high flash point, is pretty non-toxic as oil quenchants go, and is generally free. It seems to work just fine. . . . . er, avoid the fried fish places.

Pete Fels - Monday, 06/19/00 07:26:37 GMT

Tempering unknown steel (blades): When you use scrap steel you have to become your own metallurgist. Take a scrap piece forge a section similar to your blade and experiment. Use a file or HSS cutter bit to judge the hardness. Try sharpening a short piece and see how it works. You have a choice of trial and error or purchasing a known steel.

The transformation point of steel is just a tad higher than the point at which it becomes non-magnetic BUT is equal or lower on high carbon steels. But by the time you've tested (in the forge) the part will have reached the transformation point. Many alloy steels are oil quench and I start there. If it doesn't harden sufficiently then try water (it should be warm or slightly above room temperature). You cannot judge temper temperatures of alloy steels by temper colors.

The best way to get a uniform temper is to heat a larger block or slab of steel to a known temperature and then set your blade on that and let it soak up the heat. It should remain at tempering temperature for as long as you can maintain it or up to an hour. If your tempering block is fairly large just let it and the blade both cool together. Tempering temperature varies with the variety of steel. It can be as low as 350°F and as high as 1300°F. Most steels are tempered in the 500 to 600°F range. You really need to find a copy of MACHINERY'S HANDBOOK or one of the blacksmithing references such as Edge of the Anvil that has tempering data. If you are going to stay in the knife business you should purchase one of the (relatively expensive) references such as the ASM Metals Reference Book. There are just too many steels and too many combinations of treatments to cover here.

Heattreating - Exact SAE1095: Unless you know exactly what kind of steel you are dealing with there is no "exact" and trial and error is the rule. Don't assume every piece is the same. EVERY manufacturer make their own steel selections and may change them at any time. Where one manufacturer uses plain carbon steel another may use an alloy steel.

THEN there is the matter of temperature control. Unless you have calibrated temperature measurement equipment and controlled furnace/salt pots then determining the "correct" temperature will require more trial and error.

Blacksmith style heat treating is about as close to alchemy or magic as you can get. Judging heats by colors described in florid terms like "sunrise red" that can vary 200 degrees depending on ambient light and working with steels of unknown pedigree. . . .

Assuming a plain high carbon steel like 1095 you would heat until non-magnetic and then 50°F more to 1480°F. Then quench in warm water.

Temper immediately (as soon as possible) at a minimum of 450°F for up to 2 hours to obtain Rockwell 57-58. It doesn't hurt to double temper. I'd go a little hotter (say 500°F) for a more durable blade. If its a single edged blade then you can come back and draw the temper of the back some more. This is best done with a block of steel heated to the desired temperature and watching the colors "run" on a clean ground surface of the blade. - guru - Sunday, 07/09/00 02:24:59 GMT

SAE 1095 Carbon Tool Steel:

Normalizing: Heat to 1575°F (855°C) cool in air. (note no holding time).

Annealing: "As is generaly true for all high carbon steels, the bar stock is supplied by mills in spheroidized condition. . . . . When parts are machined from bars in this condition no normalizing or annealing is required."

Forgings should be normalized.

Anneal by heating to 1475°F (800°C). Soak thoroughly. Furnace cool to 1200°F (650°C) at a rate not exceeding 50°F (28°C) per hour. From 1200°F (650°C) to ambient temperature, cooling rate is not critical.

Hardening: Heat to 1475°F (800°C), Quench in water or brine. OIL QUENCH sections under 3/16" (1.59mm).

Tempering: As quenched hardness as high as 66 HRC. Can be adjusted downward by tempering.

Heat Treaters Guide - Standard Practices and Procedures for Steel. 1982, American Society for Metals, p.81

- guru  - Friday, 08/17/01 03:06:02 GMT

Definitive Heat Treat (SAE 5160 51200): No such thing. I can give you the temperatures the books call for and you will need a calibrated pyrometer and temperature controlled furnace and tempering baths. But you will still have to TEST. Why? Because every shape needs slightly different handling. To get expected results you have to take your specific shape and cause it to go through the temperature transformations in the time frame required. Time is a crucial element in heat treating and except for annealing and some generalizations for tempering very little is published. You just have to find the right way to make the process "fit the curve". Then afterwards test the piece and adjust your methods.

It is not nearly as bad as trial and error testing of an unknown steel because you start knowing the general process but if you want to be picky and want an EXACT hardness or material condition then you are going to have to test.


Anneal at 1525°F then cool rapidly to 1300°F and cool to 1200°F at no more than 20°F/h for 5 hours.

To harden heat to 1525°F and quench in oil. Temper as needed (minimum of 350°F).

Austempering at 1550°F and quench in a salt bath at 600°F and hold for 1 hr. Cool in air, no further tempering is needed.

According to the Bethlehem book "Modern Steels - Handbook 3310" the following are APPROXIMATE Rockwell C hardnesses of oil quenched 5160 for various tempering temperatures:

SAE 5160
Temperature Hardness Temperature Hardness
400°F 59 Rc 900°F 42 Rc
500°F 57 Rc 1000°F 37 Rc
600°F 54 Rc 1100°F 32 Rc
700°F 52 Rc 1200°F 28 Rc
800°F 49 Rc 1300°F 20 Rc

Use a temper color chart to get close to the hardness you require.

- Quenchcrack - Thursday, 03/27/03 13:21:32 GMT

Normalize by heating to 1625°F and cooling in air.

Perlitic structure not desired in this steel. To anneal for a predominately speroidized structure heat to 1460°F and cool rapidly to 1380°F then continue cooling at a rate not exceeding 10°F/h. to 1250°F.

To harden heat at 1550°F in a neutral salt bath and quench in oil. Temper immediately after cooling to 100-120°F at a minimum of 250°F. Normal practice is to temper at 350°F.

ASM Metals Reference Book and ASM Heat Treaters Guide, American Society for Metals International I recommend both the above books for ALL knife makers that do their own heat treating. Both books include graphs and charts with more detail than can be produced here. See our link to ASM on the links page.

- guru - Friday, 07/06/01 00:03:58 GMT


According to the ASM Metals Reference Book
Do not normalize, Anneal at 845-900C / 1550-1650F

Harden at 995-1040C / 1825-1900F (hold for 15-40 min.) then Air quench. Immediately temper at 540-650C / 1000-1200F.

On air hardening dies I use stainless foil to protect the die while heating. If using the non-magnetic test for temperature then use a small sample (not too small) of the same alloy in the forge. Remove from the forge/furnace, pull off the foil and let cool on a grate (such as a piece of bar grating) where air can circulate all around the part. Normaly I turn off my gas forge when I remove the heated dies to harden. After hardening I put them back in and use the residual heat from the fire bricks to temper. Not very scientific but it works. Use a salt bath if you want perfect control and low oxidation.

-- guru Wednesday, 04/07/99 00:41:32 GMT

H-13 (simplified):
Heat until it becomes non-magnetic then pull it out of the fire and let it cool on a brick until you can handle it (that's the air quench hardening the piece) THEN reheat it to 1100 degrees F to temper. Clean tempered H-13 has a nice plum color.

-- guru Wednesday, 04/07/99 21:05:56 GMT

H-13: H-13 makes very good Power Hammer dies. Currently that is what they use on the BULL. Those dies are machined, heattreated and then welded (with a LOT of preheat) to a mild steel base.

Latrobe Steel sells a heattreated version of H-13 under the trade name Viscount-44. The 44 is the Rockwell hardness. This steel is sold as die steel that is machinable (just barely) with ordinary machine tools. As heattreated it is a nice plum color. Our family machine shop used quite a bit of this material to avoid heat treating parts. H-13 is an air hardening steel. I would draw it back to just short of annealed for small hammer dies.

-- guru Monday, 11/29/99 15:03:07 GMT

H 13: A chrome-moly high vanadium steel. Hot Work. All specs in Fahrenheit. Forge 1950-2100, not below 1650. Anneal 1550-1650, cool per hour 40F max. Harden with a slow rising heat to 1825-1900; quench in air. Temper 1000-1200.

Frank Turley - Monday, 11/05/01 20:47:22 GMT

D2 vs D7:

To the Gurus among us. I am a toolmaker for Cessna aircraft. We have an ongoing argument which it seems we are having trouble finding answers for. The argument is on heat treating specs for D7 tool steel. No one can seem to come up with any specifications for this tool steel. The "old timers" say that the same specs that are used for D2 will be the same for D7. The "young guns" say that they are insuficient for proper treatment of this type of steel. What do you say? also can you tell me where to go look for heat treating specifications for this steel type? Thanx

Mike - Tuesday, 08/30/05 22:42:36 EDT

D7 Tool Steel:

Heat treat instructions for this grade can be found in the ASM Heat Treaters Guide. This reference does state that higher temperatures and longer times are needed to dissolve all the carbides in D7 than in other D series tools steels.

Patrick Nowak - Wednesday, 08/31/05 07:39:21 EDT

D2 vs D7:

There ARE some differences in how the steels are hardened and tempered according to the ASM Heat Treaters Guide, Standard Practices and Procedures for Steel. This is a reference that every high tech shop should have if they do or specify heat treating.

Property D2 - UNS T30402 D7 - UNS T30407
1.50 C
1.00 V
2.35 C
4.00 V
Cycle Annealing
Stress Relieving
---- Same ----
Hardening Heat Slowly. Preheat at 1500°F and austentize at 1800 to 1875°F. Hold at temperature for 15 minutes for small tools to 45 minutes for large tools. Quench in air and cool evenly on all sides. A block 3 by 6 in. will harden throughout to 62 to 64 HRC. When salt quenching, quench in salt bath at 1000°F, hold only long enough to equalize temperatures, cool in air. Heat slowly. Preheat at 1500°F and austentize at 1850 to 1950°F. Hold at temperature for 30 min. for small tools and 1 hr. for large tools. Quench in air, cooling as evenly as possible on all sides. Approximate quenched hardness 63 to 66 HRC.
Stabilizing Similar except more strongly recomended for D7 and prior to applicable cryogenic treatment.
Tempering Temper immediately at 400 to 1000°F after tool has cooled to 120 to 150°F. Double temper, allowing tool to cool to room temperature before second temper. range of hardness after tempering is 61 to 54 HRC. Temper immediately at 300 to 1000°F after tool has cooled to about 120 to 150°F . Double temper, allowing tool to cool to room temperature before second temper. Range of hardness after tempering 58 to 65 HRC.
NOTE: The above is NOT the complete composition or recomendations, it is the most obvious differences. The VERY high carbon and increased Vanadium make D7 a considerably different steel and pickier to handle.

- guru - Wednesday, 08/31/05 08:27:33 EDT
How should a file be bent? Attempting to convert some files into almost rifflers. Wanting to curve a file to clean the inside of a spoon.

Rifflers: I've made these several times. The handle end of half round files rarely gets much wear and makes great spoon files.

I heat locally to a low red with a cutting torch while the extra file is clamped in a vise, bend with tongs or pliers and then torch off the extra and quench. The torched end is ground to clean up.

Bending the half round file produces a semi-spherical surface. Since my use was on wood I didn't perform a separate heattreat. I figured it was better not to have to heat the file and chance burning the teeth more than once. That's why it was heated and torched and quenched in one quick heat.

I've used the same technique to bend triangular files also. If you want to heattreat then it would probably be best to heat in stainless foil.

- guru - Thursday, 06/08/00 20:13:27 GMT

SS Laminates:

The trickiest part of SS laminates is determining the heattreating. You have to have combinations that can be hardened and tempered with processes that work with both or where one does not effect the other. Its a real puzzle that takes research and serious thought. THEN you have to be able to actually do the heattreating within the temperature limits determined. This requires careful temperature measurement and control.

- guru - Wednesday, 10/04/00 14:25:34 GMT

Salt Baths:

Salt Baths: Don Fogg had some nice articles on his web site about building and using a small salt bath. Common table salt works and is fairly good all round medium. Special heat treating salts are sold. The only time the type of salt is critical is if the bath is to be used for hardening (yes they get that hot) OR if there is a "use no chlorides" requirement for the application.

From Grant Sarver "guru page" post in September 1998: All sorts of salts are used in "salt" pots (as they are called in the heatreat biz) For temperatures up to 1000F sodium nitrate can be used. Barium cloride is used for high temps (like 2500F). For temps to 3000F magesium fluoride can be used. Most heatreat salt pots are heated simply by passing an electric current thru, controled by thermostat. Heatreat supplies have an assortment of salts for this purpose.

Salt baths can be used to harden, temper or anneal.

Heat Treating Salts
Salt Formula Melts Max Action
Sodium Chloride
Common salt
NaCl 1473.8°F
Potassium Chloride KCl 1418°F
Potasium Nitrate
KNO3 633°F
Barium cloride
BaCl ? 1765°F
Magesium Fluoride MgFl ? °F

The melting point for common salt is high enough for annealing and hardening carbon steels.

Potasium Nitrate is easier to melt but has a narrow working range. Organics mixed with nitrates can produce dangerous situations. Small amounts of sulfur can result in explosive mixtures but saltpeter is still commonly used for various metal working processes.

Heat treating suppliers sell various salt mixtures. Some are considered "neutral" some carburizing.

- guru

Forge Furnace Size & Salt Baths:

I would much appreciate your advice on the following. I'm am just about to create my first forge, and I beleive I will eventually be using it to forge relatively large pieces such as swords. I was wondering what size I should make it and how much that matters. I will be making a propane powered forge. I know I can work on and normalize a sword with a small forge, but the problem is heat hardening. If I just move the sword back and forth in the forge (assuming it has openings at either end) will it be heated evenly enough for quenching? Beleive me, i've tried searching for the answer, but haven't found it anywhere. I appreciate your help,

You have found the crux of the problem with gas forges. You need diferent sizes for different work. In sword making you cannot work a long piece becasue when it is hot it will droop and act like a soft noodle. So forging is done in short heats.

Heat treating long pieces is a real trick. When swords are done in a short fire they are moved back and forth as you have summized. By heating JUST enough the blade is not so soft that it can be slid back and forth supported by the coals in the fire. THEN when it is pulled from the fire it must be done so in a quick smooth motion that does not alow it to sag as it is quenched. A REAL art and a true ballet. The Japanese sword smith avoided all this and only hardened a narrow strip of the edge. Needing to straighten the blade after heat treating is not unusual.

Modern smiths using gas and oil forges use different methods. Long racks with supports every few inches are used for horizontal handling. The problem is the racks heating. So the hot blade is rolled into a cold rack.

The method used by many bladesmiths is a vertical furnace or vertical salt pot. In this method the blade is suspended in the furnace from a hole in the tang. Furnaces must be designed so that the heat enters the bottom and exits the top without buidling up in one end or the other so there are no hot spots. Salt pots are often used because the liquid salt circulates in the crucible and produces an even heat. Salt baths are used for both hardening and tempering.

The salt also protects the steel from oxidation. Tall salt pots are commonly made from stainless steel pipe and heated in a special built gas furnace. Temperature controls (a significant cost) are also applied. Due to the reactivity of the salt I would recommend a integral thermowell in the pot. However, many just replace thermocouples as needed. Common salt will work, special salts are sold, some are highly toxic.

Gas forges are VERY efficient when sized for the work but very inefficient when used for work much smaller than their capacity. SO, you need more than one forge/furnace and probably specialty furnaces for heat treating.

-guru June 6, 2004


I would like to know if you have any information on cryogenic tempering with dry ice?

Freezers (-40 to -50°F) and dry ice (-109.3°F) are not cold enough

Dry ice has a surface temperature of -109.3°F (-78.5°C). It IS possible to have dry ice colder than this but transfering that cold to another object is difficult. A dry ice acetone bath typicaly provides -78°F (-61°C) which is far short of the cryogenic temperature needed for treating steel.
Not all steels are improved by cryogenic treatment. It is also part of a complete heat treatment not a replacement or simple secondary treatment.

Dry ice is solid at -108.76°F (-78.2°C). Most cryogenic treatments need colder temperatures and use liquid nitrogen at -328°F (-200°C). So don't believe those pushing products treated in dry ice. They may be blowing CO2 up your shorts. . .

From the ASM page:
Cryogenics: The Racer's Edge: Cryogenic treatment of metal parts is performed at temperatures below 185°C (300°F). If done correctly, it causes permanent changes in the material that can enhance wear resistance. This article concentrates on applications in race cars and other performance vehicles. Roger Schiradelly and Frederick J. Diekman

Cryogenic Treatment of Tool Steels; Two mechanisms are involved during cryogenic treatment of AISI D2: transformation of retained austenite and low-temperature conditioning of martensite. The former leads to an increase in hardness (and reduction in toughness), while the latter boosts wear resistance (and enhances toughness). You can choose the results you want by proper selection of the austenitizing treatment.
ASM Also sells a book titled Cryogenics, for $36.95. I would start there.

- guru - Saturday, 01/19/02 00:03:36 GMT

Freezing and Cryogenic treating of steel.

I have seen a few questions on this recently and thought I would add a little to the discussion. Cryogenic treating or "freezing" of steel with liquid nitrogen has become a very common practice in industry for such things as cutting tools, gun barrels, knives etc. If you do a search on the net for it, you will find people claiming that it can do miraculous things, like prolong the life of panty hose. (I actually found this while doing research for a school project). The main benefit of using a deep freeze cycle is that any retained austenite in the steel will transform to martenstite.

Quick Metallurgy background:
When heat treating steel, the steel is raised the its "austenitizing temperature". Blacksmiths often judge this by using a magnet. The hot steel is then quenched, transforming the austeninte to martensite. However, in many cases, and in particular with high alloy tool steels, some of the austenite does not transform to martensite-hence the name retained austeninte.
Anyway, by cooling the steel well below room temperature, the retained austenite can be made to transform into untempered martensite, which is very brittle. This is why cryogenic treatments must be followed by additional tempering. There is still a great deal of conflict as to the benefits of cryogenic treatments, because tools that are properly heat treated to begin with see very little increase in tool life. However, tools that have not been heat treated correctly will often show dramatic improvements in tool life. As to the benefit of using this process on knives, it would probably be material dependant. If you are using highly alloyed tool steels like the A, D, M, and stainless steels, it would probably be advisable. If you are using simple carbon steels, and are already getting a good quench, then you may not see much improvement. The same goes for Damascus - it is material dependant.

Patrick Nowak (engineer) 01/30/02

NOTE: The problem isn't only that each alloy must be treated differently but that the heat treatment before the cryogenic treatment is often different than normal and the post treatment also varries. There are books with recommendations on the dsubject but even then you will find that not every alloy has been tested. If you want to treat any material that the research has not already been done THEN you need a complete metalurgical laboratory or the funds to pay one for the testing. Since this is trial and error the testing can get very expensive. Can anyone say, "Government grant required".
- guru - Wednesday, 01/30/02 17:01

416 Stainless Steel

There is a 416 and 416Se. Type Se has selinium added to further improve machinability. It is considered another type of 410 SS. Capable of hardening to 42HRC or slightly higher. Can be martempered.

All heat treating of tese steels require a protective atmosphere (vacuum, inert gas or nitrogen). Heat slowly to 1700 to 1850°F, soak for up to 30 minutes, oil quench. temper at 400 to 1400°F. Cryogenic treatment improves this steel. Temper (again) immediately after.

- guru - Wednesday, 11/30/05

References and Links

Copyright © 2001 - 2005 Jock Dempsey,

Counter GSC

Page Counter