Regular Article
Effect of Ni, Mn, V, and Al on Toughness of Blade Steels
2015 Volume 55 Issue 10 Pages 2225-2228
Details
2015 Volume 55 Issue 10 Pages 2225-2228
In modern high carbon steels, Mn alloying binds S, and usually Al or V precipitates prevent grain growth. Primitive steels do not contain these alloying elements, but they may contain Ni, obtained from iron meteorites. Totally unalloyed carbon steel (0.7C–0.02Mn–0.01S–0.01P) and Ni steel (0.7C–4Ni–0.03Mn–0.01S–0.01P) were made in a laboratory foundry and their properties were compared with modern high carbon steels in a bladesmith forge. The unalloyed carbon steel suffered grain growth. Therefore, after low temperature tempering it had exceptional poor toughness, but after higher 250°C tempering it was as tough as equally hard (57 HRC) modern fine grained carbon steel. The Ni alloying increased toughness in the case of low temperature tempering, but it also reduced attainable hardness. The studied Ni steel was equal to modern carbon steels up to 62 HRC. The study showed that Mn, Al, or V alloying is not mandatory for the best quality blade steels.
Bladesmithing has thousands of years old traditions and old techniques are still used. Therefore, it is natural that some blade enthusiasts are fascinated about the idea, that in some instances primitive steels can be equal to modern steels. In this respect, chemical composition have the most vital role. Primitive steels do not contain added alloying elements, but they may contain Ni obtained from iron meteorites.1,2,3)
Hardened high carbon steel can be tough if brittle grain boundary fracture does not occur. The intergranular fracture mode increases with increasing grain size and P contents.4,5) Also free S can weaken the former austenite grain boundaries.6) Primitive steels may have as low P and S levels as modern steels.2,3) But the absence of Mn alloying and grain size control elements (Al, V) may decrease toughness. Absence of Mn alloying can be more difficult problem in Fe–Ni alloys, like meteorites, than in plain carbon steels and irons. In Fe–Ni alloys a very small amount of free S results in hot shortness and brittle grain boundary fracture.7)
Meteorites have been forged into blades since prehistoric times,1,2,3) and they are still used by modern bladesmiths. Generally iron meteorites are malleable, but blades forged from them do not have reputation of superior properties. Some present day smiths use meteorites in crucible steel. Ancient or old use of meteorites in crucible steel is a speculation, which mainly base on the facts that crucible process is very old, some modern smiths melt meteorites, and earlier smiths were as eager and curious as modern smiths. There is no archaeological or historical evidence of such steel. Crucible charge which consists of mixture of ore based steel and meteorite iron may produce primitive Ni steel. It is not known, what is the effect of Ni in such steel, and can Ni steel at all attain good toughness without Mn alloying.
In knife business stories have an important role, and there is a famous story about the use of meteorite in crucible steel. The most legendary 19th century western knife, Bowie, has been assumed to be forged by James Black. Hundreds of different stories have been told about Jim Bowie and James Black. Paul Wellman’s novel “Iron Mistress” and the related movie with the same name tell that James Black added a piece of meteorite in a crucible charge.
The current work studies how good mechanical properties a bladesmith can achieve without modern alloying elements. Also the idea that addition of Ni containing meteorite can improve the properties of primitive crucible steel is tested. For these purposes, unalloyed 0.7% carbon steel and otherwise similar but 4% Ni alloyed steel were made in a laboratory foundry and their properties were studied in a forge with present day bladesmith techniques. The results are compared with the results of modern fine grained carbon steels.
The chemical compositions and Shepherd fracture grain sizes of the steels used in this study are given in Table 1. Steels 69C, 67Ni, and 65NiMn were made by vacuum induction melting. A very small amount of deoxidant Al was needed, and the Al level remained below the value which can prevent grain growth. 69C resembles primitive carbon steel, and 67Ni simulates primitive crucible steel which is a mixture of ore based steel and meteorite iron. 65NiMn approaches to modern alloy steels because it contains Mn. 82MnAl was taken from commercial AISI 1080 steel plate, and 80MnV was made in an air induction furnace by smiths of Knife Association of Finland.
| C | Si | Mn | P | S | Ni | V | Al | N | Grain size | |
|---|---|---|---|---|---|---|---|---|---|---|
| 69C | 0.69 | 0.04 | 0.02 | 0.012 | 0.010 | 0.0 | 0.03 | 0.010 | 0.004 | 8 |
| 67Ni | 0.67 | 0.03 | 0.03 | 0.009 | 0.012 | 4.4 | 0.03 | 0.005 | 0.004 | 10 |
| 65NiMn | 0.65 | 0.04 | 0.23 | 0.011 | 0.010 | 4.1 | 0.03 | 0.004 | 0.006 | 10 |
| 80MnV | 0.80 | 0.33 | 0.71 | 0.011 | 0.004 | 0.0 | 0.29 | 0.046 | 0.005 | 10 |
| 82MnAl | 0.82 | 0.25 | 0.75 | 0.008 | 0.006 | 0.1 | 0.00 | 0.025 | 0.008 | 10 |
Pieces about 15×15×30 mm were cut from the cast ingots, and material for 82MnAl specimens was taken from a 5 mm thick plate. The pieces were forged by hand hammer and anvil. Several forging cycles were used for a specimen: the heating temperature with the first cycles was 1250–1300°C, and with the last cycles it was about 1000°C. Surface defects were ground out after forging.
Common bladesmith techniques were used for heat treatment. Specimens were heated by a gas torch, and austenitization was determined by a magnet (demagnetization). Specimens were normalized, hardened, and again hardened. The double hardening was used because, according to the author’s experience, it produces about one ASTM number smaller initial austenite grain size than normalizing and hardening. Water quenching was used, except 69C was quenched in dilute brine, because it did not attain full hardness with water quenching. Tempering for 1 h was made in an electrical furnace, and temperature was measured by a thermocouple. After heat treatment the decarburized surface layer was removed. The final size of specimens was 3.8×6×30 mm.
A smith may test toughness by bending a blade. In this study, toughness was measured by bending a specimen until fracture occurred, refitting two pieces together, and measuring the plastic deflection (angle in fractured specimen), Fig. 1. The angle can be measured by this procedure with the accuracy of ±10%. In metallurgy toughness is the amount of energy absorbed before fracture. At constant hardness level the amount of plastic deflection has a direct relationship to the absorbed energy. Thus, the used bending test also measures toughness.
Example of toughness and plastic deflection angle measurement. Specimen is bent until fracture, two pieces are refitted together, and the amount of plastic deflection indicates toughness.
Grain size was measured from the fracture surface of as-quenched specimen by Arpi–Shepherd comparison method.8) Rockwell C hardness was inspected after bending test.
Hardness versus tempering temperature for tested steels was as expected, Fig. 2. Unalloyed steel 69C softened easier than steels 80MnV and 82MnAl containing more alloying elements. Ni alloying in 67Ni and 65NiMn decreased the attainable hardness.
Hardness vs. tempering temperature for studied steels. (Online version in color.)
V or Al alloyed steels, 80MnV and 82MnAl, had fine and smooth as-quenched fracture surface, ASTM 10, finest possible in Shepherd scale. Unalloyed steel 69C suffered grain growth, and despite an exceptional careful hardening, a slightly coarsened as-quenched fracture surface, ASTM 8, was unavoidable. Toughness was determined by bending a specimen until fracture occurred and measuring the plastic deflection. The results for carbon steels are shown in Fig. 3. The modern fine grained steels, 80MnV and 82MnAl, had equal plastic deflections as function of hardness. They had some plasticity even at extreme hardness levels (63–65 HRC). The primitive slightly coarse grained steel 69C was more brittle at high hardness levels, but at lower level (57 HRC) it reached the toughness of modern steels.
Plastic deflection vs. hardness for fracture bent specimens of carbon steels with primitive (69C) and modern (80MnV, 82MnAl) chemical compositions. Plastic deflection angles for the steels are equal at 57 HRC, but at higher hardness levels modern steels are superior. (Online version in color.)
Ni steels, 67Ni and 65NiMn, had no grain growth inhibiting precipitates, but they also attained the desired smooth and fine as-quenched fracture surface, ASTM 10 in Shepherd scale. Ni alloying increased toughness at the hardness levels of 58–62 HRC, Fig. 4. Mn alloying had no influence on toughness of Ni steels (67Ni vs. 65NiMn in Fig. 4), but the absence of Mn alloying in 67Ni resulted slightly hot brittleness. Steel 67Ni, unlike the other tested steels, in few instances, hot cracked at the beginning of the forging process.
Plastic deflection vs. hardness for fracture bent specimens of carbon steel (69C) and Ni steels (67Ni, 65NiMn) with primitive chemical compositions. Ni alloying increases plastic deflection (toughness) at hardness level of 58–62 HRC. (Online version in color.)
The maximum hardness of tempered (150°C) microstructure of Ni steels was 62 HRC. Up to this value the Ni steels had at least as high toughness as the modern carbon steels (67Ni and 65NiMn in Fig. 4 vs. 80MnV and 82MnAl in Fig. 3).
When high toughness is desired in a blade, carbon steel with nearly eutectoid composition and relative high tempering temperatures (250–300°C) are traditionally used. According to the present study, this is a good practice for unalloyed carbon steels. After low temperature tempering 69C had exceptional poor toughness (small plastic deflection), but when tempered at 250°C it was very tough, being equal to modern steels 80MnV and 82MnAl (Fig. 3).
Steel 69C had no Mn alloying and its grain size was larger than that of 80MnV and 82MnAl. The exceptional poor toughness after low temperature tempering was obviously due to the result of larger grain size. This can be assumed because the absence of Mn did not decrease toughness when 67Ni and 65NiMn were compared (Fig. 4).
The best result for 69C was obtained when it was tempered at potential temper embrittlement temperature range. The temper embrittlement is often noticed in Charpy V and torsion tests, but bending test or unnotched Charpy does not necessarily indicate embrittlement.9) Charpy V is not relevant with blade steels, because hard steels are in any case notch sensitive, and blades do not have notches. Also torsion test is irrelevant, because blades are exposed to bending load. Hence, temper embrittlement is not necessarily a problem with blades. In the present study temper embrittlement did not occur.
Fine grained V and Al alloyed steels had some plasticity after low temperature tempering (Fig. 3). The Al alloying of 82MnAl was sufficient to prevent grain growth. Thus, it was as tough as 80MnV, in which the prevention of grain growth was extra secured by relative large amount of V. In practice, knife blades made of commercial fine grained Al alloyed carbon steels, according author’s experience, can be successfully tempered as low as 180°C. However, commercial carbon steels do not always have Al alloying to prevent grain growth. Then, like 69C, they need relative high tempering temperature to attain good combination of hardness and toughness.
There is also ancient solution for preventing grain growth. In ultra high carbon steels (1.5%C) cementite particles prevent grain growth. Therefore, ultra high carbon steels attain some toughness after low temperature tempering. But after tempering at higher temperatures, due to carbides, they do not achieve the toughness of eutectoid carbon steels10) Ultra high carbon steels are traditionally used in very hard objects like files and razors.
The studied carbon content 0.7% is typical for old blades, because it provides good mechanical properties, and it can be easily achieved with carburizing atmosphere. In ancient steel making, carburization occurred during bloomery process or it was a separate process. The S content of primitive steel was generally low. It did not depend on the S content of iron ore because SO2 fumes escaped when the ore was roasted before smelting. The P content of steel was about 0.01–0.5%. It depended on the used iron ore because P distributed between metal and slag. The Mn, Si, and Al contents were usually below significant level because they went into the slag.2)
69C resembles those ancient steels which were made of iron ores containing low amount of P. The Si, Mn, and S contents of 69C are low and typical for primitive steels, but the Al content should be discussed. 69C was deoxidized with Al and there is a small amount of residual Al left (67Ni and 65NiMn have less residual Al). The residual Al did not prevent grain growth; therefore, it can be assumed that it did not alter mechanical properties.
Primitive steels contained more or less slag inclusions. A skillful blade smith could reduce the amount and size of the inclusions and thus minimize their harmful effect on toughness. But the inclusions were a problem with 18th century industrial blister steel. In order to improve cleanliness Benjamin Huntsman in 1740s developed European crucible steel. He melted blister steel in a crucible and obtained slag free microstructure. These times in Britain the best blister steel was made by carburization of Swedish iron. It had low P and S contents and when it was used in crucible process the product could be close to 69C.3)
4.2. Ni AdditionIt was suspected that Ni steel is brittle without Mn alloying because Ni enhances the harmful effects of free S.7) But this study showed that superior Ni steel can be made without Mn alloying when P and S levels are low.
Ni alloying clearly increased the toughness of otherwise unalloyed steel at high hardness levels (58–62 HRC) and made the appearance of as-quenched fracture surface fine, but the fracture surface did not necessarily reveal the prior austenite grain size. The used Shepherd scale is proved to be reliable with high carbon low alloy steels, but this is not necessarily the case with high Ni steels. Literature does not claim that Ni prevents grain growth, and this was also verified by an additional overheating experiment. When 67Ni specimen was heated at 900°C for 10 minutes as-quenched fracture surface was apparently coarsened as show in Fig. 5. Thus, Ni had no similar grain growth inhibiting power to Al or V precipitates which are still very effective at 900°C. In the optimal hardening situation, perhaps Ni prevented grain boundary fracture and in that way refined the fracture surface and increases toughness.
As-quenched fracture surfaces of 67Ni after optimal and overheating situations. Optimal hardening temperature was detected by magnet and it produced fracture grain size of ASTM 10 (above), and overheating at 900°C for 10 minutes produced fracture grain size of ASTM 4 (below).
Meteorites generally contain 5–20% Ni.1) In this study Ni content 4% was studied. It simulates the melted mixture of ore based steel and meteorite and therefore the Ni content is less than that of meteorites. Ni alloying up to about 4% is used in modern high carbon tool steels and carburizing steels when high toughness is desired. Higher Ni contents are not useful with high carbon steels due to excessive amount of retained austenite. Thus, the 4% Ni steel is the most interesting from the viewpoint of hardness and toughness.
Meteorites are not necessarily good raw materials for blades due to the combination of high Ni content and impurities. The average chemical composition of an iron meteorite is 0.04C–7.9Ni–0.5Co–0.2P–0.7S. The amounts of other elements are negligible. Most of S is in large visible FeS inclusions, and the S content in solid solution and microscopic inclusions is only 0.01–0.001%. The average P content is high, but there is some exceptions. Buchwald have listed 480 analyzed meteorites and 31 of them contain 0.05% or less P.1)
The author has forged a piece of Gibeon meteorite, which is one of those having low P (7.9Ni-0.04P). Visible FeS inclusions were removed and the remaining S content was measured to be around 0.001 mass%. The meteorite was difficult to forge due to hot brittleness. Non-hardened specimens were tough (ancient blades were often non-hardened), but after carburizing and hardening the specimens were quite brittle and an electron microscope revealed intergranular fracture.
Melted mixture of meteorite iron and ore based steel may be better blade material than the meteorites themselves. If the raw materials with low P and S are used, the chemical composition can be similar to that of 67Ni and the steel may attain superior toughness. This suggests that, as Paul Wellman’s story tells, it could be possible that James Black achieved superior mechanical properties by melting mixture of blister steel and meteorite in crucible.
Totally unalloyed high carbon steel and otherwise similar but 4% Ni alloyed steel were made in a laboratory foundry and their properties were studied in a bladesmith forge. The conclusions from this study are the followings:
(1) Tough carbon and Ni steel can be made without Mn alloying.
(2) Grain growth can not be avoided in unalloyed carbon steel with nearly eutectoid composition. Therefore, it is more brittle than Al or V alloyed steel after low temperature tempering.
(3) Old bladesmith tradition advises to use high tempering temperatures (250–300°C). This can be an optimal treatment for primitive carbon steels. After relative high temperature tempering slightly coarsened grain size do not decrease toughness, and temper embrittlement does not occur with carbon steel blades.
(4) Ni alloying (in otherwise unalloyed steel) refines fracture surface and increases toughness in the case of low temperature tempering. Thus, primitive Ni steel can be tougher at higher hardness levels than primitive carbon steel, and it can be speculated that addition of meteoric Ni in crucible charge produced primitive super steel.
