Making a welded Damascus Knife

John Neeman Tools have posted a beautiful video of manufacturing a welded Damascus patterned knife.

5 layers of 3 different steels were forge-welded, folded, and forged. With each step being repeated 8 times. This produces a patterned with 320 layers. Finally twisting and forging the steel produces a more complex pattern.

Just checking the number of layers, I get their total to be different. My calculation of the number of layers is 5 × 28 = 1280, that is 4 times more than claimed (320 layers should be the result of folding 6 times (6 folds 5 × 26).

With 1280 folds, if we assume the thickness of the knife is 2 mm, that means each layer is 1.6 μm, 320 folds would be 6 μm layers. These are both lower than what can be resolved using the naked eye. It’s very close to the wavelength of visble light — if the metal were folded one more time, or the final thickness of the knife is less than 1 mm you would be there.

MS Calculator Android App

A researcher at GIFT has produced an metallurgical application for android phones for calculation of MS, ε-MS, Ae3, Ae1, and conversion of weight percent to atomic percent.

MS Calculator Android App

MS Calculator Android App, screen shot 1

Android App

MS Calculator Android App

MS Calculator App webpage

Tempering of Martensite

Stage 1
upto – 250 C – precipitation of eta-carbide, partial loss of tetragonality of martensite.
Steels above 0.25 wt% carbon precipitate hexagonal close-packed eta-carbide within the supersaturated martensite until 0.25 wt% carbon level is reached, martensite preserves some tetragonality. The orientation relationship between the laths or rodlets and the cube planes of the matrix was first described by Jack [1].

Stage 2
200-300 C – decomposition of retained austenite.

Stage 3
200-350 C – replacement of eta-iron carbide by cemeneite,; martensite loses tetragonality

Stage 4
350 C – cementite coarsens and speroidizes, recrystallisation of ferrite.

[1] Steels 2nd Ed, Honeycombe and Bhadeshia, Edward Arnold, 1981, p172.

Making Podcasts

More podcasts this weekend…

Crystallographic texture and intervening transformations

Stress of Strain affected martensitic transformation

Quantitative Metallography of Deformed grains

Steels for Fusion

New Podcasts

I made three new podcasts with Prof. Harry Bhadeshia on his latest papers on transformation texture, the new delta-Trip steels and on prediction of Hot Strength of ferritic steels.

The work on transformation texture is from Saurabh Kundu’s thesis were Patel and Cohen’s model has been shown to correctly predict the orientation relationship between ferrite and austenite after martensitic transformation. It’s shown that variants are selected by free energy differences that can be calculated depending on the orientation.

The delta-Trip steels were developed as a result of the prediction of neural networks, were after the neural network was made computer optimisation was used to try and maximise the mechanical properties. This work was done with Saurabh Chatterjee in collaboration with Murugananth Marimuthu. Both Saurabh Kundu and Saurabh Chatterjee completed their PhD’s at Cambridge while visiting from Tata Steel, Murugananth Marimuthu is a previous member of the phase transformations group, and has now also joined Tata Steel’s research and development section.

The work on Hot Strength of ferritic steels is the part of Radu Dmitriu’s topic of research. A neural network model of the hot-strength of ferritic steels. It was observed from the neural network that the strength is expected to suddenly start to decrease at 800 Kelvin, which can has been explained to be due to changes in the mobility of dislocations.


According to wordpress documentation these links should get added to the rss feed of this bainite blog as enclosures.

Podcast: Hot Strength

Podcast: Delta Ferrite

Podcast: Transformation Texture

Subcribing to RSS Feed of this webpage should get you all the podcasts, or for Podcasts Only
Subscibe to RSS Feed for posts I remember to add to Podcasts category

Martensite kinetics

Martensite transformation usually occurs on cooling, with the volume fraction of martensite depending upon the temperature as shown in the first two examples below. Martensite transformation is often thought only to depend upon the temperature, but has been shown to also depend upon the time. It should therefore be represented by it’s own C-curve on a time-temperature-transformation graph. The strong temperature dependence may be caused by the large driving force necessary before transformation can proceed.

Classification of martensite kinetics after Guimaraes

The Koistinen-Marburger equation

This equation due to Koistinen[1] and Marburger can be used to estimate the fraction of martensite as a function of temperature.

Steels were quenched to various temperatures and the austenite volume fraction determined by X-ray diffraction. The results plotted logarithmically against Ms-Tq, where Ms is the martensite start temperature and Tq is the quench temperature.

V_gamma = exp [-b(Ms-Tq)] where b is 1.1×10-2

Magee[2] later showed that an equation of this form can be justified based on martensite nucleation theory.

Martensite volume fraction predicted by Koinstenen Marburger Equation for various MS temperatures

Rearranging the formula is simple to give the volume fraction of austenite transforming to martensite on cooling to room temperature as a function of the MS temperature of the steel.

Volume fraction transforming to martensite on cooling as a function of the MS temperature.

Khan and Bhadeshia [3] have produced a modified equation which also considered autocatalytic nucleation of martensite to allow better matching to dilatometry results. This changed the goodness of the fit by regression from 0.9 to 0.94, when analysing the data of volume fraction against temperature obtained from their own experiments. This work used one alloy (300M) but changed the starting conditions by partial transformation to bainite.

[1] D. P. Koistinen and R. E. Marburger, Acta Metallurgica, 7, p59-60, 1959
[2] C. L. Magee, The nucleation of martensite, In “Phase transformations”, p115-156, Ed: H. I. Aaronson and V. F. Zackay, ASM, Metals Park, Ohio, 1970.
[3] S. A. Khan and H. K. D. H. Bhadeshia “Kinetics of Martensitic Transformation in Partially Bainitic 300M Steel” Materials Science and Engineering A, Vol. A129, 1990, pp. 257-272.

L6 Steel Kinetics

This steel is popular for use in knives and swords like the Samurai sword “Bainite Katana”.

Custom Howard Clark L6 Katana
The Bainite Katana is made of a special purpose low-alloy steel. It is very resistant to bending, to the point of near unbreakability. These blades can be made lighter and thinner and still remain stronger than conventional steel or 1086. The blades are also springy rather than soft, they will flex more than a normal blade, but the shape will not be altered. These blades are excellent for tameshigiri as well as general sword work.

Composition; I found the composition of L6 is something like: Fe-0.7C-0.5Si-1.75Ni-0.5Mo-0.25V-0.25Cu.

I calculated the time-temperature-transformation diagram, i’m not too sure about the accuracy bainite start temperature using this program, I would like to be able to find the experimental results for this alloy.

Depending on the bainite start temperature, the kinetics are fast enough to allow isothermal transformation to bainite on a reasonable timescale, at temperature of around 300 C which would give a high strength bainite. Howard Clarke sells the swords with as a martensite/bainite sword. If the bainite start temperature is higher it would be easy to get a mixed microstructure.

I should try to make a prediction of bainite start temperature – I made a neural network model for this.

AISI L6 TTT diagram calculated with MAP program MUCG83

Even without carbide precipitation transformation results in a large volume fraction of bainite.
AISI L6 Volume fraction calculated with MAP program MUCG83

Presumably if the bainite Katana is used in marteniste/bainite condition it is produced by continuous cooling or by quenching. Quenching would be good if it can be done in at a rate which gives martensite on the outside of the blade for sharpness and hardness, and bainite in the centre to give toughness. The other possibility is that the continuous cooling gives a mixture of bainite and martensite at every location. This can have a higher strength or hardness than martensite alone because austenite will get enriched in carbon as the bainite transforms, increasing it’s contribution to the strength.

I’d really like to have a look at the microstructure of these swords to see what structure the bainite has. Also it would be interesting to measure the mechanical properties of the alloy in the same condition – I need to look for some literature on L6 I guess. It looks like they should be strengthened by carbides and by copper precipitation.

Howard Clark who makes these swords has a webpage at

–edit 6 October 2007–