6 Valyrian Steel, Made in Damascus

The blade was Valyrian steel, spell-forged and dark as smoke. Nothing held an edge like Valyrian steel.

—A Game of Thrones

For the majority of the inhabitants of Westeros and medieval Europe, normal steel did just fine. It wasn’t perfect, but it got the job done. Valyria, however, stepped up the steel game. They devised a method to make steel that was harder, stronger, lighter, and apparently great for dispatching White Walkers. Though in Valyria it was made with dragon fire and spells, which helped give it its superior quality, there is a real-life analogue to Valyrian steel. In the last chapter I talked about how southern Europe and China took iron from ore to blade, but I specifically left out India—that is, the real-life Valyria. India perfected a method of creating steel that could be fashioned into blades that were light and had an amazingly sharp edge and beautifully patterned blade. This steel was sought after by the whole world, but it was Damascus that became the place to turn this steel into blades. The working of the steel created the beautiful trademark ripple pattern on the blade, and it was said the edges were sharp enough to split a human hair in half lengthwise. Like Valyrian steel, these blades were a useful status symbol, passed down through generations of families. Also like Valyrian steel, the methods for making Damascus steel have been lost over the years. Scientists, like the smiths of Westeros and Essos, have not been able to recreate the metal, although they are still trying. There are competing theories as to how one might create modern Damascus steel. The remaining blades are extremely expensive, but unlike Valyrian steel, you can buy knockoff Damascus steel on eBay.

After learning all about the steelmaking process in the last chapter and knowing the reasons why it is so complicated, what set the steelmaking in India apart? The best steel is made using a method that accurately controls how much carbon is in the steel and what atomic form—pearlite, martensite, cementite, or ferrite—the final product takes. This could be done in the steelmaking methods I talked about in chapter 5, but not with precision. The smiths of India, however, developed another method that introduced carbon into iron in a more controlled way. This so-called crucible steel had a better composition than other types of steel. They then figured out how to use heating, cooling, working, and quenching to produce the best combination of cementite, pearlite, austenite, and martensite. In addition, the smiths of Damascus and India accidentally created some of the most advanced materials science of the day. There are two groups of scientists that are in a bit of a tiff about what really made Damascus steel so special. In 1980, two materials scientists from Stanford University, Jeffrey Wadsworth and Oleg Sherby, developed the eponymous Wadsworth–Sherby recipe for making modern-day Damascus steel. J. D. Verhoeven from Iowa State University worked with a number of blacksmiths, mainly Alfred Pendray, to develop another method in 1998. The two groups went through a number of publications and rebuttals, and I can’t say the issue has been definitively solved. Taken together, the two groups’ work, as well as a more recent study by a group in Germany, gives a pretty complete picture of the process and final product.1 This argument took up hundreds of pages in scholarly publications, and only seemed to end when the combatants either died or retired. I wish I could go into the psychology of academic jousting, but alas, I have neither the time nor the training. To be clear, Valyrian steel, like real-life Damascus steel and regular steel, must be forged, not cast. At least in the books, these terms are used correctly: Ice is described as being “reforged.” From here on out, I will assume that no one on the production team of the show bothered to look up the difference. Let’s all forget that dramatic yet incorrect scene.

Raw Materials: Crucible Steel and Wootz

Damascus steel got its name much the way Greek fire did—as in, there’s no real reason it should be named Damascus steel because all anyone in Damascus ever did was sell the stuff. The raw steel cake, called an ingot, was created in India, forged into blades in Persia, and eventually sold in Damascus. Europeans first learned of the steel during the Crusades, when they were on the pointy end of these superior weapons. The process for making this type of steel and then being able to work it into blades wasn’t really figured out by the Western world until around 1982. Although blacksmiths of Western Europe purchased ingots, they were not able to work the steel into weapons due to its brittle nature. The best way to understand what made Damascus steel so special is to follow it from iron ore to finished blade. Unlike the bloomeries discussed in chapter 5, the precursor to Damascus steel blades was wootz steel, a type of crucible steel.2

The first step in creating any steel is smelting iron. In the West, this was done in bloomeries that could get hot enough to separate slag from elemental iron. They were not hot enough to melt the iron, however, unless the iron was at its eutectic point, at which point it would become cast iron. Recall that, in this method, the ore was in the same chamber as the fire, and that carbon and CO were involved in the process. By contrast, the crucible steel process starts out the same way as bloomery iron, with iron ore being smelted into bloom. The bloom was then placed in a crucible, or small clay pot about 8 inches high and 2 inches in diameter with a 0.25-inch-thick wall. Wood chips, specifically Cassia auriculata, and a fresh leaf or two from the Convolvulus laurifolius plant were also added. Later, it was determined that Sorel, an iron–carbon alloy that contained trace amounts of other elements important to the process, was also added to the crucible. These tiny details are important. I also like to bake and make candy, and I can tell you that recipes count and tiny differences in temperature and ingredients can make the difference between an exquisite end product and glop.

The crucible was then placed in a fire and the iron inside was heated to its melting point. Yes, I did indeed just spend a chapter telling you that it wasn’t possible to get a fire hot enough to melt iron. This was only sort of true. It’s not possible to get a useful fire hot enough to melt iron in a bloomery. That’s a lot of qualifiers. Remember that the environment in a bloomery had to be high in CO and low in CO₂. High CO concentrations pull the oxygen out of the iron or reduce it. Once the ratio of CO to CO₂ falls below 3:1, the reaction changes from reducing iron to oxidizing, at which point it’s the exact opposite of what they wanted to happen. The problem is that CO is created through incomplete combustion, meaning the fire isn’t burning completely and is not as hot as it could be. For a fire to reach 1400°C, it needs to be burning in an environment with a CO to CO₂ ratio of about 2:1. In a bloomery, the fire and the steel are in the same chamber, meaning that the combustion environment of the fire is the same environment as the iron. The heat of the fire has to be kept in a range that won’t create an environment that will oxidize the iron. When making crucible steel, however, the environment in which the iron is being heated is completely independent of the fire. The ratio of CO to CO₂ can be whatever is needed to achieve the desired temperature. Modern-day smiths have been able to use ancient methods to heat charcoal fires to temperatures above 1500°C.

The crucible was placed in a charcoal fire and then covered with coal. There are different accounts of how long it would be left there, but it ranges from 2.5 hours to 24 hours. I talked about the eutectic point in chapter 5 and mentioned that at approximately 1100°C the iron takes in carbon until it reaches a concentration of 4% and melts out. This wasn’t a complete explanation, but it was all that needed to be said for that chapter. Look at the boundary line between liquid steel and a liquid–austenite mix. If the temperature is heated right to that line and the steel is melted, it will take in the corresponding percentage of carbon. In the case of 2% carbon, that hits at right about 1400°C. They probably didn’t have this super confusing phase diagram back then—they just knew it worked. Damascus steel has a very high carbon content (roughly 1.5%–2%). Scientists and historians know for sure that the iron inside the crucibles was melted. If the fires couldn’t go beyond 1300°C, the liquid produced would be the eutectic, with a carbon content of about 4%. From there, carbon would have to be removed in a secondary process to get the concentration down to 1.5%–2%. If you look back at the phase diagram of steel (figure 5.2), you’ll see that at a concentration of 2% carbon, steel needs to reach a temperature of about 1400°C to melt. Initially, historians thought there must be a process to remove carbon—decarburization—involved because fires couldn’t get hot enough to melt 2% carbon steel, but there was no real historical evidence for a secondary process. Given the steel being in a crucible and a coal fire’s potential high temperatures, it was possible—and, indeed, likely—that the liquid in the crucible was steel with a carbon content of roughly 2%. After waiting 2.5–24 hours and checking to make sure the contents were liquefied, the crucible was removed from the fire and left to cool naturally; in some cases, water was thrown on them. The end product, a solidified mass the size of a hockey puck, was the ingot. It was often described as having a crystalline pattern on the top.3

Originally, historians thought it was a pure mixture of iron and carbon, but recent developments in materials science and imaging techniques have debunked that theory. The addition of Sorel, leaves, and wood chips, as well as the ore that was originally used, infused the steel with trace amounts of 14 other elements. Three different groups measured the concentration of these additional elements and came to roughly the same conclusions: the trace elements consisted of 0.15% phosphorus, 0.03% manganese, 0.06% sulfur, 0.06% silicon, and smaller amounts of nickel, cobalt, chromium, copper, molybdenum, tungsten, niobium, aluminum, and, most importantly, vanadium. In their attempts to recreate Damascus steel, Verhoeven and Pendray were able to create a similar-looking steel on only a small number of occasions. The majority of the time, the blades were missing the characteristic wavy pattern (figure 6.1). Verhoeven and Pendray realized they hadn’t been using Sorel iron, even though it had been listed in historical records as an ingredient in the ingot. In their paper, they admit this was a gross oversight. When they added the Sorel iron to the ingot, they were able to create blades with the characteristic pattern at a much higher rate of success. They wanted to find out what was so special about Sorel iron. After determining the composition of the final steel, they sought to identify which of the 14 trace elements was needed to create the pattern. Vanadium was found in both the Sorel iron and the ingot, but because it was found at such low levels, Verhoeven and Pendray ignored it for several years. Eventually, they started the process with normal iron and slowly added in the trace elements until they found the ones that best resembled the patterns on true Damascus steel. They focused mostly on the carbide-forming elements, meaning the ones that will bond with carbon. Of the trace elements, they tried vanadium, molybdenum, chromium, niobium, and manganese, with vanadium and molybdenum being the most important. In particular, the vanadium helps the cementite particles orient themselves in a single direction instead of all willy-nilly. It’s this order that creates the distinctive rippling pattern. I’ll talk more about how cementite and other structures affect the properties of the blade in just a minute.4

Working Ingot: Not Europe’s Sharpest Moment

Though many were able to purchase wootz ingots, very few were skilled enough to actually do anything with them. The Europeans were said to have had their finest blacksmiths attempt to forge (or take from ingot to finished blade) the famous swords to no avail. In fact, as far as I can tell from my research, it wasn’t until 1982 that anyone in the West was able to do it.5 The barrier was the fact that 2% carbon was just too brittle for western blacksmiths to work with. They were used to working with steel with a much lower carbon content, which made it more ductile and therefore less hard and brittle. When they tried their normal techniques on wootz steel, it simply shattered under their hammer. This issue and any potential solutions come down to the phase diagram for steel. This was included in the previous chapter, but I’m including it again here for easy reference (figure 6.1).

In Europe, they worked with steel at approximately 1300°C. If you remember from the last chapter, this was the range in which steel was in its austenite phase. When the blade is cooled from there, it drops down into the pearlite region, or it can be quenched to form an outer layer of martensite with a pearlite core. Now, look at that same temperature but move over on the x axis to the 2% carbon label. You’ll see that at 1300°C the steel would be a combination of austenite and liquid steel. When it’s hit with a hammer it shatters. You can see from the phase diagram that it would be better to work this type of steel in the austenite (abbreviated γ) and cementite (seen here as Fe₃C, its chemical formula) range of 700ºC–900°C.

In a 1999 paper, Oleg Sherby discussed in detail what he called the Wadsworth–Sherby method for making Damascus steel blades. Verhoeven gave the method that name in a paper saying how it was wrong, just to make sure the world knew exactly who was wrong.6 Verhoeven argued that although the method produced blades that on a microscopic level looked and behaved like Damascus steel, imaging techniques showed that the structures were different at the microscopic level. Wadsworth and Sherby then rebutted his assertion by saying Verhoeven didn’t know what he was doing when he followed the Wadsworth–Sherby method. In a 2001 paper, Verhoeven again asserted that the Wadsworth–Sherby method didn’t work and described his own method and the role of impurities. What is interesting about Wadsworth and Sherby’s method is that they don’t mention how impurities might give rise to the banding seen. Verhoeven gave a pretty convincing argument that the impurities were needed to achieve banding, and this was not taken into account by Wadsworth and Sherby. What they both agree on, though, is that the blade, at 2% carbon, must be forged in the region where it is austenite and cementite, between 700ºC and 900°C. But enough about the fight—let’s hear about their methods.

Wadsworth and Sherby were the first to publish results saying they had achieved Damascus steel. In a 1999 paper, they laid out a three-step process for producing a blade.7 First, the wootz is heated to about 1100°C, putting it in the austenite phase. It is then rolled to mix everything around and fully integrate the carbon. The wootz is also cooling down during this process. Second, the wootz is then reheated to 1100°C and left there for 48 hours. Being kept at this temperature for so long allows the austenite to arrange itself in bigger crystals and create longer-range structures. This means there are fewer but longer grain boundaries instead of lots of little ones. The steel is then cooled very slowly into the region of austenite and cementite. The austenite is already sitting in nice big crystals, and the newly forming cementite doesn’t want to get in the way of that. Instead, it grows in long strips along the grain boundaries. At this point, the steel has been rolled but not really worked; if it was etched, however, it would be possible to see the characteristic stripes. Third, the wootz is then heated right to the boundary between the pearlite and austenite–cementite regions. This causes the cementite that didn’t make it into the long strands to dissolve, leaving austenite and long networks of cementite. The steel is then rolled to break up these long networks. It remains ordered in strands, to some degree, but they are not as long. When cooled to room temperature, the cementite strands lay in similar directions and appear bright after etching on a background of dark ferrite. With this new structure, the blade has amazing strength and durability.

The method extolled by Verhoeven and Pendray is similar in many ways but different enough to cause a feud rivaling that between the Starks and the Lannisters. Both teams realized that working in the 700ºC–900°C range was critical, but the biggest difference lay in their explanations of where the bands came from. Verhoeven and Pendray skipped step 2 of the Wadsworth–Sherby method and went straight to forging at the same crucial temperature. As they worked the steel, it cooled. Before it fell below the needed temperature, they heated it again and worked it again. It took about 50 cycles of heating, working, and cooling to hammer the steel into a blade. They wanted to make sure to take into account the role of impurities, particularly vanadium, in their theory. They suggested that as the metal was heated, some cementite would dissolve, as theorized by Wadsworth and Sherby, but Verhoeven contested their hypothesis, instead suggesting that the formation of cementite strips was nudged along by impurities. These impurities prevented the cementite from fully dissolving in the liquid metal, allowing large particles to remain. Each time they underwent this process, the bigger cementite particles attracted more cementite and grew slowly. After about 20 cycles of this, it became possible to see the characteristic bands. In the end, both groups produced steel with cementite wires in a matrix of ferrite, causing visible bands. Wadsworth and Sherby make no mention of the role of impurities, while Verhoeven gives impurities credit as the critical ingredient for making Damascus steel. Wadsworth and Sherby give wootz with a carbon content of 1.8% as their starting point. They do not, however, say whether they are using wootz that they made in their lab containing only carbon and iron, or whether they obtained wootz that included defects, as historical wootz would.8 Answering this question is the key to figuring out which group’s theory is correct. In the end, they both end up with long bands of cementite, which turned out to be cementite nanowires, and a really cool-looking sword (figure 6.2). In 2004, a group from Germany decided to make a better image of the edge of a Damascus blade and see if everyone had been missing something. Turns out they had been.

Really Ancient Technology Meets Really New Science

The work by Verhoeven and Pendray and by Wadsworth and Sherby was carried out in the 1980s and ’90s. Materials science has advanced significantly since then. The first paper in that series was published in 1982, the year the Commodore 64 was released. (That was a computer, not a band.) Since then, processors have gotten faster, cell phones are now powerful pocket computers, and the SUV was invented. The discovery relevant to this discussion, however, is the carbon nanotube. Yes, in just two chapters we have gone from the Stone Age to the Information Age. In 1991, Sumio Iijima of Japan was the first to discover carbon nanotubes. As with many major scientific advances, there is much controversy regarding whether or not he was the first to discover them, but he was indeed the one to make the scientific community take notice. Carbon is the most abundant element on Earth and it can be structured in many different ways. Graphite, diamond, and amorphous carbon are just a few of forms it can take. The properties of these different forms vary wildly. For instance, graphite is very conductive. You can draw with a pencil on a piece of paper and use it as a wire, albeit kind of a crappy one, but who’s ever heard of conducting electricity through the diamond in your ring? Carbon also has the ability to organize into a flat sheet only one atom thick. This is called graphene, and its discovery won Andre Geim and Konstantin Novoselov the 2010 Nobel Prize in Physics. We’ve talked about BCC and FCC crystal structures, but graphene has a honeycomb lattice shape (like chicken wire) and a whole lot of interesting properties. When this single layer of carbon atoms is rolled into a ball, it becomes the famous soccer ball–shaped buckminsterfullerene (or buckyball), but if it’s rolled into a tube, it’s a carbon nanotube (CNT).

Carbon nanotubes have some pretty amazing properties. One of the pioneers of carbon nanotubes, and my personal hero, was Mildred Dresselhaus, affectionately known as “the Queen of Carbon.” She specifically looked at the electrical properties of carbon nanotubes. (This isn’t particularly relevant to a discussion of swords forged before the discovery of electricity, but she’s awesome and more people should know about her work. You might recognize her from a wonderful GE commercial.) What we’re really interested in with respect to carbon nanotubes is their strength and how they might form in steel. CNTs are actually much stronger than steel, even Damascus steel, and they are not only strong but also very springy. This is pretty much an ideal combo for weapons-grade steel. It takes five times more energy to bend a CNT than it does steel. Unlike steel, the strength comes not from the tiny dislocations and grain size but from the fact that it doesn’t have any dislocations. A nanotube is a single crystal, and its honeycomb crystal structure isn’t conducive to the slipping seen in BCC and FCC structures. The covalent bonds between the carbon atoms are extremely strong, which leads to a tensile strength that is 100 times that of steel. If you do happen to bend one, it will bounce right back into place. These little guys are practically indestructible. And, as it turns out, they are found on the edge of a Damascus steel blade, reinforced with cementite.

Alexander Levin and his group in Dresden, Germany, began examining the edges of Damascus steel blades just as nanotubes and nanotube imaging began in full swing.9 The team was not satisfied with the current explanations for the blade’s strength and the emergence of its pattern. They chose to look at how the various elements were distributed throughout a blade to in hopes of getting a better understanding about the microstructure of the steel. They convinced a nice person at the Berne Historical Museum to give them a small piece of an actual Damascus steel sabre. This is the same weapon Verhoeven looked at. Since the piece was cut, they were able to image it both from the front and through a cross section.

They started out, like Verhoeven, by identifying which elements besides iron were actually in the blade. Levin’s team ended up with numbers very similar to Verhoeven’s but different enough to make them conclude the composition of the blade was not uniform. They looked at the crystal structure using X-ray diffraction, which involves bouncing X-rays off of a structure and seeing how they are reflected. This can give an idea of how the object reflecting the X-rays looks. In the blade, they found ferrite, cementite, martensite, and graphite. Though the martensite was distributed rather evenly throughout the blade, the cementite and the ferrite were found in different concentrations throughout their sample. Cementite concentration was higher on the outside of the blade toward the back. There wasn’t much in the bulk steel, nor near the cutting edge. They also found that cementite was present in three phases: first, as part of the layered pearlite mentioned in chapter 5, and second, as grains of cementite, as expected, but also in cementite nanowires. This was the first time nanowires of cementite had been observed. Using transmission electron microscopy, a technique where a beam of electrons is shot through a sample and an image is recorded on the other side, they were able to look more closely at the newly discovered nanowires. The researchers found that all of the nanowires ran parallel to each other, but that they do not appear in all sections of the blade. They may line up locally, but they don’t have a preferred direction in the bulk. This means that the nanowires may be aligned one way in a small section of blade, but each section on the blade could be facing in a different direction. In addition, these nanowires didn’t appear in all sections of the blade. As seen in normal cementite, there is a way ferrite, cementite, and martensite crystals like to line up. This supports the theory posited by Wadsworth and Sherby. Levin’s team theorized that the nanowires were formed by a method similar to that described by Wadsworth and Sherby, but that impurities served as nucleation sites, or places where tubes would start forming. This hypothesis incorporates the findings of Verhoeven as well.

A year later, the group took even higher resolution images of the blade edge.10 They found that not only did the blade have cementite nanowires, it also had carbon nanotubes. They found these after dunking the blade in hydrochloric acid in a controlled way. This caused the cementite to dissolve, thus exposing the carbon nanotubes; however, they found that this happened only in places where the cementite hadn’t fully dissolved. They concluded that the carbon nanotubes were protecting the cementite by encasing the nanowires. Another group of scientists showed that cementite could indeed crystallize inside a carbon nanowire.11 It just so happens that molybdenum, a key impurity found by Verhoeven, has the ability to stimulate carbon nanotube growth.12 This might bring the whole argument full circle. Thanks to Levin’s and Lei Ni’s teams, we now know that Wadsworth and Sherby and Verhoeven and Pendray were both on the right track, but neither had the full picture This is why science is so cool. The forging of the steel potentially pushed the necessary impurities around along the plane in which the steel was being worked. As the impurities moved in the same direction around different parts of the blade, they were able to nucleate more nanotube formation. It’s possible that cementite was then able to crystallize inside the carbon nanotube. As soon as I learned about carbon nanotubes, I wanted to build a faster bike out of one, but I think I’d be happy to settle for a sword.

“Damascus” Steel on eBay

By this point, you might have tried to find out where you can get one of these awesome blades for yourself. You might have entered “Damascus steel” into Google and got more than a few listings for things made of “Damascus” steel, everything from hunting knives to wedding rings. You also might think I’m full of crap for going on and on about how special this stuff is, when for a few hundred dollars you, too, could own some real-life Valyrian steel. Here’s the catch: the internet stuff isn’t real Damascus steel. It has the telltale pattern and sure looks the same. It’s even called the same thing. But, unfortunately, it’s not made the same way. It’s a completely different kind of steel called welded or pattern steel. The goal is to make something pretty that is also functional, not make the world’s most kick-ass sword that also happens to look cool.

Perhaps I am not being totally fair. The first pattern welding was done with very early steel to even out the problems of inconsistency. In the last chapter, I talked about how it was not easy early on to make steel and those who did make it couldn’t usually make good steel consistently because it was hard to control both the temperature of the fire and the carbon content. The Chinese tried to overcome this problem by wrapping wrought iron in cast iron and letting them mix. It was a pretty successful practice. Pattern welding was even simpler than this co-fusion system. Blacksmiths would take multiple sheets of steel, each with slightly different properties, and heat them all up on top of each other. Then, they would be forged together into a single blade. The idea was that if one layer was hard and another was tough and they were forged into a single blade, the sword would have macroscopic properties between the two. The Celts were particularly fond of this method. Different layers had slightly different colors of steel since the final composition of the steel varied depending on how the bloomery was operated. As the steel was forged together, the different colors of steel wove together to create a beautiful pattern. The blade may then be etched or ground to further bring out the pattern. Smiths began twisting metals together instead of laying them flat to create even more intricate patterns. Eventually, as steel production improved, the creation of patterned steel weapons was a case of form over function. Seeing that it is even used in jewelry nowadays, modern pattern-welded Damascus steel is a decorative item rather than one that is prized for its cutting ability. I think the current pattern-welded steel is beautiful and a nice addition to any home or sword collection; however, one should be aware that it won’t cut through a falling human hair, nor will it kill a White Walker.

Valyrian Steel and White Walkers

Valyrian steel and Damascus steel seem very similar. Both are made from steel that is unique to a region, and both are both are highly valued because of their unusual properties and difficulty to forge. The method for creating a Valyrian steel ingot (assuming it started as an ingot) has been lost, but the crucible method of forging Damascus steel was well known. In both cases, there are few people who know the specific recipe to forge a blade. Both types of steel are known for their distinctive swirl pattern and both are worked a number of times at specific temperatures to produce the precious final product. I think it has been assumed, if not outright stated, that dragons were involved in the production of Valyrian steel, either in the original smelting or crucible process or in the actual forging. Considering Westerosi smiths were still able to work with the remaining metal, it would make sense that the dragon fire was involved in the front end of the process and not the forging. It is unclear whether dragons can fully control the temperature of their flames. Up to this point, they’d just want it as hot as possible, but if they could control the temperature, they could produce steel with well-regulated carbon content and crystal structures. Seeing as this was a major hurdle in producing real-world steel, dragon thermostats would really help. I think by now we can assume dragons give materials a healthy amount of magic with their fire, which can’t hurt either. It is also said that the blades were quenched in blood, in line with House Targaryen’s motto. This isn’t far from reality, either; some Persian texts recommended quenching the blade in the belly of a slave. There were also texts saying it should be quenched in the urine of a red-headed boy, so maybe wildlings would have a job if they moved south of the wall.13 In general, and despite the dragon fire, it’s really fascinating that GRRM’s “fictional, magic metal” isn’t all that fictional after all.

The biggest question I had about Valyrian steel, or Damascus steel, is how it would it behave at low temperatures, specifically those in which regular steel fails against White Walkers. As you can imagine, there’s no research into how temperature affects the strength of Damascus steel because it doesn’t get that cold in the Middle East. No one is going to build a ship out of it. There is probably a ductile-to-brittle transition, as with normal steel, given that most of the blade is made of a similar material. But carbon nanotubes, which protect and strengthen the cementite and allow the blade to hold an edge, becomes stronger at lower temperatures.14 Also, CNTs do not have a ductile-to-brittle transition because they are formed with single-layer honeycomb lattice rather than BCCs and thus don’t have the weaknesses brought on by dislocations. These nanotubes make up only a very small percentage of the Damascus steel blade, so the addition might not do much. However, seeing as they are protecting the cementite nanowires, they might be able to hold the crystal together, similar to how chicken wire is used to reinforce concrete. I haven’t been able to find research on this specific question, but I, for one, am going to hope it’s the case. I’d be happy if it was spells and science holding this one together, unlike the Wall. We’ll never officially know if Damascus steel could kill a White Walker, but I’m pretty glad that remains a mystery.