Tesla's Gigacasting Alloy // Patent Application & Science
Welcome back everyone! I’m Jordan Giesige and this is The Limiting Factor. Today we’re gonna review Tesla’s patent application for the cast aluminum alloy that Tesla’s using in their gigacasting process. What’s so special about the alloy? First, it has good flowability which will reduce defect rates and allow for intricate castings. Second, it’s strong and ductile enough that it doesn’t require heat treatment, which is expensive and time consuming.
Third, it’s corrosion resistant enough that it won’t require a coating to protect the casting, which reduces production costs and provides a better end product for the customer. But, the qualities and benefits of an alloy are only part of the story. What’s the science behind alloys, how difficult are they to develop, which metals provide which properties, and is this alloy really unique? Before we begin, a special thanks to my Patreon supporters and YouTube members. This is the support that gives me the freedom to avoid chasing the algorithm and sponsors, and I hope will eventually allow me to do this full time. As always, the links for support are in the description.
Back to QuantumScape. Let’s pick up with where we left off in the first gigacasting video, which is this slide from Munro and Associates. The slide refers to Tesla’s casting alloy as AA 386. AA stands for the Aluminum Association, which is the organisation that developed the naming system for aluminum alloys.
Cast aluminum alloys use a 4 digit code, and the first number indicates what some of the minor alloying elements are. Any aluminum alloy used for casting that starts with a 3 is primarily aluminum, and the next most abundant elements are Silicon plus Copper and/or Magnesium. The number trailing the decimal tells us whether the alloy is a casting or an ingot, so we can ignore anything behind the decimal. The digits trailing the three are arbitrary but done in series.
In this case those digits are 86. I couldn’t find 386 aluminum on any official lists. That may mean that this is a new alloy in the eyes of the aluminum association, but I wasn’t able to confirm it. Regardless, in my view, what’s more important is whether patent offices view this as a unique alloy. It appears they do and that will be covered at the end of the video. Why would Tesla use an aluminum alloy instead of pure aluminum and why would they need a unique alloy? The answer to the first question is simple and the answer to the second question will be answered later in the video.
Let’s build a base of knowledge and work up from there. Aluminum alloys are used in products rather than pure aluminum because even small percentages of elements such as copper and magnesium can drastically improve the mechanical characteristics of alloys vs pure aluminum. For example, aircraft aluminum is 90% aluminum yet it's an order of magnitude stronger per unit of density than pure aluminum. Why are alloyed metals stronger than pure metals? To understand that, we need a basic understanding of how strength is measured in metals, what alloying is, and then how alloying increases strength. Strength can be measured by pulling a rod of material apart.
The results of the test are illustrated with stress-strain curves in a stress-strain diagram, which tells us not only the strength of the material, but also ductility. Stress measures the amount of load applied to the rod before it snaps, which is a measure of strength, and strain measures how much the rod stretches under that load before it snaps, which is a measure of ductility. A ductile material will deform before it snaps, while brittle materials snap before they’ll deform. That is, ductility can be desirable because brittle materials are more likely to chip and crack.
Alloying means combining two different metals to form a better metal than either of the two materials had individually. This is generally done by melting and mixing the metals. So why does mixing two metals together produce a metal that’s superior to the metals you started with? Solid metals are actually crystal structures that operate at two levels of scale. First, at the sub-nanometer scale, where substitutions can be made in the atomic matrix of the crystals.
Using larger or smaller atoms creates stresses in the matrix that make it more difficult for the atoms to slide around, which makes the material stronger. Second, at the micron scale, where the size and shape of the crystal grains comes into play. For example, smaller crystal grains create more surface area, which creates more snags when stresses are applied to the metal, which results in higher strength. In both cases, making the metal stronger can decrease its ductility. This isn’t always the case, but it’s often the case.
Ductility and strength often run inverse to each other because the additional snags within a stronger crystal structure makes the material more likely to fracture rather than to bend. The trade of ductility for strength is usually acceptable because in many use cases, strength is more important than ductility. To summarise, alloys create complex metallic crystal structures that open the door to stronger materials by controlling the composition, shape, and size of the crystals. This often creates trade-off decisions between strength and ductility.
I’d like to say thanks to Billy Wu for taking the time to answer my questions about alloys and for the videos he’s posted on his YouTube channel. As usual, any errors in this video are due to my attempt to boil things down into a quick crash course. If you’d like to dip deeper into alloys and with more precision than I’m covering here, I recommend checking out his channel.
How difficult is it create a new alloy and to find the right balance between strength and ductility? This phase diagram for steel from ‘The Efficient Engineer’ YouTube Channel shows how the crystal structure of a material changes under different conditions. It uses just two variables: temperature and carbon % by weight, which results in a range of steel crystal structures, such as ferrite and austenite. If we change the two variables from temperature and carbon percentage to temperature and time, it results in a time-temperature-transformation plot. In other words, with just two elements, iron and carbon, and two dials, time and temperature, formulating alloys becomes complex.
This is part of the reason why we continue to see improvements in steel despite the fact that steel’s been around for thousands of years. The broad variability of time, temperature, and pressure combined with a large set of alloying elements creates a set of potential outcomes that’s nearly infinite. Then, you have to have the technology in place to produce these materials economically and at high quality.
This brings us full circle to the chemical analysis from Munro and associates on Tesla’s 386 alloy. There are more than 12 different alloying elements involved here, not just two. In the original gigacasting video, I made a note posing the question “Can we assume all these elements are active materials.” After looking into it, I can confirm all of these elements have the potential to be active materials, and Tesla names 10 of them as relevant to their application. If all the element percentages here are tallied, it comes to 99.67%, which means that either
the analysis excluded the impurities or the reason why it’s not tallying to 100% is due to precision deviations. Regardless, if these elements have the potential to be active materials, we need to understand what these elements do in Tesla’s casting alloy. We’ll come back that later in the video.
First, let’s start by taking a look at Tesla’s patent application titled “Die Cast Aluminum Alloys for Structural Components”, which was filed on the 20th of January 2021 and published on the 29th of July 2021. The inventors are Jason Stucki, Grant Patinson, Quinlin Hamill, Avinash Prabhu, Shiv Palanivel, and Omar Lopez-Garrity. All of the inventors have been with Tesla for quite some time, so there’s no way to know when they started working on this alloy. The patent application starts by claiming that the invention relates to a cast aluminum alloy that has improved strength, ductility, and castability. In other words, the inventors just added one more dimension to the complexity of their alloy selection process beyond strength and ductility - castability.
How did they measure strength, ductility, and castability? For strength the inventors used the stress test described earlier in the video, but for ductility they preferred a bend angle test. They bent 3 mm thick pieces of aluminum and measured the angle that the piece of aluminum snapped. Strength and ductility were important to them for obvious reasons.
The stronger the alloy, the thinner and lighter the gigacastings can be made. The more ductile the alloy, the more energy it can absorb before chipping and cracking. This is important for gigacastings because they’ll be absorbing a lot of energy from the road through the vehicle suspension.
For castability, the inventors used a flow test. They used a 700 ton high pressure die casting machine with a casting mould that contained a groove 3mm thick and 2 meters long to test how far the alloy would flow before cooling to its solid state. Flowable alloys fill casting dies more completely and quickly, which minimises defects and allows for intricate structures. At this point, the inventors add another requirement for the alloy, which that it must be useable as cast.
Typically, several steps are required after casting such as aging, heat treatments, quenching, and coatings. These extra steps upgrade the cast material so that it’s strong and durable enough to be used in vehicle structures and so that it’s non-reactive with chemicals like road salt. Let’s pause here for a moment of appreciation. As we said earlier, an alloy with two process inputs such as temperature and carbon % or temperature and time, is complex to master. That’s further compounded by the need to balance the output variables, such as strength and ductility.
Tesla is using an alloy with at least 10 elemental variables, plus temperature and time. Then, in terms of output variables they wanted an alloy with the following characteristics: 1) A combination of strength and ductility that’s better than other non-heat treated cast Aluminum alloys 2) Great flowability to fill large and intricate moulds with minimal defects 3) Useable as cast, with no heat treatments required to upgrade the casting. 4) Low reactivity with compounds such as road salt that might corrode other aluminum castings. This is why I said in the first gigacasting video that the materials science going into the gigacastings is SpaceX level stuff. In fact, after that video, I found that the Vice President of Materials Engineering at Tesla, Charles Kuehmann, actually holds the same position at SpaceX.
Furthermore, I did a bit of digging on Charles and found that he’s an expert on computational design and has a history of designing materials for aerospace applications. That is, the head of Tesla’s materials team does in fact have a history of working at the bleeding edge of materials design in aerospace applications, making him perfect for SpaceX, and Tesla also gets to leverage that talent. This is the general process he uses to develop a material. Yes, there is a lot of computation involved, but there are also trials to test the model.
A model is not reality, so it needs to be calibrated through real world testing. You can’t just punch requirements into a computer and have it spit out alien technology – at least not yet. Developing new materials is more difficult than that, and more exciting because you discover things that the model didn’t predict.
Let’s get deeper into the patent. Figure 1 in the patent shows the landscape of cast aluminum materials that were available to Tesla. The higher the bend and strength, the better. The inventors didn’t specify what was meant by the baseline that I’ve marked with the orange dot. I’m assuming it refers to an existing alloy that had most of the properties Tesla needed, but not all.
The baseline alloy has a strength of 122 megapascals and a 25 degree bend angle. The green X is the minimum viable target Tesla had for their alloy, at a strength of 135 megapascals and a bend angle of 24 degrees. The pink dot indicating Final Composition is the aluminum alloy that the inventors highlighted for the patent application. It has a strength of 143 megapascals and a 25 degree bend angle. That is, Tesla’s alloy is 17% stronger than the baseline alloy that they had before starting their investigation. It does that while maintaining the same ductility.
This would allow for castings that weigh and cost 17% less than those made with the baseline alloy. For each Model 3 Y, that would mean a weight reduction of up to around 14 kilograms and a cost savings of more than $30 at the materials level. This doesn’t include other savings from skipping the coating and heat treatment steps. Bear in mind that the final alloy composition is different than what was indicated by Munro and Associates’ alloy test results.
This means the inventors were probably able to improve upon the alloy they refer to here as the final composition. Munro’s analysis is also different than the element ranges laid out by, for lack of a better term, the recipes in the patent application. Does that mean that the alloy Tesla’s currently using in their vehicles isn’t covered by this patent? I don’t know enough about patent law to hazard a guess, but I will say that it has element ranges that mostly fall within the element ranges listed in the recipes. I also don’t know if Tesla has another patent application on the way to cover the alloy they’re currently using. Regardless, as we’ll see later in the video, the case for this patent is strong, which I would think bodes well for adjacent patents or addendums to this patent if those are possible.
Let’s review where we stand. We have a clear view of the challenge that the inventors were presented with and its difficulty, we have a general understanding of their process, and we know that the materials team was able to meet and exceed that challenge. Now we get the chewy part: What elements in the alloy are doing what and how does Tesla’s alloy compare against a benchmark? We’ll use Munro and Associate’s analysis of Tesla’s aluminum alloy because it’s the production version of the alloy and compare it to A380, which is a common cast aluminum alloy. Note that everything I’m about to cover is my best guess based on my own research and that each element I mention here has an array of effects on the alloy beyond what I’ll be focusing on. I’ll be speaking to what I believe are the important active elements and what their primary function is. I won’t be covering every element in the alloy and every impact that each element can have.
Aluminum has a relatively low melt temperature, about half that of steel, making it ideal for casting. It also has a good strength to weight ratio and low density, making it a great candidate for vehicle bodies. A380 usually runs about 80 to 85% aluminum, whereas Tesla’s alloy runs about 90% aluminum.
This should make Tesla’s alloy slightly cheaper because Tesla’s using less Nickel and Copper. It should also increase the corrosion resistance, because aluminum naturally forms an aluminum oxide layer in open air, which makes the surface less reactive. Copper is one of the primary alloying metals used in aluminum castings to add strength.
But, copper also creates corrosion hazards and costs about 5X as much as aluminum per tonne. A380 alloys usually use about 3-4% copper and Tesla was able to reduce that to .79%, which is a roughly 75% reduction in copper.
Copper also tends to contribute to brittle castings that are prone to cracking, so the reduction in copper also helps with durability. While we’re on the topic of corrosion. Tesla covered corrosion only briefly in their patent application, which is why I haven’t provided performance data on corrosion.
They simply stated that the reactivity of the aluminum alloys they tested ranged from 10% to 200% of A380 aluminum. The assumption is here that they would have picked an alloy with much lower reactivity and therefore better corrosion resistance than A380 aluminum. Silicon reduces the melt temperature of the aluminum, which helps extend the life the casting moulds. Increasing Silicon content also improves flowability because silicon helps the alloy hold heat energy as it moves through the mould, which allows the alloy to stay in its liquid state for longer.
However, as we’ll soon see, it makes for brittle castings. Iron is added to cast aluminum alloys because it strengthens the alloy and acts as a die release agent. That is, it helps prevent the aluminum alloy from sticking to the mould, which would eat away at the mould each time a casting is removed. Tesla reduced the Iron content from 1.3% down to about .3% and appears to be using Manganese and Vanadium as die release agents instead. Manganese improves strength, improves fatigue resistance, and increases corrosion resistance.
I think the combination of standard levels of manganese, high aluminum content, and low copper content are what gives Tesla’s alloy its corrosion resistance. This allowed Tesla to do away with the Zinc, which is often used to improve corrosion resistance. The Zinc content was reduced from 3% to .02%, which is 99% less Zinc than is normally used
in cast aluminum alloys. Next, let’s cover two findings noted in Tesla’s patent application to gain insights into how Tesla was able to generate better performance from their alloy than the baseline alloy. Note that I’ll only be covering the parts of the patent application that seem to have made it through to the production alloy.
In the future, if I see other elements or element ratios from the patent used in newer versions of the alloy, I’ll do a separate video to provide updates and explanations. Typically, 7-12% Silicon is used in traditional low pressure die casting to give sufficient flowability. But, what’s the ideal for the high pressure die casting that Tesla will use for their gigacastings? The inventors found that beyond 8.5% Silicon, there was little benefit in adding more Silicon. That is, using more than 9% Silicon for gigacasting doesn’t do much to improve flowability.
This is useful data because Silicon reduces ductility. Each 1% reduction in Silicon results in a 3 degree improvement in bend angle. With new knowledge of how Silicon affects casting alloys, Tesla can maximise flowability to create intricate castings, while maintaining ductility. Munro’s analysis indicated 8.5% Silicon, which is right in the sweet spot.
The second finding of the patent application that made it through to production is that Vanadium improves ductility. In the image on screen, both 9a and 9b contain Vanadium, but 9b contains more Vanadium. The arrows from label 902 point to platelets made of compounds containing aluminum, iron, silicon, and manganese.
The platelets in 9a result in a brittle alloy. I like to think of these platelets as little levers that crack apart the alloy structure when it’s put under stress. Labels 904 and 906 point to areas where vanadium has alloyed with the aluminum, iron, silicon, and manganese to form globular structures. These globular structures improve ductility because rather than acting like levers that would create internal strain and create cracks, they act more like ball bearings that reduce internal strain and reduce the likelihood of cracks. What this all means is that vanadium improves ductility while maintaining strength. The inventors didn’t qualify exactly how large this effect was.
They only advised the effect was present and that it scaled with iron impurities. That is, the more iron impurities in the crystal structure, the more vanadium would be needed. A typical A380 alloy contains no Vanadium and about 5 times as much Iron as Tesla’s alloy, which means that Tesla’s alloy will be more ductile because it contains much less iron and the iron related impurities would alloy with the Vanadium to from benign globules instead of platelets. Let’s wrap things up. In previous patent applications I explored for Tesla, I advised that there are three requirements for a patent to be granted from a patent application. Is it a new idea? Is it significantly different from other existing inventions? And, will it be useful to industry.
The lithium clay plus salt and water extraction process failed to pass muster for inventive step and so did the tabless electrode patent application. The story is different with Tesla’s new casting alloy. The patent application reviewer advised that claims 12-14 are Novel, Inventive, and have Industrial Applicability. What are claims 12-14? They’re the 3 recipes for the alloy that Tesla suggested, which is the most important aspect of the patent application. However, even if Tesla isn’t granted a patent for the alloy, the application addresses two other intellectual property concerns that are just as important: It checks if Tesla’s infringing on anyone else’s patents and whether aspects of their alloy actually can be patented, and that establishes freedom to operate even if they aren’t granted a patent. What this all means is that Tesla is using an aluminum alloy that’s never existed before.
It’s a new material, and new materials expand the design space so that new products can be produced that have never existed before. As we saw in the last gigacasting video, automakers were reluctant to switch to aluminum because advanced high strength steel had the best value proposition. This new aluminum casting alloy, combined with the gigapress, changes the equation and helps remove 3 of the 6 steps usually required for castings. Tesla’s alloy does a lot of heavy lifting to make the gigacastings possible by being strong, ductile, flowable, and corrosion resistant. It does this by greatly reducing the copper content, increasing the aluminum content, dialling in an optimal silicon content, and introducing trace metals such as Vanadium.
Although it’s true anyone can copy Tesla’s alloy simply by sending a piece to a lab, that misses the larger point: None of Tesla’s competitors appear to have the capability to create alloys and materials that never existed before to the extent Tesla does, and that means they’ll always be two steps behind. And, if Tesla does get a patent for this alloy, if they really wanted to, they could enforce that patent, but that might not be the best use of Tesla’s time. Tesla’s moat is innovation, not resting on their laurels and defending their patents.
There are cases where Tesla will defend their trade secrets and IP, but I don’t see this as one of them. In the next gigacasting video, we’ll look at the supporting cast of the gigapress and a few key points that were missed in my original gigacasting video. If you enjoyed this video, please consider supporting me on Patreon with the link at the end of the video.
I am also active on Twitter. You can find the details in the description, and I look forward to hearing from you. A special thanks to Bruno the Tunneller, Edward, Toby Xavier for your generous support of the channel, my YouTube members, and all the other patrons listed in the credits. I appreciate all your support, and thanks for tuning in.
2021-12-03 17:00
