Welcome back everyone, I'm Jordan Geesege, and this is The Limiting Factor.
大家好,欢迎回来!我是乔丹·吉斯基,欢迎收看《限制因素》节目。
In the last video of the Magnesium series, I walked you through what Idra's Thixomolding process is and how it works compared to other casting processes, like aluminum gigacasting.
With that understanding in place, today we're going to look at why Thixomolded Magnesium is in many ways better than gigacast aluminum for use in electric vehicles.
I'll cover the cost of Thixomolded Magnesium relative to other casting methods, the hidden design benefits of Thixomolded Magnesium, whether it could be used for underbody castings, why wheels may be an even better use case, and how much magnesium we should expect in vehicles and by when.
Before we begin, a special thanks to my Patreon supporters, YouTube members and Twitter subscribers, as well as RebellionAir.com. They specialize in helping investors manage concentrated positions. RebellionAir can help with covered calls, risk management, and creating a money master plan from your financial first principles.
Let's start with the cost of Thixomolded Magnesium, because all automakers, including Tesla, are doing everything they can to reduce the cost of electric vehicles.
This image from Salzberger Aluminum Group, or SAG, compares the performance of a number of metalworking technologies.
这张来自Salzberger铝合金集团(SAG)的图片比较了几种金属加工技术的性能。
Let's pair back the bubble plot to the technologies that are relevant to this video. Vacuum Assisted Diecasting, Semi-Solid Diecasting, and Forging.
让我们将气泡图配对到与此视频相关的技术上。真空辅助压铸、半固态压铸和锻造。
SAG didn't provide any assumptions or notes on the slide, but I'll add some for context.
SAG在幻灯片中没有提供任何假设或注释,但我会添加一些以便理解背景。
The first assumption is that vacuum assisted die casting is just high pressure die casting with vacuum assist, which is what Tesla uses for their aluminum gigacastings. So I'll replace the words vacuum assisted die casting with aluminum gigacasting to make the image more clear.
The second assumption is that semi-solid casting refers to rheocasting because this image came from a video on rheocasting. Rheocasting was covered in the last video, and the short story is that, like Thixomolding, casting uses a semi-molten slurry to produce high quality castings at a similar cost to Thixomolding. So I'll replace the words semi-solid casting with Thixomolding because the cost and quality are the same, but Thixomolding is what this video is about.
The third assumption is that all three of these metalworking technologies are using aluminum, because the SAG video I pulled this image from was strictly focused on aluminum.
第三个假设是,这三种金属加工技术都是使用铝材,因为我从SAG视频中提取出这张照片的视频专注于铝材。
With those assumptions in place, the question is, where do the cost and quality of Thixomolded magnesium sit relative to the quality and cost of Thixomolded aluminum?
As for quality, we can keep the quality of Thixomolding in the same position, because whether magnesium or aluminum is used in the Thixomolding process, the part quality will remain the same. Why? Because by quality, I mean porosity. Or porosity means higher quality, which in turn means stronger and more ductile parts, because the parts contain fewer voids and more solid metal. That is, quality isn't a measure of strength and ductility, but the ability of a process to maximize those properties regardless of which metal is used.
As for cost, the image shows that Thixomolded aluminum is slightly more expensive than gigacast aluminum and significantly less expensive than forging. But what happens when we swap out the Thixomolded aluminum for Thixomolded magnesium?
First, magnesium has a lower melting temperature than aluminum, a lower heat of fusion and its lower density, so it requires less energy to both heat the metal for casting and to cool the part to solidify it. That means lower production costs and potentially faster cycle times.
Second, magnesium is less reactive with the steel dye than aluminum. But lower reactivity means that magnesium eats up less of the steel dye with each casting that aluminum does. For example, at the same temperature, the dyes for magnesium last about 80% longer. Those dyes cost hundreds of thousands of dollars each, so the fact that the dyes last longer with magnesium adds to the cost advantage of magnesium Thixomolding over aluminum Thixomolding.
Third, in North American and European markets, magnesium costs significantly more per ton than aluminum. However, the strength-to-weight ratio of magnesium is significantly higher than aluminum. That means when factoring in the strength-to-weight ratio against cost per ton, magnesium is currently only slightly more expensive. As a side note, if you'd like to know why advanced high-strength steel isn't used for everything despite its low cost, watch my gigacasting economics video.
Let's be conservative and say that the current additional cost of magnesium feedstock balances out the savings from its lower energy costs for heating and cooling and lower dye replacement costs. If that's the case, the cost of Thixomolded Magnesium should be roughly on par with the cost of Thixomolded aluminum. So for the moment, I'll keep the bubble for Thixomolding in the same position and rename it Magnesium Thixomolding.
The fourth point on cost is that we have to take into account effective cost. What do I mean by effective cost? For each kilogram of weight you remove from an EV, you can reduce the size of the other components required to move that vehicle around, such as the battery, brakes, motors, suspension, and structural reinforcement. In the first video of the series, I did some very rough back-of-the-napkin math and concluded that for every 2.2 kilograms of weight removed from the vehicle, one 4680 sized battery cell and the packaging around that battery cell could be removed from the vehicle. I also assumed that currently the battery cell and packaging material would cost about $10. Given that I don't expect Tesla to start using Thixomolding for a couple of years, and battery prices will come down over that time period, let's assume $8 saved for each 2.2 kilograms of weight reduction.
The front and rear Giga castings in the average Tesla weigh about 80 and 65 kilograms, respectively, or 145 kilograms in total. If magnesium castings are 33% lighter, that means a savings of 48 kilograms. That would save 22 battery cells or $176 in just the battery pack. There would also be cost savings in the motor, inverter, brakes, suspension, and vehicle structure, which could easily push the savings to $200 to $300 per vehicle. The bubble plot just shows the part level cost, so the implication is that if we take into account the cost savings of lighter weight parts and EVs, the effective cost of the parts at the vehicle level could be at cost parity or even lower than Giga cast aluminum parts.
For the fifth point on cost, earlier I used the words current additional cost of magnesium feedstock. That is, currently magnesium is expensive and costs more than aluminum, but that'll likely change as the decade progresses. As I pointed out in the first video of the series, part of the reason why magnesium is expensive is because it's produced at lower volumes than its nearest competitor, aluminum, and we're still using processes to produce magnesium that are about 80 years old. But with the technology improvements and scaling that are expected for magnesium this decade, prices should drop. This image from S&P Global shows that magnesium should reach cost parity with aluminum by 2030.
If that happens, when factoring in the strength-to-weight ratio against cost per ton, magnesium parts will cost about a third less than aluminum parts. That means even if we don't factor in the effective cost of magnesium parts at the vehicle level, even at the part level they should be cheaper than Giga cast aluminum parts by 2030. Furthermore, cost modeling done by Magrathia indicates that they expect to produce magnesium that not only reaches cost parity with aluminum, but that's actually cheaper than aluminum by 2030. I'll talk more about Magrathia in the next video and how they intend to hit that cost target.
What all this means is that I expect even with conservative assumptions, thick-so-molded magnesium castings should be at cost parity at the part level with thick-so-molded aluminum, slightly cheaper than aluminum Giga casting when factoring in vehicle level savings and potentially significantly cheaper than aluminum Giga casting by 2030. If we factor in improvements in extraction technology and increased magnesium supply, with cost out of the way, let's move on to the design benefits of thick-so-molded magnesium.
As I said earlier, thick-so-molding uses semi-moulton metal instead of fully molten metal. That fundamentally changes the way the metal behaves in the dye through three dynamics. First, fully molten metal expands and contracts more than semi-moulton metal during the casting process. That means the castings shrink as they cool, which makes them stick to the dye and means the parts can warp as they shrink and need to be scrapped.
Second, fully molten metal is more reactive and chemically attacks the dye more aggressively than semi-moulton metal, which means the castings done with molten metal stick to the dye more strongly when they're removed. This is in addition to the fact that aluminum is more fundamentally reactive with the dye, which I mentioned earlier.
Third, as I showed in the last video, fully molten metal experiences turbulent flow when it's injected into the dye. That turbulence traps air, creates porosity, and reduces the quality of the castings. That's as opposed to semi-moulton metal, which creates a non-turbulent laminar flow when it's injected into the dye that results in less porosity.
Those three dynamics mean that fully molten metal is more difficult to cast than semi-moulton metal. And that means that parts cast with an aluminum gigacasting process have more design restrictions than parts made with a magnesium-thixome-molding process.
With regards to the first two dynamics, the fact that the parts are more likely to shrink and stick to the dye and react with and stick to the dye mean that every part that's designed for an aluminum gigacasting process has to be shaped in a way that allows it to pop out of the dye more easily. That's done with a slope on the walls of the parts called the draft angle. That means if there's an ideal that an engineer wants to design a part to, the draft angle forces the design of the part away from that ideal, which might mean a part that's not as strong or weighs more. For gigacast aluminum, the minimum draft angle is 2 to 3 degrees, whereas for thixome-molded magnesium, it's 0 to 1.5 degrees.
Next, a combination of the first and third dynamics, shrinkage and turbulent flow that leads to porosity, mean that for gigacast aluminum, a portion of the dye is set aside to channel the turbulent flow in a way that reduces defects, and to provide a reserve of molten metal to draw from to help compensate for shrinkage. This is called gating. What this means is that for each unit of metal that's used in a thixome molding process, more of it goes to the final part rather than being wasted as gating material that has to be stripped off, scrapped, and recycled. So if both an aluminum gigacasting machine and a thixome-molded magnesium machine have a maximum shot weight of 20 kilograms, the actual final part weight of the thixome molding machine will be closer to 20 kilograms than for the gigacasting machine.
Finally, as for the third dynamic, lower turbulence and porosity, by itself it has two knock-on effects. First, it means the minimum wall thickness for magnesium thixome molding is 0.8 millimeters versus 2 millimeters for gigacast aluminum. If that design benefit carries through to larger gigascale thixome-molded parts, it may allow engineers to thin out parts of the casting that weren't able to be made thinner with a gigacasting process, meaning lighter parts. The second benefit of lower porosity is higher part consistency. The location and concentration of where porosity forms in cast parts is random. That means some parts are stronger or more ductile than others. So when a part is designed, it has to be designed so that the weaker or less ductile parts are still up to spec, which to a certain extent means the part has to be overbuilt. With higher part consistency, the parts don't need to be as overbuilt, which means lighter parts.
The net effect of a smaller draft angle, thinner minimum wall thickness, more efficient use of material, and greater part consistency means that with thixome molding, engineers can make a bigger part that's more effective for a specific use case for the same kilogram of metal. And that's independent of any of the properties of the metal itself. It's strictly a design benefit inherent to thixome molding casting technology.
Now that we understand the cost and design aspects of thixome-molded magnesium, where could it be used? Onscreen is a list of parts that auto manufacturers are already using magnesium for or are exploring using magnesium for. Let's focus on what I consider to be the most interesting use cases.
Underbody castings and wheels. Chinese researchers are already experimenting with what appears to be underbody castings. I say appears to be because the part is blurred out. However, it is confirmed to be a car body in the article, and the shape and size for me indicates that it's an underbody casting that would weigh at least 40 to 50 kilograms. That's because a similar-sized aluminum, rear underbody casting in the model Y weighs about 65 kilograms, and a magnesium casting would weigh about a third less. The question is, what type of casting technology did the Chinese researchers use? The 20-kilogram shot weight that Edra is teasing is the largest that I know of for semi-solid casting like thixome molding. So my guess is that they used standard high-pressure die casting, or vacuum die casting rather than thixome molding. Regardless, overall, I think they have the right idea about where casting technology is headed for vehicles.
Larger magnesium castings. Edra's first gigascale thixome molding machine is expected to have a shot weight of 20 kilograms, and although that's not quite large enough for an underbody casting, it's likely just the beginning. As I said in the last video, just as the size of gigacasting machines nearly tripled in the last few years, the same is likely to happen with thixome molding machines. And in my view, the shot weight for magnesium thixome molding may not even need to triple from 20 kilograms to make castings that are the size of a 60 to 80 kilogram aluminum casting. And that's for three reasons. First, magnesium is less dense than aluminum, so it can produce parts that are 50% larger for the same shot weight. Second, as I said earlier, thixome molding makes more efficient use of the casting metal because it requires less gating. So a greater portion of the metal is used for the actual part than to accommodate for turbulence and shrinkage. Third, Tesla recently revealed that they're designing their newer castings in a way that encourages the flow of molten metal through the die. For the casting on screen, they used a 6,500 ton gigapress, which just a couple of years ago would have required an 8,000 ton gigapress. With all that mine, I wouldn't be surprised if by 2026 or 2027, Idra has a second or third generation gigascale thixome molding machine that can produce front and rear underbodies that are equivalent in size to the front and rear gigacastings used in the Model Y. In the meantime, it's a target rich environment for Idra's thixome molding machines that have a 20 kilogram shot weight.
On that note, why does Idra refer to the 20 kilogram shot weight for magnesium thixome molding as gigascale? If current, underbodies gigascale castings weigh at least 65 kilograms. I can't speak for Idra, but for me, it's because the 20 kilogram casting is produced on the same Idra gigapress machines as the 65 kilogram castings. The only difference is that the 20 kilogram casting requires Idra's thixome molding piston injection unit that's bolted onto the gigapress, which I covered in the last video. Lastly, before we move on to magnesium wheels, an interesting side benefit of a magnesium underbodies casting is that magnesium is great at turning vibrational energy into heat that's quickly dissipated, also known as damping. If you're curious about the science behind damping, the information on screen does a good job of explaining it. We don't need to get into that much depth today. The key takeaway is that magnesium has several times more damping capacity than aluminum. That means if it's used for structural and suspension parts, it can create a vehicle with better ride quality while maintaining a firm body structure.
Let's move on to wheels. The use of cast magnesium wheels was fairly widespread in the 1950s and 1960s because they're about 25% lighter than aluminum wheels and about 40% lighter than steel wheels. However, among other things, they were prone to corrosion, so for the most part, they were abandoned by the auto industry. The one exception was high performance race cars and other performance vehicles because corrosion resistance wasn't a key requirement. Then in the 1990s, racers started using forged magnesium wheels because they're even lighter than cast magnesium wheels. That's because, as I showed earlier, forged parts offer the highest part quality and therefore the highest strength and ductility. But the drawback of forging is that it's the most expensive metalworking technology.
You probably see where I'm going with this. Thixo molded magnesium produces parts that are nearly as high quality as forged parts but at a much lower cost. So I'm expecting to see the resurrection of cast magnesium wheels except this time using Thixo molding.
What about the corrosion concerns that caused magnesium wheels to be abandoned? A lot has changed in 50 years. As I showed in the first video of the series, with newer alloys, corrosion is now a solved problem. And if more corrosion resistance is needed, the wheels can be coated, which is often done with aluminum parts as well.
As a result, magnesium wheels on the market today can have up to a 10 year warranty and can even outlast aluminum wheels. Some people might also raise flammability as a concern, but it's the same deal. Even a dash of calcium can dramatically reduce flammability. If you'd like to know more about that, watch the first video of the series.
All that aside and getting to the main point, why are lighter wheels such a big deal? They have a big impact on vehicle efficiency because they have lower rotational inertia. In simple terms, what that means is that the lighter the wheel, the easier it is for the motors to spin that wheel, meaning greater vehicle efficiency and range and faster acceleration.
For example, the aluminum 20 inch uber turbine wheels used on performance Tesla Model 3s and Y's weigh 13 kilograms each. Cast magnesium wheels of the same size weigh about 10.2 kilograms and forged magnesium wheels 7.4 kilograms. Fix-o-molded magnesium wheels would probably weigh about 8 to 8.5 kilograms.
That means a weight savings of about 5 kilograms per wheel, or 11 pounds, by using fix-o-molded magnesium wheels instead of cast aluminum wheels. For a total weight savings of 20 kilograms or 44 pounds for all four wheels.
That's roughly in line with the test that Tesla owners Silicon Valley did, where they reduced the total tire and wheel weight by 52 pounds or 23.6 kilograms. The result was that they saw a 4-6 percent range improvement. So with a 20 kilogram weight savings, the result would probably be closer to 3-5 percent.
That 3-5 percent range improvement would likely be more than the range improvement provided by replacing the aluminum underbody castings with magnesium, which would improve vehicle range by around 1-2 percent. That is, the wheels would provide double or triple the range improvement while using roughly 1-3 the magnesium.
Furthermore, with fix-o-molded magnesium wheels, the vehicle would not only see a 3-5 percent range improvement. The wheels would be cheaper than cast aluminum wheels. They would provide better ride quality through improved damping, and the car would also be more nimble and accelerate faster because the lighter weight wheels are more responsive. Much the same way that wearing tennis shoes makes you more agile than wearing heavy boots.
Before we move on, it's worth making two more points on magnesium wheels. First, the technology to cast even one fix-o-molded magnesium wheel, which would weigh at least 8 kilograms, only arrived last year when 10 kilogram shot weight fix-o-molding machines hit the market. Now that the technology is there to cast wheels that large and even larger, in the next 3-5 years, fix-o-molded magnesium wheels could become a trend.
Second, bear in mind that the 3-5 percent improvement is for a large wheel used on a performance vehicle. For smaller efficiency vehicles, they tend to use smaller rims and narrower tires to improve range and reduce production costs. So it might not make sense to use magnesium wheels on, for example, a robo-taxi. But I can't see a good reason why higher end vehicles like the Model 3, Y, S, X, or even the Cybertruck wouldn't use fix-o-molded magnesium wheels.
Now that we've covered the benefits of fix-o-molded magnesium and its uses, how much magnesium can we actually expect to be used in vehicles? McKinsey expects that at least 5% of the weight of a new vehicle produced in 2030 will be magnesium, which on average is about 100 kilograms of magnesium per vehicle, or enough to replace the underbody castings.
On the high end, they estimate it would be up to 16% of the vehicle, which in my estimation would mean replacing all of the parts we see on screen with magnesium. Maybe more. Either way, analysts are expecting large increases in the amount of magnesium used in vehicles, because currently the average vehicle is less than 1% magnesium. That 1% tends to be used in smaller parts like the screen mount for Tesla's vehicles. But that'll change with the ability to do larger ultra-high quality fix-o-molded magnesium castings.
The question is, why are analysts so bullish on magnesium and what factors are they looking at?
问题是,为什么分析师对镁如此看好,他们看重哪些因素?
This spider chart for 2020 from S&P Global's automotive materials team in Detroit shows the performance of magnesium and aluminum across a range of properties. Magnesium is in black and aluminum is in blue. For most of the properties, aluminum outperforms magnesium. So at a glance, from a historical perspective, the typical criticisms of magnesium's cost and performance are valid. Of course, this information is now 4 years old. What does the future look like?
By 2030, thanks to factors like fix-o-molding and expanded magnesium production, the performance of magnesium in the spider chart improves dramatically, and magnesium becomes much more competitive with aluminum. But this is S&P's base case. What about a more bullish estimate? In the image on screen, Magrithia took S&P's 2030 estimate and overlaid it with their own projections with a pink dotted line. Let's take a closer look at the key properties, what they mean, and how they compare between magnesium and aluminum, and what the implications are.
First, as I said earlier, Magrithia's goal is to produce magnesium cheaper than aluminum, whereas S&P Global expects magnesium to reach cost parity with aluminum. Either way, both forecasts are bullish for magnesium, and cost is the single most important property on the spider chart. Why? Despite the drawbacks of magnesium due to its light weight, at the right price it's considered a useful replacement for aluminum. Depending on what the magnesium to aluminum price ratio is, sometimes overall, it's actually preferable to use magnesium.
So if the price of magnesium can be reduced on a sustained basis, and its performance increases, it will become the dominant light weight structural metal. The next most important property is availability. If there's not a relatively broad supply chain for magnesium with a number of suppliers, manufacturers won't build their products around it. Even in S&P's base case scenario, availability improves dramatically between 2020 and 2030. Magrithia, on the other hand, has aggressive scaling plans that I'll cover in the next video.
Next, energy absorption is likely referring to buckling. Buckling refers to how the geometry of a structure responds to compression and bunches up like an accordion. For example, if you look closely at Tesla's gigacastings, you can see wavy accordion-like geometry designed to do exactly that. Buckle. Stipness is a function of how much stress a part can take before it deforms, or for those familiar with stress strain curves, it's the slope of the line.
As for toughness, they're likely referring to fracture toughness, which is how resistant a material is to crack propagation. In 2030, energy absorption, stiffness, and toughness are close enough to aluminum that magnesium will likely be able to serve as an alternative in most use cases. And notably, magnesium in 2030 outperforms the numbers for aluminum from 2020 in both toughness and stiffness, 3 versus 2 and 4 versus 3, respectively, and it barely falls short in terms of energy absorption.
At 9 versus 10, that is, S&P assumes that by 2030, magnesium will have considerably better stiffness and toughness compared to the aluminum of 2020, while achieving 90% of its energy absorption. That should put to bed the frequent arguments I've heard against magnesium suitability for automotive use cases because it can't compete with aluminum in terms of its structural properties. That may have been true in the past, but new alloys and better processes are being developed each year.
Furthermore, it's the whole spider chart that matters not just one or two properties. The fact that magnesium will offer lighter weight parts at equal to or lower cost than aluminum means that it'll offer better value for money for EVs. Where cost and weight considerations dominate. The other material properties could be accommodated for with fractionally thicker parts, while still maintaining the cost and weight advantages.
As a side note, I've also heard criticisms of the fatigue and creep performance of magnesium. With thixomolding, magnesium performs as well as, or better than, gigacast aluminum for both properties. However, that's deeper than we need to go today, and I may do another video in the future that dives deeper into properties like creep, fatigue, stiffness, and energy absorption.
In summary, in my view, a number of factors are converging that may mean that by the end of the decade, thixomolded magnesium will replace gigacast aluminum as the best way to reduce the cost and weight of vehicle components.
First, EVs can afford to pay a premium for lightweight materials because a lighter weight vehicle needs, for example, smaller motors, batteries, and brakes, which are high cost components. That'll drive increased demand for magnesium because it's the lightest structural metal, allowing for a greater reduction in high cost components than aluminum.
Second, in the past few decades, alloys have been developed that reduce the flammability of magnesium, increase its corrosion resistance, and improve its overall performance, which further increases its utility and therefore demand.
Third, the factors that we covered in depth today, magnesium thixomolding will be able to produce parts that are higher quality, use the space and the die more efficiently, extend the die life, and allow for part designs that are closer to the optimum part design rather than a design that's been compromised by limits on draft angles and wall thicknesses. This will also increase the demand for magnesium.
Fourth, thanks to the increased demand due to EVs, better alloys and thixomolding, the supply of magnesium will increase, which will lower prices and increase availability. That in turn will mean thixomolded parts will become increasingly compelling versus gigacast aluminum parts.
So when will we see large thixomolded parts in, for example, a Tesla? First, Idra has to get their gigascale thixomolding machines on the market, which will happen in 2025 based on their presentation. That means a year or two after that before we see those machines in an actual factory at high production rates, which means 2026 or 2027.
Second, there needs to be an increase in the availability and a decrease in the cost of magnesium, which in my view are the primary limiting factors for the shift to magnesium. I expect minor increases in production and availability by 2026 and a much larger increases by the end of the decade, which should correspond with cost decreases. That is, I don't expect thixomolded magnesium parts to really hit the mainstream for another two years.
And after that, the extent to which magnesium thixomolding becomes dominant will depend on magnesium supply. The shift may start with smaller parts and then expand to larger parts as greater magnesium supply becomes available. But when magnesium supply does begin to scale in earnest, the shift to thixomolded magnesium will be as rapid and nearly as significant as Tesla's shift from stamped underbodies to gigacast aluminum underbodies.
We'll talk more about scaling in the next video when I cover magnesium from seawater.
在下一个视频中,当我介绍海水中的镁时,我们会更详细地讨论扩展问题。
As a final note, this video series is focusing mainly on the automotive use case for magnesium, but the broader use cases are endless in a world where everything is getting lighter. For example, magnesium might make a great material to make a robot skeleton out of to increase efficiency and battery life.
Needless to say, I'm excited to see what the future holds for magnesium and interested to read your comments about your thoughts and predictions below.
毫无疑问,我对镁的未来充满期待,也很想读到您对此的想法和预测的评论。
If you enjoyed this video, please consider supporting the channel by using the links in the description. Also, consider following me on X. I often use X as a test bed for sharing ideas and X subscribers like my Patreon supporters generally get access to my videos a week early.
On that note, a special thanks to Siddharth Atreya for your generous support of the channel, my YouTube members, X subscribers, and all the other patrons listed in the credits. I appreciate all of your support and thanks for tuning in.