Seems like a nice PR piece. Porsche is using 3d printing to drive tooling development to cut new shapes on pistons. Awesome for r&d but 3d printing metallurgy has a long way to go before it’s production ready.
> and feature an integrated and closed cooling duct in the piston crown that's apparently unable to be reproduced using traditional manufacturing methods
My limited understanding is that metal parts made by additive processes can end up with poorer lifespans than forged or machined parts. You get different layers which melted and cooled at different times, with weird residual stresses. Parts can fail sooner and just be weaker.
If your prototype only needs to live long enough for you to iterate on it, that might be great. But it might not be suitable for real world use for an extended period?
I’ve read that Rolls Royce is going to be using 3D metal printing on their upcoming Advance3 engine. If it’s good enough for those operating conditions then it should be good enough for anything.
Powdered metal parts are everywhere today, but it's not really an additive process - you pour a bunch of powder into a mold, compact it, and then bake it.
PM parts are also porous and fairly weak (in some dimensions) relative to machined counterparts, so you don't typically see them in critical applications.
Scroll down to the countries starting to ban fossil fuel vehicles. I think you’re fine for the next 10-15 years, after that I think it’s hazy whether fossil fuel vehicles will be permitted to be used at all if battery electrics take off. My bet is we reach a tipping point and an outright ban around 2035, if not earlier due to refinery unprofitability due to volume declines.
I do look forward to Porsche attempting a competitive EV though!
@JumpCrisscross: Throttled, can’t reply to your comment. “Progress happens slowly, and then all of a sudden.”
The minority may love the nostalgia, but sport racing cultures are primarily built on speed and ICE vehicles are already being left in the dust. Another 10 years and I would be surprised if there was a single ICE vehicle that could beat an electric.
There is some racing that lends itself to electric like stage rally, rally cross and the drag strip. I'd love to drive an electric rally cross car, especially since it's one of the most accessible forms of racing out there.
Circuit racing (in particular endurance racing) is a long way from electrics being competitive (though LMP1 cars certainly use a hybrid powertrain to boost speeds). Le Mans cars currently cover over 5000 km in 24 hours. When Porsche stretched the Taycan to the limit they managed 3300 km in 24 hours. That's roughly a 1950s Le Mans pace (and that as when the track had fewer turns and longer straights).
Torque from an electric engine is insane. Tesla prototypes literally tore the transmissions apart. This is why a Tesla has the second fastest 0-60 time of 2.28 sec, and the low-end model (Model 3) still has a quick 3.0 sec time
How? Unless the compressed air thruster is also going to take complete control over the cars handling... how is it going to overcome the limiting factor of the 4 contact patches your tires need to keep the car pointing in a given direction?
> How? Unless the compressed air thruster is also going to take complete control over the cars handling... how is it going to overcome the limiting factor of the 4 contact patches your tires need to keep the car pointing in a given direction?
I guess we'll have to wait and see how it's implemented.
My understanding is that there are really only 2 times a hypercar is traction limited: accelerating from 0 and breaking. I suspect that those are going to be the only two times the system will kick in and have meaningful impact on the car's performance.
Even normal, 4 door cars with a little bit of power end up traction limited accelerating at speed
The system can only engage when the car isn't traction limited unless it's somehow taking the tires out of the equation for handling (which I doubt they're doing)
But the thing is, it's not going to be meaningful in any way. You can already get a car that would break traction at almost every gear (there's a 1000HP Honda Odyssey minivan by Bisimoto that needs it's power output limited per gear otherwise it'd just spin in place when you put down)
If the tires don't get better, you're not really able to use power past a certain point.
Things that matter more than raw power are stuff like suspension and weight. Things EVs do not do well with and end up relying on computers to make up for.
This is going to sound like the typical Luddite screeching about how computers ruin anything, but as someone who drives cars as a hobby and has built up a track focused roadster, it really is not the same experience driving a car that needs computers working overtime pulling power and adjusting each tire's speed individually vs a less powerful car that's just well put together and easy to control for mere mortals like myself.
A million consumer miles is very different from 3000 endurance racing miles. It's not a coincidence Porsche is the most successful Le Mans team of all time and their cars can be absolutely beat on at the track without missing a step.
The FIA rules around 'refuelling' are incredibly strict and right now replacing batteries in the pit would be highly illegal. The rules can change of course, but with 400+ V and tons of mass that has to be slung around there is no way the FIA would take it lightly. When a couple extra seconds in the pit can make or break a race teams will always try to cut corners, so the FIA takes pit safety extremely seriously.
There is a reason formula-e used 2 cars rather than battery swapping when they didn't have the legs to complete a single race.
> A million consumer miles is very different from 3000 endurance racing miles. It's not a coincidence Porsche is the most successful Le Mans team of all time and their cars can be absolutely beat on at the track without missing a step.
Definitely valid points here. I didn't mean to say that Tesla's million consumer miles meant that they could automatically endure Le Mans, but more that it's clear the technology scales very well. Maybe Porsche will come out of nowhere with an even more stable and more performant electric car that has a 6000mi range and can just hold the pedal down the whole time. Who knows, but my instinct says the eventual winner will definitely be electric.
there is also a natural cliff where gas stations lose profitability. Considering many today only profit due to convenience store items, and consumer patterns have radically changed involving less random spending as well as less miles driven, this may be even sooner than we think.
I think diesel vehicles in metropolitan areas are the worst offenders.
Travelling around Europe, I think these were my biggest pet peeve, as an otherwise incredible and quaint city (such as Porto or Malaga) ends up being tainted with the disgusting smell of diesel fumes.
Petroleum only would be a huge improvement. Of course, electric only would even be better.
Think it will become a matter of having enough money. Maybe it'll cost 10X more but as with all these kinds of policies and restrictions, $ will still speak. Since most of the bans you point to are about new cars, I can already bet that the price for used combustion engine cars will skyrocket depending on model. Hmmm, makes me wonder how one could play this market
We still have 10+ years until batteries and charging infrastructure is good enough to completely replace ICE cars.
And then we have a few more decades while all the existing ICE cars age out and are actually replaced. Consider how many people today drive 10-20 year old cars. This goes doubly for high end collector cars like Porsches: Porsches from the 70s and 80s are still popular among enthusiasts, and these cars need to be repaired.
I don't think the argument is about collectibles, people still ride horses and sail sailboats across oceans. At some point EVs will be cheaper to own than any ICE so even the old cars will get scrapped.
But the question whether they really produce less pollution over their whole life time is still not settled. Production of the battery pack is quite energy intensive, lithium and neodym mining is not without problems. And you have to move around the battery pack all the time.
If electrical energy storage were the most economic one, I would expect at least one species to do that. Instead, they all store their energy in chemical storage mechanisms, i.e. fat, glycogen, ATP/ADP etc.
In order to solve the climate problems, it would be enough to synthesize gasoline from C02 in the athmosphere using renewable energy sources. A break through here, could immediately revert the whole trend towards electric vehicles. Of course this would not completely solve the pollution problem. Just CO2 pollution.
It is 100% settled. This story about EVs producing more CO2 is a myth from an MIT study that was misinterpreted by a blog paid by an oil PR firm. And it seems to still be working despite the time that it has passed and despite the fact that the MIT team came forward debunking it. A Nissan Leaf for example takes 2 years to reach the breakeven point in CO2 production.
Engineering Explained[0] on Youtube does a good job talking about improvements across both technologies.
He says he gets this question a lot when he talks about conventional engine improvements, whether or not they are going to be irrelevant soon. In one video (which I frustratingly I can't find at the moment) he lists a series of use cases that are going to rely on conventional, maybe diesel, for a long time, especially heavy hauls across remote areas.
>they weigh 10 percent less than their forged equivalents and feature an integrated and closed cooling duct in the piston crown that's apparently unable to be reproduced using traditional manufacturing methods. The decrease in weight and temperature results in an extra 30 horsepower on top of the GT2 RS's already mighty 700.
So.. 10% less weight and ~4% increase in performance. But what about the increase in price? 3D printing is magnitudes slower than forging and probably many times more expensive. I guess some Porsche owners would gladly pay for these improvements.
It's not only HP necessarily either. Lighter pistons could also improve responsiveness. You're accelerating less mass so that could mean a nicer torque curve if you increase weight savings in the moving parts of the drivetrain.
These are limited edition cars. A relative handful are made. That’s often where a manufacturer goes to more customized tooling and parts. Keep in mind Porsche owners of these types of cars will often easily spend 20-50k usd customizing the most minute details such as stitching color of the seatbelt, etc. Factoring a few thousand extra dollars in engine cost is easy.
>Factoring a few thousand extra dollars in engine cost is easy.
Totally get your point from the rest of the comment, but we could be seeing notably more expense from this. 2015-2017 I was working in a university research lab and we had an industry partner with a metal printer. They said a piece 10 inches long and 3 inches in diameter had an all-in cost of 5-8k, depending if it was steel or titanium.
I never saw a detailed breakdown of the costs they accounted for, and hope we see these manufacturing costs continue to drop.
There is a lot of challenge making the powder good for 3d printing. Tons of requirements around the size and shape - not being perfectly spherical, not being too long, etc. My understanding is this leads to much of the high material cost.
I would assume it would be for racing, not consumers. They didn't mention cost, but looking at the process, time and materials involved I'd bet you're looking at tens of thousands of dollars. I don't see consumers going for that, but GT racing clubs? Oh yeah!
I can see some overlap on the extreme consumer end [their 300K 700hp car is probably hundreds produced at most. The profits are already really high on those vehicles so I am curious if/when/how they will release and price this tech, or if it is just some random Porsche team drumming up some marketing for mostly vaporware :)
GT2 RS is already close to $200,000USD (If you can even get one.) Also ask around on Rennlist how many GT2 owners have popped the motor requiring a full replacement in the first 10,000 miles.
Price difference for relatively low run orders probably isn't that bad. I'd estimate 3d printing costs to be on the order of 250-400$/piston in quantity before post print machining for OD and wrist pin tolerance.
Well put. In general, the concerns for "massive scale" manufacturing that manufacturers like Honda and Toyota have are simply not shared by Porsche, who are known over their entire existence to manufacture < 1000 cars for a lot of their models.
Suppose you wanted to mass produce 3D printed parts as cheaply as possible. Would there be opportunities to scale up the speed of printing?
One thought is to steal an idea from traditional (2D, paper) printers. A traditional dot matrix printer has a head which sweeps side to side, but there are much faster printers called line matrix printers (https://en.wikipedia.org/wiki/Line_matrix_printer) which essentially have a head the width of the entire page, so it prints one entire row of pixels in one fell swoop. (One such printer is the IBM 6400. Video here: https://www.youtube.com/watch?v=KnPBWru2Ecg#t=6m57s)
One could imagine a similar approach with 3D printing with a wide print head that prints an entire row at once instead of a point.
Maybe there could be other optimizations or economies of scale as well.
Obviously R&D costs would be high, so you'd probably only do it if you could use for more than just pistons in one low-volume model of car.
For some types of 3d printing where the material is laid down in a bead, the path of the nozzle has an effect on the resulting part. So having a wide row of nozzles would give you less flexibility in the design. But this may have less of an effect on certain metal 3d printing technologies. For example, that could work very well for some metal sintering methods.
What I find facinating about 3d printing is how much unexplored area of research there is. Not only is there so much more to explore, a lot of it is very accessable without lots of money or technology. This is especially true on the sofware/gcode/slicing side of things.
I don't understand the point of this comment. No one is saying that 3D printing will make forging obsolete. There's always a balance of trade offs. It's up to the designer/engineer to decide how something should be made. Developing more options, and better understanding those options is part of engineering progress. It may be good to temper expectations (no pun intended) to avoid overhype, but at the same time I think it's also good to let people's imaginations run wild because that's where creative developments come from.
For example, I'm sure people will come up with ways to have more control over grain boundary and other material properties of the 3D printed material through the use of heated chambers, annealing, new alloys created for specific processes, etc.
>What I find facinating about 3d printing is how much unexplored area of research there is.
I spent about 2 years working on manufacturability analysis software during grad school. We were looking at machining, 3d printing, casting, and die casting, with the intent to give feedback to the design engineer.
There are indeed an incredible number of opportunities when it comes to additive manufacturing. And when you start dealing with integrated CNC/AM machines, you get an order of magnitude more possibilities.
The lab I worked in got one of these machines and I toured the lab last year. They were working on things including:
* Fast 3d printing of big parts. A demo part ~24x12x3 inches was printed in about 2 hours, which would have typically taken a couple days. The nozzle was much larger than most 3d printers, high precision and accuracy weren't required, but the demo part wasn't easy to manufacture with traditional subtractive methods
* Switching materials mid-print, enabling plastic and metal to be integrated in way which are otherwise prohibitively expensive or impossible.
* Integrating machining and 3d printing. Fixturing is traditionally a big challenge with machining, and every time you reset the part in a machine, you add variability to the part. 3d printing features then machining flat surfaces (for later use as datums or functionally) really cleans things up for better assembly, part function, and later fixturing.
Manufacturing capability is growing very fast in neat ways, and I'm excited to see how things will be in 5, 10, and 50 years.
You can already see some relatively high-volume components produced with additive processes, but we still need to address part strength/integrity and process reliability.
In my (admittedly limited) experience and leaving aside cost considerations, process reliability is the biggest thing holding us back from widespread adoption of AM in mass production. If you could just leave the machines running lights out with confidence that you’ll get good parts in the end, then individual process speed wouldn’t be a huge issue. We aren’t there yet, though.
I imagine this is where mass CNC machining processes were ~30 years ago.
My info is a little out of date (~10 years) but I did a MS thesis on aluminum/carbon fiber composites and holy hell are they impossible to produce. There are so many factors that just make these materials not want to go together (similar for Mg / carbon fiber). I would be really surprised to see any real MMC any time soon.
MMC parts are used in quite a few high performance vehicles today, particularly brake rotors. They're still premium parts, but it's stuff you can get from Ford or Honda on their more limited editions.
I know less about it, but as I understand it they're productionized in aerospace now too, again though, just generally for very high spec and price parts.
Pistons move very quickly, and faster and faster as engine speed (RPM) increases. Eventually, getting the piston to change direction becomes not only a parasitic loss of power on its own, but introduces a lot of design challenges around the strength and design of the crankshaft bearings, connecting rods, and piston retention pin. Furthermore, the crankshaft should balance the weight of the pistons, so losing weight from the pistons also loses weight from the crankshaft. Reducing the mass of the pistons often adds more headroom in these components and enables the engine to reach a higher maximum operating RPM (redline), and since horsepower is a function of torque x speed (RPM), increasing redline often directly increases peak output.
10% is massive in a piece of metal rotating so quickly.
Tiny pistons is how some motorcycles got incredible power out of small displacement(and ridiculous gearboxes). 20,000 RPM engines, 150mph/240kmh out of 250cc(4-stroke as well). And this was in the 60's.
It's also why racing engines (when not limited by regulations, or when only limited on total capacity) tended to have 8, 10, 12 or more cylinders. Despite the increase in the physical size of the engine, each cylinder would be smaller (for the same displacement) allowing higher revs and more power.
In 2014, Old South Carriage put a contemporary twist on our horse-drawn carriages. We began converting our fleet of wooden carriages to a carbon-fiber composite material similar to that used in Boeing’s 787 jetliner. Our reasoning? If it’s good enough for Boeing, it’s good enough for us! The result is a lighter weight, more durable carriage which requires much less maintenance than its wooden predecessor. With this innovation, we successfully increased the level of comfort for both our horses and passengers while preserving the look, feel, and authenticity of our original carriages.
The fricking media/PR around 3D printing is really annoying.
You can't match:
1) Get benefits of strength like in forging process
2) Case hardening would still need to be done
3) Shot peening and surface hardening?
4) Speed
5) Cost
6) Robustness and reliability, how many 3D printers have various ways of 3D printing, how do you validate parts for strength? It is a monumental challenge. Traditional processes are vetted and tested for decades.
7) Material wastage
8) Footprint and floorspace (1 injection molding machine can replace about a 10,000 3d printers to match the volume).
9) Power consumption
10) Human labor and attendance (may be this is already solved)
I work in manufacturing and we have a bunch of 3d printers. They make amazing tools for prototyping and testing things out. Hell, we don't even use them anymore because its just faster and cheaper to get it from Protolabs / 3D Hubs.
3D Printing has extremely useful applications where we can optimize the shape without concerning for right angles, draft angles, chamfers, and fillets. Some of the ANSYS tools can use algorithms to optimize the shape that looks like a bio-mechanical thing because we are no longer constrained by aforementioned restrictions.
But the public needs to understand that 3D printing will not replace everything. It is like saying cars replaced bicycles when they were being used in late 1890's... no, we still have bicylces 120 years later and will continue to use them. 3D printing serves a niche purpose and a huge benefit in medical/dental industries but boy the media is doing a horrible job of explaining its advantages and disadvantages. I just see misinformation all over.
In my experience it’s not the general public who is hyping 3D printing its the tech/nerd crowd. I can remember when 3D printing was getting affordable and people telling me I just didn’t understand how it would change manufacturing.
> It is like saying cars replaced bicycles when they were being used in late 1890's... no, we still have bicycles
A couple years ago, maybe on HN itself, someone bragged about how QR codes were totally obsolete because they had made a better version.
Their 'better' version relied on a web service that they controlled. So, I can't scan the code offline, and I'm at their mercy. It would have been a great joke if they weren't so serious.
To be fair, aluminum isn't typically case-hardened, at least for this application. While I lament the loss of the strength benefits from forging, they may be able to make up for a lot of them by using more material around the wrist pins than on a typical forged piston. 10% weight savings is pretty compelling, as the limiting factor in engine speed is typically con-rod stress.
3D-printed metal parts are already a thing in a number of aerospace applications--I imagine the reliability piece is either fairly well-solved for sintered metal printing, or the parts are expensive enough that we're willing to print each one more than once if required.
Bingo. Pistons have about as much work in them as a piece of jewelry. by the time your 3d printer cranks out a part, a five stage hot press and a normalizing furnace have already shot out enough raw forged parts to make your printer an expensive paperweight. depending on the metal you can still heat-treat it, but thats a big "depending."
metals for pistons and gears are an incredible trade secret. sometimes gas furnace hardening happens after shave or milling steps or even inbetween them. 3d printed parts are right up there in terms of durability with prototype parts, but miles away in terms of cost. these arent your average sintered metal molds.
I was watching a project that involved making a bunch of plastic parts. The total volume was in the low thousands. It was enough parts to start thinking about the scale but not enough to make molds worth it. It seemed like the breakeven point for molds was around 20-50k parts depending on the mold. The design that was used for the final product was made super fast custom made reprap printers with rather low accuracy. Due to the unique abilities of 3d printers it ended up being very strong and durable with a honeycomb structure inside.
3d printers are at the point were for some applications it makes sense to use them in production. The exceptions are high volume as well as very low volume that has high strength requirements( and even then for metal printers some stuff can be done post-manufacturing).
It's nice that they get some use out of 3d printing, but what's the point of R&D in ICE cars? It's a dying technology. In twenty or thirty years their sale will hopefully be banned in most interesting markets, or at least strongly disincentivized.
This is like the RISC vs CISC debate: 25 years ago some of the loudest individuals proclaimed that x86 was dead, and we’d all be on MIPS/PowerPC/SPARC/Alpha. Reality doesn’t always match up with idealism.
Engineering has a habit of advancing due to incremental improvements. The ICE still has room to grow and become more efficient.
The ICE has no room to become carbon neutral with efficiencies that are even comparable to electrics. Getting from electricity to liquid fuel to movement probably has losses exceeding 90%. Electric cars on the other hand are like 80% efficient.
You are very wrong here. Ethanol as an ICE fuel could be done carbon neutral very quickly thru yeast, and battery technology will probably never reach comparable energy densities in weight limited applications (mobile applications). We may engineer a society where the limitations of batteries are not limiting (because we/goods never travel far), but only a fool thinks you need electricity as an energy source for carbon neutral ICE.
I think ICE cars will find a new home with hobbyists who are interested primarily in driving on a track. Tons of people already throw non street-legal race cars on trailers and drive hundreds of miles every weekend. There are 30-50 cars racing at my small, local track every weekend. For a company like Porsche, with a name that car enthusiasts hold in high esteem, you have to imagine that they will continue to sell enthusiast-focused cars. Also, they made the Taycan, which many consider to be the sports car of the future (though that’s a matter of opinion).
I’m not a “business” person by any means, but I tend to doubt that every automaker in the world is making a poor decision (on a financial level) by continuing to improve their ICEs. There are incredibly smart teams of people making these decisions, and they’re doing it for a reason.
It seems to me that you could get pretty close to this design with traditional casting, forging, and then a pass with a multiple axis CNC machine. You’d get a little extra material down in some of those more complex curves, but how much really? And with a stronger part you could save material elsewhere.
97 comments
[ 2.9 ms ] story [ 172 ms ] thread> and feature an integrated and closed cooling duct in the piston crown that's apparently unable to be reproduced using traditional manufacturing methods
If they can get the process reproducible and consistency up, I don't see why it couldn't be productionized
If your prototype only needs to live long enough for you to iterate on it, that might be great. But it might not be suitable for real world use for an extended period?
PM parts are also porous and fairly weak (in some dimensions) relative to machined counterparts, so you don't typically see them in critical applications.
I suppose if will be gradual replacement by electrics but still it’s not like they would have lots of life in them.
Scroll down to the countries starting to ban fossil fuel vehicles. I think you’re fine for the next 10-15 years, after that I think it’s hazy whether fossil fuel vehicles will be permitted to be used at all if battery electrics take off. My bet is we reach a tipping point and an outright ban around 2035, if not earlier due to refinery unprofitability due to volume declines.
I do look forward to Porsche attempting a competitive EV though!
@JumpCrisscross: Throttled, can’t reply to your comment. “Progress happens slowly, and then all of a sudden.”
Zero chance of this happening in the U.S., U.K. or Germany. I imagine the Gulf, India and China will similarly retain sport driving cultures.
Mass-manufactured ICE vehicles are past their nadir. But mechanical engines will have a long life thereafter.
The minority may love the nostalgia, but sport racing cultures are primarily built on speed and ICE vehicles are already being left in the dust. Another 10 years and I would be surprised if there was a single ICE vehicle that could beat an electric.
Circuit racing (in particular endurance racing) is a long way from electrics being competitive (though LMP1 cars certainly use a hybrid powertrain to boost speeds). Le Mans cars currently cover over 5000 km in 24 hours. When Porsche stretched the Taycan to the limit they managed 3300 km in 24 hours. That's roughly a 1950s Le Mans pace (and that as when the track had fewer turns and longer straights).
Torque from an electric engine is insane. Tesla prototypes literally tore the transmissions apart. This is why a Tesla has the second fastest 0-60 time of 2.28 sec, and the low-end model (Model 3) still has a quick 3.0 sec time
Tesla is supposed to offer a "SpaceX package" with their Roadster that uses compressed air thrusters to overcome even that limit.
I guess we'll have to wait and see how it's implemented.
My understanding is that there are really only 2 times a hypercar is traction limited: accelerating from 0 and breaking. I suspect that those are going to be the only two times the system will kick in and have meaningful impact on the car's performance.
No-one wants a car that 'breaks' when they press the pedal.
The system can only engage when the car isn't traction limited unless it's somehow taking the tires out of the equation for handling (which I doubt they're doing)
But the thing is, it's not going to be meaningful in any way. You can already get a car that would break traction at almost every gear (there's a 1000HP Honda Odyssey minivan by Bisimoto that needs it's power output limited per gear otherwise it'd just spin in place when you put down)
If the tires don't get better, you're not really able to use power past a certain point.
Things that matter more than raw power are stuff like suspension and weight. Things EVs do not do well with and end up relying on computers to make up for.
This is going to sound like the typical Luddite screeching about how computers ruin anything, but as someone who drives cars as a hobby and has built up a track focused roadster, it really is not the same experience driving a car that needs computers working overtime pulling power and adjusting each tire's speed individually vs a less powerful car that's just well put together and easy to control for mere mortals like myself.
I think that really depends on the context.
For most races? You're probably right.
It'll be extremely difficult for electrics to beat gas cars at something like the 24 hours of Le Mans.
- Teslas are going to be built in such a way that every part runs for 1,000,000 miles
- “Refuelling” in the pit could be as quick as quick-releasing the battery sled, and popping in a charged one in it’s place
The FIA rules around 'refuelling' are incredibly strict and right now replacing batteries in the pit would be highly illegal. The rules can change of course, but with 400+ V and tons of mass that has to be slung around there is no way the FIA would take it lightly. When a couple extra seconds in the pit can make or break a race teams will always try to cut corners, so the FIA takes pit safety extremely seriously.
There is a reason formula-e used 2 cars rather than battery swapping when they didn't have the legs to complete a single race.
Definitely valid points here. I didn't mean to say that Tesla's million consumer miles meant that they could automatically endure Le Mans, but more that it's clear the technology scales very well. Maybe Porsche will come out of nowhere with an even more stable and more performant electric car that has a 6000mi range and can just hold the pedal down the whole time. Who knows, but my instinct says the eventual winner will definitely be electric.
Travelling around Europe, I think these were my biggest pet peeve, as an otherwise incredible and quaint city (such as Porto or Malaga) ends up being tainted with the disgusting smell of diesel fumes.
Petroleum only would be a huge improvement. Of course, electric only would even be better.
And then we have a few more decades while all the existing ICE cars age out and are actually replaced. Consider how many people today drive 10-20 year old cars. This goes doubly for high end collector cars like Porsches: Porsches from the 70s and 80s are still popular among enthusiasts, and these cars need to be repaired.
If electrical energy storage were the most economic one, I would expect at least one species to do that. Instead, they all store their energy in chemical storage mechanisms, i.e. fat, glycogen, ATP/ADP etc.
In order to solve the climate problems, it would be enough to synthesize gasoline from C02 in the athmosphere using renewable energy sources. A break through here, could immediately revert the whole trend towards electric vehicles. Of course this would not completely solve the pollution problem. Just CO2 pollution.
He says he gets this question a lot when he talks about conventional engine improvements, whether or not they are going to be irrelevant soon. In one video (which I frustratingly I can't find at the moment) he lists a series of use cases that are going to rely on conventional, maybe diesel, for a long time, especially heavy hauls across remote areas.
https://www.youtube.com/channel/UClqhvGmHcvWL9w3R48t9QXQ
EDIT: This is probably the video: https://www.youtube.com/watch?v=Hatav_Rdnno&t=27s
There's YouTube channels that get 3d car engine parts printed. This is not just the realm of multinationals.
>they weigh 10 percent less than their forged equivalents and feature an integrated and closed cooling duct in the piston crown that's apparently unable to be reproduced using traditional manufacturing methods. The decrease in weight and temperature results in an extra 30 horsepower on top of the GT2 RS's already mighty 700.
So.. 10% less weight and ~4% increase in performance. But what about the increase in price? 3D printing is magnitudes slower than forging and probably many times more expensive. I guess some Porsche owners would gladly pay for these improvements.
Totally get your point from the rest of the comment, but we could be seeing notably more expense from this. 2015-2017 I was working in a university research lab and we had an industry partner with a metal printer. They said a piece 10 inches long and 3 inches in diameter had an all-in cost of 5-8k, depending if it was steel or titanium.
I never saw a detailed breakdown of the costs they accounted for, and hope we see these manufacturing costs continue to drop.
There is a lot of challenge making the powder good for 3d printing. Tons of requirements around the size and shape - not being perfectly spherical, not being too long, etc. My understanding is this leads to much of the high material cost.
Suppose you wanted to mass produce 3D printed parts as cheaply as possible. Would there be opportunities to scale up the speed of printing?
One thought is to steal an idea from traditional (2D, paper) printers. A traditional dot matrix printer has a head which sweeps side to side, but there are much faster printers called line matrix printers (https://en.wikipedia.org/wiki/Line_matrix_printer) which essentially have a head the width of the entire page, so it prints one entire row of pixels in one fell swoop. (One such printer is the IBM 6400. Video here: https://www.youtube.com/watch?v=KnPBWru2Ecg#t=6m57s)
One could imagine a similar approach with 3D printing with a wide print head that prints an entire row at once instead of a point.
Maybe there could be other optimizations or economies of scale as well.
Obviously R&D costs would be high, so you'd probably only do it if you could use for more than just pistons in one low-volume model of car.
https://en.wikipedia.org/wiki/Selective_laser_sintering
What I find facinating about 3d printing is how much unexplored area of research there is. Not only is there so much more to explore, a lot of it is very accessable without lots of money or technology. This is especially true on the sofware/gcode/slicing side of things.
There is also shot peening which is like case hardening but with little balls of steel. Usually done on piston rods and turbine blades.
I don't think 3D printing can do these things, but I hope to be proven wrong some day.
For example, I'm sure people will come up with ways to have more control over grain boundary and other material properties of the 3D printed material through the use of heated chambers, annealing, new alloys created for specific processes, etc.
I spent about 2 years working on manufacturability analysis software during grad school. We were looking at machining, 3d printing, casting, and die casting, with the intent to give feedback to the design engineer.
There are indeed an incredible number of opportunities when it comes to additive manufacturing. And when you start dealing with integrated CNC/AM machines, you get an order of magnitude more possibilities.
The lab I worked in got one of these machines and I toured the lab last year. They were working on things including:
* Fast 3d printing of big parts. A demo part ~24x12x3 inches was printed in about 2 hours, which would have typically taken a couple days. The nozzle was much larger than most 3d printers, high precision and accuracy weren't required, but the demo part wasn't easy to manufacture with traditional subtractive methods
* Switching materials mid-print, enabling plastic and metal to be integrated in way which are otherwise prohibitively expensive or impossible.
* Integrating machining and 3d printing. Fixturing is traditionally a big challenge with machining, and every time you reset the part in a machine, you add variability to the part. 3d printing features then machining flat surfaces (for later use as datums or functionally) really cleans things up for better assembly, part function, and later fixturing.
Manufacturing capability is growing very fast in neat ways, and I'm excited to see how things will be in 5, 10, and 50 years.
In my (admittedly limited) experience and leaving aside cost considerations, process reliability is the biggest thing holding us back from widespread adoption of AM in mass production. If you could just leave the machines running lights out with confidence that you’ll get good parts in the end, then individual process speed wouldn’t be a huge issue. We aren’t there yet, though.
I imagine this is where mass CNC machining processes were ~30 years ago.
Regarding the extra 30hp, I think they probably come from the ability to slightly increase the RPMs because of the weight reduction.
https://en.m.wikipedia.org/wiki/Metal_matrix_composite
I know less about it, but as I understand it they're productionized in aerospace now too, again though, just generally for very high spec and price parts.
Tiny pistons is how some motorcycles got incredible power out of small displacement(and ridiculous gearboxes). 20,000 RPM engines, 150mph/240kmh out of 250cc(4-stroke as well). And this was in the 60's.
(loud) https://youtu.be/o57JwibqCb8?t=114 https://i.imgur.com/HMvaSYh.jpg https://i.imgur.com/BefSzRs.jpg https://youtu.be/7d3hdDgFMrQ?t=5979
*reciprocating
It's also why racing engines (when not limited by regulations, or when only limited on total capacity) tended to have 8, 10, 12 or more cylinders. Despite the increase in the physical size of the engine, each cylinder would be smaller (for the same displacement) allowing higher revs and more power.
At 6000rpm a typical piston has to be accelerated from zero to ~50km/h and back to zero 200 times per second. Every bit of weight saving helps.
https://www.oldsouthcarriage.com/about/our-carriages/
In 2014, Old South Carriage put a contemporary twist on our horse-drawn carriages. We began converting our fleet of wooden carriages to a carbon-fiber composite material similar to that used in Boeing’s 787 jetliner. Our reasoning? If it’s good enough for Boeing, it’s good enough for us! The result is a lighter weight, more durable carriage which requires much less maintenance than its wooden predecessor. With this innovation, we successfully increased the level of comfort for both our horses and passengers while preserving the look, feel, and authenticity of our original carriages.
"The Amish Horse-Drawn Buggy Is More Tech-Forward Than You Think: The tech inside this 19th-century conveyance isn't stuck in the 19th century." https://www.popularmechanics.com/cars/car-technology/a24666/...
You can't match:
1) Get benefits of strength like in forging process
2) Case hardening would still need to be done
3) Shot peening and surface hardening?
4) Speed
5) Cost
6) Robustness and reliability, how many 3D printers have various ways of 3D printing, how do you validate parts for strength? It is a monumental challenge. Traditional processes are vetted and tested for decades.
7) Material wastage
8) Footprint and floorspace (1 injection molding machine can replace about a 10,000 3d printers to match the volume).
9) Power consumption
10) Human labor and attendance (may be this is already solved)
I work in manufacturing and we have a bunch of 3d printers. They make amazing tools for prototyping and testing things out. Hell, we don't even use them anymore because its just faster and cheaper to get it from Protolabs / 3D Hubs.
3D Printing has extremely useful applications where we can optimize the shape without concerning for right angles, draft angles, chamfers, and fillets. Some of the ANSYS tools can use algorithms to optimize the shape that looks like a bio-mechanical thing because we are no longer constrained by aforementioned restrictions.
But the public needs to understand that 3D printing will not replace everything. It is like saying cars replaced bicycles when they were being used in late 1890's... no, we still have bicylces 120 years later and will continue to use them. 3D printing serves a niche purpose and a huge benefit in medical/dental industries but boy the media is doing a horrible job of explaining its advantages and disadvantages. I just see misinformation all over.
A couple years ago, maybe on HN itself, someone bragged about how QR codes were totally obsolete because they had made a better version.
Their 'better' version relied on a web service that they controlled. So, I can't scan the code offline, and I'm at their mercy. It would have been a great joke if they weren't so serious.
3D-printed metal parts are already a thing in a number of aerospace applications--I imagine the reliability piece is either fairly well-solved for sintered metal printing, or the parts are expensive enough that we're willing to print each one more than once if required.
metals for pistons and gears are an incredible trade secret. sometimes gas furnace hardening happens after shave or milling steps or even inbetween them. 3d printed parts are right up there in terms of durability with prototype parts, but miles away in terms of cost. these arent your average sintered metal molds.
3d printers are at the point were for some applications it makes sense to use them in production. The exceptions are high volume as well as very low volume that has high strength requirements( and even then for metal printers some stuff can be done post-manufacturing).
Engineering has a habit of advancing due to incremental improvements. The ICE still has room to grow and become more efficient.
I’m not a “business” person by any means, but I tend to doubt that every automaker in the world is making a poor decision (on a financial level) by continuing to improve their ICEs. There are incredibly smart teams of people making these decisions, and they’re doing it for a reason.