4 ms·

Combustion engine development was oddly stunted though. We have known the basic formula for power for a very long time (forced induction + high compression + hi

by fhood 7y ago

Combustion engine development was oddly stunted though. We have known the basic formula for power for a very long time (forced induction + high compression + high octane fuel). Superchargers are ancient, turbos aren't new either, although I guess fuel injection had to wait for computers to be reliable and cost effective. Even then I don't fully understand why it took 30 years for automakers to figure out how to produce high power engines that met emissions standards.

abakker 7y ago
It didn't take long to know how to do it. It took a long time to make cost effective parts that were reliable in doing it for the service life of a car. the bearings in a modern vw turbocharger are incredible. Imagine getting 100K Miles out of a turbocharger which can regularly spool to 200K RPMs. those bearings need to be made to incredible tolerances, which were not previously cost effective.
- Merrill 7y ago
  People generally overestimate the difficulty of designing a product and underestimate the difficulty of designing the equipment, processes, and systems needed to mass produce the product economically.
  - baq 7y ago
    musk tweeeted a few years ago that "it isn't about the car, it's about the factory". he then proceeded to fail and redesign production due to overautomation, but the point stands and perhaps was stated in bold at that moment.
  - Gibbon1 7y ago
    I think of it as a rule of five. Takes five times as much effort to design the production process as it does to design the product. And five times that to put it into production. The number five is fungible depending on volume and similarity to existing products. It's why a hobbyist in his garage can design a product but can't commercialize it.
  - nostrademons 7y ago
    Is that true? There are a number of interesting corollaries if so, notably that China should get a lot more respect for extremely cheap mass-production of products, the U.S's position as "designer of products that are built elsewhere" is quite precarious, that software and other IP (where mass production is basically free) are basically the perfect businesses, and that services that can be provided without mass-producing a product will grow in importance. I could buy all 5 of those statements, but it's interesting that the combined effect of them would be to push people out of making things and into manipulating brains (which itself is a statement I could believe).
    - snagglegaggle 7y ago
      It is true, and some Chinese production lines are truly feats of engineering. However, it's a little tricky to determine whether this is really Chinese values and engineering or the values and engineering they have imported from other countries. There's a lot of production lines in China that produce nonfunctional goods. In the microelectronic and mechanical areas the strength of the Chinese market seems to be very, very lax IP law. While most of the interesting parts don't make it over to the western market, they have very featureful, accessible, and novel components that are made by cobbling together (potentially stolen or at least unlicensed) IP. Historically US manufacturing used to be the best cost/performance optimized and this is what helped win WWII.
    - Merrill 7y ago
      It's my impression that Germany and Japan are leaders in producing the capital equipment needed to set up manufacturing, although the US is fairly capable in some areas. I think that China imports a lot of equipment from them and then integrates it into factory systems. You 5 statements are probably right with some qualification. For example, software mass production is basically free (actually just low cost) for apps distributed via the app stores. However, if you are selling software as a service at scale, setting up a cloud service with five 9s availability is complex and expensive.
- fhood 7y ago
  Sure, but the Corvair had a turbocharged option in the 60s.
- taneq 7y ago
  Turbos often use fluid bearings, and if treated well can last indefinitely. I have an old Jap sports car in the shed with over 400k km on the original turbo.
naravara 7y ago
>Even then I don't fully understand why it took 30 years for automakers to figure out how to produce high power engines that met emissions standards. They could probably have done it before, but this stuff is expensive. The current make/model of the first car I bought in the 90s costs double what it did then if you buy it new. I don't think the demand for efficiency really came in until oil prices started shooting up and the government imposed stricter CAFE standards.
post_break 7y ago
More rigorous airbags and crash safety has stunted MPGs too.
jacquesm 7y ago
Fuel injection has been done forever without electronics even (purely mechanical), and with electronics but without computers decades before the first microprocessor came around. The increase in performance goes hand-in-hand with emissions standards because - to some approximation - a more fuel efficient engine is more powerful for the same displacement than one that isn't. CFD is what really drove the last 25 years of improvements in this field, better models of what actually happens in a combustion chamber have helped tremendously in extracting the last bit of performance from the engines. The F1 engines are amazing in this respect, their longevity is terrible but their performance is nothing short of incredible in terms of thermal efficiency (> 50%; long thought to be impossible). https://autoweek.com/article/formula-one/mercedes-f1-engine-true-modern-marvel https://autoweek.com/article/formula-one/mercedes-f1-engine-...
- fhood 7y ago
  > more fuel efficient engine is more powerful for the same displacement than one that isn't Not necessarily true, but for kinda stupid reasons. Carburetors often run cooler at the intake than fuel injection, resulting in more peak power despite worse economy. I am aware that this is not actually relevant here, I just think it's interesting. My main point is that in that in the early 70s Chevy's small blocks could make damn near 400 horsepower. Fast forwards and it isn't until that late 80s that you see a consumer car with north of 250 (grand national), and that was a pretty extreme outlier. Move to the early 2000s and we finally get modern Chevy LS motors which are once again making north of 400hp from the factory. These are still single cam, 2 valve per cylinder V8s though. WTF actually happened?
  - perl4ever 7y ago
    Measurement of power changed, I think. In 1971, Car & Driver claims a Corvette with 425 hp did 0-60 in 5.3 seconds[1]. 1/4 mile was maybe 13.8 s. A 2008 Corvette with supposedly similar power did 0-60 in 4.1 seconds and 1/4 mile in 12.5.[2] Yeah, there's some difference in weight, and modern transmissions, tires, and suspension are better and so on, but there is usually such a big difference that I think most of it is that horses used to be smaller. [1]https://www.caranddriver.com/features/g15379023/the-chevrolet-corvette-a-history-in-zero-to-60-times/?slide=14 https://www.caranddriver.com/features/g15379023/the-chevrole... [2]https://www.motortrend.com/cars/chevrolet/corvette/2008/2008-chevrolet-corvette/ https://www.motortrend.com/cars/chevrolet/corvette/2008/2008...

by fhood 7y ago

abakker 7y ago

It didn't take long to know how to do it. It took a long time to make cost effective parts that were reliable in doing it for the service life of a car. the bearings in a modern vw turbocharger are incredible. Imagine getting 100K Miles out of a turbocharger which can regularly spool to 200K RPMs. those bearings need to be made to incredible tolerances, which were not previously cost effective.

Merrill 7y ago
People generally overestimate the difficulty of designing a product and underestimate the difficulty of designing the equipment, processes, and systems needed to mass produce the product economically.
- baq 7y ago
  musk tweeeted a few years ago that "it isn't about the car, it's about the factory". he then proceeded to fail and redesign production due to overautomation, but the point stands and perhaps was stated in bold at that moment.
- Gibbon1 7y ago
  I think of it as a rule of five. Takes five times as much effort to design the production process as it does to design the product. And five times that to put it into production. The number five is fungible depending on volume and similarity to existing products. It's why a hobbyist in his garage can design a product but can't commercialize it.
- nostrademons 7y ago
  Is that true? There are a number of interesting corollaries if so, notably that China should get a lot more respect for extremely cheap mass-production of products, the U.S's position as "designer of products that are built elsewhere" is quite precarious, that software and other IP (where mass production is basically free) are basically the perfect businesses, and that services that can be provided without mass-producing a product will grow in importance. I could buy all 5 of those statements, but it's interesting that the combined effect of them would be to push people out of making things and into manipulating brains (which itself is a statement I could believe).
  - snagglegaggle 7y ago
    It is true, and some Chinese production lines are truly feats of engineering. However, it's a little tricky to determine whether this is really Chinese values and engineering or the values and engineering they have imported from other countries. There's a lot of production lines in China that produce nonfunctional goods. In the microelectronic and mechanical areas the strength of the Chinese market seems to be very, very lax IP law. While most of the interesting parts don't make it over to the western market, they have very featureful, accessible, and novel components that are made by cobbling together (potentially stolen or at least unlicensed) IP. Historically US manufacturing used to be the best cost/performance optimized and this is what helped win WWII.
  - Merrill 7y ago
    It's my impression that Germany and Japan are leaders in producing the capital equipment needed to set up manufacturing, although the US is fairly capable in some areas. I think that China imports a lot of equipment from them and then integrates it into factory systems. You 5 statements are probably right with some qualification. For example, software mass production is basically free (actually just low cost) for apps distributed via the app stores. However, if you are selling software as a service at scale, setting up a cloud service with five 9s availability is complex and expensive.
fhood 7y ago
Sure, but the Corvair had a turbocharged option in the 60s.
taneq 7y ago
Turbos often use fluid bearings, and if treated well can last indefinitely. I have an old Jap sports car in the shed with over 400k km on the original turbo.

naravara 7y ago

>Even then I don't fully understand why it took 30 years for automakers to figure out how to produce high power engines that met emissions standards. They could probably have done it before, but this stuff is expensive. The current make/model of the first car I bought in the 90s costs double what it did then if you buy it new. I don't think the demand for efficiency really came in until oil prices started shooting up and the government imposed stricter CAFE standards.

post_break 7y ago

More rigorous airbags and crash safety has stunted MPGs too.

jacquesm 7y ago

Fuel injection has been done forever without electronics even (purely mechanical), and with electronics but without computers decades before the first microprocessor came around. The increase in performance goes hand-in-hand with emissions standards because - to some approximation - a more fuel efficient engine is more powerful for the same displacement than one that isn't. CFD is what really drove the last 25 years of improvements in this field, better models of what actually happens in a combustion chamber have helped tremendously in extracting the last bit of performance from the engines. The F1 engines are amazing in this respect, their longevity is terrible but their performance is nothing short of incredible in terms of thermal efficiency (> 50%; long thought to be impossible). https://autoweek.com/article/formula-one/mercedes-f1-engine-true-modern-marvel https://autoweek.com/article/formula-one/mercedes-f1-engine-...

fhood 7y ago
> more fuel efficient engine is more powerful for the same displacement than one that isn't Not necessarily true, but for kinda stupid reasons. Carburetors often run cooler at the intake than fuel injection, resulting in more peak power despite worse economy. I am aware that this is not actually relevant here, I just think it's interesting. My main point is that in that in the early 70s Chevy's small blocks could make damn near 400 horsepower. Fast forwards and it isn't until that late 80s that you see a consumer car with north of 250 (grand national), and that was a pretty extreme outlier. Move to the early 2000s and we finally get modern Chevy LS motors which are once again making north of 400hp from the factory. These are still single cam, 2 valve per cylinder V8s though. WTF actually happened?
- perl4ever 7y ago
  Measurement of power changed, I think. In 1971, Car & Driver claims a Corvette with 425 hp did 0-60 in 5.3 seconds[1]. 1/4 mile was maybe 13.8 s. A 2008 Corvette with supposedly similar power did 0-60 in 4.1 seconds and 1/4 mile in 12.5.[2] Yeah, there's some difference in weight, and modern transmissions, tires, and suspension are better and so on, but there is usually such a big difference that I think most of it is that horses used to be smaller. [1]https://www.caranddriver.com/features/g15379023/the-chevrolet-corvette-a-history-in-zero-to-60-times/?slide=14 https://www.caranddriver.com/features/g15379023/the-chevrole... [2]https://www.motortrend.com/cars/chevrolet/corvette/2008/2008-chevrolet-corvette/ https://www.motortrend.com/cars/chevrolet/corvette/2008/2008...