Comment by rayiner | Hacker News Reader

Comment by rayiner | Hacker News Reader

I found this quote interesting: "There are a decent number of people working on more efficient jet engines, but not very many civilians working on replacing the jet engine with something new."

It's not really the problem that there aren't very many people working on replacing jet engines. The issue is that the problem domain is full of physical limits. Ask yourself: why are we still using quicksort now, half a century after it was developed? Well, because theory tells us that any sorting algorithm based on binary comparisons is going to require at least O(n log(n)) such comparisons, and that is going to remain the case no matter how much engineering effort you throw at the problem. All you can really do when it comes to sorting algorithms is improve on the constant factors and take advantage of special cases (radix sort for small integer items, etc).

Software engineers don't usually have to think in terms of theoretical limits. But when I was in engineering school, as an aerospace major, every course began with: these are the theoretical limits in this particular area--all you can do is approach them more closely. A Brayton cycle has a certain efficiency at a given pressure ratio and an airfoil has a certain maximum lift-curve slope and there is nothing you can do about it. Improvements are in "hard engineering"--new materials that can withstand higher temperatures and pressures to eke out slightly more efficiency, etc. You're already seeing this in semiconductor manufacturing. As you run into the limits of lithography with ever-smaller wavelengths of light, semiconductor fabrication plants get exponentially more expensive to build.

These physical limits are what give the traditional engineering fields the shape of their "innovation curve." Much more happened in aerospace engineering from 1930 to 1960 than has (and will) happen from 1960 to 2020. The long-hanging fruit is gone, and all that's left is hard engineering.

And of course, this all plays into the "get rich quick" issue the author brings up. If you had a billion dollars lying around, where would you invest that money? A jet engine that's 0.5% more efficient? If you had $100 billion, would you spend it making an alternative to jet engines entirely (in an industry where people will fly on truly shitty airlines just to save $5 on a ticket)?

It's not widely known or understood, but there are, in my opinion, quite a few ways to make significantly superior jet engine replacements.

Factors of 2 or 3 in specific power and efficiency are possible. Due to the rocket equation, this could yield vastly longer range and more efficient aircraft (1/10 the weight, 10x the range, 10x the efficiency, 2x the speed...).

In particular, you point out the limitations of the Brayton cycle and material limitations. However, there are other cycles that are much closer to Carnot efficiency, or match it in the ideal case. There's a lot you can do when isothermal compression is possible. Additionally, if there is a thermal limitation, you can do isothermal combustion, such that much more heat addition is at the maximum temperature of your materials, and you can even design efficient reheating.

It's not a jet -- you want to pair this with a driven fan system or at least have high bypass, but it's fundamentally more thermodynamically efficient.

For supersonic applications, scramjets remain possible, but challenging. And rocketry + high altitude flight remains extremely exciting and potentially more efficient than standard jetliner transport. Just difficult.

So yes, there are physical limitations. But they are not what most people think they are, and they do not mean what people think they mean.

I didn't mean to imply that, say, the Brayton cycle was an absolute physical limit to aerospace vehicle performance that cannot be worked around. Of course, you can switch from jet engines to scramjets, you can switch from optical lithography to x-ray lithography, you can avoid the realities of supersonic drag by going into space, etc.

But: a) all that is risky and extremely expensive (and as the semiconductor example shows us, often exponentially so); and b) whatever technology you use, you will always be pushing up against physical limits--that's just the nature of the physical engineering disciplines.

These are novel concepts in a world where we see "technological progress" as the evolution of the internet from AOL in 1995 to Facebook in 2010 and wonder why other areas don't improve so fast.

PS) I'd be quite interested to see any specific papers you had in mind vis-a-vis alternative cycles for jet engines.

I think the point is to retreat farther from the idea "a better jet engine" to "a better way to fly" or "a better method of propulsion". By shopping around in that larger idea space you may be able to trade in your current physical limitations for something less constraining, or even just less known. Discussions on this topic always make me think of Neal Stephenson's article on rocketry, which in my mind is closely linked to local maximums of innovation/invention[1].

[1]: http://www.slate.com/articles/technology/future_tense/2011/0...

It's a good article, thanks for reminding me to re-read it.

I note that in it, Stephenson makes a prediction, as if it were a known fact, that I think we now have significant reason to doubt.

He says "Rockets are as close to perfect as they're ever going to get. For a few more billion dollars we might be able to achieve a microscopic improvement in efficiency or reliability, but to make any game-changing improvements is not merely expensive; it's a physical impossibility."

I think we now have reason to believe that for around a billion dollars ($1b is the estimated total R&D budget as of a year ago), SpaceX has already made better-than-microscopic improvements in efficiency. But more interestingly, there's reason to believe that with little or no additional outside investment (i.e. working on profits), SpaceX is pretty likely to soon make a very large improvement in efficiency, with Grasshopper &c. Rockets are probably not as perfect as they're ever going to get.

Without taking away from SpaceX's accomplishments, it's important to keep the performance of the Merlin engines in context. The Merlin 1d is arguably the most efficient rocket engine using a gas-generator cycle and RP-1 (kerosene) fuel. That's a very specific claim. Hydrogen-burning engines like the one on the Space Shuttle are much more efficient (measured by specific impulse), and some Soviet kerosene rockets came very close both in specific impulse and thrust to weight ratio, though they used a somewhat different cycle. Relevant discussion: http://www.flightglobal.com/blogs/hyperbola/2012/06/comment-....

Still a very impressive achievement, but looking at the performance of SpaceX's engines circa 2012 and comparing them to what the Soviets built in 1960 or 1970 helps put the pace of the industry into context.

I had assumed, in my ignorance, that the Merlin engines were not an important advancement. I don't think that specific impulse is necessarily the important sense of "efficiency" (though I do assume that it correlates with the important sense).

The relevant efficiency, to me, is "awesomeness achieved per human effort and non-renewable resource", and the useful proxy for this is "awesomeness per price". Arguably SpaceX rockets are cheaper (per given mass to given orbit, obviously) that most or arguably all of the competition. I say "arguably cheaper" because it's a little tough to figure out how to price the shoulders of giants on which they stand.

And really I only picked the word "efficiency" in the context of Stephenson's "efficiency or reliability".

Anyway, hopefully this explains why I think that true reusable rockets, Musk's often-stated ambition, qualify as a "very large improvement" in efficiency. Specific impulse is not expected to significantly improve, right?

Also, I apologize in advance for my legion failings in the area of comprehending aerospace engineering.

Not exactly efficiency, (though they have high combustion efficiency) but low cost and very good mass ratio. If they make it reusable, that where the killer improvement is. Elon says they're getting there; I'm crossing my fingers.

I utterly agree in substance, see my cousin-comment.

Do you think think it's unreasonable to call "payload per cost" a relevant "efficiency"?

In time I'll publish a paper and technical report. In a little more, I'll show a prototype. But I do have my hands fairly full!

I don't usually work from the literature (it is hard to access papers outside of academia, and they're usually weird, abstract, hard to read, and miss the point anyway) so I work from the first principles.

But at a high level, imagine either an Ericsson or Carnot cycle, (dependent on materials), where high temp, high pressure heat addition is contiguous combustion during expansion, and the isothermal compression is a secret sauce :-)

> I don't usually work from the literature (it is hard to access papers outside of academia, and they're usually weird, abstract, hard to read, and miss the point anyway) so I work from the first principles.

Amazing.

Amazing will be 30 years from now when I'm watching a plane fly over using a Fong-type engine ;).

> In a little more, I'll show a prototype

Say what you will about HN, but comment threads about seemingly impossible technical challenges that end in this make me immensely happy to be here :D

The world could use another Elon Musk. A female Elon Musk would be even better. Congratulations on Lightsail and keep thinking even bigger.

Why would a female Elon Musk be better than a male one? Are males inferior?

Same reasons it's awesome to have a black president:

1. There are people it would inspire, and

2. There are people whose expectations would be shaken up

The latter group needs as much help as the former.

There are few enough Elon Musks in the world that I think we should be grateful for whichever ones we get

Wow a black president made certainly a lot of difference. Just like the white ones : lying, deceptive, secretive, and dishonest (See NSA). Sorry if I fail to see much difference.

He's been mendacious as all of them, sure. That's not the important difference.

The important difference is that there are now millions of black American kids no longer growing up in a world where they just "know" as a first principle that "black people can't be president". That sort of example simply changes people's outlook. Inspiration and aspiration matter.

I rather see millions of people disappointed that no politician, no matter the race or creed, can ever be trusted. Obama had a chance to make a difference and he failed miserably.

Err yeah. Why would you think that a black president would be different to a white one?

That's precisely the shakeup the world needs - to realise that whatever skin colour, male or female, people are people and you shouldn't judge them negatively or positively on how much melanin or testosterone they produce :P

To be fair, it might have helped some people: black people are just the same as anyone else. They can be president and they'll be the same corrupt shit bags as white men who become president.

To serve as an inspiration to all the girls around the world to go into STEM fields. Things are definitely getting better but the current system of society expectations still discourages women from participation in technology, sometimes in very subtle ways. Even if we set aside equality of opportunities which is the goal by itself, think of it as extending the pool of potential inventors and innovators, speeding up the innovation.

Not inferior, different. Women have a different perspective, a diversity of perspectives (taken on merit) allows a group to not get stuck in a local min-max.

Are you claiming that women think differently than men? Sources, please.

In practice, the kinds of people that can discuss ideas that push the state of the art forward are so rare that you should be grateful that your society gets any, and not worry so much about their genitalia.

> Are you claiming that women think differently than men? Sources, please.

Effect of estrogen, progesterone, and testosterone on cognition.

> It's not widely known or understood, but there are, in my opinion, quite a few ways to make significantly superior jet engine replacements.

> Factors of 2 or 3 in specific power and efficiency are possible. Due to the rocket equation, this could yield vastly longer range and more efficient aircraft (1/10 the weight, 10x the range, 10x the efficiency, 2x the speed...).

I'd like to hear about these, assuming they produce comparable amounts of thrust as jet engines while being a reasonable size (so as not to increase drag excessively).

True but at the same time there are physical powers we witness in the universe that we cannot yet dream of harnessing but which, by virtue of their existence: do ultimately offer a whole new paradigm.

The main one that comes to mind is nuclear fusion... The power of the sun. We see and rely on it every day. We know it's possible to create with a trivial set-up: lot's of crap thrown together. We haven't figured out how to control it yet but once we do, it'll change everything.

So you're right: there are physical limits and some absolute of course, but we're not at the absolutes yet: it's just going to be a very step-wise function on the way up.

edit: while this may seem tangential to a discussion of jet engines: imagine if we did harness fusion... think about the aeronautical implications. We could probably have helicopter islands like in a Sci-Fi film. :)

Just an amusing note: a useful fusion power plant on Earth would probably require conditions many times more extreme than those inside the Sun.

"The power production density of the core overall is similar to the metabolic production density of a reptile. The peak power production in the Sun's center, per volume, has been compared to the volumetric heat generated in an active compost heap. The tremendous power output of the Sun is due not to its high power per volume, but rather to its gigantic size." http://en.wikipedia.org/wiki/Solar_core

This.

People romanticize the notion of Fusion. It's a great thing, no doubt, But to say:

> We could probably have helicopter islands like in a Sci-Fi film.

Feels a little bit to me like people saying the same thing when they first discovered refined petroleum.

Having said that we've made massive progress towards energy positive fusion with both magnetic confinement, and inertial confinement, And it's not crazy to think they may provide good usable power into our grid within our lifetimes.

Deuterium/Tritium fusion is vastly more efficient than light hydrogen fusion.

I think you're misunderstanding the concept here. They're saying that temperature gradients determine power production and since the sun is reasonably smooth in its gradient, there is minimal power generation possibility her volume..

But if you were able to contain a golf ball of sun, the heat it produced would be able to power far more than a reptile's brain as its heat was transfered to the environment...

_Perhaps_ you're pointing out that the sun's size is what allows it to 'contain' these many otherwise-uncontainable golf-ball size chunks. If so, I hear you... obviously the engineering challenge is developing a system on earth to contain. We've clearly shown with fusion bombs that we can harness this effect for a net energy gain for an instant.

My point was that these are engineering challenges, rather than physical limits.

>But if you were able to contain a golf ball of sun, the heat it produced would be able to power far more than a reptile's brain as its heat was transfered to the environment...

It is you who is misunderstanding the concept here. A golf ball of Sun produces heat only about as much as the reptile brain. In other way - if you put as many reptile brains in one pile so that the pile will be as large as Sun, you'll get the pile radiating with Sun's black-body radiation temperature and intensity. The trick here is that amount of energy produced is proportional to volume, while radiating surface is just square-proportional, thus black-body radiation temperature (i.e. frequency, ie. energy of a given radiated photon) of a pile of reptile brains (or golf-ball sized chunks of Sun) would be much higher than that of a one brain or the same set of brains (or chunks of Sun) spread around with significant distances between each other.

Edit: it isn't to say that fusion power can't be used as a practical source of energy on Earth. It is just some forms of [stationary] containment are ruled out, mainly by bremsstrahlung (Sun overcomes it precisely because of volume to ratio being big enough). Dynamic forms of containment are proven to be working (net-positive) at least by H-bomb.

I'm sorry, but the core of the sun is 150 times as dense as a reptile's brain and far far hotter. Thus you would need far more than the equivalent volume of brains to create the same effect. I see what you mean about the energy of escaping radiation, however.

For your point to be persuasive you need to add why temperature and density would increase power. e.g. one could have a core of extremely hot, dense inert material.

Something hot should not be thought of as inert, because something hot by definition is something with incredible internal velocity. But then there are multiple definitions of inert..

I think what you're saying, though, is that if you contain something extremely hot and dense, you cannot get any useful energy out... Because the energy needed to contain it prevents you from getting anything out. Right?

Because I'm coming from a perspective that says: if you can contain something like a small bit of solar core: and if you can let some of that heat leak to the outside environment: you can derive terrific power from the thermal dynamics at work.

The sun exists in the pressure-free vacuum and thus extends to a far less dense outer layer where the heat gradient (the argument against power potential per volume) is low. However if you could contain the sun into a smaller volume, it would have the same massive heat flux but over a much lower surface area and thus with far more power potential per volume.

I assume we're not so much disagreeing: over focusing on different issues. Is it correct that you're saying that the act of containment itself prevents the ability to harness the flux?

We might be slightly talking at cross purposes. I agree that something hot contains a lot of available energy. I'm making the point that it may not produce any available energy (i.e. inert in the sense of nuclear fusion).

You make some very good points, but if the theoretical limits are less clear cut in sofware (apart from very simple tasks like sorting), wouldn't it be all the more rational to pursue longer term research in this field?

We have no idea where the theoretical limit is in automatic natuaral language understanding for instance. Yet the rewards of any breakthroughs there would be enormous.

But I'm actually not so sure if hardware and software are really that different in terms of theoretical limits. The incentive to optimize existing approaches instead of exploring completely new directions in hardware may be be a result of much higher capital requirements for anything physical.

"X field/industry sure is slow to develop" is a favorite topic on HN, whether it's airline travel, car dealerships, or bicycle helmets. I think there's an undercurrent of comparing progress in every field to that in computation (i.e. Moore's law), without actually studying the other field.

FWIW, I've seen a trend toward merge or radix sort lately in a bunch of large systems, largely because they can be easily externalized (as datasets get bigger, you commonly need to sort data that won't fit in RAM). I also see a bunch of combined sorting algorithms that use different approaches at different stages of the algorithm, eg. timsort (mergesort + insertion sort) and combined mergesort + radixsort approaches for sorting data on disk or partially-ordered data.

Reply on news.ycombinator.com