What are your ultimate commercial rocket designs?

[Hyperion5]

So, the point is not "ultimate" as in final forever, but rather the style of design that many companies will converge on after many years of operation.

I think that's the point to be made in that discussion. Rocketry is nowehere near this state.

Also, how about a F-22 raptor? An UAV? Weather/spying high altitude balloons? VTOLS? B-2?

F-22 Raptor is a military vehicle, as are many UAVs, weather and spy balloons rarely have the kind of commercial utility of airliners, VTOL aircraft so far are mostly the province of the military, and we all know there's little commercial utility to a B-2. 

Rocketry is beginning to move towards a certain number of more standard features.  Take the number of stages for example.  When the USSR built the Sputnik-PS from the R-7, it had only one stage.  Within a few short years they were building Proton rockets with three and four stages.  The N-1 rocket had no fewer than five stages.  Atlas rockets were famous for their relatively unique staging.  In the early years there was relatively little consensus on the number of stages a rocket should have. 

Look at the situation today.  The incoming Angara & Long March 5 rocket families have a maximum of 3 serial stages, and will often fly with just two.  Adding in boosters makes them 2.5 & 3.5 stage rockets.  The Ariane 5 is a 2.5 stage rocket, as will be the Falcon Heavy, the Space Launch System, and if the Russians built the Sudrozhestvo (a 5-Zenit core HLV), that as well.  Both the Delta IV & Atlas V are 2 stage rockets in their base format, and 2.5 stage rockets with added cores as boosters.  If I had a guess as to why we're settling on 2-3.5 stage rockets as our common range, I'd guess it would be to minimize serial staging risk.  All one has to do is compare the reliability of the Soyuz family vs the Proton family over the years to see some of this risk in action.

[Andrew_W]

What's the ultimate commercial aircraft design?

bingo

Long cylindrical fuselage, swept wings with complex control surfaces and winglets, high-bypass turbofan engines in nacelles either on the wing or tail, dense seating in the cabin, all facing forward, APU in the tail to provide ground power. Next time you are at the airport, I dare you to find a commercial passenger aircraft that doesn't follow this description.

So, the point is not "ultimate" as in final forever, but rather the style of design that many companies will converge on after many years of operation.

Well said.

[Rugoz]

You should be more specific, like

The cheapest/most reliable launcher for...

"20 6t comsats into GTO orbit per year"

or

"4 20t payloads into LEO per year"

The ultimate commercial rocket design depends on market conditions, just like the "ultimate commercial airplane" design.

[watermod]

Tongue totally in cheek and wearing an evil Ming the Merciless hat...
 
It starts with a business plan that removes competition...
so...
Start with spinoff's of the original Project Orion launching as near as possible to your main competitor's factories. I should probably go up on one of the Orion launches to avoid the lawsuits and arrest.

After the radiation takes care of the  competition ... any kind of rocket for normal launches I wanted...
 
Sorry - this question was just too wide open.

[RanulfC]

As one of the folks whom Hyperion5 ran this thread-concept by for "checking" before he posted I apologize (a lot!) for missing the fact it could be seen as too "broad" a subject

However I'll take a "stab" at answering both the questions of:
What would your ideal commercial rocket designs look like?
What's the ultimate commercial aircraft design?

Since I've recently run across a company that is trying for both
(Ok, technically TransPower is ONLY the lead on one segment and an overall "management" group for the project. However their "partner" and "lead" company for "Application Analysis" including research, design and evaluation of concepts and systems is Boeing

How about a "Nuclear Hybrid Space Plane"?
http://www.transpowerusa.com/wordpress/space-transportation/commercial-space-plane/

I'm kind of "impressed" that their Operational Goals for the vehicle extend all the way out to Jupiter:
http://www.transpowerusa.com/wordpress/space-transportation/commercial-space-plane/operating-goals/

The idea is to use a "core" of Fusion Plasma (Gasdynamic Mirror, GDM) to produce neutrons to bombard a "blanket" of Thorium-232 which transmutes to Uranium-233 and undergoes fission. The thermal output would then be applied to a reaction-mass and used in Combined Cycle Engines in both Air-Breathing and Rocket modes to achieve space flight.

Nice work, if you can get it

Randy

[Elmar Moelzer]

At a very brief first look, I like the concept. I will look at it some more when I have the time. One of the problems with conventional nuclear thermal propulsion is that small amounts of nuclear fuel and fission products would escape with the exhaust gases. Not sure whether that is a problem with their system too, but it might give some people reason for concern.

[Hyperion5]

As one of the folks whom Hyperion5 ran this thread-concept by for "checking" before he posted I apologize (a lot!) for missing the fact it could be seen as too "broad" a subject

That's alright, Randy.  It's kind of hard to foresee everything, as people in my field know all too well. 

However I'll take a "stab" at answering both the questions of:
What would your ideal commercial rocket designs look like?
What's the ultimate commercial aircraft design?

Since I've recently run across a company that is trying for both
(Ok, technically TransPower is ONLY the lead on one segment and an overall "management" group for the project. However their "partner" and "lead" company for "Application Analysis" including research, design and evaluation of concepts and systems is Boeing

How about a "Nuclear Hybrid Space Plane"?
http://www.transpowerusa.com/wordpress/space-transportation/commercial-space-plane/

I'm kind of "impressed" that their Operational Goals for the vehicle extend all the way out to Jupiter:
http://www.transpowerusa.com/wordpress/space-transportation/commercial-space-plane/operating-goals/

The idea is to use a "core" of Fusion Plasma (Gasdynamic Mirror, GDM) to produce neutrons to bombard a "blanket" of Thorium-232 which transmutes to Uranium-233 and undergoes fission. The thermal output would then be applied to a reaction-mass and used in Combined Cycle Engines in both Air-Breathing and Rocket modes to achieve space flight.

Nice work, if you can get it

Randy

This design concept would almost be like suddenly going from Apollo to launching into orbit like the Millennium Falcon does in Star Wars.  If they could make it work it'd revolutionize the field and make the colonization of the moon and Mars a relative cake-walk to what it is today. 

I'll say this for my design concept of the Neptune V.  It may not be as revolutionary as that, but it's basically taking current design trends and pushing them to the extreme.  It's got common tanks on both stages like a Falcon 9, it improves its dry mass fraction by using common bulkheads on both stages (like on Saturn V's S-II), and this is all thanks to its uses a cheap easy-to-handle propellant mix (not dissimilar to Soyuz & Falcon 9).  That cuts construction and handling costs considerably. 

To max out reliability I decided it needed multi-stage engine-out capability, which is enabled by five engines per stage.  That means you're using the same quantity of engines per launch as a Falcon 9 (though half of them are much smaller) but since there are only five in action at a time, the stage failure rate is likely to be quite low if you can get the engines reliable.  To maximize reliability and minimize serial staging risk it only has two stages (like Falcon 9/Atlas V/Delta IV).  Also, due to the engine count per core, you can make this thing extremely modular like the Angara family.  1, 3, 4 and 5 core versions are all possibilities, with even the 25 engines lifting the Neptune V Ultra 2 fewer than the Falcon Heavy. 

To maximize performance it would use lithium-aluminum alloys to save weight.  Its engines are also basically cousins of each other, with the same chamber pressures (2900 Psi), and both engine types are high-efficiency staged combustion metholox engines.  They're a bit overkill for rockets of this size, but the reason for that is I want to still be able to send up useful payloads on fully reusable versions.  That's enabled by the x layout of the five engines on each stage, so once you can get a Grasshopper-style program running, these things ought to be able to match Elon's Falcon 9. 

The only issue right now is I'm trying to figure out what size Neptune V to consider.  I'm currently considering 3 versions varying in power weighing 478 mt, 510 mt and 636 mt respectively.  The smallest would be an excellent Falcon 9/Zenit 3 competitor, while the largest would be taking on the Ariane 5, Long March 5, Delta IV Heavy & Angara A5. 

[RanulfC]

At a very brief first look, I like the concept. I will look at it some more when I have the time. One of the problems with conventional nuclear thermal propulsion is that small amounts of nuclear fuel and fission products would escape with the exhaust gases. Not sure whether that is a problem with their system too, but it might give some people reason for concern.

Actually by the end of the NERVA program there was almost no measurable product escape even after being run for hours. The system described in the link would either use heat-exchanger system or run liquid thorium in a heat exchanger to move the heat to the exhaust gas.

Randy

[Hyperion5]

At a very brief first look, I like the concept. I will look at it some more when I have the time. One of the problems with conventional nuclear thermal propulsion is that small amounts of nuclear fuel and fission products would escape with the exhaust gases. Not sure whether that is a problem with their system too, but it might give some people reason for concern.

Actually by the end of the NERVA program there was almost no measurable product escape even after being run for hours. The system described in the link would either use heat-exchanger system or run liquid thorium in a heat exchanger to move the heat to the exhaust gas.

Randy

So the question then becomes, can you make a nuclear thermal rocket-powered LV politically acceptable?  If you can realize the design, you're only halfway to commercial rocket heaven.  Imagine you're the program head of this hybrid nuclear rocket spaceplane project and the design is finally getting ready for powered flight.  Your major problem is the Sierra Club and a huge crowd of anti-nuclear activists from across the world have descended on Cape Canaveral to try to stop you.  How do you win the PR war with these people?  Security will stop them from preventing the flight, but how do you deal with the Congressman demanding this sort of vehicle be outlawed from the states?  I would think you'd want to be able to fly from the major continents, not out of isolated islands. 

[RanulfC]

So the question then becomes, can you make a nuclear thermal rocket-powered LV politically acceptable?  If you can realize the design, you're only halfway to commercial rocket heaven.  Imagine you're the program head of this hybrid nuclear rocket spaceplane project and the design is finally getting ready for powered flight.  Your major problem is the Sierra Club and a huge crowd of anti-nuclear activists from across the world have descended on Cape Canaveral to try to stop you.  How do you win the PR war with these people?  Security will stop them from preventing the flight, but how do you deal with the Congressman demanding this sort of vehicle be outlawed from the states?  I would think you'd want to be able to fly from the major continents, not out of isolated islands. 

Not to get to far "off-topic" but it IS the question and the answer is actually pretty easy IF one "assumes" a reasonable world

Public education and outreach is a big factor here because you won't HAVE "huge-crowds" of anti-nuclear activists if you are open and honest in providing the public information because the majority of "anti-nuclear" activists tend to be misinformed at the least. Part of that education and outreach would of course have to target politicians as well, due to their positions. Though in truth all the education and honesty in the world won't overcome some of the "concerns" and those will have to be addressed in design or operations planning. Even then there will always be a "hard-core" segment that will never be convinced that anything "nuclear" can ever be safe in any form or fashion and there will have to be considerations made of their effect as well.

In the end it might JUST be "better" all around to at least start flying from a remote Pacific island until a decent flight record is built up and public fears begin to fade. Part of what makes this more likely in a Thorium scenerio than most others is the non-proliferation aspect of Thorium over (say) Uranium or Plutonium reactors. Thorium would "allow" some face-saving political manuevering to permit many "first-world" nations to support putting thorium reactors, processing and even "spaceport" facilities into some equatorial countries without seeming to support nuclear proliferation or weapons development yet still maintaining "control" and being part of the process for profit and political purposes.
(Of course in the long run such a "policy" and attitude could easily come back to bite the "supporting" nations in the butt given the whole convayed attitude of "better your backyard than mine" but I for one have never noted a reall "long-view" political attitude for most nations in the fist place

In the end "I" feel the PR battle is winnable and in truth it has always seemed to me to be a battle for pro-nuclear advocates to "lose" rather than for anti-nuclear advocates to "win" in the public support arena. YMMV of course

Randy

[Hyperion5]

At a very brief first look, I like the concept. I will look at it some more when I have the time. One of the problems with conventional nuclear thermal propulsion is that small amounts of nuclear fuel and fission products would escape with the exhaust gases. Not sure whether that is a problem with their system too, but it might give some people reason for concern.

Actually by the end of the NERVA program there was almost no measurable product escape even after being run for hours. The system described in the link would either use heat-exchanger system or run liquid thorium in a heat exchanger to move the heat to the exhaust gas.

Randy

So what exactly are the major engineering challenges to turning this into reality?  Which challenge is larger, the engineering or paying for the vehicle's development?  I can see a business case for a nuclear hybrid rocket spaceplane, but how much upfront investment would it require?  If it's going to cost half of Bill Gates' 60 billion plus dollar fortune, then I can see why Boeing isn't exactly flinging resources at the problem. 

[edkyle99]

If I had a guess as to why we're settling on 2-3.5 stage rockets as our common range, I'd guess it would be to minimize serial staging risk.  All one has to do is compare the reliability of the Soyuz family vs the Proton family over the years to see some of this risk in action.

There are a couple of problems with this comparison. 

First, an R-7 rocket consists of six to seven propulsion modules, depending on mission, each of which must separate from something.  Proton only has four propulsion modules (stages) at most.  So R-7, the more reliable of the two, has more separation events than the less-reliable Proton system.

Second, staging accounts for a small percentage of failure causes.  Most failures are propulsion related.  Guidance/flight control also causes losses, as do what is probably the most annoying orbital launch failure - payload shroud separation issues.  I only show one staging related cause among all of the Proton failures over the years.

In my view, launch vehicle design is substantially driven by cost - both development and operating cost.  Minimizing the number of different propulsion system types (think Merlin or YF-22 or RD-107/108) cuts development costs.  High reliability helps to keep operating costs low.  A good design builds in reliability by not pushing margins too hard (see R-7).  A troublesome design pushes margins too much (see the Briz M failures on Proton).

As for this thread's open question, I believe that a good "commercial rocket" is more than just the rocket - it is also the launch site and the testing sites and all of the industrial infrastructure that supports the program. 

Assuming that we are talking about supporting a variety of payload types and orbits (which I believe we must), then I would nominate Kourou, or something like it, as a nearly-ideal launch site.  Kourou can, all by itself, support launches to nearly any orbital plane - something that no other launch site (except Sea Launch Odyssey and air launch Pegasus) can do.  Kourou also gains the velocity advantage of its near-equatorial position for GTO launches.

As for the rocket, or, rather, rocket family, I would nominate China's Chang Zheng, or something like it, as an example of a flexible, reliable option.  CZ comes in numerous variations, but all are based on the same basic propulsion system and launch system and flight control system, etc. China Academy of Launch Vehicle Technology (CALT) builds it all, and tests it all, providing strong engineering control for the program.  The CZ family can fly in just two stages to lift 2 tonnes to polar orbit, or can be augmented with strap on boosters (using nearly identical engines) to lift Shenzhou crew spacecraft weighing more than 8 tonnes to LEO, or can be topped by a high energy upper stage to lift nearly 6 tonnes to GTO.  CZ flies from relatively basic launch sites.  Like R-7, CZ is reliable (0.98 success rate in 115 launches since 2000) because it is generally a conservative design.  No other country in the world has anything like CZ that can cover such a large payload range using essentially one rocket.

 - Ed Kyle

[go4mars]

Well, I'd like one that can significantly expand the launch market through price and capability.  Easily reusable, scalable, and cheapest fuel (methane).  Aviation is something like a 700 billion dollar industry.  I'd aim to capture 1% of that, and would want to be able to provide worthwhile destinations and adventures in space. 

Can fly 1,2,3, or 5 1st stage cores per launch.  5 would be shaped like 8o8
Fully reusable, always crossfeeds when more than 1 core. 
1st stage cores are at least 20 meters in diameter. 

Gaseous combustion plume cushion for atmospheric re-entry, and propulsive landing (liquid combustion). 

It would have optionality to fly with no upper stage (just a payload on parallel staging to orbit), or 1 or 2 upper stages, also an option depending on mission/goals.  Option for the 3rd stage to be hydrogen (or nuclear if I'm allowed to suggest it). 

Upper stages would fly themselves into position in the stacking process, or be emplaced by 'quick-detach' rocket belts that have significant cosign losses and use their own fuel.  Cosign loses so the stage below doesn't get cooked.  Fast and precise.  Maybe neutral buoyancy dirigibles. 
Parallel staging might be generally preferred compared to stacking, because of ground handling challenges. 

There would be a semi-ballistic space plane that goes on as a top stage, and seats hundreds.  It would re-enter above the ocean hypersonically, then switch to electric power, cruising at 80 thousand feet or so.  Then slow down, and glide through the atmosphere toward an exciting high-g vertical propulsive landing, 30 miles or so from city limits.  Passengers then quickly transfer to very high speed vacuum tunnel trains to population hubs on the destination continent.  Refuel, relaunch.  Or if we demand to be more boring than that, a subsonic glide to touchdown at a conventional airport then taxi tens of miles back to a launch site (I might be straying off topic there). 

Anyway, back to the rocket:  It enables colonization of space, and fast(ish) large missions to the outer solar system.

It is sized such that it's largest iteration comfortably gets the ultra-high isp fusion system, (General Fusion's secret project: were such to exist) into LEO. 

Called the Zeus (Ian Douglas novel reference).

Slightly modified versions of the 1st stage cores will be launched mostly empty, as part of the payload, and used as a landing stage on Mars, the moon, or elsewhere; available for re-fuel and re-launch.  The Mars version might also have a big ballute that can double as a greenhouse or pressurized habitation structure (only needed for entry to Mars at interplanetary speeds and not required for point-to-point travel on Mars).

Part of the reason for large diameter is for radial artificial gravity within the transit vehicles and stations.  If first stage core is 20 m diameter, the payload, packed might be 30 meters.  Station, Inflated, might be 50+ meters. 

Ground handling and payload processing would be a considerable challenge at all stages. But not insurmountable. 

The End.

[Hyperion5]

If I had a guess as to why we're settling on 2-3.5 stage rockets as our common range, I'd guess it would be to minimize serial staging risk.  All one has to do is compare the reliability of the Soyuz family vs the Proton family over the years to see some of this risk in action.

There are a couple of problems with this comparison. 

First, an R-7 rocket consists of six to seven propulsion modules, depending on mission, each of which must separate from something.  Proton only has four propulsion modules (stages) at most.  So R-7, the more reliable of the two, has more separation events than the less-reliable Proton system.

A good point, but of those six-seven separation events, four of them are of the less dangerous parallel staging variety.  The overall point I should have clarified was that adding more serial stages, not parallel stages, adds serial staging risk but more importantly ups the chances of upper stage failures on the rocket.  Of the six failures and one partial failure of the R-7 family since 1992, four out of the seven have related to upper stage issues, with 3 out of those 4 relating to issues with the third stage. 

I've looked long and hard, and so far I haven't found an example of a Soyuz suffering a bad set of booster separations since 1992.  There was one LRB explosion in 2002, but that's different than a parallel staging going awry. 

Second, staging accounts for a small percentage of failure causes.  Most failures are propulsion related.  Guidance/flight control also causes losses, as do what is probably the most annoying orbital launch failure - payload shroud separation issues.  I only show one staging related cause among all of the Proton failures over the years.

In my view, launch vehicle design is substantially driven by cost - both development and operating cost.  Minimizing the number of different propulsion system types (think Merlin or YF-22 or RD-107/108) cuts development costs.  High reliability helps to keep operating costs low.  A good design builds in reliability by not pushing margins too hard (see R-7).  A troublesome design pushes margins too much (see the Briz M failures on Proton).

Again, while it's true the actual staging risk is not as bad as I feared, I was thinking more generally as well of upper stage failures. I looked through the Proton's flight record since 1992, and I'm counting some 14 failures and partial failures.  Of those 14 failures, a full 10 out of 14 of them concern something going wrong with the final stage, and those failures include issues with others besides Briz-M stages.  By cutting down to 2-3 serial stages, the Angara rocket family that will succeed the Proton family ought to see failure rates more like that of the R-7 family.  The Russians seemed to have learned the lesson that cutting serial stages improves reliability, and looking at the Proton's record I can see why.   

As for this thread's open question, I believe that a good "commercial rocket" is more than just the rocket - it is also the launch site and the testing sites and all of the industrial infrastructure that supports the program. 

Assuming that we are talking about supporting a variety of payload types and orbits (which I believe we must), then I would nominate Kourou, or something like it, as a nearly-ideal launch site.  Kourou can, all by itself, support launches to nearly any orbital plane - something that no other launch site (except Sea Launch Odyssey and air launch Pegasus) can do.  Kourou also gains the velocity advantage of its near-equatorial position for GTO launches.

Certainly a good location for launching all kinds of payloads to all sorts of orbits.  I certainly can't fault the ESA for setting up at the location. 

As for the rocket, or, rather, rocket family, I would nominate China's Chang Zheng, or something like it, as an example of a flexible, reliable option.  CZ comes in numerous variations, but all are based on the same basic propulsion system and launch system and flight control system, etc. China Academy of Launch Vehicle Technology (CALT) builds it all, and tests it all, providing strong engineering control for the program.  The CZ family can fly in just two stages to lift 2 tonnes to polar orbit, or can be augmented with strap on boosters (using nearly identical engines) to lift Shenzhou crew spacecraft weighing more than 8 tonnes to LEO, or can be topped by a high energy upper stage to lift nearly 6 tonnes to GTO.  CZ flies from relatively basic launch sites.  Like R-7, CZ is reliable (0.98 success rate in 115 launches since 2000) because it is generally a conservative design.  No other country in the world has anything like CZ that can cover such a large payload range using essentially one rocket.

 - Ed Kyle

I'm not a fan of the current CZ family, and it comes down to two reasons.  The first is the fuels that propel current Chang Zheng rockets.  They're toxic, hypergolic propellants that cause huge issues whenever there's a failure and they fall near populated areas.  If a Falcon 9 failed shortly after launch, it would not be leaving a horrific mess and launching many horrific vapors into the air.  The Chinese know first-hand how dangerous this is from the time when a CZ veered off-course and killed 100 people in the 1990s.  Those toxic propellants also add handling costs and complexities, which can hardly help on costs.  Given the current CZ family is about to be replaced by non-hypergolic LV families like the Long March 5 series, I'd say even the Chinese agree hypergolic fueled LVs are not good long-term solutions. 

My second reason has to do with versatility.  Sure the CZ family as it is right now is more flexible and uses a lot of common parts.  However, once Russia starts launching the Angara family, even the upcoming CZ families will pale in versatility & commonality by comparison.  The Angara 1.2 will launch only 3.7 mt, while an Angara 5/KVRB will launch up to 28.5 mt.  There's even talk of Angara 7 versions launching over 40 mt and perhaps as much as 50 mt, with all but the big Angara 7 versions launching from a common pad.  All versions, except perhaps the biggest proposed Angara 5 & 7 versions, would use a common URM, and all of them use a common and proven first stage engine in varying numbers.  They all use, with the exception of the Angara A5P, a common second stage and engine.  The only real variety is in third stage, where the choice is between the Briz-M and varying numbers of RD-0146 LH2 engines (equivalent to RL-10B-2). 

Sure the Angara family is not yet proven, but I would bet long odds on it being more ideal commercial rocket family than the Chang Zheng family as it is or will be in a few years.   

Edit--correcting my faulty memory on fatality count.  The articles still count 80 houses destroyed, which makes a fatality estimate of 100 entirely reasonable by third parties.  This is an authoritarian state we're talking about here, not France or the US.  There is a tendency amongst such states to cover up the full extent of the disaster, the best example being Chernobyl in the USSR.

[edkyle99]

I'm not a fan of the current CZ family, and it comes down to two reasons.  The first is the fuels that propel current Chang Zheng rockets.  They're toxic, hypergolic propellants that cause huge issues whenever there's a failure and they fall near populated areas.  If a Falcon 9 failed shortly after launch, it would not be leaving a horrific mess and launching many horrific vapors into the air.  The Chinese know first-hand how dangerous this is from the time when a CZ veered off-course and killed 400 people in the 1990s. 

There is no basis for that claim, and the number keeps growing as the years pass.  At the time, China's official media reported six dead and 57 injured.  A U.S. Congressional report estimated that the real number might have been 100 "or more", but also provided no basis for the claim. 

The CZ problem isn't the hypergolic propellants, it is the inland launch sites.  A big kerosene/LOX rocket falling on the town would have been just as devastating as a CZ-3B.  Hyper fueled rockets have been safely launched for years from sea-side launch sites adhering to strict launch criteria.  Ariane 4 flew from Kourou, safely and very successfully, for many years, as did Titan from the Cape and Vandenberg.  (Nearly every rocket launched carries some toxics, if not for the rocket almost certainly in the satellite).

Going to RP/LOX in place of hypergolics is a good idea because it simplifies fueling processes, which helps reduce costs.

Re: the serial staging discussion.  The problem with Proton, IMO, isn't that it has a fourth stage, it is that Briz M is being made to perform very long burns.  The propulsion system wasn't initially designed for such burns, which have been facilitated by the addition of a drop tank.  Longer burns mean longer heat-soaking periods for chambers and rotating machinery, which stresses everything.  The long burns are separated by even longer coast periods, pushing mission length to 9 hours or more. 

 - Ed Kyle

[Hyperion5]

I'm not a fan of the current CZ family, and it comes down to two reasons.  The first is the fuels that propel current Chang Zheng rockets.  They're toxic, hypergolic propellants that cause huge issues whenever there's a failure and they fall near populated areas.  If a Falcon 9 failed shortly after launch, it would not be leaving a horrific mess and launching many horrific vapors into the air.  The Chinese know first-hand how dangerous this is from the time when a CZ veered off-course and killed 100 people in the 1990s. 

There is no basis for that claim, and the number keeps growing as the years pass.  At the time, China's official media reported six dead and 57 injured.  A U.S. Congressional report estimated that the real number might have been 100 "or more", but also provided no basis for the claim. 

Faulty memory on that one, sorry.  I've edited it down to 100 people as a reasonable estimate given 80 homes were destroyed.  I still think you might not want to entirely rely upon official Chinese media for your figures on this.  The official press is not exactly noted for always telling the truth when it's awkward for the CCP. 

The CZ problem isn't the hypergolic propellants, it is the inland launch sites.  A big kerosene/LOX rocket falling on the town would have been just as devastating as a CZ-3B.  Hyper fueled rockets have been safely launched for years from sea-side launch sites adhering to strict launch criteria.  Ariane 4 flew from Kourou, safely and very successfully, for many years, as did Titan from the Cape and Vandenberg.  (Nearly every rocket launched carries some toxics, if not for the rocket almost certainly in the satellite).

Going to RP/LOX in place of hypergolics is a good idea because it simplifies fueling processes, which helps reduce costs.

That last line basically sums up my case for not using hypergolic fuels--it costs less and is safer.  This is precisely why the Chang Zheng family is not staying with hypergolic fuels.  They are not good fuels for commercially cost-competitive LV families nor are they fun to handle.  Hypergolic fuels are good for vehicles like ICBMs or if you want an all-weather launcher ready to lift off at any time.  For me, commercially competitive LV families ought to be man-rated and their handling should be relatively easy.  The last man-rated hypergolic LV from the US was the Titan II, and I think there's a good reason for that.  Besides, imagine trying to sell space tourists on hitching a ride on a Chang Zheng versus a Soyuz.  I think they'd be sold on the Soyuz from the moment you described what the propellants in a Chang Zheng could do to them. 

Re: the serial staging discussion.  The problem with Proton, IMO, isn't that it has a fourth stage, it is that Briz M is being made to perform very long burns.  The propulsion system wasn't initially designed for such burns, which have been facilitated by the addition of a drop tank.  Longer burns mean longer heat-soaking periods for chambers and rotating machinery, which stresses everything.  The long burns are separated by even longer coast periods, pushing mission length to 9 hours or more. 

 - Ed Kyle

Alright, but it's not just been the Briz-M that's been failing, or else the Proton's lifetime reliability would not be so much worse than the R-7 family's.  Just counting from 1992, I count a DM-2 upper stage failing in February 1996, a Block D-2 upper stage failing to restart in November 1996, a Proton K/DM3 upper stage failing one second into its second burn in December 1997, and a Block DM upper stage failure in November 2002.  That's just the final stage failures before the Briz-M came into regular service.  There's something wrong about the way the rocket's designed to go about its missions if it has this many problems with the upper stage.  Besides that, we have the example of the Soyuz failures since 1992.  Three-fourths of the upper stage failures happened on the third stage, not the second stage.  I don't think that's necessarily just a coincidence.     


I would bet almost anyone would not design a 4-stage rocket if they were trying to create their ultimate commercial rocket design from scratch.  I seriously doubt you can make a 4-stage rocket as reliable as one with 2-3 stages even if your overall build & design quality is excellent.  More stages generally equals more things that can go wrong.  I'm fairly sure your risk isn't going down when you tack on another stage and add yet another burn to the flight plan. 

[edkyle99]

Alright, but it's not just been the Briz-M that's been failing, or else the Proton's lifetime reliability would not be so much worse than the R-7 family's.  Just counting from 1992, I count a DM-2 upper stage failing in February 1996, a Block D-2 upper stage failing to restart in November 1996, a Proton K/DM3 upper stage failing one second into its second burn in December 1997, and a Block DM upper stage failure in November 2002.
...
I would bet almost anyone would not design a 4-stage rocket if they were trying to create their ultimate commercial rocket design from scratch.

Fewer stages obviously means fewer engines, which should mean less money.  However, I'm still not ready to agree that fewer stages automatically means better reliability.  I think that in real life other factors tend to outweigh the number-of-stages factor.

Consider that Proton M/Briz M currently sports about the same reliability record as Sea Launch Zenit and China's CZ-3 series - rockets that only have three stages.  (It also has a better record than two-stage Falcon 9, but it is too soon to really make any judgements given Falcon 9's low number of flights.)  Also consider that Proton with DM-2 or DM-2M upper stages proved more reliable than Proton with Briz M (and also more reliable than Ariane 5G). 

CZ itself is very reliable.  The CZ-4 series is essentially tied with Soyuz U and Soyuz FG at the top of the reliability list for active launch vehicles. 

 - Ed kyle

[Hyperion5]

Alright, but it's not just been the Briz-M that's been failing, or else the Proton's lifetime reliability would not be so much worse than the R-7 family's.  Just counting from 1992, I count a DM-2 upper stage failing in February 1996, a Block D-2 upper stage failing to restart in November 1996, a Proton K/DM3 upper stage failing one second into its second burn in December 1997, and a Block DM upper stage failure in November 2002.
...
I would bet almost anyone would not design a 4-stage rocket if they were trying to create their ultimate commercial rocket design from scratch.

Fewer stages obviously means fewer engines, which should mean less money.  However, I'm still not ready to agree that fewer stages automatically means better reliability.  I think that in real life other factors tend to outweigh the number-of-stages factor.

Consider that Proton M/Briz M currently sports about the same reliability record as Sea Launch Zenit and China's CZ-3 series - rockets that only have three stages.  (It also has a better record than two-stage Falcon 9, but it is too soon to really make any judgements given Falcon 9's low number of flights.)  Also consider that Proton with DM-2 or DM-2M upper stages proved more reliable than Proton with Briz M (and also more reliable than Ariane 5G). 

It may not be an automatic guarantor of better reliability to have fewer stages, but it surely doesn't help reliability to have more stages either.  Of the four failures & partial failures involving the upper stages of the R-7 family since 1992, three-fourths have involved a third stage, not the second stage.  I also have to point out the reliability record of the Ariane 5G is skewed because it was the initial vehicle of the Ariane 5 family.  They hadn't ironed out the design's initial issues by then, while the Proton by that time was an established design that had been launching for decades.  If you had compared the Proton's early years with that of the Ariane 5G's early years, the comparison would not be nearly as pleasant for the Proton. 

I actually talked to Jim a bit on the Zenit.  At the time I thought the reason why the Zenit's reliability was poor (but improving) was the quad-chamber design of its first stage engine.  I contrasted that with the ever-reliable RD-180, which ULA was considering mounting two per core of for a prospective Atlas V Phase 2.  I thought the reason why ULA wanted two RD-180 engines is because it would be more reliable due to their dual-chamber design than a single quad-chamer RD-170. Jim said the following about that theory:

Wrong.  The reason they use RD-180 for advanced designs is because they have a source for engine.

RD-17X problems have nothing to do with the number of chambers.  It has to do with manufacturing quality.

So basically if the build quality were higher for the RD-170 engines used on the Zenit, you wouldn't be seeing anywhere near the number of first stages exploding, which would improve the Zenit's reliability considerably.  Imagine if an enlarged Atlas V used an RD-170 engine.  I would guess it'd suffer very few first stage failures in contrast to the Zenit because the quality control would be higher.  I still don't think the Zenit is the most ideal LV family, but it is potentially very modular.  RKK Energia's proposed creating a 2.5 stage, 5-core Zenit called the Sodruzhestvo to be Russia's HLV in the future.  http://www.russianspaceweb.com/sodruzhestvo.html
 

CZ itself is very reliable.  The CZ-4 series is essentially tied with Soyuz U and Soyuz FG at the top of the reliability list for active launch vehicles. 

 - Ed kyle

You know, there's probably a way for the R-7 family to top the Chang Zheng family in reliability and have almost as good versatility as well.  My idea is you take the Soyuz 2.1V and Soyuz 3 concept and take the best elements from both for a new R-7 family. 

http://www.russianspaceweb.com/soyuz1_lv.html
http://www.russianspaceweb.com/soyuz3_lv.html

For the base version, you simply add on an advanced ACES-type upper stage with four RD-0146E engines to the Soyuz 2.1V.  You factor a T+.1 engine-out capability into the upper stage payload number and throttle back the stage.  For the next version, you throttle the upper stage up, add on a pair of NK-33-1 boosters and lengthen the upper stage propellant tanks.  The biggest version would feature the normal 4 boosters and have the longest upper stage and still have T+.1 engine-out redundancy factored in on the upper stage.  Something like this ought to be even more reliable than the existing R-7 family and more modular, though you'd want all versions to be able to lift off a single pad like the Angara 1.2-Angara 5/KVRB. 

I factored in engine-out capability because ULA has been pretty adamant what they think the reliability effects of having it: "Significantly, the dual and quad engine ACES variants incorporate engine-out.  Engine-out capability provides the single largest lever to improving system reliability (Figure 4). It is a key basis for ULA’s desire for multiple smaller (25Klb) class engines rather than a single larger upper stage engine."

According to ULA's math, you'd see a better than order of magnitude improvement in 2nd stage failure rates with the Atlas V.  I think given how similar the RD-0146 engine is it wouldn't be too hard to copy this with a more modern and less costly engine for the R-7 family.  A Soyuz family like this might even keep Musk up at night before the Falcon 9 achieves demonstrated reusability. 

Edit--
An alternative cheaper version would be four RD-0146U engines up top burning CH4/Lox (metholox version of RD-0146).  You'd give up some payload lift but cut your handling issues significantly plus not have to worry about boil-off as much prior to additional burns. 

[Hyperion5]

Alright, but it's not just been the Briz-M that's been failing, or else the Proton's lifetime reliability would not be so much worse than the R-7 family's.  Just counting from 1992, I count a DM-2 upper stage failing in February 1996, a Block D-2 upper stage failing to restart in November 1996, a Proton K/DM3 upper stage failing one second into its second burn in December 1997, and a Block DM upper stage failure in November 2002.
...
I would bet almost anyone would not design a 4-stage rocket if they were trying to create their ultimate commercial rocket design from scratch.

Fewer stages obviously means fewer engines, which should mean less money.  However, I'm still not ready to agree that fewer stages automatically means better reliability.  I think that in real life other factors tend to outweigh the number-of-stages factor.

Consider that Proton M/Briz M currently sports about the same reliability record as Sea Launch Zenit and China's CZ-3 series - rockets that only have three stages.  (It also has a better record than two-stage Falcon 9, but it is too soon to really make any judgements given Falcon 9's low number of flights.)  Also consider that Proton with DM-2 or DM-2M upper stages proved more reliable than Proton with Briz M (and also more reliable than Ariane 5G). 

CZ itself is very reliable.  The CZ-4 series is essentially tied with Soyuz U and Soyuz FG at the top of the reliability list for active launch vehicles. 

 - Ed kyle

I just thought of something when we're talking about reliability.  I saw this claim by Spacex on their Falcon 9 page:

"Falcon 9 has nine Merlin engines clustered together. This vehicle will be capable of sustaining an engine failure at any point in flight and still successfully completing its mission. This actually results in an even higher level of reliability than a single engine stage. The SpaceX nine engine architecture is an improved version of the architecture employed by the Saturn V and Saturn I rockets of the Apollo Program, which had flawless flight records despite losing engines on a number of missions."


While I've been impressed with the reliability of the Soyuz & Chang Zheng families, I've always wondered if they might benefit from engine-out capability.  Both the Saturn V & Saturn I/IB have engine-out capability on at least one stage, and both rocket families actually wound up using that capability.  It's true none of the Saturns were really commercial LVs, but I noticed something remarkable about their reliability record.  All of the Saturns that did lift off on flights intended to make orbit accomplished the task, just as Spacex claims, even if one Saturn V made an orbit that wasn't quite in the plans. 

Which begs the question, should an ideal commercial LV have engine-out capability?  I've run across literally dozens of instances where a single engine stage, most often an upper stage, fails, and thus a mission is doomed.  Surely that has a significant cost to launch providers even if they're cutting expense and complexity with single-engine stages.  Are the extra costs and complexity be worth it if the chance of engine failure costing the mission is significantly reduced? 

[Jim]

Falcon 9 can not sustain an engine failure during second stage flight, which is more than 2/3 of the time to orbit.

Also, Soyus can not sustain an engine out as there is only one engine per booster