Ariane 6: solid vs. liquid

[Proponent]

My hate for solids is due to the inability to do engine cut off.

Thrust termination for solids isn't out of the question.  The burn rate of solids generally goes as a power of the pressure, with index greater than one.  Vent the casing and the pressure will drop.  Thrust-termination ports were planned for the solids that the Air Force was going to use on the Titan III-M when launching the Manned Orbiting Laboratory.  Blowing the nozzle off has also been suggested.  Whether this is a good idea or not is another question; I'm just pointing out that the blanket statement that solids can't be shut down isn't true.

I think the biggest arguments against solids for human missions are that their failure modes are not graceful.  The entire casing is a critical area, whereas in a liquid-propellant engine only the small engine itself (including pumps, combustion chamber and nozzle) is critical.  It's therefore easier to monitor a liquid's health and shut it down if things start to go wrong.  If a solid fails, it's more likely to be spectacular, whereas when liquids fail they don't usually blow up and tear the vehicle apart.  And if if a solid were to fail non-spectacularly, there's no possibility of re-routing its propellant to another engine, as there is with a liquid.

All of that said, small solids like those of used on Atlas V (which are not segmented and don't have thrust-vector control) have very good records.  For example, over a thousand were launched on Delta IIs with just one outright failure (and one which failed to separate).  Hence, I suspect that adding a small solid or two to a liquid stage doesn't add much risk.

[Proponent]

How do the solids proposed for Ariane 6 compare in size with Vega's first stage?  Is there a synergy?

I agree with the comments above that, when you consider all factors, this design makes a lot of sense for ESA.  It keeps production going for large solids, which the military will like.  Ariane 5, despite subsidies, is becoming uncompetitive, so if ESA's giving up on commercial launches, Ariane 6 will fly infrequently, so solids may well offer the lowest cost.

Once Ariane 5 is gone, the only launch vehicle with segmented solids will be SLS.

[sdsds]

The P7C design for Ariane 6 would reportedly use a P135 motor, as compared with the P80 motor currently used for the Vega first stage. An Ariane 6 with the new motor, “could facilitate new development of a more powerful Vega engine that could lower production costs and greatly improve its competitiveness in the commercial smallsat market.

“For Vega, we want a P120 at a minimum,” says Enrico Saggese, head of Italian space agency ASI.

http://www.aviationweek.com/Article.aspx?id=/article-xml/AW_10_15_2012_p26-505016.xml&p=3

[Proponent]

Thanks.

[mmeijeri]

1. First stage liquid propulsion knowledge will be lost

Do you work on Ariane's core stage liquid propulsion yourself?

[tobi453]

No, I don't work on Ariane at all. I'm just a poor student concerned about the european space program. So don't believe everything I'm writing.

[Rugoz]

Maybe someone working on those ariane 6 studies could present us some rough cost estimates, if its not confidential  .

If you look at the ariane 5 cost breakdown there is no way you can offer the launch of an all liquid launcher with 3 vulcain engines for 70m.

The P7C however seems at least feasible on paper for 70m.

[mmeijeri]

If you look at the ariane 5 cost breakdown there is no way you can offer the launch of an all liquid launcher with 3 vulcain engines for 70m.

Do you have a link for that? I imagine a large part of the costs are fixed costs and a large part of those fixed costs are related to the solids.

[Patchouli]

It's a fact that gets ignored.

Kerolox > solids

The difference is both performance and price now. Casting and transporting solids cost big dollars and the lost payload over time of making your rocket heavier don't make sense.

That is all nice but as far as I can remember kerolox was never seriously considered for ariane 6. In the end it was multiple vulcains vs solids.

I thought there was an option to purchase Russian engines?

Just buy RD-180 and build an Atlas V copy.

I think they want to. If they can't make a Falcon 9 or a Soyuz themselves it's a good option.

Spacex and the Russians along with ULA's Atlas V seem to suggest kerolox is the way to go if you want the cheapest and most reliable first stage possible.

If they want a jobs program Snecma should be able to easily produce a large kerolox engine.
To get started quickly they could license the RD-180 or even an engine like the Tr-107 or F-1A or AJ-26-500.

A single RD-180 or F-1 first stage plus a cheap second stage might be able to deliver on the 70M target yet still be able to offer larger payloads with the addition of a better upper stage.
Maybe even use a cut down Vega first stage for the low end vehicle's second stage.

[woods170]

To quote Jim: Rockets are not Lego's.

[Rugoz]

Do you have a link for that? I imagine a large part of the costs are fixed costs and a large part of those fixed costs are related to the solids.

Don't know what you mean by fixed costs in this context, but the facility for casting boosters with 3m diameter is already up and running in kourou, the more you use it the better.

I posted this in another thread, however the numbers are from a german blog and only estimates which may be simply wrong:

I found a nice calculation on a german site:

Ariane 5

Production cost: 114m (not launch cost, based on arianespace's purchase of 35 rockets for 4bn)

Boosters: ~25m
Vulcain: ~15m
Upper stage: ~20m
EPC (incl. vulcain), VEB, fairing: 69m

Ok lets assume 3 cheaper vulcains at 30m + 20m upper stage, that is already 50m without boosters and core. 40% cheaper sounds very difficult to achieve.

Spacex and the Russians along with ULA's Atlas V seem to suggest kerolox is the way to go if you want the cheapest and most reliable first stage possible.

The russians and SpaceX also lack the knowhow to design and manufacture large solids.

Also, why are delta IV and atlas V so costly despite resembling the optimal designs many here advocate?

[spectre9]

Atlas V 401 is a beast rocket.

Look at it's stats, no solids 

Everything else with or without solids from the EELV range just doesn't float my boat like the 401.

Kerolox 1st stage, Hydrolox 2nd stage.

That's the way to go for the best performance.

The Russians and SpaceX lose a bit not having the LH2 stage.

[mmeijeri]

Don't know what you mean by fixed costs in this context, but the facility for casting boosters with 3m diameter is already up and running in kourou, the more you use it the better.

fixed costs = costs you incur even if you produce no engines
variable costs = additional costs per engine

If you build N engines, costs would be C = C_fixed + N * C_variable

With 3 * N engines, costs would be C = C_fixed + 3 * N * C_variable

If C_{fixed,average} := C_fixed / N >> C_variable, then total costs for 3 * N engines could be much less than 3 times the costs for N engines, plus you would get to eliminate the fixed and variable costs of the solids.

A similar argument applies to the solids, so it would be interesting to know all the components to determine which option would be cheaper. Just knowing the average costs of an engine at current production numbers doesn't tell you very much.

[cheesybagel]

IMHO Ariane 6 is an interim solution to replace the use of Soyuz for launching the Galileo constellation and other tasks like that. The infrastructure to cast solids is already there in Kourou where it is used for both P230 and P80. I would not be surprised if they only needed to change the casings and the electronics. This should provide a responsive launch solution for military and other critical payloads at a low R&D cost.

However Ariane 6 does not address the long term requirement for reusable launch which is being embodied in programs like the Volga LOX/LCH4 staged combustion rocket engine and IXV.

Still this is somewhat disappointing. In my opinion liquid rocket engines are the most viable solution for space launch we have. It will also be rather pathetic to witness European liquid rocket engine capabilities crumble just as China gets their staged combustion Long March 5 rocket working. A decade from now European leaders will wonder where their space launch infrastructure went as SpaceX, Chinese, and Russians grab the launch market.

[spacediver]

Wow, only a few days and more than 50 replies!
Obviously it was an urgent problem to be discussed.

I'd like to give some more results from the NELS study but I will not publish any concrete numbers as they are confidential. I hope you understand that I don't want to put my job in jeopardy...

In the study we investigated different launcher concepts, including concepts with LOX/kerosene, LOX / hydrogen and solid first stages, all combined with a cryogenic upper stage. We also investigated different versions of the PPH approach, including the CNES concepts.

The most interesting result is that all modular concepts, using a common core approach or strap-on-boosters, are not competitive to the “clean” inline designs.

The three most promising concepts KH (kerolox first stage with NK-33 engines), HH (cryogenic first stage with Vulcain 3) and PPH (solid first and second stage in-line) have practically the same recurrent cost. The small differences are well within the accuracy of our cost estimates.
Therefore, no economic justification for a solid Ariane-6 design can be derived from these results!   

BTW, no concept is able to meet the 70M Euro benchmark for 6,5t to GTO.

Taking into account the loss of crucial technology, for me the clear winner is a HH concept using 3 Vulcain-3 engines in the first stage for the large payloads (up to 6,5 t) and two engines on the same first stage with a 35% propellant offload for the small payloads (up to 3,5 t).

The development cost for the HH concept are in the same range as for the solid concepts, but the development risk is much lower as the Vulcain engines are well understood and the V3 is directly derived from the V2. Also the main stage tank is nothing but a stretched version (with different wall thickness of course) of the existing EPC stage.

The development time is also shorter. A “Block 1” version of the HH concept, using Vulcain 2 engines with shortened nozzles instead of the Vulcain 3 and the Ariane 5ME upper stage could be launched already in 2018! It would have a GTO payload capacity of only 4,3 t but it can be used to qualify the first stage prior to the availability of the Vulcain 3.

In my eyes it is a shame that such a concept is not even given a fair chance to compete against the PPH in a phase A study.

Spacediver

[mmeijeri]

The most interesting result is that all modular concepts, using a common core approach or strap-on-boosters, are not competitive to the “clean” inline designs.

Competitive is in the eye of the beholder. You can tweak your assumptions to get the outcome you want.

The three most promising concepts KH (kerolox first stage with NK-33 engines), HH (cryogenic first stage with Vulcain 3) and PPH (solid first and second stage in-line) have practically the same recurrent cost.

What about fixed vs variable costs?

The small differences are well within the accuracy of our cost estimates.
Therefore, no economic justification for a solid Ariane-6 design can be derived from these results!   

Or against it.

BTW, no concept is able to meet the 70M Euro benchmark for 6,5t to GTO.

Interesting.

Taking into account the loss of crucial technology, for me the clear winner is a HH concept using 3 Vulcain-3 engines in the first stage for the large payloads (up to 6,5 t) and two engines on the same first stage with a 35% propellant offload for the small payloads (up to 3,5 t).

Do you work on liquid engines yourself?

The development cost for the HH concept are in the same range as for the solid concepts, but the development risk is much lower as the Vulcain engines are well understood and the V3 is directly derived from the V2. Also the main stage tank is nothing but a stretched version (with different wall thickness of course) of the existing EPC stage.

Did the study look at a triple Vulcain 2 variant?

[Rugoz]

The most interesting result is that all modular concepts, using a common core approach or strap-on-boosters, are not competitive to the “clean” inline designs.

The three most promising concepts KH (kerolox first stage with NK-33 engines), HH (cryogenic first stage with Vulcain 3) and PPH (solid first and second stage in-line) have practically the same recurrent cost. The small differences are well within the accuracy of our cost estimates.
Therefore, no economic justification for a solid Ariane-6 design can be derived from these results!   

BTW, no concept is able to meet the 70M Euro benchmark for 6,5t to GTO.

Taking into account the loss of crucial technology, for me the clear winner is a HH concept using 3 Vulcain-3 engines in the first stage for the large payloads (up to 6,5 t) and two engines on the same first stage with a 35% propellant offload for the small payloads (up to 3,5 t).

Interesting, can you provide us with a cost breakdown, core stage, vulcains, upper stage?

Have you also looked at the P7C design?

[MikeAtkinson]

spacediver,

did you look at any reusable options?

Although the KH, HH and PPH options were similar in the cost estimates for 6.5 T to GTO, where there significant differences in either payload to LEO or payload for BEO orbit missions?

[sdsds]

results from the NELS study

Thanks very much for providing this overview!

In addition to the questions from others:

for me the clear winner is a HH concept using 3 Vulcain-3 engines in the first stage for the large payloads (up to 6,5 t) and two engines on the same first stage with a 35% propellant offload for the small payloads (up to 3,5 t).

On other threads people have suggested propellant offload approaches like this, and had their ideas shot down by assertions from professionals that first stage propellant offloading just isn't done. (It came up most prominently in the AJAX thread, if I remember correctly. But also in discussions of other cores that might use 3 or 4 engines depending on the mission.) So, is this concept considered to be "low risk" in general? Or were there specific aspects of the core using Vulcain-3 engines that made it possible?

[Patchouli]

To quote Jim: Rockets are not Lego's.

They're kinda already doing that with some of the variants to get commonality with Vega.

One issue with Araine 5 it's now too small for double GTO launches as a typical geostationary satellite today is nearly twice the mass one was back in the 80s but overkill for a single comsat.

The best course would seem to be to try and emulate LVs like Falcon and Angara.