Friday, March 29, 2013

The Coming SSTO's: multi-Vulcain Ariane.

Copyright 2013 Robert Clark

 The option the ESA decided on for the planned Ariane 6 was the version using a solid propellant first stage:

CNES, ASI Favor Solid-Rocket Design For Ariane 6.

By Amy Svitak
Source: Aviation Week & Space Technology
October 15, 2010

 However,  one of the other options discussed for the Ariane 6 would also allow manned European flight capability. This would be the two Vulcain core version.

CNES is evaluating these three launch vehicle concepts for a next-generation Ariane 6: two based on solid-rocket-motor technology plus an all-liquid-fueled launcher with optional solid-motor boosters. (Credit: CNES)

 To estimate the payload capability for the twin Vulcain core I'll use John Schillings launch performance calculator:

Launch Vehicle Performance Calculator.

 In the calculations for this multi-Vulcain Ariane core stage, I used this page for the specifications on the Ariane:

Space Launch Report:  Ariane 5 Data Sheet.

 For the Vulcain 2 specifications, I've seen different numbers in different sources, though close to each other. I'll use this source:

 I'll also use the earlier Ariane 5 "G" version that is lighter than the current "E" version to be lofted by two Vulcains without side boosters. According to the SpaceLaunchReport page it had a 170 mT gross mass for the core at a 158 mT propellant load, giving a 12 mT dry mass.
 According to the Astronautix page, Vulcain 2 has a 434 s vacuum Isp and 1350 kN vacuum thrust. So two will have a 2700 kN vacuum thrust. The Vulcain's mass is listed as 1,800 kg. So adding another will bring the stage dry mass to 13,800 kg.
Now input this data into Schilling's calculator. Select again default residuals and select "No" for the "Restartable Upper Stage?" option. Select the Kourou launch site for this Ariane 5 core rocket. For the orbital inclination, I input 5.2 degrees. I gather Schilling uses this for Kourou's latitude since deviating from this decreases the payload. I chose also direct ascent for the trajectory.
Then the result I got was 7,456 kg(!) to orbit:

Mission Performance:
Launch Vehicle:     User-Defined Launch Vehicle
Launch Site:     Guiana Space Center (Kourou)
Destination Orbit:      185 x 185 km, 5 deg
Estimated Payload:      7456 kg
95% Confidence Interval:      4528 - 10898 kg

 We should be able to remove a component on the Ariane 5 core to lighten the weight for this application. Ed Kyle on his page discusses the Liberty rocket that had been planned to use a SRB first stage and an Ariane 5 core second stage. For the Liberty application, a forward skirt on the core called the JAVE ("Jupe AVant Equipée") that transmits the forces of the two solid boosters to the core would be removed. This will also be removed for our application without solid boosters.
 The JAVE massed 1,700 kg. So our payload could be increased to 9,156 kg. However, Kyle also discusses on his page on the Liberty rocket that the increased thrust from the SRB first stage would require thicker walls on the Ariane core now used as an upper stage.
 The thicker walls on the Ariane 5 core for the Liberty rocket are indicated in this video:

 The 5-segment SRB to be used on the Liberty rocket has 12 times the thrust of the Vulcain engine. Yet as seen in the video the increased thickness to handle the increased axial load is only 50%. Then only doubling the thrust by adding a second Vulcain quite likely will require a much smaller increase in thickness. I'm informed that the 158 mT propellant mass tank has a dry weight in the range of 4,400 kg. So even increasing the thickness 50% increases the weight by ca. 2,200 kg, and the payload would still be approx. 7,000 kg.  A problem with this estimate though, aside from the unknown accuracy of the video, is that it is based on the larger Evolution "E" version of the Ariane 5 core, which might not require as much strengthening to handle the higher thrust loads as the smaller "G" version.  So we'll use a formula for calculating the thickness of a propellant tank based on the axial load as given on this lecture page: Launch Vehicle Design: Configurations and Structures. Space System Design, MAE 342, Princeton University Robert Stengel  on page 9:
  From the first formula the critical buckling load without the pressurization effect is:   σc,w/o pressure = [9(t/R)1.6 + 0.16(t/L)1.3]*E  Multiplying out the second formula for critical buckling with the pressurization effect you see it's:   σc,w/ pressure = σc,w/o pressure + 0.191p(R/t).   Now use the formula on p. 8 that relates the tensile strength of the material to the thickness required of a pressurized tank:
   You see that  σhoop = p(R/t)  so that the formula above becomes:     σc,w/ pressure = σc,w/o pressure + 0.191σhoop  Now use values for the tensile strength of aluminum alloy. The aluminum alloy used on the Ariane 5 core tanks, Al 2219, happens to get stronger at cryogenic temperatures:
 Table taken from Properties of Aluminum Alloys: Tensile, Creep, and Fatigue Data at High and Low Temperatures, page 86. The table gives the aluminum alloy strength at liquid hydrogen temperatures as 685 MPa and elasticity modulus, E, as 85 GPa.  For the Ariane 5 core "G" version, the hydrogen tank walls are only 1.3 mm thick, while the oxygen's, 4.7 mm. The diameter of the tanks is 5.4 m. Because of its extreme wall thinness it's the hydrogen tank whose stress has to be limited. It's length is about 18 m. Then the formula for the critical buckling load without pressurization gives: σc,w/o pressure = [9(t/R)1.6 + 0.16(t/L)1.3]*E = [9(0.0013/2.7)1.6 + 0.16(0.0013/18)1.3]*85*109 = 3,800,000 Pa.  And the additional buckling strength due to pressurization is 0.191σhoop = 0.191*685,000,000 = 130,800,000 Pa, for a total critical buckling load of 134,600,000 Pa.  The maximum thrust of two Vulcain 2's will be 2,700,000 N. The cross-sectional area of the hydrogen tank walls is 2*π*R*t = 2(3.14)(2.7)(0.0013) = 0.022 m2 . Then the maximum axial pressure is 2,700,000/0.022 = 123,000,000 Pa.  This is indeed less than the critical buckling load of 134.6 MPa. However, for a manned launcher a safety factor of 1.4 is usually included. This will require the maximum axial pressure to be less than 96 MPa. This requires a wall thickness of 1.6 mm, about a 25% increase. This still only increases the tank weight by 1,000 kg, so the payload becomes now ca. 8,000 kg, still quite high for a SSTO. Remember also switching to aluminum-lithium alloy can save as much as 25% off the dry weight which would bring us again to the 9,000 kg payload range.

   Bob Clark


  1. What usually sets case thickness on a solid is max expected operating pressure (as hoop stress). This is especially true in larger sizes like SRB's, but largely true even in the far smaller sizes of tactical rockets.

    Man-rating with a solid merely requires a lot of tests with no problems experienced, something the 5-segment SRB/Liberty was not honestly able to claim. That "thrust oscillation" they "solved" with a damper did not fool me: that was a fundamental solid motor longitudinal-mode combustion instability, induced by the extra length of the 5-segment motor, with otherwise the same 4-segment solid propellant grain design.

    Instabilities like that tend to blow up solid motors, sooner or later. Better sooner than later, but it didn't happen (yet) in the ground tests of the 5 segment motor.

    Interesting, is it not, that the motor NASA actually did fly was 4 live segments and one inert dummy segment ??!!?? Wanna guess why?

    Solid motor combustion instability is something the liquid guys at NASA (and ULA) know nothing about: their liquid-engine combustion instabilities are quite different physics entirely.

    NASA has no long-experienced solid motor guys. And they never have. Those guys were at Thiokol, UTC, ATK, etc. Still are, if not dead or retired. Or working in some other industry, like me.

    The 4-segment SRB on Space Shuttle was indeed "man-rated", in spite of what I always considered to be a defective segment joint design mandated by a NASA not qualified to judge the merits of the design. They simply made it even worse after the Challenger accident, with that idiotic 3-O-ring design.

    Even the 2-O-ring design before Challenger was idiotic. And I really do mean idiotic. Idiotic! Truly idiotic!!!! It really was a very egregious mis-design mandated by idiots who very clearly knew no better.

    The only reason it (that awful 3-O-ring "design") flew successfully for all those flights after Challenger was that "they never again flew cold" (just as Chuck Yeager recommended to the Rogers Commission in its very first meeting), not because of the egregiously-mis-designed post-Challenger 3-O-ring SRB segment joint.

    NASA still (to this very day) does not know anything truly useful about solid rocket design. I do, I worked in a for-profit solid rocket company for 16 years.

    ESA knows incredibly-little about solid rockets, basically they know only what NASA has told them. NASA? Really?

    So, who would you believe? NASA? Or someone like me, who did it for a living in a for-profit company?


  2. Thanks for the info. I am not a fan of solids because I feel they do not advance the technology.
    Ground tests have been done of the 5-segments. I wonder what the thrust oscillations look like in those tests.

    Bob Clark

  3. Well, as near as I can tell, from what little was publicly released, the thrust oscillations were very bad with all the 5 segment motors. There was talk of lethality to an astronaut, until they added a big damping mass (a payload penalty). The real solution to combustion instability in a solid is a redesign of the propellant grain and internal volume shape in the motor. They didn't do that. Due diligence would say they should have, but pride rarely admits to an error.

    Solids have good application about the first half of a typical 1st stage burn of 2 stages. That's where thrust is a whole lot more important than high Isp. The rapid weight burnoff is also very important to vehicle acceleration in that early flight. Plus, there's also higher thrust per unit stage blockage area available, in spite of the much-higher inert weight, partly due to the internal-burning grain design, and partly due to the heavier molecular-weight gas properties.

    The best (and safest) are aluminized HTPB-AP propellants, with burn rates that fall in the 0.1 to 1 inch/sec range at 1000 psia, and Isp's in the 255-260 sec range in very large sizes. Shuttle SRB's used that stuff. So did we in the tactical missile business. That technology hasn't changed much since the 1960's. It's very mature.

    The longer 5-segment design simply brought into resonance a longer bore space to interact with the natural variability of solid surface burning. It's most likely a first longitudinal mode thing, although other, higher modes have been known.

    It was probably greatly exacerbated by the erosive-burning effects of a higher mass flux at the exit to the longer grain design. I've seen that many times before. The first thing to have tried would have been a larger (or better-yet, tapered) bore in the aft segment. But, they didn't do that.


    1. Thanks for that. I had some questions on using solids in some other applications I want to run by you that I'll discuss in an email.

      Bob Clark

  4. Research has already been done on replacing the current Al 2219 aluminum alloy of the Ariane 5 core with Al 2195, a lighter, but stronger, aluminum-lithium alloy:

    NASA Uses Twin Processes to Develop New Tank Dome Technology.
    Spherical tank dome combines friction stir welding and spun formation.
    A full-scale spherical tank dome measuring 18 feet in diameter was produced from high-strength 2195 aluminum-lithium using twin manufacturing processes. Image credit: MT Aerospace
    NASA has partnered with Lockheed Martin Space Systems in Denver, Colo., and MT Aerospace in Augsburg, Germany, to successfully manufacture the first full-scale friction stir welded and spun formed tank dome designed for use in large liquid propellant tanks.
    The NASA and Lockheed Martin team traveled to Germany to witness the first successful aerospace application of two separate manufacturing processes: friction stir welding, a solid-state joining process, and spin forming, a metal working process used to form symmetric parts.
    The twin processes were used by MT Aerospace to produce an 18-foot-diameter tank dome using high-strength 2195 aluminum-lithium. The diameter of this development dome matches the tank dimensions of the upper stage of the ARES I launch vehicle under development by NASA, as well as the central stage of the European Ariane V launcher.
    "This new manufacturing technology allows us to use a thinner, high-strength alloy that will reduce the weight of future liquid propellant tanks by 25 percent, compared to current tank designs that use a lower-strength aluminum alloy that weighs more," said Louis Lollar, project lead for the Friction Stir Weld Spun Form Dome Project at NASA's Marshall Space Flight Center in Huntsville, Ala.[/quote]

    This could bring the payload of this SSTO into the 9,000 kg range.

    Bob Clark