From Readware:
For those who fancy satellite viewing, as certain people do, they know that the Zenit rockets in flight after payload execution have provided moments of exceptional flashing brightness that can lead the poet to mistake it for the evening star blinking as it gazes on the earth. Is it because the hull is made of aluminum, or because of fuel leakage, and does one need elsets to find where the rocket is orbiting? Or is it as with the little boy, in his
huck-a-buck world: he puts down his jacks and ball, and turns his eyes heavenward. he looks through the wisps-- can he see what we have been told only telescopes privilege the eyes to see?
***
We attended the Launch Systems Design and Engineering conference, conducted by Dr. Kaplan, 9-11 March of this year, in Washington, D.C. In attendance also was Robin World. Only now have we been able to turn our attention to the findings there as Globalstar's second launch has executed successfully. Lectures on stress analysis, thermal systems design, determination of aerodynamic properties, heating, avionics, ascent guidance subsystems, internal and
GPS-assisted navigation systems, ignition ordinance systems, new developments in propulsion systems, ascent equation dynamics (both max Q and alpha Q), aerodynamic drag on fuel, and. to the investor, most important of all, launch site and LV (launch vehicle) selections, were addressed.
The findings, we believe, become important now only because the Zenit-2 12 sat Globalstar launch is the first of its kind in history. Why 12 sats all at once? What of this Zenit-2? Has it not been shrouded in secrecy for so many years? Is it an attempt (recklessly) by Globalstar management at catch-up? Could there not have been more up-to-now conventional launches-- perhaps more Delta launches, perhaps even the Proton instead of Zenit-2?
Launches fail for a variety of reasons, though the failure rate has declined dramatically in the past 4 years. That is undeniable. We will look at Ariane 5 as an example of failure, and then turn to Zenit-2. We choose Ariane since it is the most influential and largest commercial launch system on the globe.
Ariane 1 had developed over time seamlessly into Ariane 4: both had three stages, the bottom two burning solid fuel, and the third liquid hydrogen. Four boosters at their sides allowed propulsion of 4 tons of geostationary weight into orbit 22,000 miles up. Also, it had ten engines-- eight on the first stage, including the boosters, one on the next two stages each.
Ariane 5 was to be different. It would have only three engines on the lower composite, and a simple one on the upper stage. But there was a brand new engine-- the Vulcain. Ariane 5 would be powered by a main stage that used liquid oxygen to burn liquid hydrogen-- the fuel that, pound-for-pound, provides the most thrust. Chilled to minus 418 degrees F, the hydrogen would provide the effcient burn needed to raise the payload to orbital speed. A pair of
solid-fuel boosters (280 tons each) would get the launcher off the ground and through the atmosphere to an altitude of 36 miles in 150 seconds, and then falling away into the Atlantic off the Guyana coast, relieving Ariane of the weight.
Ariane had had troubles-- out of 86 launches prior to Ariane 5, seven had failed. Always the same problem-- engine difficulty-- usually the one on the cryogemic third stage: a turbo pump would fail to deliver the oxygen to the combustion chamber, or a valve would leak.
But Ariane 5 would be different. Ariane 1 had cables and wires running from ground to the pad and each individual circuit board in equipment bay. For 5, there was simply a single fiber-optic cable connecting the ground to an on-board components network, each with its own microporocessor and embedded software, and each under the control of the main on-board computer. Ariane 1's computer had 4 kilobytes of RAM. 5 had 400,000 kilobytes, enabling it to
carry on a thousand more calculations.
The entire flight control system would be redundant-- two or three of everything: two intertial reference systems (IRS) to calculate the rocket's position. The IRS permits the rocket to move from A to B without referring to the outside world at all. It calculates changes in the rocket's position by simply measuring its acceleration along three axes with accelerometers, and by measuring changes in its attiude (i.e., "heading, pitch, and roll") with
gyroscopes. To do that, though, the system must first align those internal axes with respect to Earth's: it must find north-south, up-down, and then fix the locations of A and B in the internal coordinates. the alignment module is tasked to do that.
Other redundancies were two on-board mainframes to compare the rocket's position with the planned trajectory and to send the approproate commands to the control electronics, and two sets of control electronics to swivel the engine nozzles and steer the rocket. Any problem with the first sytem would default immediately to the back-up, and then "loop out" of the problem.
That is what was supposed to happen.
Instead a multi-billion dollar tragedy occurred. There was a flaw in the IRS logic. IRS calculations of all motions must be updated uninterruptedly. The alignment module instructs the navigation system at every ingestion of binay datum as to what is next to occur. If all motions are finally acceptable, the navigaion system is instructed to enter "flight mode". At any moment the IRC can cut off those instructions. And the countdown begins again.
What happened in Ariane 5 is that a duplicate of the background IRS instructions never shut off. A ground computer had started to instruct the module to halt navigation initiation. But the background IRS-- the redundant one-- had told the navigation system to continue. Suddenly, horizontal velocity variables, thrust indications, propulsion parameters-- all their measurements became binary nonsense to the navigation system. Internal reference (IRS)
failed. The rocket launched with compromised trajectory as it began its ascent. Billions of dollars suddenly were lost.
Like Ariane 5, Zenit-2 is one of the most powerful LVs ever designed. It traces, through the shrouded N-1 project, back to the R-2 rocket of post WWII, as the successor to the German R-10* in the guise of the Russian V-2. Moving rapidly through Sputnik successes brought the R-12, renamed the Cosmos launch vehicle. It was the first to carry one megaton of nuclear payload. The target was Washington D.C.-- till 1987, then on the R-14.
With the end of the Cold War, the Zenit launch system (in the Tsyklon family) entered non-military application (with only one known classified reconnaisance payload) lifting into orbit the Tselina system (on the Cosmos rockets). The 15th, 16th, 17th, and 28th launch failed.
Since the last launch, Zenit-2 has been completely redesigned. It is because Zenit is no longer under military control, but now is an alliance of Boeing, Saab-Erricson AB, Kvaener a.s. of Norway, Korolev Rocket & Space of Russia, and NPO Yuzhnoye (the Ukraine). What has made the change most pronounced is the commitment to Sea-Launch, the fourth generation launch vehicle. That will bring about Zenit-3. Prior to 3, though, must be 2. We may note here
that Zenit 3 SL has incorporated the Proton LV, thus bringing Lockheed Martin into the effort of the new Zenit-2. Proton's overall reliability, prior to the Lockheed affiliation, in 200 launches is 96%, a posssible testament to Russian launch know-how.
Zenit-2 maintians its payload lift capability of nuclear tonnage, and thus is an ample vehicle for lifting ito orbit the weight of 12 Globalstar LEOs. The total lift weight is approximately, together, 8,700 pounds. Zenit-2 overall design was meant to lift up to between 18,000 and 19,000 pounds. On the Zenit system is the RD-171 engine, the result of a decade of Cold War research by Energomash. RD-171 provides 1.6 million pounds of initial trust in a
clean burn of liquid oxygen and kerosene to deliver a LEO payload at 51.6 -52 degrees inclination. The separation mechanism was designed for the new Zenit-2 by Saab-Erricson. It has 130 successful separations to date, with no fails. Zenit-2 employs a variant of the Lockheed DM block-- that of the Proton, which has been in one fail (AsiaSat) in the last three years. The diameter payload variant was constructed by Boeing, whose launch success is 98%.
On Delta Globalstar's initial launch azimuth (for safety reasons) was 65 degrees. The solids separated 67 seconds into mission, with dog-leg (left turn) to the desired azimuth 44.7 degrees, with final orbit at 52 degrees inclination. Deployment altitude was 1247 kilometeres circular. The satellite's on-board propulsion system raised the orbits to the operational altitude of 1414 kilometers.
Zenit-2 will launch 12 Globalstar LEOs at a time 900 kilometers. It will do so because new system software and design indicates a high likelihood of achieving such. As the orbits naturally precess, four LEOs at a time will be raised to their operational altitudes-- the quivalent of a multi-plane launch vehicle without the need of a super-booster or multiple launches. Dog-leg to the 44th degree is effected by redundancy software designed by Boeing and
all subsequent maneuvers are controlled by Space Systems Loral TT&C at the Cosmodrome and from Palo Alto.
What can go wrong? We saw in Ariane 5 that the logic of the IRS was corrupted by a failure to override duplication. So odd-- since most faults are in propulsion, not software. Ariane 5 resulted in a catastrophe. We remember the day of 4 June when it occurred.
Much has been learned since that time. Since the last Zenit mishap Boeing and others have netered its sphere. Zenit is effectively a new enterprise, an international effort along the lines of Lockheed-Intersputnik and Lockheed-Energeia. The rocket that was meant to deliver thermonuclear horror to the United States has only the thrust capabilities of those days. The software avionics, navigational reference systems, balance and thruster management,
design metallics, guidance system, RAM capability, instruction override, engineering personnel, pneumatics-- they are new or emended.
The verdict will be in 15 July 1998 as Globalstar launches 12 new LE0s. As testing continues in the Cosmodrome (a Zenit launch was rescheduled just last week for further testing), and with the arrival of SSL engineers at the end of May at the Cosmodrome, we are reasonably confident of success. Those companies involved in Zenit-2 represent combined almost 75 years of launch experience. We believe Globalstar management will permit no launch till every
certitude has been exhausted. None of the aerospace entities involved would have it any other way.
End post. |
