From: schillin@spock.usc.edu   
      
   "Allen Thomson"  writes:   
      
      
   >John Schilling wrote:   
   >> "Allen Thomson"  writes:   
      
   >> > continuous thrust >= >1 kN, ISP >= 10,000 sec   
   >> > (>= 100 km/sec Ve) and a total run time >= 1 megasecond   
   >> > or greater, preferably getting toward >10 Ms.   
      
   >> they can make you up a cluster of high-power ion thrusters   
   >> that deliver the sort of performance you are looking for,   
   >> without even stretching the state of the art.   
      
   >Could they really? That's excellent news.   
      
   >>  And yes, the ion thruster array will be heavy (kiloton)   
      
   >Apparently you have a specific design, or design in mind.   
   >What, in general, are you thinking about?   
      
   Wasn't thinking anything specific, but since you ask:   
      
   Start with the NSTAR ion thruster developed for NASA's Deep Space 1   
   mission.  Not that it's the best around or the most suited for your   
   hypothetical requirements, but it's unambiguously real and I can   
   talk about it without getting into anything proprietary or classified.   
      
   Weighs 8.3 kilograms, turns 3 mg/sec of Xenon and 2.4 Amps at 640 Volts   
   into 92 mN of thrust at a specific impulse of 3120 seconds.  Lasts at   
   least 12,800 hours, or 4.6 megaseconds.  And requires a 15 kilogram   
   power processing unit to turn 2.3 kilowatts of 28 VDC bus power into   
   the aforementioned 2.4A/640V drive power (and other important stuff   
   besides; it's not just a step-up transformer).  Marginal cost of one   
   thruster/PPU combo would be on the order of a million dollars if you   
   were buying it commercially.   
      
   The specific impulse is less than a third of your requirement, but that   
   is not a fundamental limitation of the technology or even the specific   
   hardware, just a mission optimization - power scales as thrust times   
   Isp, 92 mN at ten thousand seconds would have required more power than   
   Deep Space One had available, and a deliverable 29 mN at ten thousand   
   seconds would not have resulted in enough acceleration to get DS-1 to   
   its destination before the ops budget ran out.  In lab demonstrations   
   unconstrained by the requirements of particular missions, ion thrusters   
   using the same basic technology have delivered Isp values of 8,000   
   seconds, and ten thousand seconds would be no great problem.   
      
   In particular, if you run a stock NSTAR thruster on Argon rather than   
   Xenon, you should get 51 mN at 5,650 seconds Isp.  If we increase the   
   drive voltage to 2kV and decrease the current to 0.75A or so, that   
   gets us 29 mN at 10,000 seconds Isp.  We'll need to modify the   
   thruster to do this, increasing grid spacing, internal high-voltage   
   standoff, insulation thickness, etc.  But as we've reduced the current,   
   we can reduce conductor size as well, and all of these are minor.  As   
   the gross power has remained constant, mass will also remain constant   
   to the first order.   
      
   So, 29 mN thrust, 10000 seconds Isp, and well over a megasecond of life.   
   Two of your three requirements met.   
      
   All we have to do is gang thirty-five thousand of them in parallel, and   
   you've got your kilonewton of thrust as well.  Eight hundred fifteen   
   tons of hardware, plus some integration overhead, rounds to a kiloton.   
      
   In practice, there are economies of scale to be had in using smaller   
   numbers of larger thrusters.  Mostly in cost, but some of the NSTAR   
   components are running up against minimum-gauge limits in mass as   
   well.  OTOH, there are severe difficulties if your individual thrusters   
   wind up larger than can be realistically tested in the largest ground   
   facilities[1].  So something like 800 thrusters at 100 kW each might   
   be appropriate.  And eight power processing units with a hundred output   
   channels each.   
      
   And the eighty-megawatt power supply that we have agreed not to talk   
   about for now...   
      
      
   [1] Unless you can arrange for an advanced propulsion R&D team to take   
   up residence on ISS or one of its successors, of course.  If you *can*   
   arrange for this, may I send you my resume?   
      
      
   --   
   *John Schilling                    * "Anything worth doing,         *   
   *Member:AIAA,NRA,ACLU,SAS,LP       *  is worth doing for money"     *   
   *Chief Scientist & General Partner *    -13th Rule of Acquisition   *   
   *White Elephant Research, LLC      * "There is no substitute        *   
   *schillin@spock.usc.edu            *  for success"                  *   
   *661-718-0955 or 661-275-6795      *    -58th Rule of Acquisition   *   
      
   --- SoupGate-Win32 v1.05   
    * Origin: you cannot sedate... all the things you hate (1:229/2)