GSLV-D3 / GSAT-4 launch - April 15, 2010

• #80 by ugordan on 18 Apr, 2010 19:12
• Interesting. The pump did start but failed immediately. Wouldn't that suggest the failure mode could be reproduced without a vacuum chamber?
• #81 by Art LeBrun on 18 Apr, 2010 19:15
• Possible broken output shaft or gearbox failure............other scenarios?
• #82 by avollhar on 18 Apr, 2010 19:32
• Quote from: Art LeBrun on 18 Apr, 2010 15:57

  Since this was still a high velocity flight why do you need to settle fully loaded cryogenic propellants with ullage pressures in place?
  

  What has velocity to do with it? It's about acceleration and the lack of it (or gravity) means that there is a possibility of gaseous H2/O2 at the turbopump inlet rather than LH2/LOX.. which could result in the 'burp'.
  
  Even when fully loaded you have some gas reservoir to pressurize the tanks.
• #83 by Art LeBrun on 18 Apr, 2010 19:49
• My point is that tanks are pressurized when fully loaded which helps to seat the propellants gravity or not. I will concede that gaseous condition could occur in a feedline. After a first burn then a gaseous condition will occur.
• #84 by ugordan on 18 Apr, 2010 19:53
• Quote from: Art LeBrun on 18 Apr, 2010 19:49

  My point is that tanks are pressurized when fully loaded which helps to seat the propellants gravity or not.
  

  
  Pressurization itself does not help seat the propellants. There is still nothing forcing the ullage to remain at the top of the tank in zero G. Most vehicles today however can live with that because the ignition of the stage comes soon after cutoff of previous stage so the gas bubble does not have time to travel to the feedline inlet at the tank bottom.
  
  As to the failure in question, I wonder if cavitation could be an issue? On the ground they have the increased pump inlet pressure due to weight of the propellant, that would be missing in flight. Seems like a silly mistake to make, though.
• #85 by input~2 on 18 Apr, 2010 20:11
• Screenshot from televised video at T₀+505s when telemetry link was lost (curves and scales have been colored for better readability)
  
  - Nominal altitude vs time in green
  - Nominal relative velocity vs time in magenta
  - Observed deviations in red
  GS2 IGN = second stage ignition
  CUS IGN = third stage ignition
  CUS OFF = third stage cut-off
  
  
• #86 by tonthomas on 18 Apr, 2010 20:17
• So what happened between 290.4 s (shutdown of second stage) and 304.9 s (third stage main engine(?) ignition)?
  
  The turbine that spins the fuel pump may have failed (said in already mentioned http://beta.thehindu.com/sci-tech/technology/article402907.ece) - because of overspeed from a dry pump, other (mechanical) causes, or i.e. there was no gas to drive the turbine longer...
• #87 by seshagirib on 19 Apr, 2010 05:26
• My understanding of the cryogenic engine is H and O2 pumps are driven by the turbine, but how are the pumps driven before  ignition? battery driven?
• #88 by s^3 on 19 Apr, 2010 11:16
• Quote from: input~2 on 18 Apr, 2010 20:11

  Screenshot from televised video at T₀+505s when telemetry link was lost (curves and scales have been colored for better readability)
  
  - Nominal altitude vs time in green
  - Nominal relative velocity vs time in magenta
  - Observed deviations in red
  GS2 IGN = second stage ignition
  CUS IGN = third stage ignition
  CUS OFF = third stage cut-off
  
  
  

  
  Using the data given in earlier post
  
  [290.4   GS2
  293   GS2
  297   4.898
  298   4.898
  299   4.898
  300   4.898
  301   4.897
  304.9   CUS
  308   4.895
  309   4.895
  310   4.895
  316   4.893
  317   4.893
  318   4.893
  319   4.893
  320   4.893
  321   4.893
  322   4.893
  323   4.893
  324   4.892
  325   4.892
  326   4.892
  327   4.892
  328   4.892
  329   4.892
  373   4.899
  379   4.901
  446.5   4.951
  505   5.023]
  
  I plotted the graph and added trendlines.
  
  Velocity is continuously falling from 293 ( GS2 Shutoff/seperation ) to 329 seconds indicating that there is no force operating to increase the velocity ( Velocity falls from 4.898 to 4.892 ) .
  
  After that the kinetic energy attained by previous ignitions is finished and the free fall starts increasing the freefall velocity to 4.892 to 4.899... from 329 to 373... onwards upto 5.203 at 505 seconds.
  
  So.. even if the cryo ignited it did not impart any force towards increasing the velocity of the rocket.
• #89 by input~2 on 19 Apr, 2010 14:37
• Quote from: s^3 on 19 Apr, 2010 11:16

  So.. even if the cryo ignited it did not impart any force towards increasing the velocity of the rocket.
  

  I tend to agree. This is illustrated below when zooming in from my earlier graph on the period around CUS IGN.
  
  Edit: I have added trendlines
• #90 by sanman on 19 Apr, 2010 17:24
• Well, against the claim of ~1 sec burn of cryo engine, it's rather hard to tell. You'd have to superimpose a linear plot against a ballistic curve, and for such a short interval relative to the data points available, it's hard to interpolate whether or not some slight acceleration was imparted.
  
  If ISRO can release more detailed flight telemetry data, then more accurate analysis would be possible. Otherwise, it's hard to discern the difference between ~1sec burn and non-ignition based on the data points we now have available.
  
  I do wish that ISRO would also have cameras aboard their launch vehicles, as these can provide important information on the flight, as well as serving as a valuable historical record of a milestone achievement in the country's space progress.
• #91 by input~2 on 28 Apr, 2010 09:46
• Quote from: ISRO

  Indian launches are never insured

  (source)
• #92 by input~2 on 05 May, 2010 20:19
• An interesting article on the failure just published by the Indian Front Line magazine "Cryogenic setback"
• #93 by kanaka on 06 May, 2010 05:37
• Good Article. Can't we have real time simulators for simulating the flight test before going for actual test? This would save loss of vehicle or satellite. ISRO should have gone with a dummy satellite (rocks even !!!   of equal weight instead of trying with original satellite while doing experiemnts. This saves cost in case of failuare.
• #94 by ugordan on 06 May, 2010 09:07
• Quote from: kanaka on 06 May, 2010 05:37

  ISRO should have gone with a dummy satellite (rocks even !!!   of equal weight instead of trying with original satellite while doing experiemnts. This saves cost in case of failuare.
  

  
  And if you're launching with the logic assumption the flight will go well (otherwise why do you feel you're ready for the launch?) and it does go well, you just wasted one working vehicle on rocks. Meanwhile, there are countless payloads sitting on the ground that would take any chance they can get just to get a ride to space, no matter how risky.
• #95 by kanaka on 06 May, 2010 09:20
• satellite is more expensive than working vehicle. simulators with real time parameters are better for simulated flight testing before going for actual test
• #96 by ugordan on 06 May, 2010 10:49
• Why then do you think anyone puts a paying customer on any inaugural flight?
• #97 by input~2 on 06 May, 2010 12:59
• GSLV failure analysis report expected in mid June (source)
• #98 by seshagirib on 29 Jun, 2010 16:55
• http://www.dnaindia.com/india/report_geosynchronous-satellite-relaunch-with-indigenous-cryogenic-engine-in-one-yr-isro_1402900
  
  Looks like ISRO made some progress in the Failure Analysis:
  
  "We have come across a few scenarios after detailed analysis of the failure. Now the immediate task is to test it on the ground and we look forward to relaunch it next year," ISRO chairman K Radhakrishnan told reporters......