NASAWatch posted an interesting presentation today on propellant depot risk mitigation done at JSC last year: http://nasawatch.com/archives/2012/04/nasa-still-thin.htmlI find it interesting to note that the ISS data supports a catastrophic AR&D (autonomous rendezvous and docking) failure rate about 1000x lower probability than was used in HEFT or ESAS when analyzing propellant depot risks. I figured there was something wrong with assuming that AR&D was 10,000x riskier than an aircraft carrier landing...~Jon
Quote from: jongoff on 04/23/2012 06:01 pmNASAWatch posted an interesting presentation today on propellant depot risk mitigation done at JSC last year: http://nasawatch.com/archives/2012/04/nasa-still-thin.htmlI find it interesting to note that the ISS data supports a catastrophic AR&D (autonomous rendezvous and docking) failure rate about 1000x lower probability than was used in HEFT or ESAS when analyzing propellant depot risks. I figured there was something wrong with assuming that AR&D was 10,000x riskier than an aircraft carrier landing...~Jon3 orders of magnitude difference, how do they sleep at night
Who can now say that NASA didn't have fingers on the scale for those studies?
Hey JongoffI'm a physicist, and three orders of magnitude count over here big, and I mean being off by three orders gets one in big trouble
Quote from: BrightLight on 04/23/2012 10:26 pmHey JongoffI'm a physicist, and three orders of magnitude count over here big, and I mean being off by three orders gets one in big trouble I can't remember who said it (and if he was being tongue-in-cheek or serious), but I do remember the "wow, I'm apparently in the wrong profession" reaction he got.~Jon
Saturday I asked Jim Sensenbrenner if the House Science Committee, which he is a member of, ever got the fuel depot report. He said no.
"This {JSC depot} study does not support the perception that depots add an unacceptable amount of risk and and should not be considered due to an increased number of launches, AR&Ds, and transfers"
ESAS Risk pg 605o "Mission modes requiring three launches were eventually eliminated from consideration due to their cost and reliability issues (i.e., multiple launches, AR&D)."pg 575o "The final lunar mission architecture selected does not require AR&D. Pressurized cargo delivery to the ISS will require some level of AR&D; however, ISS crew will be available to provide backup capability. Other lunar missions that were considered did use AR&D. In many of these missions, the risk presented from AR&D was a driver"
Quote from: Eric Hedman on 04/23/2012 09:39 pmSaturday I asked Jim Sensenbrenner if the House Science Committee, which he is a member of, ever got the fuel depot report. He said no.Why don't they just subpoena what they want?
I'm sure they'd subpoena what they *want*. Why would they want this?
Quote from: QuantumG on 04/24/2012 12:04 amQuote from: Eric Hedman on 04/23/2012 09:39 pmSaturday I asked Jim Sensenbrenner if the House Science Committee, which he is a member of, ever got the fuel depot report. He said no.Why don't they just subpoena what they want?I'm sure they'd subpoena what they *want*. Why would they want this? It would further undermine the case for a giant pork fueled rocket. Congressional support for NASA efforts is not normally driven by an altruistic desire to yield the maximum amount of space and science results for a given amount of funding, it is driven by a desire to route funds to specific destinations. The funding destinations involving a giant pork fueled rocket are apparently much higher on the priority list than those involving depot based architectures.
I find it interesting to note that the ISS data supports a catastrophic AR&D (autonomous rendezvous and docking) failure rate about 1000x lower probability than was used in HEFT or ESAS when analyzing propellant depot risks. I figured there was something wrong with assuming that AR&D was 10,000x riskier than an aircraft carrier landing...
And I think something of that is going on with these numbers. There are so few cases of AR&D that you really can't put a statistically significant number on them. So, the one-sigma probability of failure probably ranges several orders of magnitude.
BTW 2 collisions out of 138 is 1.5%.