For those who don't know, the way engineers assess risk is with Risk Priority Number spreadsheets. Basically the risk is broken down into 3 parts: the possible Severity of the risk, the Frequency or Occurrence of the risk, and the ability to Detect or Prevent the risk. Each part is given a number between 1 and 10, one being the least and 10 the most. These numbers are then multiplied together, with the final number assessing the risk on a scale of 1-1000. Then risks with the highest numbers are given the highest priority for correction or reduction.
We know that MMOD strikes occur on every spaceflight, so that's also a 10.
Quote from: Hobbes-22 on 08/23/2016 06:53 pmThat's not a very good comparison. The risk of having a lethal car accident is .01%/yr in the US. Not many civilian activities carry a 1% risk of dying. Accidents of all types are 5% of US deaths.
That's not a very good comparison. The risk of having a lethal car accident is .01%/yr in the US. Not many civilian activities carry a 1% risk of dying.
Quote from: whitelancer64 on 08/23/2016 05:34 pmAgain, just because there haven't been catastrophic failures doesn't mean there is no risk.Literally nobody is saying that. Why do you keep repeating it like someone is?QuoteThe overall risk level is still high because a MMOD strike to a critical system could easily cause loss of crew or vehicle.Actually, based on the number of LOV (zero) from MMOD damage on a rather large number of flights, this is provably wrong. Depending on your definition of "high", I guess. Certainly the risk level is high compared to flying in an airplane. Compared to the Shuttle risks not associated with MMOD I'd say they are rather low.
Again, just because there haven't been catastrophic failures doesn't mean there is no risk.
The overall risk level is still high because a MMOD strike to a critical system could easily cause loss of crew or vehicle.
In fact, all Commercial Crew vehicles are docked with their heat shields pointed fore.
Quote from: whitelancer64 on 08/23/2016 06:12 pmFor those who don't know, the way engineers assess risk is with Risk Priority Number spreadsheets. Basically the risk is broken down into 3 parts: the possible Severity of the risk, the Frequency or Occurrence of the risk, and the ability to Detect or Prevent the risk. Each part is given a number between 1 and 10, one being the least and 10 the most. These numbers are then multiplied together, with the final number assessing the risk on a scale of 1-1000. Then risks with the highest numbers are given the highest priority for correction or reduction. Wow, that method is just wrong. The range should be from 0 to 10, with any real value in between (a mathematician would have any value between 0 and 1). A way better estimate of the risk isintegral from 0 to m_max integral from 0 to v_max S(m*v²/2)*A*T*P(m,v)*D(m) dm dvwhereA = cross sectional area of spacecraftT = time in orbitm = mass of debrism_max = maximum debris massv = velocity of debrisv_max = maximum debris velocityS(E) = severity as a function of impact energyP(m,v) = mass and velocity debris probability distribution per unit area and unit timeD(m) = Detect or Prevent as a function of debris massA good engineer should be able to come up with estimates of these functions.QuoteWe know that MMOD strikes occur on every spaceflight, so that's also a 10.That would result in an over estimation of the risk. Most impacts are very small which have almost zero risk.
Let's not forget that Soyuz are docked in aft of ISS, where they are more protected than the USOS fore side. In fact, all Commercial Crew vehicles are docked with their heat shields pointed fore.
Quote from: abaddon on 08/23/2016 08:26 pmQuote from: whitelancer64 on 08/23/2016 05:34 pmAgain, just because there haven't been catastrophic failures doesn't mean there is no risk.Literally nobody is saying that. Why do you keep repeating it like someone is?QuoteThe overall risk level is still high because a MMOD strike to a critical system could easily cause loss of crew or vehicle.Actually, based on the number of LOV (zero) from MMOD damage on a rather large number of flights, this is provably wrong. Depending on your definition of "high", I guess. Certainly the risk level is high compared to flying in an airplane. Compared to the Shuttle risks not associated with MMOD I'd say they are rather low.Regrettably, that's not how statistics work. The fact that Shuttle had the failures that it had, and not MMOD, might have been by pure chance. There's not enough statistical samples in all Shuttle history to say otherwise.Challenger was a particularly bad example, because at those temperatures it was to fail. It was almost a certainty. Edward Tufte book is more than clear on that. You use statistics for things that are chance, but if you get out of the specified range, you might get into straight certainties.
You've never heard of Failure Mode and Effect Analysis (FMEA)?
So how far off are Dragon and CST-100 from the requested 1 in 270 number?
Quote from: baldusi on 08/24/2016 12:20 pmQuote from: abaddon on 08/23/2016 08:26 pmQuote from: whitelancer64 on 08/23/2016 05:34 pmAgain, just because there haven't been catastrophic failures doesn't mean there is no risk.Literally nobody is saying that. Why do you keep repeating it like someone is?QuoteThe overall risk level is still high because a MMOD strike to a critical system could easily cause loss of crew or vehicle.Actually, based on the number of LOV (zero) from MMOD damage on a rather large number of flights, this is provably wrong. Depending on your definition of "high", I guess. Certainly the risk level is high compared to flying in an airplane. Compared to the Shuttle risks not associated with MMOD I'd say they are rather low.Regrettably, that's not how statistics work. The fact that Shuttle had the failures that it had, and not MMOD, might have been by pure chance. There's not enough statistical samples in all Shuttle history to say otherwise.Challenger was a particularly bad example, because at those temperatures it was to fail. It was almost a certainty. Edward Tufte book is more than clear on that. You use statistics for things that are chance, but if you get out of the specified range, you might get into straight certainties.The statistical chance of a failure mode occurring can be very low, however, the amount of risk of that failure mode will remain the same, i.e., if a MMOD strike makes a direct hit on a critical system, then you're always going to have a bad day, even if that MMOD strike on a critical system only occurs once in a thousand spaceflights. That risk level can only be reduced by engineering solutions to reduce it. An example of this is where the Shuttle's radiator got additional MMOD shielding and that directly prevented a MMOD strike from damaging a coolant loop that would have caused a mission abort: research.jsc.nasa.gov/BiennialResearchReport/2011/265-2011-Biennial.pdf
You can't assure safety even while walking across the street. 1/270 is the standard, if I understood the article, because that's what Orion is supposed to have. That the CC contractors apparently haven't met that 1/270 PBRA is causing concern.
The key will be to refine the MMOD threat data, which is based on historical flight information and may be – due to NASA requirements – overly conservative.“The MMOD damage analysis depends on the modeling of the environment, which is in many aspects speculative and quite robust,” added the minutes (from ASAP).“There are discussions regarding gathering additional historical information to determine if the environmental model is perhaps too robust. All answers are yet to be determined.”
Quote from: whitelancer64 on 08/24/2016 03:13 pmYou've never heard of Failure Mode and Effect Analysis (FMEA)?Yes, I've heard of FMEA. If that's how its done with ranges of 1 to 10, then the method is mathematically flawed.
If NASA gets uppity about it, say the astronauts have to remain suited the whole time. That'll get approval done for the early missions, then the astronaut corp would get mad and say they don't have to wear the suits.
Quote from: Robotbeat on 08/25/2016 08:44 pmIf NASA gets uppity about it, say the astronauts have to remain suited the whole time. That'll get approval done for the early missions, then the astronaut corp would get mad and say they don't have to wear the suits.That only protects against LOC from pressure vessel penetration, other factors could also result in LOC that the suits wouldn't help.
Quote from: srtreadgold on 08/25/2016 09:46 pmQuote from: Robotbeat on 08/25/2016 08:44 pmIf NASA gets uppity about it, say the astronauts have to remain suited the whole time. That'll get approval done for the early missions, then the astronaut corp would get mad and say they don't have to wear the suits.That only protects against LOC from pressure vessel penetration, other factors could also result in LOC that the suits wouldn't help.Another major one is loss of coolant. Why did you edit out the sublimator part?pressure vessel penetration is probably the biggest one, since if that happens, there's not a lot you can do if you're not suited up. With the other things, there are options, provided you're in a somewhat stable orbit (which you will be 99.9% of the time)