1. other similar class vehicles is hardly an abnormal type of test.2. They don't want it to be expendable, they want to recover the first stage. 3. And even if it has been hot fired, there were tweaks to their system to eliminate roll off the pad. Perhaps they want to test pad loads or something. Also they had a problem previously with their test stand not having the exact same configuration as the pad. What's wrong with being extra careful knowing there is precedent for issues there?
Static fires are not for software testing. They are only for propulsion system testing
+19 EELVs with only 1 pad test firings+100 Delta II's with none
+1 Delta IV H that failed due to software on first flight
And yes, static fires do provide info for the GNC folks, especially on Falcon 9 where guidance is through differential thrust (unlike Delta or Atlas)...
Quote from: simonbp on 09/18/2010 11:02 pm+1 Delta IV H that failed due to software on first flightReally?
Quote from: Jim on 09/18/2010 03:38 am+19 EELVs with only 1 pad test firings+100 Delta II's with none+1 Delta IV H that failed due to software on first flight+1 Delta III that failed due to software on first flight+1 Ariane V that failed due to software on first flightNotice a trend? And yes, static fires do provide info for the GNC folks, especially on Falcon 9 where guidance is through differential thrust (unlike Delta or Atlas)...
Yes it was a software commanded shutdown, but it had a real hardware based cause.
Quote from: Jim on 09/18/2010 03:24 amStatic fires are not for software testing. They are only for propulsion system testingEnd to end is further than you think <-- software testing maxim.Not saying that is the reason in this case.
flight software is not tested in static firing, the vehicle is not moving
Quote from: Jim on 09/19/2010 03:51 amflight software is not tested in static firing, the vehicle is not movingI don't understand the logic. Obviously the flight software has been thoroughly and continually tested since long before the static test (and if they're doing it right even before it was written). And obviously there any many important areas that don't get tested with a static test. But every test helps. The vehicle still shakes, sensors are measuring things, valves move. Things could possibly go wrong at this stage. And those are the things you test for. Again, I'm not saying this is their reason and I'm not saying other companies are doing it wrong. I'm just saying that as a software guy I would consider such a test valuable. Depending on the cost of such a test, which I don't know, it may or may not be worth it.
I. non flight software is usually use for static tests.2. Most sensors on a vehicle are not used in the guidance (flight) software, the data is just telemetered to the ground.3. There is very little of the guidance software that deals with the propulsion system. Other than thrust chamber ok switches/chamber pressures and occasionally propellant utilization, there is little interaction of the flight software with the rest of the vehicle. 4. Guidance (flight) software steers the vehicle and controls events. Most vehicles have separate engine controllers that operate the engine. The guidance system only issues start, shutdown and throttle level commands. The engine controller does all the other necessary computations for valve movements.
Thanks for that detailed information! About the software for the engine controller: in a vertically integrated company like SpaceX, is there a particular reason for the engine controller to be physically located on the engine? Couldn't the controller be run on the same processor as the flight software? It's software after all.