Orion Abort Test Completed
Test complete! Today’s ~3 minute test of @NASA_Orion spacecraft's launch abort system verified that @NASA_Astronauts can safely get away from their launch vehicle in the case of an emergency after liftoff. More: https://t.co/mUf6z9uYjA pic.twitter.com/nI9CZKT6H9
— NASA (@NASA) July 2, 2019
Just seemed odd not seeing the test capsule come down by parachute. I know they were testing the abort system and have tested the parachutes previously, and apparently the capsule was not built for spaceflight so no need to recover and study I guess. It just seems like passing up a rare opportunity to do a parachute test in a dynamic environment like this, and also wasn’t there anything that the engineers might want to look at on the capsule? Program manager Mark Kirasich said they wanted to keep the test as simple as possible so that they could hurry up and get the data for Artemis 1 next year. This thing has been going on for over a decade and now they are rushing. They actually had to spend some time designing the test capsule to break apart on water impact so that as many pieces as possible would sink. That’s up there with redesigning RS-25 engines to be non-reusable.
Probably just a case of saving money since the simulations show that the attitude control system and parachutes would work. I noticed NASA also learned from Ares I and panned the camera away before the capsule shattering impact on the water. ?
Saving money? Reminder this is Orion we are talking about.
I guess that even a pork program like Orion has it’s limits. I recall reading somewhere that each Orion capsule will cost around $1.5 billion dollars, so perhaps that is why they decided to just use a boiler plate and trust the simulation. Hopefully someone who follows the program closer will have some better figures to offer.
I suspect the first Orion would be 1.5 Billion I imagine they must get cheaper the more you build.
Saving Money truly , a parachute cost what 50K , the test vehicle was probably 50 Million , it is very weird and if it was spacex they would insist on having the chutes !
I thought it was odd too. What if something had happened with the data collection and all data was lost? They’d have to build a new test Orion just to try the abort test again.
They could make the telemetry redundant, with multiple antennas, transmitters and receivers. So that probably wasn’t a concern. But I’m also one of the people who thinks you can never have enough housekeeping data to identify and resolve problems. If you’re getting a voltage samples ten times a second, and everything looks file, I’d be wondering if something flaky could be going on between those samples. If it were possible, I’d be inclined to do something like sampling a few thousand times a second, saving that to a hard drive, and recovering the hard drive after the test. It’s very, very easy to collect more potentially useful data than you send over a radio.
https://spaceflightnow.com/…
It’s usually a good practice to test one things at a time. But it’s also a good practice to test as flight-like a configuration as possible. So, to test the abort rockets separately from the parachute could be either a good or a bad idea. But I think part of this is also programatic. Somewhere, they have a development plan written down, and somewhere in it, it says something like, “If AA-2 is successful, then proceed with X. If AA-2 is not successful, then figure out why, fix it, and if the fix is a major change, then conduct a AA-2 retest.” What if they had included the parachute and it hadn’t worked as expected? No one would want to get into a conversation about how that wasn’t really what they were testing, so it doesn’t matter. Or why they included it in the first place, if it really didn’t matter.
I’m sure they can provide a list of reasons why a parachute test wasn’t necessary and why it was not necessary to recover the capsule, or install test hardware that would need to be recovered. And as a general rule you don’t want to pile too many things into a single test. But that philosophy is normally associated with aircraft testing where there will be many test flights. In this case this was one of the very few opportunities to test capsule hardware including the parachute system in what would seem to be a relatively high fidelity simulation of the dynamics of an actual launch.
Presumably they felt that telemetry data was all that was needed, and as backup the “black boxes” that were spit out prior to impact containing recorded data (painted orange just like airplane black boxes). In some tests like the scramjet test flights telemetry is the only practical option. But in this case they could have recovered the hardware but chose not to. Perhaps for valid reasons not related to schedule pressure. It just creates an odd visual that you have to overcome with trust in how the program is going.
You see, that last bit is what bothers me.
Now, they are all smarter than me, so they know best. But, since it is the same parachute system as both CCDev vehicles, and since there have been problems with it, then I am on the side of more tests…especially process and system integration tests.
Here’s a question…would they or anyone else be satisfied with the CCDev vehicles carrying out inflight abort tests and NOT continue the test all the way to the ocean surface?
I’m fairly sure the parachutes are different for Orion, Dragon 2 and Starliner. They’d almost have to be, given the mass and shape of those vehicles. (And Dragon 2 uses four, not three.) But that’s a detail. The requirements and required reliability would be very similar.
I’ll also say I’m not sure if the logic I described is the logic the Orion program followed. It’s just consistent with other NASA project, the decision not to use parachutes and what I’ve seen on NASA projects I’ve been involved in. I don’t actually own a crystal ball.
But in terms of the expectations for commercial crew, it’s important to realize that SpaceX was the one who put a full-up abort test at max Q into their development program. In the commercial crew proposals, companies were expected to describe how they were going to design, develop, and test their vehicle. And how that whole, overall process would convince NASA that the product would work, and work reliably. SpaceX took an approach which used a design and development process which is quite different from how NASA would do things, didn’t give NASA a whole lot of visibility and little oversight, and was very heavy on realistic tests.
And that design and development process does matter to NASA. Actually, as much as testing. There is a widespread believe that the process NASA will produce a vehicle that works, and works as intended. If you look at any program’s schedule, there isn’t a loop for doing a test, identifying all the serious problems, redesigning, and retesting. That happens sometimes, but time and money come out of reserves, the expectation of failed tests isn’t the built into the schedule or the budget. The expectation is that, if it’s properly designed, the test isn’t there to demonstrate that the system works. It’s there to simply confirm the program’s conviction that the design was just fine.
By that logic, if you follow the approved process, you don’t need all that many tests, and they don’t necessarily have to be completely flight like. But if, like SpaceX, you follow a different process, then the expectations on testing are higher. Given the planned tests, I suspect whatever process Boeing is following is much more NASA-like.
Well, you know more about it than I do, but I did read somewhere that the parachute systems were not being developed separately by the CCDev contractors, but by the same NASA design team or contractor or something as Orion’s and that they shared so many common origins that a failure for one could impact all three programs.
Also, comparing the development costs of SLS/Orion vs Falcon/Crew Dragon might put a question to the claim that the NASA approach is more budget-friendly. Having said that, I agree with that approach being superior…or at least better informed and more solid.
I also don’t think that the integrity of that approach holds up as well under intense and changing calendar pressure as the SpaceX approach…making the skipping of a full systems test troubling in the context of Artemis.
I am not a fan of Artemis. I like that it excites the segment of the population that wants Trump to look good and that it applies political pressure on space exploration, but those are the only things I like about it. I don’t like skipped steps in a running plan. I don’t the direction toward “Flags and Footprints” that the newly accelerated timeline has turned the mission design. I don’t like how this whole plan now depends on the impossibility of this bipolar and Trump-toxic congress passing a budget this year.
When it comes to the parachutes, I actually have no idea who designed and developed them. But they would have to be different designs, due to the differences between the vehicles. But it’s certainly possible the same group did all three designs. Unfortunately, that approach would actually make it harder to demonstrate reliability. They would be different enough that proving one works doesn’t let you skip testing the others. But a failure in one could raise questions about common design issues.
The NASA approach to design and development is quite clearly not cost efficient, and it isn’t well suited to schedule pressure. But I don’t agree about it being a superior process. Most failures involve things no one thought of. You can spend a huge amount of time and effort hunting down and resolving every problem you can think of, but you are going to miss some things. I’d rather put the resources into more extensive and realistic testing, to catch the 1% probability problems which no one thought of, rather than fixing the 0.1% probability issues someone did think of.
Ya, my understanding was that the parachute systems were as absolutely close to the same as could be achieved on three different design capsules and that this was deliberately done to avoid the complexity of redevelopment and certification. I’m not just talking about hardware either. Since the Orion program is the source of the design, then perhaps they’ll have enough time to look at system integration in a different test…has Orion had a pad abort test yet?
I hope Boeing and SpaceX both do a full systems test all the way to the water.
Yes, there was an Orion Pad Abort Test in 2010, so no need to repeat it. Sure its a different capsule now, but the box was checked off…
https://www.youtube.com/wat…
“What if they had included the parachute and it hadn’t worked as expected? No one would want to get into a conversation about how that wasn’t really what they were testing, so it doesn’t matter. “
This has been my thinking on the lack of data from SX on their recent failure- likely unrelated, so let’s avoid more mud in the water.
It depends on the test plans, and I don’t know what SpaceX was doing during that test. There is always a balance between making a test realistic, and also making it easy to diagnose the causes of a failure. The real world is complicated, and most real failures don’t have a simple, single, easily identified cause.
On the other hand, SpaceX is probably doing the right thing by saying so little. People always complain, and it always generates bad press, but one of the last thing an incident investigation should do is release preliminary, potentially incorrect, conclusions.
I wonder what happens the first time they have to use the abort system and the parachutes don’t work mmm
Geez, I never would have thought Orion would meet a milestone (abort test) before both Dragon and Starliner. Of course, the rocket it rides on is still not available. Maybe Bridenstine was on to something when he was musing about mixing and matching capsules and rockets.
I am not sure you could call this a true abort test since they didn’t test the capsule’s attitude control system or parachutes during an abort. All it tested was the ability of the abort system to pull the capsule away from the rocket. That said remember that the Orion capsule has been in development since 2006, so it is not surprising they were first using a simple boiler plate compared to the actual capsules Boeing and SpaceX will use. But then at a billion dollars apiece it’s probably too expensive for NASA to use a real capsule instead of a cheap boiler plate one.
Before I would ride that thing, I’d want to see a full-up integrated abort test performed. At least they got some preliminary data on g-forces due to the tractor rocket’s acceleration away from the booster (about 7g). But, without the stabilization system and deployment of the chutes, one really doesn’t have a good feel for the g-forces during the tumbling, stabilization, etc.
I would expect SpaceX and Boeing to do full-up Max-Q abort tests, possibly some time this fall. Best to both of them!
SpaceX will be doing a Max Q abort test. They even rigged the Falcon 9 booster to actually fail at that point so the software will need to make a decision to make it as realistic as possible. Boeing however is just doing a simple Pad Abort Test like SpaceX did in 2015 and calling it finished.
“Of course, the rocket it rides on is still not available.”
OTOH, that boost motor they used — courtesy of the venerable Peacekeeper — worked just fine-dandy!! (Dang,I wish we’d kept that one instead)
On the other hand, Crew Dragon has actually been to orbit, and even executed active docking and undocking with the International Space Station. So this becomes a less impressive milestone “first” than it might otherwise be.
SpaceX had done this test , basically the same simulation !! The Inflight full abort test is to be done using a full falcon rocket at launch weight and speed , as insisted by NASA
13 years and probably close to $15B and they still haven’t done a flight like fully integrated test, what a boondoggle. More boilerplate and non flight hardware configurations to kluge together a test in parts to some how pencil whip they have shown during an actual flight emergency they have confidence that the system can perform the full abort profile from event trigger to getting away from launch vehicle to safely landing crew in the water. Whatever happened to test like you fly mantra for the agency?
https://arstechnica.com/sci…
Estimated $16 billion.
Hey, when your launch vehicle gets delayed, you stretch the program. Which means you keep the design team around years longer than expected. And when you get told, just after design is mostly done, “By the way, we have this new service module that you need to design in to keep ESA happy”, costs go up.
Yes mmm Old Space , you gotta keep Boeing and Lockheed in business !!
I was actually more interested in the booster.
From Wikipedia:
The booster was a repurposed Peacekeeper missile procured from the United States Air Force and modified for the mission by Northrop Grumman.
Would be cheaper to used a brand new Castor 120 from NGIS. But they get to disposed one of the old Peacekeeper motors without paying for the cleanup cost with the usual method of disposal by burning it in a static fire.
I’m not sure. Even as boilerplate, the test article had to have the same mass and mass properties as a real Orion. And the vehicle would have to be at the same altitude and speed (actually, velocity) at max Q as a real Orion/SLS would have. And the same maximum dynamic pressure at max Q. At least it that would have to be true if the test wasn’t meaningless. Using a boilerplate test article and not using parachutes are things that can be talked around. But not flight-like conditions at the time the abort rocket fired. Solid rockets can’t be throttled up and down (well, not ones we have lying around…) so that probably places some serious constraints on what, exactly, they could use as the launch vehicle.
The SR-118 motor from the Peacekeeper have a higher thrust and shorter burn time than the Castor 120 motor. So required ballast to match the acceleration profile of the SLS. The ballast necessitated additional structural reinforcement of the abort test booster in additional development cost and time.
Correct, it is a re-purposed/refurbished first stage (only) of the LGM-118, which was a 3-stage missile.
Swords into plowshares.
This photographer got a good video of the Wiley Coyote Splat at the end that isn’t on the offical NASA one. It also shows just how fast the capsule was tumbling before it hit. The mositure condensation even makes it looks like its trailing flames at one point. Great Video!
https://www.youtube.com/wat…
I noticed that there is a second smaller splash five seconds after the first impact which I am guessing was the escape tower impacting.
Someone else on YouTube zoomed in on this footage and slowed it down and pointed out that the first splash occurs six frames after the capsule disappears over the horizon, which gives you some sense of the distance involved.
The second splash might be the rebound as the “crater” of water collapses back in.
I waited and waited for the parachutes – but any crash test dummies in this test would have all expired. Come on, NASA, you can do better than this! This is not a Launch Abort System test, possibly just a test of a solid rocket engine’s power, nothing more. If SpaceX had done this you would have said, FAIL!.
Come on, this was not an In Flight Abort test. Any test dummies in the capsule would be dead. If SpaceX or Boeing tried to foist this off as an In Flight Abort test they would have gotten a FAIL, no doubt about it.
Except that Boeing isn’t doing any sort of in-flight abort test. And NASA doesn’t consider that an automatic fail. (Which, by the way, most students can’t manage at a university. Not showing up for the test is a fail.) SpaceX put in a full-up, flight-like abort test in their proposal for commercial crew, and that was to give NASA confidence that the final product would work. Boeing proposed a different approach to demonstrating Starliner would work. Internally, within NASA, the Orion program took yet another, different approach. The logic behind those approaches is different, and some of it may be bad logic. But that depends on the whole development approach, not the details of a particular test. (Although, by the way, I agree about AA-2 being a decidedly unimpressive one.)
And it cost only $256 million. Such a bargain.
Yeah they need to ditch Orion and contract SpaceX to provide them a Falcon Super Heavy or two !