There are a couple of bits I don't get from this test. Firstly, in terms of software development, is the test an actual real world test if the software they are using to abort is flight software embedded in non flight software? How does that work? Is this just a keeping people busy activity?
k? Secondly, why wouldn't you perform a parachute landing? you can never have enough knowledge of parachute performance (you could test one out scenario for example) There is a requirement to get data back physically, yet NASA have resorted to physical ejection, what lessons will be learnt from that?
Quote from: charlielowndes on 05/01/2018 11:46 amThere are a couple of bits I don't get from this test. Firstly, in terms of software development, is the test an actual real world test if the software they are using to abort is flight software embedded in non flight software? How does that work? Is this just a keeping people busy activity?It is the algorithms and processed that are being tested and not the softwareQuote from: charlielowndes on 05/01/2018 11:46 amk? Secondly, why wouldn't you perform a parachute landing? you can never have enough knowledge of parachute performance (you could test one out scenario for example) There is a requirement to get data back physically, yet NASA have resorted to physical ejection, what lessons will be learnt from that? recovery of the vehicle adds more complexity and costs.
Quote from: charlielowndes on 05/01/2018 11:46 amThere are a couple of bits I don't get from this test. Firstly, in terms of software development, is the test an actual real world test if the software they are using to abort is flight software embedded in non flight software? How does that work? Is this just a keeping people busy activity?It is the algorithms and processed that are being tested and not the software
It just seems to be a bit of a mash up, looking busy, really, we’re not doing anything of substance soon type deal. I get that you don’t need to do full up testing anymore, but this activity seems to test very little other than some subsystems that are not even being run in flight systems. I struggle to see the value, but hey, I’m no NASA engineer!
Quote from: Jim on 05/01/2018 04:18 pmQuote from: charlielowndes on 05/01/2018 11:46 amThere are a couple of bits I don't get from this test. Firstly, in terms of software development, is the test an actual real world test if the software they are using to abort is flight software embedded in non flight software? How does that work? Is this just a keeping people busy activity?It is the algorithms and processed that are being tested and not the softwareI'm a software guy so this confuses me. Software IS "algorithms and processes" ...
Quote from: Lar on 05/01/2018 10:27 pmQuote from: Jim on 05/01/2018 04:18 pmQuote from: charlielowndes on 05/01/2018 11:46 amThere are a couple of bits I don't get from this test. Firstly, in terms of software development, is the test an actual real world test if the software they are using to abort is flight software embedded in non flight software? How does that work? Is this just a keeping people busy activity?It is the algorithms and processed that are being tested and not the softwareI'm a software guy so this confuses me. Software IS "algorithms and processes" ... It sounds like what they're doing is taking some components of the flight software and putting them in a virtual machine that that makes the GNC software think it's running on the flight hardware, when actually it's running on the AA-2 hardware, making the whole thing an elaborate kind of integration test harness. As a software guy, surely you've heard of "virtual machines" and are familiar with the concept of a "test harness" for use in automated testing?
Quote from: Oberon_Command on 05/02/2018 01:50 pmQuote from: Lar on 05/01/2018 10:27 pmQuote from: Jim on 05/01/2018 04:18 pmQuote from: charlielowndes on 05/01/2018 11:46 amThere are a couple of bits I don't get from this test. Firstly, in terms of software development, is the test an actual real world test if the software they are using to abort is flight software embedded in non flight software? How does that work? Is this just a keeping people busy activity?It is the algorithms and processed that are being tested and not the softwareI'm a software guy so this confuses me. Software IS "algorithms and processes" ... It sounds like what they're doing is taking some components of the flight software and putting them in a virtual machine that that makes the GNC software think it's running on the flight hardware, when actually it's running on the AA-2 hardware, making the whole thing an elaborate kind of integration test harness. As a software guy, surely you've heard of "virtual machines" and are familiar with the concept of a "test harness" for use in automated testing? Excuse me Jim.Don’t snip others posts and make them out to be mine. Sent from my iPhone using Tapatalk
Your post 8 contains snips of post 2 (Mine) and post 5 (Lars), combined to make them look like mine.
Some years back, there was a concern that the Launch Abort Vehicle (LAV), which consists of the Orion CM + LAS, was aerodynamically unstable at transonic and low supersonic velocities. So, I'm interested in see the LAS pull the CM off the Abort Test Booster and fly straight, leaving the CM in a stable attitude from which it could deploy parachutes.After the performance of the parachute system on EFT-1, I see no need for parachutes in this test. I do think it is bad public relations to not have the parachutes, and to crash the CM on camera. That video will show up in way too many movies and TV shows. The Orion software that deals with a LAS abort is a small subset of the total Orion software. Pulling together a full software load for this test would take too much time and distract from preparing for EM-1, so I think Orion is doing the right thing, here. I'm looking forward to this test!Great article!