|
Ascension Mk1 Succeeds Despite LES Mishap
Preliminary study of the data from yesterday’s mission shows that for the most part everything went exceedingly well. The Ascension team will be spending at least the next week or maybe two analyzing all aspects of the mission – not only does the ascent of the rocket need to be studied but the performance of the capsule while in space (not just RCS but also internal systems like life support) and the performance of the capsule and heat shield on re-entry. We can already tell just by looking at the heat shield some ablator material was shed this time, how much is still to be determined.
The unintended activation of the Launch Escape System is thankfully not a serious issue and the cause was quickly tracked to a logic error in the AFCS mission script. The LES abort was set to be triggered if the computer no longer had a complete circuit from the wires that ran the length of the fuel tank – so if any part of it ruptured the wires would be cut and signal the LES to activate. Well, the script failed to take into account the fact that once the capsule was released from the lifter, the connection to the tank would also be severed and thus the LES would think it had exploded! A simple command to disable the LES tank check at separation is all that will be needed to avoid this in the future.
Progeny Mk7 Won’t Reach Orbit – Yet
Although we made a claim last week that the initial Mk7 design will be capable of attaining a 100km equatorial circular orbit, we’ve been forced to retract that in light of new results from the Launch Vehicle Designer (a tool found in KSPTOT). It was discovered that somehow the LVD had deleted Kerbin’s atmosphere entirely (the program is still under development and has bugs) so our initial claim was based upon an ascent from an airless body the size of Kerbin. Once we rebuilt the ascent with the atmosphere back in the rocket failed to achieve any type of orbit, falling short by only ~40m/s of Δv.
Although it would have been nice to been able to reach orbit with the initial Mk7 design, the inability to do so isn’t such a big deal. Note that we’ve been saying initial design. The Mk7 is planned to undergo a series of rapid iterations towards a final design that will be capable of placing independent payloads into orbit. The first Mk7, the Mk7-A, will mainly ensure that various new technologies (steerable fins, reaction wheels, 0.625 payload parts) are functioning as planned before moving on to more and more complex designs all aimed at the same purpose (hence the same Mark designation).
ATN Database
The latest update for the Asteroid Tracking Network database is available here, containing 3,440 asteroids and 2 updated with new observation data. Here are the 29 asteroids that were discovered this past week:
 |
 |
From the Desk of Drew Kerman
Out of Character Behind the Scenes stuffWritten on 4/10/19
Ascension flight
Oof that was a lot of work, but very rewarding! I hope everyone enjoyed the mission preview tweet with all the graphical goodies. The ground plot, I want to point out, was rendered entirely on data that the LVD and Mission Architect spat out. I did not fly the mission and then use the telemetry data to plot that out. I would have posted a ground track comparison except for the stupid mistake on my part of not thinking to save the damn game after the LES tower fired – so that trajectory is lost unfortunately and in the grand scheme of things I didn’t consider it important enough to take the time to recreate (especially since it would never happen the same way it did).
The LES tower firing was an actual logic error in my kOS code and after flying the mission with it fixed I decided to go back and keep the error in. Routine missions aren’t as cool as ones where something goes wrong.
So yes I flew the mission several times to work out any logic errors in the kOS script but it’s important to me that I make it clear I did not do anything to change the initial ascent plan that I had worked up in LVD. So in other words I didn’t finish the mission and decide that the rocket didn’t fly as far as I wanted or as high as I wanted or anything like that. I only fixed any script issues that would prevent the rocket from doing what I had initially planned in LVD. There was no reverse-engineering of the mission planning! LVD did a good job getting me where I wanted to go.
Luckily the new FAR developer released a fix for an issue with stock heatshields on re-entry I was able to apply before this flight. Good timing!
I did have to overcome an issue with kOS not detecting any connection with the KSC through RemoteTech because I was using a kerbed capsule. Although it’s a known issue thankfully the solution is rather simple and easily applied through a MM patch.
I did actually have Jeb along for the ride so the capsule would properly use life support and I could ensure Kerbalism wasn’t going to try to kill my kerbal.
Finally, if you recall I was griping in a previous Desk Notes about my non-rechargeable batteries no longer working for the Deuce. Well I remembered that was because I changed them to having to manually turn them on. Duh! It worked fine for the capsule too, draining whenever the engine alternator was off.
Mk7 oops
This was a real mistake I decided to not go back and fix, because the LVD is indeed a program that still has some issues and the KSA is indeed still learning how to use it properly, so there’s no reason why mistakes shouldn’t be made here and there. Having this rocket be an initial test design for the final Mk7 however was always the plan and not something I made up to cover for the shortcomings of this rocket.
Stupid twitter collections
I mentioned a while back about TweetDeck no longer allowing me to access older collections to add tweets and I have to do it manually. I used to use a JavaScript library to code up my own solution but recently that’s been having trouble completing requests. Lately it’s just plain refused to work at all, which sent me scrambling around for 2hrs or so for a replacement option. Found one, and I can modify the code and run what I need right from the testbed site. Still, I hope someday TweetDeck gets their shit together.
