Jun 08 2018

Operations Summary – Week of 6/4/18

View post on imgur.com

Ascension Program First Flight Partially Successful

Officially, the flight of the first Ascension Mk1 earlier today will be marked as a failure since it did not achieve its primary goal of making orbit. At the same time though, it can be regarded as a success in other ways: all the new ground support equipment in use functioned as designed (although a few extra days were required to work out all the kinks) and the updated Automated Flight Control Software showed it was capable of not only controlling the rocket systems but for the first time using active guidance via control surfaces to maintain a given orientation during ascent – at least while there was still enough air left! Although the rocket’s return did not seem to go as smoothly as hoped, we are glad it at least survived enough to be recoverable. Once the teams can get at the treasure trove of onboard data we can start to learn more about what went wrong during the ascent and descent, such as why the operations code seemed to partially hang up. We will of course have a full flight analysis posted thereafter with all the details.

Looking ahead now to the next Ascension Mk1 launch in July, parts have already started to arrive and Lead Engineer Simon expects to begin assembly and integration by the middle of the month. Because we can’t get immediate access to this flight’s data to determine what may need to be adjusted for the next flight, we’re aiming more for mid-July as a launch target at this time. The engine for this second flight has already been built, but work on additional engines has been halted until we can confirm that today’s incident was due to the engine being worn from multiple test firings and not from some overlooked design flaw.

New Crew Oriented & Ready to Assist Genesis Program

Flight Officers Tedman and Aldeny have spent this past week getting acquainted with their new home here at the Kerbal Space Center and getting into a regular routine. This will be the first time both of them have lived above ground and the shorter day/night cycles take some getting used to as opposed to the controlled lighting environments below ground that run on a longer 16hr/day, 8hr/night. In case it wasn’t obvious, no one above ground actually sleeps every night cycle but the times that you do sleep can be different than the times people below ground are sleeping.

Now that the new crew members are in sync with the rest of the campus and know their way around next week they will begin assisting with Genesis operations in conjunction with C7 Aerospace. First order of business will be to get the Deuce ready to mount a rocket on its back for a carry test later this month. It’s also nearly time for another Civvie flight to gather pollen data – Lead Scientist Cheranne has been saying all week she would set a date for that but so far nothing. Genesis is also always working on new contract deals for the Deuce to fly. As always, updates on planned missions can be found via twitter.

KerBalloon Crew Carries On

Not to disrespect Specialists Bill and Bob but they were never really an integral part of the KerBalloon program – the whole crew worked together to achieve mission success. So although Bill and Bob no longer take part since finally embarking fully on their journey to become astronauts the KerBalloon crew have not had any issues carrying out missions in their absence. This week they took a foray via UTV out to the west shore desert to release a high-altitude balloon, which was successfully recovered and returned to KSC. Profit!

Alaba Sets Up for 9th Intercept

This past Monday our moonlet Alaba took a swing through Mun’s SOI, just as astronomers had predicted again accurately for the second time in a row. New observations showed that it was setup for yet another intercept this coming Saturday, so we’ll see if it continues to end up where astronomers predict. If so, and if another intercept isn’t imminent, they will attempt to chart out the next several intercepts to see once more how well their predictions hold up over time.

Help Us Name the K-3X

Little known fact: the Civve and the Deuce were both internally known as the K-3X project, but had nicknames chosen prior to their public debut. The K-3X designation translates to K followed by 3 numbers that refer to engine count and wingspan. So the Civvie can also be called the K-108 while the Deuce would be designated K-213. The number assigned to our latest K-3X will be 4## – the wingspan is still not confirmed as the original design couldn’t fit through our HAB doors and the concept of folding wings are still under review by the Air Safety Administration.

If anyone has a nickname we can use for our next aircraft, let us know! Here is the most recent design render if you need some visual inspiration.

ATN Database

The weekly update for the Asteroid Tracking Network database is available here, containing 2,054 asteroids and 7 updated with new observation data.

From the Desk of Drew Kerman

Out of Character Behind the Scenes stuff

Written on 5/24/18

You know, I was this close – I was behind my 3-week lead by three days, and then the lead-up and launch of the Ascension Mk1 came around. OOOoof! Gawd, I had completely forgotten how much work I had to do to get things setup for the Progeny rockets back even as little as a few months ago for the Mk5 – the Mk6 was so similar I was able to reuse a lot of assets. Long story short, I’m back to where I started over the course of like 3 in-game days. Long story long… well ok here we go…

Going vertical

You maybe thought a photo like this of the rocket getting raised took me about 20-30 minutes to get setup and snapped. (Not counting all the work I had done previously on the carry vehicle, which I built weeks ago.) Yea no, it actually took me like 3-4 hours of total work to get right. First I tried to actually set it up that way in the game by using the robotic controls and lifting the platform and deploying the legs – oh my, physics had a field day with that. The whole thing wobbled so much I couldn’t keep it straight long enough to get a good line up with the engine collar, and the stabilizer legs visually clipped into the launch pad. So back to the SPH to just move everything into position beforehand, strut the shit out of it, then move it back to the launch pad. Carry vehicle still wanted to jitter ever so slightly and so I tried time warping to settle it but every time I came out of warp the damn launch towers and engine clamp would jump a bit higher out of the ground. FFS

Those are just the highlights of my troubles. In the end I still ended up with a slight misalignment of the engine to the clamp. Maybe you can’t see it but I can, dammit. The crew were relatively easy to composite in. I wasn’t going to bother at first because of how much time it took to get just the damn rocket photo but there are some running around the VAB without tools in their hands I can use, so I did.

Also if you didn’t notice then I’ll point out the orientation of the rocket isn’t the same as when it is seen attached to the towers in all subsequent photos. But it was too late to correct that oversight after I noticed it *le sigh*


Ok let’s start with all the assets. Needed new Ops Tracker images for the flight updates, needed a new ascent profile diagram, a new public viewing diagram of KSC, had to come up with a launch timeline taking into consideration the new rocket and its capabilities and properties – including the minutiae of how it was going to launch. But all was pretty minor compared to three main issues:

New code

Biggest deal was writing the new operations code that would fly the rocket. Okay writing it wasn’t so bad – testing it was a pain in the ass. Nevermind the fact that I can’t run-check my code outside of the game so I have to keep getting errors and restarting the flight to fix stupid syntax issues, there were several new things I was doing that introduced some tricky logic issues. One in particular hung me up so bad I had to ask for help on reddit. On the flipside, it turned out to be such an insidious logic error I figured it couldn’t have been discovered without flying the rocket so it was left in for the actual flight. It will be detailed in the flight analysis. This took me about a day and a half all told.

I’ve also given up making the code completely “realistic” in showing every behavior. The engine pressure check? Yea can’t actually do that so it’s just not there. Kicking things off just by using the STAGE command? Yep.

Launch photo

I like the stock ground smoke effects it’s just they are so ridiculously huge and I could never get anyone to really tell me how to tone them down nor did any fiddling of my own yield any decent results back when I was considering this issue for the Progeny rockets. Smokescreen works great for ground smoke but the only problem is that this rocket doesn’t have SRBs, so there won’t be any smoke once it leaves the pad. Try as I might, I couldn’t set the smoke emitter to detect enough of an air density change after lift-off to automatically disable the smoke during ascent. So for the photo I just had to have the Smokescreen window open and when I launched I manually disabled the emitter, hid everything and snapped the photo.

But that wasn’t my only problem. The way the rocket is mounted to the engine clamp in this photo? Yea, it doesn’t actually work like that. In fact if you do try to launch like that, the rocket will glitch out and go flying up a few meters into the air. The clamp is actually supposed to attach to the bottom of the engine and grab on to its shroud, but I didn’t want the rocket launching out of a shroud (fun fact, the Rokot does this since it used to be an ICBM). So I had to glitch the rocket out of the clamp, then reset it back atop with VesselMover and quick switch over to my KerbCam view to trigger its ignition sequence before very slight movement at its base put it off the clamp edge and in tipped over. Yes, the clamp itself is not actually visible in the launch photo so why bother? Well the smoke particles are physical objects and so I needed the clamp there for them to bounce off properly, not just the ground.

I also planned to do several launch photos with different smoke settings and composite them for more dynamic-looking ground smoke but then I realized I couldn’t reproduce that in a video so I didn’t bother.

Also, I found out that SmokeScreen has a particle limit per effect in addition to its global particle limit. By default its 1000 so if you have your global limit set to 8000 but the engine has only 4 effects, you’ll only see a max of 4000 particles by default. Here’s how you change that.

Launch issues

So the whole idea of the engine failure was spawned by this tweet, which was a follower asking how the engine bell was cooled and is yet another reason why I encourage people to interact with what’s going on and ask questions and poke holes! I hadn’t considered what the testing would do to the engine in the form of stress from heating and cooling not to mention just general vibration and whatnot. Ideally a new engine would have been used but as it was noted, the KSA wanted to fly sooner rather than later so they kept this engine to use for the first mission. It took me a few tries to figure out something I could do that would cause a similar effect in the game without screwing up the mission and my running code too badly. Basically I just dialed down the thrust limiter on the engine and then, since that would reduce fuel consumption, used TAC Fuel Balancer to dump the fuel so it would still get used up as if the engine were at full thrust.

Although I wrote in the launch abort as just a means to foreshadow that there were going to be issues with the engine during flight, in doing research for what changes I should make to my launch timeline I came across this article about a Falcon 9 flight that suffered a similar problem with tight constraints set to an engine. I thought that was pretty cool.

Small stuff

I’m not sure how to calculate the amount of gas I need for pressurization, so not sure how I’m going to handle that yet. Using nitrogen works now because it’s so dirt cheap I can get away without factoring it into launch costs. Eventually though I will want to switch to lighter helium, which would be more expensive, even for kerbals. I also don’t know the mass since I don’t know the amount, so if I don’t figure it out I’ll just make the whole tank part a bit lighter when switching to helium at some point. The reason I’ve bothered defining the pressurization gas is because both nitrogen and helium are actual gasses that can be mined and used for various aspects of the game in various mods, and picking between the two does have consequences given the difference in their properties.

I forgot when I actually flew the mission to account for the slight amount of fuel that would have been used up during the first launch attempt that was aborted. It’s rather insignificant, but still!

Also the alternator recharges the batteries but since I don’t have that tech yet, I just let it happen. But then, when working up the telemetry data I made note of the EC level reported at launch and just copied that over all the 100% records while the engine was burning, then subtracted that difference from the rest of the cells. So the alternator allowed the battery level to remain constant while the engine was on, then it drained again. I also had to remember as well to take into account the 40 minutes during which the rocket was no longer attached to the service towers and operating on its own power.

Finance juggling

Just another example of the crazy detail I go to – where did the weather balloons for this launch come from? Well the ones that Progenitor bought were sitting around as surplus so Ascension used them. But Progenitor wasn’t going to just let them take them for free, so I adjusted the program finance sheets to reflect that Progenitor’s last purchase of 5 balloons is now 3 and Ascension has expenses for 2 now. On the monthly expense sheet I didn’t change anything since in the overall finance scheme the expenditure is the same. If that made any sense maybe you want to be my accountant?