Box of Stardust

Okay, we have battles. And again, the leaderboard and stats sheet (linked in the first post) has been updated, if you want to know the current state of things and see how I measured each plane against each other. Very useful if you want to test against leaderboard planes and get a baseline for where any new planes you design might stand.

Okay, battles have happened. Leaderboard and aircraft stats sheet has been updated. Videos later. Here's my main points after the session, with minimal spoilers: Every plane with Ven's Stock Revamp was skipped over. Du-4A was a surprise. X-Fighter Hunter earns a rematch. Armor Bud is relegated to exhibition matches only because the armor plates are too cheesy because they still add inconsequential weight, and can be modified without gaining weight, and I'm pretty sure it's otherwise impossible to kill a plane plated with those.

Okay, I've finished my second plane update after all this time (because wow was there a lot to learn about BDA 1.2), so I guess I'll start up the battles again. The Sidewinder spam is real. What's the difference between 30 and 50?

yes am busy and little burnt out, gimme a bit more time and part of my BDA time has gone into competition planes for PFC also I was busy helping out the Kerbal Express Airlines thread, which was also time consuming and also just playing other games Plus I just received word of a fix for the incredibly annoying 'missile launch self-detonation' problem, which has been a really annoying thing in BDA 1.2, along with more AI Pilot quirks that I'm learning (trying to make the best 'manned' aircraft for the 'manned' category). I did a quick fix to the PEGASys-D6 to alleviate the whole 'occasionally detonates missiles on itself', so there's that. I know technically we're not supposed to modify aircraft once they've battled, but I think minor things like tweaking values to fix general problems in gameplay should be allowed, since it's not a change directed at beating any specific enemy.

I think it's because 'gallons per passenger mile' is an actual industry term, but we actually use the range in kilometers. It's a little obtuse, but yeah.

Well I've paid my dues. I've added a net amount of... 5 or 6 planes after this time, offsetting a lot of the submissions I made. Since there's not much others that want to step up and judge, maybe we should just call an official hold for all the judges to not feel obligated to do reviews for a while. I mean, I know we've just plain haven't, but to place a stop in an official capacity for a month or two might help. If you want to help and think you're up to the task, we'll take any we can get! I have one last review I signed up to do by the way; I'll get that out of the way later, or tomorrow.

The other ones, I mean.

Do you want to revise your entries with Ven's Stock Revamp? I think @dundun92 mentioned that VSR changes stock parts, which we can't have.

Just a final report update on experimentation with the scaled down Goliath engine. On the same aircraft with roughly 2200 kallons of fuel, using identical twin-jet configurations: STRIDE; 1.8m Goliath: 255m/s (0.13 burn) @ 7300m = 4300km OR 220m/s (0.11 burn) @ 6500m = 4400km Lotus: 219m/s (0.10 burn) @ 6400m = 4800km Wheesley: 254m/s (0.15 burn) @ 7200m = 4200km (this is also with an additional 300 units of fuel, for a total of 2500) So STRIDE is superior to the Wheesley in every way, and costs 300 less than the Lotus. However, the STRIDE's advantage over the Lotus is its speed and cost, which is offset with worse fuel consumption. The STRIDE can barely make the 0.10 burn rate consistently while at a 220m/s cruise like the Lotus; it's probably fair to call it a 0.105 burn rate, putting it at 4600km range. However, in real life, fuel efficiency is the progression being worked towards, so... well, maybe in the 1970s-ish pastiche of kerbal-land, "faster is better" probably has a place, but I don't know. These are the values I have, and I'm left to believe that the STRIDE is no more of a better engine than the Lotus is, at least by a significant margin, trading off some fuel efficiency for a higher speed limit. While a plane powered by STRIDEs will make the speed recommendation/requirement, they will lose out to an equivalent plane in KPPM and lifetime cost to a Lotus-powered plane. And Wheesleys... just suck.

Test Pilot Review: @Magzimum's MAD TART-1b Sparrow A quirky one for sure... Figures as Tested (TART-1b, 24 passengers max): Price: 15,512,000 (empty) Fuel: 760kallons Cruising speed: 129m/s Cruising altitude: 2000m Fuel burn rate: 0.05kal/s Range: 1900km Review Notes: We at Twin Crown Aerospace Industries has found an opening in our test airfield's busy schedule to fulfill another KEA outsourced aircraft inspection. But while it may be a test day for a civilian aircraft, apparently observers at the airfield will see similar acrobatics and crazy feats as during the recent fighter jet tests taking place prior to the date and will likely take place after... Magzimum Aerospace Design has designed... an interesting biplane intended for the turboprop class- with inverse stagger, so that's pretty cool. Pre-rollout analysis of the Sparrow revealed a center of lift slightly behind the center of mass. We expected the plane to have a noticeable pitch-down tendency, but for an aircraft of its size and control surfaces, should be able to compensate for. Ground observation by engineers... made notes on its very tall landing gear, but these were required to give the pusher props ground clearance. Also we're not really quite sure where the entrance is other than through the cockpit, which is a bit inconvenient, especially since it's so high up. Lastly, concern was noted over the oversized engines for an aircraft of this size. TCA engineers have not really dealt with this particular turboprop much- once only, actually, on an internal project- so we weren't really sure as to its exact performance statistics either. And with nothing else much to note down, our new test pilot, Shepry- who most call 'Shep', apparently- was given her first assignment as a civilian aircraft test pilot. We think she seemed quite excited, as the day prior, she was flying around the same airfield, but in a super-maneuverable fighter jet. No boring, large commercial liner to test yet! (Though if the Stouts were anything to go by...) Pre-flight checks revealed... interesting behavior by the front control surfaces. While they rolled together, the bottom one didn't seem to actuate for pitch correctly. We probably should have turned these two sets of control surfaces off for flight, or at least the one that interferes, but... we didn't. Oh well. Shep was cleared for takeoff, and the two large turboprops propelled the Sparrow down the runway with ease. In fact, it felt like it could pull up and take off fairly quickly- but its landing gear setup was hindering it from actually taking off until around 45m/s. Cruising altitude was reached in no time, only 2000m. But on the way there, the turboprops made a surprising noise once the plane hit around 120-125m/s- they just kinda died. Not shut off, but just stopped producing thrust. This was a bit of a problem, since cruising speed was said to be 135m/s according to the briefing. But it was not to be so, with the Sparrow topping out at 129m/s. That actually misses the turboprop class minimum speed by 1m/s, but... well, if TCA's own jet airliner designs are anything to go by, speed is a guideline that can be circumvented by performing well in all other aspects. Also, it's just 1m/s... With those cruise conditions, a maximum fuel load will actually fly 1900km, which is 300km more than the advertised range. So that's good. General maneuverability tests on the way up to cruise revealed no glaring issues; it's hard to get small craft wrong, really. During cruise, trimming the aircraft did not prove difficult, though the pitch-down tendency was felt in trimming the aircraft. Engine failure testing was next. Port engine failure- because Shep is new to the test pilot program and shut off the wrong engine- did not cause any severe loss of control incidents. It also performs well under full power loss conditions. Mid-speed and low-speed maneuverability tests were done. Control authority was found to be good in all axes, though the nature of the Sparrow does mean it's a bit squirrely. If anything of note, it's that pitch was found to be okay, but we kind of expected better. Especially since the front control surfaces actuate for pitch control also, but the incorrect actuation of one is a probable culprit. The Sparrow is stable enough, no complaints there. Critical speed was under 40m/s, though the engines aren't the fastest-responding to throttle input. Lastly, we decided to see what all of this acrobatic stuff was all about. First we tried the thrust reversers mid-flight, which was... interesting. It was here that we discovered that the throttle doesn't really like to respond quickly if the throttle was zeroed out- but the thrust reversers themselves responded quickly when toggled. It was also discovered that it didn't output the same amount of thrust as it did flying forward, which meant in reverse, the Sparrow didn't have a 1:1 thrust ratio like it did in forward flight. Shep put it through some loops, which actually didn't seem to impress her that much; it actually felt hesitant a bit, apparently, like it could have been smoother and looser. We don't know, we're just reporting what our test pilot said. She actually smacked the plane into the ground because of a combination of the strange throttle response and pitch characteristics... and with the landing gear deployed just before it hit the ground... it survived? The Sparrow impacted the ground at 20m/s, but nothing was damaged. With a bit of righting, Shep taxied the plane onto one of the site complex's streets and took off again. This durability test was done... a few more times... inadvertently. Somehow that large glass cockpit was withstanding impacts, probably because the large main landing gear stuck out enough to absorb the initial shock of impact. Speaking of surviving by landing nosed-over, it was discovered that, upon a normal landing, the Sparrow had a tendency to nose-over with the brakes applied to stop the plane. So... that's not too good. Finally, MAD made such a huge deal about the acrobatic achievements of the plane, that Shep thought the plane could do better than a hangar roof if a veteran pilot was behind the controls. So why not a helipad? Well, the first few times, it became clear the plane was a bit squirrely to control well to accurately place down, and the thrust reversers required precise timing even with the fairly fast thrust reversing response time, which she could not be bothered with anymore, so a new idea was devised based on the Sparrow's accidentally discovered quirks- its crash durability and thrust reverser flight performance. So after lining up with the helipad, Shep put the plane into a dive with the thrust reversers on to slow the descent. Significantly. And by toggling the thrust reversers, adjustments could be made to the longitudinal aiming position. This aim point for the landing was important, since we needed a little bit of distance to get up to speed again for a takeoff, lest the plane not have enough forward velocity to pass the ledge of the building below the end of the helipad. And finally, with a grand whack at a velocity between 20m/s and 25m/s, the Sparrow hit the pad at the intended spot and rebounded back onto its tailwheel, no worse for wear. Probably. It was a finely aimed landing, proving great short-field potential. Though, we don't really recommend this method, lest passengers receive concussions upon landing. Plus, you know, using the cockpit as a lithobraking system. Though Shep's hands were fast, the reverse thrust upon landing pulled the Sparrow back just a little bit more before the brakes could be locked down and the thrust reversers shut off and the throttle zeroed. But it turns out that this was actually a good thing, as it marginally increased the length of takeoff room for the Sparrow to pick up speed to dive past the edge of the building, and fly off to circle back onto the runway. Where it promptly nosed-over upon stopping. The Verdict: A strangely engineered, somewhat over-engineered, biplane. For it to be a good transport turboprop, we recommend switching to the industry-standard Kitty turboprops, which will give it better performance across the board, especially top speed, while losing only a marginal amount of thrust. The smaller blades will also allow for shorter landing gear, as well as a better landing gear layout for less nose-overs and easier takeoff. They'll also be a little more bearable for passengers. And the pusher configuration might be unnecessary. Actually, the pusher configuration is already kind of unnecessary even here. Also maybe a small adjustment to the aerodynamics. And the control surface actuations. And an easier entry point to the passenger cabins. Resolve these issues and it might make a good consideration for a turboprop aircraft in tight airspaces. It's a very compact plane that stores a respectable amount of fuel and gets good range, all for a really good price (er, what's the price between the Predator and Kitty turboprops?). And we can say for sure, after a few more shenanigans and accidentally bouncing the plane around the top of a building between various structures, that durability is certainly one of its strong points.

Well, the newer plane (AX-505) is lighter (37t) than the older ones the original STRIDEs were designed for (A-502, 44t), but that's still not very light, and it's also carrying 128 passengers. I guess this one can get away with Wheesleys, since I think a good rule for using those was 1 Wheesley per 15t. It'll feel really underpowered most of the time though. That's a test for later. I'm just trying new engines since I feel satisfied with the efficiency of the aircraft I've put out, but lack in speed. Here's a size comparison with the Lotus in the middle, and 1.8m and 1.9m Goliaths: So it's also a lot more reasonable in size for what would be the equivalent of the GE90 in real life, because stock Goliaths are maaaasssiiive. Stock Goliaths are like, real-GE90 size, but in KSP, everything else is scaled down. As for the new cabin, it's scaled against the Mk1 and Size 2 cabins, which are 1t and 2t respectively, and for our challenge purposes, I agree with @neistridlar's 16 passenger capacity, which comes out to the cheapest- but within the range of the other cabins unlike the Mk1 cabin- cost per seat, making it a good choice for smaller-yet-still-reasonably-sized jet. To balance the far-heavier stock cabins, we just say those are luxury cabins.

New projects because new parts! AX-505, assuming we decide on 16 passenger capacity (which, after thinking some more about it, I think is a good number), is a 128 passenger medium regional jet. But really, that's getting within an unofficial 'large' range. (hmm... small is 40 - 71, medium is 72-151 right now... maybe there should be a new cutoff for large starting about 110-120?) It's like a bigger A-401! Juuuust within 'normal' size dimensions, but not quite, since it's starting to get kinda passenger-noodle-like... It's very much just a prototype right now because it uses a new STRIDE (Scaled Turbofan Radius with Improved Dimensions and Efficiency), but this time, far less cheaty, with no screwing with the thrust limiters. Still, values are something like 255m/s (0.13 burn) @ 7100m = 4,300km. I looked into the Goliath scaling a little more to see how it would best compare with the Lotus, and I came up with these values: 1.80m: 300 cheaper, 21kN less nominal 1.90m: roughly same price, roughly same thrust And values in the middle are somewhere between those two extremes. But again, with the Goliath thrust curves allowing the engine to continue producing meaningful thrust above the Lotus's 220m/s limit at a higher altitude. Its only drawbacks against the Lotus is in its physical dimensions and slightly worse fuel efficiency. A Lotus would work on this plane, but it'd also be back to being stuck to the 220m/s speed limit. Also, I toyed around with the new J-34 Chevron engine, but while it's cool, the thrust curves for that engine seem to be even worse than the Lotus's, lol. I couldn't get any faster than 215m/s at a reasonable cruise altitude for a small regional jet. But it sure does help this 64 passenger jet fit the airliner look with the new cabins! The AX-404 also costs ~42,000,000, so not exactly breaking any of the price records, but, as was previously mentioned, that's mainly because the Mk1 cabins are so broken in terms of balance for values (even without fabricating the passenger numbers, they still come out the cheapest for cost-per-seat!). Oh, and this design as-is also only got about 2,500km range, which isn't really anything spectacular either (but that's probably because I always build for very long range). That is off of only 1200 fuel, though, which is good for efficiency I guess. But So might submit it just because it's a solid plane anyways. And since this new cabin is going to generate a bit of talk, might as well refer back to the in-development scoring sheet. I'd put this cabin as comfort rating 2, and push everything above it up one score. Really, it feels like the 'true' economy cabin, that fits along the others price-wise, and isn't an awkward, diminuative size for jet liners. (Big for turboprops though.)

So in combat, what is the cross-section seen by a radar-equipped craft? Is it the actual RCS as seen from the radar-equipped craft's direction?

So how does this 'compromise' work? Is there an explanation somewhere?

I suppose this answers a question I once had (but forgot to ask?) whether RCS was dynamically calculated based on the actual cross-section in a given moment of time while in-game. Cool. Just one question on the whole 'use updated parts' thing. How does that exactly affect craft? I know missiles have had values changed, but for most other things, how does it actually have any effect? If the object is essentially written the same format in the craft file as in a previous version, what is the effect of replacing old parts with new ones?

Well, we already knew the Mk1 had a ludicrously cheap seat price; that's why you made the Slinkies. I personally always felt that the Mk2 should have had 12 seats, but since we've now also done 'cabin luxury level', it's kind of fine. ... That Mk3 is a really heavy deal though. Let's compromise and give the new cabins 14 seats. Though, I guess 16 seats wouldn't be bad either, since we're a bit low on 'economy' options with reasonable capacity and of reasonable size dimensions.

There is the question, though, of how we count cabin capacity of these new parts. But I guess for this case, we don't have to flub numbers? none of my planes have been reviewed yet and I might already start designing a new A-4xx and A-5xx lol

NEW PARTS?! oooooooooooooooooooh

Not yet, according to our records. It's a chronologically recent plane relative to the rest of the queue. Though it's good that you mentioned it, since it isn't present anywhere at all on our judging sheet. Is the P-4-120 the only plane?

Well, you're listed as an editor in the doc, so I don't know what the issue is. As for part count limit, I can consider an expansion to 125 - 150, and I suppose removing weight limit wouldn't hurt either. However, we discussed removing weapon limits and came to the conclusion that laser spam is still dumb, and turrets just aren't in the spirit of this competition.

access sent

Yeah, I've assumed these based off of behavior of things like placing turrets on planes, and how weapons will try to slew onto target individually and only fire if the target is within the aim. These are some very helpful notes; pretty much what I was looking for. So I guess that answers my original question. I was wondering how it interacts with multiple fixed weaponry (assumed pointed in the same direction with 0 convergence/divergence). I had an assumption that the AI takes an average aim point from them, but testing with the debug lines active made me figure that the AI fires guns if a target is within a certain variance from a specific gun's aim. Though, I'm still trying to wrap my head around the AI logic for something like this (not my current project, but good illustrator), with fairly significantly varying translate values for the guns; all the guns fire when the plane is pointed at a target: It also brings the hypothetical of how the AI would react if one side's guns were removed, leaving only the ones on the remaining half. Construction-wise, and why I asked all this in the first place, my current project has boiled down to: is it more effective (read: more compliant with AI behavior) to have a single over-shoulder, offset-to-side position (as in most fighter jets), or two guns mirrored on each side/a single centered gun?

You still don't understand the depth of the questions I'm asking and the answers I'm looking for. Valiant effort for attempting to contribute though. Since you said you have some good fighter planes, let me give you an incremental challenge. Under AI-only dogfight competitions and using no turrets, have your planes beat my planes in this order: Challenge #1 Challenge #2 Challenge #3

So how are multiple guns pointing in the same direction calculated?

That still doesn't really answer the question. I don't need a construction solution, I'm looking for exact AI behavior. Precise AI behavior, as how it is peogrammed. I already build some of the most high-performing combat aircraft, and part of that is learning exactly how the AI works, down to the smallest behavioral quirks, which gives me knowledge of what to exploit when building and tuning the AI, and what to avoid. A slightly off-center gun, in my hypothesis, causes the AI to make minute adjustments to get the gun on the target, despite the gun being pointed perfectly center, instead of just aiming the plane at the target and letting the gun fire. I would test my question myself by pointing a fixed Vulcan off-center by an angle or so, and then try with that mirrored, but I'm not at my computer right now.