aevionknight

UPDATE: Looks like Creat already got this one, but here's a concrete example anyway. I have a solid way to reproduce the problem in the VAB. I've also experienced a frozen Engineer in flight, but I don't have a case for that. Hopefully they are the ultimately the same problem, and fixing one fixes both. STEPS TO REPRODUCE: 1. Unzip a fresh copy of 0.23. 2. Install Engineer 0.6.2.2. 3. Install KW Rocketry 2.5.52 (listed as 2.5.5 on spaceport). 4. Load up this craft file: http://www./view/pjgvc892xc1lc2g/KW_Engineer%20Broken.craft 5. Notice that Engineer is working - add a fuel tank on top of the pod. Do not touch anything else. 6. Right-click on the engine (a KW Vespa). Engineer stops updating. I was not able to reproduce this trying a couple of stock engines - it seems to be related to KW specifically.

Hmm, I\'m pretty sure I tried it with the latest. Can you be more specific?

Nice-looking spreadsheet. Oh, and look, a shuttle too. : )

I feared this with Cyclone until I made a spreadsheet and finally got the correct formula for drag. Then I could just plug in numbers all day and tune it without worrying about breaking things. Checkpointing with revision control also helps.

Pitch/yaw is rotation. It\'s broken on roll/pitch/yaw, all of them.

Hmm, looks like I didn\'t play with the chutes while re-entering enough. They move your center of mass far enough back that, combined with the bigger pod torque, you can now flip the Cyclone during re-entry. It doesn\'t break anything, just makes it easier to control than I wanted. I guess the chutes both make it tamer to fly and give you a non-propulsive landing option. If you want to be hardcore still, fly without the chutes. : ) I\'ll probably tune that behavior out in the next release along with taking another pass over the textures. That won\'t be for a couple weeks, though. For now, enjoy! And let me know if there\'s anything else you\'d like to see changed, fixed, improved. Keep in mind: RCS doesn\'t work for rotation due to a KSP problem. The flaps cause madness when using the roll keys (Q/E), also due to a KSP problem. Can\'t fix those until HarvesteR does.

Just finished Cyclone 1.1: [li]Added surface attachment to all decouplers.[/li] [li]Added a 1.75m decoupler.[/li] [li]Added an emergency chute that attaches to the Cyclone baseplate. Four (4) required for safe landing. Will only work for water landings![/li] [li]Reworked crash tolerances to avoid part exploding as much.[/li] [li]Upped pod torque to make rotation in space / entry less painful.[/li]

Just a typical hemispherical chute. I finally got the things work work after digging through the .DAE file. I\'ll model non-placeholder ones tomorrow.

I whipped up a quick parachute system. I think I\'m going to have the parachutes only be safe over water. If you want to land on dry land, you\'ll have to use the engines. I also upped the pod torque, it handles a lot better in orbit now (especially on roll) and the flight dynamics aren\'t affected. RCS is still useless for roll, I\'ll revisit it when Harv fixes the bug. [uPDATE: Hmph, having trouble getting the parachute to look right in the half-open position. It basically has no radius at the moment.]

tl;dr Bricks in spaaace: RCS is pointless for roll/pitch/yaw in 1.0. Just turn it off and only use translation (IJKLHN). I can either boost RCS thrust and give up translation, or try and boost pod torque without destroying flight dynamics. What do you guys want? Landing: Sounds like parachutes are a want. I\'ll see to it. I\'ll also post a video of a landing as proof that it\'s possible. It looks like this is difficult enough to make an interesting entry on the challenge board. : ) Textures: The decouplers are crap, I wanted to release Cyclone before I got tied up in holiday stuff. I\'ll fix them at some point. Don\'t know if I agree about the brown color, but I\'ll see if I can spice it up a bit. Exploding pod: I don\'t know why you want to look at the pressure vessel without its skin, but I\'ll stop it from exploding. : ) File host: I\'ll find a better one. [uPDATE: all transferred to MediaFire now] RCS Woes Unfortunately RCS attitude control is broken. The WSADQE keys produce far less thrust than they should (compared to IJKLHN). This means you have two options: 1. Have small RCS thrust. This gives you nice fine control for linear translation but makes it pointless for roll/pitch/yaw. 2. Have huge RCS thrust. This makes it possible to turn on RCS but makes linear translation simply outrageous. a. If you have a little fuel, translation will simply drain your tanks super fast and is useless. b. If you have a lot of fuel, RCS can .I chose #1. To make up for it, I gave the pod a little torque, but that brings me to... Brick pod For the Cyclone, it was important that the reentry flaps work nicely. This means I can\'t have torque like the mk1pod, since that allows you to flip and spin without a care in the world - adding wings is completely superfluous. I wanted to have an unpowered pod and just got with RCS, but that didn\'t work (see above). Changing this is a delicate operation, but I\'m willing to do it to make flying less painful until KSP RCS gets an overhaul. Flaps don\'t roll, son Geometrically, if you look at the flaps, there\'s no way they can cause roll. However, KSP doesn\'t seem to care, and tries to use them to roll anyway, leading to pitch/yaw madness. I\'m going to file a detailed bug on this today, but we\'re sort of stuck with it for now. Hopefully a less brickish pod will make this less of an issue.

Cyclone is released! Grab it on the original post. The only major change from what I\'ve said before is that the propellant load was de-tuned slightly. Starting from a 100km x 100km circular orbit, you can only safely boost to a 350km x 350km and have enough to de-orbit and land. Also cut down the RCS load somewhat. Stuff I\'d like to hear from you guys: [list type=decimal] [li]How does the handling feel? Too much / little lift? If there\'s too little you\'ll feel like you\'re dropping like a stone, too much and you\'ll feel like you\'re floating.[/li] [li]How is the propellant load? It\'s designed to feel 'constrained'. Let me know if it\'s too draconian.[/li] [li]Any textures/model work you think needs more attention?[/li] [li]Finally, I toyed with the idea of creating some emergency chutes in case you burn too much propellant in orbit. Shouldn\'t take long, is that something you\'d be interested in?[/li] I\'ve been playing with this thing for two weeks so I\'m pretty good at it, it\'s hard for me to tell whether others will find it too easy/hard to fly. Feedback is appreciated. FYI, this is how much fuel you need to land safely. Start your landing burn no earlier than 1000m altitude or you\'ll probably run out. I recommend killing most of your horizontal velocity with the main engines, then pointing up, engaging SAS, and killing the rest with RCS. I\'m going to post a flight manual and some glamour shots to the original thread tomorrow, and I think I might whip up a challenge or two for fun. Clear skies and good flying! [EDIT: Putting the Cyclone together is pretty easy, but here\'s the .craft for the impatient.]

Everything\'s textured, doing final tuning pass now. Then a quick scrub of the flavor text, missing .cfg params, etc. [uPDATE] Final retuning done. Scrubbing .cfg files. [uPDATE] Scrub done. Doing final flight tests.

You need about 0.2m of collider below the origin in order to avoid insta-kills.

Alright, I can\'t resist showing this off: You can see the engines have been conveniently omitted. The plumbing detail behind the RCS thruster quad and reentry control flap. Stuff left to do: [list type=decimal] [li]Texture the engines.[/li] [li]Texture the decouplers.[/li] [li]Final balancing pass, now that we have a Kerbal Theory of Drag.[/li] [li]Final .cfg cleanup / scrub.[/li] [li]Release![/li] I think I can do all of that in one day, though it may slip to two.

In theory, anyway. Doesn\'t do much of that right now.

I belive the RCS groups identify axes of thrust, not the numbered thrust vectors. So rcsGroup0 is +x or whatever (don\'t know what the actual mapping is).

Version control, man.

Hah, I was trying to be less spammy with my updates. Work has been quite busy so I\'ve only been able to put a couple hours in. Got a bit of progress on the fuel tank, maybe 50% done now. Most of the final assembled surface area has at least some non-orange texture now and I\'m quite pleased with it. Probably won\'t post any pics until release. I tried splitting the flaps, but it didn\'t help. Control surfaces oriented in this direction clearly weren\'t anticipated by the code, so we\'ll have to live with the roll weirdness until I can get Harv to look at it. I\'ll post the bug once I\'ve released the mod so he can test it. Also spent some time with the new release to see if the RCS control problems had been addressed, but no such luck. Filed a bug on it, hopefully Harv will have time to look at it this cycle. Still looking at a couple days from a release.

I\'m not sure if you\'re trying to match the real Dragon trunk or not, but the trunk carries cargo, not fuel or an engine.

It\'s actually worse than that, because the collision mesh for the terrain is below the visible terrain (that\'s why landing legs appear to sink into the ground), but the auto-kill detection happens on the visible terrain. You need to have a good 0.2m of collision mesh below the origin of the part that\'s touching the ground to avoid insta-kills.

As expected, didn\'t get through everything today. I textured the RCS tank and am about 40% of the way through the fuel tank. That leaves: [list type=decimal] [li]Finish texturing the fuel tank.[/li] [li]Texture the RCS thrusters.[/li] [li]Texture the main engines.[/li] [li]Texture the decouplers.[/li] I\'ve decided to re-model the re-entry flaps. Right now they have no way to affect roll. This would be fine, except that they *think* they can affect roll, and respond in crazy ways. So if you hit the roll controls while deep in re-entry, you\'ll get wild pitching and yawing, which really sucks. I\'m going to split the flaps into two so that they will hopefully give some roll authority and stop the madness. I can also make them a little more interesting than just a square while I\'m at it. Getting real close now. The fuel tank is the last big texturing job, the rest should go quickly.

You put my lander on a lander. Dawg. And I will name it Directest Ascent.

Heading to bed for the night, quick progress report: [li]Modeled 1m and 2m aero decouplers, they work great. They have almost no ejection force, so give the forward RCS a few bursts to clear the stage (this look pretty neat).[/li] [li]Textured the heatshield, CMG, and pressure vessel. Also laid out the UV map for the RCS tank.[/li] [li]Nailed down the propellant budget. The 'reference mission' for Cyclone is: [list type=decimal] [li]Inject Cyclone to a 100km circular orbit and separate from the booster.[/li] [li]Use Cyclone main engines to do a Hohmann transfer to a 500 Km circular orbit.[/li] [li]Do a de-orbit burn from 500 Km to a 30 Km apogee (well within the atmosphere).[/li] [li]Perform lifting re-entry, holding approximately 20 degrees angle of attack.[/li] [li]At 1000 m, light engines for a pitchover, horizontal velocity null, and propulsive landing.[/li] [/li] Cyclone can accomplish this with a few seconds of fuel remaining (and pulling less than 3 gees on entry to boot). The RCS can also give you about 150 m/s of delta-v for trimming your orbits or killing small horizontal velocity on the landing. For bigger missions (going to the Mun, geostationary, etc) you\'ll need to provide a bigger propulsion module to handle some of the lifting. It\'s possible I\'ll have something to release tomorrow but it will likely be early this week.

I\'m not sure I quite understand your problem. Here\'s what I did: [list type=decimal] [li]Gave my capsule body four nodes called [tt]node_stack_rcs1[/tt] through [tt]node_stack_rcs4[/tt].[/li] [li]Gave my capsule body [tt]stackSymmetry = 3[/tt][/li] [li]Gave my RCS thrusters a node called [tt]node_stack_top[/tt].[/li] [tt]stackSymmmetry[/tt] is vital, and your nodes need to start with [tt]node_stack_[/tt]. I believe it\'s also important to number your symmetrical nodes 1 - N with the same base name. Best of luck.

I readily admit to stealing the shape of the EXPERT RV, I love the look of it. By 'fin kind', do you mean a linear aerospike engine, like the one that was developed for the X-33? In that case, I did look, but I decided against it: 1. I wanted a docking hatch in the center of the baseplate, so couldn\'t put part of an engine there. 2. The engines are used 90% in space, with a short burst for landing. An aerospike engine requires a gas generator and other complexity, when you really just want a lightweight pressure-fed engine. The altitude compensation was also unnecessary. All parts are stacked. The RCS blocks just have an attachment node in the +z axis, rotation isn\'t required.