Jump to content

carde

Members
  • Posts

    15
  • Joined

  • Last visited

Posts posted by carde

  1. Cancer rates would skyrocket as well as spontaneous abortions and lots of really deformed and sick babies would be born.

    That's why it will never find its way into the world of commercial application, even if non-fissile isotopes are used..

    There is not enough radioactive material available in the world to cause noticeably large increases in background radiation (granted, we are doing our best with mined coal, but...).

    Also, mutations are not actually easily generated using radiation. Cancer, to some extent, yes, as an example a deadly dose or radiation has a 1 in 10 chance of causing cancer (of course, you are dead from the direct radiation poisoning in that case, but let's say you get 1% of a full-body ld50 dose, you now have once extra chance in 1000 of getting cancer during your life)

    At least if the Linear no-threshold model is true.

  2. Then we need to bug the makers of PhysX to make an API for multithreading in unity. ;)

    PhysX is fully multithreaded, and even works on GPU:s (for about a 10-100x performance increase using low-level cards, more with high-end (eg, titan class) cards).

    However, unity has choosen to remain with a very old version. And they modified it so it only uses one thread (even that version of the original PhysX SDK was actually multi-threaded, and also uses the GPU if possible)

  3. Has anyone really been far even as decided to use even go want to do look more like?

    I can probably do this challenge, but I still fail at returning to Kerbin from a far removed solar orbit.

    I might have misparsed your post, but:

    Simply do an orbit that has a period that is is an integer multiple of kerbins. If you have mechjeb to show your orbital period that is easy, otherwise try to thrust along kerbins orbit and look at your apoapsis and periapsis times.

  4. You can fairly easily do it using stock parts.

    My example:

    .craft: http://per.hedbor.se/probe1.craft

    Both the capsule and the "probe" is escaping the sun. I have mechjeb on it because I wanted a separately controllable probe, but it also makes the launch easier (and significantly less tedious, the boost is 24 minutes or so long).

    The trick is to do most of your boosting in low kerbin orbit, as previously mentioned in the thread.

    The easiest way is to launch as normal, start tilting over at 10k or so, and once you get the "orbit" indicator follow that the rest of the burn (this is made easy using mechjeb, just click the 'prograde' button)

    Using the high specific impulse engines helps too. The aerospikes are not vectored, though, so you have to be extra careful when launching.

    My launch in the movie is rather far from optimal, if nothing else i did not end up boosting in the correct direction, I was about 20 degrees off, so I lowered the periapsis some.

  5. You might want to work on the scoring.

    The "bonus" points are very low, really. There is no advantage in going for them.

    My first try gave me about 5.94km/s separation speed (with more than 3km/s dV left in the "probe".) -> 5940 points

    Minmus was reached in 1h46m (with the probe, after using 1km/s dV) and 2h with the booster.

    So 1334 or 1320 more points.

    (The time to reach minmus is really the same thing as the speed)

    I easily reached escape velocity. That seems to subtract significantly from the score, though. Since there is no apoapsis :( (or, well, at there is no maximum altitude above the sun).

    By the way, unless you aim in the wrong direction the speed is also basically the same thing as how far above the sun your apoapsis ends up. So all the scores measure the same thing.

    Except one:

    The rocket cost 32350, unless I miscalculate, so 9676 points there.

    <a href="http://per.hedbor.se/probe1.craft">.craft available</a>. All stock except for mechjeb on the probe.

    I have a <a href="

    ">movie of the whole launch</a> if anyone would like to see that...

    I did mess the launch up somewhat, it could almost certainly be done better using the same design. But to optimize <i>score</i> I think the most important part is to have a small rocket, since the $1M budget is so huge (also, avoid reaching escape velocity, unlike me...)

×
×
  • Create New...