blakemw

Members
  • Content count

    129
  • Joined

  • Last visited

Community Reputation

76 Excellent

About blakemw

  • Rank
    Spacecraft Engineer

Recent Profile Visitors

828 profile views
  1. how to get this into orbit.

    It could be that MJ is assuming you are going to decouple some mass which according to MJ's calculations is dead mass (like a spent booster), this can especially happen when the root part isn't in the final stage, something which always makes dV estimations wildly inaccurate.
  2. how to get this into orbit.

    Complex vessels can be confusing for deltaV calculators. They are not perfect. That's one of the merits of very simple serial staging, you'll get deltaV values of utmost trustability.
  3. how to get this into orbit.

    The way I would do it is to add a pair of boosters to the side of the vehicle so it can deliver itself through some 1000m/s or so (note: I always do tail-first landings, and land using a procedure I call a "hover-tip" which is what it sounds like, works well on low-gravity worlds, so the boosters should point tailwards rather than down). Then I would strap the entire thing to the *base* of a Twin-Boar SSTO, at the base it won't cause center-of-pressure problems and the Twin-Boar will have no troubles delivering it to orbit, probably Mun's orbit, a Twin-Boar can get quite a bit dV out of a single stage if the payload mass is negligible. If you want extra dV can add some hammers to give the launcher a bit of a kick off the pad, they get decoupled before the drag starts to ramp up. Btw I have a suspicion the probe core has the wrong rotation, vehicles kind of work when the probe core is facing up, but they work a lot better when it's facing forward.
  4. Low lunar/munar flyby

    Once you have the perilune you want make a small adjustment to the timing of the burn (like 0.1s) - this will change perilune (because the ship ejects at a slightly different angle and so will pass Mun closer or further), then adjust the prograde magnitude of the burn (in increments of like 0.1m/s) so that perilune returns to what you want it to be, through this process perigee will have changed, either getting higher or lower, make note of whether it go better or worse, and either do the same thing again, or the opposite thing. If you get both the timing and the prograde magnitude exactly right you won't need ANY burn at perilune, in practice the level of precision is too great to do in a single ejection burn from Earth, but a single ejection burn plus a correction burn to fix the error (might involve a radial component) should be enough to put you on a free return all the way back to Earth's atmosphere. It is much cheaper to make corrections when just leaving Earth than later on.
  5. SpacePlane Reentry?

    Passive radiation still works during reentry due to some quirks of the heating model (we're going to go down the rabbit hole here). Basically whatever isn't directly exposed to the oncoming air (is in a heat shadow of another part) is icy cold so parts (or the part of a part) which aren't exposed are still radiating away heat just fine. Now that would make radiators really good during reentry, and that would make it obvious that the heating model is borked. So there's a hack: active cooling for radiators/thermal control systems is basically just switched off during re-entry, it's not that the environment is too hot to radiate heat into, it just arbitrarily stops working (well not exactly arbitrarily, but the point is it's not the laws of thermodynamics that stops it working, it's a hack which goes "looks like re-entry! better switch off the active cooling of radiators!). Nevertheless, radiators thanks to their large surface area can still passively radiate away heat into the icy cold of the heat shadow, they just aren't actively sucking heat out of other parts, being limited to normal part-to-part conduction which is not as effective. Though more likely a radiator will actually have a greater effect from the extra drag it causes than passive heat radiation, they have a pretty extreme "drag to mass" ratio and if you want your plane to slow down better than it flies, add radiators (or for the best of both worlds, put them in a cargo bay). Note that in KSP the plasma wake does not generate heat, IRL it bathes the vessel in infrared radiation but in KSP it's a purely visual effect.
  6. Another good approach is to use the KSP built-in debug/cheat menu to menu the vessel into the target world's orbit, then use HE to do whatever you need. KSPs built-in system does a better job of doing everything by the numbers.
  7. When is a parachute safe to open?

    I've also observed the two peaks, I think the first peak is from slowing down to subsonic velocity (relative to the atmospheric density), and the second peak is from hitting denser atmosphere.
  8. SpacePlane Reentry?

    My usual approach is to set Pe at.... whatever! (seriously, -100km to 30km) but then pitch up while descending to stay above 30km until bleeding off some speed, often a pitch of around 30 degrees is appropriate. It must be said you do yourself a massive favor if you can use a mod like Mechjeb or Atmospheric Autopilot to maintain an appropriate pitch without constant babysitting. I use rocket parts (rather than aircraft parts) all the time with no problems when using this kind of reentry. It's also possible for a sufficiently light plane to "skip" across the upper atmosphere about halfway around Kerbin, at an altitude of around 40km, then once it is has skipped its way to the KSC continent just pitch down severely and glide to the runway. Temperature tolerance is only really important for the parts of the plane that takes the brunt of the heat - the nose and leading wing surfaces - everything else is at least partially shielded by the stuff in front of it which makes a huge difference to heat loading. It's a bad idea to have anything with an external or internal temperature tolerance of 2000 or less taking the brunt of the heat at leading surfaces, but behind the leading surfaces you are far more free in your choice of parts. It is wise to never use the pointy cockpits (especially Mk1) because they are extremely prone to internal overheating only use the inline cockpits, and then for the nose make it something with temperature tolerance; shockcone intake or aerodynamic nose cone aren't ideal but do tend to suffice, fairing is ideal, small nose cone is bad it will burn off due to low thermal mass.
  9. Whats the use for probes/rovers?

    The main reason to use probes is if you're playing hard difficulty (and/or with hardcore life support mods) and don't want to lose kerbals. ISRU rover can be great, I like to park one on the flats of Minmus to fuel up larger vessels (anything can be landed on Minmus, and it's a lot less work to fuel thing up in place than to move fuel around). Rovers can also be handy in the MKS/USI mod suite.
  10. Why are rovers so hard?

    Another factor is center of mass, while not perfectly relevant to KSP rovers, it's instructive to watch this really short video about center of mass: If your rover is nose-heavy it'll basically want to keep going nose-first, this means it tends to understeer (you try to steer, and it wants to keep going in the same direction as its already going) But if your rover is tail-heavy, it'll basically want to go tail-first, this means that it'll take any excuse (such as loss of traction) to try and spin around and go backwards, which will usually result in flipping as the wheels on one side dig in and try to resist the rotation. If it's otherwise stable it'll tend to oversteer (when you start steering, it'll be over-enthusiastic about turning). Obviously being mildly nose-heavy is a big stability advantage for a rover in a low-friction environment like a world with low gravity (though being too nose-heavy might encourage doing a nose-dive style of flip). Moving away from the real world, in KSP you can tweak numerous properties of wheels (if Advanced Tweakables is enabled). The most useful for stability is "friction control" which determines how much wheels "bite". Leaving friction control at auto is a bad idea! Auto is like "stupid" or something. Setting friction control to 0 makes wheels frictionless which is great for sliding around and not flipping, but you can't accelerate, brake or steer... Setting friction control to 5 will make the wheels toss the rover around, any loss of traction will cause severe forces on the rover. Usually, the best handling is found with a friction control of 0.5-0.8. Happily you can even adjust friction control on the fly, the main thing you lose with a low friction control is braking power (it's not such a problem in terms of acceleration), so if you've set friction control to a low value for improved stability, you can just crank it up for improved braking, like to make the rover stay still on a slope. It also tends to be helpful to have a lower friction control on front wheels than rear wheels, because the front wheels are more prone to cause flipping.
  11. Low lunar/munar flyby

    This is totally possible. I'll assume you're in a circular orbit around the Earth, at 303km (the lowest orbit cheat menu will put you :P) and matching inclination with the Moon. Here is an example of what it looks like: This kind of gravity assist is super "threading the needle" type business (especially when it comes to the resulting perigee), so a couple of few m/s correction burns will be needed to get it completely perfect, but it's a basic free return with a low perilune and a perigee near Earth's atmosphere. A few m/s in correction burns will make it perfect. The most important thing is the amount of overshoot you have relative to the Moon's orbit, if I rotate the maneuver node about the orbit to get the Moon out of the way, this is about how much overshoot there should be: The free return works because essentially the Moon performs an orbit-lowering gravity assist (by passing behind the moon), so first you need to overshoot to give the Moon's gravity something to lower. If you have too much overshoot (i.e. hyperbolic) you might be able to still achieve your goals, but you'll end up in a retrograde orbit around Earth after the gravity assist rather than a prograde orbit, getting the prograde orbit is unbelievably sensitive, in contrast, it's relatively easy to get a retrograde orbit (this will reduce trip time to moon though). If you have insufficient overshoot you won't be able to get both a low perilune and perigee. Since it's super-sensitive to the exact timing and the exact deltaV of the ejection burn, some kind of maneuver node editor (which I didn't actually use here) helps though you're still going to need at least some tiny correction burns. And of course, because the Moon's orbit is elliptical those exact numbers aren't going to work for you unless the moon is in exactly the same place! Nevertheless starting with about that amount of overshoot should save some time in finding the exact burn required…
  12. Reversing orbital direction?

    The cheapest way is to leave the SOI then re-encounter the Mun again about a third of the way around Kerbin, now with the desired inclination. This eliminates the 133m/s orbit-reversing burn, at the cost of ~2m/s to completely exit Mun's SOI. Generally this can be done by just getting the minimal escaping ejection, then changing the timing (i.e. pulling the node around the orbit) so the ship re-encounters, then tweak the exact timing and the exact amount of prograde dV so it ends up retrograde instead of prograde, it's not nearly as hard as it may sound since it's really, really easy to re-encounter the Mun. In theory this can all be setup with the one burn, but a minuscule correction burn will probably be required because encounters at low relative speeds are extremely sensitive.
  13. Optimum altitude to jettison fairings?

    Techcanilly, yes. : thrust * fairing_mass / rocket_mass kN of thrust is lost accelerating the fairing popping the fairing will increase drag, also measured in kN Since these numbers are directly comparable, once the rocket is wasting more kN of thrust accelerating the fairing, than the fairing is reducing drag, it is time to pop the fairing. However often in KSP you will complete the main burn to nearly gain orbital velocity, while still deep in the atmosphere ( < 40km) and as such there is no advantage in jettisoning the fairing until out of the atmosphere since you're just coasting. If I am using a steep trajectory I'd usually pop the fairing at 50-55km - I have a hunch it'd often be profitable to do so earlier (depending on payload) but it's one of those "better safe than sorry" things, keeping the fairing on for longer than necessary is just a small penalty to deltaV, jettisoning it too early can cause serious problems.
  14. Are fairings broken in 1.4.2

    Yah this is quite well known and is apparently being fixed in 1.4.3. In 1.4.2 Fairings are invisible to the aero model, I exploited this to put a fairing in orbit at a very low altitude (just above the ocean!), they experience no drag, no lift, and offer no streamlining to the part they are mounted to.
  15. Yeah I was mainly replying in the context of stock engines which are made completely redundant by MH engines, not so much whether the new engines are good or not. Maybe the Rhino should be added to the redundant list, clusters of Wolfhounds badly outperform it for any payload where a Rhino would be justified, granted the Rhino has higher power-area density if you don't like clipping engine bells together, but vacuum stages can often get away with not being sleek. Remember, it has variants. The bare variant is only 1.875m. Which means, unlike the Mainsail, it works well for clustering on bigger rockets. Really? When I try to use it it'll only ever give a 2.5m shroud regardless of variant. It's true it's a bit smaller visually though. With stack sizes there seem to be three different things: 1. The available shroud sizes, for example, Mastodon only has 2.5m, Skiff has 1.875 and 2.5m. 2. The available base sizes, for example, Skiff variants actually fit only the 2.5m and 1.25m parts: you need to use a 1.875 to 1.25m adapter piece to fit the Skiff smoothly to a 1.875m stack. 3. The visual size, how good it looks attached to a certain stack and how much it visually clips. Largely irrelevant to physics since the game doesn't mind a bit (or sometimes even a lot) of clipping. Like the Skiff will clip a little if you stuff it into a 1.25m tube, but it doesn't look disproportionate on a 1.25m rocket.