Losing bits when aerobraking with FAR - solutions?

[eddiew]

Hey guys,

First time I dropped a probe on Eve, it hit the atmosphere at a rather brisk 5km/s. Even with FAR's aerodynamic disintegration turned off, and after several attempts at graceful capture, I had a choice between not getting into orbit, or losing one of my science instruments (apparently it wasn't glued on as well as the others).

Since then, I've been wrapping fairings around them and only opening them up when nice and slow in the lower atmosphere. This works fine - except now I want to aerobrake a 2.5m probe with bulbous projections and I don't have a fairing big enough to wrap around it. Its job is to aerobrake, then land on Gilly, then come home, so by nature it's pretty chunky and I can't really downscale it without losing half the science payload.

I can however put an oversized (3.75m) nosecone on it, which will be wider than the whole probe body, and can detach before the probe widens its orbit after braking. Given that everything on the probe will be sheltered behind this makeshift heatshield, will the game physics actually respect this and not rip bits off?

Thanks!

Edited January 21, 2015 by eddiew

[Cirocco]

I don't run FAR so I'm afraid I'm only of limited help here, but I do have a question: what is the apoapsis height you aim for after aerocapture? I do have some limited experience with FAR, even though I currently run stock and I do know that trying to go for a low apoapsis on the first pass is a bad idea. You should aim for just-barely-captured on your first aerobraking pass and lower the apoapsis with 1 or more consecutive passes into the atmosphere.

Trying to get an apoapsis near gilly orbit in the same pass as braking from interplanetary travel is quite likely to result in unplanned disassembly.

As for using the nosecone, I think that would work, but again, I am no expert on FAR, better have someone else who actually runs it confirm.

[biohazard15]

I think it might work, although I've never landed or aerobraked a probe without some sort of aeroshell. You might also try to raise your Pe (like, 80 km) and use engines to help slow down the probe on the 1st pass - and then just do several passes until you're satisfied with the result.

[cicatrix]

A procedural fairings mod might help you if you do not mind the mod. Actually, if you use FAR I think procedural fairings is a must have mod, but it's up to you.

Alternatively you can try to orient your craft so that nothing 'protrudes' during the reentry or plan maneuvers so that you have lesser velocity when you enter Eve's SOI.

[eddiew]

Oh I'm absolutely happy for just barely being captured on the first pass - I usually do multi-loop captures, even when just returning from Mun/Minmus to Kerbin, because it lets me be more accurate with where the ship will eventually land

My initial problem of in-atmosphere-disassembly was probably down to a rather high arrival speed, which forced that first probe down to about 63km just to get any kind of capture. Hoping that the next mission won't have the same problem, and aiming for a gentler/slower transfer, really I'm just looking to insure myself against being a nublet and turning up way too fast again. I'm a very parallel player and it could be several real-time days between launch and arrival, so I have to get things right first time

I haven't tried procedural fairings; thus far I've managed to fit fragile things into the standard sizes. The nosecone should prevent protrusions vs the angle of insertion, assuming that the craft doesn't flip round (which is a worry, to be fair); I'm just not sure whether the game engine will actually say 'that thing's behind another thing, so air pressure won't rip it off unless the front thing goes first' ^^;

[Jovus]

Sadly, no, that won't work. FAR doesn't model airflow shielding (because that would be a serious pain in the ass). Unless it specifically is shielded by a part capable of doing so, it's exposed to the airflow.

Your best bet is to slap some part(s) on it with low mass but high drag and high structural integrity (like, maybe modular girders? I dunno) and then raise your apoapsis. In the last resort, unless your budget is way tight, it's probably best to only partially aerobrake and use rockets to finish the capture.

[cantab]

FAR includes aerodynamic visualisation options, so you can see which parts are experiencing high drag.

If redesigning the whole probe to fit in a fairing isn't practical, what about putting just the problem parts inside a cargo bay that's part of the probe?

[Renegrade]

First time I dropped a probe on Eve, it hit the atmosphere at a rather brisk 5km/s. Even with FAR's aerodynamic disintegration turned off, and after several attempts at graceful capture, I had a choice between not getting into orbit, or losing one of my science instruments (apparently it wasn't glued on as well as the others).

Wait, what? A science instrument came off? Like a gravioli or thermometer or such?

Are you SURE you had aero structural failures off? O.o

[eddiew]

Sadly, no, that won't work. FAR doesn't model airflow shielding (because that would be a serious pain in the ass). Unless it specifically is shielded by a part capable of doing so, it's exposed to the airflow.

Dangit...

If redesigning the whole probe to fit in a fairing isn't practical, what about putting just the problem parts inside a cargo bay that's part of the probe?

Well, I've got some rocket cargo bays modded in (SXT and MRS both have options I think); maybe I can reconfigure my probe around one of those. It won't be as .... looking though Or maybe I should research 3.75m rocket parts and fairings and brute force it. I have 50 game days until the Eve window opens, which might be enough time to get the points together...

Wait, what? A science instrument came off? Like a gravioli or thermometer or such? Are you SURE you had aero structural failures off? O.o

It was a DMagic RPWS instrument, actually, but basically the same deal. It was mounted at a weird angle, and the g-meter was basically in the red with the speed I'd come in at, both of which probably didn't help any xD And yeah, I always have aerodynamic structural failure turned off. I love FAR for making planes work properly, but I really don't enjoy SSTOs breaking up every time they come home

[Axelord FTW]

I've been using interstage fairing (procedural) as aeroshield and they've been working pretty well so far. By using the strucutral fairing and using the decoupler on the fairing ring part, it's fairly straight forward.

Otherwise, I have been testing cold version of my long range probe using Kerbin as a testbed. That also works well.

Running with RemoteTech, I sometime place an antenna inside a single long fairing on the smallest fairing ring size possible. Either with structural fairings and I left them there or with aero fairing which I eventually drop away to kill some of my mass. In the end, the probe has a fairly, er, suggestive protuberance. I can imagine that placing the science instrument inside the shaft would also works. I need to test which method is the lightest.

eddiew, might we see a few pictures of your probe?

[eddiew]

Tbh, I redesigned the whole thing last night so that it now uses an MK2 body with a cargo bay for all the bits Now it's got a bit of a hammerhead look on the launch pad, and I suspect that FAR with disintegration enabled would pull it apart, but it makes it to LKO, and should therefore survive Eve aerobraking. But sure, I can post some before and after pics when I get home - it was actually interesting to go through a redesign process based on answers to this thread (namely enclose everything you want to keep safe) ^^

[eddiew]

As requested... The exposed science instruments on the mk1 are a real concern. Having lost one before to aerobraking at Eve, I was wondering if that oversized nose cone could protect them. Seems not

... so along comes the mk2, designed while I was a bit drunk. Still very much a work in progress and needs aesthetic attention, but more or less fits what I need into a package with enough delta-v to circularise around Eve, drop onto Gilly (several times), and return home - all while safely shielding its payload.

Looks weird on top of a rocket, but a simulation run made it to space with just an orange tank and a mainsail, so obviously not a real problem. Twin probe cores because... I dunno, it only has vertical thrust anyway, but I guess it looks better with a face Some clipping required, but most of the 2.5m tank occupies the empty space in the cargo bay, so it's not really a double-fuel cheat. Plenty of room for many sciences (mostly courtesy of DMagic Orbital Science). Hopefully a bigger antenna on one end will survive the atmosphere, but there might be room to squeeze one in between the experiments. Worst case it all comes home anyway, as long as the instruments stay intact.

Will fit some shielded/folding solar panels to his back in case the basic ones blow off during atmospheric entry - yet to see the SP line of panels (white cases) ever detach from anything.

*edit* Final form; survived a simulated drop onto Kerbin at ~2700m/s. Transfer stage has nearly 5k delta-v, which should let me do most of my slowing down before hitting the atmosphere

Edited January 21, 2015 by eddiew

[Bryce Ring]

there was a mod that actually deleted the whole file from the physics while it was in the container, I think it was used to help with FPS. I wonder if this would help.

http://forum.kerbalspaceprogram.com/threads/88933

[DBT85]

Maybe try the ADEPT mod. Nice big extending heat shield.

[Axelord FTW]

Wait, is your rocket starting inclined on the launch pad?

Is it more efficient?

I'll try to get a picture of my probe setup tomorrow morning here so you can see what I mean.

P.S. Those crabs are nice. Not exactly a streamlined shape but they are nice looking.

[eddiew]

there was a mod that actually deleted the whole file from the physics while it was in the container, I think it was used to help with FPS. I wonder if this would help.

http://forum.kerbalspaceprogram.com/threads/88933

Interesting, I'll look into that one. Would certainly help when launching a ~60 part rover if I could take it out of the launch calculations. Even my (pretty quick) desktop can be down to 30% speed on a heavy launch.

Maybe try the ADEPT mod. Nice big extending heat shield.

Cheers, will take a look

Wait, is your rocket starting inclined on the launch pad? Is it more efficient?

I'll try to get a picture of my probe setup tomorrow morning here so you can see what I mean.

P.S. Those crabs are nice. Not exactly a streamlined shape but they are nice looking.

The slant is something I started doing because FAR requires a 'proper' gravity turn, i.e. a gentle change of direction from low altitudes. If you try the stock thing of going up to 10k then turning 45 degrees, you'll send your rocket end-over-end, or it'll just plain fall apart if you have disintegration enabled. While you can of course manually nudge it eastwards during a vertical launch, that's got a real habit of inducing roll-wobble. I find that letting the launch clamps hold things on a small incline means I can just put the SAS on prograde, tune the TWR to ~2-2.5, and it'll give me a lovely smooth ascent without any of the twitchyness on the longitudinal axis. Usually gets me to ~20km pointing between 45 and 30 degrees, which generally gets into LKO with the advertised 3200 delta-v, or very close to it.

The Cicada models above both took about 3600-3800, which I'm attributing to the higher drag of the wide frames. Nothing that can't be compensated for with a beefier launcher though!

And thank you ^^ Tbh, one of the reasons I turn off FAR disintegration is because I try to make my ships look interesting/funny, and it tends to get in my way and pull things apart that are in any way adventurous - particularly the Cicada mk2, which gets mounted crosswise to the airflow!