Aeronauts , what are your biggest gripes with the stock game/physics?

[AeroGav]

1.  the physics bug that due to bending moments, causes all aircraft to slowly roll left or right in flight no  matter how much yaw and roll stability is built in.

2.  the lack of suitabile autopilot which would make 1) more bearable.   So , you have to manually correct every heading deviation all the way to orbit.

3.  poor support for joysticks that makes hanflying handflying* needlessly difficult.

4.  excess drag from fuel ducts and struts.  Exacerbated by the fact that wing parts don't cross feed fuel

5.  .... and the way rocket engines don't have a "drain evenly from every tank" mode the way that jet engines do.  So your CG gets messed up by rocket motors draining from the front tanks first.

 

Edited May 6, 2016 by AeroGav
*of course flying like the famous smuggler and pilot of the Millennium Falcon should be difficult!

[Red Iron Crown]

I can help with 4, open your Physics.cfg file, find the line "stack_PriUsesSurf = False", change to True. Now radially attached parts are validly connected for fuel crossfeed.

You could fix 5 by editing cfg files and changing the flow modes for individual engines. This would make the above suggestion redundant though, as the drain-evenly flow mode rarely needs fuel lines.

[qromodynmc]

Shape of my plane barely effects my speed, FAR is really good at that, but way too long to set up planes in far.

[AeroGav]

3 minutes ago, qromodynmc said:

Shape of my plane barely effects my speed, FAR is really good at that, but way too long to set up planes in far.

Ah then you should see my other thread.

I really have too much time on my hands don't I  ?   Why don't I just get on and build something dang it !

 

[RizzoTheRat]

Re 2, for SSTO's I always use MechJeb, the Smart ASS tool is brilliant, you set an orientation for your craft (roll, bearing and pitch) and it keeps it pointing that way.  It'll sometimes start doing a bit of pilot induced oscillation but generally that means your aircraft design has some issues, a decent stable plane will usually respond well.  I just set it to heading East with wings level and adjust the pitch angle through the flight

3. is my main gripe, which is partly why I do the above.

 

Edited May 6, 2016 by RizzoTheRat

[AeroGav]

16 minutes ago, RizzoTheRat said:

Re 2, for SSTO's I always use MechJeb, the Smart ASS tool is brilliant, you set an orientation for your craft (roll, bearing and pitch) and it keeps it pointing that way.  It'll sometimes start doing a bit of pilot induced oscillation but generally that means your aircraft design has some issues, a decent stable plane will usually respond well.  I just set it to heading East with wings level and adjust the pitch angle through the flight

3. is my main gripe, which is partly why I do the above.

 

Can you have mechjeb control only heading and leave pitch to the player?   I did try it once but didn't like having to lock my plane into a particular pitch angle.  I'm usually making fine adjustments in pitch trim hunting the optimum lift:drag ratio on the aero data menu on my low thrust:weight designs.    Climb subsonic until AoA required to maintain lift gets excessive, lower the nose to accelerate to mach 1.3, climb supersonic hunting best lift:drag again, make speed run,  activate NERVs, hunt best lift:drag again until reaching orbit.

[RizzoTheRat]

Not that I'm aware of, but if you leave the Smart ASS interface open it's got buttons to adjust by 1 degree, so I fly it on those, climb out fairly steeply to start with but then nose down to keep the angle of attack relatively shallow.  Sounds like I'm flying a pretty similar profile to the one you describe.

Edited May 6, 2016 by RizzoTheRat

[regex]

Fot the stock game?  None, really.  It's pretty darn good for what it is and what it's supposed to do.  If I want a better sim I install FAR.

Edited May 6, 2016 by regex

[herbal space program]

I have encountered 1) and I assume they'll fix that sooner or later. The rest of it doesn't really bug me so much. I think the air is actually a bit better in 1.1.2 than it was in 1.0.5.  I seem to be able to get significantly more dV on LKO now with my latest Mk1 SSTO than I could in that version, although to my preference they went a bit too far making the lower atmosphere less draggy and the upper atmosphere less lifty. I'll gladly keep that over the excessive UA drag from 1.0.5 though.

Edited May 6, 2016 by herbal space program

[HebaruSan]

I'd like the directional SAS modes to be smarter, but I understand the difficulty and it's not a dealbreaker.

[STS Fallacy]

My biggest gripe? The way the "Friction Control" AKA "Spaceplane Messer" works right now. It crashes my craft on takeoff and landing (due to inexplicably blocking wheels and thus spinning them out of control), seems to have a negative effect on wheel strength and doubles as a Kraken bait. More than once, I've had a craft sliding sideways across the runway with the system on, as it somehow thought that it's normal. And the worst thing about it is that if you turn it off, it steals your brakes and replaces them with the ones your kid's bike uses. It's so outrageously bad that sometimes I just turn the plane around on landing and brake with my engines. 

But otherwise, the game seems good to me as it is. Squad has done a really good job. Just fix the wheels and it will be good. 

[Renegrade]

9 hours ago, AeroGav said:

4.  excess drag from fuel ducts and struts.  Exacerbated by the fact that wing parts don't cross feed fuel

^  This.  Very much this.  A single strut has more drag than like a dozen mk1 cockpits directly exposed to airflow.  A flying brick with no struts flies faster than a sleek craft that has struts.  Counter-intuitive for the win!

With the ways things are now, you'd probably be better off with the old aero than the new aero + any significant strutting.

I attached some pictures to http://bugs.kerbalspaceprogram.com/issues/9210 .

This craft:

Is faster than this craft:

by about 50%.  Take those two pairs of struts off and the second craft becomes slightly faster than the first (less than 3%).

The struts actually save the plane though, it typically overheats and explodes after a couple of minutes at mach 2.9~

 

9 hours ago, AeroGav said:

5.  .... and the way rocket engines don't have a "drain evenly from every tank" mode the way that jet engines do.  So your CG gets messed up by rocket motors draining from the front tanks first.

I don't mind this so much, although I DO wish that they'd change it so that two tanks of the same connector type (that includes adapters connecting to the correct types) would partweld together into a single component.  That would reduce part counts (moar performance, less #lolflex) and help CoM for both planes and rockets.   Probably too late in development for that sort of thing though :/

[herbal space program]

8 minutes ago, Renegrade said:

^  This.  Very much this.  A single strut has more drag than like a dozen mk1 cockpits directly exposed to airflow.  A flying brick with no struts flies faster than a sleek craft that has struts.  Counter-intuitive for the win!

With the ways things are now, you'd probably be better off with the old aero than the new aero + any significant strutting.

I attached some pictures to http://bugs.kerbalspaceprogram.com/issues/9210 .

 

 

So is this excessive drag a fixed quantity for each strut or does it depend on how much strut area is exposed to the air?

[Renegrade]

14 minutes ago, herbal space program said:

So is this excessive drag a fixed quantity for each strut or does it depend on how much strut area is exposed to the air?

I believe it's fixed per strut.  I didn't test beyond "this is wrong by MANY, MANY, MANY orders of magnitude" even with tiny struts.

Also, as an added bonus, the strut file itself has something that seems to attempt to REMOVE it's drag cube entirely, but apparently fails:

	DRAG_CUBE
	{
		none = True
	}

(The drag cube determines drag from a given facing in new stock aero)

I'm considering trying to replace the drag cube with a working one from something with really low drag.  I'd aim for zero, but I'd be worried about the code handling that badly... The cockpit-on-top plane shouldn't be 330m/s faster than the four-strut version, dammit!

 

[herbal space program]

6 minutes ago, Renegrade said:

I believe it's fixed per strut.  I didn't test beyond "this is wrong by MANY, MANY, MANY orders of magnitude" even with tiny struts.

 

Well that is certainly just wrong. I'm tempted to see how removing the two struts on my SSTO14 would affect drag, but it flies so poorly without them...

[Rabada]

41 minutes ago, Renegrade said:

snip

I feel like I'm probably being blind and missing the obvious, but I do not see any struts on either of the craft in the pictures.

[FullMetalMachinist]

3 minutes ago, Rabada said:

I feel like I'm probably being blind and missing the obvious, but I do not see any struts on either of the craft in the pictures.

The first one doesn't have any (that's the point) and the second one has two on each wing. They're really short, right where the wings meet the fuselage, one front one back. 

[Renegrade]

Just now, herbal space program said:

Well that is certainly just wrong. I'm tempted to see how removing the two struts on my SSTO14 would affect drag, but it flies so poorly without them...

Prior to the pre-release (1.0.x era), I had one spaceplane that I felt should have worked, but it miserably failed to get anywhere near orbit.  I then saw a post from GoSlash that mentioned "fuel lines and struts = bad for drag" (although it was non-specific about HOW bad) and I removed a bunch of struts and a pair of unnecessary fuel lines, and suddenly it was reaching orbit with 500 m/s to spare.

3 minutes ago, Rabada said:

I feel like I'm probably being blind and missing the obvious, but I do not see any struts on either of the craft in the pictures.

If you look REALLY close at the second plane, you'll see them attached to the main wing from the fuselage.  One's just ahead of the doge flag on the fuselage, and the other is just ahead of the flaps.   They're mirrored, so there's two more on the other side as well, naturally.

[Wallygator]

5 - huh? Fuel drain on rockets should behave that way. This assumes multi-stacked tanks are actually one giant tank. 

6 - my size3decoupler just CTD on staging. Grrr... Still????

[Renegrade]

20 minutes ago, Wallygator said:

5 - huh? Fuel drain on rockets should behave that way. This assumes multi-stacked tanks are actually one giant tank. 

No, they don't.  Bipropellant rockets usually have TWO tanks in each segment that drain at the same time, one above the other, so the CoM doesn't shift quite as dramatically as it does in KSP with a similar layout.  The top tank (be it oxidizer or fuel) will help keep the CoM up.  We get a nasty end-swap tendency in our KSP rockets due to the top tanks in a stage draining first.  Well that and some weird drag values and such.

(ex. Shuttle external tank, Saturn V - S-IC / S-II / S-IVB)

I'd actually hazard a guess that the CoM wouldn't shift at all (relative to fore/aft) if the tanks were laid out horizontally in those rockets.

 

[katateochi]

10 hours ago, RizzoTheRat said:

Re 2, for SSTO's I always use MechJeb, the Smart ASS tool is brilliant, you set an orientation for your craft (roll, bearing and pitch) and it keeps it pointing that way.

I do this as well, not just for SSTO's but most of my rockets are flown to orbit using the SmartASS. Makes for very repeatable flight profiles.

10 hours ago, AeroGav said:

Can you have mechjeb control only heading and leave pitch to the player?   I did try it once but didn't like having to lock my plane into a particular pitch angle.

You couldn't back in the day, but for some time now there is the option to toggle heading, pitch and yaw control on and off in the SmartASS. What I wasn't sure, but I just checked is if it behaved like SAS was one for the unchecked one.  It does (thankfully!).  So you can set heading and yaw, but leave pitch to be manually controlled and then it's just like you have SAS on for pitch (although the SAS light on the navball will be off).

1 hour ago, Renegrade said:

I then saw a post from GoSlash that mentioned "fuel lines and struts = bad for drag"

ARGH!!! That's it!! it's the Goram fuel lines!! Sorry, I thought either I or KSP was going mad. I'd built an SSTO in 1.1.1 Which could easily reach orbit with 1700m/s dV left, it just needed a couple fuel lines so I didn't have to transfer fuel during the ascent. Then 1.1.2 came out, I updated the craft, added the fuel lines and suddenly I could barely reach orbit with 1400m/s dV spare. I thought perhaps 1.1.2 had changed something about the aero, but my other planes worked just as well as before. It didn't cross my mind that the fuel lines would make such a difference.Several days have been wasted on this, trying different ascent profiles and scratching my head as to what was wrong.
I just took the fuel lines off and boom, back up to having between 1700-1800m/s dV once in orbit.  3-400m/s dV difference, just for having 4 tiny fuel lines!! 

 

[herbal space program]

I dunno, I just spent a couple of hours flying my SSTO14b to orbit with and without 2 pairs of struts holding the fuselage sections together. I ended up with 5.7 km/s dV on orbit with struts and 5.9 km/s without. There was definitely a bit of a hit, but I can't believe it was 50%.

[Wallygator]

6 hours ago, Renegrade said:

No, they don't.  Bipropellant rockets usually have TWO tanks in each segment that drain at the same time, one above the other, so the CoM doesn't shift quite as dramatically as it does in KSP with a similar layout.  The top tank (be it oxidizer or fuel) will help keep the CoM up.  We get a nasty end-swap tendency in our KSP rockets due to the top tanks in a stage draining first.  Well that and some weird drag values and such.

(ex. Shuttle external tank, Saturn V - S-IC / S-II / S-IVB)

I'd actually hazard a guess that the CoM wouldn't shift at all (relative to fore/aft) if the tanks were laid out horizontally in those rockets.

 

Ha! I stand corrected! I was fantasizing incorrectly. You are correct. I might just change flow mode on my engines. 

Note to self: must to check if real fuels is similar... Also don't write forum responses after a tiring evening of freezer defrosting!!!

[boxley]

i do wish ground effect was modeled, not just so i can make hovercraft but because it will keep me from smashing my larger planes to bits on landing and give some real purpose to airbrakes

[RizzoTheRat]

5 hours ago, herbal space program said:

I just spent a couple of hours flying my SSTO14b

Glad to see I'm not the only one with a lack of imaginative naming

Edited May 7, 2016 by RizzoTheRat