Squad on the recent Aero changes

[Fuzzy Dunlop]

Ok I still have 1.0 so did some testing. I used both the 1 man pod and the 3 man pod - both with (physics fixed) heat shields and parachutes.

On a 75km x 25km re-entry trajectory both slowed to less than 300 m/s by 4km - terminal velocity was about 270-280 m/s.

On a vertical suborbital launch up to 75km the pods slowed to less than 300 m/s before I had to deploy the chutes to prevent a crash (so about 500m).

I think this is pretty reasonable. Yeah some pods in real life do have a lower terminal velocities, but it's a matter of degree. In KSP 1.0 you should always have time to deploy the 'chutes at subsonic speeds before you hit the ground (unless you have a really high suborbital trajectory, or de-orbit straight into the ground - but there's no reason to do that).

Something heavier and more areodynamic than a pod might hit the ground much faster; this is totally realistic - a V2 rocket, mostly empty tank, would impact at 800 m/s (half its peak velocity) from 100 km high suborbital trajectory.

A more pressing issue occurs if we make pods slow down more, as in 1.02. Basically too much drag high in the atmosphere kills too much of the pod's speed preventing heating (which isn't linked to drag in the games physics). This makes heat shields pointless since in 1.02 a naked pod can survive a return from essentially any altitude.

[mikegarrison]

IRL they don't fall at Mach 2.

IRL they fall at hypermach speeds. http://www.grc.nasa.gov/WWW/BGH/hihyper.html

[Tarheel1999]

Yeah, I don't think Kuu was talking about speed at re-entry which is what the linked page mentions but rather speed in the middle to lower atmosphere.

[Jouni]

I think this is pretty reasonable. Yeah some pods in real life do have a lower terminal velocities, but it's a matter of degree. In KSP 1.0 you should always have time to deploy the 'chutes at subsonic speeds before you hit the ground (unless you have a really high suborbital trajectory, or de-orbit straight into the ground - but there's no reason to do that).

It's nowhere near reasonable. With 1.0 aerodynamics, the pods will crash into ground before they can slow down enough to reach speeds, where it would be reasonable to deploy drogue parachutes.

A more pressing issue occurs if we make pods slow down more, as in 1.02. Basically too much drag high in the atmosphere kills too much of the pod's speed preventing heating (which isn't linked to drag in the games physics). This makes heat shields pointless since in 1.02 a naked pod can survive a return from essentially any altitude.

If a part is designed to withstand temperatures up 2000 °C, then it doesn't require any additional shielding for a low-speed reentry. Unshielded parts should have heat tolerances in the 100-300 °C range.

[Enceos]

Here's the math of real physics:

The formula for drag force is

F_D = ÃÂ C_D A V² / 2

where,

F_D = drag force

ÃÂ = air density

C_D = drag coefficient

A = area normal to the flow

V = velocity

Since the capsule is falling at terminal velocity, the drag force equals the weight of the body. The Apollo capsules had a mass of about 5,000 kg at splashdown, thus their weight was about 5000 x 9.8 = 49,000 N. Assuming the vehicle is falling heat shield down, the area is 12 m² (3.91 m diameter). The Apollo capsule drag coefficient was about 0.85 at subsonic velocity. Air density at sea level is 1.225 kg/m³. Plugging in all these number, we get

49000 = 1.225 x 0.85 x 12 x V² / 2

V = 89 m/s (200 MPH) terminal velocity for the Apollo capsule at sea level.

This assumes the the parachutes didn't deploy at all or were torn away from the capsule.

Apollo capsule has two drag coefficients:

1.3 C_D at hypersonic velocity.

0.85 C_D at Mach 0.4.

[Fuzzy Dunlop]

It's nowhere near reasonable. With 1.0 aerodynamics, the pods will crash into ground before they can slow down enough to reach speeds, where it would be reasonable to deploy drogue parachutes.

No they won't. I did test this. Unless you're plunging vertically through the atmosphere your pod will always slow to bellow 300 m/s which is a perfectly reasonable speed to deploy drogues; it's not massively faster than the soyuz drogue deployment speed.

Also drogues are occasionally used at supersonic speed (Felix Baumgartner used one to stabalise him on his supersonic skydive). Nasa even uses gigantic supersonic parachutes, for example during curiosity's decent.

[KatzOhki]

@OP:

Can you put the transcript / summary in the first post when you're done with it? Would really appreciate it, thanks.

[sephirotic]

Ok I still have 1.0 so did some testing. I used both the 1 man pod and the 3 man pod - both with (physics fixed) heat shields and parachutes.

On a 75km x 25km re-entry trajectory both slowed to less than 300 m/s by 4km - terminal velocity was about 270-280 m/s.

On a vertical suborbital launch up to 75km the pods slowed to less than 300 m/s before I had to deploy the chutes to prevent a crash (so about 500m).

I think this is pretty reasonable. Yeah some pods in real life do have a lower terminal velocities, but it's a matter of degree. In KSP 1.0 you should always have time to deploy the 'chutes at subsonic speeds before you hit the ground (unless you have a really high suborbital trajectory, or de-orbit straight into the ground - but there's no reason to do that).

Something heavier and more areodynamic than a pod might hit the ground much faster; this is totally realistic - a V2 rocket, mostly empty tank, would impact at 800 m/s (half its peak velocity) from 100 km high suborbital trajectory.

A more pressing issue occurs if we make pods slow down more, as in 1.02. Basically too much drag high in the atmosphere kills too much of the pod's speed preventing heating (which isn't linked to drag in the games physics). This makes heat shields pointless since in 1.02 a naked pod can survive a return from essentially any altitude.

I totally second this post, as a matter of fact, as I've stated before: EVEN ON 1.0 THE DRAG IS STILL TOO HIGH FOR SPACEPLANES AT LOW ALLTITUDES.

This becomes obvious if you think that Kerbal's atmosphere has only 60% of the equivalent height of earth's, so the pods have much less time to decelerate.

It's nowhere near reasonable. With 1.0 aerodynamics, the pods will crash into ground before they can slow down enough to reach speeds, where it would be reasonable to deploy drogue parachutes.

You are not getting the bigger picture: 250m/s IS REASANOBLE to depley a drogue chute, curiosity's droge was deployed at Freaking mach 1.7. You can´t expect a deceleration identical to real life by having a atmosphere that tops out at 70km. Progress that just deorbited by drag yesterday, decayed because IT SUFFERED DRAG AT 170km above sea level! So you can´t have two words: Perfectly reallistic chutes, perfectly reallistic drag levels, perfectly reallistic drag for airplanes at low alltitude. THERE HAVE TO BE COMPROMISES>

The question is: Do we want to think that the first 10km of alltitude, are just like earth so planes fly simmilarly to real life, or do we want to have pods reaching the exact same terminal velocity at 2km alttitude? We can´t have both without some tweakings and compromisses, The onle way we can do the latter without messing with other things is by having a very condensed unrealistic soupy atmosphere. The community ALREADY SAID THEY DON`T WANT THAT IN THE PAST.

We want a better balance: Planes that fly reallistically with low drag bellow 10~15km alltitude, and pods that reaches subsonic speed at 2km.

How is squad going to do that? I don´t know, but the 1.0.2 solution IS CLEARLY BAD. Maybe they can Increase the drag of pods and non cylindrical parts, but don´t mess with excessive drag, Is it so hard to do that? And is ist so hard for people defending the 1.0.2 aero understand that?

[Jouni]

No they won't. I did test this. Unless you're plunging vertically through the atmosphere your pod will always slow to bellow 300 m/s which is a perfectly reasonable speed to deploy drogues; it's not massively faster than the soyuz drogue deployment speed.

I said "reasonable", not "works in the game due to its simplified physics". Soyuz deploys drogue chutes at 230 m/s / 9 km, which is roughly the same as 230 m/s / 5.3 km or 165 m/s / 2 km in KSP. When air density increases by 2x, the safe deployment speed of parachutes decreases by about 30%.

Also drogues are occasionally used at supersonic speed (Felix Baumgartner used one to stabalise him on his supersonic skydive). Nasa even uses gigantic supersonic parachutes, for example during curiosity's decent.

The parachutes designed to be used at supersonic speeds in dense atmospheres look nothing like the drogue chute in KSP. The parachutes used at Mars are irrelevant, because the atmosphere is extremely thin there.

The question is: Do we want to think that the first 10km of alltitude, are just like earth so planes fly simmilarly to real life, or do we want to have pods reaching the exact same terminal velocity at 2km alttitude? We can´t have both without some tweakings and compromisses, The onle way we can do the latter without messing with other things is by having a very condensed unrealistic soupy atmosphere. The community ALREADY SAID THEY DON`T WANT THAT IN THE PAST.

This is a game about building and flying rockets. Planes are just secondary content. If something has to be broken until a better solution can be found, it's better to break planes than rockets and pods.

[Aerindel]

No they won't. I did test this. Unless you're plunging vertically through the atmosphere your pod will always slow to bellow 300 m/s which is a perfectly reasonable speed to deploy drogues; it's not massively faster than the soyuz drogue deployment speed.

Also drogues are occasionally used at supersonic speed (Felix Baumgartner used one to stabalise him on his supersonic skydive). Nasa even uses gigantic supersonic parachutes, for example during curiosity's decent.

This is true of 1.02 in 1.0 this is not true, pods barely slowed at all.

This is a game about building and flying rockets. Planes are just secondary content. If something has to be broken until a better solution can be found, it's better to break planes than rockets and pods.

Agree. 1.02 aero is perfectly good for a space program game.

Edited May 8, 2015 by Aerindel

[oversoul]

All I know is my simplest rockets don't attain orbit now. What voodoo is this?

Nevermind, I see they massively nerfed some of the engines. Again.

14 thrust for the LV-909? This can't be right... re-installing...

Edited May 8, 2015 by oversoul

[CodedThreat]

Why are so many people trying to use Real World numbers and comparing it to a Game World where the planet is 1/10 the size but has 100% of the other stuff, like gravity and Aero and drag? I would not try to put the numbers in until RSS comes out and Kerbin is scaled up to Earth, then you can do your testing and see where it lands. Until then, all these points seem moot.

[sephirotic]

This is a game about building and flying rockets. Planes are just secondary content. If something has to be broken until a better solution can be found, it's better to break planes than rockets and pods.

KSP is not ultra-realistic, for that we have RSS. A large part of the community HATED the old soupy atmosphere, IT WAS A BAD BALANCE COMPROMISSE. They have to come with some OTHER way.

1.0.0 drag is STILL MUCH HIGH than real life, but was a more balanced compromisse that felt more natural than the stats at 1.0.2.

This is true of 1.02 in 1.0 this is not true, pods barely slowed at all.

Agree. 1.02 aero is perfectly good for a space program game.

How about a planet with the density of a frigging black whole like Kerbin? I think is ridiculous but you, me and a lot of people have come to accept it as a realism compromise (this is why most of us, me including like playing RSS a lot) What lots of people DIDN`T ACCEPT, was the soupy atmosphere. Myself included. This is why FAR was so poppular, alhoug it had it's problems too and was ALSO a little bit too "draggy". But was the best we got, and now that squad got right at 1.0.0 with the aero, they made a step backward by instead of fixing overpowered jet engiens, they returned the hated soupy atmosphere.

I don´t understand why it is so hard to understand, all the arguments in favor of squad decision don´t make sense to me. I wonder why.

[Pthigrivi]

Yeah they should resize the planet. Oh and take those weird green guys out and put some humans in for god's sake they're not realistic at all.

[sephirotic]

Yeah they should resize the planet. Oh and take those weird green guys out and put some humans in for god's sake they're not realistic at all.

It's not about realism, is about good balance compromisses. You missed my point.

[Jouni]

KSP is not ultra-realistic, for that we have RSS. A large part of the community HATED the old soupy atmosphere, IT WAS A BAD BALANCE COMPROMISSE. They have to come with some OTHER way.

I'm not talking about ultra-realism. I just want that things work in a way that generally feels reasonable. Right now we can't have that in aerodynamics, because there are a number of issues and their complex interactions, and nobody has figured out yet how to fix them. We can either have rockets and pods that feel reasonable, or planes that feel reasonable, but not both. Because rockets are more central to KSP than planes, I feel that Squad made a good decision with the aerodynamics changes in 1.0.1.

[Aerindel]

Honestly, off all the multitude of bad problems that 1.02 still has, the atmosphere is not one of them. I don't understand what people think it should be like. Wildly unrealistic SSTO still work, rockets still fly...what more do people want? It sounds like they wan unlimited speed, lift and no drag.

Airplanes suck...thats why this isn't a flight sim, its something a loot better, a rocket sim.

[Pthigrivi]

It's not about realism, is about good balance compromisses. You missed my point.

Sorry that wasn't at you sephirotic. I saw people suggesting this one of the hundred million other qq aero threads and was being snark.

[Fel]

Yeah they should resize the planet. Oh and take those weird green guys out and put some humans in for god's sake they're not realistic at all.

Xenophobe!

Really though, this seems to be "Space Planes vs Rockets." Thing is, space planes STILL have largely overpowered equipment which allows you to make them, but differently. If squad "fixed" space planes you wouldn't be able to make them at all. Just have fun with the game, or tweak aero how you like it.

[sephirotic]

I don't understand what people think it should be like. Wildly unrealistic SSTO still work, rockets still fly...what more do people want?

I for myself don´t understand were have you been for the last 3 years of game development as pretty much every more experienced played talked about how they just completely abandoned the stock aero model for being too soupy and switched to FAR and how this 1.0.2 is a retrogress towards what we have mostly hated for so many years about KSP after a brief nice experience with the aero of 1.0.0.

[m4v]

Why are so many people trying to use Real World numbers and comparing it to a Game World where the planet is 1/10 the size but has 100% of the other stuff, like gravity and Aero and drag? I would not try to put the numbers in until RSS comes out and Kerbin is scaled up to Earth, then you can do your testing and see where it lands. Until then, all these points seem moot.

well, that's because that's what people were doing when arguing for a less dragy atmosphere at the beginning of the thread. People would go over how atmosphere should be changed because IRL planes can reach mach speeds easily while at the same time disregarding the unrealistic terminal velocities of pods.

Edited May 9, 2015 by m4v

[Renegrade]

Since the capsule is falling at terminal velocity, the drag force equals the weight of the body. The Apollo capsules had a mass of about 5,000 kg at splashdown, thus their weight was about 5000 x 9.8 = 49,000 N. Assuming the vehicle is falling heat shield down, the area is 12 m² (3.91 m diameter). The Apollo capsule drag coefficient was about 0.85 at subsonic velocity. Air density at sea level is 1.225 kg/m³. Plugging in all these number, we get

An interesting tidbit here is that the Mk1-2 pod is something like 4120kg and 2.5m (slightly less than 2.5m possibly). (4.91m²)

..giving us a terminal velocity of 125.74ish..yes?

(and the mk1 at 113.55)

...the command pods are a bit overweight for their size it seems..

[NathanKell]

The command pods have ~40% the surface area of their real life "counterparts"--the Mk1 is a Mercury analog (~1t mass with a chute and shield) and yet 1.2m in diameter not 1.9, and the Mk1-2 is an Apollo analog (~5t kitted out) and yet ~2.4m not 3.9m in diameter. 63% the diameter means ~40% the surface area means ~40% the drag means ~18.5% faster terminal velocity.

Terminal velocity is based only on the Cd*S, mass, and air density. It has literally nothing to do with how high the atmosphere is, so can we all please leave that aside for the moment? (And note, it's 80%, not 66% or 60%--go ahead and compare 8km on Kerbin and 10km on Earth).

Terminal velocity for the pods, even in 1.0.2, is a bit low. However, drag for planes is way too high. This doesn't mean "find some average drag" it means change how Cd is calculated so that pod Cds stay about the same but plane Cds go down. One thing to do this would be to add an exponent to Cd, so Cd < 1.0 gets lower (and the lower it starts as, i..e the more streamlined the part is, the more it would get lowered by the exponent).

[Guest]

I was gutted Squad neglected to include a compartment for sausages/s'mores/Hot Pockets/etc on the outer hull of the command pods. A warm snack helps pass the time waiting for the recovery crew to show up. How could they have missed something so elementary??

[basic.syntax]

I was gutted Squad neglected to include a compartment for sausages/s'mores/Hot Pockets/etc on the outer hull of the command pods. A warm snack helps pass the time waiting for the recovery crew to show up. How could they have missed something so elementary??

LoL. I use the new cylindrical in-line service bays, for snack storage. They cry out to me: "open here for cookie." Edited May 9, 2015 by basic.syntax
test - no changes made