Airbags in space

[Pds314]

It would be cool to have inflatable airbags that would compress on impact but not easily break from raw speed, for the purpose of actually-survivable lithobraking without just strapping steel beams to your ship (though you still don't wanna hit the ground at mach 2 or it will destroy them or whatever they're attached to or both).

Edited October 26, 2019 by Pds314

[paul23]

7 hours ago, Pds314 said:

It would be cool to have inflatable airbags that would compress on impact but not easily break from raw speed, for the purpose of actually-survivable lithobraking without just strapping steel beams to your ship (though you still don't wanna hit the ground at mach 2 or it will destroy them or whatever they're attached to or both).

how would they be different from bigger landing legs?

[Jodo42]

28 minutes ago, paul23 said:

how would they be different from bigger landing legs?

The airbags on Spirit and Opportunity were designed to allow a survivable landing at >10m/s and ~400kg, and bounces after landing at over 30m/s. Try that with stock landing legs and see what happens.

[paul23]

Well yes but mechanics wise it would be the same.

[PositronLance001]

Also could let us make more compact designs. sometimes u just can't fit legs. I mean the mini ones are ok but break SUPER easily.

[paul23]

Well that's just a request to make "stronger gear". Stronger "landing equipment" in the form of airbags. (In reality landing legs could withstand much higher landing speed than airbags btw, it's easy to create strong landing legs since metals play really nice).

[EchoLima]

On 10/26/2019 at 6:09 PM, paul23 said:

Well that's just a request to make "stronger gear". Stronger "landing equipment" in the form of airbags

I'd say its easier to add landing legs to a craft than to encase it in airbags. It's a tradeoff- different items have different benefits.

[Cavscout74]

On 10/26/2019 at 7:33 AM, Pds314 said:

It would be cool to have inflatable airbags that would compress on impact but not easily break from raw speed, for the purpose of actually-survivable lithobraking without just strapping steel beams to your ship (though you still don't wanna hit the ground at mach 2 or it will destroy them or whatever they're attached to or both).

If you are interested, there are a few mods that add functional airbags to the game (USI Exploration, SXT-Continued & I think Comfortable Landing).  Both SXT & USI work pretty good at least up to 1.7.3 (I haven't jumped to 1.8 yet).

[Gargamel]

On 10/26/2019 at 11:23 AM, paul23 said:

how would they be different from bigger landing legs?

 

On 10/26/2019 at 12:26 PM, paul23 said:

Well yes but mechanics wise it would be the same.

It would be a different mechanic, or at least something other than just adjusting the tolerances of a part. 

1 hour ago, EchoLima said:

It's a tradeoff- different items have different benefits.

Yup, this...   Sometimes you just want to throw a probe or rover at the ground and let it bounce around for a bit, you don't always need a controlled landing. 

[kerbiloid]

On 10/26/2019 at 6:23 PM, paul23 said:

how would they be different from bigger landing legs?

Compact enough to be hidden (say, a flat disk between the capsule and the heatshield, expanding into 6 bags after the shield is jettisonned).
And single-use, disappearing after usage, automatically or on a button.

Edited October 28, 2019 by kerbiloid

[Pds314]

On 10/26/2019 at 3:09 PM, paul23 said:

Well that's just a request to make "stronger gear". Stronger "landing equipment" in the form of airbags. (In reality landing legs could withstand much higher landing speed than airbags btw, it's easy to create strong landing legs since metals play really nice).

I'm gonna disagree somewhat here.

A lightweight airbag is pretty much always gonna have an advantage over a lightweight landing leg in impact tolerance. The limiting factor is the tensile strength of your balloon material, which if it is kevlar or some similar material will considerably exceed the compressive strength of steel even though it doesn't have to, because a spindly little leg will happily buckle under excessive force. Something a balloon can't do.

 

This also gets at a more important difference though. They should not permanently inflate, nor should they be good for stably mounting a spacecraft. We already have steel bars for that.

What they should be good at is taking very high speed impacts and turning them into a chaotic bounce across the surface, and also at taking horizontal impacts more easily than vertical ones.

 

I would even say don't give them an impact tolerance (okay, 300 m/s is fine) but a mechanic like wheel stress where overly harsh impacts compress and pop them. If you just throw a lone airbag at the ground, it should survive extremely high impact speeds. But obviously if you try to land a reuseable booster on one tiny airbag, you're gonna need to be gentle. (And more to the point, it'll tip over).

Edited October 31, 2019 by Pds314

[paul23]

12 minutes ago, Pds314 said:

I'm gonna disagree somewhat here.

A lightweight airbag is pretty much always gonna have an advantage over a lightweight landing leg in impact tolerance. The limiting factor is the tensile strength of your balloon material, which if it is kevlar or some similar material will considerably exceed the compressive strength of steel even though it doesn't have to, because a spindly little leg will happily buckle under excessive force. Something a balloon can't do.

 

This also gets at a more important difference though. They should not permanently inflate, nor should they be good for stably mounting a spacecraft. We already have steel bars for that.

What they should be good at is taking very high speed impacts and turning them into a chaotic bounce across the surface, and also at taking horizontal impacts more easily than vertical ones.

 

I would even say don't give them an impact tolerance (okay, 300 m/s is fine) but a mechanic like wheel stress where overly harsh impacts compress and pop them. If you just throw a lone airbag at the ground, it should survive extremely high impact speeds. But obviously if you try to land a reuseable booster on one tiny airbag, you're gonna need to be gentle. (And more to the point, it'll tip over).

Landing as a whole could (should) get a revamp and stress would need to be calculated in a "mitigation" factor. If there is a very stiff beam below a fragile science lab, it makes little sense that the fragile science lab "survives" the impact. So it would be best if each component has a certain stiffness, which determines how much of the momentum (! momentum, not velocity) it absorbs at landing.

 

In reality landing legs have a much higher absorption due to shock absorbers, and at higher speed one can even use one-time-only crumple area, which is what I meant with efficiency in momentum absorption.

 

Airbags indeed are only used with small crafts, where location is of little importance (they however have internal gyroscopes to keep the internal craft vertical). Legs + shock aborbers are however better at gentle touchdown.

[kerbiloid]

7 hours ago, paul23 said:

In reality landing legs have a much higher absorption due to shock absorbers, and at higher speed one can even use one-time-only crumple area, which is what I meant with efficiency in momentum absorption.

Any shock absorption is limited with one factor: the brain inside the skull is not a solid and rigid structure.
On a hit (from any direction) it moves inside the skull and hits against its wall from inside.
It compresses at the place of hit, but it's full of incompressible water.
So, the water inside the brain causes two injuries: at the place of hit, where the tissues suddenly compress around and pop, and in opposite side of the brain, where the wave of pressure reaches the outer brain edge.

As you can neither fix the brain inside the skull, nor dry it, the hit tolerance is always limited with braking distance.

***

Generally, you have to limit the acceleration on braking with ~20 g.
30..40 g require medical assistance .

The only way to limit the acceleration is to increase the braking distance.
There is no difference how this distance is provided, by legs or by bags.
Also there is no need in the dampener strengthness, unless you want to reuse it after crash, after washing the cabin.
Vice versa, a leg must be equipped with a fragile single-use dampener crashing on a strong hit.

Also a leg should not be vertical, like it seems to be in Crew Dragon, because it will pierce the cabin bottom and impale the dragonriders.
It should be foldable, not telescopic, to crash and fold out on a strong hit. Unless it's attached to the wall outside the cabin of a lunar lander.

***

As we can see, the braking distance provided by legs is  limited with the capsule clearance, i.e. vertical distance between the ground and the bottom.
The clearance can not exceed radius of the capsule, otherwise it's both hard to fold or retract the legs and the capsule can overturn.

So, for a 3..5 m wide capsule the clearance is limited with ~1 m.
Minus foot and joint, and you have just 0.7 m of braking distance.

Absolutely the same you can get with bags. But unlike the legs, the bags take much less place until being inflated, are lighter, and more reliable because of no mechanics.

***

Add to that 0.7 m of braking distance from bags or legs another 0.3 m from the special seat which extends its pistons at last seconds before the touchdown.
So, you have total 1 m braking distance in any case, bags or legs.

As we know, a = v²/(2*L)
So, Velocity = sqrt(2 * Acceleration * TotalBrakingDistance).
For TotalBrakingDistance = 1 m as above,
0.5 g → sqrt(2 * 0.5 * 9.81 * 1) ~= 3 m/s
2 g → sqrt(2 * 2 * 9.81 * 1) ~= 6 m/s
20 g → sqrt(2 * 20 * 9.81 * 1) ~= 20 m/s
40 g → sqrt(2 * 40 * 9.81 * 1) ~= 30 m/s
1000 g → sqrt(2 * 1000 * 9.81 * 1) ~= 140 m/s

So, the combination of the pistonned seat and no matter bags or legs:
is comfortable at up to 3 m/s landing
is safe at up to 6 m/s landing
lets the crew survive and probably repair at up to 20 m/s landing
gives chances to survive and maybe even repair up to 30 m/s landing
lets the strongest electronics survive at 140 m/s lithobraking.

***

Also as we can see, the latter case makes for an uncrewed strong lithobraking probe both bags and legs having no sense in the Earth atmosphere, as even without a chute the terminal velocity in the air is about 150 m/s, so you can just drop the probe.

But as the 1000 g is an artillery shell class electronics, you can't use this value for fragile scientific equipment.
It anyway should withstand just 20..100 g or so.

So, if surround a lunar probe with 2 m thick layer of airbags, the velocity limit is sqrt(2 * (20..100) * 9.81 * 2) ~= 30..60 m/s.

So, any landing dampeners are useful just at <= 60 m/s (and better < 30 m/s) even for uncrewed probes.

***

For a crewed capsule the choice between the legs and the bags is not a question of preference, but a question of landing tmethod.

When you capsule is landing by parachutes, you must have the parachutes enough large to provide ~6 m/s descent (see above, the 2 g shock).
This is absolutely comparable to the typical wind speed on the Earth, 3..5..8 m/s.

This means that an effective parachute is at once an effective sail.
So, as there is rarely no wind at all, a pure parachute landing capsule always lands at 3..6 m/s vertical speed (parachute + short-impulse retrorockets) and 3..5 horizontal speed, so always hits the ground diagonally.

This in turn means that:
1) If the capsule is more ore less high, it will overturn. Soyuz or Crew Dragon (if land the latter on ground) will overturn almost usually, even the relatively flat Apollo-shape capsule will overturn with high probability.
2) The crew inside the capsule receives a strong horizontal hit in unpredictable random direction.

This is more or less appropriate for a small 2 m capsule of Soyuz, so small that this is just a acrobatic overturn (just 2/2 = 1 m of rolling radius).
But it's hardly appropriate for large multiseat capsules.
Because
1) you have to send more casual humans, not only acrobats;
2) the rolling radius is 2..3 m, so even an acrobat will feel bad;
3) if the capsule stays overturned, they will be hanging at 2..3 m above th floor, and try to release from the seats and not fall down and break a bone.

***

You can't withstand this with legs, they will just bend sideways and break.
The only two options you have to safely land a multiseat capsule on Earth are:
1) let it land on engines, make the horizontal speed zero, and the verical speed <= 3 m/s (more complicated and heavy equipment is required);
2) let it slip along the ground on landing (less comfortable and still has a chance to overturn, but lightweight, reliable, and easy)

The obvious way to slip is a layer of inflatable airbags beneath the bottom.
CST-100 follows this way.

In this case you need inflatable airbags and jettisonnable heatshield (as the airbags should be right beneath, rather than around).
The legs are inappropriate here in this case at all.

***

Or you can land by rockets.
The advantage is: you can land at zero speed and forget about the wind.
The disadvantages are numerous and horrible.

The greater is the ignition altitude, the more fuel you have to spend while descending, so you must engage them at the lowest possible altitude.
You should prefer high-G landing impulse rather than slow smoking.
Say, you are going to land at 4 g. I.e. 1 g of gravity + T/W = 3.
The terminal velocity of aerobraking without a chute is ~150 m/s.

So, you have to engage the engines at 150² / (2 * 3 * 9.81) ~=400 m altitude to finish braking near the ground.

But what if the engine fails?
You can't land on legs or bags at 150 m/s.
You can't use a backup chute at that altitude, such big chute requires about 2 km to open and brake.
So, you have either to carry a backup set of engines which increase the capsule mass and its complexity, or engage them at 3 km to test and keep descending without switching off.
In the latter case you have to spend several times more fuel than if engage them at 400 m.

Say, you have a twin set of engines, and are sure they can't fail all at once.
You engage the engines at 400 m altitude, feeling appropriate 4 g.
You need delta-V = 150 m/s, to zerize the landing velocity.
But at T/W = 3 the braking will take 150 / (3 * 9.81) ~5 seconds. I.e plus 5 * 9.81 ~= 50 m/s of delta-V

So, you need ~200..250 m/s of delta-V to be taken away with engines.
(Plus some delta-V for final hovering and descending, but let's omit this).

At atmospheric ISP*g ~3 km/s, the mass ratio = exp(200..250 / 3000) ~= 1.08.
So, at least 8% of the capsule mass must be hypergolic fuel. For  9 t capsule it's a 700 kg hypergolic tank you're landing on.
That's not very good if you have a crash. If the tanks break, the remains of the fuel will cause a fire and poison.

Also this raises the complexity of the capsule and decreases its reliability.

***

But anyway, if you land by rocket engines, you hardly can use airbags because they will be damaged with fire.
Either legs like in Dragon, or nothing at all, but then tens of small engines around the top of the capsule like in Zarya.

Actually, only legs.

So, the choice between the legs and the bags is equal to the choice between pure chute landing (airbags only) and rocket landing (legs only).

***

Obviously, the parachute + airbags combination is more simple, and CST-100 uses it.

And the rocket landing is bad because it's equal to having an additional rocket stage, and other bad things. The early Crew Dragon animations of the rocket landing are a false way.

But the parachutes+airbags scheme has several disadvantages:
1) It still can overturn or hit the ground diagonally, as the capsule is still moving horizontally on touchdown.
2) The parachutes occupy volume inside the capsule shape, and they occupy the narrowest part of the capsule, exactly where you place a docking port and visual controls.
3) It moves horizontally while descending, so you can't be sure if it won't hit a barn or electric wires, or fall from/into something.

So, these are the arguments to prefer rocket landing.

***

Hence we get to the combined landing, partially used in PTKNP project.

Have just a small drogue chute instead of main sail.
Descend on it at 20 m/s speed down to the very ground.

As the chute is small, it's a poor sail, and the wind affects it insignificantly, so the horizontal speed is lower than if you use a big parachute.

At some very low altitude ignite a small solid motor beneath the cabin to slow down.

If the engine fails, you hit the ground at 20 m/s with you drogue chute on top.
As you can see from the table, any dampener together with pistonned seats will save the crew, so no need in backup engines or early ignitions.

If the engine ignites, you jettison the drogue chute, zerize the total speed at, say, 2 m above the ground, and slowly land at 1..3 m/s.

***

The ignition altitude is: 20² / (3 * 9.81) + 2 ~= 15 m.
So, you burn the solid motor at 15 meters from the ground, a second before the 20 m/s touchdown.

If it's ok, you have a 3..4 seconds long fire show instead.
If engine fails, you break the capsule legs, but stay healthy.

***

In this case you need delta-V ~20 m/s to be taken away with the engine.
The deceleration at T/W = 3 takes about 1 second, so +10 m/s of gravity work.
Descending from 2 m hovering to the ground at 1..3 m/s is also about a second.
So, you need total delta-V ~40 m/s instead of hundreds of m/s in the case of pure rocket landing.

For a solid motor the mass ratio = exp(40/2000) ~= 1.02. So, the solid fuel ~= 200 kg.

You don't need many solid motors, just a strong ringed pipe with several nozzles around (a single detail in KSP), and 1..2 powder charges in cylindric cases attached to it.
In case if a charge is failed, you can have 3 charges with 2 of them necessary, igniting them one by one. This lets you safely land without capsule breaking even if one of them failed.

So, in this case the whole landing retrorocket is ~300..400 kg heavy, instead of ~2 000 kg in case of pure rocket landing.
And the fuel is solid, and can be made inert without ignition, unlike the hypergolics.

Also you don't need so much volume.

***
The drogue chute should provide the descending speed ~20 m/s instead of 6 m/s of a big chute.
The drag force ~S * v²~ D² * v². so D * v = const

This means that the drogue chute can be 20/6 = 3 times smaller than a big chute.
So, it can weight 25 times less and occupy 25 times less space in your capsule top.

In fact, you can just have 3..4 tiny mortars with drogue chutes instead of the large parachute boxes occupying the whole top.

***

Obviously, the combined landing requires legs. But much smaller legs, because the capsule is not overmassed with fuel tanks and liquid engines.

***

So:

Either full parachutes + airbags (appropriate)
Or landing engines + stronger legs (the worst)
Or drogue parachutes + landing solid ring-shaped multi-nozzle retrorocket + weaker legs (the best)

(Though, no need in these perversions for "1.25" or "1.875" capsules.)

Anyway, about the airbags:
A flat disc part "capsule bottom with six cylindric airbags" - must have (at least for "2.5 m" and "3.75 m" sizes). To survive 10..15 m/s landing
A "3..4 spherical airbags to be attached from the side to a lunar/martian probe" - must have (0.625 wide spherical balloons grouped by 3, attach several and hide the probe in balloons). To survive 10..15 m/s with crew or 30..50 m/s without crew.
All of these must inflate on demand and deflate after landing.

 

7 hours ago, paul23 said:

Airbags indeed are only used with small crafts, where location is of little importance

Not that small.

Spoiler

 

 

Edited November 1, 2019 by kerbiloid

[paul23]

uhh, what are you talking about.. I stopped once you started talking about "impaling dragon riders"....

 

Is this a joke post or what? Most certainly you can have telescopic legs. And most certainly legs will use less volume compared to airbag which need to encase a whole craft. If you talk about reality, both wouldn't be enough to provide a safe margin for humans. You'd either use active landing using rockets, or some kind of lift/drag device (parachutes, wings). If you're using retroburns you already need "kind of" very good attitude control. So airbags seem strange for the final meters.

 

If you use wings/parachutes you probably don't need either since you plan to land differently anyways.

 

 

If you really wish to use airbags/landing legs solely and keep max g below 20, so less than 200 m/s^2 deceleration, you'd need a lot of room:

Terminal velocity is typically around 200 m/s. So you'll need 1 second to "break": this means 200 meter of airbag distance. Even at 50 m/s you'd need 12.5 meter of airbag distance. To get realistic values you'd wish to first reduce velocity to 20-30 m/s.. But at that point you might as well use landing legs or fully use parachutes to break to <5 m/s

[Gargamel]

22 hours ago, kerbiloid said:

Any shock absorption is limited with one factor: the brain inside the skull is not a solid and rigid structure.

Nice post.  It's one of the most misunderstood aspects of brain injury, the coup contrecoup impacts the brain suffers with a deceleration. 

14 hours ago, paul23 said:

Is this a joke post or what?

Most clearly it isn't, as @kerbiloid will put his joke posts in spoilers.   That, and obvious amount of effort put into it. 

[kerbiloid]

14 hours ago, paul23 said:

what are you talking about.. I stopped once you started talking about "impaling dragon riders"....

I talk about vertical steel bars enough strong to carry a capsule, piercing on crash the capsule bottom and the crew. You don't want them, you want foldable legs like a ladybug has.
Dragon riders are those who rides the Crew Dragon, obviously. That's not me who gave the fantasy name. I prefer names like CST-100 or TKS.

14 hours ago, paul23 said:

Most certainly you can have telescopic legs.

Most certainly that's how correct under-capsule legs look like.

Spoiler

They should fold aside on crash rather than stick into the cabin bottom.
 

Btw the landing solid engine to finish the 20 m/s descent.

Spoiler

14 hours ago, paul23 said:

And most certainly legs will use less volume compared to airbag

This doesn't matter at all.
Either you have horizontal speed on touchdown, then no legs, only airbags. Legs will be broken (the best case) or the capsule will overturn on them (the worst case).
Or you land vertically by engines, then only legs.

14 hours ago, paul23 said:

If you really wish to use airbags/landing legs solely and keep max g below 20

14 hours ago, paul23 said:

I stopped once you started talking about "impaling dragon riders"....

This explains why you got nothing.

[Ol’ Musky Boi]

On 11/1/2019 at 7:29 AM, kerbiloid said:

So, in this case the whole landing retrorocket is ~300..400 kg heavy, instead of ~2 000 kg in case of pure rocket landing.
And the fuel is solid, and can be made inert without ignition, unlike the hypergolics.

Great post, but it's worth noting that in the case of Dragon 2 the landing engines would have also been the LES, so that offsets some of the mass penalty. If we assume that an expendable LES would've been jettisoned at first stage separation (which occurred at ~1,900m/s for the DM-1 mission) and would've had a mass of 2,000 kg, then the payload penalty of bringing the LES to orbit  (~7,500m/s) is about 1,600 kg, not 2,000 kg. So about 1,200 - 1,300 kg more massive than the drogue + retrorocket system.

Spoiler

Calculated with:

m_p = (m_i / e^{∆v / v_e}) - m_d

Where:
m_p = mass of payload = 9,500 kg with LES jettison or 11,500 kg without LES jettison

m_i = initial mass = 125,500kg with LES jettison or 127,500 kg without LES jettison

∆v = change in velocity = 5,600 m/s

v_e = exhaust velocity = 3,414 m/s

m_d = dry mass = 14,000 kg with LES jettison or 16,000 kg with LES jettison

 

This is all based on estimated specifications for the Block 5 Falcon 9 which I found here: https://www.spacelaunchreport.com/falcon9ft.html

This doesn't change the picture much in terms of which is more lightweight though, the landing system you mentioned is still 4-5 times lighter. But I think using the LES as a landing system has an advantage in cost savings and reusability. As you'd still need an LES of some sort for any crew capsule, and if you wanted to re-use that LES (and re-use is half the reason you'd want to land with retrorockets anyway) you'd still need to bring that 2,000kg system back with you, in which case why wouldn't you use it to land with? I believe that's why SpaceX initially went with the latter system, as it actually saved mass, additionally, keeping your LES with you all the way to orbit gives you quite a lot of abort options.

With all that being said, SpaceX still ditched the system, I imagine it was because of the safety issues you mentioned and the difficulty they would've had in getting that approved by NASA.

Edited November 2, 2019 by Ol’ Musky Boi

[kerbiloid]

30 minutes ago, Ol’ Musky Boi said:

But I think using the LES as a landing system has an advantage in cost savings and reusability.

I don't want to argue here about Dragon delta-V budget, as the thread is about bags & legs, so all right.
But I would notice that this way Dragon has to carry two full-featured landing systems to land (engines and chutes) if it spends the fuel for launch abort, and that typical LES delta-V is ~300 m/s, as well as delta-V of deorbiting from the ISS orbit (~250 m/s).
So, in sense of economy I would prefer to have a single-use LES/deorbit solid booster (with delta-V = 300 m/s) between the capsule and the expendable trunk, to use it for deorbiting in normal mission, or as LES in case of abort (the only difference: to ignite 3 powder charges all at once or one by one for proper acceleration). Then we anyway spend it in any flight on purpose and don't need to carry hypergolic fuel onboard the crew capsule.

Edited November 2, 2019 by kerbiloid

[Ol’ Musky Boi]

1 minute ago, kerbiloid said:

I don't want to argue here about Dragon delta-V budget, as the thread is about bags & legs, so all right.
But I would notice that this way Dragon has to carry two full-fetured landing systems to land if it spends the fuel for launch abort, and that typical LES delta-V is ~300 m/s, as well as delta-V of deorbiting from the ISS orbit (~250 m/s).

I imagine a back-up parachute system would have a similar mass to a drogue + retrorocket system, if not slightly less. Based on this document the Apollo CM's parachute system had a mass of ~100kg (If I've interpreted the data right), so I still think that it would be a lighter way to go even accounting for that. You're right though, we are getting a little off-topic.