Inflatable heat shields

[IncongruousGoat]

There's been some talk recently in the SpaceX thread about ULA's SMART reuse plan. Part of this plan involves using an inflatable heatshield to protect the (now-separated) first stage engines on re-entry, due to the high velocity at MECO. This led me to wonder... what's the current state of inflatable, deployable heatshield technology? It would require some highly flexible material that ablates slowly enough that you only need shielding that's as thick as canvas. A heatsink design seems out of the question - heatshield mass just doesn't seem like it'd be high enough to support that, at least not using any real material. I can't think of any material that would fit the bill. On the other hand, I am not a materials scientist.

So, what research and testing has been done on this tech, and how mature, reliable, and cost-effective is it?

[Bill Phil]

The technology exists, though it likely isn't mature at this point.

[Nibb31]

Inflatable heat shields do not exist. There have been some (not very successful) tests of inflatable hypersonic decelerators, which are made to slow down a vehicle from orbital speed to something like supersonic speed. These have some heat resistant properties, but they are not heat shields.

Edited August 25, 2018 by Nibb31

[StrandedonEarth]

But... But.... It works in Kerbal Space Program...! 

[55delta]

Okay then...what do we know about the current state and development of inflatable hypersonic decelerators? Also, how does being an hypersonic decelerator make any different compared to a heatshield outside of avoiding re-entry heating versus managing re-entry heating?

[IncongruousGoat]

5 hours ago, Nibb31 said:

Inflatable heat shields do not exist. There have been some (not very successful) tests of inflatable hypersonic decelerators, which are made to slow down a vehicle from orbital speed to something like supersonic speed. These have some heat resistant properties, but they are not heat shields.

Yeah, that's where I thought the technology was. Although... if it's slowing down all the way from orbital speed to around Mach 1, it's going to have to deal with a lot of waste heat. Is the idea that the vehicle is separately shielded, and the inflatable thing is just a giant fancy airbrake?

[Bill Phil]

6 hours ago, Nibb31 said:

Inflatable heat shields do not exist. There have been some (not very successful) tests of inflatable hypersonic decelerators, which are made to slow down a vehicle from orbital speed to something like supersonic speed. These have some heat resistant properties, but they are not heat shields.

Well, many people within NASA refer to those decelerators as heat shields. They still have to shield the payload from whatever heat crops up during reentry. They have Thermal Protection Systems.

NASA did a successful demonstration test in 2009, and another in 2012. The technology exists. The TRL likely isn't at the necessary level for deployment, but it exists alright.

[magnemoe]

21 hours ago, Bill Phil said:

Well, many people within NASA refer to those decelerators as heat shields. They still have to shield the payload from whatever heat crops up during reentry. They have Thermal Protection Systems.

NASA did a successful demonstration test in 2009, and another in 2012. The technology exists. The TRL likely isn't at the necessary level for deployment, but it exists alright.

As I understand the large diameter make them slow down faster and higher than standard, also the large surface area spread heat over an large area so the heating get much lower and you don't need advanced heat shielding.

Note that NASA readiness requirements is for billion dollar probes who has taken years to make. 
For spaceX it would be to try to reuse upper stages who else would burn up. 

[wumpus]

1 hour ago, magnemoe said:

As I understand the large diameter make them slow down faster and higher than standard, also the large surface area spread heat over an large area so the heating get much lower and you don't need advanced heat shielding.

Note that NASA readiness requirements is for billion dollar probes who has taken years to make. 
For spaceX it would be to try to reuse upper stages who else would burn up. 

Note that the reduced heating is presumably due the hottest part of the compressed gasses in front of the heatshield being pushed further away and around the heatshield.  The total heat generated should be higher as it is proportional to the area and speed of the heatshield.

I think I was only aware of the failed 2016 test.  I'm impressed that there were two previous successes.  I still don't expect any recovery effort from ULA to be more than "checkbox engineering" where they go through the motions because the contract demanded it (if you want them to recover, you better put "use a previously used engine" requirement for some flights).

[Gargamel]

On 8/25/2018 at 6:52 AM, StrandedonEarth said:

But... But.... It works in Kerbal Space Program...! 

Obligatory:

[StrandedonEarth]

1 hour ago, Gargamel said:

Obligatory:

Where did you think I got the line from? 

[Mad Rocket Scientist]

But SMART doesn't separate the engines at orbital velocities. SpaceX has already demonstrated that essentially cork can withstand reentries from suborbital velocities.

[ment18]

We don't know how much engineering work had to go into making the engines themselves survive reentry nozzle first.

[StrandedonEarth]

24 minutes ago, ment18 said:

We don't know how much engineering work had to go into making the engines themselves survive reentry nozzle first.

That may have been one of the reasons for the entry burn: to keep the bells "inflated"

[Mad Rocket Scientist]

51 minutes ago, ment18 said:

We don't know how much engineering work had to go into making the engines themselves survive reentry nozzle first.

No, but the dancefloor protects the powerheads, octaweb and everything in it (LOXtapus can't miss a chance to use that word), and the tanks from reentry. And part of it has to be flexibly to allow gimbal on the M1Ds, so it's conceivable that you could run propellant lines through something like it.

[IncongruousGoat]

5 hours ago, Mad Rocket Scientist said:

But SMART doesn't separate the engines at orbital velocities. SpaceX has already demonstrated that essentially cork can withstand reentries from suborbital velocities.

Well, there's suborbital, and then there's suborbital. SpaceX deliberately has MECO at a low velocity to make booster recovery easier. ULA, however, has a legacy of very high MECO speeds. IIRC, Atlas V has MECO at above 5 km/s, and we have no reason to think that Vulcan will be any different. That engine pod will be separating at high speeds that will necessitate some sort of heat shielding far beyond anything SpaceX has to use.

[wumpus]

12 hours ago, IncongruousGoat said:

Well, there's suborbital, and then there's suborbital. SpaceX deliberately has MECO at a low velocity to make booster recovery easier. ULA, however, has a legacy of very high MECO speeds. IIRC, Atlas V has MECO at above 5 km/s, and we have no reason to think that Vulcan will be any different. That engine pod will be separating at high speeds that will necessitate some sort of heat shielding far beyond anything SpaceX has to use.

I generally point out to new KSP players that having MECO 1/2 your orbital delta-v (or more specifically, having each stage with equal delta-v) is generally a good place to start your rocket design (unfortunately, there is no clear optimal and you have to iterate any design).  Early (and current non-recoverable) Falcon 9 rockets appeared to have a much later MECO (obviously that last 10% or so provides significant delta-v), and Falcon Heavy returns the center stage after a significantly delayed MECO (and the first crashed).

I'm curious to see if BFR has a MECO above 4 km/s, as that is likely to be more efficient (and probably necessary for that claim of "SSTO with no cargo, and presumably no return").  And if SpaceX wants to return the second stage (as planned), they will need to deal with significant speeds (although probably through backburns as they are doing now).

[Racescort666]

12 hours ago, IncongruousGoat said:

Well, there's suborbital, and then there's suborbital. SpaceX deliberately has MECO at a low velocity to make booster recovery easier. ULA, however, has a legacy of very high MECO speeds. IIRC, Atlas V has MECO at above 5 km/s, and we have no reason to think that Vulcan will be any different. That engine pod will be separating at high speeds that will necessitate some sort of heat shielding far beyond anything SpaceX has to use.

It's worth noting (by my calculation) that a max payload Atlas V 551 upper stage only has 4.6 km/s dV for GTO missions. It's likely that the MECO is much higher than 5 km/s for max payload GTO missions. 

[IncongruousGoat]

12 minutes ago, Racescort666 said:

It's worth noting (by my calculation) that a max payload Atlas V 551 upper stage only has 4.6 km/s dV for GTO missions. It's likely that the MECO is much higher than 5 km/s for max payload GTO missions. 

Yeah, I couldn't remember the exact number off of the top of my head, and so put 5 km/s per second as a conservative estimate. For heavy LEO payloads, the MECO speed is almost certainly even worse, probably up near 6.5 km/s. For anything other than light interplanetary probes, that Centaur upper stage is very underpowered.

[Wjolcz]

I really wouldn't be surprised if they ditched the whole idea and just went for the whole stage recovery in a couple of years.

[IncongruousGoat]

5 hours ago, Wjolcz said:

I really wouldn't be surprised if they ditched the whole idea and just went for the whole stage recovery in a couple of years.

I would be very surprised. Full-stage recovery with Vulcan would be far, far harder to accomplish than it is with Falcon 9, even if it had an engine arrangement that could support said recovery (which it doesn't). Again, MECO speeds for Vulcan are going to be a lot higher than Falcon 9, which makes the amount of propellant one would need to reserve for braking problematic.

[Racescort666]

So I was looking into this a bit more and NASA has conducted several tests with inflatable decelerators. HIAD (Hypersonic Inflatable Aerodynamic Decelerator) was the one I was thinking of which is also called Inflatable Reentry Vehicle Experiment (IRVE). The latest flight test was in 2012 which was IRVE-3 and it looks like it would be fairly similar to what ULA wants to do. This project looks to be spearheaded by NASA Langley although an insider told me that Ames also had a hand in it. The latest update I could find was that they were doing some TPS testing with Boeing in 2017.

The other inflatable heatshield project is LDSD (Low Density Supersonic Decelerator) which had flight tests in 2014 and 2015 and is lead by JPL. This project also includes the SIAD (Supersonic Inflatable Aerodynamic Decelerator) which doesn't appear to be related to HIAD despite the similarities. LDSD was tested at the US Navy's Pacific Missile Test Range and both missions had successful SIAD deployment however they both also suffered from parachute failure. This I think is maybe a source of confusion for this thread. 

With regard to inflatable decelerators/heatshields, Ian Clark, principal investigator for LDSD at JPL had this to say after the 2015 LDSD test:

Quote

We also saw another successful inflation of our 20-ft SIAD and another successful deployment and inflation of our supersonic ballute (an inflatable drag device that extracts the parachute). Both of those devices have now had two great flights, and we have matured them to the point where they can be used, with confidence, on future missions

With that statement, I'd say this is around TRL 6. It would be up to ULA to decide whether they want to go ahead with using it for engine recovery since NASA seems to be mostly interested in using it to land bigger and better things on Mars. I think it's unlikely that we would see it on Atlas V and probably just on Vulcan but I would be happy to be surprised.

Edit:

More research yielded this tweet from Jeff Foust:

LOFTID is the follow on to HIAD at Langley. Although there isn't much information available on it, it's basically an orbital test of HIAD. 

Edited August 28, 2018 by Racescort666
more info

[Wjolcz]

9 hours ago, IncongruousGoat said:

I would be very surprised. Full-stage recovery with Vulcan would be far, far harder to accomplish than it is with Falcon 9, even if it had an engine arrangement that could support said recovery (which it doesn't). Again, MECO speeds for Vulcan are going to be a lot higher than Falcon 9, which makes the amount of propellant one would need to reserve for braking problematic.

I was thinking more about whatever comes after Vulcan. They will probably have to figure something out eventually.