Project BFR for Orbiter 2016

[francisdrake]

This is a concept study based on the presentation of Elon Musk in Adelaide on September 29, 2017. It contains the booster and the spaceship. Data on the BFR are scarce and sometimes contradicting. You could help me if you post data and flight experiences here.

The aerodynamic model is a biconic reentry vehicle. When fully pitched up it flies at an angle of attack of +30°. It is not very stable, don't go beyond 10x time acceleration while in atmosphere.

The 4 vacuum Raptors are the main engines of the spaceship, the 2 sea level Raptors are hover engines. For this tailsitter a landing MFD is recommended.

 

[Pipcard]

Okay, so I tried imitating that entry profile from the presentation (35:30) in which the front of the BFR spaceship is pointing down (approximately -90 degrees).

http://youtube.com/watch?v=tdUX3ypDVwI?t=35m30s

however it's from Mars orbit (3.5 km/s) instead of interplanetary velocities (7+ km/s); I also set the periapsis to 0 km (altitude).

At around 20 km, it starts pitching up (so that the heat shield is pointing towards the sky), and I can't really fight it; trying to turn using the numpad-1 and 3 keys doesn't really work either. Trying to pitch up to +90 degrees brings me back down as well. And once you get deep into the atmosphere (even for a Shuttle-style pointing-upwards entry), it's pretty much impossible to pitch over backwards to fire the descent engines.

I feel sorry for you because I understand that this is a very hard thing to recreate. I just wish we had more data, because this is meant to be the future of space travel, something that is supposed to make every other launcher obsolete.

 

[francisdrake]

Thanks for the feedback!

To control the pitchup the elevator trim can be varied with the Insert/Delete keys. I rather aim for a periapsis of 20-40 km, the atmpsphere is thinner here, so the up and down swaying is not so wild. Still, the pitchup to an engines-front attitude seems not possible with this control regime. I use pairs of super Draco thrusters for the RCS, but they seem too weak for that. Maybe SpaceX intends to rely on the gimballing of the main engines.

I could manage a fully propulsional Mars landing using the 4 main thrusters, steered by the PursuitMFD. But this needs a full tank and is obviously not the intended way.

 

[Pipcard]

You're welcome!

Is that launching from the pad or landing on the pad?

 

[vchamp]

Great! Thanks for doing this!

I checked BFR params and found some differences from the values in IAC presentation.
Slide with mass
Slide with engines params
- dry mass should be 85 t, Orbiter shows 135 t
- specific impulse of vacuum engines should be 375 s, in Orbiter I see 3820 m/s / 9.8066 = 389 s

Could you make these basic params configurable in a text file? I think the specs will change often during BFR development, it would be great to try different values without recompiling the dll.

 

[francisdrake]

Update Vers. 02: Now with booster and spaceship.
After launch perfom a roll to the right, to come on an Easterly heading. Then slowly pull back into a heads-down attitude.

On the mass and performance: The typical return payload is 50 ton, so I added this value to the 85 ton dry mass, that is where the 135 ton come from.
On the ISP values: I tend to up them by 10% for playability. Will have a look into that.

The Mars picture shows indeed an automated landing, starting with the Mars scenario, using the PursuitMFD for steering.

 

[BrianJ]

Cool! Thanks. I had a couple of runs at Mars entry-landing carrying 10% fuel. As noted, it's very difficult to control the pitch once you get into the atmosphere. In the presentation animation at around 30mins, it looks to me like the BFR is at a high AoA of ~60-70deg for most of the Mars entry, impossible to maintain even with max.trim + KillRot. Maybe could be better "balanced" for stable high-AoA?
Add-on has Shuttle-ish Lift/Drag of ~1, seems good to me.
One of the presentation Mars entry slides seemed to show quite a marked deceleration between 40km-30km alt, which I don't get with Orbiter.

I found the "heads-down" Mars entry attitude caused me to dig too deep into the atmosphere too quick and the vessel became too difficult to control. A 90deg roll attitude worked well at highest AoA possible(60-15deg),control rate of descent to ~200m/s by banking, then null out vertical speed at around 20km alt. Lose some velocity then descend to 5km. Use the the last useful lift at ~Mach4 to climb to 7-8km before descending again. Start flip over for propulsive landing at ~2.5km alt, Mach2.5. (Crash! - I haven't managed to land it nicely yet. Couldn't find an MFD that helped)

If it was more stable at high-Aoa (Shuttle-style) and maybe autopilot for flip-over and propulsive landing, would be no problem landing on Mars, anyway.

Here's my Mars entry test scn, (entry velocity 7300m/s, PeA 20km):

BEGIN_DESC

END_DESC

BEGIN_ENVIRONMENT
  System Sol
  Date MJD 57670.8294512259
END_ENVIRONMENT

BEGIN_FOCUS
  Ship BFR
END_FOCUS

BEGIN_CAMERA
  TARGET BFR
  MODE Extern
  POS 4.000000 117.684887 -60.368343
  TRACKMODE GlobalFrame
  FOV 40.00
END_CAMERA

BEGIN_HUD
  TYPE Orbit
  REF AUTO
END_HUD

BEGIN_MFD Left
  TYPE Orbit
  PROJ Ship
  FRAME Equator
  ALT
  REF Mars
END_MFD

BEGIN_MFD Right
  TYPE Surface
  SPDMODE 1
END_MFD

BEGIN_SHIPS
BFR:BFR
  STATUS Orbiting Mars
  RPOS -2014281.709 2742669.263 -746871.826
  RVEL -2872.3033 -5107.9357 -4654.7810
  AROT -91.993 -21.297 44.692
  VROT -0.0064 0.0004 -0.0001
  AFCMODE 7
  PRPLEVEL 0:0.1
  IDS 0:589 100
  NAVFREQ 20 0
  MODE 1
  SOLAR 0
END
END_SHIPS

BEGIN_ExtMFD
END

 

[francisdrake]

Brian, thanks for the extensive test!
I will try to achieve a higher AoA with earodynamic control surfaces. But the current wing flap is already assumed to be 2 m deep. I guess a considerable higher AoA needs a strong RCS to maintain the attittude.

My experience is, when the atmosphere gets denser (especially at Earth reentry) the vessel yaws and rolls a lot. I even ran into a stable hypersonic corkscrew spin, which I was unable to break out.

I will increase the aerodynamic damping, maybe this helps, but makes on the other hand achieving a high AoA even more difficult, as the weather-vane effect tends to orient the vessel nose-forward. Will need some trial and error.

 

[Pipcard]

The Mars picture shows indeed an automated landing, starting with the Mars scenario, using the PursuitMFD for steering.

Oh, okay. I was too busy paying attention to that precise landing to notice that it had been done fully propulsively.

 

[vchamp]

Update Vers. 02: Now with booster and spaceship.

Looks great and ascends to orbit beautifully! I didn't expect to have the booster so soon. BTW, it's the first time I've seen a suggested layout of engines.
The rocket is too powerful, it clears the tower very quickly. I don't know what plays a major role here, higher thrust or lower weight.
On the visual part I think the BFS material must have a brighter ambient color, it's too dark on a not-sunlit side in space.

 

[francisdrake]

Glad you like it! Placing the engines was indeed a challenge.

After the first implementation, now comes the fine tuning. I have to look into the reduced sea-level thrust, etc. For the performance of the first stage, the amount of flyback-propellant would be interesting.

While some values for the spaceship were given in the presentation, there is nothing on the booster. Right now I assume:

Total lift-off mass 4400 ton (from presentation)
Spaceship: 85 (dry) + 150 (payload) + 1100 (fuel) = 1335 ton
Note: The dry mass is only 6% of the total mass. I would call this unrealistically low, for a ship that has to withstand aerodynamic stresses during reentry.

Booster mass = 4400 - 1335 = 3065 ton
Now the guessing starts:
Booster dry mass: approx. 10% = 300 ton (??)
Booster flyback fuel: 200 ton (??) I have no idea ...
Useable booster fuel = 3065 - 300 - 200 = 2065 ton

These figures may be way off! Please feel free to post your estimations here.

 

[Pipcard]

The original presentation (at around 34:00) said the booster would have a dry mass of 275 t and a propellant mass of 6700 t.

Based on that, the new booster dry mass is (275/(6700+275))*3065 = 120.842 t.

 

[vchamp]

Also from the last year's presentation we know that the booster will use 7% of its propellant for landing: https://i.imgur.com/SdCCdQk.png. I suppose it should be about the same for the new version. So using Pipcard's calculation of the dry mass we have the booster's total propellant mass = 3065 - 120 = 2945 t from which 2945 * 7 / 100 = 206 t is used for landing.

---------- Post added at 11:03 AM ---------- Previous post was at 09:31 AM ----------

People on Reddit and NSF forum posted their calculations which take the 4400 mass as not including the payload. Then the first stage mass is 3215, dry mass is 126.76, propellant for landing is 216 t.

Also they increase the dry/total mass ratio of S1 because they think that 0.0394 from the presentation is too optimistic. Doing the same we can also get a more realistic performance for BFR in Orbiter.

DanHeidel's calculation from reddit:
https://docs.google.com/spreadsheets/d/17n9ynWx_04Tw5AZkgQoHItruoR1oAfhw3FkGOvtQlYk/edit#gid=1894457952
https://forum.nasaspaceflight.com/index.php?topic=43920.msg1732872#msg1732872

 

[Urwumpe]

Also they increase the dry/total mass ratio of S1 because they think that 0.0394 from the presentation is too optimistic.

That's just barely less than 4%. I have doubts its really "too optimistic". Considering the recent technological advances in structural engineering, I would not use this adjective above 3%. Especially the data for the existing A350 XWB and the 787 show pretty well, that the claims of SpaceX are not unrealistic. Challenging, yes. But not impossible with already available technology.

 

[francisdrake]

Vers. 03: Performance update

Launch:
I followed approximately the figures discussed on Nasaspaceflight.com. Now it is more difficult to achieve orbit. After spaceship seaparation pitch up to +40° to prevent falling back into the atmosphere. Activate the sealevel (hover) engines briefly to augument the lifting force, until enough fuel is burned.

Scenarios:
ISS: There is a half-filled tanker a few kilometers out. Undock from the ISS and dock to the tanker. There is docking port at the rear end of the spaceship. Reverse docking is challenging, like parking a car.

Mars: Wait until the southernmost part of the current orbit. Turn retrograde and burn to a periapsis height of 40 km.
Before atmospheric entry turn prograde, righten vessel to horizon and trim fully up. Dampen oszillations with Killrot.
For the final landing, activate a landing MFD. When switching control to the landing MFD, press the [A] key to deactivate the aerodynamic steering.

Note: The atmospheric behaviour has been modified. The AoA is now +30°, the vessel has a slightly better lift coefficient.

The [A] key toggles the aerodynamic steering. A debug string is displayed: Aero on = airfoils are active, Aero off = legacy aerodynamics, no lift, only drag. With Aero off it is easy to righten up the vessel for a propulsive landing.

Model:
A brighter but paler texture is used for the lower part of the spaceship. Let me know if this or the old darker one looks better.

 

[francisdrake]

Uploaded the slightly updated vers. 04 here and on OrbitHangar.
Translation RCS now does no more lead to unwanted rotation.
Solar panels added . Chose the darker belly texture for more contrast.

 

[fatcat]

L key on Launchpad

Just a minor thing.
If you hit the L key on the launchpad, the rocket falls down below the pad.

 

[francisdrake]

That's a good one! Thanks for reporting.
Indeed I had major problems with setting up the launch scenario. The rocket was always toppled over, looking invclined into the ground. No matter if new or old touchdown points.

In the end I copied the values from the default Atlantis launch scenario. If someone could give me a tip how to properly set up a scenario with a launch heading due east (90°) I would be happy!

 

[Axel]

The booster and the spaceship are not able to stand at the surface. After short thrust at launch pad the booster falls down into the surface mesh. The spaceship is looking invclined into the ground, it has even no effect when i use scenario editor to set the ship to a planet.

 

[francisdrake]

Yes, sorry for that, I don't know how to improve that.
Below are the touchdown points of the booster. 3 at the rear end (z = -60, one in the nose). Tried it this way or the other way around (sequence inverted to indicate a different normal vector). For now: Don't test fire the engines, just launch, when it is time.

static const DWORD ntd1 = 4;
static TOUCHDOWNVTX td1[ntd1] = {
    { _V(0, 10, -60), 2e6, 1e5, 3.2, 0.8 },
    { _V(5,-5, -60), 2e6, 1e5, 3.2, 0.4 },
    { _V(-5,-5, -60), 2e6, 1e5, 3.2, 0.4 },
    { _V( 0, 0,  60), 2e6, 1e5, 3.2, 0.4 }
};