Project Starship Mk.1

[BrianJ]

Hi,
thanks so much for the input - glad you're having some fun with it. Yes, the aerodynamics are very simplistic at the moment, just to see how this thing behaves. Without any data from SpaceX, it's hard to know what their design strategy is. I'm getting the impression that the Starship is going to fly at very high AoA pretty much all the way, so I'm tempted to give it a little stability at that attitude. The CoG and CoP are where they are(on the add-on) because thats where the wings and canards "balance" (1:3 distance to CoG ratio) if the constraint is realistic wing/canard area and position - if I simply shift the CoG forward by 3m (gives a 1:2 ratio) then I'd have to have the wings at ~50deg deflection to balance it. Or change the CoP for the main hull aerofoil, etc. But maybe it does work that way!
Only clue I really have is the "Mars Reentry" animation where I think I can see the CoG is just about where the wings start, but that's on a previous design.

If you or anyone wants to have a go at tweaking the aerodynamics, please do! And if possible let me know how it goes

I'm still wondering if the "default" position for the "fins" is full-deflection, and steering done by extending them. I'm doing it the other way round.

In the meantime, I've made the RCS a bit better(more balanced) and added an auto-throttle to the "Retro Attitude" function, so now it's pretty much a "Flip'n'Land" function. Works well from around 4km alt. or maybe even lower. I'm gradually getting an idea of how the final landing zone targeting is going to have to work.

And I've made a very simple .cfg based "BocaChica" base to fly from and back to, so I'll put it all in a new beta test .zip very soon.

Cheers,
Brian

 

[francisdrake]

I'm still wondering if the "default" position for the "fins" is full-deflection, and steering done by extending them. I'm doing it the other way round.

After flying a few reentries I think the fins will be partially swept back during hypersonic flight. There are some reasons for that:

The shockwaves will extend from the nose and belly at some angle upwards and backwards (relatively to the hull). If the fins are in the slipstream inside of these shockwaves, the mechanical stresses on the fins, the rotating joints and actuators will be much lower. They may extend only a little out of these protected zone if corrective actions are required.

During hypersonic flight the hull is a lifting body, providing enough lift and drag. No additional 'wings' are required.

The swept-back fins augment roll stability. If the hull wants to roll unintendedly, the fin on the forward-rolling side gets more exposed to the airflow, thus creating a self-aligning torque to righten the ship.

As the angle of shockwaves decreases when the vessel slows down I would expect the fins will be more and more extended, until they are nearly fully extended during the subsonic skydiver fall, to maximise the drag area.
Still I would assume they keep a slight v-shape for roll stability.

 

[BrianJ]

Hi,
yes, that all makes sense to me. At the moment all I have implemented is the "Reentry Mode" with fins fully swept back, there is some stability at 60deg AoA, and I've added the capability to crank the rear fins in or out by about 5deg so you can trim the stable attitude from 45deg - 75deg. Steering is currently done by RCS only.

Do you think it would be feasible to use the fins for a limited amount of steering during reentry phase? I had just imagined that would be too much stress for the fins - but my intuition about hypersonic/high altitude flight is probably not good! Using the fins for steering as much as possible saves RCS propellant which is in short supply anyway.

I was trying some experiments with the "Skydive" mode (fins fully extended) aerodynamics yesterday, just trying to shift the CoP to see if I could get a bit of stability, but didn't hit on anything useful yet.

It did occur to me that I should have some "vertical" aerofoils (again with variable area) to represent the fins when they are retracted - tried to implement that but it made the whole thing go crazy! Am going to have another try, though, since it seems like a reasonable idea. I can't help thinking it should improve stability if I can get it right.

And maybe an "Intermediate" mode with fins partially swept back would be a good thing to look at and see how it works. Need to do some work on the fins steering logic for that.

Thanks again for your thoughts and input
Cheers,
Brian

 

[francisdrake]

The vessel seems more stable with the fins swept back. I see a change in the acceleration between extended / retracted which I assume is done by changing the wing area?

Unfortunately Orbiter seems to allow only horizontal and vertical wings, defined by LIFT_VERTICAL and LIFT_HORIZONTAL. I would add vertical fin(s) with the area of the main fins x sine of the sweep angle. If you keep their lift function low, there should be no crazy spinning.

Could the fins being used during reentry? Well, 'normal' lifting bodies like ESA's Space Rider use body flaps to trim the AoA during reentry. So I would assume a slight deviation from the swept-back orientation would be ok to assist the RCS.
By the way, I guess the swept-back angle is more than 45°, maybe like 70° for the initial reentry phase, to keep the forces on the fins low.

A note on future versions: The hinges on the current SpaceX Starship look makeshift. I doubt they will be the same on the final production versions. They might either be more blended-in into the body, or the body will get a sharp edge just below the hinges to force airflow separation.

Looking forward to your next iteration of this very special flying machine!

 

[BrianJ]

Hi,

The vessel seems more stable with the fins swept back. I see a change in the acceleration between extended / retracted which I assume is done by changing the wing area?


Unfortunately Orbiter seems to allow only horizontal and vertical wings, defined by LIFT_VERTICAL and LIFT_HORIZONTAL. I would add vertical fin(s) with the area of the main fins x sine of the sweep angle. If you keep their lift function low, there should be no crazy spinning.

Could the fins being used during reentry? Well, 'normal' lifting bodies like ESA's Space Rider use body flaps to trim the AoA during reentry. So I would assume a slight deviation from the swept-back orientation would be ok to assist the RCS.
By the way, I guess the swept-back angle is more than 45°, maybe like 70° for the initial reentry phase, to keep the forces on the fins low.

Yes, what I have so far is a vertical and a horizontal aerofoil for the Hull 450m^2 each, two Canard aerofoils 18m^2 each, two Wing aerofoils 54m^2 each. The Canard and Wings are scaled by cos(deflection) angle.
There is now also a "vertical" aerofoil for the wings,scaled by sin(deflection). This adds a bit more yaw/roll stability.

I moved the hull CoP up and back a tiny bit, which gives an AoA of ~58deg. And you can trim that up or down by setting a small deflection in the wings for extra or less drag.

Default fin deflection is 60deg in Reentry mode.
Aerofoil steering in Reentry mode now possible.

A note on future versions: The hinges on the current SpaceX Starship look makeshift. I doubt they will be the same on the final production versions. They might either be more blended-in into the body, or the body will get a sharp edge just below the hinges to force airflow separation.

Yes, the Mk.1 is obviously not designed to go much above 20km!
I'm just distracting myself thinking about doing reentries with it, I know


Added throttle control to the Anti-groundspeed attitude function - so now it's a "Flip'n'Land" function - use from 4km to 2km alt. depending on vertical speed. No targeting though.

Also added BocaChica base (very simple) and a couple of scenarios including a recorded REPLAY of my best landing at BocaChica.

If you're doing any flights from Texas, I can recomend grabbing the Hi-Res Earth textures for the BocaChica area here:
https://mirror.orbiter-radio.co.uk/orbiter/assets/packages/Earth/EarthHi_10_07.zip

Anyway, here's the latest beta .zip
View attachment starship_mk1_beta_191021.zip

Cheers,
Brian

 

[francisdrake]

I took the Starship for a trip from LEO to lunar landing, using the 'Challenge 2 - Lunar transfer' scenario, replacing the Delta Glider with Starship.
It worked, after some trial and error. The picture above shows the ship incoming using the Flip'n'Land function.

Very nice trip! Started with full tanks from LEO and arrived with 12 ton fuel remaining on the lunar surface.

Two remarks:
Maybe the wing movement could be turned off completely for space operations (e.g. as an additional 'B' aerofoil mode).

I had occasional Orbiter crashes using the Killrot command in deep space. Maybe this was related to the wings flapping out of bounds, or their control algorithm exceeding a value, dividing by zero, or so.

 

[BrianJ]

Nice! So much for "not designed for going above 20km" :thumbup:
You can turn off the aerfoil steering with ALT + NUMPAD /.
I'll have a look in the code for any possible CTD causes - could well be a math problem, away from Earth. Can't see why it would be related to KillRot function, but will check.

I learned from this talk that it's not a "Flip'n'Land", it's a "Reorientation Maneuver"

Cheers,
Brian

 

[francisdrake]

Played around to dock the Starship to a tanker. Added a docking port to the rear of the Starship through the cfg-file.

Looking backwards while docking is possible by changing the view with 'E'. The backward/forward translational RCS is a little sluggish, as the thrusters mainly blow outward, only little in axial direction.

 

[Jeremyxxx]

When is the rocket for the Starship (BFR) going to be made? because I have been wanting to see an unmanned variant for cargo scenarios.

 

[BrianJ]

Played around to dock the Starship to a tanker. Added a docking port to the rear of the Starship through the cfg-file.

Looking backwards while docking is possible by changing the view with 'E'. The backward/forward translational RCS is a little sluggish, as the thrusters mainly blow outward, only little in axial direction.

Cool! In future, I might add a tanker+fuel-transfer variant.
My Mk.1 RCS is definitely not designed for orbital operations
But I hope we'll get some clear pics of what SpaceX are actually doing with the RCS, soon.

When is the rocket for the Starship (BFR) going to be made? because I have been wanting to see an unmanned variant for cargo scenarios.

At the moment, I'm still having fun trying to figure out some kind of landing guidance for the Starship (I hope SpaceX are doing better than I), so might be quite a while before I get to doing boosters.


Cheers,
Brian

 

[BrianJ]

Just a shout out to ask if anyone has any recent info or photos of the latest Starship build (SN.4?) regarding RCS placement (if any!) and aft blisters (will there be any, now the landing gear folds up inside?). I'm not finding much via google. Educated guesses welcome

Finally made some progress with landing guidance algorithm (flies like a whale, steers like a whale). It's happy without Orbiter's "Atmospheric Wind Effects"*, but can just about cope with it on.

*I do wonder if Orbiter's surface winds are a bit extreme - some time ago, I tried calculating the "Wind shear" values up to 10km alt. and they seemed quite high, but I'm no pilot, and not sure if I was calculating correctly. It's just that you get hammered by a 50mph SW wind at ~3km alt., and then it swings round to 50mph N at ~1.5km alt. Any vessel with low lateral speed gets blown about somewhat!

Anyway, I think I can start making a launch pad for Boca Chica finally
Cheers,
Brian

 

[DaveS]

AFAIK, SN-4 is just the main propellant tanks plus the thrust section equipped with a single Raptor engine. Only static fires and a single 150 m hop (very similar to what the Star-Hopper did last year) planned for it. No nose cone, fins or thrusters. This is thread on NASASpaceflight.com that you should keep an eye on as far as Starship development is concerned: https://forum.nasaspaceflight.com/index.php?topic=48895.2540

 

[4throck]

:hesaid:

Just to add that SN5 will have 3 engines and probably a more complete attitude control system. That's the one to look out for.
SN4 is mostly an engine/tank tester.

 

[BrianJ]

Thanks for the pointers, guys NASA forum page bookmarked.
So, I guess I'm chasing the SN5, with an eye on the "20km hop" test (if that's still in the plan). Interestingly, I find I only need RCS to finesse the landing final approach from 6km alt., and only if Orbiter's wind is active. If SpaceX have better aerodynamic control and wind compensation strategy, they might not need RCS at all, for the 20km hop.
Cheers,
Brian

 

[DaveS]

Thanks for the pointers, guys NASA forum page bookmarked.
So, I guess I'm chasing the SN5, with an eye on the "20km hop" test (if that's still in the plan). Interestingly, I find I only need RCS to finesse the landing final approach from 6km alt., and only if Orbiter's wind is active. If SpaceX have better aerodynamic control and wind compensation strategy, they might not need RCS at all, for the 20km hop.
Cheers,
Brian

Plans seem to change daily, so what is currently the plans for SN-5 could change tomorrow. It seems to be linked to the fact that despite outward appearances, they still don't have the manufacturing down. This is further evidenced by the fact that they barely got SN-4 through its cryogenic pressure test. They were aiming for 123 PSI (8.5 bar) but called it done at 71 PSI (4.9 bar).

 

[BrianJ]

Thanks. Yes, not easy to keep up with SpaceX plans ;-)

 

[BrianJ]

3D modelling fun with many girders......

I guess they probably won't use that stand for anything other than test firing, but it'll do for me for now.

Another day at (virtual) Boca Chica....


@Donamy
It was nice of BlueOrigin to lend SpaceX their Z6034 cherry-picker ;-)
I updated it so it works in Orbiter2016, although it tends to slide on slopes.
Zip file attached for you, at the bottom of this post.
Cheers,
Brian

View attachment z6034_200505.zip

 

[BrianJ]

SpaceX Starship Mk.1 v.200507 now on OrbitHangar:
https://www.orbithangar.com/showAddon.php?id=258256fb-677e-4266-97c6-7dd598816891

I had a lot of fun thinking up another crazy guidance algorithm for this one I didn't really get anywhere until I figured I had to maximise horizontal velocity in order to maintain a decent amount of control authority. Even so, it still has trouble with Orbiter's winds if approaching your landing target from the East (although it will try to avoid that situation). Anyway, I think it's good enough for me to start thinking about an orbital version.

Also, I see SpaceX have published a "Starship User's Guide" .pdf:
https://www.spacex.com/sites/spacex/files/starship_users_guide_v1.pdf

...and we've had a Raptor test fire already:

Cheers,
Brian

 

[MarsGlider]

I've been loving this mod since I've seen it and I have done a 186 km hop with it landing (of course I modified the fuel amount if your wondering), pretty crazy!

 

[BrianJ]

Cool! Hopefully, if I can make a reentry program that gets it from orbit to 30km alt. somewhere near the target, the final approach and landing are already pretty much taken care of