Project CST-100

[francisdrake]

Great picture, good looking model!

Did an internet search. If launched on a [ame="http://en.wikipedia.org/wiki/Atlas_V"]402 Atlas V[/ame] the payload may be 12 500 kg.
This website says a 412 Atlas V was selected for the maiden flight, estimating a payload of 13 000 kg.
A [ame="http://en.wikipedia.org/wiki/Falcon_9"]Falcon 9 v1.1[/ame] can lift 13 150 kg, so around 13 ton seems to be a good ballpark estimation.

Based on a total mass of 13 000 kg and an ISP of 3100 m/s for hypergolics, 2000 kg of fuel would result in a deltaV of 518 m/s. For comparision a Soyuz TMA is listed with 390 m/s.

 

[Kyle]

This doesn't look good! D3D9 R10

 

[N_Molson]

This doesn't look good! D3D9 R10

Well I'm using D3D9 R10, and have no problems. Maybe ask to Jarmo ?


Yeah, 500 m/s is probably a bit too generous. 400 m/s is probably more realistic. 13 tons or a little less makes sense too.

But it seems that very little is known about that spacecraft. Do the Boeing engineers even know the definitive numbers ? Not sure. I guess that the capsule should be in the 4000-5000 kg range, given it carries 7 crewmembers and the required systems to keep them alive.

Well, we'll try to make for the best. Maybe, to begin with, something like :

- CM : 6500 kg including
-- capsule structure, heatshield, dock, chutes and systems : 4500 kgs
-- 7 crewmembers ~= 500 kg
-- cargo supplies (1500 kg, maybe more ?)

- SM : 6000 kg including
-- 1200 kg of LOX/Ethanol for LES
-- 1500 kg of hypergolic propellants
-- 300 kg ? of helium or other pressurization gases
-- 2000 kg for the other systems, the umbilical and the structure itself

---------- Post added at 10:28 PM ---------- Previous post was at 10:04 PM ----------

I sent to a 220x240 km orbit a CST-100 with a total mass of 12700 kg, using the Atlas-V 412. The Centaur stage had 300 m/s of Dv remaining (if you admit it can burn the fuel to the lasts drops). So I'd say that's pretty good.

 

[Urwumpe]

Same results here, but RCS of both centaur and DemoCap are sluggish beyond belief... at least the DemoCap should have some mobility.

Also, the autopilot of the Centaur seems to be a bit coarse... I only see -20° pitch, 0° pitch, +20° Pitch. And the first stage of Atlas would better run by using a gravity turn for the early phase.

 

[francisdrake]

Hi, one point on the RCS and abort fuels: You are considering two different fuels, but this paper suggests only one common fuel system:

-- quote--
The SM features an innovative integrated bi
propellant system that uses common tanks to supply hypergolic
propellant for both the conventional on-orbit Reaction Control
System (RCS) and pusher Launch Abort Engines (LAE). For
abort, the propellants are pressurized to approximately 1000 psia
feeding high thrust Pratt and Whitney Rocketdyne BANTAM
derived engines. This unique use of a common propellant
storage system reduces total mass and provides additional
propellant for reboost of destination space complexes or
additional operational flexability.
--unquote--

Both the Orion and the proposed Dragon LAS shall provide around 200 m/s escape speed, with an initial acceleration of 7-8 g. This deltaV would be well within the orbital RCS deltaV budget, so no additional fuel should be required. The only thing required is a high-speed fuel flow to the abort thrusters, which Boeing seems to achieve by a high initial tank pressure. 1000 psia translates to 69 bar (if I did the math correctly )

 

[N_Molson]

Same results here, but RCS of both centaur and DemoCap are sluggish beyond belief... at least the DemoCap should have some mobility.

Also, the autopilot of the Centaur seems to be a bit coarse... I only see -20° pitch, 0° pitch, +20° Pitch. And the first stage of Atlas would better run by using a gravity turn for the early phase.

I was thinking that a space tug like the Centaur isn't designed for manoeuvrability. When the engines are engaged, it doesn't use RCS fuel (engines gimbal). When full with the capsule attached, we almost have a 35-tons vehicle.

The code of the capsule is not finished at all, RCS is not really implemented yet, so this is going to improve a lot, of course.

Also, the autopilot of the Centaur seems to be a bit coarse... I only see -20° pitch, 0° pitch, +20° Pitch. And the first stage of Atlas would better run by using a gravity turn for the early phase.

Yes, I may refine the Centaur guidance, though it works with some efficiency despite of its flaw, so it isn't a top priority. Notice that, concerning the Atlas CCB, you can modify the serie of pitch targets to bend the trajectory at your convenience.

Hi, one point on the RCS and abort fuels: You are considering two different fuels, but this paper suggests only one common fuel system:

Ah, this is very interesting and of course makes much more sense. Also I found this :

This hardware demonstration (CCDev-1 configuration shown in Figure 7) utilizes a 58,000 lbf class engine previously built under the BANTAM demonstrator program and operated using hypergolic propellants. CCDev-2 testing will demonstrate engine start transients, performance, and combustion stability using bi-propellant fuels.

- 58,000 lbf is more than the number I found previously. So it will boost up the LES a little bit, which isn't a bad thing.

- The engine is, AFAIK, the RS-88. I've seen on Astronautix it uses LOX/Alcohol. But looks like they made it running with hypergolics. Also they seem to say that they will evolve to a bi-propellant system. Is it possible to use that for the RCS ? And what would be the prop/oxydizer combination, then ? Not something cryogenic, any idea ?

 

[Urwumpe]

Yes, I may refine the Centaur guidance, though it works with some efficiency despite of its flaw, so it isn't a top priority. Notice that, concerning the Atlas CCB, you can modify the serie of pitch targets to bend the trajectory at your convenience.

Yes, but a gravity turn is rather a function between two pitch targets. By acceleration, gravity and initial pitch rate, you calculate the new pitch rate for letting gravity turn you towards your target pitch.

Usually, this also means almost zero AOA and is a very effective maneuver.

 

[N_Molson]

Yes, anyway I will implement later a loop that checks the dynamic pressure against the AoA, and will destroy the rocket when there is a high dynamic pressure and too high (a few degrees) AoA, pitch & yaw. So I'll have to improve the guidance anyway, there will be only a narrow "valid" corridor especially around MaxQ. Configurations with more SRMs will be even more sensitive to that, since they typically have an higher MaxQ.

 

[Urwumpe]

Yes, anyway I will implement later a loop that checks the dynamic pressure against the AoA, and will destroy the rocket when there is a high dynamic pressure and too high (a few degrees) AoA, pitch & yaw. So I'll have to improve the guidance anyway, there will be only a narrow "valid" corridor especially around MaxQ.

Would be more important for the capsule, since you can assume that the player will use the launcher as if it is flying on rails. And a capsule can also be destroyed by a too high dynamic pressure

 

[N_Molson]

Well, the idea is that if the capsule is destroyed by dynamic pressure, the rocket is then destroyed in a fraction of a second. And this is convenient for other kind of payloads. As the user can set some parameters to change the ascent course and trajectory, it will add some difficulty and realism. Also I plan to add some random failures that could spice things up. Ideally, the rocket should be able to "detect" the CST-100, and the DnP check settings be adjusted if you use the capsule, a 4-meter diameter fairing, or a 5-meter fairing. I've seen interesting graphs about this in the Atlas V manual.

But yes, it could be interesting to have this option enabled for the capsule too, because one of the strengths of the CST-100 is its high-compatibility and supposed ability to be launched by the Atlas V, the Delta IV, or the Falcon 9. Modelling the two other launchers is out of the scope of this addon, and there are already excellent addons for those two.

---------- Post added at 02:08 PM ---------- Previous post was at 11:32 AM ----------

I found interesting data about the maximum dynamic pressure endured by the Apollo missions. Gives an idea.

http://history.nasa.gov/SP-4029/Apollo_18-20_Ascent_Data.htm

Now I have trouble with the conversions (I'm terrible at maths)

The maximum is 776.938 lb/ft² for Apollo 8.

But how to express that in kPa/m² (I guess it is the unit Orbiter uses) ? :sos:

 

[Urwumpe]

Orbiter uses Pascal (Pa) as unit of dynamic pressure. 1 Pa = 1 N/m²

1 lb/ft² = 47.8803 Pa

776.938 lbf/ft² = 37.2000245214 kPa

EDIT: Just in case you also need those one day:

1 psi = 6,894.76 Pa

1 inHg = 3,386.39 Pa

 

[N_Molson]

Thanks, and very interesting, as the rocket currently reaches a maxQ of nearly 30kPa. Seems we're on the right order of magnitude. :yes:

 

[Urwumpe]

Some improvements that I would like to see to the orbit insertion:

• No rotation during capsule separation
• Instead firing of forward linear RCS thrusters for separation.
• Possibility to subtract the DV by the capsule separation from the Centaur burn target
• (Semi-) Automatic CAM of the Centaur after capsule separation

 

[N_Molson]

No rotation during capsule separation

- Easy, can be defined in the .scn file, see readme.txt

Instead firing of forward linear RCS thrusters for separation.

- I don't get that one, seems a word is missing ?

Possibility to subtract the DV by the capsule separation from the Centaur burn target

- I'm not sure I get it, you mean knowing how much Dv is required to reach orbit ? Interesting idea, but I'm not sure to know how to calculate that.

(Semi-) Automatic CAM of the Centaur after capsule separation

- Yep, planned and will be done.


Now a pic of the 20 main propulsion thrusters firing at the same time :

Of course white-ish particle streams will look more realistic, but standard exhaust is nice to get the thrusters locations. Also I set a common fuel for RCS and LES, which can become an "emergency" (and probably suicidal) de-orbiting device :lol:

 

[Urwumpe]

No word missing there... simply using the "20 main propulsion thrusters" as you call them for a few seconds for getting faster distance between capsule and centaur. This burn will accelerate the capsule and add an additional DV of a few meters per second. By subtracting this DV from the targets of the centaur, the capsule would be in the target orbit after the separation burn, and the centaur slightly below.

 

[N_Molson]

OK, got it

 

[Urwumpe]

I don't know how you currently calculate the Centaur trajectory - did you consider implementing a PEG?

http://www.orbiterwiki.org/wiki/Powered_Explicit_Guidance

 

[N_Molson]

I don't know how you currently calculate the Centaur trajectory - did you consider implementing a PEG?

Uh well this looks very good and I get the point of it, but again that's far above my loosy math skills, not saying I would have to express that in C++. :shifty:

Currently, the guidance system is hybrid. First stage works by a simple definition of pitch targets against time, and also engine throttling at different steps of the ascent. This is activated once the heading is determined by the initial roll. Any pitch/yaw/roll error is corrected using some dampening system to avoid over-reactions and oscillations (the closer the target, the smaller is the allowed max angular velocity).

Second stage uses a somehow more autonomous system. It tries to reach a set apoapsis, and adjusts pitch up or down in regard of the altitude and vertical velocity. It is possible to modify some parameters in the .scn file to get different target orbits. It works, but lacks of precision and as you said it is a bit rough around the corners. Also, it can follow a pre-set heading.

I guess it would look absolutely barbaric to a professional coder :embarrassed:

 

[Urwumpe]

Uh well this looks very good but again and I get the point of it, but again that's far above my loosy math skills, not saying I would have to express that in C++. :shifty:

Is not that hard... the worst that happens to you there, is the need to have a favorite matrix solver (for linear equation systems) and implement it. That is usually the point where you wish for somebody implementing the [ame="http://en.wikipedia.org/wiki/OpenFOAM"]OpenFOAM[/ame] math library for Orbiter. :lol: :lol:

I have some old lecture notes here for matrix solvers, also you can of course look into the source code of some add-ons that already implemented PEG.

I guess it would look absolutely barbaric to a professional coder :embarrassed:

Actually, it is just that looking out of the capsule during such a bumpy ride makes the christmas dinner become weightless in me. :lol:

(You don't want to know how the code of a professional coder looks like... time pressure can turn the worst out of nice people)

 

[N_Molson]

Actually, it is just that looking out of the capsule during such a bumpy ride makes the christmas dinner become weightless in me.

That's nothing. Boeing has a solid stockpile of "little paper bags" :lol: