SpaceX Falcon Mission Simulations

[OneSpeed]

They had no crew and just a few hundred kg of cargo.  There should have been literally tonnes of extra margin.  Was it *all* used to reduce MaxQ?

I suspect most of the margin was spent on the increased capsule mass, which I estimate is a difference of about 3t, including the crew mass simulators. DM-1 would also have incurred significant gravity losses through Max-Q, and the sim suggests there might have been about 2t of propellant left in S2 at SECO (a tonne more than ullage might require). So with a more aggressive ASDS profile, F9 + Crew Dragon may well be able to loft 4 crew and another couple of tonnes of cargo in the trunk. Using RTLS, I doubt there would be much payload at all, but it will be fascinating to see what DM-2 can do.

[IainMcClatchie]

Since this was a demo flight, you'd think they would have wanted to demonstrate, if not the maximum possible mass to station, at least the nominal amount of mass to station.  And though crew mass simulators in the seats seems like it might be needed to demonstrate that the seats won't break, it seems odd that you'd not at least pack water in those seats.

Isn't the point of Dragon-2 to eliminate the need to get rides on Soyuz?  That thing is packed nearly solid on the way up.  If Dragon-2 is going to replace it, it'll need to bring the same amount of stuff, at minimum, right?

So how was this a demo of that capability?

[Kansan52]

NASA likes slow steps.

They Demo-ed a new capsule (with lots more room than a Soyuz).
Plus new docking mechanism.
Plus new (use of) control room.
ect...

Lots of Demos.

[Alexphysics]

Since this was a demo flight, you'd think they would have wanted to demonstrate, if not the maximum possible mass to station, at least the nominal amount of mass to station.  And though crew mass simulators in the seats seems like it might be needed to demonstrate that the seats won't break, it seems odd that you'd not at least pack water in those seats.

Isn't the point of Dragon-2 to eliminate the need to get rides on Soyuz?  That thing is packed nearly solid on the way up.  If Dragon-2 is going to replace it, it'll need to bring the same amount of stuff, at minimum, right?

So how was this a demo of that capability?

Soyuz looks packed but it is because it is tiny. It doesn't carry too much cargo, less than 100kg. Crew Dragon has carried 180kg on this flight and just with a few boxes of cargo below the seats...

[Slarty1080]

Has anyone ever tried simulating how far a Starship could get without Superheavy for the purposes of Point to Point? If not how easy would this be?

[OneSpeed]

Has anyone ever tried simulating how far a Starship could get without Superheavy for the purposes of Point to Point? If not how easy would this be?

The answer is extremely sensitive to the Starship dry mass fraction, and the Raptor Isp and thrust. If you happen to have those numbers, I'll post a sim in the SpaceX 'Star series' simulation thread by the weekend

https://forum.nasaspaceflight.com/index.php?topic=47179.0

[AlesH]

I agree that the results (for point to point "hops") strongly depend on the actual Starship parameters. I tried to use 85 tons of dry weight, 1080 tons of fuel for take-off and 30 tons of fuel for landing. Then a theoretical range with a payload of 100 tons is about 3000 km, with a payload of 50 tons about 5000 km, with a payload of 30 tons nearly 7000 km and with payload of 10 tons about 10000 km, maybe more. Height of the hops is 500 to 2000 km. The flight time is from 20 minutes to max. 1 hour.

I believe the inaccuracy of my estimates is less than 20%, but I will be happy if someone independently verifies it.

All above requires sufficient thrust for take-off. So the 7 Raptors must have a total thrust over 14 MN (1400 tons) [2 MN (200 tons) thrust each of the 7 engines].

[IainMcClatchie]

These calculations are bizarre.  In suborbital flight the gravity losses are higher than orbital flights.  The Starship isn't very well designed for unboosted operation in the Earth's gravity well.  It's fully fuelled weight is 1185 tonnes, full thrust is 14 MN, for a takeoff acceleration of about 2 m/s^2.  Raptor is expected to have an excellent thrust-to-weight ratio.  If you were actually going to operate suborbitally, you would add a couple of Raptors to reduce gravity losses.

[OneSpeed]

From the In Flight Abort (IFA) Draft Environmental Assessment (attached), the abort test sequence does not have distinct timings, altitudes, distances and velocities. Rather, it specifies ranges for all of those values, as per the table below.

So, what I've attempted to simulate here is the mean of all of those ranges. Interestingly, the separation event range is from 83-100 seconds. All of these values are well after MaxQ, which most of us assumed was the logical time to perform the abort. There is another consideration though, and that is that the acceleration being provided by the booster at MaxQ is actually quite low, around 1.6gs. In the sim, the abort is at 90 seconds, when acceleration has increased to 2.4gs.

However, if S1 thrust terminates at 90 seconds, acceleration quickly drops to 0g, and dynamic pressure on the capsule has dropped well below that for MaxQ, from 27.4kPa back to 12.3kPa. Unless there is some overlap between S1 shutdown and SuperDraco ignition, it is difficult to see the point of the test.

Anyway, at 90 seconds the capsule appears to have sufficient thrust to provide around 4.4gs of acceleration, increasing to over 5gs as propellant is expended. This would be sufficient to escape S1 regardless of whether or not S1 shuts down.

In the video, I've provided a 'picture in picture' (PIP) view of the capsule, hoping to reduce repeated zooming in and out. It's not perfect, but let me know if you think it helps or not.

[S.Paulissen]

It was my understanding that they were conducting the abort at Max drag, not max q.  Which is typically a little after max q.

[OneSpeed]

It was my understanding that they were conducting the abort at Max drag, not max q.  Which is typically a little after max q.

Yes, the simulation predicts MaxQ at 59 seconds, and max drag at 60 seconds. Oddly, both of these events are well before the separation event range which is from 83-100 seconds.

[wes_wilson]

It was my understanding that they were conducting the abort at Max drag, not max q.  Which is typically a little after max q.

Yes, the simulation predicts MaxQ at 59 seconds, and max drag at 60 seconds. Oddly, both of these events are well before the separation event range which is from 83-100 seconds.

IANARS - But is max q and max drag the same for the full stack as it is for the capsule accelerating alone?  (naively I'd think no?) Is there any possibility the discrepancy is because they're looking at max loads on the capsule alone after it begins to abort away and is accelerating harder than the separated 1st and 2nd stage?

[IainMcClatchie]

For MaxQ to be different than max drag, the drag coefficient would have to be changing.  Are you modelling that?  How do you guess the drag coefficients at different speeds?

[OneSpeed]

For MaxQ to be different than max drag, the drag coefficient would have to be changing.  Are you modelling that?  How do you guess the drag coefficients at different speeds?

The way SpaceSim works is that each stage of the vehicle is assigned a base Cd and Cl. Each value is considered a constant up to Mach 1, or is multiplied by an exponent above Mach 1, e.g.

public virtual double GetBaseCd(double baseCd)
{
   if (MachNumber > 1.0)
   {
      double exp = Math.Exp(0.3 / MachNumber);
      return baseCd * 1.4 * exp;
   }

   return baseCd;
}

The base Cd depends on the fineness of the stage, and if it is connected to other stages that could extend the fineness, or increase the frontal area if they are side by side. E.g. for a base Cd of 0.8, the curve up to Mach 25 is of the form shown below. The coefficients are also modulated by the AoA. Here is a simplified example:

public override double FormDragCoefficient
{
   get
   {
      double alpha = GetAlpha();
      double baseCd = GetBaseCd(0.;
      return Math.Abs(baseCd * Math.Sin(alpha));
   }
}

For rockets that have flown, the coefficients used are the result of years of analysis (literally) of real rocket telemetry. Although it is just a model, it can make surprisingly accurate predictions.

[Slarty1080]

Elon Musk stated that the second Falcon Heavy would have 10% more thrust than the first flight. Some people have said that this will have minimal effect <200m/s gain avoiding some gravity losses by burning out earlier, but I’m not so sure. It should be possible to throttle down the central booster engines to preferentially conserve fuel in the centre stage. In fact I think that this technique is already used.

So my question is if all of the throttling of the stack could be achieved by the central booster, what effect would this have on its capability? And with 10% extra thrust available what then?

I’m not sure that the central core could be throttled sufficiently to achieve this, but two options might make it possible. Firstly all engines in the core stage could be made throttle capable and throttled and secondly some engines could be shut down. It has been suggested that shutting down engines would pose a risk, this is arguable, but might be worth investigating if it gave sufficient advantages.

So what might be achieved by these techniques? 93% total thrust is already achievable if all three throttled core engines are throttled to minimum (39%). 80% thrust is achievable if all 9 core engines were throttle capable and 67% thrust would be achievable if all core engines were shut down.

[acsawdey]

Well, one data point is that from the timeline released by spacex, the booster sep is at 2:30 (similar to F9) but the core booster doesn't cut off until 3:30. So the overall average throttle on the core booster has to be 71% of that of the side boosters to make that work.

[Barley]

(3 * 39% + 6 * 100%)/9 = 79.7% > 71%

[Slarty1080]

(3 * 39% + 6 * 100%)/9 = 79.7% > 71%

My thoughts entirely. so what gives?

[niwax]

(3 * 39% + 6 * 100%)/9 = 79.7% > 71%

My thoughts entirely. so what gives?

Speculation on other threads is that some booster engines cut out early to limit acceleration further, there was a callout along those lines

[Slarty1080]

(3 * 39% + 6 * 100%)/9 = 79.7% > 71%

My thoughts entirely. so what gives?

Speculation on other threads is that some booster engines cut out early to limit acceleration further, there was a callout along those lines

That would be significant 71% = 6 engines out and 3 engines at 39%! Do they really cut the core engines, I would be very surprised? Or is there some other more mundane explanation?