SpaceX COTS Demo 1 Updates

[meekGee]

Yeah, that's why I was asking about the bearing type...  but, since it's built for reusability, I don't think it's dumping the momentum that quickly, and as you show, it's parallel anyway.

So, I agree, most likely an impulse, and probably intentional, action.

So look at the picture I posted a few posts ago - clearly the interstage is catching a good chunk of the plume.  Is that healthy for the chutes inside?  Maybe the quite hurried tumbling is also to get the front of the interstage out of the direct line of fire, while still firing the second stage asap?

[jongoff]

Actually, even though the interstage is definitely catching some of the exhaust plume, the rolling definitely starts before the second stage lights up.

Though the Merlin-Vac engine might be running a LOX lead or purge before main ignition (just based on my experience at Masten--if you watch the Masten in-air relight video you can hear what I'm talking about right before the engine lights again), which probably wouldn't be visible, but could impart a decent amount of force on the upper part of the first stage.  And as someone earlier pointed out, it would be pretty easy for a force like that to cause a no-longer 3DOF controlled stage to tumble.  Just a guess though.

Here's a question for someone who might know - when the 9 Merlins shut down, what happens to the angular momentum stored in the pumps?  Do they spin on regular bearings or something fancier like air bearings?   Any chance some knows how much the rotors weigh and how fast they are spinning, and most importantly - in what direction they are mounted?

That's another option. 

Edit: or as you and others have said, they may have a way of intentionally tumbling the stage. 

~Jon

[Toner Soprano]

Maybe it's just me.... but I'm not seeing the interstage part of the video or the separation from inside the interstage.  Was the video edited again? 

[stockman]

Maybe it's just me.... but I'm not seeing the interstage part of the video or the separation from inside the interstage.  Was the video edited again? 

it appears, as someone mentioned earlier, that SpaceX has changed up the video... yesterday they had the HD flight highlights and today it looks like regular def with a different mix of views - none of the interstage camera's for example... Not sure why they would take it down now that its out in the public domain? I have my copy and I am sure most others do as well... nothing to hide now really... I wonder if someone posted the wrong video and had to do a quick swap before the boss found out...

[Lars_J]

I downloaded the HD version of the clip with the interstage split screen. Sharing it would be tricky, though, unless someone has an easy place to upload it.

As for as the turbopump orientation, it must be vertical. Remember F9 flight #1, where the startup torque caused the whole LV to rotate at release. So I don't think that is the source of the stage rotation after separation. It looks intentional.

[ChefPat]

I downloaded the HD version of the clip with the interstage split screen. Sharing it would be tricky, though, unless someone has an easy place to upload it.

YouTube.

[ugordan]

I wonder if someone posted the wrong video and had to do a quick swap before the boss found out...

Come to think of it, we had a situation here recently when a drawing of Taurus II first stage was removed because of... well, apparently ITAR. In light of that this move appears less surprising. We were essentially given engineering camera views of the innards of a new U.S. vehicle. Sensitive/proprietary stuff? Probably.

[simcosmos]

Edit:  Go for it, start the thread.

Hi Martin (all),

I'm not ready yet to open a new thread focusing on the Falcon9-Dragon COTS Demo1 trajectory + performance reconstruction because that kind of subject contents really takes time and extra effort to properly reference and share the input data (and to do all the related results contextualization + discussion in public, if the objective would be to try to understand a little better the flight and, answering to TimL to try to keep speculation within some constraint, at least for what I have called of a given set of input parameter 'families'  ).


I. Playing with Falcon9 + Dragon: Mass Breakouts + Performance Specs

As a side comment, I have also built a spreadsheet to play a little with preliminary Falcon9 'Block1 vs Block2' mass / performance data as well to play with several conceptual Dragon spacecraft configurations (example: COTS Demo1 vs COTS Dragon configurations for Falcon Block1 and Block2 vs same exercise but for crewed Dragons with the addition of assuming a more standard LAS vs integrated LAS, etc, etc). The 'results' of such preliminary spreadsheet are then used as departure points for simulation work (both on the launch vehicle and on the spacecraft performance implementation + trajectory for a given mission type).



II. SpaceX COTS Demo1 Virtual Reconstruction

As mentioned earlier on page 82, I think that was more or less able to replicate the COTS Demo1 flight, which is not, by itself, a sign of a given set of  input parameters being all 100% correct but, at least, helps to constraint a little more some specific parts of the generic puzzle (I mean, it helps with questions formulation). 

Only as example, one of the COTS Demo1 SpaceX vehicle 'possibilities' that has been simulated here had a total pre-ignition mass of ~313200 kg and ~48470 kg total vehicle mass at simulated staging (including upper stage mass + trunk + cubesats + Dragon capsule + nose cover). From there, I was able to more or less meet altitude vs velocity vs MET calls (although extra brainstorm and clarification would be needed on a number of trajectory / ascent events details and also on some of the input vehicle parameters... there are a few things for which I could have questions / comments to make but not much time to share those doubts now...).

Anyway, despite this is far from completed (haven't touched much on that simulation since the last time I wrote in this thread), one of the sets of files that have 'coded' achieves the following approximated performance (not yet an exact replication of real life results and without entering, for now, in extra details about the mass breakout, although I did further differentiated the masses on the simulation files):

288 x 301 km, 34.5 inclination, 10647 kg (US+USprop+Trunk+cubesats+DragonCapsule)

After releasing the Dragon capsule and the cubesats, the final burn state was simulated as something like:

  4110 kg : Upper Stage + Trunk
  2777 kg : Upper Stage Prop.
---------------
  6887 kg

Which, after a burn of a little more than 20s or so (for a dV of ~1567m/s or so), resulted in 281 x 11100 km, 34.61 inc:

 4110 kg : Upper Stage + Trunk
   121 kg : Upper Stage Prop.
---------------
 4231 kg
 


III. Upcoming SpaceX COTS Demo Flights
(previsions, by using demo1 data?)

To end, will soon attach to this post, one of the virtual telemetry outputs (Altitude vs MET) that resulted from using one of the COTS Demo1 estimation / simulation data to extrapolate the performance, guidance / trajectory for the delivery of  ~6800 kg cargo Dragon spacecraft into something like ~250 km circular orbit / ISS inclination (with ~11.1t – 6.8t (spacecraft) = 4.3t being upper stage mass and related remaining prop.)

This is all very preliminary (December last year) but it seems that COTS Demo1 flight used a lightweight Dragon (both in terms of capsule and trunk mass): from my clumsy preliminary research + 'spreadsheet' math (focusing mostly on spacecraft and F9 Block2 delivery expectations, I haven't simulated Falcon9 'block2'), a fully loaded cargo Dragon, on a Falcon9 'block2' could be ~2.5x as heavy as the mass breakout that I have assumed for the simulated Dragon COTS Demo1. On another note, something between COTS Demo1 and a full cargo Dragon on F9 'Block2', a kind of Progress equivalent (in terms of rough total mass) Dragon spacecraft flying on a Falcon9 'Block1' could perhaps be ~1.6x or so the mass of COTS Demo1 Dragon.

But again, all these are just and only very loose notes based on personal interpretations  and related personal simulation work of available scattered data: extra details would require additional work + presentation preparation + discussion. As a minor final comment, and before someone asks (once the ascent graphic is attached), I note that liftoff has not been simulated here at T-0 (although that is the moment mentioned on official press-kits and Falcon9 Users Guide, in reality there seems to exist a delay in the engines ignition / verification sequence(?) and then on lift-off time, which is what I'm doing too, at least for 'now').

António

Edit: and here is the preliminary ascent graphic for something like a Falcon Block1 (hopefully with a complete upper stage nozzle @ ~336s vac. ISP?) delivering ~6.8t Dragon (Capsule+CapsuleProp+Trunk+Payload) into ~250 km / ISS inc. orbit (have not tweaked the files yet for extra injection precision: getting ~24X x 25X km and ~51.6X inc.)

[ugordan]

I played around with the 2nd burn parameters as well. From the video it looks like the burn was precisely 19 seconds. I assumed an MVac propellant flow rate of 145 kg/s, not sure if it's even close to the mark, but looks like that also agrees with your expended propellant mass. Given the deltaV was about 1570 m/s and assuming an Isp of the truncated nozzle of 325 s (again, a WAG number), it gives a burnout mass of 4300 kg for the stage + trunk.

The problem is uncertainties in the flow rate and Isp will change this a lot. A 10 kg/s difference in flow rate is enough to change the burnout mass by 300 kg and a 5s Isp difference will change it by about 100 kg.

Oh and that gives a burnout acceleration of over 9 Gs. Yikes.

[Ben the Space Brit]

FWIW, I thought that Mvac's Isp was in the range of 305-310s, but I admit it's been a while since I've bothered to check SpaceX's Falcon-9 page for updates on engine performance.

[kkattula]

FWIW, I thought that Mvac's Isp was in the range of 305-310s, but I admit it's been a while since I've bothered to check SpaceX's Falcon-9 page for updates on engine performance.

IIRC, it was originally 342, then listed as 338 before the second F9 flight. Presumably first flight performance was a little lower than expected.

Non-vacuum Merlin's vacuum Isp is 304.

That puts the trimmed Mvac somewhere between 305 and 337.  Given the exit plane area wasn't hugely reduced, I would guess somewhere north of 325.

[simcosmos]

With this talk about engine specs the readers can understand why I'm taking time to prepare a thread about the COTS Demo1 Virtual Reconstruction: on the initial post of such eventual thread I would like to explicitly reference / link to several types of input data, in an organized and clear way (data which is currently very scattered) and then, based on such references, present the numbers + ascent procedures / etc that used as input data + constraints for the simulation(s)...

… All the above in order to properly offer a better context for the obtained simulation(s) results... and I might also be talking about doing a similar thing regarding a more generic topic (than the demo1 flight), about both Falcon9 (Block1 vs Block2) and several Dragon configurations.

But again, this kind of data gathering - or better, the related text and comments - takes time to prepare (the data gathering and references are already done, what takes time is the 'playtime' with such data, the Math and simulation brainstorms in order to try to constraint a little more the comments / doubts that I might eventually present and so that others might also have a better data set to also comment / interpret / correct / clarify as needed)...


In what relates to Falcon9 engines, I have found three references for the first stage and two main references for the upper stage.

Regarding the first stage, one of the references is about the 'block2' performance goals (as far as I'm aware, SpaceX is not there yet). The other reference seems to be about 'block1'... And then, I also found another (partial) reference that seems to be somewhere between 'block1' and 'block2'...

Regarding the upper stage, I have found two main references: one, with an higher thrust and vac. ISP of 342s seems to be the expectation for 'block2'. The other, with a lower thrust and a vac. ISP of 336s seems to be close to the current state... The 'block1' references was what I have used to produce the ascent graphic that have shared on my previous post (about an eventual Falcon9 – 6.8t Dragon COTS demo  configuration (?))

Going back, for the purpose of SpaceX COTS Demo1 virtual performance and trajectory reconstruction, there is the upper stage nozzle 'fix' to take in account: my admittedly very clumsy  guess for the simulation that resulted in the numbers presented on my previous post was ~328s vac. ISP (which resulted in a little more than 20s or so burn duration, after Dragon + cubesats release and when taking in account ignition + shutdown thrust transients).



As mentioned in the start, I would like to present the data references that have here in a single thread because, although focusing in the COTS demo flights, the spreadsheet I'm using is also about both Falcon9 and Dragon (and other payloads) eventual configurations... But that might take time (given that, for example, the Dragon mass breakouts are related with launch vehicle 'configurations', mission type, etc...)


… Hummm, if there is interest in having a slightly sooner discussion about what I have here, please someone let me know and I might then perhaps try to split the Dragon part into a separated thread (sooner than the eventual thread dedicated to the Demo1 flight + Extrapolations for  other demo flights, Falcon9 launch vehicle simulation, etc). Else, I will take my time and present / reference all in a single thread, when possible.

António.

[Nomadd]

FWIW, I thought that Mvac's Isp was in the range of 305-310s, but I admit it's been a while since I've bothered to check SpaceX's Falcon-9 page for updates on engine performance.

IIRC, it was originally 342, then listed as 338 before the second F9 flight. Presumably first flight performance was a little lower than expected.

Non-vacuum Merlin's vacuum Isp is 304.

That puts the trimmed Mvac somewhere between 305 and 337.  Given the exit plane area wasn't hugely reduced, I would guess somewhere north of 325.

I know their combustion efficiency for the 1C was a couple percent below what they hoped for. The 342 might be from the earliest days.

[simcosmos]

Falcon9 Propulsion Assumptions on SpaceX COTS Demo1 vs Updated Merlin Specs?

Not sure if this is the correct thread to continue this topic but given that what I will write is still related with the SpaceX COTS Demo1 flight... Will attach a quick screenshot from a small part of an Excel file where I'm playing with Falcon9 + Dragon (Dragon spacecraft is in a separated sheet of the same book) in order to get early 'departure points' of overall vehicle / payload properties for a given set of injection case(s) and for eventual further simulation (performance + trajectory implementation) and related study / refinement (by comparing with 'real life' records / results vs some of the information on Vehicle's Users Guide, Press Kits, etc).

The important thing, for the moment, is the top of that spreadsheet, the propulsion numbers: I'm not sure if what I'm calling of 'block2' is still the current expectation or not. In the same way, I'm not sure if what I'm calling of 'block1' is or not close to the current development state of the engines (as used in the recent COTS demo1 flight).

I ask the reader to please do not mind about the four decimals (they are there just because of conversion factors and because are input data). Feel free to check the information (please see the provided sources) and comment / correct as needed.

I would like to note that this is just one of the pieces of the puzzle, when trying to make a virtual reconstruction of the COTS demo1 flight or when trying to use such mission information to extrapolate upcoming flights...

Another thing that I would be very interested in studying is the flight envelope vs recovery conditions of the first stage... I might be incorrect in what will write next but the very preliminary impression here is that Falcon9 first stage might be going a bit high... (a good bit more than AresI first stage expected apogee, for example, although the two stages recovery are not directly comparable because, among several other key differences - such as the longer burn time - SpaceX stage is also lighter, smaller, etc).

There are also other pieces of the puzzle which are equally or more challenging to research (but I guess that it is better to leave that for other eventual future specific  thread(s)).

Meanwhile, if someone has a fresh input or further comments, research directions, etc about Merlin current specs and updates on expected maximum goal (for block2), it would be much appreciated.

Thanks,
António

[Toner Soprano]

I forgot to download the interstage video from spacex.  Does anyone have a youtube link to that?  I can't find any on youtube.

[martin hegedus]

Does anyone know what the story is behind the different seal level and vacuum thrust specs for the first stage Merlin 1C engine?

From the COTS1 Press Kit http://www.spacex.com/downloads/cots1-20101206.pdf
Sea Level, 423 kN
Vacuum, 483 kN

From Falcon 1 page, http://www.spacex.com/falcon1.php
Sea Level, 512 kN
Vacuum, 569 kN

From Falcon 9 page, http://www.spacex.com/falcon9.php
Sea Level, 556 kN
Vacuum, 617 kN

Is some (or all?) of the data presented incorrect?  Are the cores (i.e. thrust chamber and bell) for the three Merlin 1Cs different?  Or, are the cores the same and the difference is because of different chamber pressures?  Will future COTS missions be upping the thrust or is the 423kN what the Falcon 9 was designed for?

Thanks.

[starsilk]

http://www.spacenews.com/civil/110107-nasa-boosted-cots-funding.html

looks like an additional $20 million already paid out under COTS to SpaceX for more risk reduction testing of Dragon (in $5 million increments): vibration testing, Dragon solar array deployment, LIDAR testing (DragonEye), and some thermal vacuum testing.

also apparently there are a few extra $5M tests NASA wants by March, full vacuum testing of Dragon, more LIDAR work, etc.

(worth noting they're paying out similar sums to Orbital, too, so it's all nice and fair)

[martin hegedus]

...

The important thing, for the moment, is the top of that spreadsheet, the propulsion numbers: I'm not sure if what I'm calling of 'block2' is still the current expectation or not. In the same way, I'm not sure if what I'm calling of 'block1' is or not close to the current development state of the engines (as used in the recent COTS demo1 flight).

...

Thanks,
António

When you use the terms block1, block2 etc. it seems like there are different engines.  And that may be the case.  However, it could also be that the engines are throttled back for redundancy sake.  Maybe the max thrust of a Merlin 1C when configured for Falcon 9 is 556kN at sea level but they throttle it back to 423kN.  It is interesting to note that, when using vac thrust, there is about a 2 engine redundancy, i.e. 483*9/617 = 7.045.  Using sea level thrust the math results in 6.847.

[Antares]

Turbodynamics are highly non-linear. It shouldn't be assumed a lower throttle setting has higher margin.

[go4mars]

Seems there is a lot of process engineering bickering going on about Spacex. Just reminder that Nasa HSF is not immune from process failure...

SpaceX has about 140 positions on their 'now hiring' pages.  Some of them relate to "configuration management" and Configuration analysis", a bunch of "quality engineer" and "quality engineering analyst" positions.  Many titles include words like "Integrator" or "Inspector".  There also seems to be a lot more "manager" positions than there used to. 

Causes? 

Is this for the switch to mass production phase from experiment phase?  Are they there yet?
Is it to address the "process engineering concerns" noted by Jim?  Will it properly address those concerns?
Other causes?
With the reportedly long work hours there, is employee attrition higher than elsewhere?

Effects?

Extra overhead and oversight for ensuring things are done consistently across the board clearly adds a financial burden ($7-12M extra per year at a guess based on their available positions) .  Is this enough to effectively render and stabilize processes? 
There are a lot of technicians shifts for around the clock work.  Could they possibly get 5 F9's and 1 or more F1e in the sky in 2011 if they fill their roster soon?  Or is there no hope of that?