SpaceX F9 : Starlink v0.9 : May 23, 2019 - DISCUSSION

[joseph.a.navin]

Upper level winds look much better for tonight, with a max of 80kts and only 10% prob of violation! 

[smoliarm]

...
Something else that I haven't seen noted yet: 227 kg is what you get when you convert exactly 500 pounds to kg and round to 3 significant figures. I think that the number started as "about 500 pounds" and magically acquired two more significant digits in the conversion to metric.

...

Exactly,
I had the same thought: actual satellite weight - in pounds - was rounded to hundreds ("500") and converted to kg.
If true, we have to threat it with 10% rounding error:
500 ± 50 lb
or
227 ± 23 kg

One more thing to keep in mind:
LSP has following note on heavy payloads for Falcon 9:

Payload mass greater than 7,250 kg (15,983 lbm) may require mission-unique adapter/accommodations, resulting in cost and/or performance impacts

As I recall, SpaceX has something similar in their Falcon's "user manual".

It means, that for such a heavy payload  they have to use stronger PAF and (may be) stronger second stage.
I expect this to be the greater factor deviating their actual performance from the numbers given by LSP-calculator.
In other words, for this case we should treat LSP-calculator results as an UPPER limit.

[envy887]

One more thing to keep in mind:
LSP has following note on heavy payloads for Falcon 9:

Payload mass greater than 7,250 kg (15,983 lbm) may require mission-unique adapter/accommodations, resulting in cost and/or performance impacts

As I recall, SpaceX has something similar in their Falcon's "user manual".

It means, that for such a heavy payload  they have to use stronger PAF and (may be) stronger second stage.
I expect this to be the greater factor deviating their actual performance from the numbers given by LSP-calculator.

The PAF shown in the tweeted photo of the payload is not the standard Falcon PAF, so the limitation mentioned in the PUG is irrelevant.

In other words, for this case we should treat LSP-calculator results as an UPPER limit.

Why? SpaceX could have a custom Starlink PAF that is both lighter and stronger and allows a higher payload mass.

[Wolfram66]

http://tropic.ssec.wisc.edu/real-time/atlantic/movies/wg8shr/wg8shrjava.html

Wind Shear potential still looks "Iffy" for tonight... fingers crossed and ohhmmmmmmmm... for calm...

[joseph.a.navin]

http://tropic.ssec.wisc.edu/real-time/atlantic/movies/wg8shr/wg8shrjava.html

Wind Shear potential still looks "Iffy" for tonight... fingers crossed and ohhmmmmmmmm... for calm...

And hope @upperlevelwinds doesn't strike back a second time   

[eeergo]

ARTICLE: First Starlink mission to be heaviest payload launch by SpaceX to date

https://www.nasaspaceflight.com/2019/05/first-starlink-mission-heaviest-payload-launch-spacex/

- By Chris Gebhardt

"More impressively, the Startracker system will allow each Starlink to independently track on-orbit debris and autonomously fire its Hall thrusters to avoid debris."

Note this has proven to be wrong: the sats are able to perform COLAs based on NORAD data, but not active tracking or avoidance.

[eeergo]

Ok, here is my dime on question "Why Kr?"

As far as I know, Xe is a usual choice as propellant in Hall thrusters because it gives more *thrust per kWt*.
On the other hand, Kr gives better Isp.
Plus, Kr is about 3 times cheaper than Xe (approximately - prices here vary with purity grade).

So, AIUI, for heavy GEO-satellites the higher thrust is the most important (especially if they want to use ion engine for GTO-GEO transfer). Also, for a quarter-billion satellite, the propellant price does not matter.

Continuing the research about Kr, came across this interesting concise study by ESA/Sitael about such an ion engine from 2017: https://artes.esa.int/projects/kht-krypton-hall-thrusters

As other studies pointed out, Kr crucially causes more erosion than Xe on the thruster's ceramic material, but other effects can be offset by its advantages (thrust vs Isp...). It concludes:

"Project activities have been completed. The overall outcome is that propellant different from xenon can provide significant economic benefits in the long term for commercial telecom applications. In particular, krypton would allow for a major reduction of qualification and operation costs with minor performance drawbacks."

[oldAtlas_Eguy]

Ok, here is my dime on question "Why Kr?"

As far as I know, Xe is a usual choice as propellant in Hall thrusters because it gives more *thrust per kWt*.
On the other hand, Kr gives better Isp.
Plus, Kr is about 3 times cheaper than Xe (approximately - prices here vary with purity grade).

So, AIUI, for heavy GEO-satellites the higher thrust is the most important (especially if they want to use ion engine for GTO-GEO transfer). Also, for a quarter-billion satellite, the propellant price does not matter.

Continuing the research about Kr, came across this interesting concise study by ESA/Sitael about such an ion engine from 2017: https://artes.esa.int/projects/kht-krypton-hall-thrusters

As other studies pointed out, Kr crucially causes more erosion than Xe on the thruster's ceramic material, but other effects can be offset by its advantages (thrust vs Isp...). It concludes:

"Project activities have been completed. The overall outcome is that propellant different from xenon can provide significant economic benefits in the long term for commercial telecom applications. In particular, krypton would allow for a major reduction of qualification and operation costs with minor performance drawbacks."

So not only is propellant 1/4 cheaper but the engines are cheaper to manufacture.

It is possible that the combined savings per Sat could be as high as $100K for using Krypton thrusters vs Xenon. Or as high as $6M on this first set of 60.

$100k here and a $100K there and you end with a Sat costing less than $500K.

[envy887]

ARTICLE: First Starlink mission to be heaviest payload launch by SpaceX to date

https://www.nasaspaceflight.com/2019/05/first-starlink-mission-heaviest-payload-launch-spacex/

- By Chris Gebhardt

"More impressively, the Startracker system will allow each Starlink to independently track on-orbit debris and autonomously fire its Hall thrusters to avoid debris."

Note this has proven to be wrong: the sats are able to perform COLAs based on NORAD data, but not active tracking or avoidance.

They do autonomous active avoidance:

https://twitter.com/ChrisG_NSF/status/1128787802093694985

Satellites receive NORAD debris tracking data to maneuver autonomously around what NORAD is tracking.

[eeergo]

ARTICLE: First Starlink mission to be heaviest payload launch by SpaceX to date

https://www.nasaspaceflight.com/2019/05/first-starlink-mission-heaviest-payload-launch-spacex/

- By Chris Gebhardt

"More impressively, the Startracker system will allow each Starlink to independently track on-orbit debris and autonomously fire its Hall thrusters to avoid debris."

Note this has proven to be wrong: the sats are able to perform COLAs based on NORAD data, but not active tracking or avoidance.

They do autonomous active avoidance:

Satellites receive NORAD debris tracking data to maneuver autonomously around what NORAD is tracking.

Sure, but "autonomously avoid" is not the wrong bit - the part that is incorrect, as proven by Chris G. himself in a later tweet from the recent telecon with Musk, is that the star tracker (sorry, not capitalizing it) allows on-orbit debris tracking. With "active tracking and avoidance" I meant the satellite seeing debris and moving away from them, which is quite different from the actual capability.

[Robotbeat]

...
Something else that I haven't seen noted yet: 227 kg is what you get when you convert exactly 500 pounds to kg and round to 3 significant figures. I think that the number started as "about 500 pounds" and magically acquired two more significant digits in the conversion to metric.

...

Exactly,
I had the same thought: actual satellite weight - in pounds - was rounded to hundreds ("500") and converted to kg.
If true, we have to threat it with 10% rounding error:
500 ± 50 lb
or
227 ± 23 kg

One more thing to keep in mind:
LSP has following note on heavy payloads for Falcon 9:

Payload mass greater than 7,250 kg (15,983 lbm) may require mission-unique adapter/accommodations, resulting in cost and/or performance impacts

As I recall, SpaceX has something similar in their Falcon's "user manual".

It means, that for such a heavy payload  they have to use stronger PAF and (may be) stronger second stage.
I expect this to be the greater factor deviating their actual performance from the numbers given by LSP-calculator.
In other words, for this case we should treat LSP-calculator results as an UPPER limit.

Not at all! The LSP is for an earlier version of Falcon 9. It is a minimum number which NASA is comfortable with. It does not mean SpaceX cannot exceed it. This is true of other launch vehicles on the LSP as well.

[Robotbeat]

ARTICLE: First Starlink mission to be heaviest payload launch by SpaceX to date

https://www.nasaspaceflight.com/2019/05/first-starlink-mission-heaviest-payload-launch-spacex/

- By Chris Gebhardt

"More impressively, the Startracker system will allow each Starlink to independently track on-orbit debris and autonomously fire its Hall thrusters to avoid debris."

Note this has proven to be wrong: the sats are able to perform COLAs based on NORAD data, but not active tracking or avoidance.

The avoidance is done autonomously using just data from NORAD. That means in principle they don’t need a full time team devoted to collision avoidance, which could become onerous for such a ridiculously big constellation.

[Robotbeat]

...
Elon Musk ✔ @elonmusk
...

If all goes well, each launch of 60 satellites will generate more power than Space Station

...

ISS's 8 solar wings produced about 248kilowatts beginning-of-life (less than that now, and less also if averaged over the orbit). That implies these satellites do about 4kW nameplate apiece. Not bad for a 227kg satellite built for a song. The whole constellation (~12000 birds) will be 50 Megawatts or more, then.

- here is a chance for another big discrepancy. Only a few people know this total (248 kW for all 8 wings, or 31 kW max power for a single wing).
Most will look ISS electric power in WIKI which says "Altogether, the arrays can generate 84 to 120 kilowatts".
And there is no "discrepancy" here, there is another word - redundancy

BUT - we do not know, what Elon was referring to in his tweet -
a. Max output of all 8 arrays combined (248 kW)
or
b. Max power generation by the ISS electric system (120 kW)

Therefore, just to be on a safe side, I would say Starlink nameplate has something like 2 kW.

Which is still - very impressive - for a sat with approx. dimensions of 3.5 m by 1.5 m

Agreed.

[AUricle]

http://tropic.ssec.wisc.edu/real-time/atlantic/movies/wg8shr/wg8shrjava.html

Wind Shear potential still looks "Iffy" for tonight... fingers crossed and ohhmmmmmmmm... for calm...

And hope @upperlevelwinds doesn't strike back a second time   

I have a question. What about the ULW and "shear" is it that is a threat to a rocket. I've seen people say it's the difference in wind speeds over a few thousand meters of atmosphere. But F9 only occupies 70 meters of airspace at any one moment. So wind velocity can't possibly be that different from the top of Falcon to the bottom.
Or is it the wind speed that poses the threat. And just how much can they tolerate??

[LouScheffer]

Elon is surely using short tons in his statement.  First, that's the convention - if he meant metric tonnes, he'd use that.  This can be ambiguous in speech, but this was written.  Just ask google  "18.5 tons in tonnes" to get 16.8 tonnes.

Second, we can work backwards from GTO performance, as in this thread.   We can calculate the mass of the payload in orbit before the GTO injection burn.  This works out to a payload of 17.1 tonnes (to a 28^o orbit).  16.8 tonnes to a 53^o orbit is a little better, but plausible.   18.5 tonnes is not, or they'd be doing a heck of a lot better on their GTO launches.

So Elon is quoting in short tons, presumably to be able to quote the higher number.

[intelati]

http://tropic.ssec.wisc.edu/real-time/atlantic/movies/wg8shr/wg8shrjava.html

Wind Shear potential still looks "Iffy" for tonight... fingers crossed and ohhmmmmmmmm... for calm...

And hope @upperlevelwinds doesn't strike back a second time   

I have a question. What about the ULW and "shear" is it that is a threat to a rocket. I've seen people say it's the difference in wind speeds over a few thousand meters of atmosphere. But F9 only occupies 70 meters of airspace at any one moment. So wind velocity can't possibly be that different from the top of Falcon to the bottom.
Or is it the wind speed that poses the threat. And just how much can they tolerate??

Straight line winds force the rocket one direction. If the winds are strong enough one would expect them to possibly counteract the maneuvering capabilities of the rocket. Considering the rocket is able to vector the engines, this isn't very common.

Shear is more important. When you get hit by opposing winds, that force is transfered suddenly to the structure causing high stresses.

[rsdavis9]

http://tropic.ssec.wisc.edu/real-time/atlantic/movies/wg8shr/wg8shrjava.html

Wind Shear potential still looks "Iffy" for tonight... fingers crossed and ohhmmmmmmmm... for calm...

And hope @upperlevelwinds doesn't strike back a second time   

I have a question. What about the ULW and "shear" is it that is a threat to a rocket. I've seen people say it's the difference in wind speeds over a few thousand meters of atmosphere. But F9 only occupies 70 meters of airspace at any one moment. So wind velocity can't possibly be that different from the top of Falcon to the bottom.
Or is it the wind speed that poses the threat. And just how much can they tolerate??

I don't know how accurate an analogy it is but:
Airbumps when flying in an airplane happen because of different vertical speed air masses that are being flown through.
Is it because the plane is straddling 2 different air masses or more probably going up vertically when it hits that down draft.
So in other words change of velocity perpendicular to craft and inertia of craft.
I have seen wings flex pretty good from airbumps!

[envy887]

http://tropic.ssec.wisc.edu/real-time/atlantic/movies/wg8shr/wg8shrjava.html

Wind Shear potential still looks "Iffy" for tonight... fingers crossed and ohhmmmmmmmm... for calm...

And hope @upperlevelwinds doesn't strike back a second time   

I have a question. What about the ULW and "shear" is it that is a threat to a rocket. I've seen people say it's the difference in wind speeds over a few thousand meters of atmosphere. But F9 only occupies 70 meters of airspace at any one moment. So wind velocity can't possibly be that different from the top of Falcon to the bottom.
Or is it the wind speed that poses the threat. And just how much can they tolerate??

Dynamic load is a function of angle of attack and of vehicle speed. Angle of attack is a function of both direction of vehicle travel and the motion of the surrounding air. So a rapid change in the relative motion of the surrounding air (aka wind shear) combined with a high vehicle speed, induces a high dynamic aero load because the vehicle can't change it's direction of travel or angle of attack nearly as quickly as it can transit a shear layer while going supersonic. At Mach 1.5 the Falcon 9 travels it's own length in about 100 milliseconds.

[Robotbeat]

Elon is surely using short tons in his statement.  First, that's the convention - if he meant metric tonnes, he'd use that.  This can be ambiguous in speech, but this was written.  Just ask google  "18.5 tons in tonnes" to get 16.8 tonnes....

...That's not accurate. "Tons" is correctly and regularly used to mean metric units as well. Musk prefers metric.

So I acknowledge the units are more ambiguous here than we'd like, but I do believe he intended metric tons.

[Kabloona]

Elon is surely using short tons in his statement.  First, that's the convention - if he meant metric tonnes, he'd use that.  This can be ambiguous in speech, but this was written.  Just ask google  "18.5 tons in tonnes" to get 16.8 tonnes....

...That's not accurate. "Tons" is correctly and regularly used to mean metric units as well. Musk prefers metric.

So I acknowledge the units are more ambiguous here than we'd like, but I do believe he intended metric tons.

OTOH, if you Google "Elon Musk twitter metric tons" you'll find at least 6 different instances where Elon has specifically cited "metric tons" in his tweets.

Then there are at least three other tweets citing only "tons" without the "metric."

 So it does look to me like he's careful to say "metric" when he means metric, to avoid confusing the general public.