Author Topic: SpaceX's Autonomous Spaceport Drone Ship Discussion Thread 3  (Read 712280 times)

Offline CyndyC

Those 4-tons might present a bang-bang type scenario where they're practically freezing the equipment.

I was told during my HVAC repair last January that SpaceX keeps their drone ship electronics at the minimum temp A/Cs can provide before requiring refrigeration equipment.
"Either lead, follow, or get out of the way." -- quote of debatable origin tweeted by Ted Turner and previously seen on his desk

Offline woods170

  • IRAS fan
  • Senior Member
  • *****
  • Posts: 7911
  • IRAS fan
  • The Netherlands
  • Liked: 3884
  • Likes Given: 1182
Re: SpaceX's Autonomous Spaceport Drone Ship Discussion Thread 3
« Reply #1321 on: 07/13/2016 03:41 PM »
Those 4-tons might present a bang-bang type scenario where they're practically freezing the equipment.

I was told during my HVAC repair last January that SpaceX keeps their drone ship electronics at the minimum temp A/Cs can provide before requiring refrigeration equipment.
Unsurprisingly. The lower the temperature, the lower the amount of water vapour in the air, and thus less salt. Electronics hate water. Electronics hate salt water even more.

Offline kaiser

  • Member
  • Posts: 51
  • Liked: 70
  • Likes Given: 20
Re: SpaceX's Autonomous Spaceport Drone Ship Discussion Thread 3
« Reply #1322 on: 07/13/2016 05:07 PM »
Those 4-tons might present a bang-bang type scenario where they're practically freezing the equipment.

I was told during my HVAC repair last January that SpaceX keeps their drone ship electronics at the minimum temp A/Cs can provide before requiring refrigeration equipment.
Unsurprisingly. The lower the temperature, the lower the amount of water vapour in the air, and thus less salt. Electronics hate water. Electronics hate salt water even more.

Eh, it's more for equipment reliability.  There's a pretty linear relationship between equipment temperature and longevity.  Lots of data centers are making the trade to go warmer, because AC is expensive and equipment is cheap and easy to replace.  But mobile systems, replacing the equipment is expensive if not impossible, so you typically go nice and low on the temperature.  On the flip side, if you keep things that cold then you have to have a warmup period before you go in.  Otherwise, you open that door and immediately get condensation on all types of equipment.  So, we keep things cool, but not too cool, because otherwise you can't open them up :)  But for autonomous operations, definitely go as cold as you can essentially.

With those 4-ton units though, I'm surprised that they're not routinely icing them up (maybe they are and they let the other one de-ice and run the secondary, or they've installed secondary de-icers on the coils).  While at sea, a 3-ton AC can keep telemetry and some satcom gear chilled so low that you're down to min temperature 2 minutes into the cycle, but you have a minimum 5-minute or so run time on the AC, but there's not enough heat load so the external coils ice up even while on the equator.  So you either adjust the run times so that it's not an issue, allow large temperature swings, or add on a secondary kit to help prevent icing.

With respect to humidity, a 1-ton AC unit takes a 20' ISO container that's not even fully sealed very well down to single digit humidity percentages really, really quickly, even while at sea and getting splashed by waves.  Humidity is really not a concern.  Most containers, if they've had a couple of cycles of chilling and are decently chilled, you've bottomed out the humidity sensor's range and it stays there.

Offline launchwatcher

  • Full Member
  • ***
  • Posts: 339
  • Liked: 231
  • Likes Given: 285
Re: SpaceX's Autonomous Spaceport Drone Ship Discussion Thread 3
« Reply #1323 on: 07/13/2016 06:37 PM »
Eh, it's more for equipment reliability.  There's a pretty linear relationship between equipment temperature and longevity.  Lots of data centers are making the trade to go warmer, because AC is expensive and equipment is cheap and easy to replace.  But mobile systems, replacing the equipment is expensive if not impossible, so you typically go nice and low on the temperature. 
it's.. more complicated than that.   Large-scale data center gear has limited lifetime because newer more power efficient gear comes along quickly enough that it doesn't really make economic sense to keep the old stuff running.

And it's not always better to go colder.    In 2007, Google published research showing that in google's data centers (which are large warehouses), hard drives were measurably more reliable at 35-40C than they were at 15-30C.   
See "Failure Trends in a Large Disk Drive Population" (pdf), particularly section 3.4 and figures 4 and 5.

Stability of temperature and humidity - inherently easier in large buildings compared with shipping-container-scale installations - may well be more important than the absolute temperature.


Offline kaiser

  • Member
  • Posts: 51
  • Liked: 70
  • Likes Given: 20
Re: SpaceX's Autonomous Spaceport Drone Ship Discussion Thread 3
« Reply #1324 on: 07/13/2016 10:50 PM »
Eh, it's more for equipment reliability.  There's a pretty linear relationship between equipment temperature and longevity.  Lots of data centers are making the trade to go warmer, because AC is expensive and equipment is cheap and easy to replace.  But mobile systems, replacing the equipment is expensive if not impossible, so you typically go nice and low on the temperature. 
it's.. more complicated than that.   Large-scale data center gear has limited lifetime because newer more power efficient gear comes along quickly enough that it doesn't really make economic sense to keep the old stuff running.

And it's not always better to go colder.    In 2007, Google published research showing that in google's data centers (which are large warehouses), hard drives were measurably more reliable at 35-40C than they were at 15-30C.   
See "Failure Trends in a Large Disk Drive Population" (pdf), particularly section 3.4 and figures 4 and 5.

Stability of temperature and humidity - inherently easier in large buildings compared with shipping-container-scale installations - may well be more important than the absolute temperature.

Definitely agree with you about the limited lifetime of datacenter gear.

With reliability and temperature, as always as you said it's complicated, so there are contradictory reports! :-D  Backblaze, Microsoft, UVA, US DOD and a couple of others have lots of data to show otherwise on hard drives.  Google may be right though for the way that their data centers are designed , their operational profiles and the exact hardware they use.  There's more to it than just hard drives though, caps on motherboards and other components do generally like to be chilly.

Back on topic:

That said, temperature and humidity stability is pretty darn important, which is why I mentioned that a 4-ton with their likely limited heat load might bang-bang.  Oversizing the AC can really cause some big swings, versus getting it sized just right so that it idles and keeps temperatures in a narrow range.  Sometimes we mix/match sizes, like a 2-ton and a 3-ton for this reason.  The 2-ton is appropriately sized, and can keep a stable temperature whereas the 3-ton would tend to overcool and shut down, but you kick the 3-ton on when you're really pushing the power or on a really hot day or after you've expanded your operations now the 3-ton is appropriately sized, and the 2-ton can kick on to help out as needed without causing too large of a temperature gradient.

Offline CameronD

  • Full Member
  • ****
  • Posts: 1328
  • Melbourne, Australia
    • Norton Consultants
  • Liked: 425
  • Likes Given: 320
Re: SpaceX's Autonomous Spaceport Drone Ship Discussion Thread 3
« Reply #1325 on: 07/14/2016 12:03 AM »
Those 4-tons might present a bang-bang type scenario where they're practically freezing the equipment.

I was told during my HVAC repair last January that SpaceX keeps their drone ship electronics at the minimum temp A/Cs can provide before requiring refrigeration equipment.

After a long day sun-baking on the dark-painted deck of his mobile island I would think Elon would want plenty of refrigeration to keep the drinks cold for his guests..
 
With sufficient thrust, pigs fly just fine - however, this is not necessarily a good idea. It is hard to be sure where they are
going to land, and it could be dangerous sitting under them as they fly overhead.

Offline CyndyC

At any rate, the discussion gives me some good questions to ask next time, which will actually be sometime soon. They had to order a part for me, so if I get someone else who helped with drone ship repairs, maybe he'll know how the units are coordinated.

Can anyone tell if whatever SpaceX has been doing, it has seemed to be working? I just realized I don't even know why they would need much more than GPS telemetry for the ship itself, assuming the rocket sends most of the landing data.
"Either lead, follow, or get out of the way." -- quote of debatable origin tweeted by Ted Turner and previously seen on his desk

Offline cscott

  • Senior Member
  • *****
  • Posts: 2848
  • Liked: 1970
  • Likes Given: 664
Re: SpaceX's Autonomous Spaceport Drone Ship Discussion Thread 3
« Reply #1327 on: 07/14/2016 01:09 AM »
My wild guess would be: GPS and power electronics to control the hydraulics for the station-keeping (if that's in the container it would account for a decent heat load over "mere" telemetry), plus telemetry receivers and a full redundant high-reliability independent storage system to record the raw signals received in case of incident.  Plus a small transmitter for post-landing operations, but I think we've seen this already in dock operations and it's not large.  There's some additional control electronics for the fire-fighting gear, but it doesn't seem terribly effective.

Anyway, that's my wild guess at the contents of the container; hopefully better-informed folks can use this as a starting point to tell me exactly how I'm wrong.
« Last Edit: 07/14/2016 01:09 AM by cscott »

Offline CameronD

  • Full Member
  • ****
  • Posts: 1328
  • Melbourne, Australia
    • Norton Consultants
  • Liked: 425
  • Likes Given: 320
Re: SpaceX's Autonomous Spaceport Drone Ship Discussion Thread 3
« Reply #1328 on: 07/14/2016 01:25 AM »
My wild guess would be: GPS and power electronics to control the hydraulics for the station-keeping (if that's in the container it would account for a decent heat load over "mere" telemetry), plus telemetry receivers and a full redundant high-reliability independent storage system to record the raw signals received in case of incident.  Plus a small transmitter for post-landing operations, but I think we've seen this already in dock operations and it's not large.  There's some additional control electronics for the fire-fighting gear, but it doesn't seem terribly effective.

FWIW, the station-keeping GPS, control and hydraulic systems (assuming they are as outlined earlier in this thread) are all marine-rated and, for that reason, would not require any kind of air-conditioning nor would generally be supplied with any by design.

I would guess the A/C systems reported to be aboard would be solely for the satellite, telemetry and radio communications gear, some of which could be very temperature-sensitive indeed... and a shipping container, even a white-painted one, can get pretty warm in the hot sun.
With sufficient thrust, pigs fly just fine - however, this is not necessarily a good idea. It is hard to be sure where they are
going to land, and it could be dangerous sitting under them as they fly overhead.

Offline CyndyC

Maybe no one has to assume the containers are very full to begin with. The size and weight of the containers by themselves might help stabilize the ship.

Edited to pluralize since I was told there are 2 containers w/ 4 cooling units each, one set at the bow and one at the stern. 
« Last Edit: 07/14/2016 03:15 AM by CyndyC »
"Either lead, follow, or get out of the way." -- quote of debatable origin tweeted by Ted Turner and previously seen on his desk

Offline the_other_Doug

  • Senior Member
  • *****
  • Posts: 2709
  • Minneapolis, MN
  • Liked: 1764
  • Likes Given: 3327
Re: SpaceX's Autonomous Spaceport Drone Ship Discussion Thread 3
« Reply #1330 on: 07/14/2016 03:36 AM »
Well... ISTR there is always a call-out on ASDS recovery attempts that the ASDS acquires the telemetry from the first stage as the stage comes over its horizon.  Wouldn't that indicate that the ASDS is both relaying the telemetry back to Hawthorne and recording it for later analysis?

Seeing as how the regular tracking and telemetry systems are all busy tracking, and keeping tabs on the telemetry from, the second stage, it would make sense that the ASDS might be serving a lot of those functions for the first stage during its final descent and its landing burn.  For RTLS landings, there are communications systems in place at the launch site to handle those functions.

That would mean there's a lot of electronics for processing and retransmitting telemetry that needs cooling, I would think...
-Doug  (With my shield, not yet upon it)

Offline cscott

  • Senior Member
  • *****
  • Posts: 2848
  • Liked: 1970
  • Likes Given: 664
Re: SpaceX's Autonomous Spaceport Drone Ship Discussion Thread 3
« Reply #1331 on: 07/14/2016 05:06 PM »
I've always suspected that go quest is doing most of the telemetry. It had its own electronics container, and the FCC license mentions some of its gear.  To play devil's advocate: the ASDS container could contain nothing but video camera relay/recording and post-launch command-and-control, with go quest handling all telemetry duties and the station-keeping hardware all outside in the marine environment.

Offline CameronD

  • Full Member
  • ****
  • Posts: 1328
  • Melbourne, Australia
    • Norton Consultants
  • Liked: 425
  • Likes Given: 320
Re: SpaceX's Autonomous Spaceport Drone Ship Discussion Thread 3
« Reply #1332 on: 07/15/2016 12:32 AM »
I've always suspected that go quest is doing most of the telemetry. It had its own electronics container, and the FCC license mentions some of its gear.  To play devil's advocate: the ASDS container could contain nothing but video camera relay/recording and post-launch command-and-control, with go quest handling all telemetry duties and the station-keeping hardware all outside in the marine environment.

Well.. we know there is at least some telemetry on the ASDS: the flight controllers back in MCC-X need to not only know the ASDS is 'on-station' with 'all systems green', but need to shift it remotely via changing the GPS target set-point (as easy as sending a NMEA command string over a serial comms link from their on-board telemetry server no doubt) in the event of a last-hour decision to water-land.

Plus the video camera transmissions.. and the remote activation of the deluge system.. there's a bit there.

« Last Edit: 07/15/2016 12:34 AM by CameronD »
With sufficient thrust, pigs fly just fine - however, this is not necessarily a good idea. It is hard to be sure where they are
going to land, and it could be dangerous sitting under them as they fly overhead.

Offline MarekCyzio

OCISLY earlier today.

Offline Jim

  • Night Gator
  • Senior Member
  • *****
  • Posts: 32003
  • Cape Canaveral Spaceport
  • Liked: 10629
  • Likes Given: 317
Re: SpaceX's Autonomous Spaceport Drone Ship Discussion Thread 3
« Reply #1334 on: 07/16/2016 11:04 PM »

Well.. we know there is at least some telemetry on the ASDS: the flight controllers back in MCC-X need to not only know the ASDS is 'on-station' with 'all systems green', but need to shift it remotely via changing the GPS target set-point (as easy as sending a NMEA command string over a serial comms link from their on-board telemetry server no doubt) in the event of a last-hour decision to water-land.

Plus the video camera transmissions.. and the remote activation of the deluge system.. there's a bit there.

Go Quest can do those vs MCC

Offline CyndyC


Well.. we know there is at least some telemetry on the ASDS: the flight controllers back in MCC-X need to not only know the ASDS is 'on-station' with 'all systems green', but need to shift it remotely via changing the GPS target set-point (as easy as sending a NMEA command string over a serial comms link from their on-board telemetry server no doubt) in the event of a last-hour decision to water-land.

Plus the video camera transmissions.. and the remote activation of the deluge system.. there's a bit there.


Go Quest can do those vs MCC

So in other words, using Facebook's Pirate English translation, the Go Quest crew is a mix of landlubbing black-hearted scallywags and seafaring yellow-bellied freebooters. You have to hope they're doing more than eyeballing the landing trajectory to adjust the ASDS position if necessary.
"Either lead, follow, or get out of the way." -- quote of debatable origin tweeted by Ted Turner and previously seen on his desk

Offline Bob Shaw

  • Full Member
  • ****
  • Posts: 1031
  • Liked: 410
  • Likes Given: 365
Re: SpaceX's Autonomous Spaceport Drone Ship Discussion Thread 3
« Reply #1336 on: 07/17/2016 12:25 AM »

Well.. we know there is at least some telemetry on the ASDS: the flight controllers back in MCC-X need to not only know the ASDS is 'on-station' with 'all systems green', but need to shift it remotely via changing the GPS target set-point (as easy as sending a NMEA command string over a serial comms link from their on-board telemetry server no doubt) in the event of a last-hour decision to water-land.

Plus the video camera transmissions.. and the remote activation of the deluge system.. there's a bit there.

Go Quest can do those vs MCC

Low bit-rate comms will suffice for most C&C of the drone ship, but there's no reason to assume that there isn't also higher capacity microwave data being sent. It's not that big an ask - I've participated in artillery exercises where fall of shot video was transmitted by a forward observation vessel to the firing vessel, with a low-bitrate link from the mothership controlling camera positioning. This sort of stuff isn't at all new, and requires almost trivial levels of kit.

Offline cscott

  • Senior Member
  • *****
  • Posts: 2848
  • Liked: 1970
  • Likes Given: 664
Re: SpaceX's Autonomous Spaceport Drone Ship Discussion Thread 3
« Reply #1337 on: 07/17/2016 04:07 AM »
Folks, we're missing the point: the question was what gear is likely to be in the air-conditioned SpaceX containers on the ASDS deck.  Two strawmen were raised, a "maximal" one which consisted  of all the equipment which could possibly be present, and a "minimal" one with as little equipment as possible (nothing that could plausibly be aboard Go Quest instead, for example).  We have no way of knowing which is nearer to correct, but it's interesting to explore both directions.
« Last Edit: 07/17/2016 08:33 PM by cscott »

Offline miki

  • Member
  • Posts: 14
  • Serbia
  • Liked: 24
  • Likes Given: 8
Re: SpaceX's Autonomous Spaceport Drone Ship Discussion Thread 3
« Reply #1338 on: 07/17/2016 05:14 PM »
Not sure if someone has posted this, but here is a video in real speed of CRS-8 landing on ASDS:

https://www.youtube.com/watch?v=VOagay_opLQ?t=56m20s (at 56m and 20s in video).
« Last Edit: 07/17/2016 06:22 PM by miki »

Offline Req

  • Full Member
  • ****
  • Posts: 400
  • Liked: 413
  • Likes Given: 2581
Re: SpaceX's Autonomous Spaceport Drone Ship Discussion Thread 3
« Reply #1339 on: 07/17/2016 06:43 PM »
Not sure if someone has posted this, but here is a video in real speed of CRS-8 landing on ASDS:

https://www.youtube.com/watch?v=VOagay_opLQ?t=56m20s (at 56m and 20s in video).

The forum keeps replacing &t with ?t when it caches that url or something breaking the link, this should work:

http://www.tubechop.com/watch/8187976

Tags: