4

A lot of commercially-produced electronic equipment is designed for a 19-inch rack (or less commonly, 23-inch) with a particular spacing for mounting holes. Much of the pressurized volume of the International Space Station is used for equipment racks such as the International Standard Payload Rack.

Would standard rackmount electronics fit on any of the ISS racks? Consider not only the equipment width, but also the depth and mounting holes.

DrSheldon
  • 47,946
  • 12
  • 162
  • 329
  • 1
    Just about any electronics can mount on a 19 inch rack unless it is too wide, if it is narrow you just put ears on it or a custom shelf. – GdD Oct 21 '20 at 08:22
  • 2
    It seems the Wikipedia page you linked and the sources listed there contain all necessary information. Width is 2x 19", holes are "standard pitched" and depth is whatever it is - there is no default depth for 19" modules. – asdfex Oct 21 '20 at 08:53
  • 1
    If it fits, it fits. If it doesn't fit, adapters exist or could be manufactured. Issues such as heat, circulation, security, hardening against radiation, power draw, and outgassing would be the larger concerns trying to put commercial rack servers on the ISS. – Schwern Oct 21 '20 at 20:52

1 Answers1

4

A standard EXPRESS Rack single locker payload is 18.125" external width, with an internal width of 17.340". So COTS rack mount hardware would not fit facing the front. I suppose you could mount it sideways, since the available depth is 20.195".

enter image description here Image Source (PDF)

If you really needed more width, there have been some double locker payloads that are a single unit high and two units wide. That's what HP did with their Spaceborne Computer payload, which

contains two servers and no modifications to its hardware. A water-cooled enclosure will further help it deflect radiation while the software will have to meet Nasa’s strict standards, in place to protect spacecrafts and astronauts. (source)

enter image description here

Speaking of Spaceborne Computer, you'll need some form of radiation hardening, either software (like SC is demonstrating) or hardware.

However, the size isn't the biggest issue with flying COTS hardware. All powered payloads must have some form of smoke detection. Payloads in an EXPRESS rack must either have air circulating through a closed volume that exhausts through the rack smoke detector, or if they're water cooled they need their own internal ECW (Emergency, Caution, Warning) enunciation that notifies the rack that the hardware is overheating.

The EXPRESS Rack supplies 28V DC power and an ethernet connection through the front of the payload. If your rack mount hardware isn't compatible with 28V, you'll need a power supply to convert the voltage too.

Doresoom
  • 1,724
  • 15
  • 21
  • This post is hidden. You deleted this post just now.

    I can't follow your reasoning. If 19" rack equipment has an external width of less than 17.75" (that's the internal width of the rack), why shouldn't it fit into a place that can accommodate a 18.125" payload? Obviously it won't match many other design criteria, but width doesn't seem to be one.

     – asdfex Jan 29 '24 at 13:11
  • I'm not sure what you mean about deleting the post, which I didn't do. The 17.75" would need to fit IN a locker, or have its own custom enclosure made for smoke detection purposes. – Doresoom Jan 29 '24 at 13:30
  • 1
    Writing a comment is hard... first I mistyped as an answer, then after deleting it I copied the "deleted" message as well.. – asdfex Jan 29 '24 at 13:30
  • 1
    The need for radiation hardening is often exaggerated. NICER on the ISS collects its x-ray events with ordinary automotive microcontrollers. There have been a handful of upsets in over six years, with negligible impact on productivity. – John Doty Jan 30 '24 at 21:08
  • 1
    Hi John! I was on console as an OC for the initial NICER deployment. Glad to hear your hardware is working well. When I worked at MSFC POIC, we had other payloads that would regularly have single event upsets when passing through the South Atlantic Anomaly, and required the PD to check into console to recover. One of them sent us a map of the ISS ground track corresponding to all their SEUs, and it was a huge cluster around the SAA with just a few outliers in other locations. So I guess the answer is "radiation is a problem for some computer systems, but not all." – Doresoom Jan 30 '24 at 21:22