One experienced engineer`s approach to better/cheaper/faster satellite testing (philosophies and lessons learned) Page: 3 of 11
This article is part of the collection entitled: Office of Scientific & Technical Information Technical Reports and was provided to Digital Library by the UNT Libraries Government Documents Department.
The following text was automatically extracted from the image on this page using optical character recognition software:
somehow commanded to a nonsensical/non-recoverable
I have observed operational failures due to EMI (e.g. a
payload which responded to the magnetic field
generated by a nearby earth sensor, a command uplink
receiver which locked onto a harmonic of a nearby
subsystem and prevented further commanding, and high
level noise which was generated in only one mode of a
subsystem due to higher currents/faster rise times).
I have observed working payloads partially or
completely ruined by environmental overtesting and
mishandling: (e.g. payloads vibrated so long they
physically fell apart, blown fuses caused by probing,
connectors mismated, contamination spread on satellites
by an air conditioning system, dropped satellites,
"purple plague" basically caused by inadequate
cleaning, condensation due to a vacuum chamber
backfill being below the dew point, latent failures
caused by ElectroStatic Discharge (ESD), mop water
slopped onto flight equipment, a trailblazer satellite
model which was dropped coming out of a thermal
vacuum facility, a payload in a shipping container
which was dropped off a loading ramp leading into the
cargo compartment of an aircraft, oil contamination
from a thermal vacuum diffusion pump, water damage
from a failed sprinkler head, and wood shipping
containers which were impossible to clean and left
wood slivers all over a flight payload).
I have also observed some things that did not cause
failures, but which could have (e.g. a wasp inside a
payload, non-pneumatic tires on airport cargo vehicles
causing a severe shock environment, and a coffee cup
ring on top of a flight box).
The point I am trying to make with the preceding list is
there are a lot of things that can wrong during the
process of designing, manufacturing, assembling,
testing and delivering a flightworthy satellite.
What we need to do with our testing is to detect the bad
parts, manufacturing problems, and design errors
without hurting the satellite by overtesting or
mishandling, and to do so at a reasonable cost.
Since the trend is toward having fewer on-board
redundancies, using fewer high-reliability parts, and
trying to reduce time/budget spent during design and
manufacturing, it seems that the importance of testing is
increasing; therefore, the relative percentage of the
project budget that we are willing to spend on testing
should increase. In the case of a failure, there will not
be a redundant subsystem to switch to! Fortunately, the
same improved technologies that have given us more
complex systems has also given us higher commanding
and data rates which do speed up testing from long ago
(try checking out a system with only one and a half
BITS PER SECOND of data transmission).
The fact that many of the subsystems are no longer
redundant shortens overall test time during payload-,
spacecraft-, and satellite-level testing.
Some of the failure types (e.g. incorrect parts, parts
mounted backwards, infant mortalities) should be
quickly detectable during box-level testing; others (e.g.
inadequate cable service loops) require environmental
stress testing to show up (either box-level or higher-
level); and some (e.g. latent ESD damage, timing races)
may take hundreds of hours of accumulated testing
before they ever show up (if they show up at all!).
Unfortunately, on many projects, by the time you start
satellite-level testing, the remaining time and budget is
In our present economic environment, all parts of
projects, including testing, will be under pressure to
conserve funds for the projects to remain marketable.
Nevertheless, I think that a rigorous test program is
required in order to deliver a good-quality satellite. For
a satellite that is designed for a lifetime of one year, I
suggest that it must have at least 500 hours of powered
satellite-level testing. (It will get almost 20 times that
amount of operation during the first year on orbit and
may get more before it fails, is turned off, or operating
funding is withdrawn.) I think that the following
recommendations will help optimize the testing. As far
as eliminating testing in order to save money, note that I
recommend performing environmental testing only at
the satellite level and using the flight structure to
perform protoflight level environmental testing. (The
more traditional approach also contains environmental
testing at the payload and spacecraft levels and requires
a mass structural model to perform qualification level
The following are philosophies that I suggest be used to
maximize the chance of detecting the type of failures
Here’s what’s next.
This article can be searched. Note: Results may vary based on the legibility of text within the document.
Tools / Downloads
Get a copy of this page or view the extracted text.
Citing and Sharing
Basic information for referencing this web page. We also provide extended guidance on usage rights, references, copying or embedding.
Reference the current page of this Article.
Greenwood, C.M. One experienced engineer`s approach to better/cheaper/faster satellite testing (philosophies and lessons learned), article, October 1, 1995; Albuquerque, New Mexico. (digital.library.unt.edu/ark:/67531/metadc665817/m1/3/: accessed November 14, 2018), University of North Texas Libraries, Digital Library, digital.library.unt.edu; crediting UNT Libraries Government Documents Department.