Thermal Engineering

1
Thermal Engineering

• Testing

2
Some Links
Paper on the importance of testing http//www.aero
.org/conferences/ats/ThrmlVac.pdf
photogrammetry A Photogrammetry System for Use
in Thermal Vacuum Testing http//esapub.esrin.esa.
it/bulletin/bullet81/bouman81.htm Paper on
vacuum vs in air testing http//techinfo.jpl.nasa.
gov/practice/1409.pdf Smithsonian Astrophysical
Observatory Thermal Vacuum Chamber and Solar
Simulator http//ce-www.harvard.edu/technical_reso
urces/solar_simulator.html JPL http//etl-web.jp
l.nasa.gov/ MARIE instrument testing http//mari
e.jsc.nasa.gov/Testing.html Mars Lander
http//mars.jpl.nasa.gov/2001/lander/thermvac/tv14
-MAR-00/ Test procedure http//sprg.ssl.berkele
y.edu/sprite/tech_1/imagepdf/8310-w7a.pdf
3
More Links
22nd Space Simulation Conference, Oct. 2002
http//home.earthlink.net/ssc2002/space22/confere
nce.htm Radarsat http//www.ccrs.nrcan.gc.ca/cc
rs/data/satsens/radarsat/photos/radpix_e.html
Voyager http//voyager.jpl.nasa.gov/image/assem
bly.html Ball Aerospace http//www.ball.com/aero
space/fac_tvac.html Johnson Space
Center http//lmso.houston.external.lmco.com/seat
/testfacilities/pages/thermal.htm IMAGE
spacecraft http//pluto.space.swri.edu/IMAGE/slid
e/slide_13.html Spacecraft Thermal Control NASA
paper http//trs.nis.nasa.gov/archive/00000152/01/
sp8105.pdf
4
Why Test?

• Testing is a lot of work
• Testing is expensive
• Takes a lot of preparation
• Not very flexible
• Limited range of cases and conditions
• Actual operational range and environment may not
  be possible to recreate under test conditions
• Results are difficult to interpret
• May be dangerous

5
Analysis
A x T q

• All thermal analysis comes down to setting up
  this system of equations and trying to solve it,
  usually numerically. May contain non-linear
  elements
• From where does the coefficient matrix come?
• Sometimes from first principal physics
• Other analysis
• Usually from testing
• In general what we put in A is pretty "soft"
• Contact resistances
• Heat transfer coefficients
• Materials properties
• Only cases which is even near to be deterministic
  are
• Some types of radiation interchange in a vacuum
• Conduction in a continuous solid
• q often also has a high uncertainty
• EEs aren't interested in predicting thermal
  losses, only performance.
• Analysis predicts end up with "margins" added on
  to them

6
Hand-in-Hand

• Analysis and Testing have to work together in the
  design process
• Testing feeds into A and then the model has to
  be correlated against test results
• In an ideal world, testing and analysis would
  continue in a ye until the results of each close
• In real life might have one retest
• Big assumption
• A correlated model can make predictions for
  conditions that can't be tested.
• After a test, the "margins" are reduced?
• Tests have to be designed so that so that test
  data can be used to correlate models
• Location of instrumentation is the great
  challenge of testing.
• Soft parts of the model should be identified
  before the test.
• Besides correlation, testing is also used to
  validate workmanship and performance.
• Testing can be performed on engineering units or
  production hardware.

7
Testing

• Preparation is everything
• Requirements flow down
• Reviews
• Safety concerns
• Personnel
• Hardware
• Instrumentation
• Test fixturing
• Timelines
• Logbooks
• Data logging
• Document