NASA Jet Propulsion Laboratory California Institute of Technology JPL Home Earth Solar System Stars & Galaxies Science & Technology Bring the Universe to You JPL Email News RSS Podcast Video
Follow this link to skip to the main content

Awards / Achievements
National Academy Members
Professional Society Fellows
SRS List
News and Events
ST icon
Office of the Chief Scientist and Chief Technologist


<< Back To List Page

Juno: Thermal Math Model Meshing
06.23.2009 2:54 PM
Ruwan Somawardhana

By Ruwan Somawardhana
Spacecraft Thermal Engineering

For the Juno Telecom panel, I am currently focused on creating my thermal math model (TMM). This model will be used to predict the thermal performance of the hardware under various operating conditions. Initially, these predictions will help to plan the combined thermal balance and qualification test that is scheduled for the telecom panel in early 2010. We can use the model to determine whether our test setup is adequate and to help define what typical temperatures to expect during testing. If any AFT violations are expected (either due to the panel configuration or the test setup), it will allow for minor changes to be made before testing (such as increasing the radiator size on the panel, adding or changing the locations of heat exchangers, and adding test heaters) to try and avoid the off-nominal performance.

After testing is completed, the model will then be modified and tuned to match actual performance data. The tuned model can then be used to make in-flight predictions which are critical to flight operations to ensure the thermal health of spacecraft components.

In order to make the TMM, the first step is to get the geometry of all the parts. I was able to get this information from my mechanical/structural counterpart on the team. The next big step is to mesh all the parts to create the finite element model. This requires careful attention and can be the difference between a good model and a poor model. If the mesh size is too coarse, you lose resolution and may not get critical temperatures. If the mesh size is too fine, the computation time becomes impractical. I have learned early on that thermal meshes don’t need to be nearly as fine or detailed as a structural mesh. This often means that the geometric parts to be meshed must first be simplified and often manual meshing results in faster model run times. Although manual meshing does take longer than using auto-mesh features, it is often worth it for thermal models to reduce computation time.

Site Manager:  Brian Knosp
Webmasters:  Cornell Lewis, Maryia Davis

JPL Clearance:  CL#08-4147