Fraunhofer-Gesellschaft

Publica

Hier finden Sie wissenschaftliche Publikationen aus den Fraunhofer-Instituten.

Fast view factor determination for thermal modelling

 
: Gulde, Max; Montemayor Mancías, Javier; Schimmerohn, Martin; Schäfer, Frank

:
Volltext (PDF; )

European Space Agency -ESA-, Paris:
European Space Thermal Engineering Workshop 2018. Proceedings : ESA/ESTC, Noordwijk, The Netherlands, 29-31 October 2018
Noordwijk: ESA, 2018
ISSN: 1022-6656
S.297-312
European Space Thermal Engineering Workshop <2018, Noordwijk>
Englisch
Konferenzbeitrag, Elektronische Publikation
Fraunhofer EMI ()

Abstract
While we equip today’s small satellites and in particular CubeSats with progressively sophisticated payloads and let them perform increasingly complex tasks [1], [2], thorough thermal modelling is usually not compatible with development time scales and resources. For a thermal modelling approach to be suitable for small missions, it must mirror their design spirit: It should (a) easily adapt to design modifications, (b) be resource efficient, and (c) be performant enough to allow fast simulations of the whole system.
One of the main computational bottlenecks of thermal modelling is the view factor determination to calculate radiative heat transfer [3]. Conventionally, we perform this step using Monte-Carlo Ray Tracing (MCRT) methods. For small spacecraft, however, this approach can prove too time- and resource consuming.
Here, we introduce two methodologies for very fast view factor determination commonly used in the field of graphics-processing-unit-based 3D rendering [4]. For insolation, we employ an orthographic projection algorithm, allowing for the computation of several thousands of view factors per second. Specifically, we substantially rely on parallel processing to achieve very high performance.
Similarly, we compute intra-spacecraft view factors by a radiosity hemicube approximation [5]. This method natively incorporates the correct angular and distance relations required. Furthermore, we use MIP mapping to quickly extract the view factors from the resulting images [6]. Comparison with analytical solutions show good agreement.
In order to demonstrate the feasibility of the approach, we simulate the thermal environment of the nanosatellite ERNST [7] currently in development at Fraunhofer and validate the results with commercial off-the-shelf thermal modelling software as well as experimental thermal vacuum tests.
We believe that this approach is particularly well suited for the thermal modelling of small spacecraft and will help to propel future development efforts.

: http://publica.fraunhofer.de/dokumente/N-534703.html