How-to Articles
Thermal Sensitivity Study
Cost and Mass Trade Study
Thermal Model Correlation
Leveraging the speed of reduced-order models (ROMs), thermal analysis teams have previously had access to rapid optimization, sensitivity studies, rapid model correlation, and uncertainty quantification, to name a few. An advantage of the developed approach is its relative robustness, which enabled expansion to include structural models in addition to thermal models leading to both rapid thermal-structural and rapid structural-only analyses.
In this paper, thermal-structural ROMs were developed for several applications including a complex NASA mission. The results show how ROMs could be used to quickly evaluate system uncertainties. Finally, these methods were applied to a structural-only application and used to better optimize and understand the structural design.
Advanced SmallSat Thermal Analysis Using Reduced-Order Models
This paper examines the use of Reduced-Order Models (ROMs) for Small Satellite applications. Use cases include rapid thermal model correlation, sensitivity analyses, and Uncertainty Quantification.
Efficient thermal architecture for large space telescopes
This paper examines the Large UV/Optical/Infrared Surveyor (LUVOIR) deployable sunshade and explores alternative ideas for accomplishing the thermal conditions needed for this or similar missions. In this work, Reduced Order Models (ROM) using Veritrek were developed from training data generated using Thermal Desktop. This ROM enabled rapid exploration of the operational envelope. Veritrek results were “both striking and insightful” and “surprising and beneficial to the LUVOIR architecture”.
An Improved Thermal Model Correlation Process Using Veritrek’s Reduced-Order mOdeling Software
Often, model correlation processes involve institutional knowledge and an iterative “guess and check” method that can become time-consuming and costly. Veritrek’s Correlation Analysis feature provides a new approach to correlating thermal models to test data by using reduced-order modeling to explore thousands of parameter combinations in a few seconds.
Engineers at Ball Aerospace used the Veritrek software to help determine key parameter values that made model outputs match test data for an Internal Research and Development (IRAD) effort involving cryo instruments. With Veritrek, over 20,000 combinations of parameter values were quickly explored and produced a few dozen viable solutions for correlating the Thermal Desktop® model. In total, this model correlation effort would have typically taken an estimated 3-4 weeks to complete; but with Veritrek, a best solution was determined in an automated and repeatable fashion within a few days. Not only did the Veritrek approach save time, but it provided much more confidence in the chosen best solution.
Optimizing thermal radiator designs using the Veritrek software
A common problem exists in the aerospace industry when the thermal design is driven by minimizing survival power while maintaining compliance with allowable flight temperatures (AFTs) in a variable thermal environment. Finding an optimal design solution can be time-consuming and requires a large number of simulations. ATA used the Veritrek software to create a ROM to efficiently solve this complex problem and avoid an overdesigned thermal system. In all, it took about five days to reach an optimal design solution using Veritrek, an effort that would have taken about one month using traditional thermal analysis techniques.
Assessment of the Mars helicopter thermal design sensitivities using the Veritrek software
The Mars Helicopter will be a technology demonstration conducted during the Mars 2020 mission. A reduced-order model (ROM) created with the Veritrek software was utilized to explore the sensitivities of the thermal system’s drivers, such as electronics dissipations, gas gaps, and heat transfer coefficients, as well as to assess and verify the final thermal design. The results produced by Veritrek were utilized to study the effect of the major thermal design drivers and Mars environment on the Mars Helicopter in as little as 10 days, an effort that would have taken over 4 months using traditional thermal analysis techniques.
Automatic creation of Reduced-Order Models using Thermal Desktop
A reduced-order modeling approach to predict spacecraft output responses for a set of input factors was developed. It is based on Latin Hypercube sampling and Gaussian Process regression modeling. This approach was successfully applied to a broad range of applications including the Orion Crew Exploration Vehicle and a nominal Hex Spacecraft Bus. Results compared favorably to the underlying Thermal Desktop® model.
Enhanced data exploration through Reduced-order Models
A reduced-order modeling approach to predict spacecraft output responses for a set of input factors was developed. It is based on Latin Hypercube sampling and Gaussian Process regression modeling. A test case, based on a simplified Orion Crew Exploration Vehicle Thermal Desktop® model, was developed and included nine input factors and seven output responses. Output response residuals, found for predicted temperatures, hydraulic power, and pressure, had means of 1.6 K, 0.2 W, 1.6 kPa and standard deviations of 5.0 K, 1.93 W, 18.2 kPa, respectively.