No abstract

This article is a review of optimization in relationship to experiments in four aspects. The first is the use of optimization for designing experiments to provide the maximum amount of insight and information on the phenomena being investigated. Second, the use of experiments to perform optimization is reviewed. Third, the use of techniques developed for experimental optimization in numerical optimization is discussed. These first three aspects of optimization in relationship to experiments are discussed with an emphasis on reducing the number of required experiments or analyses. Finally, the importance of experimental validation of optimization is presented. Applications from a variety of fields are provided for each of these aspects. This review article contains 119 references.

The supersonic aerodynamic and aeroelastic characteristics of a tension cone inflatable aerodynamic decelerator were investigated by wind tunnel testing. Two sets of tests were conducted: one using rigid models and another using textile models. Tests using rigid models were conducted over a Mach number range from 1.65 to 4.5 at angles of attack from -12 to 20 degrees. The axial, normal, and pitching moment coefficients were found to be insensitive to Mach number over the tested range. The axial force coefficient was nearly constant ( ! C A = 1.45 ± 0.05) with respect to angle of attack. Both the normal and pitching moment coefficients were nearly linear with respect to angle of attack. The pitching moment coefficient showed the model to be statically stable about the reference point. Schlieren images and video showed a detached bow shock with no evidence of large regions of separated flow and/or embedded shocks at all Mach numbers investigated. Qualitatively similar static aerodynamic coefficient and flow visualization results were obtained using textile models at a Mach number of 2.5. Using inflatable textile models the torus pressure required to maintain the model in the fully-inflated configuration was determined. This pressure was found to be sensitive to details in the structural configuration of the inflatable models. Additional tests included surface pressure measurements on rigid models and deployment and inflation tests with inflatable models.

On May 25, 2008, the Mars Phoenix Lander successfully landed on the northern arctic plains of Mars. An overview of a preliminary reconstruction analysis performed on each entry, descent, and landing phase to assess the performance of Phoenix as it descended is presented and a comparison to pre-entry predictions is provided. The landing occurred 21 km further downrange than the predicted landing location. Analysis of the flight data revealed that the primary cause of Phoenix's downrange landing was a higher trim total angle of attack during the hypersonic phase of the entry, which resulted in Phoenix flying a slightly lifting trajectory. The cause of this higher trim attitude is not known at this time. Parachute deployment was 6.4 s later than prediction. This later deployment time was within the variations expected and is consistent with a lifting trajectory. The parachute deployment and inflation process occurred as expected with no anomalies identified. The subsequent parachute descent and powered terminal landing also behaved as expected. A preliminary reconstruction of the landing day atmospheric density profile was found to be lower than the best apriori prediction, ranging from a few percent less to a maximum of 8%. A comparison of the flight reconstructed trajectory parameters shows that the actual Phoenix entry, descent, and landing was close to pre-entry predictions. This reconstruction investigation is currently ongoing and the results to date are in the process of being refined.

*Two Disk-Gap-Band model parachute designs were tested in the NASA Langley Transonic Dynamics Tunnel. The purposes of these tests were to determine the drag and static stability coefficients of these two model parachutes at various subsonic Mach numbers in support of the Mars Exploration Rover mission. The two model parachute designs were designated 1.6 Viking and MPF. These model parachute designs were chosen to investigate the tradeoff between drag and static stability. Each of the parachute designs was tested with models fabricated from MIL-C-7020 Type III or F-111 fabric. The reason for testing model parachutes fabricated with different fabrics was to evaluate the effect of fabric permeability on the drag and static stability coefficients. Several improvements over the Viking-era wind tunnel tests were implemented in the testing procedures and data analyses. Among these improvements were corrections for test fixture drag interference and blockage effects, and use of an improved test fixture for measuring static stability coefficients. The 1.6 Viking model parachutes had drag coefficients from 0.440 to 0.539, while the MPF model parachutes had drag coefficients from 0.363 to 0.428. The 1.6 Viking model parachutes had drag coefficients 18 to 22 percent higher than the MPF model parachute for equivalent fabric materials and test conditions. Model parachutes of the same design tested at the same conditions had drag coefficients approximately 11 to 15 percent higher when manufactured from F-111 fabric as compared to those fabricated from MIL-C-7020 Type III fabric. The lower fabric permeability of the F-111 fabric was the source of this difference. The MPF model parachutes had smaller absolute statically stable trim angles of attack as compared to the 1.6 Viking model parachutes for equivalent fabric materials and test conditions. This was attributed to the MPF model parachutes' larger band height to nominal diameter ratio. For both designs, model parachutes fabricated * Aerospace Engineer; Juan.R.Cruz@nasa.gov † Aerospace Engineer; Raymond.E.Mineck@nasa.gov ‡ Aerospace Engineer; Donald.F.Keller@nasa.gov § Aerospace Engineer; Maria.V.Bobskill@nasa.gov This material is declared a work of the U.S. Government and is not subject to copyright protection in the United States. from F-111 fabric had significantly greater statically stable absolute trim angles of attack at equivalent test conditions as compared to those fabricated from MIL-C-7020 Type III fabric. This reduction in static stability exhibited by model parachutes fabricated from F-111 fabric was attributed to the lower permeability of the F-111 fabric. The drag and static stability coefficient results were interpolated to obtain their values at Mars flight conditions using total porosity as the interpolating parameter. SYMBOLS AND ACRONYMS