Experimental designs and regression models for calibrating nonmonolithic (multiple piece) internal wind-tunnel force balances were investigated through a case study that demonstrated fundamental deficiencies with a typical test schedule. Itwas found that the current calibration point selectionmethod,which swept thedesign space two factors at a time, introduced a degree of correlation among model terms, depending on the model form. While using the statistical design of experiment performancemetrics to analyze the deficiencies in the experimental design, it was also found that there were problems inherent to the model form itself that were independent of the design. An analysis of the calibrationmodel, endorsed by the AIAA recommended practices document for nonmonolithic balances, lead to correlated response model terms due to overparameterization. Four new modeling strategies are proposed to overcome these challenges for nonmonolithic force balances.
[1]
Norman R. Draper,et al.
Third Order Rotatable Designs in Three Factors: Analysis'
,
1962
.
[2]
N. Draper,et al.
Applied Regression Analysis
,
1967
.
[3]
N. Draper,et al.
Applied Regression Analysis: Draper/Applied Regression Analysis
,
1998
.
[4]
Robert W. Mee.
Optimal Three-Level Designs for Response Surfaces in Spherical Experimental Regions
,
2007
.
[5]
S. Huda.
Some Third-Order Rotatable Designs
,
1982
.
[6]
N. Draper,et al.
Applied Regression Analysis
,
1966
.
[7]
R. H. Myers,et al.
Response Surface Methodology: Process and Product Optimization Using Designed Experiments
,
1995
.
[8]
Roger E. Kirk.
Elementary Statistics
,
1984
.
[9]
Norman R. Draper,et al.
A Third Order Rotatable Design in Four Dimensions
,
1960
.