Tactical defenses against systematic variation in wind tunnel testing

DeLoach, Richard

doi:10.2514/6.2002-885

Cited by 27 publications

(19 citation statements)

References 3 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…An earlier study 12 estimates that the 95% confidence interval enclosing the percentage of polars afflicted with systematic covariate effects to range from 15% to 35% for that study, while the current study reports somewhat higher frequencies-56% for lift and 69% for drag. The probability that any given data sample will be substantially free of systematic effects is thus expected to be in the range of 31% to 44% by the current study and 65% to 85% by the earlier one, so it is not difficult to miss evidence of such effects, especially when few quality assessment data points are budgeted.…”

Section: Attitudes Toward Systematic Unexplained Variancementioning

confidence: 44%

“…A tunnel with such behavior would be said to be in a state of "statistical control." Unfortunately, statistical control is elusive in the real world; random fluctuations in wind tunnel response measurements are often found to occur about mean values that change systematically (not randomly) with time [10][11][12][13] . Box, Hunter, and Hunter address this in the second edition of their seminal text on experiment design 3 where they say, "The idea of a process in a perfect state of control contravenes the Second Law of Thermodynamics: Thus an exact state of control is unrealizable, and must be regarded as a purely theoretical concept."…”

Section: A Unexplained Variance and Its Impactmentioning

confidence: 99%

“…The sample standard deviation for the 12 differential lift coefficient values was 0.0079, which we divide by the square root of the sample size (12) to compute the standard error in the estimate of the mean. Thus, we would report a mean difference of 0.0008, with a value of 0.0023 as the standard error in the estimate of that mean.…”

Section: A Hypothesis Testing With Paired T-testmentioning

confidence: 99%

See 2 more Smart Citations

A Practical Methodology for Quantifying the Random and Systematic Components of Unexplained Variance in a Wind Tunnel

DeLoach

Obara

Goodman

2012

50th AIAA Aerospace Sciences Meeting Including the New Horizons Forum and Aerospace Exposition

View full text Add to dashboard Cite

This paper documents a check standard wind tunnel test conducted in the Langley 0.3-Meter Transonic Cryogenic Tunnel (0.3M TCT) that was designed and analyzed using the Modern Design of Experiments (MDOE). The test designed to partition the unexplained variance of typical wind tunnel data samples into two constituent components, one attributable to ordinary random error, and one attributable to systematic error induced by covariate effects. Covariate effects in wind tunnel testing are discussed, with examples. The impact of systematic (non-random) unexplained variance on the statistical independence of sequential measurements is reviewed. The corresponding correlation among experimental errors is discussed, as is the impact of such correlation on experimental results generally. The specific experiment documented herein was organized as a formal test for the presence of unexplained variance in representative samples of wind tunnel data, in order to quantify the frequency with which such systematic error was detected, and its magnitude relative to ordinary random error. Levels of systematic and random error reported here are representative of those quantified in other facilities, as cited in the references.

show abstract

Section: Attitudes Toward Systematic Unexplained Variancementioning

confidence: 44%

Section: A Unexplained Variance and Its Impactmentioning

confidence: 99%

Section: A Hypothesis Testing With Paired T-testmentioning

confidence: 99%

See 1 more Smart Citation

A Practical Methodology for Quantifying the Random and Systematic Components of Unexplained Variance in a Wind Tunnel

DeLoach

Obara

Goodman

2012

50th AIAA Aerospace Sciences Meeting Including the New Horizons Forum and Aerospace Exposition

View full text Add to dashboard Cite

show abstract

“…The problem of systematic unexplained variance is most effectively attacked during the acquisition of the data, by invoking certain quality assurance tactics in the design of the experiment that eliminate the systematic error by converting it to another component of random error. [30][31][32] These tactics are key elements of the Modern Design of Experiments and are reflected in Data Sets 3 and 4. )…”

Section: Phase Four: Diagnostic Evaluationmentioning

confidence: 99%

A Comparison of Two Balance Calibration Model Building Methods

DeLoach

Ulbrich

2007

45th AIAA Aerospace Sciences Meeting and Exhibit

View full text Add to dashboard Cite

Simulated strain-gage balance calibration data is used to compare the accuracy of two balance calibration model building methods for different noise environments and calibration experiment designs. The first building method obtains a math model for the analysis of balance calibration data after applying a candidate math model search algorithm to the calibration data set. The second building method uses stepwise regression analysis in order to construct a model for the analysis. Four balance calibration data sets were simulated in order to compare the accuracy of the two math model building methods. The simulated data sets were prepared using the traditional One Factor At a Time (OFAT) technique and the Modern Design of Experiments (MDOE) approach. Random and systematic errors were introduced in the simulated calibration data sets in order to study their influence on the math model building methods. Residuals of the fitted calibration responses and other statistical metrics were compared in order to evaluate the calibration models developed with different combinations of noise environment, experiment design, and model building method. Overall, predicted math models and residuals of both math model building methods show very good agreement. Significant differences in model quality were attributable to noise environment, experiment design, and their interaction. Generally, the addition of systematic error significantly degraded the quality of calibration models developed from OFAT data by either method, but MDOE experiment designs were more robust with respect to the introduction of a systematic component of the unexplained variance. = Difference between total sum of squares and explained sum of squares Significance = Risk of erroneously rejecting a null hypothesis SVS = Single Vector System t-Limit = Minimum acceptable probability that a given term is significant t-Value = Measured quantity expressed as multiple of standard error in measurement Total SS = Total sum of squares. Measure of variability in data set VIF = Variance Inflation Factor/ A measure of multicollinearity

show abstract

“…3,4 Such variations can be due to any number of factors, but among the most common are instrument drift, warm-up effects, thermal expansion, and even operator fatigue (or operator learning effects, by which the operator's performance improves after an initial acclimation period). These variations have in common the fact that they are not random, so that ordinary replication will do little to cancel them out.…”

Section: Introductionmentioning

confidence: 99%