Date of Award

Spring 2021

Document Type

Dissertation

Degree Name

Doctor of Philosophy (PhD)

Department

Mechanical & Aerospace Engineering

Program/Concentration

Mechanical Engineering

Committee Director

Drew Landman

Committee Member

Colin Britcher

Committee Member

Oleksandr Kravchenko

Committee Member

Louis Centolanza

Abstract

In this work, a new strategy is presented to wire, calibrate, and measure strain gages for rotor blade testing that will provide more information and is robust to individual gage loss. The additional information can be used in several ways. including reducing redundancy, offering rapid identification of damage locations, and in some cases reducing risk allowing tests to continue to collect calibrated data after one or more sensors have failed. This strategy replaces the classical four-gage full Wheatstone bridge with four separately wired quarter bridges that are combined into a full bridge in the data acquisition system using a calculated channel. This strain gaging concept also provides the engineer with data from the four individual stresses, as well as the calibrated full bridge output that has been converted into engineering units. The Design of Experiments (DOE)-based calibration approach was used to increase the model accuracy and to provide better evaluation of calibrated model fit.

This strategy was evaluated first by instrumenting, calibrating, and testing a well predicted isotropic aluminum I-beam and then by repeating this process on a metal composite hybrid helicopter tail rotor blade. Both articles were instrumented using both the proposed quarter bridge and the classical full bridge method. After instrumentation, the articles were calibrated using a Central Composite Design, second order regression modeling, and ANOVA for parameter significance testing. Lastly, several tests were conducted on the tail rotor blade article to demonstrate the validity of the quarter bridge concept and to prove the concept’s robustness to individual gage loss.

The DOE calibration approach provided a significant improvement over the traditional One Factor at a Time calibration approach, yielding an order of magnitude better fit results when comparing the Coefficient of Determination. A comparison of the residual sum of squares for confirmation points showed a three-fold decrease using the Central Composite Design. Results of thermal drift testing did not reveal any significant difference between the current and proposed strategy, and the error analysis showed a potential increase in error (due to the data acquisition system) in the range of 0.25 to 0.50% using the available hardware. A simplified first order plus interaction model was used in each of the calibrations of the test articles to support rapid matrix inversion. Use of a quadratic model could improve the accuracy by an additional 0.28%. The virtual repair demonstration also showed a slight increase in error when using a two-gage half bridge after failure of a gage in the full bridge. After three consecutive random virtual repairs of real damage, the error only increased by 1.1% of the full-scale output. In a cost analysis, it was determined that a single virtual repair could save as many as 70 manhours of labor and could eliminate two weeks of down time in rotor blade ground testing. In flight testing, even more cost and time could be saved.

DOI

10.25777/z643-1g53

ISBN

9798515245979

ORCID

0000-0002-0344-6505

Share

COinS