Date of Award

May 2018

Degree Type

Thesis

Degree Name

Master of Science

Department

Engineering

First Advisor

Anoop K Dhingra

Committee Members

Nidal Abu-Zahra, Nathan Salowitz

Abstract

Any structure which is used in service can be subjected to different kinds of loads such as static, dynamic, moving or impact loads. It is very much necessary such structures are reliable and robust enough to be used for service. Hence accurately estimating the load acting on the structure is very important. In case of impact loading, it is very difficult to estimate the load because it acts for a short period of time. For such loading, the structural response of the system can be used as a medium to estimate the applied load. The structural response used could be strain or displacement. This is known as the "inverse problem''.

The response cannot be measured at the locations on the structure as this will be cost-prohibitive. The inverse problem techniques encounter some limitations such as usage of limited number of sensors due to financial constraints, inaccessible location to place the sensor and the influence of sensor on the structural response. Due to these limitations, it is very much obvious that the load estimates have few errors. By choosing the correct optimum locations to place the sensors, it is possible to minimize these errors.

The impact load is recovered in this work by using the strain values at the chosen optimum locations. The technique used to choose the optimum location to place the sensors is called D-optimal technique. In this thesis, the D-optimal technique is used extensively. The recovery of impact loads through measurement of structural response at a finite number of optimally selected location is demonstrated. This optimum sensor locations are identified using the D-optimal design Algorithm. Separate algorithms are developed to recover the impact load. ANSYS APDL 17.2 and MATLAB are programming software were used to estimate the applied loads. Based on the results obtained from several numerical examples, it is seen that the technique presented gave fairly accurate load estimates.

Share

COinS