1.3 Advances in Experimentation
Effects of Actuator Tuning on Pseudo-Dynamic Tests
J.T. O’Hagan & Q.T. Ma
ABSTRACT: Experimental error is an unavoidable feature of all experimental investigations. The effects of this on the accuracy of pseudo-dynamic (PSD) testing have been the subject of many studies but it remains not well understood and it has been rarely validated experimentally. An important consideration when investigating experimental errors in PSD tests is the performance of the actuator. In general, the actuator motion will lag the displacement command. It is important to acknowledge this delay in the PSD algorithm to ensure correct restoring forces are measured. Incorrect restoring force arising from systematic measurement errors can propagate through the PSD solution process and invalidate the results. This paper reports on a series of experiments assessing the optimum wait time for a default tuned actuator to established a stabilised feedback value for PSD tests. Force feedbacks were intentionally measured prematurely to quantify the error propagation in the algorithm. It was experimentally shown that errors of this kind did not significantly alter the time history response for a harmonically excited SDOF system with damping. The results showed that a constant hold time is not practical as actuator lag time can vary dramatically. It is recommended that actuator displacement is checked iteratively within each load step to ensure the command displacement is achieved before feedback values are measured. This would reduce systematic error and in turns minimise numerical damping in the PSD algorithm.
A Model-Based Predictor-Corrector Algorithm for Substructure Hybrid Test System
Tao WANG & Chun CHENG
ABSTRACT: The online hybrid test, commonly combined with substructure technique, has been developed for more than thirty years, and promoted significantly in the past decade. Proposed in this paper is a hybrid test framework that is able to accommodate multiple tested and numerical substructures. A coordinator program is designed to integrate and update the stiffness matrix associated with the boundaries. It also solves the dynamics of the overall structure using an equivalent static form of the equations of motion. The boundary stiffness matrix is originally obtained from a pre-static test, while updated in each step by means of a quasi-Newton procedure. A non-iterative predictor-corrector algorithm is designed to advance the time-varying process step by step. One unique feature of the proposed framework is its model-based predictor which is much desirable to solve the difficulty in dealing with static degrees of freedom at the boundaries. The proposed online hybrid test system has been demonstrated effectively by one round numerical simulation using a three-storey braced frame, and a physical test of a one-bay one-storey frame.
Damage Assessment of Seismically-Excited Buildings through Incomplete Measurements
G.L. Lin, C.C Lin & K.S Hsieh
ABSTRACT: A real structure possesses large number of degrees of freedom. It is impossible and impractical to have measurements at all degrees of freedom. This study presents a damage assessment technique for seismically-excited buildings through incomplete measurements. The system realization using information matrix (SRIM) identification technique was firstly employed to estimate the modal properties such as frequencies and damping ratios of the instrumented building. However, the complete mode shapes can not be obtained due to incomplete measurements. This study proposed an optimal mode shape recovery technique to construct the complete first mode shape of the building system. An optimization process was performed to minimize a prescribed objective function representing the error between the measured and estimated outputs at the instrumented locations. A storey damage index (SDI) computed through the recovered first mode shape was applied to express the degree of storey damage. A five-storey shear building considering measurement noise was conducted for numerical simulations. Finally, a three-storey benchmark building was considered to examine the accuracy and applicability of the proposed damage assessment technique via experimental data. It is shown that the proposed method obtained results agreeing fairly well with the results of full measurements and is of value in practical application.
Image-Based Measurement Techniques for Earthquake Engineering
Y.S. Yang, C.W. Huang & C.L. Wu
ABSTRACT: This paper presents a simple image-based measurement method to measure surface strains of earthquake engineering experimental specimens. The method is cost-effective, easy to apply, and provides satisfactory measurement accuracy. This method integrates camera calibration, stereo triangulation, image metric rectification, and image template matching techniques. Applications of this method for a zero-strain test and an RC-wall experiment are presented in this paper. The zero-strain test results showed that the measurement accuracy achieves 0.04 pixels. That is, the relative displacement accuracy achieves 0.005mm and the strain accuracy achieves 0.001, when using eight-mega-pixel digital cameras, measurement region sized 16cm by 23cm, and a 31-by-51 measurement grid. In the application for the RC-wall experiment, concrete cracks with 0.02mm width can be found by examining measured displacement fields. Falling of surface concrete can be detected by examining the measurement results of this method.
Experimental Testing of Full-Scale Timber Floor Diaphragms in Unreinforced Masonry Buildings
A.W. Wilson, P.J.H. Quenneville & J.M. Ingham
ABSTRACT: Unreinforced masonry (URM) buildings in New Zealand are typically constructed with rigid clay brick perimeter walls and comparatively flexible timber floor diaphragms. URM construction represents the predominant architectural heritage of our nation but the preservation of these buildings is threatened due their well established inadequacy to withstand earthquakes. Timber floor diaphragms are widely recognized to significantly influence the seismic response of URM structures and the accurate assessment of diaphragms is therefore crucial in any seismic assessment and retrofit of URM buildings. As part of a wider research program, a series of full-scale diaphragm tests were performed to generate the much needed data to critique the current state-of-the-art desktop procedures. In this contribution, the NZSEE and ASCE 41-06 procedures are used to predict full-scale diaphragm performance and are compared against experimentally determined values. It was found that inconsistency exists between the two assessment documents with considerable differences found in strength, stiffness and ductility predictions. The procedures published in NZSEE and ASCE 41-06 poorly predicted diaphragm response and require updated and representative values that include provisions to address the highly orthotropic nature of timber diaphragms. These documents should also be harmonized to ensure that transparency and consistency exists between international assessment procedures.
The Role of Spandrels within Masonry Walls with Openings: An Experimental Investigation
F. Parisi, N. Augenti & A. Prota
ABSTRACT: Quasi-static lateral loading tests were carried out on a full-scale masonry wall with an opening to assess the contribution of spandrels to the in-plane nonlinear behaviour of masonry walls. The wall response was first assessed through a monotonic test up to the attainment of moderate cracking in the spandrel. Second, the pre-damaged wall was cyclically tested up to about the same lateral drift reached in the monotonic test. Finally, the wall was repaired and upgraded through inorganic matrix-grid (IMG) composites applied to the spandrel. A cyclic test was then performed up to a near-collapse state. In this paper the experimental results are discussed with special emphasis to parameters that quantify the in-plane nonlinear response. In all tests, the piers provided high displacement capacity through a rocking response. Both the load-carrying capacity and the post-peak strength degradation were affected by the nonlinear behaviour of the spandrel panel. The IMG strengthening system provided high energy dissipation capacity to the spandrel panel, restoring the load-carrying capacity of the as-built wall, and delaying strength degradation at larger drifts. Low residual displacements and full reversibility could make the IMG strengthening system attractive for historic buildings.