Experimental Evaluation of System Level Properties of Porcelain Post Insulators Based on a Large Set of Full-Scale High-Voltage Insulators
Shakhzod Takhirov1, Frank Blalock2, Jerry Stewart3
1 Structures Laboratory, CEE, UC Berkeley
Berkeley, California, USA
takhirov@berkeley.edu
2 Southern States LLC.
f.blalock@SouthernStatesLLC.Com
3 NGK-Locke, Inc.
gstewart@ngkli.com
Abstract. The IEEE693 document allows qualification of high-voltage substation equipment by numerical analysis. For proper modelling of equipment assembled from porcelain insulators, material properties of the porcelain are required. In many cases, these properties are difficult to obtain. In addition, the attachment fixture of the insulator is quite complex. It usually consists of a metal end fitting with an oversized cylindrical space that is filled with grout between the porcelain and the end fitting’s walls. Since the properties of the interface are also unknown, providing properties of the porcelain will have limited benefits for modelling. Instead, it is better to provide a force versus displacement curve similar to the one provided by the manufacturers of polymer insulators. This curve includes the performance of the porcelain and the interface (porcelain, grout, local deformation of the end fitting). As a representative example, typical 245-kV post insulators were experimentally studied at the University of California, Berkeley. They were loaded on top in a horizontal direction to correlate the strain at the bottom of the insulator with the cantilever strength provided by the manufacturers. Commonly used high-strength porcelain insulators were used in the study. The post insulator was constructed of two insulator sections: bottom and top. A total of 6 specimens were studied. The posts were rigidly mounted for these tests and the loading was taken to 25% of the rated cantilever strength. The load was applied via a hydraulic actuator that was displacement controlled. The load was applied in both horizontal directions. The insulator was instrumented by strain gages at the bottom section and the displacement of the insulator’s top was recorded. The tests were performed to calibrate insulator strain gages and allow calculation of equivalent loads (in seismic qualification tests) from the strain data. After these calibration tests the posts were tested in a 245-kV disconnect switch on a 6DOF shaking table. The calibration tests demonstrated that the correlation between the cantilever load and the strain at the bottom of the insulator is very close to a linear relationship. It was shown that it depends on the strain’s sign: it was less for positive strains and greater for negative strains. The force versus displacement relationship was not linear. The finite element modelling with limited information usually provided by a manufacturer was evaluated and compared to the test data.
Keywords: IEEE693, numerical modelling, high-voltage porcelain insulator, experimental study, substation equipment.
