| Abstract: | A mathematical model for a three-component accelerograph is presented that accounts for the effects of transducer misalignment and cross-axis sensitivity. The former refers to departures in the alignment of the transducer penduli (typically of the order of few degrees) from an ideal orthogonal system, and the latter to a transducer recording components of motion in directions other than its principal sensitivity axis. Transducer misalignment magnifies the effects of cross-axis sensitivity. These effects are significant only for large amplitude recordings (peak acceleration close to 1g), for example in the near-field of the 1994 Northridge, California, earthquake. The misalignment angles and the angular amplification constant can be determined by a static tilt test, performed in the field, and solving a generalized inverse problem. This paper presents an algorithm for solution of this inverse problem on a routine basis. Practical implementation of the algorithm and results for 76 instruments of the Los Angeles Strong Motion Array are presented in a separate paper. © 1998 John Wiley & Sons, Ltd. |
