Vibration monitoring analysis of tup – coe building: correlating raspberry shake 4d sensors with finite element modeling for structural health assessment using machine learning/
Brix N. Buenaventura, Joshua Ryan G. Ferrer, Isabel A. Marcelino, Joshua R. Mayo, and Desiree G. Robiso.--
- Manila: Technological University of the Philippines, 2024.
- x,177pages: 29cm.
Bachelor's thesis
College of Engineering.--
Includes bibliographic references and index.
The increasing frequency of seismic events underscores the need for advancements in Structural Health Monitoring (SHM), especially in regions like the Philippines, where complex seismic activities pose significant challenges for the development of real-time, cost-effective monitoring systems. This study explores the use of Raspberry Shake 4D (RS4D) sensors to promote efficient and affordable SHM in the College of Engineering Building at the Technological University of the Philippines. Using vibration analysis and Finite Element Modeling (FEM), the research theoretically examines the building's structural response. The acceleration data from the RS4D sensor were processed using the Savitzky-Golay filter and Fast Fourier Transform (FFT) to assess the building's dynamic properties, including its period and frequency. The natural frequencies at the peak amplitude of the event were found to be approximately 1.1 Hz, 1.3 Hz, and 1.6 Hz, corresponding to periods of 0.909 s, 0.769 s, and 0.625 s, respectively. The acceleration values generated for the fifth floor using FEM and the actual fifth-floor sensor data were 0.333 in/sec2 and 0.324 in/sec2, respectively. In addition to monitoring structural health, the study also explores the potential for these sensors to detect subtle vibrations and motions, thus contributing to activity monitoring and occupancy detection. The integration of modern technologies with traditional engineering approaches is emphasized. This research validates RS4D sensors as a reliable tool for SHM and explores their potential for innovative applications in activity monitoring, enhanced through the integration of machine learning techniques.