Modal analysis is a dynamic response-based experimental structural identification procedure widely used for condition assessment and structural health monitoring (SHM). Piezoelectric sensors are smart material-based dynamic strain sensors being progressively implemented in vibration-based monitoring techniques. However, these have been so far limited to preliminary investigations on 1D and 2D structures with regard to modal analysis. Strain modal analysis using piezoelectric sensors has so far utilized only lower bending modes for frequency and mode shape-based SHM. This paper aims to evaluate the piezoelectric sensors for modal response measurement, modal parameter identification, and structural health monitoring of 3D structures, especially using torsional modes. It reports a comparative experimental study between the piezoelectric sensors and the accelerometers for dynamic response measurement, modal parameter extraction and damage identification and localization. Experimental modal analysis using impact hammer-based excitation is performed on two lab-sized specimens under different boundary conditions. Displacement mode shapes from both the sensors are used to determine the location and the severity of the damage. Torsional and bending modes are studied distinctly for sensitivity towards damage monitoring. Modal assurance criteria (MAC), coordinated MAC (CoMAC), and modal strain energy (MSE) are adopted as damage indicators. Results of the study show that the piezoelectric sensors are equivalent to accelerometers for response measurements and modal parameter extraction. They possess enhanced sensitivity towards damage monitoring as compared to accelerometers. Torsional modes perform better than bending modes for damage localization and severity quantification. Bending modes from both sensors are ineffective in locating the damage precisely. Modal strain energy (MSE) based damage index from torsional modes acquired from piezoelectric sensor performs best in terms of sensitivity towards damage location and exhibits peak severity, especially for incipient damage. The study establishes piezoelectric sensors as reliable and cost-effective modal response sensors.