Experimental Setup for vibration analysis

Project Details

Project Overview

I was tasked with developing an experimental setup to characterize blade vibrations in centrifugal-compressor impellers of various scales. From mechanical design through software control, I created an integrated system allowing precise excitation and non-contact measurement of vibrational modes—validating simulation results and informing operational limits.

Core Contributions

Experimental Rig Design

• Mechanical Fixture & Shaker Integration: Engineered a robust base to support impellers weighing up to 1 t, mounting an electrodynamic shaker beneath the hub.
• Laser Vibrometer Deployment: Selected and configured a laser Doppler vibrometer, installing reflective targets at multiple blade locations for non-contact velocity measurements.
• Multi-Point Scanning: Devised a manual scanning protocol around the impeller circumference, ensuring coverage of all blades and enabling spatial mode mapping.

Signal Generation & Data Acquisition

• MATLAB-Driven Excitation: Wrote MATLAB scripts to output sine, step, chirp, and custom waveforms to the shaker controller, tuning frequency ranges to isolate resonant peaks.
• NI DAQ Integration: Employed National Instruments hardware and LabVIEW/Python interfaces to synchronize shaker drive signals with laser-vibrometer outputs and amplifier channels.
• Calibration & Amplification: Selected appropriate signal amplifiers; performed calibration routines to correct for gain and phase shifts in the measurement chain.

Data Analysis & Visualization

• Modal-Analysis GUI: Developed a MATLAB GUI to display vibration amplitude vs. frequency for each measurement point, overlaying results on 3D impeller models for intuitive mode-shape visualization.
• Resonant-Frequency Extraction: Applied signal-processing techniques (FFT, bandpass filtering, peak detection) to identify dominant modes and compare against ANSYS simulation predictions.
• Operational Envelope Recommendations: Analyzed frequency-response curves to recommend safe operating speeds—avoiding excitation bands that could induce high-amplitude blade bending.

Automation Proposal

• Rotating Platform Concept: Designed a motor-driven turntable to index the impeller in precise angular increments, paired with a pan–tilt laser mount for automated scanning.
• Vision-Guided Targeting: Proposed integrating a machine-vision camera and OpenCV routines to locate reflective targets dynamically, reducing setup time and human error.
• Software Workflow: Outlined an end-to-end control architecture: from CAD-driven scan-point generation through automated test execution and data-archival in a database.

Technologies & Tools

• Software & Scripting: MATLAB (scripts & GUI), LabVIEW, Python
• Hardware: Laser Doppler Vibrometer, Electrodynamic Shaker, NI DAQ modules, Signal Amplifiers
• Mechanical Design: CAD for base fixture (SolidWorks/Fusion 360)
• Signal Processing: FFT, windowing, peak detection algorithms
• Simulation Comparison: ANSYS modal-analysis results
• Vision & Automation (proposal): Pan–tilt laser mount, OpenCV, stepper-motor turntable

Outcomes & Impact

• Established a repeatable procedure for capturing blade-vibration profiles across a wide range of impeller sizes.
• Identified optimal excitation signals and minimal measurement-point sets for accurate modal characterization.
• Delivered a MATLAB-based visualization tool that streamlined data interpretation and simulation validation.
• Laid the groundwork for a fully automated test bench—promising faster, more reliable vibration testing in future developments.