AE data (Hsu-Nielsen tests) from overwrapped composite pressure vessel type-IV
Description
Vallen files: - Contains the raw acoustic emission data for each of the 40 marked areas on the pressure vessel - This data was directly recorded using the Vallen measurement system Matlab files and h5 files: - These folders contain processed and filtered data, with data for each measurement area provided separately within each folder type - For each area, the data is structured into two main arrays: -- TRAI_TIME_MAP (variable mapping): This is a numerical array with dimensions (2, N, 4) --- The first dimension (size 2) serves to differentiate between two types of information for each measurement event and sensor: ---- Index 1 (Matlab m files)/ Index 0 (Python HDF files): Contains the Transaction Record Identifiers (TRAIs). Each TRAI is a unique identifier associated with a specific acoustic emission event detected by a sensor ---- Index 2 (Matlab m files)/ Index 1 (Python HDF files): Contains the Time of Arrival (ToA) of the acoustic emission signal at each sensor, measured in microseconds or other consistent time units from the start of the recording epoch for that event --- N represents the total number of distinct acoustic emission measurements (events, e.g., pencil lead breaks) recorded for that specific area --- The third dimension (size 4) corresponds to the four acoustic emission sensors used in the setup --- For example, TRAI_TIME_MAP[1, i, j] would provide the TRAI for the i-th measurement event as captured by the j-th sensor, while TRAI_TIME_MAP[2, i, j] would provide the corresponding Time of Arrival (using Matlab) -- TRAI_WAVEFORMS: This array with dimensions (L, M) stores the complete acoustic emission waveform data for all recorded signals. --- L is the sample length of each individual signal (e.g., 4096 samples) --- M is the total number of unique recorded waveforms (identified by TRAIs) for that specific measurement area (e.g., M = 610). --- This array serves as a "pool" of all waveforms, where each column m is a complete, unique signal - Explicit Mapping Logic: -- The two arrays are linked via an index-based lookup. To find the waveform for a specific event i and sensor j, you must perform the following two steps: --- Get the Waveform Index: First, retrieve the mapping index, let's call it m, from TRAI_TIME_MAP. This index m corresponds to the specific TRAI for that measurement --- m = TRAI_TIME_MAP[1,i,j] (Matlab usage)/ m = TRAI_TIME_MAP[0,i,j] (Python HDF usage) --- Retrieve the Waveform: Use this index m to select the correct column from the TRAI_WAVEFORMS array. The complete signal is the entirety of that column --- Waveform Signal = TRAI_WAVEFORMS[:,m] -- Matlab uses numbering from 1, HDF files are created in Python therefore index numbers start from 0
Files
Steps to reproduce
Experimental Setup & Methodology Facts Vessel Configuration - The vessel surface was divided into 40 marked areas - Areas were labeled with letters (A, B, C, D) in the circumferential direction - Areas were numbered from 0 to 9 in the axial direction Acoustic Emission (AE) Sensors - Four AE sensors were utilized - Sensor details: -- Two VS75-SIC sensors: --- Resonance frequency: 75 kHz --- Equipped with 34-dB embedded preamplifiers -- One VS150-RIC sensor: --- Resonance frequency: 150 kHz --- Equipped with a 34-dB embedded preamplifier -- One VS150-M sensor: --- Resonance frequency: 150 kHz --- Utilized an external 34-dB preamplifier - A grease-based couplant was applied between sensors and the specimen for optimal acoustic coupling - Sensors were securely fixed using tension straps and grippers attached to the vessel's surface - Sensor positioning criteria: -- Ensure different signal propagation paths for each sensor -- Enable detection of AE sources in all 40 marked areas Experimental Procedure - The procedure involved three main steps: -- Generation of AEs at various locations -- Acquisition of signals using AE equipment -- Processing of the acquired signals AE Signal Generation - AE signals were generated using the Hsu-Nielsen method -- This method involves the breakage of a 0.5 mm 2H type pencil lead - 110 signals were generated for each of the 40 areas - Signal generation occurred at a random location within each specific area Minimizing Interference - The vessel was suspended using cables attached to an aluminum frame structure - This suspension was to minimize reflections and dissipation of acoustic waves Signal Acquisition System - The four AE transducers were connected to a Vallen AMSY-5 measurement system. - Recorded signals were exported to HDF files, Matlab files and stored in Vallen raw files for post-processing - Vallen Systeme software acquisition settings: -- Sampling rate: 2 MHz. -- Sample length: 4096 samples. -- Pre-trigger sample count: 500 samples. -- Amplitude threshold: 30.6 dB.
Institutions
- Universita degli Studi di Bologna
- Ceske Vysoke Uceni Technicke v Praze Fakulta strojni