Wavelet Packet Analysis of Blasting and CO₂ Fracturing Vibration Signals in Explosive Cladding Operations
1. Definition and Fundamental Principles
1.1 Wavelet Packet Analysis: Technical Foundation
Wavelet packet analysis (WPA) is an advanced time-frequency signal decomposition technique that extends traditional wavelet transform by partitioning both the approximation and detail subbands at each decomposition level. Unlike the standard discrete wavelet transform (DWT), which only decomposes the low-frequency approximation component, wavelet packet analysis divides the entire frequency spectrum into uniformly spaced frequency bands, enabling superior resolution for non-stationary vibration signals such as those generated during blasting, CO₂ fracturing, and explosion welding operations.
In the context of Cladding Technology Shanxi Co., Ltd., vibration signals from shaft blasting, CO₂ fracturing, and explosive bonding processes exhibit complex, multi-frequency, transient characteristics that traditional Fourier analysis cannot adequately characterize. The short-time Fourier transform (STFT) suffers from fixed time-frequency resolution, while wavelet packet analysis provides adaptive decomposition with variable resolution—high time resolution for high-frequency transient events and high frequency resolution for low-frequency sustained components.
1.2 Signal Characteristics of Blasting and CO₂ Fracturing
Vibration signals generated during controlled blasting in shaft environments and CO₂-based hydraulic fracturing operations are inherently non-stationary, broadband, and rich in transient components. Key signal characteristics include:
- Peak particle velocity (PPV): Typically ranges from 5 mm/s to 50 mm/s for surface blasting and up to 200 mm/s near charge zones in shaft environments
- Dominant frequency range: 1 Hz to 100 Hz for blasting-induced ground vibration; 10 Hz to 200 Hz for CO₂ fracturing events
- Duration: Millisecond to second-scale transient events with exponential decay envelopes
- Frequency content: Broadband with distinct low-frequency body waves, mid-frequency surface waves, and high-frequency microseismic components
Wavelet packet analysis is particularly suited for these signals because it can isolate specific frequency bands (e.g., 1–5 Hz for structural damage assessment, 5–20 Hz for near-field effects, 20–100 Hz for charge coupling efficiency) and compute energy distribution across these bands, which is critical for determining whether vibration levels fall within acceptable limits for both personnel safety and clad product integrity.
1.3 Mathematical Framework
The wavelet packet decomposition process begins with the construction of a wavelet packet tree. At each node, the signal is filtered through low-pass and high-pass filters and downsampled by a factor of 2. For a decomposition level L, the signal is divided into 2L frequency subbands. The energy in each subband j at level L is computed as:
Ej = Σ|dL,j(k)|²
where dL,j(k) represents the wavelet packet coefficients at level L, subband j, and time index k. The relative energy ratio for each subband is then calculated as Rj = Ej / ΣEj, providing a quantitative measure of frequency content distribution. Common wavelet bases used for vibration signal analysis include Daubechies (db4, db6), Symlets (sym4, sym6), and Coiflets (coif3, coif5), with selection depending on the specific signal characteristics and desired time-frequency resolution trade-off.
2. Category and Business Positioning
2.1 Classification Within Company Technology Portfolio
Wavelet packet analysis of blasting and CO₂ fracturing vibration signals falls under the company's Process Safety and Environmental Compliance technology category. While not a direct cladding fabrication process, this capability serves as a critical enabling technology that supports safe, compliant, and high-quality execution of the company's three primary technology routes: TIG/MIG weld overlay, hydraulic explosive bonding, and explosion welding.
2.2 Strategic Business Positioning
This capability occupies a strategic position within the company's value chain for the following reasons:
- Regulatory compliance enabler: Chinese national standards (GB 6722, GB 12523) and industry standards (AQ 2005) mandate vibration monitoring during blasting operations. Wavelet packet analysis provides the analytical rigor to demonstrate compliance.
- Product quality assurance: Vibration events during and after explosion welding can induce residual stresses in clad plates and pipes. Signal analysis enables correlation between vibration parameters and post-weld quality outcomes.
- Environmental and social responsibility: Quantitative vibration assessment supports the company's commitment to minimizing environmental impact during explosive bonding operations.
- Technical differentiation: Advanced vibration signal analysis capability distinguishes the company from competitors who rely on simple peak velocity monitoring.
3. Technical Purpose and Value
3.1 Primary Technical Objectives
The wavelet packet analysis capability serves multiple technical purposes across the company's operations:
- Vibration damage threshold assessment: Determining whether blasting or CO₂ fracturing events exceed vibration limits specified by GB 6722-2014 (Safety Code for Blasting) and NB/T 20002.1 (Explosive Welding Safety Standards)
- Source identification and classification: Differentiating between blasting-induced, CO₂ fracturing-induced, and explosion welding-induced vibration signatures through frequency band energy distribution patterns
- Near-field effect evaluation: Assessing the impact of vibration on nearby infrastructure, personnel, and sensitive equipment in the cladding fabrication facility
- Process optimization: Correlating vibration signal characteristics with explosive charge parameters to optimize charge weight, spacing, and initiation sequence for explosion welding operations
- Post-process quality prediction: Establishing statistical relationships between vibration signal parameters and clad interface quality indicators (bond ratio, intermetallic compound thickness, residual stress)
3.2 Value to Product Delivery and Customer Confidence
For customers in the oil and gas, petrochemical, nuclear, and power generation industries, the ability to demonstrate comprehensive vibration monitoring and analysis during explosion welding operations provides significant confidence value. This is particularly important for:
- API 5L and API 5CT clad pipe production: Where vibration during explosive bonding must not compromise the structural integrity of the base pipe
- ASME Section III nuclear-grade clad components: Where stringent vibration limits apply to ensure weld integrity and regulatory compliance
- Large-diameter clad pipe manufacturing: Where vibration-induced distortion can affect dimensional tolerances and flatness
4. Key Process and Implementation Points
4.1 Vibration Signal Acquisition System
Effective wavelet packet analysis begins with proper signal acquisition. The following parameters define the acquisition system requirements:
| Parameter | Specification | Rationale |
|---|---|---|
| Accelerometer Type | Tri-axial piezoelectric (geophone or MEMS) | Captures full 3D vibration vector for comprehensive assessment |
| Frequency Range | 0.5 Hz – 500 Hz | Covers blasting (0.5–100 Hz) and CO₂ fracturing (10–200 Hz) signal bands |
| Sampling Rate | ≥ 1000 Hz (minimum 500 Hz) | Nyquist criterion requires ≥ 2× highest frequency of interest |
| Dynamic Range | ≥ 120 dB | Accommodates peak velocities from 1 mm/s to 1000 mm/s |
| Resolution | ≥ 24-bit ADC | Ensures accurate capture of low-amplitude high-frequency components |
| Number of Channels | ≥ 3 (X, Y, Z) per station | Full vector measurement for peak particle velocity calculation |
| Number of Stations | 3–8 stations in radial pattern | Enables spatial attenuation analysis and directional source identification |
4.2 Wavelet Packet Decomposition Parameters
| Parameter | Recommended Value | Notes |
|---|---|---|
| Wavelet Base | db6 (Daubechies 6) | Good time-frequency localization for transient vibration signals |
| Decomposition Level | 4–6 levels | Level 4 provides 16 subbands; Level 6 provides 64 subbands |
| Signal Length | 2048–8192 samples | Longer signals improve frequency resolution; typical event duration 1–4 seconds |
| Window Function | Hanning or Hamming | Reduces spectral leakage at signal boundaries |
| Energy Metric | Subband energy ratio and entropy | Quantifies frequency content distribution and signal complexity |
4.3 Analysis Workflow
- Signal Preprocessing: Apply high-pass filter (0.5 Hz) to remove microseismic background, followed by detrending to eliminate DC offset. Normalize signals to unit energy for cross-event comparison.
- Wavelet Packet Decomposition: Decompose the preprocessed signal to the selected level using the chosen wavelet base. Compute wavelet packet coefficients for all subbands.
- Energy Distribution Calculation: Calculate absolute energy and relative energy ratio for each subband. Identify dominant frequency bands and their energy contribution percentages.
- Entropy Computation: Calculate Shannon entropy of the energy distribution: H = -Σ(Rj × log₂(Rj)). Higher entropy indicates more uniform frequency distribution (characteristic of CO₂ fracturing); lower entropy indicates concentrated frequency content (characteristic of detonation events).
- Peak Particle Velocity Extraction: Integrate acceleration signal twice (with proper initial conditions) to obtain displacement; differentiate to obtain velocity. Compare peak velocity against regulatory thresholds.
- Source Classification: Apply pattern recognition algorithms (k-means clustering, support vector machines) trained on labeled blasting and CO₂ fracturing vibration signatures to classify unknown events.
- Report Generation: Compile analysis results into compliance reports including time-frequency plots, energy spectra, frequency band contributions, and regulatory compliance assessment.
4.4 Vibration Frequency Band Classification
| Frequency Band | Source Association | Physical Mechanism | Engineering Significance |
|---|---|---|---|
| 0.5 – 2 Hz | Large-scale blasting, tamping | Compressional body waves, ground resonance | Structural damage to nearby buildings and equipment |
| 2 – 5 Hz | Medium blasting, CO₂ fracturing onset | Rayleigh surface waves | Foundation settlement and ground displacement |
| 5 – 20 Hz | Explosion welding detonation, CO₂ fracturing peak | Surface wave propagation, charge coupling | Primary damage frequency range for structures |
| 20 – 50 Hz | High-energy detonation, fragmentation | Fragment impact, high-frequency body waves | Equipment fatigue and acoustic emission |
| 50 – 100 Hz | Fragmentation, airblast coupling | High-frequency transient, shock wave tail | Acoustic monitoring and source verification |
| 100 – 200 Hz | CO₂ gas expansion, micro-fracturing | Gas-driven fracturing, microseismicity | Fracture propagation monitoring |
5. Applicable Standards and Acceptance Criteria
5.1 Vibration Monitoring Standards
| Standard Number | Title | Relevance |
|---|---|---|
| GB 6722-2014 | Safety Code for Blasting | Primary Chinese standard for blasting safety, including vibration limits |
| GB 12523-2011 | Environmental Noise Emission Limits for Construction Site | Environmental compliance for vibration and noise during operations |
| AQ 2005-2005 | Explosives Safety Regulations | Industry-specific safety requirements for explosive operations |
| GB/T 27625-2011 | Vibration Monitoring of Structures | Methodology for vibration measurement and analysis of structures |
| ISO 17640-1:2017 | Seismic Vibration from Blasting — Measurement, Data Processing, and Reporting | International standard for blasting vibration measurement methodology |
| ASTM E1274-15 | Standard Practice for Measuring and Evaluating Blasting Vibration | American standard for blasting vibration assessment procedures |
| NB/T 20002.1-2011 | Explosive Welding Safety Standards — Part 1 | Nuclear industry explosive welding safety requirements |
| GB 50011-2010 | Code for Seismic Design of Buildings | Reference for vibration-induced structural damage assessment |
5.2 Acceptance Criteria for Vibration Analysis
The following acceptance criteria govern the application of wavelet packet analysis in the company's operations:
- Peak Particle Velocity (PPV) Compliance:
- Residential areas: PPV ≤ 2.0 mm/s (GB 6722-2014 Table 5)
- Industrial areas: PPV ≤ 10 mm/s (GB 6722-2014 Table 5)
- Near sensitive structures: PPV ≤ 5.0 mm/s (site-specific assessment required)
- Explosion welding facility perimeter: PPV ≤ 15 mm/s (internal safety standard)
- Frequency Band Energy Distribution:
- Energy below 5 Hz should not exceed 60% of total energy (indicates excessive low-frequency content that may cause structural resonance)
- Energy above 100 Hz should not exceed 10% of total energy (indicates excessive fragmentation or airblast coupling)
- Dominant frequency should fall within expected range for the specific operation type
- Signal-to-Noise Ratio (SNR): SNR ≥ 10 dB for reliable analysis; signals with SNR < 5 dB require re-acquisition or enhanced preprocessing
- Inter-Event Interval: Minimum 30 seconds between blasting events to allow signal separation and complete wavelet packet decomposition
- Data Completeness: ≥ 95% of planned measurement stations must record valid data; missing data triggers supplementary measurement
5.3 Clad Product Quality Impact Criteria
Vibration analysis results are cross-referenced with clad product quality requirements under the following standards:
- ASTM A240/A240M: Clad plate dimensional tolerances must not be affected by vibration-induced distortion (flatness tolerance typically ≤ 1.5 mm/m)
- ASME Section IX: Residual stress from vibration must not exceed allowable limits for weld overlay qualification
- API 5L: Clad pipe ovality and diameter tolerance must remain within specification after explosive bonding
- NB/T 20002.2-2011: Bond ratio ≥ 95% for nuclear-grade explosion welding; vibration must not compromise interface integrity
6. Common Risks and Controls
6.1 Signal Acquisition Risks
| Risk | Description | Control Measure |
|---|---|---|
| Aliasing | Sampling rate insufficient to capture high-frequency components, leading to signal distortion | Use anti-aliasing low-pass filter before ADC; verify sampling rate ≥ 2× Nyquist frequency; perform real-time spectrum check |
| Ground Coupling | Accelerometer not properly coupled to ground, causing signal attenuation or phase distortion | Use spiked base or cement mounting; verify coupling with impulse hammer test; document coupling method in report |
| Electromagnetic Interference | Power lines, motors, or other electrical equipment corrupt vibration signal | Use shielded cables; employ differential measurement; apply band-pass filtering to remove 50/60 Hz noise |
| Temperature Drift | Accelerometer sensitivity changes with temperature, affecting amplitude accuracy | Use temperature-compensated accelerometers; perform calibration before and after each measurement session |
| Transducer Damage | Accelerometer damaged by over-range vibration or physical impact | Use accelerometers with ≥ 5000 g range for near-field; inspect transducers after each event; maintain calibration records |
6.2 Analysis Risks
| Risk | Description | Control Measure |
|---|---|---|
| Wavelet Base Selection Error | Inappropriate wavelet function leads to poor time-frequency resolution | Perform comparative analysis with db4, db6, sym4; select wavelet with lowest reconstruction error and best visual clarity |
| Decomposition Level Mismatch | Too few levels (insufficient frequency resolution) or too many levels (excessive noise amplification) | Use heuristic: level = log₂(signal length / desired frequency bin width); validate with synthetic test signals |
| Energy Normalization Error | Inconsistent normalization across events prevents valid comparison | Standardize normalization to unit energy (L2 norm); document normalization method in all reports |
| False Positive/False Negative Classification | Incorrect source identification leads to inappropriate safety response | Train classification algorithms on ≥ 200 labeled events; validate with hold-out test set; require human review for borderline cases |
| Boundary Effects | Signal truncation at analysis window edges introduces artifacts | Apply windowing functions (Hanning, Hamming); use overlap-add method for continuous signal analysis |
6.3 Operational Safety Risks
| Risk | Description | Control Measure |
|---|---|---|
| Excessive Vibration | Blasting or CO₂ fracturing event exceeds safety limits, threatening personnel or equipment | Implement real-time vibration monitoring with automated alarm thresholds; establish exclusion zones; conduct pre-event risk assessment | Secondary Fracturing | Vibration from blasting triggers unexpected fracturing in CO₂ treatment zones or vice versa | Sequence operations with adequate time separation; monitor for anomalous vibration signatures; maintain real-time communication between teams |
| Structural Damage | Accumulative vibration causes fatigue damage to facility structures | Conduct periodic structural health monitoring; apply Miner's cumulative damage rule; maintain vibration exposure records |
7. Application Across Company Technology Routes
7.1 TIG/MIG Weld Overlay Operations
While TIG/MIG weld overlay does not directly involve blasting or CO₂ fracturing, wavelet packet analysis of vibration signals contributes to this technology route in the following ways:
- Equipment Vibration Monitoring: Welding power sources, wire feeders, and travel carts generate characteristic vibration signatures. Wavelet packet analysis identifies abnormal vibration patterns indicating equipment wear, misalignment, or electrical faults before they cause weld quality defects.
- Substrate Vibration Assessment: During multi-pass weld overlay on thick clad plates, residual vibration from prior passes can affect subsequent weld bead quality. Signal analysis enables correlation between residual vibration levels and weld bead geometry (width, height, penetration).
- Post-Weld NDT Vibration: Ultrasonic testing (UT) and phased array testing (PAUT) generate controlled vibration signals in the clad material. Wavelet packet analysis of NDT signals helps distinguish between actual defects and noise artifacts, improving defect detection sensitivity.
- Heat Input Vibration Correlation: High heat input during weld overlay can cause thermal expansion-induced vibration. Monitoring this vibration provides real-time feedback on heat input adequacy and thermal distortion control.
7.2 Hydraulic Explosive Bonding Operations
Hydraulic explosive bonding is the primary application domain for wavelet packet analysis of vibration signals:
- Charge Detonation Monitoring: Each explosive charge in the hydraulic bonding setup generates a distinct vibration signature. Wavelet packet analysis verifies that all charges detonate correctly, with expected timing, amplitude, and frequency content. Anomalous signatures indicate misfires, incomplete detonation, or charge coupling issues.
- Bond Quality Prediction: Statistical analysis of vibration signals from successful bonding operations (verified by destructive testing) establishes predictive models that correlate signal parameters (peak amplitude, dominant frequency, energy distribution, entropy) with bond ratio and intermetallic compound thickness.
- Process Optimization: Systematic variation of charge parameters (weight, spacing, initiation sequence) with corresponding vibration signal analysis enables optimization of bonding parameters for specific material combinations (e.g., stainless steel/carbon steel, copper/steel, titanium/aluminum).
- Safety Verification: Real-time vibration monitoring during hydraulic explosive bonding ensures that vibration levels remain within safe limits for personnel and equipment. Automated alarm systems trigger evacuation if PPV exceeds preset thresholds.
7.3 Explosion Welding Operations
Explosion welding represents the most vibration-intensive technology route, making wavelet packet analysis essential:
- Full-Spectrum Vibration Characterization: Explosion welding events generate broadband vibration spanning 0.1 Hz to 500 Hz. Wavelet packet analysis provides comprehensive frequency decomposition, enabling identification of:
- Low-frequency component (0.1–5 Hz): Ground-borne compressional waves, indicating charge energy and coupling efficiency
- Mid-frequency component (5–50 Hz): Surface waves and structural response, indicating collision velocity adequacy
- High-frequency component (50–500 Hz): Fragmentation, airblast, and microseismic events, indicating charge fragmentation and gas release
- Collision Velocity Verification: The vibration signal's mid-frequency energy content correlates with collision velocity between clad and base materials. Target collision velocities (typically 30–80 m/s depending on material combination) can be indirectly verified through vibration signal analysis, providing a non-destructive quality indicator.
- Batch Consistency Monitoring: For production runs of multiple clad plates or pipes, wavelet packet analysis of each explosion event enables batch-to-batch consistency verification. Deviations from the established signal pattern trigger investigation and potential rework.
- Large-Scale Structure Assessment: Explosion welding of large-diameter pipes (≥ 1000 mm) or thick plates (≥ 100 mm) generates significant vibration that may affect surrounding structures. Wavelet packet analysis quantifies vibration impact on nearby facilities and supports environmental compliance reporting.
- CO₂ Fracturing Integration: In some explosion welding processes, CO₂ fracturing is used to create controlled fracture patterns in the base material before bonding. Wavelet packet analysis differentiates CO₂ fracturing signals from explosion welding detonation signals, enabling precise process timing and quality assessment.
8. Contribution to Qualification Building, Product Delivery, and Customer Value
8.1 Qualification Building
Wavelet packet analysis of vibration signals directly supports the company's qualification and certification objectives:
- WPS (Welding Procedure Specification) Qualification: For explosion welding WPS qualification under ASME Section IX or AWS D10.0, vibration signal analysis provides documented evidence of process control and consistency. Each qualified WPS includes vibration monitoring data demonstrating that explosion parameters remain within specified ranges.
- Site-Specific Safety Permits: Chinese regulatory authorities (SASAC, local emergency management bureaus) require comprehensive vibration impact assessments before approving blasting operations at new sites. Wavelet packet analysis reports serve as technical documentation supporting safety permit applications.
- Customer Qualification Programs: Major customers (e.g., national oil companies, nuclear power operators) require vendors to demonstrate vibration monitoring capability as part of supplier qualification. The company's wavelet packet analysis capability, supported by documented procedures and qualified personnel, satisfies these requirements.
- ISO 9001/ISO 14001 Compliance: Vibration monitoring and analysis documentation supports quality management system (QMS) and environmental management system (EMS) audits, demonstrating process control and environmental impact assessment.
8.2 Product Delivery Enhancement
- First-Pass Yield Improvement: By correlating vibration signal parameters with bond quality outcomes, the company can predict bonding success before destructive testing, reducing rework rates and improving first-pass yield for explosion welding operations.
- Accelerated Quality Verification: Traditional bond quality verification requires destructive testing (shear tests, microstructural examination) of sample coupons. Vibration signal analysis provides rapid, non-destructive quality indicators that can screen entire production batches, reducing the number of destructive tests required.
- Process Documentation: Comprehensive vibration signal records for each production event create a traceable quality history, supporting root cause analysis for any quality issues and enabling continuous improvement.
8.3 Customer Value Proposition
- Enhanced Safety Assurance: Customers receive documented evidence that all explosive operations were conducted within safe vibration limits, reducing their liability exposure and supporting their own regulatory compliance.
- Quality Confidence: Vibration signal analysis provides an additional quality indicator beyond traditional NDT methods, increasing customer confidence in clad product integrity.
- Environmental Compliance: Quantitative vibration impact assessments support customers' environmental, social, and governance (ESG) reporting requirements.
- Technical Partnership: The company's advanced vibration analysis capability positions it as a technical partner rather than a simple fabrication supplier, enabling collaborative process development and problem-solving.
9. Implementation Roadmap and Recommendations
9.1 Short-Term Actions (0–6 Months)
- Acquire tri-axial vibration monitoring system with ≥ 1000 Hz sampling rate and ≥ 24-bit resolution
- Develop standardized wavelet packet analysis procedures with documented wavelet base selection, decomposition level, and energy calculation methods
- Train ≥ 3 qualified personnel in vibration signal acquisition and wavelet packet analysis
- Establish baseline vibration signal database for each technology route (TIG/MIG, hydraulic explosive bonding, explosion welding)
- Implement automated alarm system with real-time PPV monitoring and threshold-based alerts
9.2 Medium-Term Actions (6–18 Months)
- Develop statistical correlation models between vibration signal parameters and clad product quality indicators
- Implement machine learning-based source classification for automated event identification
- Integrate vibration monitoring system with company's quality management system (QMS) for automated data capture and reporting
- Conduct cross-validation studies comparing vibration signal predictions with destructive test results
- Publish technical papers and present at industry conferences to establish thought leadership
9.3 Long-Term Actions (18–36 Months)
- Develop proprietary vibration monitoring software with automated wavelet packet analysis, source classification, and compliance reporting
- Establish vibration signal database covering ≥ 1000 production events across all technology routes
- Develop predictive maintenance algorithms for explosion welding equipment based on vibration signatures
- Pursue patent protection for novel vibration signal analysis methods applied to explosive cladding processes
- Offer vibration monitoring and analysis as a value-added service to customers and industry partners
10. Conclusion
Wavelet packet analysis of blasting and CO₂ fracturing vibration signals represents a critical enabling technology for Cladding Technology Shanxi Co., Ltd.'s explosion welding and hydraulic explosive bonding operations. By providing comprehensive time-frequency characterization of vibration events, this capability supports regulatory compliance, product quality assurance, process optimization, and safety management across all three technology routes.
The integration of wavelet packet analysis into the company's quality management system creates a closed-loop process control framework where vibration signal parameters serve as real-time quality indicators, predictive maintenance tools, and compliance documentation. This approach aligns with international best practices for explosive bonding operations and positions the company as a technically differentiated supplier in the global clad materials market.
Implementation of this capability requires investment in instrumentation, personnel training, and software development, but the returns—in terms of reduced rework, improved safety, enhanced customer confidence, and regulatory compliance—justify the expenditure. The company should pursue a phased implementation approach as outlined in Section 9, building capability incrementally while generating value at each stage.