Signal Measurement of Projectile Penetration Overload Based on Charge Sensor

Debiao Zhang, Jie Li, Xiaokai Wei, Kaiqiang Feng, Yu Wang, Jihe ZhaoIEEE Access 2019 · DOI

Abstract

In modern warfare, missile warheads are often required to explode after penetrating deep into a target. Consequently, measuring the mechanical signals during projectile penetration is critical for design and evaluation. This paper investigates the dynamic design principles of a projectile penetration overload test system based on a charge sensor. We design a measurement circuit tailored to the characteristics of steep-front signals and high-frequency structural impulses. The system incorporates a charge sensor, high-impedance amplifier, and a specialized filtering stage. To evaluate the system objectively, we utilize dynamic uncertainty estimation rather than traditional frequency domain metrics. Furthermore, an improved de-noising algorithm based on Ensemble Empirical Mode Decomposition (EEMD) is proposed to enhance measurement accuracy. Experimental results demonstrate the effectiveness of the proposed system, achieving a peak error of only 1.4% compared to laser interferometry.

Introduction

Accurately capturing the "overload" signal—the extreme deceleration and mechanical stress experienced by a projectile upon impact—is a significant challenge in ballistics engineering. These signals are characterized by a "steep front," representing a near-instantaneous rise in force, followed by complex high-frequency oscillations caused by the projectile's internal structural response. Traditional measurement systems often struggle with these transients, either introducing electronic oscillations or failing to distinguish between the actual penetration data and structural noise. This research addresses these issues through a combined hardware and software approach.

System Design

The hardware architecture is centered on a high-precision charge sensor. The core of the measurement circuit is a high-impedance charge amplifier utilizing the OPA129 operational amplifier to minimize signal loss and noise. To ensure robust signal integrity, the output is converted into a fully differential format before being processed by a high-speed Analog-to-Digital Converter (ADC).

Circuit diagram of the charge sensor and high-impedance amplifier system.

A critical design choice involves the selection of low-pass filters. While Butterworth and Chebyshev filters are common in many applications, they often introduce significant ringing or phase distortion in response to step-like impact signals. This study demonstrates that Bessel filters are superior for this application. Due to their linear phase characteristics, Bessel filters provide the best step response with minimal oscillation and the fastest stabilization time, preserving the integrity of the steep-front impact pulse.

Signal Processing and De-noising

To further refine the captured data, we implemented an adaptive de-noising algorithm based on Ensemble Empirical Mode Decomposition (EEMD). Unlike static Fourier-based filters, EEMD decomposes the complex, non-stationary impact signal into a series of Intrinsic Mode Functions (IMFs) based on the signal's inherent timescales. By selectively filtering these IMFs, the system can effectively strip away high-frequency structural vibrations while maintaining the underlying deceleration curve.

We also propose moving beyond traditional evaluation metrics like amplitude-frequency "flatness." Instead, we use dynamic uncertainty estimation (root-mean-square error) as a more objective standard for evaluating how accurately the test system reproduces the transient impact event.

Experimental Results

The system was validated using a Hopkinson bar test setup, where a 16mm diameter titanium alloy bar simulated high-velocity impact. The results were benchmarked against a high-precision laser interferometer.

The experimental data showed exceptional accuracy:

  • Peak Acceleration: The system measured a peak of 33,320g, compared to the 33,800g ground truth from the laser interferometer—a peak error of just 1.4%.
  • Pulse Width: The measured pulse width of 18.9μs resulted in a 1.5% error relative to the benchmark.

Visual analysis of the processed signals confirms that the EEMD algorithm successfully smoothed the massive oscillations caused by the installation structure, providing a clean and usable deceleration profile for engineering analysis.

Conclusion

This research provides a robust framework for measuring extreme mechanical transients in ballistic applications. By combining high-impedance hardware design, phase-linear Bessel filtering, and adaptive EEMD de-noising, we achieved high-fidelity measurements of projectile penetration overload. Future work will focus on miniaturizing the system for field-fire tests and optimizing the EEMD algorithm for real-time processing.

Summary updated

BibTeX
@article{zhang2019signal,
  author = {Debiao Zhang and Jie Li and Xiaokai Wei and Kaiqiang Feng and Yu Wang and Jihe Zhao},
  title = {{Signal Measurement of Projectile Penetration Overload Based on Charge Sensor}},
  journal = {IEEE Access},
  year = {2019},
  doi = {10.1109/access.2019.2958387},
}
Figure