PSI - Issue 77

2

Kumar C. Jois et al. / Procedia Structural Integrity 77 (2026) 405–412 Jois, et al./ Structural Integrity Procedia 00 (2026) 000–000

406

Despite these advantages, the manufacturing of CPVs inevitably introduces defects, even under strict quality control conditions. Among these, voids—air pockets or trapped volatiles within the matrix—are among the most common and detrimental. They compromise load transfer between fiber and matrix, degrade interlaminar properties, and accelerate failure initiation. Accurately accounting for such defects remains a key challenge, as traditional design approaches often assume ideal material behavior and thus yield optimistic performance predictions. Filament winding is one of the most common manufacturing techniques for cylindrical composite pressure vessels, enabling precise fiber placement and efficient load distribution. However, variations in process parameters—such as winding tension, resin viscosity, curing temperature, and winding pattern—affect void formation and distribution [1]. These factors are often interdependent, complicating the assessment of individual influences. In this study, we focus exclusively on the influence of voids on the structural performance of filament-wound composite pressure vessels. Previous studies have established that voids can significantly reduce mechanical properties of fiber-reinforced polymers (FRPs) [2–4]. Interlaminar voids are particularly critical, as they reduce fracture toughness and promote delamination [5]. Experimental data highlight substantial strength losses: Wisnom et al. [6] reported up to 31% reduction in interlaminar shear strength (ILSS) for voids of approximately 3 mm, while Asp and Brandt [7] observed mode-dependent delamination behavior in carbon–epoxy laminates. Finite element analyses confirm that void morphology—shape, aspect ratio, and orientation—strongly influences local stress concentration, with flat or elongated voids causing greater reductions in out-of-plane stiffness and fracture resistance [4]. Although fiber dominated longitudinal properties remain relatively stable, transverse and interlaminar properties degrade rapidly once void content exceeds roughly 2% [8]. Recent investigations emphasize that void clustering and architecture also govern damage initiation and fatigue life. Large voids (>400 µm) act as preferential crack initiation sites [9], and fatigue life can drop by several orders of magnitude as void content increases from 1.6% to 3.1% [5]. These results suggest that fatigue behavior is substantially more sensitive to void content than static strength, underscoring the need for microstructurally informed design approaches. To address these challenges, the present study employs computed tomography (CT) to characterize void morphology and layer-dependent distribution in filament-wound composite pressure vessels. The experimentally derived statistical void data are incorporated into finite element (FE) models to simulate the influence of void content and clustering on burst performance. A sensitivity analysis across random void realizations is conducted to quantify performance variability and identify dominant parameters influencing failure. The outcomes of this work aim to provide a clearer understanding of how voids affect the structural integrity and reliability of composite pressure vessels. Moreover, the findings support the development of performance-based acceptance criteria, enabling a more rational balance between manufacturing cost, safety, and structural efficiency in next-generation composite vessel design.

2. Experimental method and numerical model 2.1. Void characterization using computer tomography

Computed Tomography (CT) is a non-destructive imaging technique widely used for characterizing internal defects such as voids and delaminations in composite materials. In CT scanning, a specimen is irradiated with X-rays from multiple angles, and the transmitted intensities are recorded by a detector. The resulting projections are then reconstructed into a three-dimensional volumetric image using algorithms such as filtered back-projection or iterative reconstruction. Differences in X-ray attenuation between fiber, matrix, and air allow for the identification and quantification of voids. This enables precise measurement of void volume fraction, morphology, and spatial