Stochastic thermal buckling investigation of optimized variable stiffness composite laminates incurred internal flaws
Topic(s) : Material and Structural Behavior - Simulation & Testing
Co-authors :
Prateek CHANDRAKAR (INDIA), Narayan SHARMA , Dipak Kumar MAITIAbstract :
The automated tow placement technology enables the production of modern composites featuring curved fibers named variable stiffness composite laminates (VSCL). The structural ability of VSCL renders it appropriate for high-speed launch and reentry vehicles undergoing aerodynamic heating. The structural integrity of supersonic and hypersonic vehicles may be compromised by material degradation at high temperatures, reducing safety and longevity. In the present study, the VSCL plates are exposed to thermally varying environments, and the buckling performance is evaluated concerning optimized curvature profiles of eight distinct materials [1]. The operating circumstances of composites may be responsible for defect formation; this study considers the incursion of defects, leading to the undesired mechanical response of structures. The finite element model (FEM) is built on the improved first-order shear deformation theory [2], which incorporates an effective shear correction factor to offer parabolic shear stress distribution for the satisfaction of free edge transverse shear stress conditions. The FEM utilizes a material stiffness reduction-based damage model [3] and estimates the critical buckling temperature of VSCL plates undergoing three levels of damage severity (Λ=3,4,5) to the two most influential locations (corner and mid). Composite manufacturing is an intricate task, and this complexity may lead to variability in the composite properties; hence, this study introduces the randomness to various composite characteristics to estimate the variability in the buckling performance of VSCL. To address the significant computational expenses involved with traditional Monte-Carlo simulations (MCS), an RBFN-based surrogate model is implemented. As per the convergence study, the developed surrogate model is capable of predicting the stochastic outcomes with significantly less (≈95%) data sets than MCS. Moreover, estimating reliability characteristics is critical in developing a safe and reliable design. The developed surrogate model is coupled with a probabilistic failure estimation algorithm to provide the failure probability of VSCL plates to various scenarios.
The investigation is intended to reveal the resistance capability of various optimized VSCL plates towards damage and the severity of damage location depending on multiple circumstantial conditions (Fig.1). Furthermore, the sensitivity to uncertain composite properties is a different aspect; despite incorporating similar levels of randomization in the characteristics of eight VSCL, the uncertainty in buckling strength can be substantially reduced by using appropriate composite materials. The damages at the mid location of VSCL plates happened to be more severe as it caused the most significant reduction in the critical buckling temperature and the early failure of VSCL plates; additionally, a lower range of failure probabilities indicates complete failure in a shorter temperature span (Fig.2).
The investigation is intended to reveal the resistance capability of various optimized VSCL plates towards damage and the severity of damage location depending on multiple circumstantial conditions (Fig.1). Furthermore, the sensitivity to uncertain composite properties is a different aspect; despite incorporating similar levels of randomization in the characteristics of eight VSCL, the uncertainty in buckling strength can be substantially reduced by using appropriate composite materials. The damages at the mid location of VSCL plates happened to be more severe as it caused the most significant reduction in the critical buckling temperature and the early failure of VSCL plates; additionally, a lower range of failure probabilities indicates complete failure in a shorter temperature span (Fig.2).