Push forward the understanding of the rare-event stochastic failure of advanced aerospace structures and materials;
Develop super reliable lightweight composite structures for extreme and uncertain space environments.
Most engineering structures (rockets, satellites, airframes, micro-electronics, etc.) shall be designed for failure probability at least not exceeding one in a million per lifetime. Due to the extremely low required failure risk and catastrophic consequences, the unexpected failures of crucial aerospace structures are precisely what are referred to as black swan events. Understanding the statistics of the rare-event failures could fundamentally fuel the development of super-reliable lightweight structures, which could further improve the reusability of launch vehicles, satellites, and airplanes, ultimately decreasing the cost of entering space and air.
Characterizing the lower tail of strength distributions is essential for reliability assessment of advanced materials, such as additively manufactured (AM) composites. Yet conventional single-specimen tensile testing is inefficient for probing rare-event failures. To address this challenge, we developed a STiffness-Augmented Serial High-throughput (STASH) tensile testing framework that leverages automated 3D printing with serial sample coupling to encode the material strength distribution directly into the system-level load-displacement response. By decoding this global response, the method enables large-scale statistical characterization of strength from a single test for brittle and quasi-brittle materials. A key feature of this approach is its ability to mitigate hysteresis caused by nonuniform cyclic loading, which arises inevitably from the serial coupling of samples. This is achieved by implementing controlled partial unloading after each individual sample failure, rather than fully unloading the system. By directly sampling the weakest specimens, the STASH test substantially enhances efficiency in resolving lower-tail failure probabilities. The framework is scalable to hundreds of samples per run and provides a promising route toward rapid probabilistic failure characterization in quasi-brittle materials.
It is well known that the buckling of thin-walled shells, especially the cylindrical shells, is highly stochastic and imperfection sensitive - a slight geometric imperfection could lead to a substantially lower load capacity compared to the theoretical buckling load of a perfect shell. While the cause of this phenomenon has become qualitatively clear, the accurate assessment of a shell's load capacity is still challenging in that the imperfections are random, and no two shells can be made exactly the same. To address imperfection sensitivity in the design of thin-walled shells, the concept of buckling knockdown factor (KDF) is widely adopted. However, the KDF is an empirical lower bound with unclear underlying physics. Depending solely on this empirical safety factor can be either risky or uneconomical and can also constrain the designer's creative possibilities.
This research seeks to integrate nonlinear mechanics, additive manufacturing, high-throughput testing, and uncertainty quantification to comprehensively analyze the statistical characteristics of imperfection sensitivity in thin-walled shells, with special attention given to the lower tail region of the buckling load distribution. It will pave the road for the development of a scientifically justified probabilistic buckling knockdown factor for shells, ultimately enabling the design of significantly safer lightweight structures.
Architected materials (usually additively manufactured) promise transformative weight reduction and multifunctional performance in critical applications from turbine blades to morphing wings and medical stents. Although most architected materials are based on periodic unit cells, they remain susceptible to random manufacturing defects. Building on this challenge, our prior work on fishnet statistics has revealed an unexpected opportunity: imperfections can be deliberately harnessed to promote distributed damage, delay crack localization, and ultimately enhance resilience. These findings point toward a transformative paradigm in which engineered disorder is not a liability but a design lever, opening the door to architected materials with exceptional combinations of strength and reliability. This project aims to unravel the failure risks of 3D-printed lattices to enable architected materials that are not just lightweight and strong, but also super reliable.
This research studies the snap-through buckling that occurs ahead of the coiled region in thin, linear-elastic, isotropic coilable cylindrical shells with a sudden change in the thickness of the shell cross section. The study is focused on Triangular Rollable And Collapsible (TRAC) booms wided used for space applications. It is shown that coiling of these shells leads to longitudinal compression of the inner flange mid-surface, which in turn leads to the formation of a buckle in the transition region between the fully coiled and fully deployed parts of the inner flange. This buckle grows to reach a steady-state configuration and is then pushed along the shell without changing its shape when the shell is coiled. Although the present study has focused on a specific TRAC boom cross-section, the insight that has been gained is applicable to coilable shell structures in general.