Research
Fibers are in every aspect of our lives, and we rely on them in a wide variety of applications, from flexible body armor, to composite overwrapped pressure vessels that are being used everywhere from our vehicles to outer space, to structural members in buildings and bridges.
Despite all these uses, there is not a full understanding of how individual fiber strength relates to the overall strength of the final object, and how this relationship changes for different types of materials. This is due in large part to the large variability in fiber failure strength, which introduces some elements of stochastic processes into the mechanical failure. Frequently the fiber and matrix used have time and temperature dependent properties, creating an additional challenge. Predicting overall bulk material reliability from constituent material properties thus requires not only careful experimental measurements and characterization at multiple rates, but also theoretical modeling and Monte-Carlo statistical simulation.
For certain applications, an understanding of the strength distribution is critical. Examples are body armor and COPVs, where the ‘acceptable’ probability of failure is very low (less than one in a million), as either the rupture of a pressure vessel or a penetration into a piece of body armor could lead to loss of life, as well as potentially a large monetary cost in the case of COPVs. For example, the 2016 failure of a Falcon 9 rocket led to more than a billion dollars in damage, and was caused by a COPV failure. It is not feasible to measure a one in a million probability of failure experimentally, as that requires testing at least ten million specimens. Only by combining experiment, theory and simulation can we ensure the overall reliability of such large scale engineered components.
Body armor “ballistic witness coupon”
Body armor is currently certified for 5 years from its manufacturing date, due to concerns about the material ageing over time. However, how the body armor is used or stored has a large influence on the amount of degradation in the armor. Armor with minimal use may still have enough strength to protect the wearer for many years longer than the stated 5 year certification period. Currently, however, the only method to test if armor is still safe is to ballistically test it, after which the armor is no longer useable. Ideally a “ballistic witness coupon” could be included in the armor package, exposed to the same conditions as the armor itself, and could be removed and tested to re-certify armor. This coupon would need to be small, to add as little additional weight to the armor as possible. Thus, re-certification relies on being able to accurately predict the ballistic properties of the armor from small scale testing on its constituent materials.
Variable rate and temperature fiber characterization
Understanding the inherent properties of the fibers is necessary in order to make predictions for composites comprising those fibers. As many fibers are temperature and rate dependent, their mechanical properties must be determined over a range of loading rates and temperatures.
Theoretical modeling of composite failure
In order to determine the failure distribution of large scale composites, it is not sufficient to fully characterize the constituent materials. The failure process of the composite must be considered, the key elements identified and then modeled in a tractable manner. This approach complements that of Monte-Carlo simulation.
Monte-Carlo simulation of tensile material properties
By instantiating random strengths and assuming load transfer occurs according to deterministic formulas, Monte-Carlo simulation can be used to determine material property distributions for larger composite samples. Currently we are working on predicting the stress-strain curve for large bundles, both loose and twisted, based on measuring the single fiber properties. This approach complements FCRL’s theoretical modeling.