Power and uncertainty reduction in the process

As an illustration, one case objective was to determine the final shape of a component fabricated of carbon fiber/epoxy unidirectional prepreg tape. A simple angle structural element is desired, but the well-known “shrinkage deformation” phenomenon resulting from thermal and chemical shrinkage of the epoxy during cure can produce a final geometry significantly different from the designed shape. The first step in the process is to design the geometry of the structural element, including the laminate lay-up, to meet the geometric constraints of the desired element. Use of software modeling provides the necessary functionality. Next, a finite-element analysis (FEA) of the structural element is carried out to satisfy stiffness and strength performance requirements. This process involves creating an FEA model using a software program. When the geometry of the structural element involves curvilinear surfaces, the conformation of the unidirectional prepreg tape to the tool surface is determined from another simulation tool.

UQ is an established methodology to predict the range in expected outcomes through simulation. It can combine simulation and experimental test results to reflect the actual range in expected performance and thereby assure confidence in performance with fewer tests. Today, the production of composites design variables, such as “A” and “B” allowables, is an entirely experimental task. The science of UQ is currently utilized by the National Nuclear Security Administration to certify weapons performance in the absence of testing as governed by global treaties. This body of knowledge can be transferred to the composites industry to guide the development of the new certification paradigm for composite materials and structures. Combining UQ with composites simulation tools can provide the foundation for certification of composite products with fewer physical tests.

One example of the UQ approach to determine composite performance variability is illustrated by incorporating the variability in the microstructure on failure in fiber-reinforced composite materials such as the open hole tensile strength. The uncertainty in the fiber-volume fraction is propagated to determine the uncertainty in the failure location and stress or magnitude at the onset of damage. The results in the table on page 18 show that the variability of the microstructure, in particular of the fiber-volume fraction, determines variability in failure location and load magnitude. As can be seen, the experimental and simulation results complement one another so that a smaller number of experimental results can be combined with analytical predictions of variability to achieve knowledge appropriate for certification. Other material defects, such as surface imperfections in the hole and debonded or broken fibers may be incorporated with a similar approach. Further, this framework can be used to establish variability in both performance and manufacturing results such as macro/micro geometry and resin cure variability.