The question for Moore, his Sandia National Laboratories colleague Timothy Briggs in California and their teams is whether the impact caused significant, hidden damage inside the composite. They're developing nondestructive ways to detect damage in composites, using traditional medical inspection techniques such as X-rays and sonograms and advanced methods including infrared imaging, ultrasonic spectroscopy and computed tomography. Sandia began studying composites several years ago to see whether the lightweight materials could be used in national security applications. While a composite in a cell phone needs to last only a couple of years, "typically materials for national security applications must survive for decades. This makes you think differently about where and why you would use a material," said Moore, who works in the Structural Dynamics and X-ray/Nondestructive Evaluation department. "We need to study the lifecycle of a component. We tend to think deeply about the consequences of fracture or deformation and how we can verify what happened." The work supports many aspects of Sandia's national security mission, including energy efficiency and performance improvements in lightweight vehicles or wind turbine blades, said Briggs of Sandia California's Lightweight Structures organization. Composites join together separate materials with different characteristics. They often consist of a soft polymer matrix with reinforcing fibers like carbon, Kevlar or glass. Composites can use bundles of thinner-than-a-hair carbon strands that yield a high strength-to-weight ratio. The final shape and strength is obtained after heating the part in an industrial oven that sets the polymer resin and yields the qualities necessary for a structural component. Composites are increasingly important in aerospace and other industries because they're strong and weigh less than metals. Most can be bonded to metal for such uses as aircraft wings, making planes lighter and less expensive to fly. Outside surface of a composite doesn't hint at what's inside "We have a rich history of understanding metals and their failure mechanisms," Moore said. "Composite materials are very different." If a service truck backs into a composite aircraft fuselage, an examination of the impact site might not detect damage under the surface. That highlights the reason for nondestructive techniques that can fully evaluate how composites react in various circumstances. The research team is assessing the accuracy of nondestructive methods and how they could be used on a production floor. "You have to know what could go wrong in the processing steps and how to circumvent those, and then you want to make sure if you're going to make one or a hundred or a thousand that you're making them the same way all the time," Moore said. "Once we establish the limits of detectability, the threshold of good, bad and questionable, we'll be able to say, 'We want this composite bonded to this material with a defined quality and it shall be inspected with this technique,'" he said. Sandia's Lightweight Structures Lab defines and consolidates materials to study, using particular stack sequences of composite material layers to tailor strength and stiffness. It works in concert with the National Security Campus in Kansas City, Missouri, in everything from developing process methods to building prototypes to qualifying designs for particular applications. After making composites, the fabrication lab cuts out specimens for instrumented experiments—abusing the carefully made sample to study deformation, fracture and damage growth. Briggs "pulls, stretches, torques and crushes them. He performs these mechanical experiments so we can understand the fracture mechanisms around failure," Moore said. "Then we try to detect some of those failure modes." "The fundamental characterization of composites measures material properties and structural characteristics, which in turn provides information to validate computer modeling and simulation," Briggs said. Data from these destructive tests is correlated with nondestructive evaluations from Moore's team to understand what caused the material to respond the way it did, Briggs said. The effort hinges on close collaboration with Sandia's materials characterization groups and modeling and simulation colleagues to help validate their computer-based simulations. Composites must support a particular weight and size. "Anyone can build a structure to carry the load, but we have to design our structures to fit within a geometric envelope and be lightweight," Briggs said. "We cannot simply over-engineer to unrealistic levels. We have to be very smart and efficient with our designs, yet provide enough margin for long-term reliability." Materials bonded in an oven or autoclave often have different thermal expansion rates—aluminum expands more than fiber-reinforced plastics, for example. Once a composite cools after curing, residual stresses can build up inside, particularly at interfaces. If the composite can't handle those stresses, the bonds can fail. Sandia is developing advanced techniques to complete sample inspections in less than 5 minutes in some cases. "This is a way to gain a lot of information very quickly about the quality of the bonds," Moore said. His team uses the deliberately mistreated composites to assess such inspection techniques as advanced ultrasonics, flashed or active thermography and computed tomography. Ultrasonic testing has been around for 60 years, but computers and other improvements now allow study of more complex applications. Technologist Andrew Lentfer demonstrates, scanning a piece of composite with a handheld ultrasonic roller that resembles a small paint roller with a hollow, water-filled barrel. As he scans the composite's layers, a computer screen maps them in color: Yellow-green is OK; blue indicates weakness. Rollers can scan curved surfaces, even large ones like airplanes. Fibers and interfaces in a composite scatter ultrasonic waves moving through the material. Moore compares it to ocean waves: "If a wave hits a rock face in the ocean it moves around it; if a wave washes up on the sand it gets absorbed; and if it hits a seawall the wave energy is redirected quickly. Those are the same fundamentals we investigate: ultrasonic energy moving through a composite matrix." Knowledge gained from characterizing materials helps develop new nondestructive techniques "so when we establish an inspection criteria, we have a better feel for what we can detect and what we cannot," he said. Flashed thermography, commercially available for more than 20 years, flashes very high-energy light onto a surface for 15 microseconds, then an infrared camera watches how the surface cools. The process takes only minutes. "It's very fast, but you have to understand the fundamentals of heat flow and how the material surface either gives off heat to its surroundings or transfers heat within itself," Moore said. The research uses composites with high-quality bonds and others deliberately made with weaker bonds. Differences in results help the team improve detection of defects or damage. A computer screen shows light or dark spots indicating possible problems, and an overlaid graph ties the depth of the potential problem to the time indicated on the image. "The question becomes, is that a concern? Is that a crack or not?" Moore said. "We'll be able to answer those questions. If the defect propagates deep into the material, we may not detect it. It's wise to understand the capabilities of the technique and then perform the math and science behind it." Computed tomography systems are efficient for finding small defects. The technique rotates a sample 360 degrees while taking 1,000 images, similar to a medical CT scan, and generates an image from each thin slice of the object. Since each image taken is two dimensional, computer algorithms reconstruct, calculate, locate and display everything to represent the object in three dimensions. "Once the 3-D image is reconstructed, you look at the front surface and then start moving through the thickness to view what is below the surface," Moore said. "This technology gives us a knowledge baseline and validates how the other techniques are performing." Sandia must ensure designs meet requirements by specifying and qualifying an inspection technique, and now is writing inspection procedures for the National Security Campus, Moore said. "Once we establish criteria and limits of acceptability, the product definition can be established," he said. Explore further: Seeing below the surface: New way to inspect advanced materials used to build airplanes
« Germany pumping €1B into plug-in vehicle subsidies, infrastructure; base price cap of €60K | Main | DLR presents HY4 4-passenger fuel cell hybrid electric aircraft at 2016 Hannover Messe » Lawrence Livermore National Laboratory (LLNL) material scientists have found that 3D-printed foam works better than standard cellular materials in terms of durability and long-term mechanical performance. Foams, also known as cellular solids, are an important class of materials with applications ranging from thermal insulation and shock-absorbing support cushions to lightweight structural and floatation components. Such material is an essential component in a large number of industries, including automotive, aerospace, electronics, marine, biomedical, packaging and defense. Traditionally, foams are created by processes that lead to a highly non-uniform structure with significant dispersion in size, shape, thickness, connectedness and topology of its constituent cells. As an improved alternative, scientists at the additive manufacturing lab at LLNL recently demonstrated the feasibility of 3D printing of uniform foam structures through a process called direct-ink-write. The material is built up layer-by-layer, with each layer consisting of equally-spaced parallel cylinders of the same uniform diameter. However, since 3D printing requires the use of polymers of certain properties, it is important to understand the long-term mechanical stability of such printed materials before they can be commercialized. This is especially vital in applications such as support cushions, where the foam material is subjected to long-term mechanical stresses. To address the stability question, the LLNL team performed accelerated aging experiments in which samples of both traditional stochastic foam and 3D-printed materials were subjected to a set of elevated temperatures under constant compressive strain. The stress condition, mechanical response and permanent structural deformation of each sample were monitored for a period of one year and, in some cases, even longer. A method called time-temperature-superposition was then used to quantitatively model the evolution of such properties over a period of decades under ambient conditions. This study convincingly demonstrated that 3D-printed materials age slowly—i.e., they better retain their mechanical and structural characteristics—as compared to their traditional counterparts. Interestingly, native rubber (i.e. elastomer) comprising each foam showed exactly the opposite effect—i.e., the rubber in the printed material aged faster than the corresponding rubber used in the traditional foam. To gain further insight into why the printed cellular material displayed superior long-term stability, the team imaged the 3D micro-structure of each foam sample with X-ray computed tomography, and performed finite-element analysis of the stress distribution within each micro-structure. They found that there is a much wider variation in local stresses within the stochastic foam, with points of extreme stress significantly higher than the maximum stress points within the more uniform 3D-printed foam. This paper represents the very first study of this nature. Here the long-term mechanical characteristics of a 3D printed polymer foam is carefully compared with that of a traditional stochastic foam through the analysis of multi-year-long accelerated aging data using a time-temperature-superposition procedure based on geometric arc-length minimization. The resulting master curves predict clearly superior long-term performance of the AM foam, both in terms of compression set and load retention. This result is remarkable given that the AM foam is created out of rubber with three times the stronger propensity for permanent deformation as compared to the rubber constituting the stochastic foam. To gain insight, we have imaged the microstructures of both foams with X-ray computed tomography and carried out Finite-element analysis of stress distribution. Such analysis leads us to conclude that the superior long-term behavior of the AM foam is due to a more uniform local stress distribution pattern relative to the stochastic foam, which develops more extreme stress points within its microstructures. The latter is likely responsible for irreversible damage to the foam structure including pore collapse, strut fracture, and permanent deformation of the cell wall. An open-access paper on the research appears in the journal Scientific Reports. The group also acknowledges contributions from the National Security Campus, Missouri (formerly Kansas City Plant) where some of the early aging experiments were performed.
Everhart W.,National Security Campus |
Sawyer E.,National Security Campus |
Neidt T.,National Security Campus |
Dinardo J.,National Security Campus |
Brown B.,National Security Campus
Journal of Materials Science | Year: 2016
Additive Manufacturing (AM) has significantly increased the design freedom available for metal parts. Many novel designs rely on the existence of surfaces that are not accessible and therefore rely on the surface finish of the parts directly from the AM equipment. Work has been performed to characterize the difference between AM, then machined tensile samples and AM tensile samples with an unimproved surface finish. This work utilizes surface analysis, fractography, and finite element analysis (FEA) to expand on this by investigating the effects of the unimproved surfaces on local tensile behavior and fracture mechanics in AM materials. Results show that measurement error in cross-sectional area is the main source of variation between unfinished and machined strength measurements. Results also indicate that a ductile material may demonstrate the same tensile strength regardless of post processing. Fractography shows that stress concentration near the surface of the samples leads to changes in fracture behavior likely explaining the difference in elongation of the samples. Finally, FEA work did not successfully show a difference in fracture initiation, though this is likely due to inaccurate representation of the samples surface. © 2016, Springer Science+Business Media New York (outside the USA).