For effective structural health monitoring in fuselages and in engines, it is important to quantify the damage tolerance of a candidate structure. New aircraft designs, both commercial and military, are increasingly relying on composite structures to reduce aircraft weight and fuel consumption. Fiber composites, by their multiscale nature, are able to arrest cracks and prevent self-similar crack propagation, and composite structures have received a great deal of consideration for designs that emphasize damage tolerance. However, the number of design parameters involved--such as fiber-orientation patterns, choices of constituent material combinations, ply drops, and hybridization--make the design of composite structures very complex. Thus, designers must evaluate damage initiation in a composite structure,as well as fracture propagation characteristics, in the early design phase in order to achieve a rational damage-tolerant design.
Compared with homogeneous materials, fiber composites have much more complicated damage initiation and progression characteristics. Composite structures often contain some preexisting flaws or flaws that were induced in the matrix and fibers during fabrication. At lower stresses, the matrix is likely to be cracked because of flaw-induced stress concentrations, and these flaws can propagate across the composite. By using established material modeling and finite element models--and considering the influence of local defects, through-the-thickness cracks, and residual stresses--researchers can use computational simulation to evaluate the details of progressive damage and fracture in composite structures.
In a computational simulation, a damage evolution quantifier such as the damage volume, exhausted damage energy, or the damage energy release rate (DERR) are used to quantify the structural damage tolerance at different stages of degradation. Low DERR levels usually indicate that degradation takes place with minor resistance by the structure. Structural resistance to damage propagation is often dependent on structural geometry and boundary conditions as well as on the applied loading and the state of stress. In certain cases, such as the room-temperature behavior of composites designed for high-temperature applications, internal microcracks initiated in the matrix grow large enough to be externally visible. Thus, matrix cracking and its effect on damage propagation and damage tolerance need to be evaluated.
Damage initiation, growth, accumulation, and propagation to fracture were studied at the NASA Glenn Research Center. Since the complete evaluation of ply- and sub-ply-level damage and fracture processes is the fundamental premise of computational simulation, a microstress-level damage index was added to identify and track sub-ply-level damage processes. Computed damage regions were similarly correlated with ultrasonically scanned damage regions. Then, the simulation was compared with data from acoustic ultrasonic testing to validate it. Results showed that computational simulation can be used with suitable nondestructive evaluation (NDE) methods for credible inservice monitoring of composites.
Definitions of regions for ply microstress calculations. The numbers 2 and 3 indicate material axes.
At Glenn, we evaluated the microstresses by considering the model on the preceding page. In this model, the different regions where failure will occur are labeled. For example, A represents an all-matrix region, and B represents a mixed region of matrix, interface, and fiber. The concept is illustrated in the following graphs where the acoustic signal is plotted. The top graph shows the envelope curve for an epoxy-resin-bonded specimen, and the bottom graph shows the envelope curve for a prepreg-bonded specimen. The area under the envelope represents the total damage energy detected by the acoustic emission during the period monitored. Thus, we could correlate the microstress damage energy through computational simulation with the relative total damage energies represented by the envelope areas. The envelope area in the top graph is 1.22986×10-3 and that in the bottom graph is 1.30145×10-3. The ratio of the energies from the bottom graph to those of the top graph is 1.058. Comparatively, the ratio of damage energies computed from the NDE test results (ultrasonically scanned damage regions, ref. 1) is 1.168, which is 10 percent higher. Results showed that computational simulation can be used with suitable NDE methods for credible inservice monitoring of composites.
Energy envelope of acoustic emission signals for bonded specimens. A special curve “spline” was fit through several distinct points for the interpolation. Top: Epoxy resin. Bottom: Prepreg.
Last updated: August 31, 2007
Responsible NASA Official:
Gynelle.C.Steele@nasa.gov
216-433-8258
Point of contact for NASA Glenn's Research & Technology reports:
Cynthia.L.Dreibelbis@nasa.gov
216-433-2912
SGT, Inc.
Web page curator:
Nancy.L.Obryan@nasa.gov
216-433-5793
Wyle Information Systems, LLC
