Search NASA⌕ Search

Engineering topics

William M Johnston

Publications and source records attributed to William M Johnston.

8 records
All sourcesNASA NTRSDOE OSTI
NASA NTRSDate not supplied

TPSAS-NF1676L-32227-DND

The ability of a recently proposed cohesive fatigue model to predict the initiation and propagation of skin/stiffener separation was evaluated using a three-point bend (3PB) configuration. The 3PB specimen consists of a quasi-isotropic laminate representing the skin of a fuselage. A shorter doubler representing the flange of a stiffener is co-cured at the center of the specimen. A simplified modeling technique was applied, in which the complex damage mechanisms involved in the separation, such as delamination migration, matrix cracking, bridging, and crack delving are represented by a mixed-mode trilinear cohesive law at the interface between the skin and the doubler. The quasi-static properties of the trilinear law were characterized. The main advantage of the proposed fatigue model is that it relies on quasi-static properties as well as engineering assumptions to compose the shape of the S-N diagram. The model does not require Paris law data, and it is capable of predicting the initiation of separation, the rate of propagation, as well as the effects of crack resistance curves and the stress ratio.

Carlos G Davila↗
NASA NTRSDate not supplied

Initiation and Propagation of Skin/Stiffener Separation in Postbuckled Structures Subjected to Cyclic Loads

A methodology is proposed to evaluate experimentally and numerically the initiation and propagation of skin/stiffener separation in postbuckled panels subjected to cyclic loads. The response mechanisms that characterize skin/stiffener separation were investigated using a three point bend (3PB) test specimen, which is a simple specimen consisting of a laminate representing the skin and a doubler representing the stiffener flange. The 3PB tests were performed under quasi-static and cyclic loads. A fatigue cohesive model was used to analyze the initiation and propagation of skin/stiffener separation in the 3PB. The results indicate that the fatigue cohesive model with model parameters obtained from quasi-static fracture tests can provide accurate predictions of initiation from a pristine state, and the rate of fatigue crack propagation for a variety of mode mixities, load levels, and stress ratios.

Carlos G Davila↗
NASA NTRSDate not supplied

Validation of Delamination Growth Predictions under Fatigue Loading using the Single Leg Bending Specimen

The Single Leg Bending Specimen (SLB) was chosen for a study to compare experimentally determined delamination growth with analysis predictions. First, fatigue tests were performed on SLB specimens with several different starter delamination lengths. Specimens were cycled in load control and delamination length increase was continuously recorded with increasing number of cycles. Test results from different load levels were combined to capture the entire delamination growth law (Paris Law) over a wide range of energy release rates. Second, a recently developed interpolation scheme was used to predict the Paris Law associated with the SLB specimen. This interpolation scheme allows the calculation of the growth rates for a given mixed-mode ratio from known single-mode growth laws. Third, finite element analyses were performed using two-dimensional models of the SLB specimen. New Virtual Crack Closure Technique based analysis tools for automated delamination growth analysis under cyclic loading in Abaqus were used for growth predictions. Fourth, the predictions from analyses were compared to experimental fatigue data. Remaining analysis shortcomings were highlighted.

Ronald Krueger↗
NASA NTRSDate not supplied

Challenges in Scaling Analysis from Coupons to Elements in 3D Woven Composites for a Launch Vehicle End-Ring Structure

The limitations of structural design with 3D woven composites based on homogenized material models characterized using coupon test data, representative unit cell analysis, or a combination of the two are discussed in the context of a space launch system (SLS)-scale end ring. Selected building block test configurations at the coupon, element, and subcomponent scales are described in detail. Highly instrumented tests and X-ray computed tomography inspections are used to characterize the structural behavior and associated damage progression with increasing load. It is shown that a single homogenized material constitutive relationship, applied at coupon and element scale, is not sufficient to model the structural response observed in both tests. In fact, substantially different stress-strain laws are required to predict the response of the coupon and element scale specimens, despite similar loading conditions. A model with discrete representation of the binder tows and cohesive interfaces to model matrix cracking shows much better agreement with the test data at the two length scales. Thus, evidence is provided showing that predictive structural analysis of 3D woven composites requires discrete representation of the tows, matrix cracking morphology, and their interactions in the structural configuration. With accurate knowledge of the element scale behavior, engineering approximations readily enable analysis scale-up to the subcomponent level.

Andrew C Bergan↗
NASA NTRSDate not supplied

Interlaminar R-Curve Effects in Thermoplastic Composites: Experiments and Model Correlations

The interlaminar fracture behavior of thermoplastic composites under Mode-I and Mode-II static loading is investigated in this research. The material system under consideration is a carbon fiber reinforced low melt semi-crystalline resin TC1225 LMPAEK T700G (T700/LMPAEK)†. Characterization experiments were conducted to measure the load-displacement response, peak load, and propagation of delamination from a pre-implanted Kapton® insert. The Mode-I response was characterized using a double cantilever beam (DCB) test according to the ASTM D5528-22 standard. It was observed that the propagation exhibited significant resistance to delamination growth, i.e., an R-Curve effect. The test data indicated that the steady-state fracture toughness, G SS I , was approximately twice the mean initiation fracture toughness, G Ic . The Mode-II response of an end notch flexure (ENF) specimen tested according to the ASTM D7905-19 standard exhibited an unexpected stable behavior upon attaining peak load. Prior to attaining peak load, the Mode-II response exhibited some non-linearity, which may be attributed to nonlinear bulk material response, e.g., hyper-elasticity, viscoelasticity, plasticity, etc., and potentially friction at the load/support rollers and R-Curve effects. Finite element models based on the Cohesive Zone Method (CZM) and the Virtual Crack Closure Technique (VCCT) were used to predict the DCB and ENF specimen responses. DCB analysis results obtained using G SS I predicted propagation consistent with experimental data. However, results obtained with the as characterized G Ic were inaccurate in predicting delamination growth onset and peak load. ENF analysis results obtained with the as-characterized Mode-II fracture toughness, G IIc , were inaccurate in predicting the peak load and propagation. Hence, R-Curve effects were included in both DCB and ENF simulations, which enabled improved predictions. The experimentally determined Mode-I R-Curve was used for the DCB simulations, whereas an R-Curve was derived for Mode-II assuming that the change in compliance of the specimen was associated with crack growth only.

Prabhakar M Rao↗