Search

A general framework is described for coupling digital image correlation (DIC) with finite element analysis to determine material parameters from measurements of nonuniform displacement fields in a specimen. The approach is to minimize the difference between measured and computed displacement fields and external forces applied to the specimen by repeatedly correcting the material parameters in a chosen constitutive model. The minimization can be accomplished analytically, and the resulting nonlinear system of equations solved using a Newton–Raphson procedure that integrates naturally with standard nonlinear finite element computations. As a representative application, the method is applied to measure the flow behavior of tensile specimens beyond uniform elongation. The results of a series of tensile tests on tapered HSLA 590FB steel specimens show that the effective stress-strain behavior can reliably be extracted to a strain of 60%, well beyond the limits of 20% commonly achieved using traditional engineering analyses on straight-gauge specimens.

An experimental investigation of spatio-temporal characteristics of the Portevin–Le Châtelier (PLC) effect in austenitic steel with twinning induced plasticity (TWIP) is presented. Post-processing of high resolution digital images captured from specimens in quasi-static, room temperature tensile tests was conducted with a digital image correlation (DIC) method. This provided direct measurement of strain fields during all stages of the tests. Variable rate digital image capture, enabled with a custom image acquisition algorithm, guaranteed a suitable number of images recorded during serrations in load–time records. Nucleation, propagation, and morphology of individual PLC bands in both straight gage and tapered specimens were quantified with strain rate contours computed with a backward differentiation scheme. Time histories of strain evolution in the PLC band wakes were extracted from cumulative strain contours. Of the three types of PLC bands, only the continuously propagating Type A bands were observed. Band nucleation, which occurred at serration crests in flow curves derived from the DIC results, was not limited to regions of geometry-induced stress concentrations. Due to its importance in finite element springback predictions and to support theoretical model development of inelastic behavior in TWIP steel, we measured Young’s modulus variation with strain in periodic loading–unloading tests. Implications of the experimental results for theoretical modeling of the PLC effect in TWIP steel are discussed.

A problem formulation and solution methodology for design optimization of laminated composite I-sections is presented. Objective functions and constraints are given in the form of beam stiffnesses. Two different objective functions are considered, maximum beam bending stiffness and maximum beam axial stiffness. Fiber directions in the beam walls are treated as design variables. It is assumed that the beam is constructed using unidirectional tape, and manufacturing issues associated with the use of unidirectional tape are discussed and included as constraints in the problem formulation and solution. The paper demonstrates that the design optimization of composite thin-walled beams is a complex global optimization problem that cannot be solved by means of traditional convex programming. Therefore, the solutions described are found using a global search algorithm, Improving Hit-and-Run, which allows the design variables to be either continuous or discrete with a user-specified discretization interval. Numerical results for two material systems and nine different design families for manufacturing composite I-sections are presented.