Multiaxial Testing for Material Characterization
The recent years have seen an increasing attention to structural composites and functional composites for smart materials applications such as piezoelectric polymers, and bio-materials. These materials exhibit anisotropic behavior and thus provide more capabilities for mission and application tailored design, as well as functional flexibility to final structures, than regular isotropic materials. The directional properties, introduced by their anisotropic makeup, however complicate their deformation as a compensation. In the characterization of materials, uniaxial tests are dominantly used regardless of the directionality of materials. When anisotropic materials are characterized, a specimen is fixed to a uniaxial testing machine so that a material coordinate aligns to the axis of the testing machine, and its strength with respect to a different material coordinate is examined by re-fixing the specimen or replacing it by another identical specimen so that this coordinate coincides with the machine axis. Such an experimental procedure resultantly identifies all the elastic and inelastic properties of the anisotropic material, but the identified properties are often unreliable due to the complication of their deformation and do not contribute to the comprehensive understanding of material behavior. This has given rise to the need for material characterization through multiaxial tests and, accordingly, has seen the recent developments of multiaxial testing machines.

Objective
This presents a methodology that updates the loading path of a multiaxial testing machine to identify the elastic moduli reliably every time a new set of sensor readings are obtained. In accordance to the setup of the multi-axial testing machine, equipped with sensors to measure the boundary displacements/forces and the full-field strain of the specimen, the proposed methodology derives the increments of the external work and the strain energy from every set of sensor readings. The material properties are identified iteratively by equating the increments of the external work and the strain energy at every sensor reading.

Figure 2: Material characterization via biaxial testing

Results
Two performance measures, distinguishability and uniqueness, have been defined to quantify the loading path and the reliability of material parameters identified. Figure 2 shows the control and identification system developed for biaxial testing. The automatic feeder is used to provide specimens consecutively so that the material parameters can be identified from a large number of experimental data. Boundary displacements/forces and full-field strain are measured and the measured data are used to evaluate distinguishability and uniqueness. The optimal loading path are updated at every sensor reading such that both the distinguishability and uniqueness are maximized. Figure 3 shows a result of identification via optimal loading. It is shown that both the distinguishability and uniqueness are well increased.

Figure 3: Distinguishability, uniqueness, ellipse for asymmetric specimen