Page 130 - Mechanics of Asphalt Microstructure and Micromechanics

122 Ch a p t e r F o u r

DIC matching approaches have been developed involving different types of object-
based patterns such as lines, grids, dots, and random arrays. Common use of DIC in-
volves random patterns and compares sub-regions throughout the image to obtain a
full field of measurements. As a promising tool for accurate experimental measure-
ment, the DIC method and its application in general mechanical experimental testing,
particularly the asphalt mixture testing, are briefly presented as follows. The DIC is
superior to conventional displacement and strain measurement methods, such as the
linear variable differential transformer (LVDT), in that it has the ability to capture full-
field deformation of the specimen and does not entail contact mounting on the speci-
men surface. The general information on DIC in the following sections is abstracted
from an excellent book by Sutton et al. (2009).

4.4.1 History of DIC
Pioneering work in image correlation dates back to the early 1950s. Gilbert Hobrough
compared analog representations of photographs to register features from various
views. With the emergence of digitized images in the 1960s, vision-based algorithms
were developed and applied in artificial intelligence and robotics. In the 1970s, image
correlation application centered on character recognition, microscopy, medical radiol-
ogy, and photogrammetry photography. Extensive applications in engineering shape
and deformation measurements were performed (Rosenfeld, 2001). The 1970s also saw
the use of laser technology in image correlation, including mainly the laser speckle
(Dainty, 1975), laser speckle photography (Archbold et al., 1970), and laser speckle in-
terferometry (Mallik and Roblin, 1972). Significant progress has been made since the
1970s in 2D and 3D DIC development for surface characterization, and in 3D DIC for
volumetric characterization. The following sections detail the developments in these
regards.

4.4.2 Use of DIC in 2D Surface Analysis
As the experimental testing of materials and mechanical systems gets more complex,
researchers have developed methods for digitally recording images of measurement
data, algorithms for analyzing images and extracting measurement data, and the ap-
proaches for automating the entire process. One of the early works that used computer-
based image acquisition and deformation measurements in material testing was con-
ducted by Peters and Ranson (1982). Sutton et al. (1983) developed numerical algo-
rithms known as 2D digital image correlation (2D-DIC). Davidson and Lankford (1983)
extended 2D-DIC concepts to the micro-scale for measuring large deformation near fa-
tigue flaws. Han et al. (1994) developed a high magnification optical system to measure
deformations around stationary and growing crack tips under nominal Mode I loading.
In addition to the work in fracture mechanics, investigators used 2D-DIC to understand
the deformation behavior of materials including metals, plastics, wood, ceramics, and
tensile loading of papers. In the late 1990s and 2000, investigators applied 2D-DIC to
study damage in composites and concrete. The 2D-DIC was applied to measure surface
deformations in planar components in the 1990s. Most of the studies prior to 2000 used
direct image correlation principles for matching subsets and extracting full-field dis-
placements.
The use of 2D-DIC has gained rapid growth since 2000. 2D-DIC measurement proce-
dures have received various modifications, including the search procedure, correlation
approach, and registration method. As to the registration methods, Cheng et al. (2002)