October_AMP_Digital

ADVANCED MATERIALS & PROCESSES •

OCTOBER 2014

TECHNICAL SPOTLIGHT

Using

Digital Image Correlation

to Measure Full Field Strain

ith new and advanced materials

being developed continually, both

research efforts and materials test-

ing must adapt to keep up. Physical testing is

not simply a matter of following a standard to

collect data, but is a tool to analyze and opti-

mize properties. Traditional strain measure-

ment requirements involve characterizing

axial strain at one gauge length in the center of

the specimen. In some cases, transverse strain

is also measured to determine Poisson’s ratio,

and strain gauges are used where it is difficult

to use traditional extensometers.

However, existing strain measurement

techniques do not offer enough information

about how failures occur, so alternative meth-

ods are required. One technique involves

using digital image correlation (DIC) to

measure full field strain over the entire mate-

rial surface. In addition to the strain map pro-

duced, virtual strain gauges and extensometers

can be placed on the specimen after the test

and replayed multiple times. Further, the

technique can be used for almost any mate-

rial; polymers, metals, composites, rubbers,

foams, textiles, and other materials have been

tested using DIC.

How digital image correlation works

Digital image correlation is a strain

measurement technique that works by cap-

turing a series of images throughout a test

and analyzing them afterwards. A typical

setup includes a camera system, lighting,

and software package to control image cap-

ture and conduct post-test analysis. Re-

cently, integrated DIC systems (Fig. 1),

which use existing video extensometers,

have been introduced to provide a stream-

lined package tailored to the materials test-

ing market.

Pre-test samples are usually treated with a

speckle pattern (Fig. 2) added by spraying,

stamping, or sticking decals to their surfaces.

In some cases, the sample’s natural surface pat-

tern is sufficient without the need to apply any

additional marking. The number of images

captured during a test depends on time, speed,

and the sample itself, but 50 to 100 images are

usually adequate. The first image—known as

the

reference image

—is captured when there is

no strain on the sample. The image is then split

into small subsets and the patterns within each

subset of subsequent images are compared to

the reference image, and displacements are

calculated from which a strain/displacement

map is produced. Only one sample is required

to determine axial and transverse strains and

displacements, shear strain, and maximum

and minimum normal strains.

These maps are similar to FEA-type im-

ages (Fig. 3), which leads to a very useful ap-

plication for DIC—proving whether or not

an FEA simulation is correct. In the past, en-

gineers used simple test data from exten-

someters and strain gauges to verify

predictions from an FEA simulation. With

the new technique, subtle changes in a ma-

terial that might be missed in a simulation

can now be physically measured and identi-

fied, allowing operators to compare what

they believe is happening to what is actually

happening.

DICsoftwarealsoallowstheinclusionofvir-

tual extensometers andstraingauges that canbe

placedanywhereonthesample.Thisenablesan-

alyzing the sample near failure points, so a posi-

tion does not have to be selected before

conductingthetest.Italsoenablescomparingex-

isting datawithnewresults.

DIC as a teaching tool and more

Having this type of visual method to

observe material failure in action, as well as

Fig. 2

—

Typical

speckle

pattern

applied

to a test

sample.

Fig. 3

—

Sample DIC strain map.

Fig. 1

—

Integrated digital image correlation (DIC) system.

Axial strain,