In-Situ Tensile Testing with EBSD: Watching Metals Deform

A conventional tensile test gives a stress-strain curve: yield point, work hardening, ultimate strength, elongation. It says nothing about what the grains were doing to produce that curve. Post-mortem microscopy shows the end state, but not the order in which things happened or which grains gave way first.

Running the test inside the SEM, with EBSD mapping at chosen points along the curve, connects the two. You can see which grains start to slip first, how they rotate, where misorientation builds up, when twins appear and where cracks start, all on the same patch of material.

What in-situ EBSD shows

Some of the questions in-situ EBSD answers well:

• Grain rotation and texture development. As a polycrystal deforms, grains rotate towards preferred orientations. Following individual grains shows how texture develops, and which neighbourhoods resist it.
• Slip activity. Slip traces on the surface, combined with each grain’s orientation, show which slip systems switched on and in what sequence.
• Strain localisation. Measures such as kernel average misorientation (KAM) and grain reference orientation deviation (GROD) map where lattice curvature, and so dislocation density, builds up. It typically concentrates near grain boundaries, triple junctions and hard particles, often well before anything shows in an SE image.
• Twinning. In magnesium, titanium, zirconium and twinning-induced plasticity (TWIP) steels, deformation twins nucleate and grow during loading. In-situ maps show which grains twin, at what strain, and which twin variants form.
• Phase transformation. In TRIP steels, retained austenite turns to martensite under strain. Interrupted EBSD maps track how fast it disappears, and where.
• Damage. Mapping around voids and microcracks as they form shows the local orientation and strain conditions that set them off.

At elevated temperature the same approach extends to recovery, recrystallisation, grain growth and high-temperature deformation.

The stage: fitting a tensile test to EBSD geometry

EBSD needs the sample surface tilted about 70° towards the detector. Tilting the whole SEM stage that far with a tensile stage on it is often impractical, so EBSD tensile stages build the tilt in. Deben’s Microtest EBSD stages hold the sample at a 70° pre-tilt for EBSD, or flat for horizontal SEM imaging, while the SEM stage itself stays level. On the 5 kN stages, exchangeable 70° EBSD jaws leave a clear line of sight to the SEM lens and the EBSD detector.

There are four models in the range:

• MT2000ER: 2 kN tension and compression at room temperature.
• MT2000EH: 2 kN with heating to 600 °C.
• MT5000ER: 5 kN dual-motor stage at room temperature.
• MT5000EH: 5 kN with heating to 1000 °C, water-cooled.

On the dual-motor 5 kN stages the jaws can move together, so the middle of the gauge length stays roughly in place as the sample stretches, which helps keep the mapped area in view. Load, extension and speed are controlled from Deben’s Test Suite software, and the specifications for each stage are on its product page and in Deben’s EBSD stage datasheet.

Interrupted or continuous?

Most in-situ EBSD tests are interrupted: the stage pulls to a set strain, holds, and EBSD maps the area before the next step. That gives sharp patterns at each step, but it has a known side-effect. While the sample is held, the load relaxes somewhat as the material creeps and dislocations rearrange. The shorter the hold, the less it matters.

This is where detector speed comes in. With a conventional camera, a useful map could take tens of minutes per step, so a test had only a handful of steps and each hold was long. A direct-detection EBSD detector such as Bruker’s eWARP, part of the QUANTAX EBSD system, collects hundreds of thousands of points in under a minute (see our article on ultra-fast EBSD). Holds become short, more strain steps fit into a session, and the deformation sequence is captured in much finer detail. Bruker also points to fast forescatter and backscatter imaging, at up to 350,000 patterns per second, for following changes almost as they happen, and its ESPRIT software can acquire time-resolved maps automatically during a test.

Practical tips

Sample design. The gauge section has to suit the stage’s clamping lengths and maximum width, and the area of interest needs to be where it can be mapped at 70°. Check the sample dimensions on the stage’s product page before machining.

Surface preparation. EBSD patterns come from the top few tens of nanometres, so the surface needs to be damage-free before the test starts. Electropolishing or broad-beam argon ion milling, for example flat milling with the COXEM IP-10K, removes the layer that mechanical polishing leaves. As the sample deforms, the surface roughens and patterns degrade, which is another reason to start with the best possible surface.

Fiducial markers. A few FIB marks or micro-indents around the mapped area make it far easier to find the same grains at each step and to align maps afterwards.

Combine with DIC. Digital image correlation of SE images, using a fine speckle pattern on the surface, gives a full-field strain map. Overlaid on EBSD, it links local strain to local crystallography.

Plan the steps. Spend more steps where things change quickly, around yield and the onset of necking, and fewer in steady work hardening.

Beyond tension

The same stages run in compression, and Deben also makes stages for testing inside X-ray CT systems. See our tensile and compression testing overview, or browse all tensile and compression stages.

If you are planning in-situ EBSD work, we can help match a stage to your SEM chamber, EBSD detector geometry and sample sizes. Contact us to start the conversation.

Stage details from Deben UK; detector details from Bruker. Photos: Deben UK and Bruker.