Heat waves in Digital Image Correlation
Posted by Micah Simonsen, Last modified by Micah Simonsen on 13 October 2016 01:05 PM
When heat sources or hot surfaces are present in a DIC testing setup, refractive heat waves may distort your image; this distortion is not always apparent but it can create major errors in your results. These heat waves are similar to the heat waves visible on hot asphalt; they are caused by convection of hot air from the surface mixing with cooler ambient air.
The severity of effects will be dependent on the magnification of your system. In general, heat waves will present as large, shifting bands of strain and displacement. The bands will shift from image to image and typically have a period of around one second. False strains will be in the range of a few hundred microstrain to a few thousand microstrain.
In a 3D test, these heat waves will also increase the projection error for an analysis.
To check for heat waves, you can simply zoom in to the live image and watch it. The image should appear completely steady. Severe heat waves will cause image displacement and they will also make the image appear to "swim" in and out of focus.
Less severe heat waves will not be visible to the eye; if in doubt, simply capture a few images of your static specimen over a period of several seconds. Analyze the images and look for any significant false displacement or strain.
Heat waves can be caused by either a hot source (usually, a light source); a hot specimen (for example, a hot coupon in a high-temperature tensile test); or a surface that is being heated by your light source (for example, the bottom platen of a test frame).
If the cause is a hot source like a tungsten light, the simplest fix is often to simply move the light. The light, and the area directly above it, should be out of the optical path. This means moving the light either behind the cameras, or above the optical path so that the waves do not pass through it.
If the specimen or surrounds are being heated, one option is to move to a cool light source like an LED light or a fiber optic illuminator. If very bright light is required for a high speed test, you can set the test up, and then turn off the lighting and allow everything to cool. When it's time to test, turn on the light and immediately run the test; the light and specimen won't have time to get hot enough to generate heat waves.
If the heat waves can't be corrected in these ways, use a small desk or stand fan to mix the air around the specimen and optical path. This will even out the index of refraction and the heat waves should disappear. After applying any of these fixes, use the static check described above to confirm the absence of distortion.
Time Averaging Results
If heat waves cannot be eliminated - perhaps the specimen is inside a furnace and the air cannot be mixed or the heat source removed - we can mitigate them with the Time Averaging feature in Vic-3D.
To take images for time averaging, you must take n images at each test position or load stage. n can be any number greater than or equal to 2 but it must be the same for each step. For example, in a tensile test, we would take 10 images unloaded; 10 images at 1kN; 10 images at 2kN; etc. Then, in Vic-3D, run the data as usual. Finally, select Time average data from the Postprocessing menu. Select the correct value for n and click to process. Each set of images will be averaged; this should reduce or eliminate the random noise induced by the heat waves.