Reproducing the Velador Experiment
Measuring Beam Deflection
As with thermal refraction, beam deflection can be artificically produced by applying external loads. These deflections are an important source of error in the experiment and must be controlled for. Optimally, no experimental trial with this instrument will be considered complete unless the projected image is measured at least twice at each orientation, preferrably with a full 360 degree rotation between each measurement.
Example of Deflection Due to Lateral Loading
This is an adjusted false color image image of the central region of the laser beam image taken using the camera 2 setup. Most of the diffraction pattern rings have been edited out, with the exception of one arc which could not be reliably separated. The upper portion of each colored area is the actual laser area of incidence, while the arc below is a part of a diffraction pattern. All of these images were taken with a south orientation during the same trial, with a 360 degree rotation and other images being taken at intermediate orientations. Only the southward laser orientation images are shown. The light green and black images match relatively closely but differ by an amount similar to image differences over 180 degrees measured in earlier calibration runs. This is due to uncorrected pendular motion of the apparatus. The dark green image does not match position in comparison, although it is also a south orientation image taken during the same trial. It's position error is also due to uncorrected pendular motion.
All three images are aligned at the corner positions.
Example of Image Distortion Under Beam Deflection
In the picture above, the adjusted three south images are superimposed over each other by aligning the central image areas of the laser beams rather than aligning the image fields. This alignment reveals a large distortion in the diffraction pattern of the image which is moved the farthest by beam deformation.
It is unclear at this time whether the image distortion is due to thermal refraction or some unidentified geometric effect. However, if this is a geometric effect, it strongly suggests that measurements should be taken from the central area of the laser beam image as much as possible, and measurements based entirely on the surrounding diffraction pattern should be avoided. This should not be a sacrifice, as the central area will already have the lowest level of position error at its edge. It also raises the possibility that major deflections due to beam deformation can be identified by careful analysis of the surrounding diffraction pattern.
The above is an example of a single image from the set which was superimposed for these examples, using the same raw, unadjusted colors recorded by the camera. This image has been reduced 25% in size. The surrounding diffraction pattern is visible.
This surrounding pattern can be mostly edited out for position averaging. The portion which cannot be removed can be partially accounted for by subtraction and/or splitting the image.