Peter A. A. M. Somers

Maintaining sub-pixel alignment for a single-camera two-bucket shearing speckle interferometer

In a phase stepped speckle interferometer, phase steps can be realized temporally or spatially. The spatial approach has the advantage of simultaneous acquisition of all phase stepped interferograms which eliminates phase errors due to phase changes that can occur between phase steps when phase is stepped in time. For spatially phase stepped systems based on multiple optical channels, it is important that interferograms are well aligned to the others. Alignment can be achieved and maintained by a system that does not rely strongly on high mechanical stability but is based on measuring the alignment errors and correcting them. The alignment system presented in this work uses speckle correlation to quantify alignment errors with sub-pixel accuracy, and comprises a piezo-electric controlled mirror to achieve precise alignment. The alignment method has been implemented for a two-channel shearing speckle interferometer. It allows for precise initial alignment both for translation and rotation of the two phase stepped speckle patterns on the CCD. Translation adjustments to maintain sub-pixel alignment during operation can be realized very quickly, preceding a test session. The importance of adequate alignment, and the effects of insufficient alignment, are illustrated by experimental results.

Polarization plane rotator used as a phase stepping device in a 2-channel shearing speckle interferometer

A polarization phase stepping method is presented based on the use of a polarization plane rotator that establishes a relative phase shift between two counter-rotating circularly polarized beams. The phase step can be made relatively accurate, since it just depends on the accuracy with which the rotator is manufactured, and not on its orientation. The phase stepping method has been implemented in a single-camera two-channel shearing speckle interferometer, with two optical channels, and a relative phase step of π/2 between them.

A two-bucket phase-stepped shearing speckle interferometer: Why does it work?

It is generally acknowledged that it takes at least three phase-stepped speckle patterns to obtain quantitative phase information. However, if only a phase change has to be determined, a two-bucket approach can be followed, under certain conditions. The background of the two-bucket algorithm, and the requirements with respect to the use of it, are explained. Examples of modeled and measured phase stepped speckle intensity data, showing numerical instability when conditions are not met are presented. It is also shown that two phase-stepped speckle pattern pairs, taken before and after an event can be sufficient to determine phase changes invoked by the event.

Handling unfavourable polarization states in a polarization-based shearing speckle interferometer

A polarization-based shearing speckle interferometer requires the incoming scattered light to have proper polarization characteristics: the light should be equally distributed over two orthogonal polarization states in order to obtain two interfering beams with equal intensities and the highest possible modulation. Many surfaces scatter linearly polarized light randomly, providing equal intensities for the interfering beams on average. However, when a diffusely reflecting metallic surface is illuminated by a linearly polarized laser beam, the polarization direction of the illuminating beam is retained to a large extent in the scattered light. As a result light entering the interferometer will not be randomly polarized, which may lead to low modulation due to unbalanced object and reference beams. A solution that handles the problem effectively is proposed. It consists of a quarter-wave plate positioned in front of the interferometer, oriented at 45°. It is shown that unfavourable predominant polarization states encountered when testing unprepared metallic surfaces can be converted into favourable ones, thereby obtaining well balanced object and reference beams, irrespective of the polarization direction of the incoming light.

Temporal phase unwrapping with two or four images per time frame: a comparison

A simple scheme is presented that enables the measurement of phase changes by recording only two phase stepped images per time frame. Such a scheme is useful for the detection of defects or to monitor vibration characteristics in constructions. The differences between the here presented method and the four recordings per time frame method are discussed. Finally some experimental results are shown.

Implementation of temporal-phase unwrapping in a real-time phase-stepped shearing speckle interferometer

The implementation of temporal phase unwrapping within a real- time phase stepped shearing speckle interferometer is presented. Speckle phase maps obtained with a shearing speckle interferometer, representing an object before and after deformation, reveal sub-surface defects or damage, after subtracting the images. Phase information is only known modulo 2(pi) , and has to be unwrapped for a true representation of the deformation map. Results are easier to interpret during the deformation process when the images are unwrapped; phase unwrapping also facilitates automatic detection of suspected areas. It is shown that, compared to the more common spatial* phase unwrapping methods, temporal phase unwrapping is much faster, and can be implemented in a real-time system. In addition, this method offers an increased measuring range, reduces sensitivity for speckle decorrelation, handles discontinuities in the object, and is very reliable, even when used with noisy data. Processing strategies for the selective removal of unwanted image components, and for automatic defect detection are presented. Examples of results obtained for artificial defects in metallic and composite aeronautical components are shown. Measurements have been carried out with our phase stepped shearography system under laboratory and industrial conditions, showing improved performance under non- ideal conditions. It has been shown during these experiments that shearography cannot only be used for detection of defects, but also for characterization.

Three-bucket quadrature phase stepping in a shearing speckle interferometer

Phase stepping algorithms are mostly based on three or more interferograms that can either be acquired sequentially, involving some temporal phase stepping mechanism, or in parallel. When a phase step is applied between acquisitions, object phase changes may cause phase errors when calculating phase. A new system that allows the object to change in between arbitrary temporal phase steps is proposed. It comprises a relatively simple polarization based two-channel speckle interferometer that acquires two π/2 phase stepped interferograms simultaneously with a single camera. This quadrature pair is phase stepped with a temporal phase stepping system that is also polarization based. Simulations and experimental results are presented that illustrate the improvements achieved with quadrature phase stepping compared to results obtained with the two-channel speckle interferometer operated without additional temporal phase stepping.

Determination of modulation and background intensity by uncalibrated temporal phase stepping in a two-bucket spatially phase stepped speckle interferometer

A new phase stepping method is presented, based on the combination of spatial and temporal phase stepping. The method comprises one fixed spatial phase step of π/2 and two arbitrary, unknown temporal phase steps. The method prevents phase errors caused by phase changes during temporal phase stepping. It is therefore in particular useful for dynamic applications, but will also improve system performance for quasi-static applications in non-ideal environments.

Speckle Decorrelation: Observed, Explained, and Tackled

It is commonly known that speckle decorrelation limits the possibility to measure large deformations by electronic speckle pattern interferometry (ESPI). We present here a method to lift this limitation and show that ESPI can now be used to measure large deformations, even in a hostile environment. An experimental result is shown that clearly demonstrates the problem due to speckle decorrelation and it is shown that we can remedy that problem. Some simulations are presented showing the cause of the observed problems.