Loren Chang’s first undergraduate research experience led to a dead end when he learned that plasma physics was not his passion. However, when he began work in the wavefront sensing lab under Professor Chanan, he knew he had found a match to his interests. Loren’s research in the testing of telescope phasing algorithms developed during his sophomore year and eventually resulted in his current pursuit of graduate studies in atmospheric remote sensing systems at the University of Colorado, Boulder. In his free time, Loren enjoys a wide range of leisure activities including astronomy, computers, swimming, cycling, hiking, and traveling to new places.

With the introduction of large telescope mirrors comprised of many individual segments, the problem of insuring a smooth continuous mirror surface (i.e. phasing) becomes critical. The vertical displacements (piston errors) between the individual segments must be reduced to a small fraction of the wavelength of incoming light. In one proposed technique, light from the telescope mirror is split between the two arms of a Mach-Zehnder interferometer, and the two outputs are subtracted from one another. The piston error of each mirror segment can then be determined from the resulting fringes that occur at the segment edges of this “differential interferogram.” By implementing this algorithm via computer simulation, it can be demonstrated that the dependence of fringe intensity on the piston error is a sine function when the light is monochromatic. In addition, the original algorithm may be modified to work using wavelengths of light in a Gaussian bandpass. Unlike the monochromatic case, the intensities in the broadband case behave as a sine function modulated so they decay to zero as the piston error increases. This allows for a more robust algorithm that is much more effective at detecting piston errors greater than one wavelength.

Gary Chanan School of  Physical Sciences

Loren Chang’s research was concerned with aligning the next generation of giant optical telescopes, which will consist of segmented primary mirrors with hundreds or even thousands of mirror segments. We know how to solve this problem for the current generation of large telescopes, such as the two Keck telescopes on Mauna Kea. However, these have only 36 segments each, and it is not clear that the same techniques will work with much larger numbers of segments. A group at the European Southern Observatory (ESO) has developed an alternative technique. Loren confirmed that the ESO simulations were substantially correct (already an admirable achievement for an undergraduate thesis). However, he went beyond this and showed how an instability in the ESO method could be addressed by using “white light,” in which multiple wavelengths are sampled in parallel, rather than in series. Loren’s work (among other ideas) will be featured in a collaboration meeting between UC and ESO astronomers planned for this summer.