Optical path distortion data taken as a function of time for 30 watts of power flowing through a nichrome ring of radius 9.5cm, which is suspended 7.1cm off of the face of the optic. The data are azimuthally averaged over the apparatus' clear aperature to yield the resulting curves. Error on the curves is typically +/-10nm, dominated primarily by systematic errors (imperfect optics along with slight misalignments which creep into the system during the 4 hour data run). The peak-to-valley error in optical path across the clear aperature of the wavefront sensor is typically <5nm.

Corresponding curves found by numerically solving the heat equation (in two dimensions) for the calculated heating pattern of the ring, then integrating over the optical axis to find the resulting optical path distortion. The model uses an emissisvity of 0.9 (the ring temperature at 30 watts is ~1000 degrees Kelvin, hence the radiation is primarily in the far IR), and an index derivative (dn/dT) of 11.8e-6 per degree Kelvin.

Time series for the optical path distortion at a radius R=2 cm on the face of the optic of the data shown above.