Homogeneity of the linear thermal expansion coefficient of ZERODUR measured with improved accuracy

For future extremely large telescope projects like OWL or TMT with at least several hundreds of mirror blanks the homogeneity of the coefficient of linear thermal expansion (CTE) within a single blank is an important issue. The telescope designers are not only interested in the global CTE homogeneity but also in measuring the axial CTE gradient to the highest precision. It has been proven in the past in many projects like GTC and Keck that ZERODUR(r) itself is a material of highest homogeneity even in large dimensions and huge quantities. About 95.5% of all 2m class mirror segments of all projects exhibit a peak to valley homogeneity of better than 0.015*10-6K-1. The actual homogeneity of the material is even better because the results so far are largely influenced by the restrictions of the CTE measurement repeatability in the past. This paper introduces an advanced method for the measurement of the CTE of ZERODUR(r) exhibiting a significantly improved reproducibility. The dilatometer setup was especially optimized to cope with the demand of highly accurate homogeneity measurements of 2 m class ZERODUR(r) segments for giant astronomical telescopes. Detailed measurement results out of a single 1.5 m class ZERODUR(r) segment based on the current state of production will be shown. The results show CTE distributions in radial, angular and axial direction. SCHOTT has already improved the production capacity for ZERODUR(r) immensely, thereby the results represent the current status of quality of the available mass production facilities at SCHOTT.