Angle calibration on precision polygons
Project Description
Two precision polygons, one with 24 and the other with 7 faces, will be circulated among the participants. Standard calibration methods using either one autocollimator and an index or measuring table or two autocollimators and a rotary table should preferably be applied. It is recommended to measure the polygons also in the reversed (upside down) position to detect optical asymmetry errors in conjunction with the autocollimator(s). Harmonic errors of the circular divisions can be derived from the results with the 24-faced polygon, whereas the 7-faced polygon can also reveal shortrange calibration and interpolation errors of the instruments. Additional interferometric measurements of the polygon face flatness will be performed by the PTB.
Optionally, only 12 faces of the 24-faced polygon can be measured.
*BNM-LNE will join the project at the end of 1997.
Final Report 2000-04-12
Two precision polygons, one (P7) with 7 faces and the other (P24) with 24 faces, were circulated among the participants. The angle measurements were carried out in both the normal and the inverted (upside down) position of the polygons. In several cases, the measurements on P24 were restricted to only 12 faces. The associated standard uncertainties of the results reported ranged from 0,05 to 0,40 arcsec for P7 and from 0,02 to 0,45 arcsec for P24. For evaluation, these uncertainties have been enlarged by the standard deviations of the differences between the two polygon positions as measured by each participant. These enlarged uncertainties are between 0,06 and 0,40 arcsec for P7 and between 0,02 and 0,45 arcsec for P24.A mean of all results, weighted by the inverse square of the enlarged uncertainties, was calculated for each polygon face as a reference value of the reduced angle deviation. The mean values of the results of the two polygon positions deviate from the reference values in a range between –0,32 and 0,32 arcsec in the case of P7 and between –0,57 and 0,84 arcsec in the case of P24.
The distribution of the deviations revealed rather large systematic influences for P24, which were nearly identical for three results. Additional independent investigations in the pilot laboratory proved that this effect was attributable to a certain autocollimator type which was used in the respective laboratories. This could also be confirmed by the En-criterion and the Rb-ratio consistency test evaluated with the results.
The main conclusion of this intercomparison is that precision angle measurements on polygons are largely dependent on optical sensing characteristics of the autocollimators in conjunction with flatness errors of the polygon faces. It remains, however, difficult to predict or correct this systematical influence.
Further information is available in a detailed report of the coordinator, issued by the PTB.