Measurement uncertainty in gear metrology

Abstract

Gears play an important role in mechanical power transmission systems. They enable the prime mover characteristic (a gas turbine for example) to be matched to the characteristic of the driven load (say, a slow speed propeller), thus reducing the cost of both manufacturing and operating the system. The customer requirements for higher power density and lower noise demands more accurate gears. This imposes more stringent requirements on the measuring equipment that controls the quality of the manufacturing machines. Many gears have flank form and tooth spacing tolerances that are less then 10μm, and according to the so called `Golden rule', measuring equipment on the shop floor should have a measurement uncertainty of between 1 to 2μm. These are stringent requirements that demand the highest standards of metrology. Thus the need to accurately quantify the measurement uncertainty of inspection machines is of paramount importance if costly mistakes are to be avoided. The work reported in this thesis was completed as part of the activities undertaken by the author in his role as head of the UK National Gear Metrology Laboratory (NGML). The laboratory is accredited by the United Kingdom Accreditation Service (UKAS) for gear measurement and on-site calibration of gear measuring machines. The work is mainly experimental in nature. In fact, much of what is reported is centred on work undertaken with two artefact sets: one set consisting of 100mm diameter lead and profile artefacts and a second set of 200mm diameter artefacts. These gear artefacts are probably the most valuable in the world because of the volume and quality of the calibration data associated with them.