"Measurement uncertainty in gear metrology (lies, damn lies and statistics)"
Rob Frazer (MSE: Design Unit)
16 May 2007
11am, Stephenson Building, room F16 (first floor)
All measurement processes are prone to errors, even those in the National Gear Metrology Laboratory here in Newcastle University. Gears require expensive, complex instruments for effective measurement but to properly interpret the results we need to estimate the uncertainty of the measurement process. This seminar will discuss the calibration of complex measuring instruments, review conventional methods of estimating measurement uncertainty, focus on some of the limitations and demonstrate how some these can be overcome by applying a Monte Carlo Simulation (MCS) process to model the measurement process. Results will be presented from international comparisons involving the UK with Japan, Germany and the USA. Most of the processes discussed will be applicable to any measurement or test process.
Rob Frazer has worked in the gear industry for over 20 years. He is a senior engineer in the Design Unit and is head of the UKs National Gear Metrology Laboratory, here st Newcastle University. Before this he worked at Reliance Gears in Huddersfield and served his time as a technician apprentice with GEC Turbine Generators in Stafford. He knows nothing about statistics but does not let that stop him applying it on a daily basis.