Title :
Identification of time-varying resistance during welding
Author :
Garza, Frank J. ; Das, Manohar
Author_Institution :
Oakland Univ., Rochester, MI, USA
Abstract :
The time-variation of resistance across the workpieces (weldment) during a weld, provides a nondestructive indication of weld quality. However, inductive noise, power factor, time-varying parameters and random noise complicate the measurement of resistance. Existing measurement methods have focused on the noise and bias problem, approximating the time variation by a single resistance value for each half cycle (8.33 msec). The goal of this paper is to present a method to identify time-varying resistance (and mutual inductance), with finer resolution and accuracy than existing methods. Sampled voltage and current measurements are used with a single-input (current), single-output (voltage), continuous, time-varying model
Keywords :
electric resistance measurement; inductance; mean square error methods; process monitoring; recursive estimation; resistance welding; time-varying systems; MSE criterion; SISO continuous time-varying model; forgetting factor; mutual inductance; nondestructive indication; parameter estimation; recursive least squares; resistance spot welding; sampled voltage-current measurements; system identification; time-varying resistance; weld circuit measurements; weld quality; Current measurement; Electrical resistance measurement; Electrodes; Force measurement; Inductance; Resistance heating; Surface resistance; Time measurement; Voltage; Welding;
Conference_Titel :
Instrumentation and Measurement Technology Conference, 2000. IMTC 2000. Proceedings of the 17th IEEE
Conference_Location :
Baltimore, MD
Print_ISBN :
0-7803-5890-2
DOI :
10.1109/IMTC.2000.848729