Resistance strain gauges for the measurement of steady strains at temperatures above 650°C

Abstract

The paper considers the problem of obtaining steady strain data using electrical resistance strain gauges at temperatures above 650°C, in the presence of moderate temperature gradients, from turbojet components operating in oxidizing environments. The characteristics of forty-six metal and alloy systems are briefly outlined and the properties of platinum, and alloys of face centred cubic platinum metals and metals in the VIa group of the periodic table of elements are discussed in detail. The behaviour of colloidal sols, acid and alkali bonded ceramic cements and flame-sprayed oxides at high temperatures is considered and their effect on the short and long term stability and strain sensitivity of strain gauges presented. Only a limited number of pure oxides possessed an adequately high electrical resistivity at the higher temperatures for use as strain gauge bonding media. Such oxides, when flame-sprayed on to suitably prepared surfaces, developed a bond of sufficient shear strength to transfer the applied strain in a consistent and repeatable manner to the gauge filament. The porosity was low enough to offer some degree of oxidation protection to the strain sensitive grid. The temperature coefficient of resistance, drift rate and strain sensitivity of flat grid, dual-element gauges bonded in this manner to Nimonic specimens are shown and it is concluded that grids of this type, wound from platinum and a commercially available platinum-tungsten alloy, would permit the detection of steady strains within an accuracy of about ±0.5 ton/in2 over the temperature range 600—850°C and within ±1.0 ton/in2 over the approximate range 600—950°C from Nimonic specimens, provided the mean gauge temperature could be determined. Methods of determining this temperature in the presence of moderate temperature gradients along the longitudinal axis of the gauge are discussed and some field data are presented. Drift rates equivalent to less than 0.1 ton/in2/h were noted. It is thought that a coil wound gauge would offer improved performance over a flat grid gauge for temperature gradients across the element, but the problem of providing a suitable former has not been resolved satisfactorily.