An experimental program was conducted to investigate the noncontact filter paper technique for total suction testing. Calibration curves for seven different batches of Whatman #42 filter paper were obtained to evaluate the variation in the calibration characteristics from one batch of paper to another. Additional tests using salt solutions for controlling suction were conducted to evaluate the measurement uncertainty as a function of suction magnitude. Calibration curves for the various papers varied significantly from batch to batch. This variation can be as much as 11% for filter paper water content, which results in a potential measurement error as high as 92% for suction. Scatter in the measurements appeared to increase with decreasing total suction. At relatively high values of suction (32,000 kPa), the standard deviation for five independent measurements was as much as 11.6% of the mean. At relatively low values (270 kPa), the deviation appeared to increase to 23.2% of the mean. One hundred measurements over a wide range of values for an expansive clayey soil confirmed the observation that the scatter in the measurements increases as total suction decreases.
Abstract