Experiments were conducted to investigate the destruction of theice-substrate bonds by chemical undercutting. the undercutting experiments determined the undercut area, as a function of time, of three deicers at three temperatures on highway core samples and laboratory-produced specimens. tests were conducted at temperatures of 25 deg f (-4 deg c), 15 deg f (-9 deg c), and 5 deg f (-15 deg c) using all the substrates in the as-received, but cleaned, condition. the three deicers--nacl, cacl2, and ethylene glycol--were used for all temperature-substrate combinations. the substrates included portland cement concrete, dense-graded asphalt, open-graded asphalt, rubber-modified asphalt core samples, and laboratory-produced specimens made of portland cement concrete and dense-graded asphalt. undercutting action was also investigated on smooth highway core samples and smoothlaboratory-produced specimens at selected temperatures. linear regression models with good predictive power were developed to predict the undercutting behavior of deicer material for given combinations of pavement type, surface condition, and temperature as a function oftime. using these regression models, the following conclusions weredrawn: (a) at 25 deg f, sodium chloride produces larger undercut areas than does calcium chloride when applied to as-received core samples of dense-graded and open-graded asphalt; (b) however, at 25 deg f, calcium chloride produces more undercutting action than does sodium chloride when applied to as-received core samples of portland cement concrete and rubber-modified asphalt; (c) at 15 deg f and 5 deg f, calcium chloride produces more undercutting than does sodium chloride for all four as-received pavement core samples; and (d) at 25 deg f and 15 deg f, the as-received laboratory-produced specimens tended to be undercut more extensively than the corresponding core samples at these temperatures. this paper appears in transportation research record no. 1304, highway maintenance operations and research 1991.
Abstract