Refinement of the Hot-Mix Asphalt Ignition Method for High-Loss Aggregates

Four methodologies for determining the asphalt content of mixtures containing high-loss aggregates in the ignition furnace were evaluated: the standard method using the Thermolyne furnace (control), the Troxler NTO infrared furnace, the Ontario method, and a Tempyrox glass-cleaning oven. Six aggregate sources with high ignition furnace aggregate corrections were obtained from around the country: four dolomites, a basalt, and a serpentine/chlorite. Calibration factors were determined for each method at optimum asphalt content. Additional samples were then tested at optimum plus 0.5% asphalt content, and the measured asphalt content was calculated by using the correction factor determined for that method and aggregate source. The Tempyrox Pyro-Clean furnace, commonly used for cleaning laboratory glassware, produced the lowest aggregate correction factors. The standard method and the Ontario method, both using the Thermolyne ignition furnace, produced the smallest bias or error in measured asphalt content. The standard deviation of the corrected asphalt contents for these high-loss sources was higher than the within-laboratory standard deviation reported for AASHTO T308. The only exception was the Alabama source using the standard method. The Ontario method and Tempyrox oven generally reduced the variability of asphalt content measurements for high-loss aggregates. None of the methods evaluated statistically reduced aggregate breakdown on the nominal maximum aggregate size and 4.75-mm sieves. The Ontario method significantly reduced, but did not eliminate, aggregate breakdown on the 0.075-mm sieve. The Ontario method is the best method for immediate implementation for determining the asphalt content by the ignition method for high-loss aggregates.

Evaluation of Infrared Ignition Furnace for Determination of Asphalt Content

The Troxler Model 4730 infrared ignition furnace was compared with a standard Thermolyne ignition furnace. Comparisons conducted with a single unit of each furnace type were based on the correction factor for aggregate loss during ignition, accuracy, and the variability of the measured asphalt content and aggregate degradation during ignition. Forty-eight samples representing two nominal maximum aggregate sizes (9.5 and 19.0 mm), four aggregate types (granite, crushed gravel, limestone, and dolomite), and two asphalt contents (optimum and optimum plus 0.5% asphalt content) were tested in each furnace. The results indicated that the correction factors for aggregate loss during ignition were significantly different for each type of furnace, thus requiring a separate calibration for each type of furnace. In practical terms, the differences for all but the 9.5-mm nominal maximum aggregate size (NMAS) limestone and both dolomite mixtures were less than 0.1%. The samples with the optimum plus 0.5% asphalt content were tested by using the calibration factors developed for a particular mix–furnace combination. The results were analyzed in terms of accuracy (bias) and variability (standard deviation). Neither the measured biases nor the standard deviations for the two types of furnaces were significantly different. The results obtained with four sieve sizes (NMAS and 4.75, 2.36, and 0.075 mm) were evaluated for aggregate breakdown. A comparison of the aggregate gradations recovered from both furnaces indicated no significant difference in the degree of aggregate degradation. A round-robin investigation should be conducted to confirm that the precision of the infrared furnace is similar to the precision of the standard furnace.

Effect of Lime on Ignition Furnace Calibration

This study investigated mix components, other than aggregate source, that were perceived to affect the ignition furnace mixture calibration. Four sets of experiments with one aggregate were designed and run to evaluate the effects of the amounts of lime, sulfur, calcium carboxylates, and fines. Five asphalt binders with different chemistries were evaluated. Two binders had high and low sulfur contents, respectively; two other binders contained relatively high concentrations of carboxylates. Various concentrations of hydrated lime (0 to 3 percent by weight) were added to mixtures and were found to have a significant effect on the ignition furnace correction factor for all five binders. The magnitude of the effect was large enough to cause the quality control tests to fail the tolerances established by the Virginia Department of Transportation. The lime appears to react with the sulfur dioxide (SO2) formed from the combustion of organic sulfur to generate calcium sulfate. The amount of sulfur present in the asphalt can significantly affect the ignition furnace correction factor. Initial thoughts that calcium carboxylates are reacting with carbon dioxide (CO2) to produce some form of calcium carbonate were not substantiated. The influence of carboxylic acid groups, however, does come into effect in the presence of lime. This suggests that lime reacts with the combustion gases to produce some form of carbonate. Basalt fines and portland cement did not have a significant effect on the ignition furnace correction factor. Additional testing should be conducted to assess the effects of fines representing other aggregates, particularly carbonates.