Measurement of air void content in hot mix asphalts: method and core diameter dependence

Assessing the air void content of hot mix asphalts (HMAs) is a vital and crucial factor for quality control
and quality assurance (QC/QA) procedures. Layer geometry, mix composition, aggregate and asphalt binder
specific gravities, and compaction characteristics (i.e., energy) together determine the air void content.
The experimental measurements used to determine specific gravity may influence the results.
Asphalt samples of a given core diameter and thickness must be tested to verify compliance with contract
requirements and specifications. To decrease the time required for the survey and to reduce costs, it
would be advantageous to collect smaller cores. The measurements are essential for assessing the economic
value of the HMA as a function of air void content. Payment may be adjusted according to the
extent of on-site air void content, and the costs associated with the QC/QA procedures depend on the
method used. In light of these concerns, attention was focused on the assessment of the impact of measurement
methodology and specimen diameter on HMA air void content. The analysis showed that both
measurement methodology and core diameter can substantially affect the specific gravity determination
and, as a consequence, the air void content estimate. The dependence of the results on measurement
methodology and core diameter was discussed.
Method-derived and diameter superposition effects substantially biased the results. Differences in estimates
of the air void content were as high as 10%. Large diameter, dimensionally derived (DIM) values
and small diameter, paraffin-coated values (FIN) were found to be the most distant and least comparable
measurements in the factorial plan of experiments. Based on the results, it is imperative that very specific
guidance for the derivation of AV be provided to minimize the potential for between-user differences and
QC/QA consequences (acceptance procedures, pay adjustment). Further evaluation of the predictive capabilities
of the two-zones model is needed. Further enhancement of the micro and macrotexture analysis
of lateral and upper surface is recommended.