Evaluation of Non-Nuclear Density Gauges for HMAC: Year 1 Report (0-4577-1)

This report describes research efforts to determine if commercially available non-nuclear density gauges can be used for TxDOT's density profile and joint density testing procedures. Researchers conducted testing with a Troxler 3450 nuclear gauge (operated in the thin-lift mode) and the Pavetracker (PT) and Pavement Quality Indicator (PQI) non-nuclear gauges. In a laboratory setting, research showed all the gauges could be affected by mix temperature, where gauge readings typically decreased with decreasing mix temperature. All gauges' readings were also impacted by moisture, with the nuclear gauge least impacted. The precision of all gauges in the lab was good, with standard deviations below 0.5 pcfwith the non-nuclear gauges and less than 1.0 pcfwith the nuclear gauge. Field-testing showed the PQI was a suitable alternative to the nuclear gauge for density profiling and joint density testing. The Pavetracker performed erratically and should only be used for density profiling if it is calibrated to the mix; this gauge should not be used for testing joint density acceptance. All the gauges could exhibit bias in the field, and due to the sporadic nature of observed mean errors, gauge bias could not be estimated. If the gauges were unbiased, the following average levels of accuracy were indicated from the projects tested: Troxler 3450 nuclear gauge (in thin-lift mode): :t4.1 pcf; Pavetracker: :t5.7 pcf; PQI: :t2.6 pcf. All the gauges should be modified to include a slope calibration function for better accuracy. The Pavetracker performed erratically in the field; no additional testing should be performed with this gauge unless an upgraded model is obtained.