Intelligent Asphalt Compaction Analyzer Abbreviated Work Plan Description
The Intelligent Asphalt Compaction Analyzer (IACA)1 is a device based on Neural Network Technology that can measure the density of an asphalt pavement continuously in real-time during its construction. This technology was developed at the University of Oklahoma (OU) and was tested to a limited extent in the field. These field trials demonstrated that, after calibration, the IACA can measure the density in real-time with accuracy comparable to existing point-wise measurement technologies (e.g., a nuclear density gauge or a non-nuclear density gauge such as Transtech's PQI 301).
Prior to the start of this project, the IACA training and calibration techniques were performed manually, thereby limiting the utility of the IACA. Further, the IACA was implemented on a computational platform that was primarily designed for testing in the laboratory. The use of the IACA technology in the field would require the porting of the IACA application to a rugged embedded hardware that is capable of withstanding extreme vibrations and the harsh environment typically found at construction sites. Further, the development of automated, simple to use calibration techniques, and the validation of the performance during exhaustive field testing are necessary for the IACA technology to gain acceptance.
A. Phase 0 - Refinement of Research Prototype (September - December 2007)
Results for Phase 0 (full report available on website)
a. A low cost WAAS GPS receiver, Trimble ProXT10, capable measuring position with sub-meter accuracy, was integrated with the IACA and automatic calibration procedures were developed. The calibration and training procedures developed were verified using construction data from prior research.
b. Preliminary studies were conducted during the construction of a full-depth pavement at three sites (base, intermediate layers and shoulder) and the calibration procedure was validated. Delays in the work schedule prevented the team from validating the IACA on the surface course. Test results show that the density estimated by the IACA compares well with the density measured from roadway cores. Results from the 14 core locations indicate that the mean error between the IACA measurement and the actual density as measured form the core is less than 1 pcf (pound-per-cubic-feet) with a 95% confidence interval6 of less than 3 pcf.
The results of the Phase 0 of the project indicate that the IACA can estimate the density of the asphalt pavement with an accuracy that is suitable for quality control during the construction of the asphalt pavement. The integration of low-cost GPS sensor also improves the commercial feasibility of the technology. Based on the results of Phase 0, the project team is ready to start the next phase of the project, i.e. the development of a IACA prototype as described in the Phase 1 of the proposal. It is anticipated that in the Phase 1 of the study, the calibration technique outlined in this report will be rigorously evaluated at several construction sites involving overlay of base, intermediate, and surface courses, as well as the construction of full-depth asphalt pavements across the state of Oklahoma.
B. Phase 1 - Development of Commercial Prototype (February - November 2008)
In this phase of the proposed development, the research prototype will be used to develop a commercializable prototype of the IACA. The first task is the selection of an electronic module in a rugged enclosure that is suitable for extended operations in the field. The project team will coordinate with MathWorks, Inc., and Volvo Road Machinery to determine a target electronic control module (ECM). Simultaneously, additional sensors to record properties (e.g., temperature) of the asphalt mat will be incorporated into the system. Extraction and testing of the roadway cores will be undertaken by EST Inc., an independent testing agency. At the completion of this task, a rugged commercial prototype will be available for systematic testing of the repeatability and accuracy of the invention.
The accuracy of the density estimates and their repeatability will then be systematically tested over a period of 9 months. Sites with different mix designs, construction types, lift thicknesses, etc. will be identified and the performance of the device will be studied carefully. Appropriate refinement in the calibration and operation procedures will be made to meet the desired performance. The use of calibration data from different construction sites on the density measurements will also be studied. This is required because in current construction practice, a control strip is constructed in less than 10% of all the projects in Oklahoma.
B.1 Project goals for Phase 1
The following are the goals for this phase of the project:
- Determine an electronic platform that is tested to SAE J14552 standard to withstand vibrations and temperature variations encountered in heavy duty off-road applications. Port the IACA application to this platform and test functionality in the laboratory.
- Incorporate temperature sensor, odometer, vibration switch, travel direction sensor, GPS sensor and test the commercial prototype during compaction in the field.
- Identify five sites for systematic testing of the IACA. The sites selected will be representative of typical construction encountered in the field and will involve full depth construction as well as rehabilitation of existing pavements. The sites will also be selected to cover different soil types to address varying liquid limit and plasticity indices.
- The performance of the IACA will be validated during the construction of the base, intermediate, and the surface layers of the asphalt pavement.
C. Phase 2 - Field Validation of the Commercial Prototype
(November 2008 - October 2009)
The successful deployment of the technology will involve systematic and rigorous testing of the prototype. To accomplish this objective, a total of five commercial units will be developed and installed on different vibratory compactors owned by Haskell Lemon. The performance of the units in the field will be studied over a period of 10 months. Comparative studies will also be conducted on the quality of construction with and without the use of the developed IACA technology. The impact of the technology on the productivity of the crew will also be analyzed. In this phase of the research, the team will coordinate the activities very closely with the contractor and with the Oklahoma Department of Transportation.
C.1 Project goals for Phase 2
The following are the goals for this phase of the project:
- Build 5 prototype units and test their functionality in the lab. Install the tested units on test compactors.
- Coordinate with FHWA and Volvo Road Machinery for field training and demonstration of the IACA.
- Validation of the IACA during construction of asphalt pavements across Oklahoma. Perform statistical sampling to estimate accuracy of the measured densities. Identify key factors affecting the performance of the IACA. Refine the implementation if necessary to account for variability in the field.
- Study the variability in the calibration between sites.
- Determine variation between minimum and maximum density achieved over 100 feet of construction with and without the use of IACA.
- Repeat step (d) at each construction site.
Selection of Test Sites
The validation of the Intelligent Asphalt Compaction Analyzer in Phase 1 will be carried out over several test locations during the months of April - December 2008. The test locations will be selected to offer a good mix of construction projects: new construction vs. rehabilitation of existing asphalt pavements, construction of surface course vs. multi-lift full depth pavement, different site geomorphologies, and different asphalt mixes. The validation will be performed during the construction of each asphalt layer (base, intermediate, and surface) and the accuracy of the IACA estimated density will be verified through the density of roadway cores measured using AASHTO T 1667 method. The following three steps would be undertaken to study the accuracy of the estimates and the uniformity in compaction achieved.
- Three cores will be extracted at random per sub-lot of asphalt mix (typically 10-12 cores per 1000 tons of asphalt mix). The locations of the cores will be recorded using GPS sensor and the corresponding density estimates from the analyzer will be studied for accuracy.
- Three different 300 feet stretches of the pavement will be selected and the minimum and maximum estimated density and their corresponding location on the mat will be identified. One core each will be extracted at the location corresponding to the maximum and the minimum density. The density measured from the roadway cores will be compared with the estimated density.
- Prior to the extraction of the cores, the density at each of the core locations will be measured using a non-nuclear as well as a nuclear density gauge and the correlation between the estimated density and the measured density will be studied.
A. Test Site 1. (I-35 (North of Purcell), McClain County (Oklahoma))
The project involves the construction of 3.016 miles of Interstate I-35 in McClain County, Oklahoma. The subgrade soil at this site primarily consists of grant silt loam, having two to five percent slopes. Typically, the surface layer is reddish brown silt loam (ML, CL, CL-ML) about ten inches thick, having liquid limit and plasticity index in the range of 20-30, and 1-10, respectively. The upper part of the subsoil is reddish brown and red silty clay loam (ML, CL) to a depth of 44 inches, and the lower part is red silt loam to a depth of 65 inches. The liquid limit and the plasticity index in this zone vary from 30 to 43, and 8 to 20, respectively. The underlying material is red soft sandstone.
The construction involves the cold milling and removal of 2 inches of existing pavement and resurfacing with 2.5 inches of type S3 (PG 76-28 OK) mix followed by a further 2-inch surface course of type S4 (PG 76-28 OK) asphalt mix. The shoulder reconstruction involves the removal of six feet width of the existing shoulder (ten feet) and the laying of six inch base of asphalt concrete type S3 (PG 64-22 OK), followed by a 3 inch lift of type S3 (PG 76-28 OK), and finally a 2 inch surface course of type S4 (PG 76-28 OK) asphalt mix.
The project also involves the construction of half mile full depth pavement. This portion of the construction involves the removal of the existing north and south bound roadway sections and first preparing an eight (8) inch subgrade stabilized with 12% fly ash. A separator fabric is then placed and an 8 inch aggregate base is compacted to 95%. Two lifts of 2.5 inch type S3 (PG 64-22 OK) asphalt mix are then compacted on top of the aggregate base. Finally, a 2 inch surface course of type S4 (PG 76-28 OK) asphalt mix is laid with slope rotation between -2% and -3/16 inch/feet.
B. Test Site 2. (US Highway 62, McClain County (Oklahoma))
The project involves the construction of 1.689 miles of US Highway 62, near Blanchard in McClain County, Oklahoma. The subgrade soil at this site consists mostly of Stephenville fine sandy loam, having two to five percent slopes (Soil Survey of McClain County2, Oklahoma, 1979). Typically, the surface layer is brown fine sandy loam (SC, SM, ML, CL), six inches thick and having liquid limit and plasticity index in the range of 0-30 and 0-10, respectively. The subsoil is red sandy clay loam (SC, CL) to a depth of 36 inches, where liquid limit and plastic limit vary 25-37 and 7-16, respectively. The underlying material is red soft sandstone.
The project involves the construction of a full depth pavement consisting of 4 driving lanes, a turn lane, and 8 feet shoulders on either side of the road. The existing pavement is removed and an eight inch modified (15% fly ash) is first compacted. The modified subgrade has a variable width from 68 feet in some sections to approximately 71 feet in others. The base consist of 3 inch type S3 (PG 64-22 OK) asphalt mix with a further 3 inch intermediate layer of type S3 (PG 64-22 OK) on the outer lane and type S3 (PG 76-28 OK) on the inner lane. Finally, a 2 inch surface course of type S4 (PG 76-28 OK) asphalt mix is laid.
Some sections (0.8 miles) have an eight inch surface layer of dowel jointed concrete laid on top of the base layer. The project also involves the construction of detours during the paving process. These detours are typically twenty four feet wide and comprise of two driving lanes. The detours are constructed using two 4 inch lifts of type S3 (PG 64-22 OK) asphalt mix on a six inch modified (15% fly ash) subgrade.
C. Test Site 3. (Tecumseh Road, Norman, Cleveland County (Oklahoma))
The project involves the construction of 1.881 miles of two lane road in Cleveland County, Oklahoma. The subgrade soil at this site primarily consists of Renfrow silty clay loam, with 1 to 5 % slopes (Soil Survey of Cleveland County3, Oklahoma, 1987). Typically the surface layer is dark grayish brown silty clay loam (CL) about 5 inches thick, liquid limit and plasticity index varying from 33-49 and 12-26, respectively. The subsoil is brown silty clay loam (CL) to a depth of about 9 inches, having a liquid limit of 37-49 and a plasticity index of 15-26. There is reddish brown silty clay (CL, CH) to a depth of about 27 inches, and red silty clay to a depth of about 73 inches. The soil in this region has liquid limit and plasticity index in the range of 37-70 and 15-38 respectively. The underlying material is red and light gray, weakly laminated, calcareous siltstone and shale to a depth of 80 inches or more.
The construction involves the construction of an undivided two lane extension of Tecumseh Road between NW 12th Avenue and NE 12th avenue in Norman, Oklahoma. The construction involves clearing the vegetation and stripping the top soil and re-spreading the soil to achieve the desired slope. Approximately fifty five feet of the subgrade is modified with 12% fly ash to a depth of eight inches and is then compacted. The modified subgrade is then sealed by an application of prime coat and a base layer compacted using a type S2 (PG 64-22 OK) asphalt mix. This is followed by the construction of a four inch intermediate lift using type S3 (PG 70-28 OK) asphalt mix and surfaced with a two inch type S2 (PG 70-28 OK) asphalt mix. Each driving lane is 12 feet wide and the shoulder is 14 feet wide. Left turn lanes of 12 feet width are also constructed on select sections of the road.
D. Test Site 4. ( SH-74, Oklahoma / Logan County, Oklahoma))
The project involves the resurfacing of 5.83 miles of SH-74 from 0.20 miles north of Waterloo Road in NW Oklahoma City. The subgrade soil at this site primarily consists of ashport silt loam, with a very mild (0-1%) slope, which indicates that frequent flooding is possible at this site (Soil Survey of Oklahoma County4, Oklahoma, 2007). Typically, the surface layer is reddish brown silt loam (ML, CL, CL-ML), about ten inches thick. The liquid limit and plasticity index are in the range of 22-37, and 2-13, respectively. The upper part of the subsoil is reddish brown silty clay loam (CL) to a depth of 64 inches; the liquid limit and the plasticity index in this zone vary from 30 to 43, and 8 to 20, respectively. Moreover, a different zone of subgrade soil exists along the length of the road, and the subgrade soil in this region consists of miller silty clay, with 0-10 % slopes. The subgrade layer in this region consists of reddish brown silty clay (CH, CL) to a depth of 12 inches, having a liquid limit and a plasticity index of 41-60 and 18-35, respectively. Subsurface consists of reddish brown silty clay to a depth of 84 inches. The liquid limit and plasticity index of the soil at this depth range from 35-60 and 15-35, respectively.
The roadway pavement is 32 feet wide (two 12-feet travel lanes and the remaining as paved shoulder). The whole width will be milled to a depth of 1 inch and repaved with 2.5 inch asphalt concrete type S4 (PG 70-28 OK).