United States Department of Transportation
ROSA P serves as an archival repository of USDOT-published products including scientific findings, journal articles, guidelines, recommendations, or other information authored or co-authored by USDOT or funded partners.
As a repository, ROSA P retains documents in their original published format to ensure public access to scientific information.
i
Use of SmartRock Sensors to Monitor Pavement Condition for Supporting Maintenance Decision Making
-
2022-04-01
-
[PDF-5.29 MB]
Details:
-
Creators:
-
Corporate Creators:
-
Corporate Contributors:
-
Subject/TRT Terms:
-
Publication/ Report Number:
-
Resource Type:
-
Geographical Coverage:
-
Edition:Final Report 03/1/2019 – 03/1/2022
-
Corporate Publisher:
-
Abstract:The in-situ modulus of asphalt mixtures directly determines the stress-strain relationships of asphalt pavement under external loading and its structural conditions. The evaluation of dynamic modulus over time has also been related to the deterioration of pavement performance. Although crucial, accurate determination of in-situ dynamic modulus has always been difficult. The most direct method is to take field cores and perform laboratory dynamic modulus tests, which damages the pavement. Non-destructive testing, as with the Falling Weight Deflectometer, can also obtain the modulus properties of the pavement periodically, yet the modulus data are scarce and the data collection process is time- and labor-intensive. With recent advancements in data science and sensing technologies, it is technologically promising to develop an efficient and reliable method of estimating the in-situ modulus; the development of such a method became the motivation of this study. This project studied the methodology of using particle-size wireless sensors to conduct in-situ dynamic modulus tests and collect data under vehicular loading. It then applied an artificial neural network (ANN) model to predict the in-situ pavement modulus and quantify vehicular speed. A variety of engineering responses, including triaxial stress and Euler angle, loading frequency, and pavement temperatures, were collected by the wireless sensors in three pavement sections and used as input data to develop the ANN model. Laboratory dynamic modulus tests and MMLS3 tests using the same material as the paving projects were also performed on specimens with embedded wireless sensors. Those laboratory data, in combination with a small portion of randomly selected early-stage field sensing data, were used as the training dataset for the development of the ANN model. The remaining field data were used to test the ANN model, using the laboratory dynamic modulus master curve as a reference. The results show that the developed ANN model, when adequately trained with particle-level sensing data, is feasible and robust to predict the in-situ dynamic modulus of asphalt pavement. Future studies are recommended to include more data on the material and pavement structure variations over extended service lives so that the in-situ modulus can be used to assess pavement conditions and support maintenance decision-making.
-
Format:
-
Funding:
-
Collection(s):
-
Main Document Checksum:
-
Download URL:
-
File Type:
