标题：基于探地雷达检测道路和桥面的深层特征及病害
Detecting Subsurface Features and Distresses of Roadways and Bridge Decks with Ground Penetrating Radar at Traffic Speed
作者：Hao Liu *a, b, Ralf Birken c, Ming L. Wang b
期刊：Proc. of SPIE Vol. 10168
abstract
This paper presents the detections of the subsurface features and distresses in roadways and bridge decks from ground penetrating radar (GPR) data collected at traffic speed
This GPR system is operated at 2 GHz with a penetration depth of 60 cm in common road materials. The system can collect 1000 traces a second, has a large dynamic range and compact packaging. Using a four channel GPR array, dense spatial coverage can be achieved in both longitudinal and transversal directions.
特别指出是以交通流速度搜集数据。交代一些参数，例如2GHZ，以及天线深度60厘米，每秒1000trace。
The GPR data contains significant information about subsurface features and distresses resulting from dielectric difference, such as distinguishing new and old asphalt, identification of the asphalt-reinforced concrete (RC) interface, and detection of rebar in bridge decks.
指出探地雷达能够识别因介电常数差异而显著的深层特征和病害。比如，区分新旧沥青，沥青混凝土界面的识别和桥面钢筋识别。
For roadways, the new and old asphalt layers are distinguished from the dielectric and thickness discontinuities. The results are complemented by surface images of the roads taken by a video camera. For bridge decks, the asphalt-RC interface is automatically detected by a cross correlation and Hilbert transform algorithms, and the layer properties (e.g., dielectric constant and thickness) can be identified. Moreover, the rebar hyperbolas can be visualized from the GPR B-scan images. In addition, the reflection amplitude from steel rebar can be extracted. It is possible to estimate the rebar corrosion level in concrete from the distribution of the rebar reflection amplitudes.
reflection amplitude 反射振幅
rebar corrosion level 钢筋腐蚀水平
1 introduction
Identifications of subsurface features and potential distresses are important and necessary for health monitoring on roadways and bridge decks.
健康监测 health monitoring
For example, different types of asphalt on pavement roads can be distinguished, such that we can know the paving history of the roadways.
On the other hand, the damages of bridge decks are the most expensive for bridges, such as rebar corrosion and its consequences.
To map the steel rebars and estimate their corrosion level, the prerequisite is to detect the rebar locations in bridge decks and to extract rebar reflection amplitude. In this paper, large GPR data sets are collected by a multi-channel GPR system at traffic speed on urban roads, interstate highway and bridge decks. The authors focus on the identifications of subsurface features from such GPR data sets
为了定位钢筋位置并且估计其腐蚀水平，先决条件是监测桥面钢筋位置并且提取钢筋的反射振幅。
GPR is a geophysical method that can use radar wave to map the subsurface of roadways and bridge decks [1]. It has several advantages over the other methodologies: it is non-destructive, cost-effective, efficient, has no impact on the structures of pavement roads or bridge decks, and carries significant amount of information about the subsurface conditions.
探地雷达的优势：无损、高效、提供海量的浅层信息。
It is demonstrated that GPR has good potential to identify subsurface characteristics, such as asphalt thickness [2, 3, 4, 5, 6, 7], voids [8, 9], ==rebar locations ==[10, 11], and railroad blast [12, 13]. Nevertheless, the use of GPR cannot meet the demand to get the comprehensive subsurface conditions of roadways and bridge decks in network level. This is mainly due to the limit of the survey speed of the current commercialized GPR systems, such that only limited amount of information can be provided.
应用场景：沥青厚度监测、脱空监测、钢筋定位等
研究局限性：由于调查速度的局限性，难以获取更多有效信息
Therefore, a four-channel GPR system is developed by the VOTERS (Versatile Onboard Traffic-Embedded Roaming Sensors) project led by Northeastern University, Boston, MA. Each channel of the GPR system is named Channel 0, 1, 2 and 3, from left to right (Figure 1). It can be operated very fast (1000 traces per second per channel), has compact packing (26cm 22cm 8 cm), light weight (~1 kg per channel), and low power draw (~ 10 W). With this GPR system, dense coverage can be achieved in the longitudinal direction —— a trace can be collected every 2.7 cm at the survey speed of 100 km/hr. This multi-channel GPR system has been used for mapping the asphalt thickness and estimating the transversal profile in city wide, and the repeatability of the GPR data is verified [7].
transversal profile: 横向轮廓
The subsurface features and distresses of pavement roads and bridge decks are identified through the GPR data from dielectric difference. As it is known, the reflection of EM waves happens at interfaces with dielectric difference. The asphalts with different ages, aggregate sizes and moisture contents may have different dielectric constants [14]. In addition, for bridge decks, the asphalt, concrete and steel rebar have distinct dielectric values [14, 15, 16]. Hence, the subsurface features with dielectric inhomogeneity can be detected from the GPR data.
This paper presents the engineering applications of GPR on subsurface feature detections from the data collected in an aircoupled mode at traffic speed. The subsurface features include thin asphalt overlay, partial-depth patch, the asphalts with different aggregate sizes, the interface between asphalt and reinforced concrete (RC), and steel rebars in bridge decks. The results indicate that the multi-channel GPR system can provide comprehensive subsurface conditions of roadways and bridge decks at the network level.
地下特征包括沥青覆盖层，部分路基的填补，具有不同尺寸的沥青，沥青与钢筋混凝土（RC）之间的界面以及桥面板中的钢筋。
2 detection of subsurface features in roadways
This section is mainly about the results to detect thin asphalt overlay and the asphalts with different gravel sizes from the air-coupled GPR data. The results are validated by drilling cores and supported by surface images taken by the video camera mounted on the VOTERS van. These features are detected by searching inhomogeneities or discontinuities in the subsurface GPR data.
drilling cores:钻孔
第二节主要监测沥青层厚度和种类。检测结果通过钻心取样和图片进行验证。这些特征是通过搜索地下GPR数据中的不均匀性或不连续性来检测的。
总结
本文主要基于探地雷达监测到的桥面和道路浅层b-scan 图像，总结沥青层新旧、厚度、砾石大小、裂缝修补补丁等图像特征，建立一系列识别得规则。
