7. Survey and monitoring techniques, diagnostics

7.1. General

Road condition monitoring techniques can be generally described as a) functional condition monitoring, that measures the service level to the road users and b) structural condition monitoring which can be related more to the asset value of the road. Road service level monitoring techniques have been earlier described in the chapter 3 wearing course, so this chapter will discuss more about the structural monitoring techniques of gravel roads and forest roads and how these results can be used in the strengthening design. Some of the structural condition techniques are also described in chapter 5 “Roads on weak subgrade”, chapter 5 “Drainage” and Chapter 10 “Seasonal changes”.

The ROADEX webinar presented on 14 December 2022 gave a brief summary of the new technologies in gravel and forest road surveys and this can be seen in the following link.

7.2. Structural and functional condition

Gravel or forest road structures and their quality beneath the wearing course are normally measured only when there are already some visual problems that could be related to poor bearing capacity. In forest roads structural condition surveys should be made if there are plans for especially big harvesting operations in the area that require continuous timber haulage without accessibility problems.

7.2.1. Excavator surveys, sampling

The traditional way to survey a gravel road or forest road structure thickness is to open the road structure with an excavator on a few points on a cross section or opening the whole cross section. This technique is very usual for instance to define if the problem can be related to Mode 1 or Mode 2 rutting. The layer thickness can also be verified and samples taken using a canvas sheet and shovel.

A photograph of a cross section from a well performing forest road built over peat. There are only slight indications of Mode 1 rutting taking place on the top of road structure. This can be seen as a slightly thicker wearing course on the wheelpaths.

A typical road cross section indicating Mode 2 rutting where the displacement has taken place at the road structure subgrade interface.

Measuring the structure thickness at different points on the road cross section and taking a sample using a canvas sheet and shovel for laboratory testing if needed.

7.2.2. Ground Penetrating Radar

While excavator pits can give good information at specific points, the benefit of GPR (Ground Penetrating Radar) is the continuous profile that is created along the gravel road or forest road. However the problem is that these structures can change substantially in the transverse direction of the road. That is why data collection and analysis using a 3D GPR system enables a full 3D analysis to be made of the gravel road structures. Cross sections with GPR also give valuable information about the road structures in 3D. More information about the theory of GPR technology is given in ROADEX “Permanent Deformation” eLearning package.

A typical 2D GPR system for gravel road and forest road surveys consisting of a 1-2.5 GHz air coupled antenna to measure thinner layers and their properties on the top part of the road structure, and a 400 MHz ground coupled antenna to collect information from deeper structures and the subgrade. In addition digital video or photos from the road and its surroundings is also collected. Positioning is made currently using GPS systems.

Cross sections made with handheld GPR system always give valuable information about the structures and root causes of rutting problems – in this case Mode 2 rutting due to a weak subgrade

The benefit of 3D GPR system is that it produces a 3D view of the gravel road or forest road structures

There are a few issues that however should be known when designing GPR surveys. First of all the extensive use of dust suppressant salts in a wearing course increases the conductivity of the material and this can cause GPR signal attenuation resulting in good quality data being collected only from the surface layers. If this is the case, GPR surveys should be made during the wintertime when the surface materials are frozen and conductivity is close to zero. When measurements are done during the wintertime the presence of frost line and ice lenses also give valuable information for differential frost heave and bearing capacity problems during the spring thaw weakening. The rule of thumb is the deeper the frost line is in the winter the less will be spring thaw weakening problems during the spring. Finally, earlier ROADEX tests with a 3D GPR system gave very promising results in detecting large boulders in the road structure that sooner or later will heave up to the road surface requiring removal.

2D GPR data collected during the winter and its interpretation presented together with spring thaw weakening history. The top field (1) presents 2 GHz air coupled antenna data, from where wearing course, base course (road structures) and embankment have been interpreted. The middle field (2) presents 400 MHz data with total road structure thickness interpreted together with bedrock surface and frost line. The spring thaw weakening history (bottom field) shows clearly that the problems can be related to the presence of bedrock. Video is collected from the road during the summer.

The use of GPR data and laser scanner data in gravel road diagnostics. The top field presents the interpreted road structure thickness measured with a 400 MHz GPR antenna (black line) and ditch bottom levels compared to road surface (blue: right ditch bottom, red: left ditch bottom). The bottom field presents the gravel road surface dielectric value measured with a 2 GHz horn antenna. The green area presents optimum moisture content. The data shows that if the ditch bottom level is much higher than the bottom of the road structure it can have a clear effect on the moisture content of the wearing course and further increases risk for permanent deformation.

3D GPR data analysis of a gravel road in Finnish Lapland. The data view consists of one longitudinal section and a cross section at the point of the red line. The third (and coloured) field is a horizontal time slice from the depth of 1.20 m showing a silty sand “pocket” under the road structure. Because data has been collected during the winter the cross section shows a clear frost line at the depth 1.6 – 2.0 m. The frost line is always deeper at the road centre because on the road shoulders the snow is acting partly as an insulation structure.

The GPR data analysis also shows ice lenses at the depth of 0.5-0.7 m in the road structure. In this case the second data field shows unfrozen water in the frozen ground indicating the presence of ice lenses (blue reflections in the time slice data). The third field presents spring thaw damage history and the bottom data presents IRI during the spring of 2006. The integrated analysis shows that the spring thaw problems can be related to the transition areas of frost susceptible and non-frost susceptible subgrades. If the problem becomes worse the solution will be most likely soil replacement or frost insulation. But the first reaction in these case is always drainage improvement.

The depth of frost under a gravel road measured in a ROADEX test in February 2005. The data shows clearly a problem section from 700-780 m where the frost line is high up (1.2 – 1.7 m) because the frost line has not penetrated deep into the wet and frost susceptible subgrade. The frost line is deep starting from 800 m indicating that the subgrade is dry and/or the road drainage system is working well

3D GPR data showing potential large boulders in the subgrade at the depth of 0.5 – 0.7 m. This data was collected early in the winter when the frost depth was only around 0.35 m. This data can be used when removing the boulders from under the gravel road.

7.2.3. Deflection surveys using FWD and LWD techniques

Information how a gravel road or forest road structure will deflect under a heavy truck axle is essential in the structural diagnostics of roads. The theory of falling weight deflectometer (FWD) and light weight deflectometer (LWD) is explained in more detail in the ROADEX “Permanent deformation” eLearning package.

Basic principles of FWD technique. The FWD device consists of a weight that drops from a pre-specified height on to a plate that is supported by rubber dampers that rest on a road-based circular plate with a specified diameter. The drop of the weight is designed to simulate the load produced by a passing heavy vehicle. The load applied normally to gravel and forest roads is a load of 50 kN on a 300 mm loading plate. The deflection is measured by multiple geophones placed radially starting from the centre of the load plate.

A typical FWD system on a gravel road survey. The point interval is normally 20-50 m and on known problem spots it is recommended to have measurements with shorter point intervals.

FWD data is more useful in gravel road diagnostics as it provides valuable information about the deflection bowl. From the deflection bowl data different bearing capacity indexes can be calculated indicating if the problems can be related to poor quality road surface materials (Mode 1 rutting), or if the main problem is that the road structure cannot spread the truck load wide enough over a weak subgrade soil (Mode 2 rutting). The benefit of the FWD data is also that it can be used, together with thickness data from GPR, to back-calculation or forward calculation of the road structure and subgrade moduli values. These can be further used in the bearing capacity design in the road strengthening projects. There are a few issues that should be remembered when ordering FWD surveys on forest or gravel roads. First of all these surveys should not be made in the spring or early summer when frost is still in the subgrade because results will be too good. Also in the dry and warm summer days when road structure is dry, the results might be high. The best time for surveys is late summer and fall when the structures have more moisture and results are more representative.

LWD measures only the E2 value on the road surface but it is cheaper and easy to use and gives information about the general bearing capacity level and its changes in the road surface. LWD has proven also to be a good tool in compaction control when building gravel and forest roads.

On gravel roads, and on forest roads, another option for stiffness measurements other than FWD is the portable Light Weight Deflectometer (LWD). This instrument measures the surface modulus, and some instruments also have supplementary geophones that can help to identify if any bearing capacity problems are related to the layers close to the road surface.

In gravel and forest road bearing capacity basic diagnostics ROADEX recommends using bearing capacity indexes SCI (surface curvature index) and BCI (base curvature index). SCI is calculated from the difference between geophones 0 mm and 200 mm. High SCI values indicate a high risk for deformations close to road surface and Mode 1 rutting. BCI is calculated from the difference between 900 mm and 1200 mm. High BCI values indicate a high risk for deformation at the road structure / subgrade interface and for Mode 2 rutting. Presenting SCI and BCI values makes it possible to locate the road sections with different types of bearing capacity and deformation risks.

Principles of calculation for Surface Curvature Index (SCI) from the FWD deflection data and SCI classification for gravel and forest roads.

Principles of calculation for Base Curvature Index (BCI) from the FWD deflection data and BCI classification for gravel and forest roads.

FWD deflection bowls (top field) and bearing capacity indexes (bottom field) from a forest road. In the top field the blue bowls are measured from the right side of the road and the red bowls from the left side of the road. The green line (1) presents the 200 µm deflection limit and if the deflection bowl end (1200 mm deflection) is beneath this the road is most likely to be resting on peat. The blue line (2) presents the limit of an extremely poor BCI value with a very high risk for Mode 2 rutting and the red line (3) presents the limit of an extremely poor SCI value with high risk for Mode 1 rutting. This case presents a very good road section with minor risks for road failures.

In addition, a further useful bearing capacity parameter is the surface bearing capacity (E2) that gives a general overview of the bearing capacity of a single gravel road or forest road or road network. Another good parameter is the subgrade moduli value that gives information about the stiffness of the subgrade soil or embankment at the depth of 1.0-1.4 m. Combining these two pieces of data enables engineers to define those sections where the problems are subgrade related and more expensive to strengthen, and sections where bearing capacity can be improved in much easier way.

GIS map and statistics showing a bearing capacity analysis (E2) of a forest road network in Finland. These maps can be used also in planning timber cutting timetables.

GIS map and statistics showing the subgrade moduli a forest road network in Finland. These maps can be used also in planning timber cutting timetables. For instance on the sections with subgrade moduli < 30 MPa, timber cutting and transport can be conducted during wintertime if the strengthening costs of the sections are considered too high. The short weak section close to the main road needs to be strengthened in any event.

7.2.4. Laser Scanners and gravel and forest roads

Since the 2010’s laser scanner technology, also called lidar, together with improved positioning systems such as GPS and Inertial Motion Units (IMU) have become more precise and cheaper. In addition more powerful PC’s and better software packages can process larger laser scanner data sets in a more efficient and easier way.

Laser scanning is a technique whereby distance measurement is derived from the travel time of a laser beam from the laser scanner to the target and back. When the laser beam angle is known, and beams are sent to a range of directions from a moving vehicle with a known position, it is possible to make a 3D surface image, a “point cloud”, of a road and its surroundings. The point cloud can have millions of points, with every point having x, y & z coordinates and additional reflection or emission characteristics. The laser scanner should be place as high as possible on the survey vehicle so that data can be collected also from the inner slope and ditch bottoms.

The great benefit from the laser scanner data is that it is possible to measure all the critical parameters and dimensions needed for a road: road width, cross slopes, ditch bottom levels, inner slope shapes angles and outer slope angles. And if there are, for instance, vegetation or other obstacles in the road area, these too can also be detected.

A single cross section of a gravel road made using a laser scanner. In addition to laser scanners the system should also have an IMU to record the exact position of the scanner and GPS. With these three systems it is possible to record the exact coordinates for each laser point.

Some of the key parameters of gravel and forest road geometry that can be automatically calculated from the laser scanner data.

In addition to the above standard parameters there are many other new parameters that can be calculated from laser scanner data that can be useful in the gravel road or forest road condition management. For instance rut depth measurements can be used for indirect information about any bearing capacity problem sections. Experience has shown that after grading operations rut growth is always the highest on the same sections with permanent deformation problems.

Rut depth measurement results on a gravel road in Sweden measured in May, in July soon after the grading, and in August. Based on this data it is easy to detect the sections with the biggest deformation problems and focus the first maintenance measures on them.

Another useful laser scanner application in road maintenance is detecting verges to be removed. As reported in the drainage chapter, verges keep water in the road surface and this can lead to potholes, deformations and even erosion during periods of heavy rain.

Detecting verges at the side of a gravel road. Because of problems with vegetation verge detection surveys should be made in early spring or late fall.

Finally, laser scanners can be used to monitor dusting problems on gravel roads and where dust suppressant should be spread.

Gravel road dust surveys made in May before the spring maintenance and before the application of dust suppressant, and in August. The map shows up the sections where there is no need for dust suppressants (green areas in both fields) and where more salt should be used (red areas both in both field

7.2.5. 3D Accelerometer techniques

A three-dimensional (3D) accelerometer is an electromechanical device that detects and measures non-gravitational accelerations. These forces can appear as motion, vibration, or orientation of the equipment. Such forces include static and dynamic accelerations outside the range of normal gravity (G). Thanks to their low price 3D accelerometers have become more common and these sensors can be found for instance in modern cellular phones. Also most of the new generation vehicles have several 3D accelerometers installed and already in 2005 the ROADEX project reported about the use of yaw sensors in detecting black ice and other friction problems on roads.

On gravel and forest roads 3D accelerometers can be used in monitoring the functional condition, ie. the service level of these roads. Due to dusting problems traditional laser based systems cannot be used in unsurfaced roads. These 3D sensors can also be used during wintertime when point lasers have problems with snow smoke.

Gravel or forest road roughness can be measured with different cellular phone applications, but if acceleration forces need to be measured in a repeatable and comparable way, specially calibrated 3D accelerometers should be used. These systems consist of a high quality 3D accelerometer sensor installed on the floor under the driver’s seat and a camera in the front (or back) window. In addition the system has data collection unit working with 12V DC as well as GPS and 4G/5G antenna. The system can measure automatically when the vehicle is running, or it can be started remotely by cellphone. Because acceleration forces are speed dependent the systems need to be speed calibrated and because different vehicle have different suspensions systems this calibration should be done with different artificial or known “bumps”. The collected data is then stored in the cloud and can be viewed and analysed using special web browser software packages.

Parameters and their names that can be measured with 3D accelerometer systems and surveys results of a gravel road measured in May, July and August. IRI can also be calculated from 3D accelerometer data.

3D accelerometer system components. These systems send the collected data in roughly two second delay to the cloud, so the system is almost “real time”.

The greatest benefit of these systems is that they can be installed on normal vehicles used by engineers responsible of the follow-up of gravel road maintenance contracts or by the supervision staff of the contractors who are driving regularly on the road network. In this way it is relatively easy to collect valuable data about changes in gravel road network roughness in different seasons and years. Results of using these systems so far indicate that a single measurement each year will not provide any kind of reliable general picture of the gravel road condition over the year.

Chart showing the relative amount of different roughness class problems in Karstula maintenance area in Central Finland from April 2021 to November 2022. The results show that gravel roads are smoothest in December when the road surface is frozen and thanks to that the bearing capacity is also good. The unevenness starts to grow when frost penetrates deeper into the road structures and subgrade and segregation ice lenses start to grow. The relative amount of roughness grows monthly until it reaches a peak in April when the Spring thaw weakening starts. Then, thanks to spring maintenance, the roughness improves substantially in May but again is worse in June mainly due to those sections that have thawed after the spring thaw maintenance measures. Roughness varies substantially during the summer until it starts to improve again during the fall.

By presenting the 3D accelerometer results as GIS maps it is easy to make objective comparisons of the average condition of roads in specific areas and thereby focus any special maintenance measures on those road sections with the biggest problems.

Classification of gravel road roughness in the Äänekoski maintenance area Central Finland in summer 2022 based on their 500 m average conditions. Experience has shown that this good for prioritizing the roads to be improved first.

The roughness problems of a particular gravel road can be analysed by a condition history over spring, summer and fall. In this analysis problems related to differential frost heave are normally those ones with decreasing roughness problems from spring to summer, and no problems during the late summer and fall. Wearing course related problems can be seen mainly during the summer and they can change based on how well the maintenance measures have been made.

Classification of gravel road roughness problems based on the appearance in spring-summer-fall period from March to November. Class 1 presents mainly differential frost heave related problems, while class 2 and 3 roughness appear mainly in the summer and are mainly related to wearing course and drainage issues (washboarding or potholes). Class 3 present a case where problems in June have been removed for instance by good grading. Class 4 problems from summer to late fall are often related to drainage problems or settlements. In this case classification is made based on maximum vertical roughness in one second.

Photographs can also be a great help in 3D accelerometer based roughness condition analyses. They can give a quick understanding of the reasons for the roughness problems in each area. View the photos from the map.

Photographs with vertical acceleration values also help the preparation of roughness classifications for gravel roads or forest roads.

Examples of gravel road / forest road conditions on different vertical acceleration levels and their preliminary diagnostics.

Finally 3D accelerometer data can be used also on the classification of the three most common roughness problems on gravel road, ie. differential frost bumps, potholes and washboarding. Their root cause and also their repair method is totally different.

The three most typical roughness problems on gravel roads in the ROADEX area: differential frost heave bumps (left), potholes (middle) and corrugation or washboarding (right).

The best indication of differential frost heave related bumps are 3D accelerometer pitch values and their variation, with additional comparisons of the spring values to early autumn values. The autumn pitch values should be much lower than the spring values.

Detection and classification of differential frost heave bumps by using 3D accelerometer pitch data. This data has been collected from a gravel road in Finnish Lapland in late April when the frost bumps roughness was the worst. These bumps can be confirmed to be related to frost heave by comparing these locations to data from August - September when they are not seen.

Potholes and washboarding can often be separated with a wavelength analysis of the vertical acceleration data. Potholes can be mainly related to drainage problems and high water content in the road structure. That is why sections with potholes will also have high amounts of long wavelengths whilst washboarding is mainly surface related problems without long wavelength components.

ROADEX Network