1 Introduction

Road traffic is very useful for the displacement of people, products, and services. This is mainly noticeable in countries where railways, aviation, and marine transport systems remain largely underdeveloped [1]. Many roads especially in developing countries start to deteriorate as soon as they are put into service, and this phenomenon has become recurrent, manifesting itself in pavement collapses. Road failures are frequently observed in Cameroon, particularly in the western part of the country. Examples of road collapses include (Fig. 1): Kekem (October, 2007), Bafoussam-Bamendjou (September, 2017), Bafang-Banja (October, 2019), Melong-Dschang (August, 2020) and Babadjou-Bamenda (March, 2021). Geological, geomorphological, geotechnical, climatic, and human factors are involved in the mass movement [2, 3]. Geological factors depend on the bedrock at the origin of undeveloped or poorly developed soils and clay soils. In terms of geomorphological factors, the landscape is characterized by hills with high slopes and a surface drainage system. Through seepage water during the rainy season, the climate changes the physical and mechanical characteristics of the materials in place. Other considerations such as road use, construction, and maintenance practices are taken into account [4]. The quality of the studies is often carried out by inexperienced design offices does not always take into account the rheology of the materials used in the pavement in relation to the immediate environment.

Geophysics is widely used with the development of measuring instruments and improved data processing software. In addition, geophysical methods are cheaper than in-situ geotechnical surveys, rapid to investigate large areas of land, and are non-destructive [5]. Geophysical research provides the knowledge necessary to understand a wide range of geological, hydrological, environmental, and geotechnical problems. Electrical resistivity imaging methods are an effective tool to detect anomalies and determine the complexity of the subsurface geology of areas affected by mass movement [6, 7].

In this study, electrical resistivity tomography combined with geotechnical techniques was applied to investigate the two sections of road failures in western Cameroon. The main objective of this work is to identify the nature, subsurface structure, water conditions and geotechnical properties of the road foundation soil, and the likely causes of recorded mass movements.

Fig. 1
figure 1

Location of the main road failure sites in western Cameroon on a topographic map (QGIS 3.32-https:/www.download.qgis.org)

Full size image

2 Description of the study area

Two sections of road were selected in Bafoussam, in the Mifi division, West Cameroon region (Fig. 2). These two roads were constructed by the Cameroonian government as part of projects for the economically viable paving of low-traffic roads while the African Nations Championships (CAN) were being held in Cameroon in 2016 and 2019. The Site 1 is the regional road 615 between Bafoussam and Bamendjou towns. It was built in December 2016 over a total length of 27 km. The pavement structure consists of a surface dressing layer and a base layer of natural lateritic gravel of 20 cm. It was destroyed in September 2017 at KM 2 + 200 and KM 2 + 300, and studies for reconstruction were carried out in October 2017. The Site 2 is an urban road built in October 2018. The failure of this road was recorded in August 2019 during the rainy season, and studies for rehabilitation then took place in September 2019.

The town has a transitional tropical climate with two distinct seasons: a dry season and a rainy season. The average temperature is 25 °C and the average rainfall is 1871 mm (Fig. 3) The highest rainfall occurs between July and October and causes high soil water saturation, leading to mass movements [3]. The altitudes of the study area vary between 1200 and 1500 m, but some talwegs are found at altitudes of ~ 700 m.

The geological formations consist of basalts, ashes and lapilli on a granite-gneissic basement of the Precambrian age [8]. Cretaceous to present basalt flows are fractured with typical orientations of N35°E and N140°E, probably related to the Cameroon Volcanic Line trends [9]. Recent basalts and pyroclastic materials occupy most of the study site. Alluvial deposits and migmatitic gneisses outcrop in the north and east, porphyritic gneisses and granites in the center, and volcanic materials and granites in the west [9]. The weathering materials are mostly from basaltic deposits. [3] described significant amounts of fine elements, high porosity, cohesion, and friction angle in these weathering materials.

Fig. 2
figure 2

Location of the study area including geophysical profiles and geotechnical tests (Surfer 21-https://surfer.software.informer.com/21.2; Google Earth Pro 7.8 -https://google-earth-pro.fr.Uptodown.com/windows/telecharger

Full size image
Fig. 3
figure 3

Rainfall distribution map for Bafoussam and nearby areas (QGIS 3.32-https://www.download.qgis.org)

Full size image

3 Data and methodology

The methodology adopted to study the mass movements of failed road sections was carried out in three stages: (1) geological and geomorphological reconnaissance of the site, (2) subsurface conditions using the electrical resistivity tomography technique, and (3) the geotechnical analysis of the materials.

The first step was to determine the geological and geomorphological conditions related to the mass movement event. The field observation of the disorder organized by the road collapse and the description of weathering profiles were carried out. For the topographic elements controlling these mass movements along the studied roads, a geomorphological analysis of the sites was examined. Longitudinal profiles highlight anomalies along roads such as slope breaks. These profiles were obtained using the [10] equation, recalibrated by [11]. Slope breaks are therefore identified along the profile by the SL index peaks.

The second step was carried out using Electrical Resistivity Tomography (ERT). This geophysical test provides a rapid 2D or 3D image of the soil layer nature and structure from near-surface, by showing lateral and vertical variations of electrical resistivities, using electrodes at the surface [12]. This technique uses a multi-electrode system to collect many measurements of different pairs of current input electrodes and potential measurement electrodes [13]. During the data acquisition sequence, the resistivity instrument automatically selects the electrodes used for current input and potential measurements. The readings are captured in a computer. The potential difference depends on the intensity of the injected current, the configuration of electrodes, and the electrical resistivity of the material. The resistivity in the soil and near-surface structure can then be deduced from the shape of the potential field. The geophysical data acquisition procedure was carried out using a multi-electrode system of the Syscal R2 switch 48 from the company Iris Instrument®. Six ERT profiles were carried out according to the Wenner alpha and Wenner-Schlumberger configurations (Table 1; Fig. 2). These arrays are very sensitive to horizontal and vertical variations in resistivity, therefore suitable for imaging the geological structure of the near-surface of the study area. The inter-electrode spacings (a) were 2.5 m, 3.5 m, and 5 m, for a better resolution. Different combinations of (a) and (n) were used. The order (n), which translates the number of acquisition levels, was a maximum of 16. This factor defines the depth of investigation reached by the array, that was 17–38 m in this instance. After collecting the apparent resistivity data, the next step was to process and invert the data using Prosys II®, X2IPI®, and Res2Dinv® software. Propsys II® was used to filter the data, and the X2IPI® program was used to remove the failed electrodes. Prosys II® software was used to input the elevation of data for each electrode. The inversion of resistivity data was carried out using Res2Dinv software [14, 15] following the parameters given in Table 2. The inversion process consists of minimizing the addition of the differences between the measured and calculated apparent resistivity on a model. The quality of the inversion is controlled by the RMS (Root Mean Squared) between 5 < < 11 for site 1 and 8 < < 21 for site 2. The RMS measures the difference between the calculated apparent resistivity and the measured electrical resistivity. In order to analyze and interpret geophysical models obtained after inversion, the resistivity data were grouped into resistivity range classes. The layout of the ERT profiles at Site 2 (Fig. 2) gave a 3D resistivity model of the subsurface from RockWorks® software.

Table 1 Main acquisition parameters of the ERT survey profiles
Full size table
Table 2 Main Inversion parameters used for the calculation for all presented inverse models
Full size table

In the last site investigation stage, geotechnical tests were carried out on site 1 to identify lithology layers and determine their mechanical and physical properties. In situ and laboratory investigations into two different types of geotechnical issues have been performed. Four hand pits (PT), four static penetration tests (SP) to standard NF P 94–116, and four dynamic penetrometer tests (SPD) to standard NF P 94–115 were drilled to a depth of 6 m (Fig. 2). Four soil samples reworked were collected from the pits for geotechnical laboratory analysis (PER1, PER2, PER3, and PER4). The following properties were determined: soil grain size to standard NF P 94 − 056 and NF P 94 − 057, void index (e), natural moisture content (w) to standard NF P 94 − 050, saturated density (γh) to standard NF P 94 − 053, dry density γd) to standard NF P 94 − 053, specific gravity (γs) to standard NF P 94 − 054, water saturation (w) to standard NF P 94 − 053, degree of saturation to standard NF P 94 − 053, Atterberg limits (liquid limit: LL, plastic limit: PL and plasticity index: PI) to standard NF P 94 − 051 and rectilinear shear to standard NF P 94-071-1. The Swelling potential was calculated as a function of plasticity index according to [16].

4 Results

4.1 Field observations of failed roads

Field observations revealed failure affecting the pavement and embankment slopes of the roads at the two sites studied (Fig. 4).

The failed section of road at site 1 is ~ 8 m wide and trends generally NE-SW (Fig. 4a). On the surface, this disaster caused meter-sized fissures and the displacement of road sections. The NW section of road failed with the maximum failure located to the west of the site with vertical movement of 3 m decreasing towards the SW. The foundation soils are clayey, reddish and they contain fragments of basalt rock (Fig. 4c). The pavement base layer and embankment slopes are formed by lateritic soils. The basaltic bedrock outcrops at the SW of the section. The slope has failed on the road. Damaged pipes are spraying water.

The section of the secondary road in site 2 is 10 m wide and E-W oriented with larger fissures (Fig. 4b and d). The southern section of road failed and the northern section remained in situ. The vertical offset between the two is ~ 1 m. The road’s foundation is built using the pavement body embankments. Structures for drainage and slopes stabilization are missing.

Fig. 4
figure 4

Map showing the damage observed on the different failed road sections: a and b represent site 1, c and d represent site 2

Full size image

4.2 Geomorphological analysis

The sites generally present steep slopes and large elevation differences of 1054.98–1666.54 m (Fig. 5a). The longitudinal profile of Site 1 is oriented NW-SE oriented and slopes between 7–70°, with two slope breaks at 1326.58 and 1349.36 m (Fig. 5b and c). The slope gradients increase from NW to SE. The longitudinal profile of Site 2 is NE-SW oriented with slope gradients between 4–40°, and a slope break at 1395.15 m (Fig. 5d and e). These slope gradients increase from NE to SW.

Fig. 5
figure 5

Morphological profiles of the study area: a DTM locating sites and longitudinal profiles, b and c 3D and 2D longitudinal profile showing slope break at site 1, d and e 3D and 2D longitudinal profile showing slope break at site 2

Full size image

4.3 Geophysical characterization of foundation soils

ERT data were recorded along six profiles (L1, L2, L3, L4, L5, and L6) with a total length of 1106 m (Fig. 2). Profiles L5 and L6 are perpendicular and the others parallel to the road alignment. The profile depths of the ERT varies between 17 and 38 m. These profiles show a vertical and horizontal variation in resistivity, characteristic of the heterogeneity of the subsurface materials (Figs. 6 and 7). Resistivity values range from < 5 to > 1000 Ohm m. The subsurface model indicates three distinct zones of resistivity values: conductive C1 (ρ < 30 Ωm), semi-resistive C2 (30 < ρ < 200 Ωm), and, resistive C3 (ρ > 200 Ωm). The C1 zone is found in the lenticular form in the C2 zone, with dimensions that can extend tens meters. The C1 zone is very wide in profiles L2 and L4. The C3 zone appears on the surface and at depth in the ERT profiles. Its thickness is significant in profiles L1 and L6.

Fig. 6
figure 6

Modeled resistivity sections along with the ERT for profiles L1 and L2 (C1: Conductive Zone, C2: Semi-resistive Zone, and C3: Resistive Zone)

Full size image
Fig. 7
figure 7

Modeled resistivity sections along with the ERT for profiles L3, L4, L5, and L6 (C1: Conductive Zone, C2: Semi-resistive Zone, and C3: Resistive Zone)

Full size image

4.4 Geotechnical properties of soil foundations

The geotechnical surveys were carried out in the proximity of the ERT profiles at site 2 to confirm the results of the geophysical survey. These geotechnical tests were used to describe the nature, conditions, and quality of the materials used for the foundation of the road failures studied. The lithological logs of the static penetration tests carried out outside the pavement reveal clayey-silt soils with weathered basalt blocks covered by a thin organo-mineral layer (Fig. 8). The geotechnical tests carried out on soil samples taken from the pit yielded the physical and mechanical parameters (Table 3). The void index values of samples are higher than e > 0.8. The natural moisture content values of the materials vary between 22 < Sr < 34%, very close to or even higher than the values of the plasticity limit 30 < PL < 40%. The saturation values are very high: 70 < Sr < 99%, indicating saturated conditions. The plotting of plasticity limits and liquidity limits values per sample on the Casagrande chart (Fig. 9) place the materials in medium to very plastic clay zones. PI values are between 15 and 40%. These results support the particle size analysis which confirms a very high fines fraction (clay) ~ 64%. The apparent internal friction angle (φuu) varies between 24 and 37°. The water saturation after shearing is 35 < ω < 50%. The swelling potential values of the samples are between 0.4 and 3.4%.

Fig. 8
figure 8

Lithological logs obtained from static penetration tests on site 1

Full size image
Table 3 Physical and mechanical parameters of soil samples from the Bafoussam-Bamendjou section
Full size table
Fig. 9
figure 9

Plasticity chart for soil samples [17]

Full size image

5 Discussion

Geological field observations at Site 1 show heterogeneous materials of silty clay contains fragments of basalt rocks. In site 2, the foundation soils are lateritic embankment gravel. The geophysical tests identified the different layers of road foundation soils and have also confirmed the degree of heterogeneity of these materials. The correlation between the ERT profiles and the lithological logs obtained shows three types of subsoil material. Site 1 consists mainly of embankment materials to the investigated depth (Fig. 10a). The high resistivity values (ρ > 200 Ωm) correspond to dry lateritic gravel embankment (DLGE). Resistivity values between 30 and 200 Ωm correspond to wet lateritic gravel embankment (WLGE) and the very low resistivity values (ρ < 30 Ωm) corresponds to a water-saturated zone (WSZ). At site 2 (Fig. 10b), the resistive zone corresponds at the surface to embankment material (EB) and at depth to weathered basalt bedrock (WB). The semi-conductive zone reveals a dry silty clay (DSC) and the conductive zone represents the water-saturated zone (WSZ). The water-saturated zone is covered with a sliding surface. From site 2, a simple illustration of the 3D model of subsurface soils is shown in Fig. 11.

The particle size analysis shows clay content as 64% and the abundant clay in the slope materials implies stability [18]. The Casagrande diagram shows clayey, highly plastic, and compressible soils. These soils are susceptible to compaction or compressible [5, 19]. According to the HRB (Highway Research Board) classification, these soils are classified as A-7-6 and reflect a medium to poor subgrade. Slopes composed of clayed or unconsolidated materials are responsible for the landsides [20].

Steep slopes (> 40°) and slope breaks at both sites can create the gravity contrast and induce landslides. The materials of steep slopes are naturally erosion and landslide-prone areas [2, 21]. These mass movement phenomena, therefore, occurred slowly as a result of a reduction in the stability of slopes. Also, the study area is characterized by an average annual rainfall of around 1600–2000 mm/year, which corresponds to a very wet region, thus a source of mass movement in the roads (Fig. 11). The majority of failure roads occurred during the rainy season (July–October) resulting in high soil water saturation [2, 3]. These observations on rainfall data would indicate that heavy rainfall could cause mass movement of roads. The seepage water into the soil increases the moisture content of the materials (21–34%), often favoring mass movement of roads. The high void index values (e > 0.80) of materials, allowed the appearance of water circulation networks between the particles [22,23,24]. The swelling potential of these clay materials is poor to medium. The meteorological conditions result therefore in the degradation of the mechanical parameters. Friction angles ranging from 24° to 37° show that these collapsed soils are highly heterogeneous in terms of cohesion. The roof of the water-saturated zone is the sliding surface that caused road damage. A slope break always occurs when the soil water content near the bottom of the slope is completely saturated, even if other parts of the foundation materials are still moving [25–27].

Fig. 10
figure 10

Inverse model of resistivity sections along profile line L1 (site 1) and L4 (site2). EB, embankment; DLGE, dry lateritic gravel embankment; WLGE, wet lateritic gravel embankment; DSC, dry silty clay; WSZ, water satured zone and WB, weathered basalt

Full size image
Fig. 11
figure 11

Deep geological structure of the road collapse soils in site 2: a water-saturated zone, b wet lateritic gravel embankment, c dry lateritic gravel embankment, and d 3D Model

Full size image

6 Conclusion

The purpose of this work was to evaluate the damage observed on two sections of road collapsed by mass movement in wersten Cameroon. The methodology used is based on soil testing along degraded road profiles. To achieve this objective, Electrical resistivity tomography, manual drilling, laboratory geotechnical tests were used to carry out in-depth investigations. These techniques have determined the electrical and mechanical properties of the soil layers in place. According to the investigations carried out, these roadway collapses are caused by seepage water during the rainy season, the nature of the materials and the steep slopes of the unprotected embankments. The seepage water has modified the nature of the materials and making them unstable. On the basis of the results obtained, pavement design did not take into account all the natural conditions of the project site. therefore, safety pavement platforms must be built or rehabilitated to support future road mass movements. he proposed solutions are: construction of a surfaced pavement in the fill and excavation zones, stabilization of the sub-base layers of pavement platforms. As geosynthetics have demonstrated their efficacy in stabilizing road pavements, they should also be used.