Residual soils, covering large areas of the earth, have been discovered for the last 30 years in most volcanic tropical and subtropical countries. The soils are mainly formed from the physical and chemical weathering of the parent rocks and minerals. They are found in the following African countries: Cameroons, Ethiopia, Tanzania, Rwanda, Kenya, Spanish Guinea, and the Canary Islands. Countries in the East African region, such as Kenya, usually experience high rainfall during which red-brown nitosols soil is derived from volcanic rocks of steep slopes. The soils generally have an acidity of pH 4.0 to pH 6.0, particularly in the Murang'a, Nyeri and Nyandarua districts of central Kenya and several parts of the Eastern province of Kenya [1]. East Africa includes six distinct climate zones: the Lake equatorial climate, the modified equatorial climate, the Kenya highland sub-tropical climate, the tropical climate, the coastal tropical climate, and the desert and semi-desert climate. Kenya has a subtropical and a coastal tropical climate, which results in a diverse range of tropical red clay soils formed over various igneous (both volcanic and intrusive) rocks.

Geotechnical engineers usually assume that the residual soils conform to preconceived patterns of soil in terms of mechanical behaviour. However, applying concepts from other soil types, such as sedimentary soils, may or may not be relevant to these soils.

Red soil is a type of residual soil and one of the most problematic soils characterized by swelling due to wetting and shrinkage due to the high clay content inherent in its structure [2]. It is very hard when dry; however, it loses this strength when in a wet condition.

The relationship between residual soil structure and engineering qualities does not provide a qualitative explanation for the soil's unusual properties in practice. As a result, the mineralogical composition and structure are usually used to characterize the residual soil [3-6]. The unusual properties of some forms of residual soil are attributed to their specific clay minerals, which cause their mechanical and physical properties to be strongly linked to their mineralogy composition [7]. Residual soils' mineralogy composition influences their behaviour more than stress history [8, 9]. Physical and chemical weathering produces particular clay minerals that can control its behaviour. Residual soil, unlike sedimentary soil, does not consider stress history. Therefore, the terms' normally consolidated' and 'over consolidated' are irrelevant to residual soils.

According to Wesley [6, 8], the residual soils are divided into three main groups (A, B, and C) mentioned in detail by Wesley. (2009) [8]. The current study will focus on group (C) - tropical soil, specifically tropical red clays, categorized under three sub-groups that included the group of residuals characterized by unusual properties. In engineering literature, tropical soils are sometimes described as falling into two broad categories: tropical black and red.

The uniqueness of residual soils has been demonstrated in previous studies. Despite the availability and importance of residual soils, few dedicated efforts have been made to develop a comprehensive document that characterizes their behaviour. This would allow the decision-makers to think accordingly and in time to avoid unanticipated problems. Physical and chemical weathering produces particular clay minerals that can control its behaviour. This type of soil has recently received special attention from researchers due to its unusual behaviour attributed to its remarkable properties. The red soil in this paper was studied in Ruaka, a suburb of a large African city, Nairobi, where relevant data on physical and mechanical properties are lacking. Kenya is one of the developing countries under the pressure of economic development. The country is currently involved in massive infrastructure expansion projects including the construction of bridges, roads, water supply pipes, underground structures, and other infrastructure throughout the country especially in Nairobi. Therefore, more research on red soils is necessary to meet the requirements of the ever-increasing construction works. This study investigated the effect of the chemical and mineralogy composition of Ruaka soil on its engineering properties. The present study results will inform infrastructure development in the area and the general behaviour of tropical red soils.

The first systematic study of this soil in Kenya was conducted between 1937 and 1945 for the Ruiru Dam site [10]. A large earth dam came to be built in the tropical areas of Kenya. It was observed for the first time that the properties of this residual soil were considerably different from those seen in temperate climate sedimentary clays. The optimum moisture content was around 50%, while the maximum dry density was approximately 72 lb/ft³ (11.53 kN/m³). Therefore, Karl [11] reassessed the integrity of this dam on this soil; the results showed that the abnormal properties were assets rather than liabilities, and the clay did not conform to conventional behaviour. Between 1947 and 1949, another site investigation for Sasumua Dam took place, Halloysite was a principal constituent, and the soil showed similar properties. Furthermore, the liquid limit was high, the natural moisture content was high, the maximum dry density varied from 65 to 75 Ib/ft (10.20 to 11.77kN/m³), and the optimal moisture level was between 55 and 45%. The soil cohesion (c') and the friction angle (ϕ') were generally between 26 and 30 kN/m² and 34֯, respectively.

In yet another study, Newill [12]. investigated the properties of two samples of red clay soil, one from Nairobi's Kabete district and the other from a site near the Sasumua Dam, 40 miles northwest of Nairobi; his research highlighted the unusual properties of Kenya's red clay soils and the difficulties in interpreting laboratory tests such as the plasticity and compaction tests. The same conclusion was presented by B. Huat et al. [9], Dixon & Robertson [10], and Karl [11] when they studied the red clay soil from Nyeri and Ruiru Dam in Kenya. The soil possessed unusual properties, including excessively high moisture absorption, increased index values and abnormally low Maximum Dry Density (MDD); they attributed these properties to its mineralogy composition and structure.

Several other studies [5, 13-16] were carried out to investigate the impact of the mineralogy composition on the physical properties of tropical clay and to obtain further laboratory evidence of the abnormal behaviour of this soil. The comprehensive investigation was launched in 1993 by the Overseas Development Administration for some tropical regions in Kenya and Indonesia [17-19]. They observed that the increase in clay content is consistent with the variation in mineralogical composition, which makes tropical soils' mechanical and physical properties unstable. Additionally, these soils can exhibit swelling characteristics in an intermediate weathering stage.

Wesley [6, 8] emphasized that geotechnical engineers should have open minds about how residual soils may behave and not assume they will conform to preconceived patterns, especially when working with tropical soils. Based on this engineering background, more and more attention has been paid to understanding the behaviour of tropical soils. Further studies [17, 19-25] were carried out to determine the behaviour of red clay and to solve practical engineering problems such as the settlement deformation of buildings. More recently, a study by Xue et al. [26] in southern China investigated red clay's shear strength and crept behaviour to evaluate red clay's long-term shear strength and deformation when subjected to various structural loading. Their study provided promising results that can help solve the settlement problem in this type of soil.

Although there is a universal need to develop a method for indexing the abnormal properties of tropical soils, conventional index tests, particularly Atterberg limits, remain essential tools for the adequate characterization of these soils [6]. Furthermore, Wesley [6] criticized the argument that soil drying affects the results of Atterberg limits tests, stating that the influence of soil drying on the Atterberg limits is confined to shifting the point on the plasticity chart parallel to the A-line. So, this is insignificant because the only criterion for evaluating the soil is the distance above or below the A-line.

However, Gadzama et al. [20] investigated the effect of temperature variation ranging from 25 to 150◦C on the geotechnical properties of selected black cotton soils. Their results revealed decreased liquid and plastic limits and increased MDD. There was no increasing effect on Plasticity Index (PI) and Optimum Moisture Content (OMC).

According to Attah & Etim [2], the reaction of clay mineralogy to an elevated temperature significantly impacts its engineering properties in service. Even though their study was conducted on lateritic soils, their results showed a decrease in the Atterberg limit and an increase in MDD, which was consistent with Gadzama et al. [20].

In the tropics, soil formation is very varied and frequently complex. Because of its high clay content, it is an expansive soil that offers a considerable engineering problem, necessitating its stabilization before being employed for any engineering applications. As a result, the recent investigations by George et al. [21], Hanuma et al. [22], Sani et al. [23], Osinubi et al. [24], and Abner [25] on red clay have contributed to improving its properties. The studies [21-25] concluded that such soils, due to their unpredictable behaviour, are unsuitable for engineering construction, emphasizing the need for soil stabilization to enhance engineering properties.

The difficulty and abnormal behaviour would appear when applying 'conventional' soil mechanical concepts regarding sample drying and preparation. Particularly with tests that require remoulding the samples, such as the triaxial test, controlling the optimum moisture content with the MDD becomes challenging.

The relationship between the mineralogical composition of this soil and its engineering properties is complex and understudied in geotechnical engineering because the properties of this soil are strongly reliant on the weathering condition of the parent rocks, which varies by region. The behaviour of tropical red soil from Ruaka was defined and analyzed using standard testing to determine the usefulness of these tests in the engineering evaluation of tropical red clay soil properties. The results were compared to the chemical, mineralogical, index, and mechanical properties of typical tropical red soils from different tropical climatic regions in Kenya.

The soil investigated in this study is red clay soil collected from Ruaka. Ruaka town is located in Ndenderu Ward, Kiambaa Sub-County, Kiambu County. The suburb is located northwest of Nairobi and is easily accessible via the Northern Bypass and Limuru Road (Fig. 1). Ruaka is part of the larger Nairobi Metropolitan Area, with easy access to Kiambu, Limuru, Nairobi, Thika, Kajiado, Machakos, and Muranga Towns [26]. The study area is located between Longitude 36° 48' 0.51”, 36° 48' 6”, and 36° 48' 10.1”, East, and latitudes 1° 12' 33.6”, 1° 12' 36.7” and 1° 12' 38.7”, South of the Equator which is characterized as having tropical red soils as shown in Fig. (2).

Three boreholes were used for laboratory investigation to characterize the soil of Ruaka with depths of 10.5m, 8.95m, and 8.95m; the investigation could not go more profound due to the presence of rocks that are out of the scope of this study. The subsoil characteristics of each borehole are studied in detail.

The mineralogy and structure of the soil determine the engineering behaviour, which in turn, is dependent on the original rock type and climatic and drainage conditions both during and after weathering. Grain size, unit weight, strength, and compressibility vary significantly due to these parameters.

Undisturbed and Disturbed soil samples were extracted from three boreholes, as shown in Fig. (3), at various depths (8.95m and 10.5m). The basic properties obtained from these undisturbed and undisturbed soil samples were determined using a variety of tests. A light excavator with a 100 mm diameter and 500mm long cylindrical corer was pushed into the soil to the desired depth to extract the undisturbed soil (U100) sample; this static push method prevents the corer from rotating, resulting in high-quality samples (Figs. 4 and 5). Disturbed samples were collected using a drill and stored in plastic containers. The samples are then sent to the laboratory in sealed containers to preserve their natural moisture contents, preparing them for subsequent laboratory analysis.

Fig. (1). The location of the area of study (Ruaka).

Fig. (2). Location of sample site (three boreholes).

Fig. (3). Soil sample collections, including disturbed sample (right) and undisturbed sample (left).

Fig. (4). Apparatus used during geotechnical drilling.

Fig. (5). Extruding the undisturbed samples.

The collected soil samples were subjected to two laboratory tests: soil classification tests (moisture content, Atterberg limits, specific gravity, and particle size distribution) and soil strength tests (standard proctor compaction and consolidation tests). After that, the samples were fully air-dried, and the lumps were pulverized and sieved for different tests performed using the British Standards, as summarized in Table 1. The chemical and mineralogical composition tests were also carried out to determine the characteristics of this soil and their impact on the physical and mechanical properties.

This study was designed to investigate the effect of mineral and chemical composition on soils' physical and mechanical properties for the red clay soils selected from Ruaka, Kenya. The investigation showed that the soils' microstructures, chemical, and mineral composition for tropical residual soils, particularly those formed from volcanic rocks, have a distinct mineralogical composition that governs their behaviour and makes them unpredictable. Therefore, working with this soil type might prove more challenging for civil engineering applications. Based on the laboratory findings, the following conclusions are drawn:

• According to XRD analysis, the most abundant minerals in Ruaka clay were Quartz, kaolinite, Halloysite, and Mica. The chemical testing revealed high concentrations of AL₂O₃ and SiO₂; the analysis also revealed the presence of a cementing agent represented by free iron oxide, which has a significant amount (16.40%) in the chemical composition.
• The structure and physical properties of tropical red clays may be altered by the mineralogical composition of these soils and partial air-drying at ambient laboratory temperatures. Some of these changes are not reversed when the soil is re-mixed with water. These structural changes can be seen in the index properties such as Atterberg limits, particle size, and particle density (Gs) tests.
• The preliminary tests for the samples in their natural state (in situ) showed a high convergence in their properties (the unit weight, the specific gravity, and the bulk density). However, Ruaka soils showed a high Liquid Limit between 63-68% which increased with increasing depth, and high moisture content varied between 37- 46% convergence with their plastic limit with a slight increase as depth increased. Other soil properties, such as clay content, increased, ranging from 59 to 79%.
• Ruaka clay was found to have a high void ratio of 1.145 to 1.63, with saturation values ranging from 69 to 99%. This increased permeability and lowered MDD by an average of 1247 kg/m³ with 35.5% OMC.
• The presence of kaolinite, Quartz, and oven drying of hydrated Halloysite significantly reduces the specific gravity (Gs) values.
• A good positive relation was observed from the oedometer test between the void ratio, e, and the compression index, Cc, with increased pressure and depth. However, the high amount of clay content made this soil mechanically unstable.
• Compared to the other nearby regions with the same tropical soil features in Kenya, Ruaka clay showed a higher fraction of clay-sized grains, which appear to have higher plasticity; however, Atterberg limits revealed a good agreement with the trend of other regions. This type of correlation is highly recommended because understanding the behaviour of a specific group of soils aids in predicting the behaviour of a new type of soil within that zone.