Comparative study of the effects of natural fibers on the physical, thermal and mechanical properties of soil building blocks

2.1. Materials used

2.1.1. Soil 

 ⌅

The soil studied was obtained from the Sous Massa region in southern Morocco. The coordinates of the site of origin of the raw material are (N 30°23’ 20.8314’’, W 8°41’ 27.9126’’), and the location is illustrated in Figure 2b. This village is known for its traditional clay construction techniques (Figure 1a). The clay soil of the Souss region was selected for this study because of its abundance and its traditional use by the local population for the production of high-quality adobes. The material was extracted by collecting samples from several locations in the selected area, at depth, after removing the top layer containing organic matter. This step was essential to ensure that the collected soil accurately represents the village soil. A sufficient quantity of raw material was then collected for the study. This study needs to characterize the soil and examine its impact in terms of mechanical strength, as well as its suitability for use with other natural additives.

Figure 1.  Raw material extraction site, a) soil extraction site, b) juncus fibers extraction site, c) delivered coconut fiber stock.

2.1.3. Coconut fibers

 ⌅

Coconut fibers, in general, are natural fibers that are extracted from the outer husk of coconuts. The planting of these fibers requires special climatic conditions, which makes them rare in Morocco, mainly grown in tropical regions. The main producing countries include India, Brazil, Indonesia and the Philippines. The coconut fibers used in this research are presented on the market in the form of compacted rolls (Figure 1c). For use under concrete screeds to improve acoustic performance by minimizing the effect of vertical noise in special-purpose buildings such as singing rooms and dens, these commercialized fibers have been prepared and dispersed manually to obtain elements of the same dimensions Figure 2a. Coconut fibers are a natural material with multiple applications, offering significant advantages in terms of performance, durability and respect for the environment. Due to their scarcity, compensating for this lack in different paid by another alternative, constitutes a real need in developing countries. The aim of this comparative research is to identify a fiber with characteristics similar to those of coconut, in order to propose a more accessible ecological alternative better suited to local contexts, where coconut fiber is not naturally available and is mainly obtained through commercial channels.

Figure 2.  Different fibers after preparation, a) manually dispersed coconut fibers and juncus fibers after grinding, b) location of soil extraction site.

2.2. Characterization of prepared samples

2.2.1. Geotechnical and physico-chemical soil characterization

 ⌅

The clay sample designed for the study was subjected to geotechnical identification (Figure 3), i.e. atterberg limit in accordance with standard NM 13.1.007 [28], methylene blue index and specific surface following NF P 94-068 [29] and sand equivalent following NM 10.1.147 [30]. Plasticity index (PI), plasticity limit (PL) and the liquidity limit (LL) of the soil studied were found to be equal to 38.05%, 23.34% and 14.71% respectively, these values are found within the range of soils recommended for the production of adobes, as shown in Table 1. According to the plasticity diagram, we can conclude that the soil studied can be considered a moderately plastic inorganic clay (2929. Rahmani A, Hazzab A, Aimer H. 2018. Identification and geotechnical classification of Ghardaïa loess (southern Algeria). Algerian J. Arid Environ. 8(1):88–103.). Table 2 also shows the physical and chemical properties of the soil studied. PH and salinity tests measured respectively by an inoLab™ 7110 benchtop pH meter and a Thermo Scientific Orion portable conductivity meter.

Table 1.  Geotechnical properties of the soil studied.

Atterberg limit	Value (%)	Recommended limits of adobe (3030. Le reglement parasismique pour les constructions en terre et instituant le Comite national des constructions en terre. 2013.)
Plastic limit PL	23.34	10%-25%
Liquid limit LL	38.05	25%-45%
Plasticity index PI	14.71	10%-25%

Table 2.  Geotechnical, chemical and physical characteristics of soil.

MBV	Specific surface (m2/g)	Porosity (%)	Bulk density (kg/m3)	SE	Absolute density (kg/m3)	PH	Salinity (ppT)
1.48	30.97	39.54	1615	2.77	2671	7.96	0.91

*MBV: Methylene blue value, *SE: Sand equivalent

Figure 3.  Soil identification tests carried out in the laboratory.

2.2.2. Mineralogical and chemical composition of the soil

 ⌅

The soil under study was ground, passed through a 2 mm mesh sieve, and then dried at 60°C for 24 hours, then prepared in sample molds for mineralogical characterization using a Bruker D8 Advance Twin diffractometer (Figure 6). The aim of this analysis is to roughly determine clay minerals qualitatively and semi-quantitatively (3131. El hammouti A, Channouf S, Charai M, Horma O, Miri H. 2023. Resource deposit, characterization and energy saving potential of olive pomace as a promising aggregate for energy efficient earth bricks in eastern Morocco. Constr Build Mater. 393:131989. https://doi.org/10.1016/j.conbuildmat.2023.131989). The diffractogram (Figure 4) shows that the soil studied contains two principal minerals, quartz (SiO₂) and calcite (CaCO₃) with percentages 34.3 % and 39.4% respectively, and the other minerals chlorite and murdochite are almost present but less abandoning. The chemical composition of the soil was determined by dispersive X-ray fluorescence spectrometry using the Panalytical 4 Kw Axios Dy 1856 (Figure 6) (Table 3). The oxides present in the soil are SiO₂ (24.78%), CaO (33.59%), however (Al₂O₃,Fe₂O₃, MgO...) are present in small quantities. The loss on ignition was evaluated after calcination of the sample at 1000°C, the value obtained is 31.37%.

Figure 4.  Mineralogical composition of the soil studied.

Table 3.  Chemical analysis of the soil studied.

Oxides	SiO2	Al2O3	Fe2O3	CaO	MgO	SO3	K2O	Na2O	P2O5	L.O. I
Mass %	24.78	5.91	2.30	33.59	2.13	0.04	1.07	0.10	0.06	31.37

2.2.3. Microstructure of the soil studied

 ⌅

The structural morphology of the raw materials was visualized by scanning electron microscopy (SEM) using by JEOL JSM IT-100 scanning electron microscopy (Figure 6). SEM enables the surface topography of the sample to be observed by scanning its surface with an electron beam and collecting the image formed. As the samples (soil and fibers) are non-conductive, surface metallization is required by coating them with a thin layer of gold. This can be achieved by two methods. For this study, sputtering was used. This involves depositing atoms torn from a piece of metal (gold) onto the sample using ionized argon in a partial vacuum chamber. The SEM image of the soil in Figure 5 illustrates a porous structure which is one of the main characteristics that allow a high sorption capacity and low thermal conductivity of unfired clay bricks. SEM analysis was combined with energy dispersive X-ray spectroscopy (EDS) to determine the chemical composition of the sample analyzed. As the spectra in Figure 5 show, SiO₂ and Al₂ O₃ contents are generally related to clay contents. SiO₂ is also a function of quartz percentages. EDS results indicate that the soil contains an average content of around 48.05% Si, Fe and Al. with an average content of around 49.36% SiO₂, 10.44% Al₂O₃ and 20.25% CaO. The high oxygen and silicon content explains the high quartz (SiO₂) peaks.

Figure 5.  SEM results and EDS analysis of the soil studied.

Figure 6.  Soil characterization tests carried out in the laboratory.

2.2.4. Characterization of fibers

 ⌅

X-ray diffraction analysis has been adopted to obtain qualitative information on the cellulose content of plant fibers. Cellulose is an essential parameter for evaluating the effectiveness of fibers in the mechanical stabilization of building materials, the diffraction pattern of coconut and juncus fibers shown in Figure 7 shows three main peaks at 2θ of 15.6, 22.19 and 34.7, corresponding to the main cellulose signals of plant fibers. The crystallinity index is one of the main properties of cellulosic fibers, which can be calculated empirically from the XRD spectrum as follows:

$C_{r}I=\frac{I_{200}-I_{am}}{I_{200}}\times 100$

 [1]

With I₂₀₀ corresponds to the maximum intensity of the crystalline part 2θ of the cellulose I, and Iam corresponds to the intensity of the peak which corresponds to the amorphous.

The crystalinity index of juncus is 42% compared with coconut fibre, which has a crystalinity index of 38.52%. This finding indicates that juncus fibers contain a significant amount of cellulose, which could be advantageous for developing effective building materials.

Figure 7.  Mineralogical composition of coconut and juncus fibers.

Table 4 below summarises some of the mechanical and physical characteristics of the fibres that were used in this research, and compares the various characteristics, i.e. tensile strength, thermal conductivity, water absorption and apparent density of each fiber, this comparison prioritizes the choice of the most suitable fiber for use as a construction material, with coconut fiber presenting more satisfactory characteristics than juncus, while the hydrophilic behavior of juncus fibers in the matrix may improve the composite’s characteristics, a hypothesis that will be treated as part of this study.

Table 4.  Mechanical and physical characteristics of the fibres employed in this research.

Property	Fiber type
	Coconut	Juncus
Apparent density (in kg/m3)	80	130
Tensile strength (in MPa)	110	40
Water absorption (in %)	175	260
Thermal conductivity (in W/m.K)	0.041	0.092

2.2.5. Microstructural characterization of fibers

 ⌅

The microstructure of juncus and coconut stems was determined by JEOL JSM IT-100 scanning electron microscopy (SEM), and a semi-quantitative analysis of chemical elements was carried out using an energy dispersive X-ray spectrometer (EDS). The microstructure of the outer surface of each rod presented in Figure 8 (R) shows the dimensions and external appearance of the coconut fibers, we observe linear lines in the form of roots parallel to the length of the fiber, which provides a rough surface, which can improve the mechanical coupling of the fibers with the clay matrix, influencing the behavior of the composite material after pointing. SEM observations of the cross-section of each rod show a porous structure for juncus fibers Figure 8 (W) and a firm structure for coconut, the pores contributing effectively to the thermal performance and lightness of the composites (Figure 12 and Figure 13), which explains the lower conductivity of the composite incorporated by juncus fibers (Figure 13a). The cross-section of coconut fibers remains smaller, with a difference of around 140 µm between the two cross-sections. Figure 9c and Figure 9d show the curves obtained from the EDS cross-section spectra of juncus and coconut respectively, each fiber revealing a significant amount of oxygen and carbon as the main constituents of these fibers. The amount of these two components remains lower for coconut fibers, which may be due to the porous structure of juncus fibers, and chlorine is another component observed in the coconut cross-section, which may be a contaminant from the outside.

Figure 8.  Fiber microstructure, (R) coconut fiber microstructure, (W) juncus fiber microstructure.

Figure 9.  SEM results of studied fiber stems, (a) longitudinal section of juncus stem, (b) longitudinal section of coconut stem, (c) EDS analysis of juncus, (d) EDS analysis of coconut fiber stem.

2.3. Adobe formulation and preparation procedure

The adobes were prepared by hand-mixing dry clay soil (< 2 mm) with different percentages by volume of the fibers (coconut and juncus fibers). The composites were prepared using the following process: First, the coconut fibers were immersed in water for 24 hours to saturation before being blended and for the local juncus fibers were dried naturally for 25days, then broken and ground, then the fibers were sieved through a sieve with a mesh size of 2 mm and a maximum length of 1.5 cm. first, the soil and fibers were mixed homogeneously, then water was added. The mixture was then stirred until a homogeneous blend was obtained. This operation was repeated for each type of mixture. The tests were carried out in the laboratory under conditions of temperature between 19°C and 23°C and relative humidity between 60% and 64%. Details of the proportions used in the composition of the mixture are given in Table 5. Contrary to current studies on agro-based (3232. Charai M, Mghazli MO, Channouf S, El Hammouti A, Jagadesh P, Moga L, et al. 2023. Lightweight waste-based gypsum composites for building temperature and moisture control using coal fly ash and plant fibers. Constr Build Mater. 393:132092. https://doi.org/10.1016/j.conbuildmat.2023.132092, 3333. Charai M, Sghiouri H, Mezrhab A, Karkri M. 2021. Thermal insulation potential of non-industrial hemp (Moroccan cannabis sativa L.) fibers for green plaster-based building materials, J Clean Prod. 292:126064. https://doi.org/10.1016/j.jclepro.2021.126064) small fiber/clay ratios were taken into account in this study (0, 10, 20% by volume). The composites were produced manually in the laboratory using parallelepiped molds with dimensions (4 × 4 ×16) cm³ and cylinder molds with dimensions (D= 100 cm and thickness = 4 cm). For compacting, the same ratio water/soil (w/s) was used for each mixture. As discussed by (3434. Ouedraogo M, Dao K, Millogo Y, Aubert JE, Messan A, Seynou M, et al. 2019. Physical, thermal and mechanical properties of adobes stabilized with fonio (Digitaria exilis) straw. J Build Eng. 23:250-258. https://doi.org/10.1016/j.jobe.2019.02.005) the required (W/s) ratio was calculated using Equation [2]:

With : PL and LL are the calculated plasticity limit and liquidity limit, respectively, in %. The steps involved in preparing biobased adobes are illustrated in the Figure 10.

3.1. Physical test 

In the last few years, the lightness of building materials has become a critical parameter in minimizing greenhouse gas emissions associated with transportation. Measured in the dry state, the apparent density of the samples was calculated as the ratio of the dry mass to the apparent volume of the sample, which was deduced from the dimensions measured using 0.01 mm precision calipers. Three samples were prepared for each percentage to conduct various tests and obtain multiple analyses.

3.2. Thermal test 

The thermal conductivity of the composites developed is measured using a TPS 1500 hot disk apparatus marketed by Thermo-Concept. This method is based on a transient planar source (TPS) technique, which involves studying the rapid transient response of samples in accordance with ISO 22007-2 (3535. Hot Disk AB, Hot disk thermal constants analyser instruction manual. Revision 2015–04-15, 2015.). The measurement protocol is based on tracking the time-dependent evolution of temperature across samples via a double spiral sensor, which serves both as an electrical heater and as a temperature sensing element (Figure 11).

First, the sensor is inserted between two samples of the material to be characterized. The dimensions of the material must be sufficient to allow the assumption of a semi-infinite medium. Then, after thermal stabilization, the samples are heated through the sensor with a direct electric current. The theoretical framework as well as the governing equations of the hot disk method are detailed in (3636. Mihiretie B, Cederkrantz D, Rosén A, Otterberg H, Sundin M, Gustafsson S, et al. 2017. Finite element modeling of the Hot Disc method. Int J Heat Mass Transf. 115:216-223. https://doi.org/10.1016/j.ijheatmasstransfer.2017.08.036, 3737. Gustafsson SE. 1991. Transient plane source techniques for thermal conductivity and thermal diffusivity measurements of solid materials. Rev Sci Instrum. 62(3):797-804. https://doi.org/10.1063/1.1142087).

3.3. Mechanical Test 

For each specimen developed, mechanical strengths were obtained at 28 days of age, on three 16x4x4 cm³ prisms for the bending test and 4cm*4cm*8cm for the compression test, using Controls Automax5 machine equipped with the Microdata Autodriver software, according to standard NF EN 196-1 (3838. Association Française de Normalisation (AFNOR). 2016. NF EN 196-1: Métodos de ensayo de cementos. Determinación de resistencias. Saint-Denis (Francia): AFNOR.). The specimens were dried in an oven at a uniform temperature of 60°C until a constant mass was obtained. For bending tests, the specimens were subjected to a uniform loading rate of 50 N/s until failure. The flexural strength of the specimens was calculated using the following expression:

With,

The transverse half of the specimen have been subjected to compressive stress. A compression test plate 40.0 mm (± 0.1) in width and 40.0 mm (± 0.1) in length receives the specimen halves, with a load rate of 2400 N.s-1 (± 200) applied. Compressive strength is determined using the following formula:

With :

4.1. Apparent density

The variation in dry composite bulk density as a function of fiber volume is shown in Figure 12. As fiber volume increases, the dry density of the composites decreases. For composites incorporated with coconut fibers, the density varies from 1598.83 kg/m³ for the composite without fibers to 1549.63kg/m³ for the composite containing 20% of fibers, corresponding to a reduction of around 3.07%, and for composites incorporated with juncus fibers, the density varies from 1598.83 kg/m³ for the reference sample to 1501.83 kg/m³ for the composite incorporated with 20% of fibers, with a rate of decrease of 6.06%. Similar trends have been observed in previous studies (2525. Omrani H, Hassini L, Benazzouk A, Beji H, ELCafsi A. 2020. Elaboration and characterization of clay-sand composite based on Juncus acutus fibers. Constr. Build. Mater. 238:117712. https://doi.org/10.1016/j.conbuildmat.2019.117712, 3939. Martínez ML, Eliche D, Cruz N, Corpas FA. 2012. Utilization of bagasse from the beer industry in clay brick production for building. Mater. Construc.62(306):199-212. https://doi.org/10.3989/mc.2012.63410) earth blocks/bricks reinforced with other natural fibers. The decrease in density was expected, as the fibers have a low fiber density (130 kg/m³ for juncus fibers and 80 kg/m³ for coconut fibers) compared to the density of the soil (1615kg/m³) (Table 2), Consequently, the increase in fiber content displaced the heavier soil content, resulting in a decrease in adobe density. In fact, the bulk densities of composites incorporated by juncus fibers improved more than those incorporated by coconut fibers, and these results are due to the greater water absorption capacity of juncus fibers (260%) compared to that of coconut fibers (175%), which would cause high matrix porosity during the drying process.

4.2. Thermal properties 

Thermal conductivity is a crucial thermophysical characteristic of construction materials (4040. Kazmi SMS, Munir MJ, Patnaikuni I, Wu YF, Fawad U. 2018. Thermal performance enhancement of eco-friendly bricks incorporating agro-wastes. Energy Build. 158:1117-1129. https://doi.org/10.1016/j.enbuild.2017.10.056), It quantifies how readily heat is able to transfer through a material under steady-state conditions (4141. Ali ME, Alabdulkarem A. 2017. On thermal characteristics and microstructure of a new insulation material extracted from date palm trees surface fibers. Constr Build Mater. 138:276-284. https://doi.org/10.1016/j.conbuildmat.2017.02.012), Thermal conductivity for both juncus and coconut fibers is illustrated in Figure 13a. According to the conductivity curves previously quoted, it is evident that conductivity decreases with increasing fiber volume, varying from 0.6349 W/m·K for the reference sample to 0.5565 W/m·K for the composite incorporated with 20% juncus fibers, and 0.5676 W/m·K for coconut fibers. Consequently, the addition of 20% juncus fibers improved the thermal performance of the material by around 12.34%, and 10.6% for coconut fibers. This can be due to the low thermal conductivity of the juncus and coconut plant fibers compared to that of the clay matrix, and by the increased porosity of the material when the fibers were added. The use of additives with low thermal conductivity leads to the manufacture of low-conductivity composites (4242. Arezki S, Chelouah N, Tahakourt A. 2016. The effect of the addition of ground olive stones on the physical and mechanical properties of clay bricks. Mater. Construc. 66:0465-2746. https://doi.org/10.3989/mc.2016.00815

43. Ali M, Alabdulkarem A, Nuhait A, Al-Salem K, Iannace G, Almuzaiqer R, et al. 2020. Thermal and acoustic characteristics of novel thermal insulating materials made of Eucalyptus Globulus leaves and wheat straw fibers. J Build Eng. 32:101452. https://doi.org/10.1016/j.jobe.2020.101452-4444. Ali M, Alabdulkarem A, Nuhait A, Al-Salem K, Almuzaiqer R, Bayaquob O, et al. 2021. Thermal analyses of loose agave, wheat straw fibers and agave/wheat straw as new hybrid thermal insulating materials for buildings. J Nat Fibers. 18(12):2173-2188. https://doi.org/10.1080/15440478.2020.1724232). In fact, the decrease in thermal conductivity is relatively small for juncus incorporated adobes compared with those incorporated by coconut fibers, as we have seen from the microstructure of coconut and juncus fibers (Figure 9), juncus fibers have a more porous structure than coconut fibers, allowing juncus fibers to absorb a significant amount of water (Table 4). The presence of water in the pores of the fibers increases the contact surface between fiber and matrix due to the stress exerted by the water trapped in the pores after drying. The stripping of the water causes the contact surface of the fibers to detach from the matrix, a large gap between the fiber/matrix contact surface is noted for juncus fibers, indicated by the microstructure Figure16b. The development of porosity in composites through the insertion of natural fibers reduces the apparent density of composites. However, the lightness of the samples contributes efficiently to the decrease in thermal conductivity. This dependency between lightness and thermal conductivity (conductivity as a function of density) is always confirmed for materials based on mineral matrix and vegetal fibers. According to Khoudja et al. (4545. Khoudja D, Taallah B, Izemmouren O, Aggoun S, Herihiri O, Guettala A. 2021. Enhanced thermo-physical properties of gypsum composites using olive pomace waste reinforcement. Constr Build Mater. 270:121824. https://doi.org/10.1016/j.conbuildmat.2020.121824) and Saghrouni et al. (2424. Saghrouni Z, Baillis D, Jemni A. 2020. Composites based on Juncus maritimus fibers for building insulation. Cem. Concr. Compos. 106:103474. https://doi.org/10.1016/j.cemconcomp.2019.103474) which used a clay matrix and Portland cement respectively. The corresponding relationships between dry bulk density and the thermal conductivity of composites with different volumes of coconut and juncus fibers are shown in Figure 13b and Figure 13c respectively. The thermal conductivity of the composites decreases with increasing bulk density. The variation observed is comparable to that described by various authors in past studies on lightweight materials, using different types of vegetal materials as palm, coconut, oil palm, bagasse, wood chips and jute fibers (4646. Danso H, Martinson DB, Ali M, Williams JB. 2015. Physical, mechanical and durability properties of soil building blocks reinforced with natural fibres, Constr Build Mater. 101:797-809. https://doi.org/10.1016/j.conbuildmat.2015.10.069).

4.3. Mechanical properties

Figure 14b illustrates the results of mechanical tests on composites based on clay incorporated with juncus and coconut fibers, while indicating their errors, calculated from the difference between the maximum and average values of the measurements obtained for each percentage. According to these figures, a decrease in flexural strength was observed for all fiber-reinforced samples Figure14a. The highest strength value was recorded for the clay-based composite with 10% coconut fiber, reaching 1.166 MPa, followed by the same blend with 20% coconut fiber reaching 0.84 MPa, a decrease of 0.3 MPa compared to the reference sample. However, the clay incorporated by the 10% and 20% juncus fibers has a value close to that recorded for 20% coconut, which remains quite modest with a decrease of 0.5 MPa compared to the value obtained by the samples reinforced with 10% coconut. Nevertheless, in Figure 14b. The compressive strengths of the two samples with 10% and 20% are lower than those of the reference sample, and the values obtained by the two fibers approach each other for the same percentages with a reduction rate of 24.71% and 32.57% for a percentage of 20% of coconut and juncus fibers respectively compared to the reference sample.

The contact surface between the clay and the coconut fibers and their orientation in the matrix, as well as the good tensile strength of the coconut fibers, are key factors contributing to the improved flexural strength of coconut reinforced composites compared to juncus reinforced samples, this observation is justified by the microstructural analysis in Figure 16, with the coconut fibers protruding from the matrix oriented perpendicular to the applied stress, making the coconut reinforced samples more cost effective in terms of flexural strength compared with juncus fibers.

From the results concluded, a decrease in the mechanical strengths of composites reinforced with coconut and juncus was observed, the decrease in flexural strength when adding natural fibers, has been detected by several authors (2727. Amazal M, Mounir S, Souidi A, Atigui M, Oubeddou S, Maaloufa Y, et al. 2024. Production and characterization of a composite based on plaster and Juncus maritimus plant fibers. Fluid Dyn. Mater. Process. 20(9):2059–2076. https://doi.org/10.32604/fdmp.2024.050613, 4747. Atigui M, Maaloufa Y, Souidi A, Amazal M, Oubeddou S, Demrati H, et al. 2024. Enhanced thermo-physical properties of gypsum composites using olive pomace waste reinforcement. Rev Compos Matér Avancés. 34(1):67-75. https://doi.org/10.18280/rcma.340109, 4848. Souidi A, Atigui M, Maaloufa Y, Amazal M, Oubeddou S, Mounir S, et al. 2024. Comparative study of gypsum composite materials reinforced with date palm and polyester fibres. Rev Compos Matér Avancés. 34(2):133-142. https://doi.org/10.18280/rcma.340202). The decrease in compressive strength observed for both composites containing juncus fibers is justified by the presence of pores and voids in the volume of the matrix, which accentuates its porosity reduction in densities Figure 12. Also poor fiber distribution has a direct impact on the mechanical stress performance of the samples, as illustrated by the microstructures shown in Figure 15, Several studies have shown that increased porosity redu has been cited in the literature, the characterization of earthen materials is important to preserve our architectural heritage and keep the same appearance and regidity of constructionsces mechanical strength while improving thermal performance (4949. Charai M, Salhi M, Horma O, Mezrhab A, Karkri M, Amraqui S. 2022. Thermal and mechanical characterization of adobes bio-sourced with Pennisetum setaceum fibers and an application for modern buildings. Constr Build Mater. 326:126809. https://doi.org/10.1016/j.conbuildmat.2022.126809, 5050. Serebe YAA, Ouedraogo M, Sere AD, Sanou I, Zagre WKJE, Aubert JE, et al. 2024. Optimization of kenaf fiber content for the improvement of the thermophysical and mechanical properties of adobes. Constr Build Mater. 431:136469. https://doi.org/10.1016/j.conbuildmat.2024.136469).

The non-destructive samples recovered after the mechanical test are put under observation to explain the mechanical results obtained, we noticed bar-like fibers protruding from the inner surface of the matrix as shown in the images in the Figure 15. These bars make a major contribution to the adhesion of the transverse faces of the clay composite, the coconut fibers are detached from the matrix, this observation can be explained by the external face of coconut which is smooth compared to that of juncus as shown in the Figure 16b, the morphology of the fibers impacts on the degree of adhesion of the additives and composites, also the dosage of the different fibers and their orientations destroys the homogeneity of the matrices as shown in Figure 15 below.

4.4. Microstructure

Figure 16 shows SEM images of the internal surfaces of the clay-based composite sections obtained from the bending test. The clay composite reinforced with coconut fibers Figure 16a shows a better bonding surface, illustrated by negligible cavities and pores between the fiber surface and the matrix. An approximately uniform distribution of fibers in the matrix, with no noticeable accumulation at specific points. This is due to the behavior of coconut fibers towards water, which tend to absorb some water during the preparation of the blends, unlike juncus fibers. The absorption of water by these juncus fibers generates cavities after evaporation of the post curing water. This phenomenon has been observed by several researchers and can be attenuated by chemical treatment of the fiber contact surface (5151. Saghrouni Z, Baillis D, Jemni A. 2020. Composites based on Juncus maritimus fibers for building insulation. Cem Concr Compos. 106:103474. https://doi.org/10.1016/j.cemconcomp.2019.103474).

The microstructures (SEM images) of the samples after bending tests are shown in Figure 16 in the mechanical testing section, the images show a large void between the clay matrix and the juncus fibers Figure 16b which contributed to the decrease in adhesion that negatively affects mechanical strengths, on the other hand image Figure 16a a shows a small void between the clay and the coconut fiber, Juncus fibers have a porous structure with a large geometry and high water absorption. After drying, the water escapes from the clay matrix, leaving voids between the matrix and the fibers. These observations explain the reduction in mechanical properties, which depend essentially on adobe porosity. The higher the number of pores in the adobe, the greater the reduction in mechanical strength (2525. Omrani H, Hassini L, Benazzouk A, Beji H, ELCafsi A. 2020. Elaboration and characterization of clay-sand composite based on Juncus acutus fibers. Constr. Build. Mater. 238:117712. https://doi.org/10.1016/j.conbuildmat.2019.117712). The increase in pores as the percentage of fibers in the clay matrix increases, as observed in these SEM images, also justifies the decrease in bulk density and the thermal properties of the composites obtained.