Keywords

1 Introduction

Revival of interest in the use of earthen construction materials can be seen over the last few decades due to the advantages such as thermal comfort, greenness and low environmental impacts (Cataldo-Born et al. 2016; Donkor and Obonyo 2015; Millogo et al. 2014; Minke 2006). However, earthen materials also have disadvantages such as low mechanical strength, poor erosion resistance and drying shrinkage cracking (Avrami et al. 2008; Minke 2000, 2006). The incorporation of industrialized fibers, such as polypropylene fibers (Donkor and Obonyo 2015; Yilmaz 2009), and of natural fibers, such as straw and even wool (Aymerich et al. 2012; Quagliarini and Lenci 2010), has been studied to mitigate some of the disadvantages of earthen materials. Among natural fibers, there have been recent studies addressing the use of jute fibers. Islam et al. (2008) studied the impact of natural fibers, including jute fibers, on strength and toughness of earthen blocks as well as shrinkage cracking of mortars and jute was found to be the most effective fiber, improving strength and toughness as well as preventing shrinkage cracking. Güllü and Khudir (2014) studied the effect of natural fibers, including jute fibers, on the unconfined compressive strength of fine-grained soil subjected to freeze-thaw cycles, and it was found that the incorporation of jute fiber increased both non-freeze-thaw and freeze-thaw unconfined compressive strengths compared to unreinforced soil. Saleem et al. (2016) assessed the effect of jute fiber on the compressive strength of earth bricks finding increments of compressive strength due to the incorporation of jute fiber. Although the contribution of the studies addressing the reinforcement of earthen materials with jute has been significant, there are still some properties that have not been studied intensely and quantitatively, such as impact resistance.

The novelty of this research resides in the incorporation of jute fiber into earthen materials such as adobe mixes, addressing quantitatively the impact of the incorporation of jute into flexural toughness, impact resistance and drying shrinkage cracking. This study has the following objectives: to evaluate the influence of different lengths and dosages of jute fiber on (i) flexural strength, (ii) flexural toughness, (iii) drying shrinkage performance and (iv) impact resistance in adobe mixes.

2 Materials and Methods

2.1 Materials

The present study uses a clayey soil from a district in Santiago, Chile. Araya-Letelier et al. (2018) conducted a characterization of the same soil, obtaining an 11%, 69% and 20% of clay, silt and sand, respectively. The liquid and plastic Atterberg limits, plasticity index and specific gravity of solids were also characterized, and their values are 29.1%, 17.4%, 11.7% and 2.51% respectively.

Jute is used as a natural reinforcing fiber for adobe mixes, and Table 24.1 summarizes some of the most relevant morphological, physical and mechanical properties of this vegetal fiber reported by Daniel et al. (2002).

Table 24.1 Morphological, physical and mechanical properties of jute fiber after Daniel et al. (2002)
Full size table

The jute fiber used in this study was processed to obtain three different lengths (7, 15 and 30 mm) to perform a sensitivity analysis on the impact of different lengths and dosages of fibers on the performance of adobe mixes. Considering the range of diameters (0.1–0.2 mm) reported by Daniel et al. (2002) and the three different lengths used in this study, the ranges of aspect ratios to be used are 35–70, 75–150 and 150–300 for fiber lengths of 7, 15 and 30 mm, respectively. Figure 24.1 shows a microscopy of the jute fibers.

Fig. 24.1
figure 1

Microscopy of jute fibers

Full size image

2.2 Adobe Mix Proportions and Specimen Preparation

A water-to-soil ratio (i.e., the weight of water to the weight of dry soil) of 0.307 was determined based on the required workability of the mixes. Seven different adobe mixes were cast for this investigation, six of them incorporating fiber and one plain mix, to make comparisons between them. The six jute-reinforced mixes had two dosages (0.5 and 2% by mass of soil) and three different fiber lengths (7, 15 and 30 mm). The different adobe mix ID numbers and fiber dosages are presented in Table 24.2, where the identification number (ID) indicates first the dosage and second the length of the fiber.

Table 24.2 Adobe mix ID number and material proportion
Full size table

The adobe mixes and the specimens were carried out manually. The fibers were mixed initially with the dry soil, until all clusters were disassembled. After all the fibers were included into the soil, water was added into the mixture in four equal portions, mixing after each incorporation. Finally, the adobe mixes were covered with a plastic bag to prevent water evaporation. The following specimens were cast for each adobe mix: (i) six notched RILEM beam specimens (160 mm × 40 mm × 40 mm, according to ISO/R679 (1968)) for impact strength test, (ii) six RILEM beam specimens for flexural strength, toughness indices and residual strength factors and (iii) two cylindrical specimens for drying shrinkage cracking. After 48 h, the specimens were demolded and kept at laboratory conditions (22 °C and 45% relative humidity) for 28 days until testing, rotating to the adjacent side 90° every 7 days. A fine coat of glue and sand was applied in the molds of the cylindrical specimens for ease of de-molding. The specimens for shrinkage crack testing were kept at laboratory conditions (22 °C and 45% relative humidity) for seven days in the molds, and cracking was measured afterward. Sketches in Fig. 24.2 show dimensions of the specimens.

Fig. 24.2
figure 2

Cylindrical specimen (right) and RILEM beam specimen (left) after Araya-Letelier et al. (2018)

Full size image

2.3 Testing

Flexural strength of each adobe mix was assessed at 28 days after casting using a three-point bending configuration over six RILEM beam specimens with a span of 130 mm between supports and using a displacement control protocol with a rate of 1 mm/minute.

To assess the energy absorption capacity of each adobe mix, flexural toughness indices were calculated according to ASTM C1018 (1997) procedure. The flexural toughness indices were calculated as the area under the load-deflection curve up to a specific deflection value divided by the area under the load-deflection curve up to the deflection where the first crack was observed (δ). Specifically, flexural toughness indices I5, I10 and I20 are the values obtained using deflections of 3δ, 5.5δ and 10.5δ, which are stated in ASTM C1018 (1997), respectively. Moreover, the residual strength factors, representing the average post-crack load retained over a deflection interval as a percentage of the load at the first crack, were estimated directly from the flexural toughness indices using Eqs. (24.1) and (24.2) (for further information, see ASTM C1018 (1997)).

$$ R_{5,10} = 20\left( {I_{10} - I_{5} } \right) $$
(24.1)
$$ R_{10,20} = 10\left( {I_{20} - I_{10} } \right) $$
(24.2)

where R5,10 and R10,20 are the residual strength factors between the intervals of 5.5δ and 3δ, and 10.5δ and 5.5δ, respectively. Mean and standard deviation of toughness indices and average residual strength factors for each adobe mix were obtained from the individual flexural load-displacement curves at 28 days after casting.

To measure the drying shrinkage cracking, widths and lengths were measured using a crack width comparator and a caliper, guided by a 20 × 20 mm grid, which was drawn on the surface of specimens. Crack width reduction ratio (CWRR) was calculated with respect to plain adobe, as well as crack width average (CWA) for all dosages.

To assess the impact energy absorption of the adobe mix after an age of 28 days, a projectile was thrown to the center of a RILEM beam specimen with a notch in the middle section as presented in Fig. 24.3. To avoid damage due to rebound, the specimens were glued with silicone to the steel frame. The impact energy was calculated using Eq. (24.3).

Fig. 24.3
figure 3

Impact test setup after Araya-Letelier et al. (2018)

Full size image
$$ E_{\text{C}} = n*m*g*h $$
(24.3)

where EC stands for the total energy at failure of the specimen, n is the required number of blows to cause failure to the specimen, m is the mass of the projectile (0.047 kg), g is the gravitational constant (9.8 m/s2), and h is the height of the fall of the projectile (0.496 m).

3 Results and Discussion

3.1 Flexural Strength, Toughness Indices and Residual Strength Factors

The results for flexural strength are presented in Fig. 24.4a, along with the error bars, considering one standard deviation above and below the mean value. Plain adobe had the worst flexural strength (0.37 MPa), whereas adobe mix ID 2.0–30 had the strongest flexural strength (0.88 MPa). Adobe mix ID 0.5–7 presented the less scattered data (SD = 0.05 MPa), whereas the most scattered data was obtained for adobe mix ID 2.0–30 (SD = 0.19 MPa). Analyzing the error bars, if the SD value of mix ID 0–0 is added to the mean, and the SD value of mix ID 2.0–30 is subtracted from the mean, the resulting value is still larger for ID 2.0–30. The latter might indicate that the length and dosage of jute fiber in adobe mixes do increase its flexural strength. The increase in flexural strength when jute is incorporated might be explained by the adequate mixing of the fibers and the satisfactory bonding between long fibers and the matrix.

Fig. 24.4
figure 4

a Flexural strength results and b compressive strength results

Full size image

Typical load-deflection curves are presented in Fig. 24.4b for each adobe mix, tested as mentioned in Sect. 2.3. Figure 24.4b shows that initially, the flexural stress increases linearly at the same stress-deflection rate until a visible crack is observed growing from the lower surface of the sample (under tension). The later occurs regardless of the fiber dosage, which indicates that up to the peak load, the flexural response of the adobe mixes depends on the mechanical behavior of the matrix mainly, which is considered as brittle. After reaching the peak load, adobe mix ID 0 presents a brittle failure mode, whereas fiber-reinforced adobe mixes present an enhanced post-peak performance.

Table 24.3 presents average and SD values of I5, I10 and I20 and the average values of residual strength factor for each adobe mix. It can be seen that all the values of toughness indices (I5, I10, and I20) for adobe mix ID 0-0 are approximately 1.0, which means that this adobe mix has a brittle mode of failure where its flexural strength is followed by a sudden drop of the mechanical load joined by unstable macroscopic crack formation. In contrast, all the fiber-reinforced adobe mixes exhibit a load recovery after the first crack, joined by an increment of the mid-span deflection before failure, and toughness indices as well as residual strength factors increase mainly due to the presence of jute fibers.

Table 24.3 Toughness indices and residual strength factors at 28 days after casting
Full size table

3.2 Restrained Drying Shrinkage Distributed Cracking

The results of the drying shrinkage cracking can be seen in Fig. 24.5, where one of the two shrinkage specimens is shown. As can be seen, cracking is consistently reduced as the fiber length and dosage is increased. The CWA values of mix ID 0-0 and ID 2.0–30 were 1.2 and 0.2 mm, respectively.

Fig. 24.5
figure 5

Restrained drying shrinkage cracking results

Full size image

CWA and CWRR results for each adobe mix are shown in Fig. 24.6 in the left side axis and right side axis, respectively. As was qualitatively observed in Fig. 24.5, the average width of the cracks, as well as the reduction ratio, is sensitive to the dosage and length of the fiber. ID 0–0 presented average values of crack width of 0.84 mm, whereas results for adobe mixes with jute fiber oscillated between 0.26 mm (ID 0.5–7) and 0.13 mm (ID 2.0–30). The corresponding CWRR values were 68 and 86%, respectively. As can be observed in Fig. 24.6, the CWRR results might indicate a higher sensitivity for fiber dosage than fiber length, where the increment for length is two percentage points and approximately 15 percentage points for increment of dosage. Antico et al. (2012) studied the behavior of material matrixes and the failure mechanisms for fibers, which is consistent with the effects observed by naked eye in this experiment. As the fiber dosages were increased, the characteristic length between the fiber and the matrix is reduced, increasing the chances that fibers are scattered throughout the matrix.

Fig. 24.6
figure 6

CWA and CWRR results

Full size image

3.3 Impact Strength

Mean results of the impact test energy absorption are shown in Fig. 24.7, as well as error bars corresponding to one standard deviation above and below the mean. It can be observed that as the fiber length and dosage is increased, the energy absorption also increases. Adobe mix ID 0.5–15 and ID 2.0–30 show average results of 2.17 and 38 J (16 and 281 times the energy required to collapse plain adobe), respectively. As was studied by Araya-Letelier et al. (2017), quasi-brittle materials like mortars are expected to enhance its toughness when natural fibers are incorporated into the matrix. Like mortars, reinforced adobe could behave as a quasi-brittle material, and likewise, as fibers are incorporated it is expected that its energy-absorbing capacity is improved.

Fig. 24.7
figure 7

Cumulative collapse impact energy

Full size image

4 Conclusions

The results show that there is an improvement in the flexure strength of jute fiber-reinforced adobe mixes when compared with those of plain adobe specimens. The jute fiber reinforcement also resulted in an increase of flexural toughness. Incorporation of jute fibers reduced crack widths. The energy absorption due to impact was considerably enhanced by the incorporation of jute fibers. The impact energy absorption capacity of adobe mixes with jute fibers is sensitive to the length as well as fiber content.