<p>Soil tensile strength plays an important role in the hydro-mechanical behaviour of earth structures and slopes interacting with the atmosphere. Shrinkage-induced cracking may be generated by drying/wetting cycles, with consequent faster water infiltration from the top of slopes and reduction of the safety factor. Vegetation roots were proven to increase soil shear strength, but less is known about their effects on soil tensile strength. For this purpose, new equipment has been designed and used to induce plant growth in compacted soil samples and to perform uniaxial tensile tests on the reinforced material. The equipment is composed of two cylindrical moulds linked by a soil bridge in which the tensile crack is induced due to geometrical restraints.</p><p>For this study, silty sand was chosen and compacted at a low dry density (&#961;<sub>d</sub> = 1.60 Mg/m<sup>3</sup>) and at a water content w = 15%. After compaction, samples were gently poured with water up to a high degree of saturation (S<sub>r</sub> &#8776; 0.95) and low suction (s &#8776; 1 kPa). Then, six of them were seeded with Cynodon dactilon, adopting fixed seeding density and spacing. Plants were irrigated and let to grow for three months: during this period, suction was monitored by a tensiometer. Seven fallow specimens were prepared following the same procedure, for comparison purposes.</p><p>When ready, samples were dried in a temperature/relative humidity-controlled room and left in the darkness for three hours, to attain and equalise the desired value of initial suction. Finally, the tensile stress was induced on the soil by a displacement rate of 0.080 mm/min. For each test, suction was continuously monitored by a tensiometer while the water content was checked at the beginning and at the end. Moreover, the void ratio and the root volume and area ratio were assessed close to the crack generated, at the end of each test.</p><p>The hydraulic state affected the soil mechanical response upon uniaxial extension: an increase of strength and a more brittle behaviour were observed as suction was increasing. At the same suction, a higher strength was systematically observed in the vegetated soil. In fact, even at very low suction (i.e. s = 1 kPa), vegetation roots induced a considerable increase in soil tensile strength (i.e. 10 kPa). The soil hydraulic state also affected the root failure mechanism. In wet soil, the roots subjected to tension were stretched and pulled-out whereas in dry soil they experienced a more immediate breakage (i.e. in concomitance with the cracking of the surrounding soil). Some preliminary PIV (Particle Image Velocimetry) analyses showed differences among dry/wet and fallow/vegetated soils. Indeed, a more diffuse strain field was observed in vegetated samples, thanks to the redistribution of stresses induced by the roots.</p><p>Results were successfully interpreted by a well-established shear strength criterion for partially saturated soils, considering the degree of saturation, suction and soil microstructure. An increase of the soil shear strength was observed and correlated to the presence of roots and to their geometrical and mechanical features. Moreover, good consistency was detected with results coming from other equipment.</p><p>&#160;</p>
Root tensile strength and root distribution of typical Mediterranean plant species and their contribution to soil shear strength
