The balls of Posidonia are collected from Hergla’s beach in Tunisia. It was manually frayed and placed on a horizontal opener. The balls opened manually are driven by a rolling lurking and then they are engaged in a threshing cylinder. Subsequently, they are driven by means of a toothed roller in order to separate fibers. By centrifugal force and aspiration the good fibers are driven upwardly and the waste falls down. After this mechanical treatment we obtain the fibers shown in Figure 1 and considered as raw fibers.

Then, in order to improve the adhesion fiber-composite, the raw Posidonia fibers surface was modified using a combined chemical treatment: sodium hydroxide and hydrogen peroxide.

For the combined chemical treatment, raw Posidonia fiber was treated using a Datacolor AHIBA MSTRI. Then, 5 g of raw Posidonia fibers were immersed in a bath with a Liquor ratio of 1 by 40. This bath contains 25 ml/L of Hydrogen peroxide and 20 g/l of Sodium hydroxide. The whole mixture has been taken to temperature of 100˚C under agitation and pressure during 45 min. The treated fibers were thoroughly rinsed with water several times and there are dried to the ambient air for 48 h.

The treated and untreated fibers were characterized by means of physical and mechanical analysis in order to define their properties. The tests must carry out on a batch of conditioned fibers to a normal atmosphere (relative humidity: 65% ± 4%, temperature: 20˚C ± 2˚C).

The Posidonia fibers were used to reinforce an unsaturated polyester matrix. The composites were manufactured manually by placing fibers and resin inside a glass mould. In order to remove en-trapped air bubbles and

also enabled manufacturing all uniform thickness composites, the mould was closed and pressed. Three percentages by weight of fibrous reinforcement (raw and treated fibers) were made: 5%, 10% and 15%.

After composite material preparation, test samples were cutted from composite panel to use in the mechanical characterization with the three-point bending flexure test according to the standard EN ISO 14125 (1998).

The flexural properties of pur-polyester samples and polyester reinforced by Posidonia fibers were made using a universal testing machine LLOYD. A speed of 5 mm/min and a measurement cell of 5 KN were used. The values were reported as averages of 10 test replicates

The flexural strength (σ), the flexural modulus (E) and the surface elongation at break (ε) are calculated respectively by:

Flexural Strength Equation (1)

Flexural Modulus Equation (2)

Elongation at Break Equation (3)

where F is the maximum load (N), L is the range (mm), h is the specimen thickness (mm), b is the specimen width (mm) and w indicates the defection (mm).

To observe the surface topography and shrieking surface of composites, Scanning Electron Microscopy (SEM, HITACHI TM 3000) was then used at low and high magnification. The metallized specimens were analyzed in partial vacuum conditions (0.1 - 0.15 torr), and under an accelerating voltage ranging from 10 to 15 KV.

The mechanical and physical properties of raw and treated Posidonia fibers are presented in Table 1.

It is clear that the physical and mechanical properties were influenced after the chemical treatment. In fact, the fibers linear density decreases after this combined chemical treatment. This reduce of mass per unit of length could be attributed to the removal of waxy and gummy substances cross linking the ultimate fibers. Also the treated Posidonia fibers present a lower diameter than those untreated. This decrease in fiber diameter could be attributed to the influence of combined chemical treatment. Besides there is a decrease in fibres tenacity and elongation after combined chemical treatment (hydrogen peroxide and sodium hydroxide). This decrease attributed to the substantial delignification and degradation of cellulosic chains during chemical treatment.

Hence, change of the physical properties form depends strongly on the treatment of the fiber. It is important to indicate that good cohesion between fibers and matrix is assured by many parameters such as the surface area, the roughness and the surface tensile of fibers [18] [19] .

The realization of composite materials requires the use of a matrix and reinforcement. In this case, we test the potential of using Posidonia fibers as reinforcement with a polyester matrix.

The three point bending flexure tests were carried out on pure polyester specimen (Comp PES), polyester reinforced by raw Posidonia fibers and polyester reinforced treated Posidonia fibers.

Figure 2 shows the shape of the PES bending curves and PES/Posidonia fibers. It is noted from this figure that the failure mechanism for the single polymer is the same of the composite. The curve was clearly linear for the PES and Posidonia fibers reinforced polyester composites. Besides, the flexural properties increase with the addition of fibers.

Table 2 shows the results of three points bending flexural test of polyester matrix reinforced with the mixture of treated and untreated Posidonia fibers of a weight ratio of 10% of fibrous reinforcement.

According to this table, it can be noted that the reinforcement of polyester matrix with Posidonia fibers changes the flexural properties of composite material. Indeed, whatever the nature of the fiber reinforcement (raw or treated) added to the matrix, the modulus of elasticity has an increase. Thus, the rate of increase of this mechanical parameter is about 52.9% for the raw Posidonia fibers and 90.5% for the treated Posidonia fibers compared to pure polyester. This affirms well the trend widely illustrated in the literature: improving the mechanical response of the system after aggregation reinforcements in polymer matrices [20] [21] . Also we can notice that the flexural modulus and stress of composites reinforced with treated fibers is more important compared with those of untreated fibers. This increase of mechanical properties could be explained by the good adhesion between fiber and matrix. Then, the combined chemical treatment has enhanced the bonding between fiber and polymer tanks to the increase of the number of possible reaction sites (OH and CH groups) when increasing the cellulose amount exposed on the fiber surface. In fact, this could be proved by the broad band observed at 3000 - 3500 cm⁻¹ in the spectra (Figure 3) indicating the presence of OH group. Moreover, the second band was observed at 2857 - 2926 cm⁻¹ indicating the stretching vibration of the groups ?CH and ?CH₂ of cellulose and another band at 1450 cm⁻¹, which also indicates the presence of ?CH produced by a symmetrical deformation of lignin and alpha cellulose. Furthermore, the ratio of the intensities of the transmittance peaks at 3343 cm⁻¹ (?OH)

and 2900 cm⁻¹ (>CH₂, >CH-) for the raw (0.9937); (0.9985) and treated (0.985); (0.994) fibers indicated the presence of more ?OH groups in the treated fiber than in the virgin sample.

It is noted also, that the tenacity increases upon the addition of the reinforcement while the elongation decreases. Which means that the composite become more resistant but less deformable and this can be explained by the fact that adding fibers to the matrix let the bending load shared throughout all the composite including fibers that oppose bending to improve the load supported by the composite while blocking structure which reduces deformation.

The SEM micrographs of the fracture surface of composite specimens reinforced with treated and untreated Posidonia fibers presented in Figure 4 showed that adhesion between fiber and matrix enhances with the chemical treatments. In fact, as shown in Figure 4(a1), before modification, the wettability between Posidoniafibers and unsaturated polyester matrix seemed to be poor because of the presence of a gap between the two componements. This gap is much less visible in the case were the fibers were treated with combined process (Figure 4 (a2) and Figure 4(b2)). Alsowe can remark that the matrix surrounding the fibers led to a good adhesion. This result confirmed the enhancement of mechanical properties of composite reinforced with treated fibers. Thus, when the Posidonia fibers were treated with sodium hydroxide and hydrogen peroxide, the gap is much less pronounced and an improved interface corroborated well the enhancement of the previously mechanical properties observed in Table 2.

Whatever the nature of the fiber reinforcement (raw or treated) added to the polymer matrix, it is evident that the flexural modulus and strength increases when increasing content fiber. In fact, the polyester resin transmits and distributes the applied stress to the Posidonia fibers resulting in higher flexural modulus and strength. As result, the composite can sustain higher load before failure compared to the unreinforced polyester.

In fact, the flexural modulus increases when increasing weight ratio of raw fibers. However, for the treated fibers the modulus achieved a threshold of 4.61 Gpa at a fiber weight ratio of 10%.Then the flexural modulus is almost doubled compared to that of single matrix. As well as, exceeding this percentage of reinforcing we notice a reduction in this property (4.21 Gpa for a treated-fiber weight ratio of 15%). This decrease can be explained by the poor cohesion between fibrous reinforcement and matrix by increasing the fiber weight ratio (Figure 5).

We can also indicate that the addition of fiber reinforcement (raw or processed) increases the flexural strength of the material. This increase is significant in the case of the treated fibers. This can be explained by the fact that the treatment has improved adhesion between fiber-matrix. This constraint is very important in the case of reinforcement with a rate of 15% for the treated and untreated fibers.

However these tests demonstrate that the addition of fibers decreases the deformation at break of reinforced

composites compared to those non-reinforced. This decrease can be explained by the rigidity of the materials following the reduction of the free volume in the matrix alone. Also we can notice that the elongation at break is more important in the case of treated fibers with a fiber weight ratio of 10% and 15% compared to non-treated fibers.