**3. Results**

#### *3.1. Change in Density*

The variation of particleboard density was determined by the recording of weight of all particleboard at the begin and at the end of the curing period in the wood mold (from 0 to 3 days). A reduction of weight of about 5% during that period occurred due to water evaporation through the mold. After removal from the mold, the particleboard mass typically reached a plateau at about 6 days, meaning that most of the free water in the mortar had evaporated in the conditioning chamber at 23 ◦C and 60% R.H by then. At 14 days, the particleboards had a specific gravity ranging between 0.68 and 0.70.

#### *3.2. Scanning Electron Microscopy*

According to the results shown in Figure 5, both materials show a low porosity and pore sizes smaller than 10 μm. Based on SEM examination, the difference of microstructure of a mixture of neat cement-wood and a mixture containing 15% of SP in replacement of cement is not significant. Both materials show a uniform and dense microstructure.

**Figure 5.** Scanning electron microscope images of cement-wood particles (**a**) and cement +15% replacement of powder steatite+ wood particles (**b**).

#### *3.3. Bending Properties*

Figure 6 presents the evaluation of bending properties obtained for WCSP and GB. Three replicates per products were tested. The average values of elastic modulus and bending strength for each tested specimen are shown in Table 4.


As mentioned previously, 28 days after casting, the particleboards were tested in static bending using in each case six specimens. Separate test series were carried out with the front or back face subjected to bending test. Typical load-deflection curves obtained for these different test configurations are displayed in Figure 6. WCSP were tested with the load being applied on the front or on the back face of the samples. As explained before, due to the settling of the SP, the mortar at the front face of the WCSP has a less porous, denser microstructure than at the back face. Therefore, the bending strength when the load is applied on the front face of the sample is lower than on the back face of the WCSP. The WCSP with the load on the back face exhibited a bending strength of 5.1 MPa, which is over 1.9 times more than it is on the front face. The analysis of the stress–displacement curves indicates that there are three material behavior stages in the course of the test that corresponds to the mechanical behavior in the constituent composite materials (wood-cement-steatite). Each experimental curve included a linear period at the beginning of the test and a non-linear period later. The linear period represents the elastic behavior of the material. The tangent elastoplastic modulus decreases in the second period, which corresponds to a non-linear plastic behavior, the third period corresponds to the last part of the curve when the WCSP began to fracture. The MOE of WCSP is about 1.7 GPa with the load applied on the front face and about 2.1 GPa with the load applied on the back face.

**Figure 6.** Characteristic stress–displacement curve for a three-point bending test of WCSP and GB in accordance with ASTM D1037-12.

For the GB, separate bending tests were conducted with the samples oriented perpendicular and along the direction of the overlay paper fibers (Figure 6). The respective stress–displacement curves in bending were entirely different. In the overlay paper fiber direction, the GB exhibit a fragile behavior, while it is more ductile in the perpendicular direction. This is due to the different tensile properties of the overlay paper in the two orthogonal directions. Therefore, for samples oriented perpendicular to the paper fiber direction, the paper failed at the beginning of the test and did not have a significant effect on the bending test. For samples oriented in the paper fiber direction, the overlay paper had a significant contribution to the mechanical properties and played the role of a reinforcement. Table 4 shows that the MOR in the overlay paper fiber direction is 5.4 MPa, which is 3.4 times higher than in the perpendicular direction. It is approximately equal to the MOR of WCSP in the case of a load applied on the back face and two times higher than in the case of a load applied on the front face. In fact, the mechanical quality of the GB depends significantly on the gypsum core properties. Therefore, the whole GB failed as the reinforcement failed. That is why the behavior is fragile (Figure 6). This mechanical comportment of GB was also noticed in the study of P. Tittelein et al. [6]. The GB bending MOE in the overlay paper fiber direction is 1.9 GPa, which is 1.5 times higher than it is in the

perpendicular direction, 1.1 times lower than the bending MOE of WCSP in the case of loading on the back face and 1.1 times higher than it is in the case of loading on the front face. The results reveal that the MOR and MOE of WCSP in the case of loading on the front face are lower than in the GB overlay paper fiber direction and higher than across the GB overlay paper direction. However, the WCSP still could replace GB when adjusting the distance of the studs in the wall composition. A good bending strength of WCSP in the case of loading on the back face is an advantage for transportation and installation.

#### *3.4. Screw Withdrawal and Nail Pull Test*

The results of the screw withdrawal and nail pull resistance tests are shown in Table 5. According to these results, WCSP has better resistance to screw and nail withdrawal than GB. The recorded screw withdrawal resistance and nail pull resistance of WCSP are respectively 37 and 11% higher than the corresponding values recorded for the GB. The resistance values of GB show less variation, as it is a more homogeneous material than WCSP.


