2.1. Materials

As a part of the composite, a high viscosity unsaturated polyester orthophthalic thermoset matrix was used, marked under the reference P-2002 by the company Industria de Resinas S.A.S. The manufacturer recommends the use of 2% p/p methyl-ethyl ketone peroxide (MEK peroxide) as a catalyst.

As a reinforcement, ceramic waste was used that comes from clay brick fragments, which were supplied by the warehouse of materials La Flecha, located in the municipality of Pamplona, Colombia. Initially, sintered clay powder with grain size (GS) less than 0.85 mm (< sieve #20) was obtained, processing the brick fragments in a ball mill.

Then, a sample of 5.37 kg of sintered clay powder (Fig. 1) was refined, under the conditions of ASTM C136/C136M, using sieves #40, #60, #80, #100, #140, #200, and the bottom [¹⁰].

Source: The Authors.

Figure 1 Obtaining process of the sintered clay powder. 

Finally, the granulometry of 100 (m > GS > 150 (m (under sieve #100 and over sieve #140) was selected, which had a presence of 9.31% in the total of the sample, without apparent macrometric impurities. The sample was dried in an oven at 100 ° C for 24 hours, before being mixed with the polyester resin.

2.2. The manufacturing process

The manufacturing process used was cast molding, as shown in Fig. 2. Initially, the amounts of resin and reinforcement were weighed, taking as a reference a percent of 10% p/p of sintered clay powder and 2% p/p of catalyst. Later, mixing was done with the aid of magnetic stirring at a temperature of 80°C, in order to achieve optimum dispersion of the particles in the polymer matrix [¹¹].

Source: The Authors.

Figure 2 Scheme of the manufacturing process. 

Finally, the contents in the molds were emptied and allowed to cure at 20°C for 7 days.

2.3. Flexion mechanical characterization

The three-point bending was executed in a universal testing machine SHIMADZU UH-600kN, following the conditions of ASTM D790 [¹²]. The dimensions of the specimen were 150x10x6 mm (length x width x thickness), with a length of 75 mm between supports (support span). The rate of crosshead motion R was 1.80 mm/min and was calculated with Eq. (1):

When, Z is rate of straining in mm/mm/min, L is support span in mm, and d is depth of beam in mm (thickness of material) [¹²].

In addition, the test was performed on five specimens of polyester resin (PR) and five specimens of the polyester resin composite with sintered clay powder (PR-SCP), for comparative purposes.

2.4. Scanning electron microscopy

The fracture surface obtained in the flexural test of the PR-SCP sample was analyzed using a scanning electron microscope TESCAN, model MIRA 3 FEG-SEM. The sample was metallized in gold plating to improve electrical conductivity.

2.5. Break lever simulation

The geometry of the pieces was drawn in Solidworks computer-aided design software, based on the dimensions of a sport-tourism motorcycle brake lever. These mechanical parts are made of aluminum casting or forged steel, and their function is to push the brake fluid to produce the braking of the front wheel. The mechanical requirements of these parts depend on the force exerted by the driver's hand [¹³].

An image of the rear lever is shown in Fig. 3a and Fig 3b corresponds to its virtual model (original CAD). Fig. 3c represents a modified CAD, which was necessary to redesign with respect to the original geometry, in order to facilitate the manufacturing process and meet the mechanical requirements by varying the assignment of material in the finite element method simulator.

Source: The Authors.

Figure 3 Geometries: (a) Brake lever of sport-tourism motorcycle. (b) Original CAD. (c) Modified CAD. 

Operating conditions for a brake lever were parameterized by real performance in limited circumstances, taking into account that the braking system is the most important component in vehicles to maintain the integrity of the passengers [¹⁴]. Consequently, it was assumed that the time needed to reach the maximum grip strength was less than 2 seconds, and the optimal values of maximum voluntary contraction are obtained within a period of 3 seconds [¹⁵]. Likewise, according to an anthropometric study conducted by Lee & Sechachalam (2016), a value of 320 N (32.60 kg) was taken as the grip strength maximum exerted by the dominant hand [¹⁶].

In accordance with the above, three simulation scenarios were proposed by the finite element method (FEM), taking into account the same operating conditions, to analyze the behavior in the design of a brake lever, with the help of the ANSYS R16 software:

For the three aforementioned cases, the material was assumed as homogeneous and isotropic. In addition, the mechanical properties of the aluminum alloy were taken from ANSYS Engineering Data (Young modulus of 71 GPa, yield strength of 280 MPa, and ultimate strength of 310 MPa), and those of the PR-SCP composite were taken from mechanical characterization to bending.

Table 1 shows the most important characteristics of the mesh for each of the geometries (CAD discretization).

Table 1 Data of the geometry mesh. 

Source: The Authors.

Tetrahedral and hexahedral elements with average size of 1mm and curvature normal angle center of 15°, was used for mesh of original and modified CAD, respectively. Also, proximity and curvature advanced size function with relevance center and smoothing high, was implemented.

These parameters describe the mesh that converges to the solution, using the appropriate number of elements to reduce computational expenditure, with the Mechanical APDL solver.

The setup of the simulation parameters was performed as shown in Fig. 4. When, A is standard earth gravity of 9.81 m/s², B is hand force of 320 N (32.60 kg), C is a frictionless support, D is a cylindrical support free tangentially, and E is an elastic support (5 N/mm³).

Source: The Authors.

Figure 4 Setup of the simulations parameters. 

The simulated operating time was 10 seconds with an application period for the maximum grip force of 3 seconds, as shown in Fig. 5. The other parameters remained constant with respect to time. In each of the proposed scenarios, the mass of each geometry, the equivalent stresses (Von Mises) and the static load safety factor were determined, using Eq. (2) for the latter.

Source: The Authors.

Figure 5 Grip strength during time. 

When, SF is safety factor of application,σ _max is flexural strength of material in MPa, and σ _equ is equivalent stress in Mpa (Von Mises).

The purpose of the simulation is mainly to determine the influence of the change of material and the geometry of the piece, in the mechanical behavior of the levers studied.

3.1. Flexural properties

Fig. 6 and Table 2 show the results of the flexural test, where a linear-elastic behavior is evident in both materials. It is important to note that the nature of the thermoset matrix of PR is fragile and, therefore, the PR-SCP composite material adopts the same behavior.

Source: The Authors.

Figure 6 Stress-Strain diagram of PR and PR-SCP composite. 

Table 2 Flexural properties of materials. 

Source: The Authors.

However, the RP-SCP composite obtained a flexural modulus of 1.17±0.19 GPa, a flexural strength of 48.32±5.89 MPa, and a deformation to fracture of 7.52±1.28%. This demonstrates that sintered clay powder reinforcement reduces stiffness and fragility, and increases mechanical strength, compared to PR without reinforcement, which has a flexural modulus of 1.30±0.25 GPa, a deformation to fracture of 5.89±1.72%, and a flexural strength of 45.23±5.34 Mpa.

3.2. Analysis of morphology of PR-SCP composite

In Fig. 7 the fracture surface is observed characterized by cracks in different planes that form the river lines (black arrows), and corroborate the fragile behavior described in the destructive test [¹⁷,¹⁸]. In addition, two phases are distinguished: a continuous one that has smooth surfaces (thermosetting resin), and the sintered clay powder microparticles that are uniformly dispersed and wrapped within the matrix.

Source: The Authors.

Figure 7 SEM micrograph for fracture surface of PR-SCP composite. 

The micrograph of Fig. 8 shows big particles with irregular shapes, which have a size distribution that averages a value of 18.83±9.38 µm. It could be considered that these particles originate as a product of the union of finer grains during the manufacturing stage, giving rise to the formation of agglomerates that would act as stress concentrators, and consequently to the premature failure of the material.

Source: The Authors.

Figure 8 Micrograph with magnitude 500x and scale 100 µm. 

In Fig. 9 the presence of finer particles of spheroidal shape with sizes ranging in average of 0.89±0.43 µm. These micrometric fragments have an optimal dispersion in the matrix, due to the good aspect ratio (l/d), whereby a greater interfacial area per unit volume is generated between the clay grains and the polyester resin, and by both is reflected in the mechanical properties.

Source: The Authors.

Figure 9 Micrograph with magnitude 100x and scale 5 microns. 

3.3. Lever validation by finite element method

In Fig. 10 the values obtained in the simulations are shown, with respect to the distribution of stresses and the critical areas that are presented in each of the scenarios. It should be noted that the equivalent stress and the safety factor are directly related and are inversely proportional, as shown by Eq. (2).

Source: The Authors.

Figure 10 Maximum stress distribution at t = 5s. (a) Simulation 1. (b) Simulation 2. (c) Simulation 3. 

Table 3 shows the quantitative results of the simulations. In simulations 1 and 2, a similarity in stress distribution is evidenced, with maximum values of 105.97 and 105.88 MPa, respectively (Fig. 10a and 10b). However, there is a decrease in the safety factor, from 2.64 to 0.46, and a reduction in the mass of 44.51 g. Consequently, there is a decrease in the performance of the original part, when the properties of the PR-SCP composite are assigned, and therefore, it is not recommended to manufacture the lever with this geometry using the polyester composite material/10% waste.

Table 3 Quantitative data of simulation results. 

Source: The Authors.

On the other hand, in simulation 3, the modified geometry reveals a significant improvement in stress distribution, with a maximum value of 21.70 MPa (Fig. 10c). This result is directly attributed to the simplification of geometry and the increase in volume in critical areas. Accordingly, the minimum safety factor in the proposed piece reaches a value of 2.23. In accordance with the above, it was possible to validate compliance with the minimum mechanical requirements at static load required by the application, and additionally there was a reduction in the mass of 21.43 g, with respect to the original geometry made of aluminum alloy.

3.4. Manufacturing of brake lever

According to the results obtained in the simulations, the manufacturing of a lever prototype was made, taking as reference the modified geometry and the PR-SCP composite material proposed in this research project. The above was executed in three phases:

In the first phase the mold design was carried out in Solidworks, to later be manufactured in MDF (medium-density fibreboard). The layers were laser cut, in order to obtain accuracy in the geometry of the lever (Fig. 11).

Source: The Authors.

Figure 11 Phase 1: design and manufacture of the mold. 

In the second phase (Fig. 12), the mold was coupled and prepared, impregnating a thin layer of mineral oil on the walls. Then, the polyester resin composite material with 10% sintered clay powder was emptied into the mold, guaranteeing the complete filling of the material in the interior. Subsequently, it was allowed to cure for 7 days, under standard conditions at room temperature and pressure.

Source: The Authors.

Figure 12 Phase 2: preparing of the mold and casting. 

In phase 3, the piece was assembled, making adjustments for the final geometry, without apparent surface imperfections, and with options for improvement from the ergonomic point of view. Finally, the part was assembled on the motorcycle handlebar, correctly fulfilling its position (Fig. 13) and function. However, it is necessary to carry out exhaustive tests of the prototype, so that in the future its real mechanical behavior can be validated, and not only virtually.

Source: The Authors.

Figure 13 Phase 3: demolding, final form and assembly.