MATERIALS AND METHODS

Ethical approval

This study did not involve live animals, animal tissues, or biological specimens and therefore did not require approval from an Institutional Animal Care and Use Committee or any other institutional ethics committee. The study was conducted exclusively on synthetic materials for the development of canine CBMs. The research was undertaken in accordance with the principles of the 3Rs (Replacement, Reduction, and Refinement), with the objective of reducing the use of cadaveric specimens in veterinary surgical education by developing a reproducible and cost-effective synthetic training model.

Study period and location

The study was conducted from November 2021 to May 2022 at the Faculty of Veterinary Medicine, Rajamangala University of Technology Tawan-ok, Chonburi, Thailand. Mechanical testing was performed at the Department of Mechanical Engineering, Faculty of Engineering, Kasetsart University Sriracha Campus, Chonburi, Thailand.

Study design

This experimental materials engineering study was designed to optimize the mechanical performance of hand-pour fabricated fiber-reinforced polymer composites for canine CBMs. R and E were evaluated as cortical bone analog materials, whereas PU foam was evaluated as a cancellous bone analog. Four fiber reinforcement levels (0, 1, 3, and 5 wt%) were investigated for each resin system. Tensile mechanical properties were determined according to American Society for Testing and Materials (ASTM) D638-14 and compared to identify the formulation providing the best balance between mechanical performance, manufacturing reproducibility, and casting feasibility.

Specimens

R and E were selected as candidate materials for the cortical bone analog, whereas PU foam was selected as the cancellous bone analog. R (R804) was obtained from Super Silicone & Resin Art, Bangkok, Thailand. E (EP-089), E-glass fiber filament, and PU foam were obtained from Concrete Composite Co., Ltd., Bangkok, Thailand. The principal properties of the polyester and Es provided by the manufacturers are summarized in Table 1.

Properties | Polyester resin R804 | Epoxy resin EP-089
Appearance | Turbid pink liquid | Clear transparent viscous liquid
Viscosity (cP at 25°C) | 300–400 | 11,500–15,000
Density (g/cm³ at 25°C) | NA | 1.16
Total solid content (%) | 56 ± 2 | NA
Gel time (min) | 15–25 | 30
Curing time (h) | 1–2 | 4–6
Appearance after curing | Pink solid | Nearly transparent solid
Cost (USD/kg) | 7.56 | 19.01

Five specimens were fabricated for each material composition using a custom silicone rubber mold (Figure 1C) at room temperature (25°C) according to ASTM D638-14 Type I specifications [25] (Figure 1A and B). For the cortical bone analog, chopped 6-mm E-glass fibers were incorporated into the resin matrix at fiber reinforcement levels of 0, 1, 3, and 5 wt%.

For E (EP-089), Parts A and B were mixed at a volumetric ratio of 2:1. For R (R804), catalyst was added at 1.5% (w/w). From the second mention onward, Super Silicone & Resin Art and Concrete Composite Co., Ltd. are referred to by company name only.

For fiber-reinforced specimens, chopped fibers were gradually incorporated into the base resin before addition of the hardener or catalyst while continuously stirring with a flat spatula to ensure complete wetting of each fiber increment. After reaching the desired fiber weight fraction, the hardener (E) or catalyst (R) was added, and the mixture was stirred until visually homogeneous (approximately 1–2 min).

To minimize air entrapment, mixing was performed using a continuous figure-of-eight motion while maintaining the spatula below the liquid surface. The resin mixture was subsequently poured into paste wax-treated molds from a height of approximately 30 cm using a thin continuous stream to facilitate bubble release. Vacuum degassing was not performed. Fiber distribution within the cured specimens was verified by visual examination of the fracture cross-section after tensile testing. Initial curing before demolding required approximately 2 h for R and 8 h for E. Specimens exhibiting visible voids, trapped air bubbles, or surface defects were excluded from subsequent analyses.

For the cancellous bone analog, a two-component PU foam consisting of polyol (Part A) and isocyanate (Part B) was prepared at a 1:1 volumetric ratio according to the manufacturer's instructions. The mixture was stirred for approximately 30 s until homogeneous. The system expanded to approximately 25 times its original liquid volume, producing a low-density closed-cell foam with a free-rise density of 27–31 kg/m³. Following casting, the foam reached its maximum expansion within approximately 1 min.

The apparent density of each cured PU specimen was determined gravimetrically by dividing specimen mass (measured to 0.01 g) by specimen volume calculated from ASTM D638-14 Type I dimensions using digital caliper measurements of specimen length, width, and thickness (Figure 1B). Following demolding, all specimens were conditioned for at least 48 h in a digitally controlled dehumidifying cabinet maintained at 23 ± 2°C and 50 ± 5% relative humidity before mechanical testing.

Figure 1: Preparation and characterization of tensile test specimens. (A) Template for mold fabrication. (B) Standard specimen dimensions according to ASTM D638-14 (mm). (C) Silicone mold. (D) Dumbbell-shaped tensile specimens prepared using the custom mold. From left to right: R+1%RF, E+1%RF, R+3%RF, E+3%RF, R+5%RF, E+5%RF, R, and E. R = polyester resin; E = epoxy resin; RF = fiber reinforcement weight fraction. Scale is shown in (E). (E) Top-view photograph of tensile specimens (scale bar = 1 cm). (F) Specimens stored in a digitally controlled dehumidifying cabinet.

Mechanical testing

All specimens were subjected to uniaxial tensile testing at 25°C using a universal testing machine (Instron 5982; Instron Inc., Norwood, MA, USA) equipped with a 100-kN load cell and self-tightening wedge grips. Dumbbell-shaped specimens prepared according to ASTM D638 Type I with a 50-mm gauge length (Figures 1D and E) were tested at a constant crosshead speed of 5.0 mm/min. Force and crosshead displacement were continuously recorded at a sampling frequency of 10 Hz until specimen failure (Figures 2A and B).

Engineering strain (ε) was calculated from crosshead displacement as:

\documentclass{article} \usepackage{amsmath} \usepackage{amssymb} \begin{document} \[ \varepsilon =\frac{\Delta L}{{L}_{0}} \] \end{document}

where ΔL is the change in gauge length after testing and L₀ is the original gauge length.

Because strain measurements were derived from crosshead displacement rather than a contact or non-contact extensometer, the calculated strain values may slightly overestimate absolute strain because of machine compliance and grip slippage. Nevertheless, all specimens were tested under identical experimental conditions, allowing reliable relative comparisons among groups. Therefore, strain-related parameters (εy and εu) should be interpreted comparatively rather than as absolute material constants.

Engineering tensile stress (σ) was calculated as:

\documentclass{article} \usepackage{amsmath} \usepackage{amssymb} \begin{document} \[ \sigma =\frac{P}{{A}_{0}} \] \end{document}

where P is the applied load and A₀ is the original cross-sectional area, determined from the mean of three digital micrometer measurements obtained at the center of the gauge section.

Stress–strain curves were generated for all specimens. The elastic modulus (Emod) was calculated from linear regression of the maximum slope within the linear elastic region when R² exceeded 0.90. Yield strength (σy) and yield strain (εy) were determined using the 0.2% offset method [26, 27]. Ultimate stress (σu) and ultimate strain (εu) were recorded for each specimen (Figure 2C).

Statistical analysis

A minimum of five specimens per group was selected according to ASTM D638 recommendations for material characterization studies [26] and was also limited by the fabrication capacity of each molding batch. The unreinforced E group contained six specimens because one additional specimen met all predefined quality criteria and was therefore included in the analysis.

Mean ± standard deviation were calculated for Emod, σy, εy, σu, and εu. Data normality was evaluated using the Shapiro–Wilk test, whereas homogeneity of variance was assessed using Levene's test. Differences among groups were analyzed using one-way analysis of variance (ANOVA), followed by Tukey's honestly significant difference (HSD) test for multiple comparisons when appropriate. Cohen's d was calculated to estimate effect sizes for statistically significant pairwise comparisons.

Mechanical properties between the E+3%RF and PU groups were additionally compared using Welch's independent-samples t-test. All statistical tests were two-sided, and statistical significance was established at p < 0.05. Statistical analyses were performed using R statistical software version 4.5.3 [28].

Figure 2: Tensile testing procedure. (A) Specimen mounted before testing. (B) Specimen after tensile failure. (C) Representative stress-strain curve showing the linear regression (dashed line) fitted to the initial elastic region. Emod = elastic modulus; σy = yield strength; εy = yield strain; σu = ultimate stress; εu = ultimate strain.