Minimizing corner cracking during the de-moulding process of industrial-size GFRP components: a case study

Abstract

This article, through an industrial-level case study, presents workflows employed for decision-making to mitigate cracking of glass fibre reinforced polymer (GFRP) parts in tight radii corner locations, often resulting from displacement-controlled de-moulding processes. Namely, using process simulation to evaluate the cure cycle of the GFRP composite parts, it was possible to optimize the time of de-moulding and reduce the potential for part damage. It was observed that the most significant factors influencing the corner defect were boundary conditions of the part during de-moulding, the workshop temperature and part thickness. The poorest process design case was identified as hot workshop temperature, a laminate with thickness on the upper end of tolerances and a boundary condition where most sides are free, allowing for the development of larger moment forces at the tight corners. Further to this, a de-moulding time chart was developed to account for the changes in material properties as a function of temperature and material thickness, allowing for the in situ decision-making of technicians to reduce the occurrence of corner cracks.

Appendix: a closer look to the effect of resin curing

The development of the resin modulus throughout the manufacturing process primarily influences the bulk mechanical properties of the composite laminate prior to de-moulding. Specifically, once the initiator compound is added to the resin mix, the generation of free radical particles begins and chemical bonding between monomers into long polymer chains begins, as depicted in Fig. 17.

Fig. 17

Illustration of how resin modulus develops due to lower molecular mobility as the cure cycle progresses

As the chains grow longer with more monomers added, the overall mobility of the molecules diminishes, and bulk rigidity becomes higher. This leads to an eventual transition from a liquid into a solid state, which becomes glassy as the available mobility effectively reduces to zero. Similarly, as this progresses, so does the measurable modulus of the material. Figure 18 shows both an example relationship between a cure temperature, degree of cure and glass transition temperature of an unsaturated polyester resin and the generic relationship between degree of cure and developed elastic modulus.

Fig. 18

(a) Example of a cure temperature (T_c), degree of cure and glass transition temperature (T_g) relationship for an unsaturated polyester resin [19], (b) the generic relationship assumed between degree of cure and elastic modulus [19]

It is important to note that the cure process is dependent on not only temperature but also time, as the chemical reaction has many factors that make the timescale great enough for de-moulding activities to take place during material property changes. The development of cure and hence material properties can be modelled in many ways, typically done empirically using a differential scanning calorimeter (DSC) and then curve-fitting the results to an assumed phenomenological model. There are many that have been developed for particular material systems, each with their own mechanics. For a chemical system with a heavy dependence on diffusion-based mechanisms, Cole et al. [20] developed Eq. (4) shown below, which has been effective in other published works [19].

$$ \frac{d\alpha}{d t}=\frac{K{\alpha}^m{\left(1-a\right)}^n}{1+{e}^{C\left\{\alpha -\left({\alpha}_{C0}+{\alpha}_{CT}T\right)\right\}}} $$

(4)

$$ K=A{e}^{\left(-E/ RT\right)} $$

(5)

where α is the degree of cure, K describes the Arrhenius function for the reaction, α_C0 is the critical degree of cure at T = 0 K, α_CT is a constant accounting for increase in critical resin degree of cure with temperature and m and n are constants obtained from a curve-fitting analysis.