Figure 3 shows the variation in the creep coefficients at 14, 30, and 90 days with respect to the load ratio applied during the creep tests. Higher load ratios implied higher creep coefficients at all ages. However, the effect of increasing the load ratio was less noticeable on the creep coefficient at 14 days: an increase of 10 % in IFa led to an average increase of the creep coefficient in 0.2, whilst this increase was of 0.35 at 90 days.

These average trends were not significantly affected by the fibres length or aspect ratio. However, interesting differences were observed when the interaction between fibre material and load ratio was analysed in addition to the simple effect of increasing the load ratio. Figure 4 shows the contrastive effects plots with respect to the load ratio but distinguishing between synthetic and non-synthetic fibres.

At lower load ratios, creep coefficients were observed to be systematically higher in those cases where synthetic fibres had been used. The gap between them, however, was gradually reduced for increasing load ratios, and differences due to the fibre material were negligible for load ratios of 60 % or higher. This pattern was consistent for all three ages considered.

However, the trends represented in Fig. 4 can be misleading: it is important to emphasise that Fig. 4 shows an association between the fibre material and higher creep coefficients, but this is not necessarily of a cause-effect nature. In fact, the analysis of variance on the squared residuals of these MLR models revealed that only one variable had a statistically significant effect on the variability of creep coefficients, and therefore on the differences represented in Fig. 4. And it was not the fibre material, but the flexural toughness T (p-values = 0.056, 0.020, and 0.016 at 14, 30, and 90 days respectively). In consequence, the differences between synthetic and non-synthetic fibres in Fig. 4 could not be directly attributed to the fibre material: the role played by the flexural toughness in these models and its interaction with the fibre material needed to be explored more closely, and that is precisely what is reported in the following section.

Fig. 4 Effects of fibre material and load ratio on creep coefficients