Issue 78

M. S ł owik et alii, Fracture and Structural Integrity, 78 (2026) 393-408; DOI: 10.3221/IGF-ESIS.78.23

Y Y   

T

1



i

ip



MPE

(8)



 

i T n    

T n

Y

i

in which: T – total number of calculation and forecast periods, N –number of forecast periods, Y i –the predicted value of the variable f ct15/15 in period i, Y ip – the predicted value of the variable f ct15/30 in period i.

fct,sp (gravel, C/W = 2.6, Dmax=16mm) = 2.7935+0.1273*x-0.0043*x^2 fct,sp (granite, C/W = 2.6, Dmax=16mm) = 2.9378+0.1546*x-0.0047*x^2

fct,sp (gravel, C/W = 1.8, Dmax=16mm) = 2.3034+0.0361*x+0.001*x^2 fct,sp (granite, C/W = 1.8, Dmax=16mm) = 2.9412+0.0688*x-0.0018*x^2

3.8

4.4

3.6

4.2

3.4

4.0

3.8

3.2

3.6

3.0

3.4

2.8

fct,sp

fct,sp

3.2

2.6

3.0

2.4

fct,sp (gravel, C/W = 2.6, Dmax=16mm) fct,sp (granite, C/W = 2.6, Dmax=16mm)

fct,sp (gravel, C/W = 1.8, Dmax=16mm) fct,sp (granite, C/W = 1.8, Dmax=16mm)

2.8

2.2

0 1 2 3 4 5 6 7 8 9 1011121314151617 2.6

0 1 2 3 4 5 6 7 8 9 1011121314151617 2.0

Figure 7: Line plots of variables f ct,sp : for weaker concretes w/c = 0.55 with granite and gravel aggregate (a) and for stronger concretes w/c = 0.38 with granite and gravel aggregate (b). The MPE calculated between regression models for concretes with granite and gravel aggregate was +20.1% for weaker concretes (w/c = 0.55 - series L16N and L16G) and +8.7% for stronger concretes (w/c = 0.38 - series H16N and H16G). Therefore, it was concluded that the effect of the aggregate type on the splitting tensile strength of concrete was more evident in weaker concrete characterised by a lower strength of the cement matrix. To verify whether the observed differences between the investigated concrete series were statistically significant, a two-way analysis of variance (ANOVA) was performed. A significance level of α = 0.05 was adopted for all statistical analyses. Prior to the analysis, the assumptions required for the application of ANOVA were verified. The Shapiro–Wilk test indicated no significant deviation from normality within each experimental series, as the calculated test statistics exceeded the critical value of w crit = 0.887 (for n=16): for H16N w = 0.894, for L16N w = 0.950, for H16G w = 0.902, and for L16G w = 0.929. In addition, Levene's test confirmed the homogeneity of variances among the experimental series (p = 0.222). Therefore, the assumptions required for the application of a two-way ANOVA were satisfied. The analysis demonstrated that both the aggregate type (p < 0.001) and the cement matrix strength (p < 0.001) had a statistically significant effect on the splitting tensile strength of concrete. Furthermore, a statistically significant interaction between these two factors was identified (p = 0.018), indicating that the effect of the cement matrix strength on the splitting tensile strength varied depending on the aggregate type. For the purpose of the final comparative evaluation, the average splitting tensile strength for all specimen series was compiled and compared with the corresponding cement mortar strength and aggregate crushing values, as presented in Tab. 6. Additionally, Tab. 6 includes the increase in cement mortar strength along with the associated increase in the splitting tensile strength of concrete. An increase in cement mortar strength of 28.1% resulted in a comparable increase in splitting tensile strength of approximately 30% for concrete containing gravel aggregate. In contrast, for the same increase in cement matrix strength (28.1%), the splitting tensile strength of concrete with granite aggregate increased by only 13.4%. Considering the influence of aggregate crushing resistance, it should be emphasised that the granite’s crushing value was about twice that of the gravel aggregate. The comparatively lower strength of granite aggregate resulted in a smaller improvement in splitting tensile strength than that observed for concretes incorporating gravel aggregate. Overall, the results suggested that cement matrix strength governed the behaviour of gravel-based concretes. In contrast, in the case of

402