PSI - Issue 22

Elise Zgheib et al. / Procedia Structural Integrity 22 (2019) 25–32 Elise Zgheib – Wassim Raphael / Structural Integrity Procedia 00 (2019) 000 – 000

31

7

Table 2 shows the parameters of the selected tests while Table 3 shows the prior and posterior distribution parameters respectively in the case of concrete with silica fume (SF).

Table 2. Parameters and constant results for concrete with silica fume. n t 0 (days) h 0 (mm) RH (%)

t-t 0 (days)

Cst

12

7

40

70

3000

21.70847

12 tests were selected from the database to quantify the correction coefficient E implemented to Eurocode 2 model in the case of concrete with silica fume. These specimens with a common notional size of 40 mm are loaded at the age of 7 days. They are maintained in an environment where relative humidity is equal to 70%. Also, this study targets to update the long-term creep which is considered to be equal to 3000 days in this case. Table 3. Parameters of the prior and posterior distribution and the results of the correction coefficient in the case of concrete with silica fume (SF). Prior parameters Posterior parameters Correction coefficient μ θ σ θ 2 μ θ p σ θ p 2 E 1.171239 0.018686 2.05257 0.008809 2.4 To evaluate the correction coefficient efficiency, the updated creep compliance is calculated for all tests in database and the results are compared to experimental measurements using statistical method as shown in the below table. Table 4. The CEB mean deviation (M CEB ) and the CEB mean square error (F CEB ) results for concrete creep predictions before and after correction in the case of concrete with silica fume (SF). Concrete with silica fume M CEB (expected value 1) F CEB (expected value 0) Before correction 1.9 108 After correction (E = 2.4) 1.04 39.75 The correction coefficient efficiency evaluation shows that M CEB and F CEB have decreased towards the expected values by adding correction coefficient and considering the percentage of admixtures into the Eurocode 2 creep compliance expression for concrete with SF (Table 4). It can be noticed that after applying the correction coefficient E = 2.4, the mean deviation M CEB and the mean square error F CEB have decreased (M CEB = 1.04, F CEB = 39.75). In the case of concrete with SF, the creep can be predicted more accurately by adding the above correction coefficient to the creep compliance formula and by considering the percentage of silica fume. 5. Conclusion The addition of admixtures to the mix composition of concrete affects its behaviour and properties especially creep deformations which is the aim of this study. But design codes and specifically Eurocode 2 model do not consider the effect of admixtures while predicting concrete creep. Therefore, correction coefficients are implemented to Eurocode 2 formula to consider the admixtures’ effect, more specifically, the effect of silica fume. These correction coefficients were implemented according to different expressions but the one considering the percentage of admixtures led to the most accurate results. In this paper, the Bayesian Linear Regression method was applied to identify these correction coefficients added to the Eurocode 2 compliance formula in the case of concrete with silica fume. It can be noticed that the implementation of correction coefficients to Eurocode 2 formula allows to predict accurately creep and to consider the effect of admixtures. These predictions are more accurate by considering the percentage of silica fume into the correction coefficient expression. This study shows that the Bayesian model assessment is an important procedure applied to update the Eurocode 2 creep model. The long-term serviceability of structures subject to creep is well improved by adopting such a design approach.