PSI - Issue 64

Nima Kian et al. / Procedia Structural Integrity 64 (2024) 1049–1056 Author name / Structural Integrity Procedia 00 (2019) 000 – 000

1054

6

the Saatcioglu and Razvi (1992) model is implemented. In the developed OpenSees model, the strain penetration originating from the slip rotations at the fixed end of the column-foundation joint is covered. Seven Gauss-Lobatto integration point along the column length are considered to take in to account distributed plasticity. Proposed buckling model of Dhakal and Maekawa (2002) is employed. The tension part of envelope curve consists of three parts including elastic range, yield plateau, and hardening region. The compression part of which consists of a linear elastic portion and nonlinear buckling part. To implement quasi-static analysis under constant axial load and reversed cyclic displacement excursions, the Newton-Raphson algorithm is implemented. Considering strain penetration effect in OpenSees, numerical results show good accuracy. However, the lateral load capacity is overestimated with respect to test around 20% for S specimen at same displacement. The reason for higher lateral load capacity estimation only in S specimen is neglection of shear deformations. Since shear deformations were negligible in other specimens (i.e. C and RS), the lateral load bearing capacity is estimated with a well accuracy. The hysteresis curves obtained by OpenSees models are represented in Figure 7.

(b)

(a)

Fig. 6. Constituent material model in OpenSees for a) concrete and b) steel rebars (Mazzoni et al. 2006).

(a)

(b)

(c)

Fig. 7. Lateral load-displacement hysteresis curves of experiment and OpenSees, a) Specimen S, b) Specimen C, c) Specimen RS.

3.3. Performance of the specimens Specimen S displayed lower ductility mainly from its poor confinement. The drift ratio capacity of S specimen was 66% less than C. After retrofitting, specimen R ’s drift ratio capacity increased by 124% with respect to specimen S. For the sake of comparison, load-displacement backbone envelope curves for all specimens are shown in Figure 8a. The discrepancy between two numerical models could be attributed to the contribution of the slip rotations occurring at top displacements. In this regard, specimens modeled through ATENA have a higher (min. 5 %, max. 30%) lateral capacity at the same displacement level when compared to the specimens modeled using OpenSees. Moreover, cumulative energy dissipation (CED) comparisons of numerical and experimental results of all specimens are shown in Figure 8b. Results indicate that specimen RS dissipated energy up to 3.4 times larger than specimen S. As seen for all specimens, the agreement of OpenSees model is more than the ATENA model mainly from covering more parameters in the analysis. Performance levels of the specimens based on strain limits of concrete and rebars defined in TBEC (2018) are obtained using the data acquired from OpenSees. The performance levels associated with the damage in TBEC (2018) are i) limited damage (SH), ii) controlled damage (KH), and iii) collapse prevention (GO). Figure 9 shows the envelope load displacement curves of all specimens obtained by OpenSees in pushing direction. As seen, the performance levels of specimen RS have been improved significantly when compared with specimen S. Table 5 shows the collapse prevention (GO), controlled damage (KH), and limited damage (SH) service limits for the specimens. In the following table, , L , and represent core concrete strain, lateral load, and lateral displacement of the specimens.