PSI - Issue 37

6

Abdalla et al./ Structural Integrity Procedia 00 (2019) 000 – 000

Jamal A. Abdalla et al. / Procedia Structural Integrity 37 (2022) 660–667

665

1.5

P R = P u /P o = -1.619E R +1.3450 R 2 = 0.901

1

P R = P u /P o = -1.4747E R +1.2661 R 2 = 0.947

Capacity Ratio (P R )

P R = P u /P o = -0.823E R +0.9152 R 2 = 0.986

P R = P u /P o = -1.1026.6E R +0.8873 R 2 = 0.890

0.5

0S0W 4S2W 4S4W 8S2W

0

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

Eccentrity Ratio (E R )

Fig. 3. Effect of eccentricity ratio in uniaxial loading case.

Table 4. Empirical linear models of capacity ratio ( P R ) and eccentricity ratio ( E R ) Specimen Linear Empirical Model

R 2

MAPE

RMSE

NMSE

0S0W (no Strip, no Wrap) 4S2W (4 Strips, 2 Wraps) 4S4W (4 Strips, 4 Wraps) 4S8W (4 Strips, 8 Wraps)

P P

o u

1.1026.6

0.8873

PR

E R

= = −

+

0.8902

34.34%

0.1078

0.1081

P P

o u

1.4747

1.2661

PR

ER

= = −

+

0.9472

20.260%

0.0979

0.0533

P P P P

o u o u

1.619

1.3450

PR

ER

= = −

+

0.9012

30.74%

0.1477

0.0962

0.823

0.9152

PR

ER

= = −

+

0.9863

4.12%

0.0604

0.0179

Table 5 shows a comparison between the normalized values of experimental results and empirical predictions. The values are in good agreement with average MAPE of 6.6%. 5. Discussion of Results and modes of failure Nonlinear prediction models developed for the concentrically loaded specimens (e x /b x = 0) shows close agreement with the experimental results with an average MAPE of 3.9%. On the other hand, linear models had an average MAPE of 9.4% for the same group. Such results implied the adequacy of both models in predicting axial load capacity for columns subjected to concentric loading with a favor to the aforementioned model.