Data-Driven Decision-Making in the Design Optimization of Thin-Walled Steel Perforated Sections: A Case Study

Table 5

Comparison among the real measured values, predicted values, and finite element numerical values.

Column type	Measure (N)	Predict (N)	Relative error (%)	FEM (N)	Relative error (%)
M45-1.5	48385.62	48429.20	0.09	49856.37	3.04
M60-1.8	79136.73	83558.23	5.59	80079.43	1.19
M60-2	107142.80	104077.64	2.86	116348.72	8.59
M75-1.8	120057.01	115392.37	3.88	120201.36	0.12
M75-2	125481.44	126375.42	0.71	133166.13	6.12
M90A-1.8	131412.35	131070.89	0.26	131635.72	0.17
M90A-2	136364.28	140316.60	2.89	144199.20	5.75
M90B-1.8	144144.65	135171.32	6.22	145158.93	0.70
M90B-2	144746.99	146277.65	1.06	156700.55	8.26
M100A-2	183358.52	180049.49	1.80	189905.92	3.57
M100A-2.5	205056.90	207739.80	1.31	206744.95	0.82
M100B-2	159740.04	162860.88	1.95	167594.50	4.92
M100B-2.5	219082.27	218138.34	0.43	219416.63	0.15
M100C-2	196641.13	198992.22	1.19	207407.37	5.48
M100C-3	340630.69	342183.74	0.45	341267.72	0.19
M120A-2.5	219349.62	217817.09	0.69	238998.79	8.96
M120A-3	268689.38	271741.82	1.13	270660.52	0.73
M120B-3.5	413894.01	417135.66	0.78	414820.20	0.22
Mean relative error (%)		1.69		3.28
Correlation coefficient, R		0.99		0.97
Cases with over 5% error		2		6