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

Table 2

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

Column type	Measure (N)	FEM (N)	Absolute error (%)
M45-1.5	48735.72	49856.30	2.31
M60-1.8	79771.43	80079.38	0.39
M60-2	107829.81	116348.70	7.90
M75-1.8	120201.30	120906.59	0.58
M75-2	126332.15	133166.10	5.41
M90A-1.8	131635.70	132295.27	0.51
M90B-2	145770.64	156700.50	7.50
M100A-2	184553.71	189905.90	2.91
M100B-2.5	219416.60	220570.64	0.54
M100C-3	341267.70	343081.97	0.53
M120A-2.5	219843.50	238998.70	8.71
M120B-3.5	414820.20	417018.34	0.53
Mean relative error (%)	3.10
Correlation coefficient, R	0.97
Proportion of cases with over 5% error	4/12

In M45-1.5, M45 is the column type, and 1.5 is thickness of the tested stub column.