openturns / otbenchmark

A benchmark library for OpenTURNS
https://openturns.github.io/otbenchmark/master
GNU Lesser General Public License v3.0
1 stars 7 forks source link

The steel column problem is missing #82

Open mbaudin47 opened 3 years ago

mbaudin47 commented 3 years ago

https://github.com/sofianehaddad/UQ_benchmark_models/blob/master/steel_column.py https://www.sfu.ca/~ssurjano/steelcol.html

image

Description: Dimensions: 9

The Steel Column function models the trade-off between cost and reliability for a steel column.

The cost for the steel column is: Cost = bt + 5h, where b is the mean flange breadth, t (mm) is the mean flange thickness (mm), and h is the mean profile height (mm). The column length L is 7500 mm. Eldred et al. (2008) use the values b = 300, d = 20 and h = 300.

Fs ~ Lognormal(mean=400, standard deviation=35) yield stress (MPa)
P1 ~ N(μ=500000, σ=50000) dead weight load (N)
P2 ~ Gumbel(mean=600000, standard deviation=90000) variable load (N)
P3 ~ Gumbel(mean=600000, standard deviation=90000) variable load (N)
B ~ Lognormal(mean=b, standard deviation=3) flange breadth (mm)
D ~ Lognormal(mean=t, standard deviation=2) flange thickness (mm)
H ~ Lognormal(mean=h, standard deviation=5) profile height (mm)
F0 ~ N(μ=30, σ=10) initial deflection (mm)
E ~ Weibull(mean=210000, standard deviation=4200) Young's modulus (MPa)
var P := Pd + P1 + P2
var Eb = _pi^2 * E*B*D*H^2 / (2*L^2)
y := Fs - P*(1 / (2*B*D) + F0*Eb / (B*D*H*(Eb-P)) )