A platform for research: civil engineering, architecture and urbanism
A platform for research: civil engineering, architecture and urbanism
Abutment scour depth modeling using neuro-fuzzy-embedded techniques
Bridge pier scour hole is formed during major flood events due to a complex vortex flow around the abutments and causes the failure of foundation. Therefore, accurate prediction of the scour hole depth is a major factor in the hydraulic structure design. This paper offers an adaptive neuro-fuzzy-embedded subtractive clustering (ANFIS–SC) method for estimating abutment scour hole depth under clear water condition with uniform bed sediments. The accuracy of ANFIS–SC method is compared with two other ANFIS methods embedded with fuzzy C-mean clustering and grid partitioning (ANFIS–FCM and ANFIS–GP). The decisive factors on the abutment scour hole depth include the ratio of the average diameter of particle size to abutment transverse length (d50/l), excess Froude number of the abutment (Fe), shape factor (Ks), and the ratio of approach stream depth to abutment transverse length (h/l). Eleven different input combinations were evaluated with four, three, and two input variables to survey the effect of each dimensionless parameter on the abutment scour depth. The results indicated that the use of all effective input variables (Fe, d50/l, h/l, Ks) results in the best prediction of abutment scour depth. The comparison of ANFIS models showed the superiority of ANFIS–SC over the other methods.
Abutment scour depth modeling using neuro-fuzzy-embedded techniques
Bridge pier scour hole is formed during major flood events due to a complex vortex flow around the abutments and causes the failure of foundation. Therefore, accurate prediction of the scour hole depth is a major factor in the hydraulic structure design. This paper offers an adaptive neuro-fuzzy-embedded subtractive clustering (ANFIS–SC) method for estimating abutment scour hole depth under clear water condition with uniform bed sediments. The accuracy of ANFIS–SC method is compared with two other ANFIS methods embedded with fuzzy C-mean clustering and grid partitioning (ANFIS–FCM and ANFIS–GP). The decisive factors on the abutment scour hole depth include the ratio of the average diameter of particle size to abutment transverse length (d50/l), excess Froude number of the abutment (Fe), shape factor (Ks), and the ratio of approach stream depth to abutment transverse length (h/l). Eleven different input combinations were evaluated with four, three, and two input variables to survey the effect of each dimensionless parameter on the abutment scour depth. The results indicated that the use of all effective input variables (Fe, d50/l, h/l, Ks) results in the best prediction of abutment scour depth. The comparison of ANFIS models showed the superiority of ANFIS–SC over the other methods.
Abutment scour depth modeling using neuro-fuzzy-embedded techniques
Moradi, Fatemeh (author) / Bonakdari, Hossein (author) / Kisi, Ozgur (author) / Ebtehaj, Isa (author) / Shiri, Jalal (author) / Gharabaghi, Bahram (author)
Marine Georesources & Geotechnology ; 37 ; 190-200
2019-02-07
11 pages
Article (Journal)
Electronic Resource
English
Neuro-genetic Approach to Predict Scour Depth Around Vertical Bridge Abutment
| Springer Verlag | 2014
Maximum Abutment Scour Depth in Cohesive Soils
| British Library Conference Proceedings | 2011
Maximum Abutment Scour Depth in Cohesive Soils
| ASCE | 2010