Scholarlink

<p>Earth fills are necessary in bridge construction, and they cause complicated hydraulic interactions which undermine the structural stability and modify river behavior. This study quantifies the hydraulic impacts of earth-fill construction near bridge foundations by using a hybrid modeling method which involves HEC-RAS and machine learning techniques as a case study at the Al-Sarafiya Bridge over the Tigris River in Baghdad, Iraq. Three discharge conditions (490 m<sup>3</sup>/s, 700 m<sup>3</sup>/s, and 3050 m<sup>3</sup>/s) were simulated at three bank conditions (no fill, left-bank fill, and right-bank fill) with the calibrated and validated one-dimensional hydraulic model (R<sup>2</sup> = 0.84, RMSE = 0.30 m). Findings show that, although earth fills have a negligible influence on water surface heights, which ensures there is sufficient capacity in bridge waterways, they generate vastly different distributions of flow energy, and thus, water flow speeds at the bridge location are augmented by up to 23.4% (right-bank fill) and 14.1% (left-bank fill). This increase in velocity is directly proportional to the increase in scour depth, with the greatest effect on right-bank fill: a scour depth of 5.91 m in flood conditions (3050 m<sup>3</sup>/s), an increase of 17% over no-fill conditions (4.91 m). The hydraulic threshold of about 875 m<sup>2</sup>/s indicates the change in scouring mode, as at this point the fill of the right bank takes up the domineering scouring character as a result of the curvilinear alignment of the thalwegs to the right bank. The flow velocity was determined as the most important scour predictor (important feature = 0.765) by machine learning analysis, with the near-perfect predictive power being established by the use of second-degree levels of regression (R<sup>2</sup> = 0.9998, RMSE = 0.0104 m). The power-law regression provided had succinct predictive equations (Scour = a·Q^b) and exponents (b = 0.25–0.27) that were consistent with the theoretical mechanics of scour and scaling coefficients (a = 0.542–0.741) that quantified the multiplicative amplification by bank encroachment. The results can offer engineers and river managers quantitative methods to evaluate the proposals of floodplain development in meandering river systems near bridge constructions.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Hydraulic impact of earth fill on bridge stability and river hydraulics: a hybrid modeling approach at Al-Sarafiya Bridge, Baghdad, Iraq

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2026年6月3日

查看全文
  • Layth Abdulameer,
  • Riyadh Jasim Mohammed Al-Saadi,
  • Khabeer Al-Awad,
  • Mahmoud Saleh Al-Khafaji

摘要

Earth fills are necessary in bridge construction, and they cause complicated hydraulic interactions which undermine the structural stability and modify river behavior. This study quantifies the hydraulic impacts of earth-fill construction near bridge foundations by using a hybrid modeling method which involves HEC-RAS and machine learning techniques as a case study at the Al-Sarafiya Bridge over the Tigris River in Baghdad, Iraq. Three discharge conditions (490 m3/s, 700 m3/s, and 3050 m3/s) were simulated at three bank conditions (no fill, left-bank fill, and right-bank fill) with the calibrated and validated one-dimensional hydraulic model (R2 = 0.84, RMSE = 0.30 m). Findings show that, although earth fills have a negligible influence on water surface heights, which ensures there is sufficient capacity in bridge waterways, they generate vastly different distributions of flow energy, and thus, water flow speeds at the bridge location are augmented by up to 23.4% (right-bank fill) and 14.1% (left-bank fill). This increase in velocity is directly proportional to the increase in scour depth, with the greatest effect on right-bank fill: a scour depth of 5.91 m in flood conditions (3050 m3/s), an increase of 17% over no-fill conditions (4.91 m). The hydraulic threshold of about 875 m2/s indicates the change in scouring mode, as at this point the fill of the right bank takes up the domineering scouring character as a result of the curvilinear alignment of the thalwegs to the right bank. The flow velocity was determined as the most important scour predictor (important feature = 0.765) by machine learning analysis, with the near-perfect predictive power being established by the use of second-degree levels of regression (R2 = 0.9998, RMSE = 0.0104 m). The power-law regression provided had succinct predictive equations (Scour = a·Q^b) and exponents (b = 0.25–0.27) that were consistent with the theoretical mechanics of scour and scaling coefficients (a = 0.542–0.741) that quantified the multiplicative amplification by bank encroachment. The results can offer engineers and river managers quantitative methods to evaluate the proposals of floodplain development in meandering river systems near bridge constructions.