Dams & Hydro
Geotech
Advanced Numerical Analysis
Calibrated numerical modelling of the stress-strain response of the 275 m high Yusufeli concrete arch dam during first filling
92nd ICOLD ANNUAL MEETING & INTERNATIONAL SYMPOSIUM 29 Sep- 03 Oct 2024
By Ryan Cassells, ARQ Dams Pty Ltd
-co-author Q.H.W. Shaw ARQ Consulting Engineers Limited, Mauritius
-F.V. Ҫevik Limak Inşaat A.Ş., Ankara, Turkey
D. Aydoğan General Directorate of State Hydraulic Works, Ankara, Turkey
With a significant increase in computing capacity over the past two decades, substantial development has ensued in Finite Element (FE) numerical, structural models, allowing designers and analysts a substantially improved ability to simulate the actual behaviour of concrete arch dams and their foundations.
Implementing the Final Design for the 275 m Yusufeli double-curvature concrete arch dam on the Çoruh River in north-eastern Turkey, it was soon discovered that the geotechnical conditions for the foundation rock mass were substantially more complex than had been assumed and a redesign of the dam and its foundation system to function appropriately was required before construction work on the foundation treatment and on the dam structure could be initiated. Detailed investigations subsequently allowed the development of what is likely to be the most complex geotechnical foundation model yet for an arch dam. Critically, a verification system was developed to allow the necessary rock mass improvement to be confirmed through in-situ measurement as works progressed, further enabling an accurate calibration of the complete rock mass model.
Subsequently, the accuracy of the FE model of both the foundation rock mass and the dam structure was comprehensively confirmed through the accuracy of the measured performance compared to that predicted through analysis during the 6-stage first impoundment process. In this paper, the authors describe the process of model development and calibration and demonstrate the elastic behaviour of the dam and foundation system, in complete alignment with predictions made in analysis modelling.
Return