09-10-2012, 04:59 PM
DAM ANALYSIS
History
Original work was sponsored by the Bureau of Reclamation (Howard Boggs) to perform a fracture mechanics (one of my specialty) based investigation of Upper Stillwater Roller Compacted Concrete dam.
This subsequently led to a major contract from the Electric Power Research Institute, EPRI for the preliminary (Static) development of a nonlinear fracture mechanics based 3D finite element analysis program for dam investigation, and numerous laboratory tests for quantification of fracture properties, field tests (for in-situ measurements of some physical properties), and centrifuge tests for preliminary verifications of our numerical code. Doug Morris was our Project Director, and Howard Boggs was our Project Monitor (weekly meeting for 5 years!).
This effort was subsequently carried on through the support of the Tokyo Electric Power Company (which is the second largest utility company in the world), through its engineering division the Tokyo Electric Power Service Company, Mr. Uchita (Dam Specialist at TEPCO) and Noguchi have been the driving forces in this project. Through this support, the computer program Merlin underwent major extensions for the dynamic nonlinear analysis of dams subjected to strong motion. There is absolutely no doubt that at present, Merlin is one of the most sophisticated specialized code for the dynamic analysis of concrete dams. Centrifuge tests (static with effect of Uplift) and dynamic ones (dam model mounted on a shaking table inside a centrifuge) have also been carried out in Boulder and Tokyo respectively. These tests further validated our numerical model.
Finally, and most recently we completed another major project for the Swiss Dam Safety Agency on the development of a constitutive model for Alkali Aggregate Reactions, its implementation in our (Merlin) finite element code, and its application to investigate the long term behavior of an existing dam.
Alkali-Aggregate Reactions in Dams There is worldwide concern about AAR in dams. Rather than limiting ourselves to a phenomenological/empirical approach to dam simulation, we have developed State of the Art models based on extensive research performed at the LCPC (Laboratoire Central des Pont et Chaussées) in Paris.
1. This is an overview paper which summarizes our model and its application. It was presented at the ASDSO conference in 2005.
2. A State of the Art Review was commissioned by our sponsor (see above). Though it is a very broad subject, we have attempted to focus on the most relevant (for our work) literature.
3. Details of the model we developed is shown in this power point presentation.
Dynamic Analysis of Dams
1. Dynamic fracture properties of concrete are reported here (EPRI Project).
2. During earthquake it is important to determine the dynamic uplift inside a crack. This paper summarizes a series of laboratory tests, and numerical model which we developed.
3. To validate the dynamic predictive capabilities of our program, a dam model was mounted on a shaking table, and itself was mounted inside a centrifuge. This a unique (and most complex) test.
Fracture of Concrete
1. Understanding of the so-called Fracture Process Zone which surrounds a crack is essential. In this test, we have use Bragg Grating optical fibers, cast inside the concrete to measure the strain around the crack and improve our understanding of concrete fracture.
2. Modeling of concrete/rock interface is most important. One approach is to consider it as a crack between two dissimilar materials (in the classical linear elastic fracture mechanics sense).
3. Major results of tests undertaken through the EPRI project on the fracture of concrete can be found here.
4. Various pictures from the EPRI project.
Modeling Joints The primary, if not only, source of nonlinearity in dam response (whether static or dynamic) is the formation and propagation of cracks/joints. Hence, it is absolutely essential that we fully understand cracking of concrete, and the opening of existing joints.
1. We performed what may be the largest direct tension test of concrete under displacement control (to capture the post-peak softening).
2. Description of tests on concrete joints subjected to compressive confinement and shear displacement (to model rock/concrete interfaces and concrete cracks).
3. Description of our first numerical model for interface cracks, extensively used in Merlin.
4. This is another paper describing essentially the same original joint model described above.
5. Paper under review for our latest joint/crack model when reverse cyclic load is applied on a confined joint.
Centrifuge Testing of Dams Computer models must be validated by laboratory model testing. However for dams, since gravity plays such an important role, we must use a centrifuge where the dam model is subjected to a gravity acceleration inversely proportional to its size scale.
1. ASCE paper describing first tests on centrifuge modeling of concrete gravity dams.
2. Inter. Journal of Water Power and Dam Construction describing centrifuge tests in which uplift was modeled, and test results were properly predicted by our finite element analysis.
3. To validate the dynamic predictive capabilities of our program, a dam model was mounted on a shaking table, and itself was mounted inside a centrifuge. This a unique (and most complex) test.
Selected Case Studies
1. ETL 1110-2-344 This is an Engineering Technical Letter from the US Army Corps of Engineers which analyses Lock & Dam 27 on the basis of ETL 1110-8-16(FR) whcich stipulates that A Fracture mechanics based investigation should be conducted prior to major rehabilitation.
2. ICOLD/Slazburg; Schlegeis Dam The detailed 3D analysis of an arch dam using both a statistical approach as well as a 3D nonlinear finite element deterministic one.
3. ICOLD/Bucharest; Scalere Dam An extensive and detailed (99 pages) nonlinear analysis of an arch-gravity dam.
4. Palagnedra Dam Analysis of a dam trying to explain the formation of a crack through a combined thermal and nonlinear stress analysis.
5. Poglia Dam: Preliminary analysis of a dam subjected to AAR. Emphasis is on the nonlinear continuum model subjected to simple thermal expansion to simulate AAR(more refined analysis is to be found in the next publication). In Italian.
6. ICOLD/Wuhan; Poglia Dam
Computer codes (short description of Merlin)
Reinforced Concrete
1. EdF (Electricite de France) organized a Benchmark workshop to assess various concrete models. Our contribution is shown here.
2. Merlin was also used to perform the nonlinear analysis of a nuclear reactor containment vessel (only a ring was modeled). In this study, leaking gas generates an internal pressure in the crack analogous to the uplift pressure in a dam.
3. Full report of the EdF study, and the analysis of the so-called "San-Diego" column (analyzed by many researchers) can be found here.
Dam Engineering Literature Over the years I have accumulated many pdf files related to Dam Engineering from around the world. In the near future, I will try to share those hundreds of pdf file with the dam engineering community.