EXPERIMENTAL STUDY OF HYDRODYNAMIC PRESSURE INSIDE NARROW CAVITIES

Tatsuo Ohmachi, Hongyuan Zhang, Nobuyoshi Yabuki, N TsukadaDamu 1998 · DOI

Abstract

This study investigates the behavior of hydrodynamic pressure within narrow cracks of concrete gravity dams during seismic events. Using shaking table experiments on acrylic models with varying crack widths (1.5 mm to 10 mm), the research demonstrates that internal pressure increases linearly with depth and exhibits resonance at specific frequencies. These findings suggest that dynamic water-crack interactions can significantly amplify stresses, potentially accelerating structural failure beyond what is predicted by standard static models.

Introduction

When a massive concrete gravity dam experiences an earthquake, the brittle nature of the concrete often leads to cracking, particularly along horizontal joints or at the foundation interface. While engineers have long known that the reservoir water exerts pressure on the dam's face—a phenomenon famously described by Westergaard—the behavior of water inside a crack is significantly more complex.

If water enters a crack, the interaction between the vibrating dam walls and the fluid can create high hydrodynamic pressures. This pressure concentrates stress at the crack tip, potentially driving the crack deeper and compromising the dam's structural integrity. Historically, most insights were based on purely analytical models; experimental data measuring pressure directly inside these narrow cavities has been severely lacking.

Experimental Methodology

The authors designed a physical simulation using a shaking table to observe these "water-crack" interactions in a controlled environment.

Model Design

The researchers constructed three rectangular acrylic boxes, each featuring a 30-cm long horizontal notch representing a crack. To investigate how crack width affects pressure, three different opening widths ($w$) were tested:

  • Model I: $w = 10.0$ mm
  • Model II: $w = 5.0$ mm
  • Model III: $w = 1.5$ mm
Schematic and cross-section of the experimental model showing crack dimensions and water level.

Setup and Instrumentation

The models were placed inside a large water container (195 cm long) linked to a shaking table. To prevent unwanted wave reflections, foam-rubber cushions were installed at the container ends. Hydrodynamic pressure sensors were embedded at multiple points along the top and bottom surfaces of the artificial cracks to measure spatial distribution.

Layout of hydrodynamic pressure sensors (s1 to s6) along the upper and lower surfaces of the artificial cavity.

The table was subjected to sinusoidal motions ranging from 1.0 Hz to 10.0 Hz with varying acceleration amplitudes (up to 700 gal). This generated primary experimental data across the three physical models, frequency sweeps, and various acceleration levels.

Analysis of Hydrodynamic Behavior

The Depth Factor and Resonance

Experimental observations revealed that pressure is not uniform throughout the crack. It increases almost linearly from the "mouth" of the crack toward the interior tip. This confirms a basic dynamic equilibrium: the deeper the water column inside the crack, the more mass is being accelerated, leading to higher pressure.

A critical finding is the "resonance phenomenon." Hydrodynamic pressure inside a crack has a peak frequency. For example, Model III (the narrowest crack) showed a pressure peak at 5.0 Hz, while Model II peaked at 7.0 Hz. This suggests the cavity acts as a resonator; if seismic excitation matches these frequencies, internal pressure could spike far beyond static predictions.

Phase Lag and Width Effects

The water pressure does not always push in perfect sync with the dam's vibration. At higher frequencies, a significant phase difference exists between the table's acceleration and the resulting water pressure, complicating stress patterns on crack walls.

Furthermore, while wider cracks generally showed higher pressures, at high frequencies (10 Hz), the narrowest crack (1.5 mm) exhibited higher pressure than the wider ones. This suggests that at high-speed vibrations, fluid confinement in tiny spaces significantly alters the mechanics.

Results and Discussion

The researchers successfully validated the linear relationship between distance along the cavity and pressure increase, matching their derived 1D dynamic equilibrium equation. By normalizing pressure against input acceleration ($p/acc$), they demonstrated that while absolute pressure changes with intensity, the underlying distribution pattern remains consistent for a given frequency.

Experimental results at the mouth of the crack agreed well with the classical Westergaard solution. However, internal pressure gradients were found to be governed by the specific geometry and frequency of the crack-water system, which Westergaard's theory does not account for.

Limitations and Future Directions

While providing valuable insights, the study has several limitations:

  • Simplified Geometry: The experiment used smooth, rectangular acrylic notches, whereas real concrete cracks are jagged and irregular.
  • Scale Effects: Scaling laws for gravity and viscosity were not strictly applied, making the results a proof-of-concept for mechanics rather than direct predictions for full-scale dams.
  • Input Motion: Pure sinusoidal waves were used instead of stochastic, multi-directional seismic waves.

Future work should focus on integrating these findings into safety evaluation protocols for dams, moving beyond static water pressure models to include frequency-dependent dynamic interactions.

Summary updated

BibTeX
@inproceedings{ohmachi1998experimental,
  author = {Tatsuo Ohmachi and Hongyuan Zhang and Nobuyoshi Yabuki and N Tsukada},
  title = {{EXPERIMENTAL STUDY OF HYDRODYNAMIC PRESSURE INSIDE NARROW CAVITIES}},
  booktitle = {Damu},
  year = {1998},
  doi = {10.11315/jsde1991.8.35},
}
Figure