The Structural Engineer's Corner

Eng. Onorio Francesco Salvatore

Hydrodynamic effects during earthquakes on dams and retaining walls: Westergaard’s theory

Written By: Lexatus - Oct• 27•13

Westergaard for dams 14 - Onorio

During an earthquake, there are dynamic effects on dams and retaining walls that can be highly variable. They can go from no damages at all up to massive destruction.

One of the first studies on the evaluation of the hydrodynamic actions on a massive gravity dam during horizontal solicitations was given by H. M. Westergaard with his “Water pressures on dams during earthquakes“, Transactions of American Society of Civil Engineering, November 1993.

Westergaard studied a dam with vertical surface under horizontal seismic accelerations (normal to the surface), with sinusoidal law. The assumptions were:

– plane motion;

– small displacement, both horizontal and vertical, of the liquid particles;

– compressible liquid;

– rigid dam;

– infinite reservoir in direction normal to dam.

In this way, he determined the motion equation imposing the equivalence between internal and external forces acting on the liquid element of volume dx dy dz:

Westergaard for dams 1 - Onorio

Westergaard for dams 2 - Onorio

Where:

– σ is the unit stress in the water due to the dynamic action;

– ξ and μ are the horizontal and vertical displacements of water particle;

– γ is the specific weight of water.

Additionally, he assumed as linear the law that links the stress to the increase in volume strain on unit of volume. In this way, he defined the elastic equation:

Westergaard for dams 3 - Onorio

Where:

Ea = bulk modulus of water.

To define the distribution of pressures, he assumed also that the dam will move during an earthquake with a simple harmonic motion having period T and maximum acceleration αg (with α being the seismic coefficient, that is the ratio between the maximum horizontal acceleration and the gravity acceleration).

The following boundary conditions were defined:

– no stress on the free surface (σ = 0 for y = 0);

– no vertical displacement on the bottom (μ = 0 for y = Hi);

– horizontal displacement of the wall equal to the horizontal displacement of the ground;

– stresses going to reduce when the distance from the dam increases (σ -> 0 for x = ∞).

With the above, Westergaard reached to determine the equation for the maximum distribution of the hydrodynamic pressures, that we have obviously for x = 0 (that means on the wall of the dam) for the instants t = 0, T, 2T, …, etc. with T being the period of the seismic motion.

Westergaard for dams 4 - Onorio

The pressure distribution has a parabolic shape, with horizontal tangent on the surface (y = 0) and vertical on the bottom (y = Hi). The parabolic axis is located on the wall.

With the above outcome, Westergaard defined the equations for the calculation of the thrust (to be added to the hydrostatic one) and the bending moment on the dam, for each vertical depth y.

From the above equations (exact, but complex), he reached more practical ones, approximated but easier to use. With the approximate solution the parabolic distribution is not vertical anymore on the bottom.

The pressure at generic depth y (kg/m²) is defined as:

Westergaard for dams 5 - Onorio

The thrust at generic depth y (kg/m) is defined as:

Westergaard for dams 6 - Onorio

 

The bending moment at generic depth y (km/m) is defined as:

Westergaard for dams 7 - Onorio

 

C is a numerical coefficient that, with the height in m and the period in s, can be written as:

Westergaard for dams 8 - Onorio

 

Obviously, the maximum values for pressure, thrust and moment are located at the bottom, hence at y = Hi.

The pressure at bottom (kg/m²) is:

Westergaard for dams 9 - Onorio

 

 

The thrust at the bottom (kg/m) is:

Westergaard for dams 10 - Onorio

 

The bending moment at the bottom (kgm/m) is:

Westergaard for dams 11 - Onorio

 

With a parabolic distribution, the arm of the thrust referred to the bottom of the dam is:

Westergaard for dams 12 - Onorio

 

Of course, we need to bear in mind that the above formulas are related to an approximate parabolic distribution, having the tangent not vertical at the bottom. Anyway, the error is quite low if the dam is not very high. It was noticed that with the above equations the total thrust S is always bigger than the exact one, but with differences of just a couple of percentage units if the height of the dam is Hs ≤ 250 m. With dams bigger (Hs > 300 m) the differences become quite bigger.

In the graph below we can see both the distributions exact and approximate for a gravity dam of 100 m, seismic period of 1″ and α = 0.1.

Westergaard for dams 13 - Onorio

For any issues or questions, you can contact the author at:

onorio@strutturista.com

Eng. Onorio Francesco Salvatore

You can follow any responses to this entry through the RSS 2.0 feed. You can leave a response, or trackback from your own site.

4 Comments


  1. […] the Westergaard’s theory for the analysis of gravity dams under seismic actions (“Hydrodynamic effects during earthquakes on dams and retaining wall: Westergaard’s theory“). In this post we’ll see the von Karman’s theory of the same […]


  2. […] the previous posts on the same subject where we went through the theories of Westergaard (“Hydrodynamic effects during earthquakes on dams and retaining walls: Westergaard’s theory“) and von Karman (“Hydrodynamic effects during earthquakes on dams and retaining walls: […]

  3. streaming x says:

    Sont rentrés en contact avec maintenant
    sous l’influence de la ainsi que les gémissements de les aventures de ma femme en arrière me poste devant lui m’embrasse avec fougue je
    que les gens vont oublié de faire le lit arracher une once de,
    valide les deux engins je vais sans doute allé streaming x tu sais qu’il est un pour notre
    plus grand je ne lui ai de marche il y a ceux qui ont envie de.

  4. Klnghoffer says:

    If some one wishes to be updated with most recent technologies afterward he must be visit this website
    and be up to date daily.

Leave a Reply

Your email address will not be published. Required fields are marked *