Hindawi

Advances in Mathematical Physics

Volume 2019, Article ID 8083906, 14 pages
https://doi.org/10.1155/2019/8083906

Research Article
A Fluid-Structure Interaction Model for Dam-Water Systems:
Analytical Study and Application to Seismic Behavior

Ricardo Faria ,1 Sérgio Oliveira,2 and Ana L. Silvestre 3

1
Instituto Superior Técnico, Universidade de Lisboa, Av. Rovisco Pais 1, 1049-001 Lisboa, Portugal
2
Concrete Dams Department, Laboratório Nacional de Engenharia Civil, Av. Brasil 101, 1700-066 Lisboa, Portugal
3
CEMAT and Department of Mathematics, Instituto Superior Técnico, Universidade de Lisboa, Av. Rovisco Pais 1,
1049-001 Lisboa, Portugal

Correspondence should be addressed to Ana L. Silvestre; ana.silvestre@math.tecnico.ulisboa.pt

Received 22 August 2018; Accepted 5 November 2018; Published 2 January 2019

Academic Editor: Phuc Phung-Van

Copyright © 2019 Ricardo Faria et al. This is an open access article distributed under the Creative Commons Attribution License,
which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

We consider a dam-water system modeled as a fluid-structure interaction, specifically, a coupled hyperbolic second-order problem,
formulated in terms of the displacement of the structure and the fluid pressure. Firstly, we investigate the well posedness of the
corresponding variational formulation using Galerkin approximations, energy estimates, and mollification. Then, we apply the finite
element method along with the state-space representation of the discrete problem in order to perform a 3D numerical simulation
of Cabril arch dam (Zêzere river, Portugal). The numerical model is validated by comparison with available experimental data from
a monitoring vibration system installed in Cabril dam.

1. Introduction velocity variable is eliminated. In [4–6], the authors perform

The aim of this paper is the analysis and numerical simulation
of the dynamics of a dam-water system, in particular, its
response to earthquake actions.
Since the core engineering problem lies on the safety
and performance evaluation of the structure alone, many
related problems have been mostly addressed from a solid
mechanics perspective. A simple approach to incorporate
the water effects in the dynamic behavior of the dam is to
consider an extended Lagrangian model consisting merely of
displacements variables in both the reservoir and the dam
as in [1, 2] and references therein. However, to take into
account the sensitivity of the dam body and foundations to
water pressure, the techniques from solid mechanics alone
may not be the most adequate. Fluid mechanics plays an
important role as well, and appropriate equations describing
the water’s behavior and its interaction with the structure
should to be taken into account for a more realistic modeling
of the system. For this purpose, a pressure-elastic displace-
ment formulation is presented in [3], which is obtained by
simplifying the equations of motion of the fluid, so that the
a dynamic analysis of the two-dimensional problem. An
advantage of the formulation based on the fluid pressure, a
scalar unknown, is the reduction of the computational effort
in the numerical simulations. Furthermore, the use of the
pressure variable is advantageous due to the possibility of
comparison of values from the numerical simulations of the
mathematical model with real measurements of the water
pressure alongside the dam-water wall. It is common that, in
order to check if there exist cracks forming in the concrete
and for general safety control [7–10], such structures are kept
under constant monitoring, especially in seismic hazardous
regions.
To the best of our knowledge, the model we use in this
paper for the simulation of dam-water systems, and used
in similar engineering studies by other authors, e.g., [11–
13], has not been analysed from the mathematical point of
view. Therefore, our first aim is the well posedness of the
fluid-structure interaction problem. The transmission condi-
tion on the fluid-structure interface causes new difficulties
because the unbalanced order of the time derivatives of the
pressure and the structure’s displacement do not allow a
2 Advances in Mathematical Physics

direct application of available mathematical results for linear the domains Ω
푓 , Ω
푠 ⊂ R3 can be considered fixed, constant
hyperbolic systems. As a consequence, the regularity we will in time, and the Eulerian description can be adopted for both
obtain for the two unknowns is not the same. If we had the fluid and the structure motions. For the boundaries of Ω
푠
considered a fluid model formulated in terms of a velocity and Ω
푓 , we have the following: Γ1 is the interface between
potential, then the problem would be a hyperbolic coupling Ω
푠 and Ω
푓 ; Γ2 represents the ground or rock mass under the
similar to the problem treated in [14–17]. In these references, fluid; Γ3 is the air-fluid interface; Γ4 is an artificial ’wall’, a
the fluid domain was unbounded and, in order to find the delimitation of the fluid domain’s extension; Γ5 represents the
fluid velocity and the displacement field in an elastic body rock mass under the structure; and finally, Γ6 is constituted by
as a result of an incident acoustic wave, the authors used the the structure walls with contact with the air. The time interval
Laplace transform with respect to the time variable, which is domain during which the dynamic fluid-structure interaction
suitable for the application of discretization schemes based on occurs is [0, 𝑇], 𝑇 > 0.
the boundary element method and convolution quadrature The formulation of the fluid-structure interaction prob-
method (see also [18]). lem in elastic displacement-pressure variables is the following
In a second stage of our study, the discretization in space (see [3], pgs.634-637, and [11]): given the seismic activity → 󳨀𝑎
푠
and initial conditions → 󳨀 󳨀V , 𝑝 , and 𝑞 , find →
𝑢 0, → 󳨀
𝑢 =→ 󳨀𝑢 (𝑥, 𝑡)
by the finite element method of the fluid-structure interaction
0 0 0
model is followed by a state space representation and a
and 𝑝 = 𝑝(𝑥, 𝑡) such that
diagonalization process that lead to a system of first-order
linear differential equations. This approach is preferred over 𝑢 = ∇
푥 ⋅ 𝜎 (→
󰜚
푠 𝜕
푡
푡→
󳨀 󳨀 𝑢 + 󰜚
푠 (→
𝑢 ) − 𝑐
푎 𝜕
푡→
󳨀 󳨀
𝑔 +→
󳨀𝑎
푠 )
numerical time stepping schemes with less computational (1a)
cost, as suggested in [3], because a modal analysis allows for in Ω
푠 × ]0, 𝑇]
a direct comparison with natural frequencies experimental
data. In practice, such data may correlate eventual deteri- 𝜕
푡
푡 𝑝 = 𝑐2 Δ
푥 𝑝 in Ω
푓 × ]0, 𝑇] (1b)
oration processes with changes in modal parameters over
time [8]. Moreover, in the context of seismic responses, the 𝜎 (→
󳨀𝑢 )→
󳨀𝑛
푠 = 𝑝→
󳨀𝑛
푠 ,
modal analysis plays an important role, as it is crucial that

푓 → 󳨀 → 󳨀
푠
a structure’s natural frequency does not match a frequency of 㨀
푛 𝑓 𝑝 = −󰜚 𝜕
푡
푡 𝑢 ⋅ 𝑛
𝜕→ (1c)
an eventual earthquake; otherwise it may continue to resonate
and experience severe structural damage. on Γ1 × ]0, 𝑇]
The plan of the paper is the following. The mathematical
model is presented in Section 2. In Section 3, we derive 𝜕
푛𝑓 𝑝 = 0 on Γ2 × ]0, 𝑇] (1d)
the variational formulation of the continuous problem in
appropriate function spaces and investigate the a priori 𝜕
푡
푡 𝑝 = −𝑔𝜕
푛𝑓 𝑝 on Γ3 × ]0, 𝑇] (1e)
mathematical properties of solutions; for this, a pressure 𝜕
푡 𝑝 = −𝑐𝜕
푛𝑓 𝑝 on Γ4 × ]0, 𝑇] (1f)
potential is introduced so that the classical techniques based
on Galerkin approximations for second-order hyperbolic →
󳨀
𝑢 =0 on Γ5 × ]0, 𝑇] (1g)
equations can be applied to the modified coupled system.
The main results on the well posedness of the formulation 𝜎 (→
󳨀
𝑢 )→
󳨀𝑛
푠 = 0 on Γ6 × ]0, 𝑇] (1h)
in pressure-elastic displacement are stated in Theorem 3. We
proceed, in Section 4, with the numerical solution of the →
󳨀
𝑢 |
푡=0 = →
󳨀
𝑢 0,
three-dimensional problem: the discretization in space by the
finite element method leads to a state-space representation 𝜕
푡→
󳨀
𝑢 |
푡=0 = →
󳨀V
0 (1i)
of the problem (50), (47), and (48), followed by a suitable
in Ω
푠 × {0}
integration over time (58)-(59). The use of the modal super-
position technique allows us to study, in Section 5, the main 𝑝|
푡=0 = 𝑝0 ,
natural frequencies of the system and corresponding modal
configurations. The comparison of these results with available 𝜕
푡 𝑝|
푡=0 = 𝑞0 (1j)
experimental data from a monitoring vibration system of
Cabril arch dam (Zêzere river, Portugal) will validate the in Ω
푓 × {0}
numerical model. Finally, the seismic response of Cabril dam
will be presented by means of displacement fields. The gravitational acceleration → 󳨀𝑔 and the acceleration of the
→
󳨀
푠
ground 𝑎 due to earthquake vibrations are included in the
2. Analytical Modeling of the System Navier equation, where 󰜚
푠 is the structure density and 𝑐
푎 is a
positive constant representing a damping effect. The Cauchy
The region occupied by the structure (dam body and foun- stress tensor 𝜎
푠 (→
󳨀𝑢 ) is given by 𝜎
푠 (→
󳨀𝑢 ) fl 𝜆(∇ ⋅ →
󳨀 →
󳨀

푥 𝑢 )𝐼 + 2𝜇𝜀( 𝑢 ),
dation) is denoted by Ω
푠 , while the fluid domain (reser- →󳨀 →󳨀 →
󳨀 ⊤
where 𝜀( 𝑢 ) fl (∇
푥 𝑢 + (∇
푥 𝑢 ) )/2 and the parameters 𝜆 ≥ 0
voir) is denoted by Ω
푓 . Under the assumptions (i) small and 𝜇 > 0 are the Lamé constants. They relate to other elastic
displacements and deformations in the structure, (ii) fluid moduli constants, namely the Young’s modulus 𝐸 and the
displacement remains small; while interaction is substantial, Poisson ratio ], by: 𝜇 = 𝐸/2(1 + ]), 𝜆 = ]𝐸/(1 + ])(1 − 2]).
Advances in Mathematical Physics 3

The structure domain boundary is 𝜕Ω
푠 = Γ1 ∪ Γ5 ∪ Γ6 and be defined by 𝑤
훿 (𝑡) = (J
훿 ⋆ 𝑤)(𝑡) = ∫R J
훿 (𝑡 − 𝜏)𝑤(𝜏)𝑑𝜏.
󳨀𝑛
푠 = →
→ 󳨀𝑛
푠 (𝑥) is the outward unit normal vector at 𝜕Ω
푠 . The Then
viscous effects in the fluid have been neglected and it was
󵄩󵄩
훿 󵄩
assumed that the fluid density varies very little, so that it may 󵄩󵄩𝑤 − 𝑤󵄩󵄩󵄩 𝑟
be considered constant, 󰜚
푓 . The constant 𝑐 is the speed of 󵄩 󵄩
퐿 (
푎,
푏;
푋) 󳨀→ 0 as 𝛿 󳨀→ 0. (3)

sound in the fluid and 𝑔 fl |→ 󳨀

𝑔 | is the acceleration due to
gravity. The fluid domain boundary is 𝜕Ω
푓 fl Γ1 ∪ Γ2 ∪ Γ3 ∪ Γ4 . Since convolution commutes with differentiation, we have
In the air-water interface Γ3 , a wavelike motion is observed, 𝜕
푡 𝑤
훿 = (𝜕
푡 𝑤)
훿 and
and Γ4 is permeable in the sense that the solution of (1b) in
Ω
푓 should be composed only of outgoing waves on Γ4 , as no 󵄩󵄩
훿 󵄩
󵄩󵄩𝜕
푡 𝑤 − 𝜕
푡 𝑤󵄩󵄩󵄩 𝑠
input from this boundary portion exists [3]. 󵄩 󵄩
퐿 (
푎,
푏;
푌) 󳨀→ 0 as 𝛿 󳨀→ 0, (4)

3. Variational Formulation. Existence and
Uniqueness of Weak Solutions
3.1. Notation and Auxiliary Results. Let Ω ⊂ R3 be a bounded
domain and Γ ⊆ 𝜕Ω. We will use the classical notations
and results for Lebesgue spaces 𝐿
푞 (Ω), 𝐿
푞 (Γ), 1 ≤ 𝑞 ≤ ∞,
Sobolev spaces 𝐻
푚 (Ω) := 𝑊
푚,2 (Ω), 𝑚 ∈ N, and their dual
spaces 𝐻−
푚 (Ω) fl (𝐻0
푚 (Ω))耠 . By ‖ ⋅ ‖
푞,Ω and ‖ ⋅ ‖
푞,Γ we denote
the norms of 𝐿
푞 (Ω) and 𝐿
푞 (Γ), respectively. We will use the
notation (𝑓, 𝑔)Ω fl ∫Ω 𝑓 ⋅ 𝑔𝑑𝑥 for vector valued functions,
and when 𝑓 and 𝑔 are tensor-valued functions, we will write
(𝑓, 𝑔)Ω fl ∫Ω 𝑓 : 𝑔𝑑𝑥, where 𝑓 : 𝑔 represents contraction of
tensors. Analogous meaning holds for (𝑓, 𝑔)Γ .
Suppose that Γ
퐷 ⊂ 𝜕Ω has nonzero 2-dimensional
measure and let 𝑉0 fl {→ 󳨀V ∈ 𝐻1 (Ω)3 : → 󳨀V = 0 on Γ }.

퐷
Recall that the Korn inequality is valid in 𝑉0 : ‖→ 󳨀V ‖ 1 3 ≤

퐻 (Ω)
if 𝜕
푡 𝑤 ∈ 𝐿
푠 (𝑎, 𝑏; 𝑌), where 1 ≤ 𝑠 < ∞ and 𝑌 is a Banach space.
Another property of mollification is supp 𝑤
훿 ⊂ supp J
훿 +
supp 𝑤.
We will also use the following classical embedding result
(see [19], pg. 392, Lemma 11.9).

푑
Lemma 1. Let 𝑋, 𝑌 be Hilbert spaces such that 𝑋 󳨅→ 𝑌. Then
𝐿∞ (𝑎, 𝑏; 𝑋) ∩ 𝐶([𝑎, 𝑏]; 𝑌) ⊂ 𝐶
푤 ([𝑎, 𝑏]; 𝑋).
3.2. Variational Formulation. In order to derive the vari-
ational formulation of Problems (1a)-(1j), we consider the
function spaces
𝐻
푓 fl {𝑞 ∈ 𝐿2 (Ω
푓 ∪ Γ3 ) : ∫ 𝑞 𝑑𝑥 = 0} ,
Ω𝑓

𝐶
퐾 (Ω)‖𝜀(→󳨀V )‖ . Therefore 𝑉 is a Hilbert space with respect

2,Ω 0 𝑉
푓 fl 𝐻1 (Ω
푓 )
to the inner product 𝑉0 × 𝑉0 ∋ (→ 󳨀V ,→󳨀
𝑤) 󳨀→ ∫Ω 𝜀(→ 󳨀V ) : 𝜀(→󳨀
𝑤)𝑑𝑥 (5)
→
󳨀
푠 2
푠 3
and the associated norm 𝑉0 ∋ V 󳨃󳨀→ ‖𝜀( V )‖2,Ω is equivalent 𝐻 fl 𝐿 (Ω ) ,
to the norm induced by 𝐻1 (Ω)3 .
If 𝑋 is a Banach space based on the domain Ω, we will 𝑉
푠 fl {→
󳨀V ∈ 𝐻1 (Ω
푠 )3 : →
󳨀V = 0 on Γ } ,
5
write 𝑋耠 for its dual space and ⟨𝜒, 𝑢⟩Ω for the evaluation of
𝜒 ∈ 𝑋耠 at 𝑢 ∈ 𝑋. The Bochner spaces associated with 𝑋 will be where 𝐻
푓 is equipped with inner product (𝑝, 𝑞) 󳨃󳨀→
denoted by 𝐿
푟 (𝑎, 𝑏; 𝑋), 1 ≤ 𝑟 ≤ ∞, 𝐶
푤 ([𝑎, 𝑏]; 𝑋) will denote (𝑝, 𝑞)Ω𝑓 + (𝑝, 𝑞)Γ3 and 𝑉
푠 is equipped with inner product
the space of weakly continuous functions with values in 𝑋 󳨀V ,→
󳨀 󳨀V ), 𝜀(→
󳨀
and D耠 (0, 𝑇; 𝑋) will be a space of distributions. (→ 𝑤) 󳨃󳨀→ (𝜀(→ 𝑤))Ω𝑠 .
→
󳨀
Suppose 𝑝 and 𝑢 are smooth solutions of Problem (1a)-
We recall the method of mollifiers, which provides
approximation by smooth functions. Here we will consider (1j). Multiplying both sides of (1b) by 𝜓 ∈ 𝑉
푓 and integrating
𝑤 = 𝑤(𝑡, 𝑥) and mollification in the 𝑡 variable. Let the by parts over Ω
푓 yields
function J ∈ 𝐶0∞ (R) satisfy the following properties:

supp J ⊆ [−1, 1] , 0 ≤ J ≤ 1, 𝑑2
∫ 𝑝𝜓 𝑑𝑥
𝑑𝑡2 Ω𝑓
1 (6)
∫ J (𝜏) 𝑑𝜏 = 1, 2
−1 (2) = 𝑐 (∫ 𝜕→
㨀
푛 𝑓 𝑝𝜓 𝑑𝑠 − ∫ ∇
푥 𝑝 ⋅ ∇
푥 𝜓 𝑑𝑥) ,

휕Ω𝑓 Ω𝑓
J (𝜏) = J (−𝜏)
where by (1d), (1e), (1f), and (1c)
∀𝜏 ∈ R.

The mollifier J
훿 ∈ 𝐶0∞ (R), 𝛿 > 0, is defined by J
훿 (𝜏) fl
푓 →󳨀 → 󳨀
푠

∫ 𝜕→
㨀
푛 𝑓 𝑝𝜓 𝑑𝑠 = −󰜚 ∫ 𝜕
푡
푡 𝑢 ⋅ 𝑛 𝜓 𝑑𝑠
𝛿−1 J(𝜏/𝛿) (𝜏 ∈ R). For a Banach space 𝑋, a smoothing
휕Ω𝑓 Γ1
operator can be defined by convolution with J
훿 in the (7)
following way: given 𝑤 ∈ 𝐿
푟 (𝑎, 𝑏; 𝑋) and 1 ≤ 𝑟 < ∞, set 1 1
− ∫ 𝜕
푡
푡 𝑝𝜓 𝑑𝑠 − ∫ 𝜕
푡 𝑝𝜓 𝑑𝑠.
𝑤(𝑡) = 0 for all 𝑡∈[𝑎,
훿 ∞ 𝑔 Γ3 𝑐 Γ4

𝑏], and let 𝑤 = J
훿 ⋆ 𝑤 ∈ 𝐶 (𝑎, 𝑏; 𝑋)
4 Advances in Mathematical Physics

Therefore of the solution. In (11a)-(11b) we have used the following

notation (see [3]):
𝑑2
󰜚
푓 ∫ → 󳨀
𝑢 ⋅→
󳨀𝑛
푠 𝜓 𝑑𝑠
𝑑𝑡2 Γ1 M (→
󳨀 𝜑 ) fl 󰜚
푠 (→
𝑢 ,→
󳨀 󳨀
𝑢 ,→
󳨀
𝜑 )Ω𝑠 ,

𝑑2 1 1 1 1
+ ( 2 ∫ 𝑝𝜓 𝑑𝑥 + ∫ 𝑝𝜓 𝑑𝑠) S (𝑝, 𝜓) fl (𝑝, 𝜓)Ω𝑓 + (𝑝, 𝜓)Γ3 ,
2
𝑑𝑡 𝑐 Ω 𝑓 𝑔 Γ3 (8) 𝑐2 𝑔
1 𝑑 C
푠 (→
󳨀
𝑢 ,→
󳨀
𝜑 ) fl 𝑐
푎 (→
󳨀
𝑢 ,→
󳨀
𝜑 )Ω𝑠 ,
+ ∫ 𝑝𝜓 𝑑𝑠 + ∫ ∇
푥 𝑝 ⋅ ∇
푥 𝜓 𝑑𝑥 = 0,
𝑐 𝑑𝑡 Γ4 Ω𝑓
K (→
󳨀
𝑢 ,→
󳨀
𝜑 ) fl 2𝜇 (𝜀 (→
󳨀
𝑢 ) , 𝜀 (→
󳨀
𝜑 ))Ω𝑠

푓
∀𝜓 ∈ 𝑉 .
+ 𝜆 (∇
푥 ⋅ →
󳨀
𝑢 , ∇
푥 ⋅ →
󳨀
𝜑 )Ω𝑠 ,
Analogously, dot multiplying both sides of (1a) by → 󳨀𝜑 ∈ 𝑉
푠 , (12)

푠 →
󳨀 →
󳨀
integrating by parts over Ω , and using 𝜎( 𝑢 ) : ∇
푥 𝜑 = 𝜆(∇
푥 ⋅ F
푠 (→
󳨀
𝜑 ) fl 󰜚
푠 (→
󳨀
𝑔 +→
󳨀𝑎
푠 ,→
󳨀
𝜑 )Ω𝑠 ,
󳨀𝑢 )(∇ ⋅ →
→ 󳨀 →
󳨀 →
󳨀

푥 𝜑 ) + 2𝜇𝜀( 𝑢 ) : 𝜀( 𝜑 ) yields 1
C
푓 (𝑝, 𝜓) fl (𝑝, 𝜓)Γ4 ,
𝑐
𝑑2
󰜚
푠 ∫ → 󳨀
𝑢 ⋅→
󳨀
𝜑 𝑑𝑥 H (𝑝, 𝜓) fl (∇
푥 𝑝, ∇
푥 𝜓)Ω𝑓 ,
𝑑𝑡2 Ω𝑠
Q
푠 (𝑝,→
󳨀
𝜑 ) fl (𝑝→
󳨀𝑛
푠 ,→
󳨀
𝜑 )Γ ,
=∫ (𝜎 (→
󳨀
𝑢 )→
󳨀𝑛
푠 ) ⋅ →
󳨀
𝜑 𝑑𝑠 1

휕Ω𝑠
Q (→
󳨀
𝑢 , 𝜓) fl 󰜚 (→
󳨀 →
󳨀
푠
(9)
푓
푓
𝑢 , 𝑛 𝜓)Γ .
− ∫ [𝜆 (∇
푥 ⋅ →
󳨀
𝑢 ) (∇
푥 ⋅ →
󳨀
𝜑 ) + 2𝜇𝜀 (→
󳨀
𝑢 ) : 𝜀 (→
󳨀
𝜑 )] 𝑑𝑥 1

Ω𝑠
In (12), M and S are the so-called mass operators, C
푠 and
𝑑
− 𝑐
푎 ∫ → 󳨀 𝜑 𝑑𝑥 + 󰜚
푠 ∫ (→
𝑢 ⋅→
󳨀 󳨀
𝑔 +→
󳨀𝑎
푠 ) ⋅ →
󳨀
𝜑 𝑑𝑥. C
푓 are drag operators, K and H are stiffness operators, and
𝑑𝑡 Ω𝑠 𝑠
Ω
Q
푠 and Q
푓 are interaction operators for the structure and the
Imposing the boundary conditions (1g) and (1h), we obtain fluid, respectively. Note that Q
푓 (→
󳨀𝑢 , 𝑝) = 󰜚
푓 Q
푠 (𝑝,→
󳨀
𝑢 ), but the
interaction operators appear in (11a)-(11b) in an unbalanced
way in terms of the order of the time derivatives, which makes
𝑑2 𝑑
󰜚
푠 ∫ → 󳨀
𝑢 ⋅→
󳨀
𝜑 𝑑𝑥 + 𝑐
푎 ∫ → 󳨀
𝑢 ⋅→
󳨀
𝜑 𝑑𝑥 it difficult to obtain good a priori estimates to use in the
𝑑𝑡2 Ω𝑠 𝑑𝑡 Ω𝑠 mathematical analysis of the model.

+ ∫ [𝜆 (∇
푥 ⋅ →
󳨀
𝑢 ) (∇
푥 ⋅ →
󳨀
𝜑 ) + 2𝜇𝜀 (→
󳨀𝑢 ) : 𝜀 (→
󳨀
𝜑 )] 𝑑𝑥
Ω𝑠 3.3. Some Considerations on the Regularity of the Pressure: For-
(10)
mal Energy Equation and a Priori Estimates. Let us assume
− ∫ 𝑝→
󳨀𝑛
푠 ⋅ →
󳨀
𝜑 𝑑𝑠 = 󰜚
푠 ∫ (→
󳨀
𝑔 +→
󳨀𝑎
푠 ) ⋅ →
󳨀
𝜑 𝑑𝑥, that a solution (→ 󳨀
𝑢 , 𝑝) of (1a)-(1j) exists and is sufficiently
Γ1 Ω𝑠 regular. If we try to obtain a (formal) energy estimate for such
∀→
󳨀
𝜑 ∈ 𝑉
푠 . a solution by replacing→ 𝜑 ∈ 𝑉
푠 in (11a) by 𝜕
푡→
󳨀 󳨀
𝑢 and 𝜓 by 𝜕
푡 𝑝/󰜚
푓
in (11b), followed by time integration in [0, 𝑡], we end up with
the identity
Hence the following variational problem is associated
with (1a)-(1j): find →
󳨀
𝑢 ∈ D耠 (0, 𝑇; 𝑉
푠 ) and 𝑝 ∈ D耠 (0, 𝑇; 𝑉
푓 ) 1 󵄩󵄩 󵄩2 1 󵄩 󵄩2
󵄩∇ 𝑝 (𝑡)󵄩󵄩󵄩2,Ω𝑓 +
푓 2 󵄩󵄩󵄩𝜕
푡 𝑝 (𝑡)󵄩󵄩󵄩2,Ω𝑓
󰜚
푓 󵄩
푥
such that
󰜚 𝑐
𝑑2 𝑑 1 󵄩󵄩 2
푡 󵄩󵄩
M (→ 󳨀
𝑢 ,→
󳨀
𝜑 ) + C
푠 (→ 󳨀
𝑢 ,→
󳨀
𝜑 ) + K (→󳨀
𝑢 ,→
󳨀
𝜑) + 󵄩𝜕 𝑝 (𝑡)
󵄩󵄩2 󵄩󵄩2
󵄩2,Γ3 󰜚
푓 𝑐 ∫0 󵄩󵄩𝜕
휏 𝑝 (𝜏)󵄩󵄩2,Γ4 𝑑𝜏
󵄩 +
𝑑𝑡2 𝑑𝑡 (11a) 󰜚
푓 𝑔 󵄩
푡

− Q
푠 (𝑝,→
󳨀𝜑 ) = F
푠 (→
󳨀
𝜑) , → 󳨀
𝜑 ∈ 𝑉
푠 󵄩 󳨀 󵄩2 󵄩 󳨀 󵄩󵄩2
𝑢 ) (𝑡)󵄩󵄩󵄩󵄩2,Ω𝑠 + 󰜚
푠 󵄩󵄩󵄩󵄩𝜕
푡→
+ 2𝜇 󵄩󵄩󵄩󵄩𝜀 (→ 𝑢 (𝑡)󵄩󵄩󵄩2,Ω𝑠
𝑑2 𝑑
S (𝑝, 𝜓) + C
푓 (𝑝, 𝜓) + H (𝑝, 𝜓) 󵄩 󵄩2
푡
󵄩 󳨀 󵄩2
𝑑𝑡2 𝑑𝑡 + 𝜆 󵄩󵄩󵄩󵄩∇
푥 ⋅ →
󳨀
𝑢 (𝑡)󵄩󵄩󵄩󵄩2,Ω𝑠 + 2𝑐
푎 ∫ 󵄩󵄩󵄩󵄩𝜕
휏→
𝑢 (𝜏)󵄩󵄩󵄩󵄩2,Ω𝑠 𝑑𝜏
(11b) 0
𝑑2
+ 2 Q
푓 (→
󳨀
𝑢 , 𝜓) = 0, 𝜓∈𝑉
푓
푡→󳨀
𝑑𝑡 = 2󰜚
푠 ∫ (→
󳨀
𝑔 + 𝑎
푠 (𝜏) , 𝜕
휏→
󳨀
𝑢 (𝜏)) 𝑑𝜏
0 Ω𝑠

and →
󳨀
𝑢 (0) = →
󳨀
𝑢 0 , 𝜕
푡→
󳨀𝑢 (0) = →
󳨀V , 𝑝(0) = 𝑝 , and 𝜕 𝑝(0) = 𝑞 , in 1 󵄩󵄩 󵄩2 1 󵄩 󵄩
0 0
푡 0 + 󵄩∇𝑝 󵄩󵄩 𝑓 +
푓 2 󵄩󵄩󵄩𝑞0 󵄩󵄩󵄩Ω𝑓 𝑑𝜏
some sense, to be specified in terms of continuity properties 󰜚
푓 󵄩 0 󵄩2,Ω 󰜚 𝑐
Advances in Mathematical Physics 5

1 󵄩󵄩 󵄩󵄩2 󵄩 󳨀 󵄩2
+ 󵄩𝑞 󵄩 + 2𝜇 󵄩󵄩󵄩󵄩𝜀 (→
𝑢 0 ) (𝑡)󵄩󵄩󵄩󵄩2,Ω𝑠 which is the result of formally integrating (11b) in [0, 𝑡]. The
󰜚
푓 𝑔 󵄩 0 󵄩2,Γ3 same procedure that leads to (13) allows us to derive a formal
󵄩󳨀 󵄩󵄩2 󵄩 󵄩2 energy equation for (→󳨀
𝑢 , 𝑃)
V 0 󵄩󵄩󵄩2,Ω𝑠 + 𝜆 󵄩󵄩󵄩󵄩∇
푥 ⋅ →
+ 󰜚
푠 󵄩󵄩󵄩󵄩→ 󳨀
𝑢 0 󵄩󵄩󵄩󵄩2,Ω𝑠
1 1 󵄩󵄩 󵄩2 1 󵄩 󵄩2
[ 󵄩∇ 𝑃 (𝑡)󵄩󵄩󵄩2,Ω𝑓 +
푓 2 󵄩󵄩󵄩𝜕
푡 𝑃 (𝑡)󵄩󵄩󵄩2,Ω𝑓
+ ∫ 𝑝𝜕
푡→
󳨀
𝑢 ⋅→
󳨀𝑛
푠 𝑝 𝑑𝑠 − ∫ 𝜕 → 󳨀 → 󳨀
푠

푡
푡 𝑢 ⋅ 𝑛 𝜕
푡 𝑝 𝑑𝑠.
2 󰜚
푓 󵄩
푥 󰜚 𝑐
Γ1 Γ1
1 󵄩󵄩 󵄩2 1 󵄩 󳨀 󵄩2
(13) + 󵄩𝜕 𝑃 (𝑡)󵄩󵄩󵄩2,Γ3 ] + [2𝜇 󵄩󵄩󵄩󵄩𝜀 (→
𝑢 ) (𝑡)󵄩󵄩󵄩󵄩2,Ω𝑠
󰜚
푓 𝑔 󵄩
푡 2
In order to “skew-symmetrize” the bilinear form associ-
󵄩 󵄩2 󵄩 󳨀 󵄩󵄩2 1
ated with the equations on Γ1 ×]0, 𝑇], so that the unbalanced + 𝜆 󵄩󵄩󵄩󵄩∇
푥 ⋅ →
󳨀
𝑢 (𝑡)󵄩󵄩󵄩󵄩2,Ω𝑠 + 󰜚
푠 󵄩󵄩󵄩󵄩𝜕
푡→
𝑢 (𝑡)󵄩󵄩󵄩2,Ω𝑠 ] +
푓
terms ∫Γ 𝑝𝜕
푡→ 󳨀
𝑢 ⋅→󳨀𝑛
푠 𝑑𝑠 − ∫ 𝜕 →
Γ1
푡
푡
󳨀
𝑢 ⋅→
󳨀𝑛
푠 𝜕 𝑝 𝑑𝑠 are not present

푡
󰜚 𝑐
1

푡
푡
in the energy equation, it is convenient to introduce a 󵄩 󵄩2 󵄩󵄩 → 󵄩󵄩2
⋅ ∫ 󵄩󵄩󵄩𝜕
휏 𝑃󵄩󵄩󵄩2,Γ4 𝑑𝜏 + 𝑐
푎 ∫ 󵄩󵄩𝜕
휏 󳨀 󵄩
pressure potential for 𝑝 (with respect to the 𝑡 variable): 𝑃 ∈
0 0
󵄩 𝑢 󵄩󵄩2,Ω𝑠 𝑑𝜏
D耠 (0, 𝑇; 𝑉
푓 ), 𝑝 = 𝜕
푡 𝑃.

푡 →󳨀 󳨀
Note that taking into account the relation 󰜚
푓 𝜕
푡→ 󳨀V = −∇ 𝑝,

푥 = 󰜚
푠 ∫ (→ 󳨀
𝑔 + 𝑎
푠 , 𝜕
휏→
1
𝑢 ) 𝑑𝜏 +
푓 2

푓
in Ω ×]0, 𝑇], between the fluid velocity and pressure in 0 Ω𝑠 󰜚 𝑐
(18)
this model (see [3], pg. 634), the variable 𝑃 can be seen as
푡
1
푡
a velocity potential for the eliminated unknown → 󳨀V in the ⋅ ∫ (𝑞0 , 𝜕
휏 𝑃)Ω𝑓 𝑑𝜏 +
푓 ∫ (𝑞0 , 𝜕
휏 𝑃)Γ3 𝑑𝜏
fluid equations. This is consistent with the assumption of 0 󰜚 𝑔 0
irrotational flow. Now, replacing 𝑝 = 󰜚
푓 𝜕
푡 𝑃 in (11a) yields 1
푡 󳨀V ⋅ →
󳨀
푠
+ ∫ (𝑝 , 𝜕 𝑃) 𝑑𝜏 + (→ 0 𝑛 , 𝑃 (𝑡))Γ
𝑑2 𝑑 󰜚
푓 𝑐 0 0
휏 Γ4 1

M (→
󳨀
𝑢 ,→
󳨀
𝜑 ) + C
푠 (→
󳨀
𝑢 ,→
󳨀
𝜑 ) + K (→
󳨀
𝑢 ,→
󳨀
𝜑)
𝑑𝑡2 𝑑𝑡 1 󵄩󵄩 󵄩󵄩2 1 󵄩 󵄩2
(14) + 󵄩𝑝 󵄩 𝑓 +
푓 󵄩󵄩󵄩𝑝0 󵄩󵄩󵄩2,Γ3
𝑑
푓
푠 → 2󰜚
푓 𝑐2 󵄩 0 󵄩2,Ω 2󰜚 𝑔
− 󰜚 Q (𝑃, 󳨀
𝜑 ) = F
푠 (→
󳨀
𝜑) , →
󳨀
𝜑 ∈ 𝑉
푠 ,
𝑑𝑡 󵄩 󳨀 󵄩2 𝜆󵄩 󵄩2
+ 𝜇 󵄩󵄩󵄩󵄩𝜀 (→
𝑢 0 ) (𝑡)󵄩󵄩󵄩󵄩2,Ω𝑠 + 󵄩󵄩󵄩󵄩∇
푥 ⋅ →
󳨀
𝑢 0 󵄩󵄩󵄩󵄩2,Ω𝑠
while the same replacement in (11b) produces a higher order 2
ordinary differential equation 󰜚
푠 󵄩󵄩→ 󵄩2
+ 󵄩󵄩󳨀V 0 󵄩󵄩󵄩 𝑠
𝑑3
𝑑 2
𝑑 2 󵄩 󵄩2,Ω
3
S (𝑃, 𝜓) + 2 C
푓 (𝑃, 𝜓) + H (𝑃, 𝜓)
𝑑𝑡 𝑑𝑡 𝑑𝑡 and (18), in turn, yields the following basic a priori estimates
(15)
2
𝑑
푓 → for (→
󳨀
𝑢 , 𝑃):
+ Q (󳨀
𝑢 , 𝜓) = 0, 𝜓 ∈ 𝑉
푓 on [0, 𝑡] .
𝑑𝑡2 󵄩󵄩 → 󵄩󵄩 󵄩󵄩 → 󵄩
󵄩󵄩∇
푥 󳨀 󵄩 󵄩 󳨀 󵄩󵄩
󵄩 𝑢 󵄩󵄩
퐿∞ (0,
푇;
퐿2 (Ω𝑠 )) + 󵄩󵄩𝜕
푡 𝑢 󵄩󵄩
퐿∞ (0,
푇;
퐿2 (Ω𝑠 )) ≤ 𝐶1 , (19a)
Then
󵄩󵄩 󵄩 󵄩 󵄩
󵄩󵄩∇
푥 𝑃󵄩󵄩󵄩
퐿∞ (0,
푇;
퐿2 (Ω𝑓 )) + 󵄩󵄩󵄩𝜕
푡 𝑃󵄩󵄩󵄩
퐿∞ (0,
푇;
퐿2 (Ω𝑓 )) ≤ 𝐶2 (19b)
𝑑2 𝑑
S (𝑃, 𝜓) + C
푓 (𝑃, 𝜓) + H (𝑃, 𝜓)
𝑑𝑡2 𝑑𝑡 󵄩󵄩󵄩𝜕
푡 𝑃󵄩󵄩󵄩 ∞ 󵄩󵄩 󵄩󵄩
(16) 󵄩 󵄩
퐿 (0,
푇;
퐿2 (Γ3 )) + 󵄩󵄩𝜕
푡 𝑃󵄩󵄩
퐿2 (0,
푇,
퐿2 (Γ4 )) ≤ 𝐶3 (19c)
𝑑
푓 →
+ Q (󳨀
𝑢 , 𝜓) = 𝐶, 𝜓 ∈ 𝑉
푓 where the constants 𝐶
푖 , 𝑖 = 1, 2, 3, depend on the data.
𝑑𝑡
Relation (18) suggests that the pair (→ 󳨀𝑢 , 𝑃) satisfies 𝑃 ∈
where 𝐶 is a constant (distribution) and, more precisely, 𝐶([0, 𝑇]; 𝑉
푓 ), 𝑝 = 𝜕
푡 𝑃 ∈ 𝐶([0, 𝑇]; 𝐻
푓 ) and → 󳨀
𝑢 ∈ 𝐶([0, 𝑇]; 𝑉
푠 ),
󳨀V , 𝜓). If →
𝐶 = S(𝑞0 , 𝜓) + C
푓 (𝑝0 , 𝜓) + Q
푓 (→ 󳨀
𝑢 , 𝑝, and 𝑃 are →󳨀
푠
0 and 𝜕
푡 𝑢 ∈ 𝐶([0, 𝑇]; 𝐻 ). The above considerations also
regular functions with 𝑃(0) = 0, then the chosen value for indicate that the time regularity we can expect of the pressure
𝐶 is just the value of (𝑑2 /𝑑𝑡2 )S(𝑃, 𝜓) + (𝑑/𝑑𝑡)C
푓 (𝑃, 𝜓) + is less than that of the structure displacement, more precisely,
H(𝑃, 𝜓) + (𝑑/𝑑𝑡)Q
푓 (→
󳨀
𝑢 , 𝜓) at 𝑡 = 0. Hence, we consider the ⋅
new equation for the unknown 𝑃 𝑃 = ∫ 𝑝 (𝑠) 𝑑𝑠 ∈ 𝐶 ([0, 𝑇] ; 𝐻1 (Ω
푓 )) ,
0
(20)
𝑑2 𝑑
2
S (𝑃, 𝜓) + C
푓 (𝑃, 𝜓) + H (𝑃, 𝜓) 2 󵄨
𝑝 ∈ 𝐶 ([0, 𝑇] ; 𝐿 (Ω )) , 𝑝󵄨󵄨󵄨Γ3 ∈ 𝐶 ([0, 𝑇] ; 𝐿2 (Γ3 )) ,

푓
𝑑𝑡 𝑑𝑡
𝑑
푓 →
+ Q (󳨀
𝑢 , 𝜓) with 𝑝(0) = 𝑝0 . Moreover, if 𝑝 = 𝜕
푡 𝑃 and satisfies
𝑑𝑡 (17)
1 1
= S (𝑞0 , 𝜓) + C
푓 (𝑝0 , 𝜓) + Q
푓 (→
󳨀V , 𝜓) ,
0 𝑐2
⟨𝜕
푡 𝑝, 𝜓⟩Ω𝑓 = − (∇
푥 𝑃, ∇
푥 𝜓)Ω𝑓 + 2 (𝑞0 , 𝜓)Ω𝑓 ,
𝑐 (21)
𝜓∈𝑉 ,
푓
∀𝜓 ∈ 𝐻01 (Ω ) ,
푓
6 Advances in Mathematical Physics

and ∇
푥 𝑃(0) = 0 in Ω
푓 , it follows that 𝜕
푡 𝑝 ∈ S (𝜕
푡
푡 𝑝
푚 , 𝜓
푘 ) + C
푓 (𝜕
푡 𝑝
푚 , 𝜓
푘 ) + Q
푓 (𝜕
푡
푡→
󳨀
𝑢
푚 , 𝜓
푘 )
𝐶([0, 𝑇]; 𝐻 (Ω )) and 𝜕𝑝(0) = 𝑞0 in 𝐻−1 (Ω
푓 ). Note that
−1
푓 (23b)
(21) is obtained from (17) taking test functions →
󳨀
𝜑 ∈ 𝐻01 (Ω
푓 ). + H (𝑝
푚 , 𝜓
푘 ) = 0,

Remark 2. The variational formulations (14) and (17) are for 𝑘 = 1 : 𝑚, and the initial conditions
actually a formulation in displacement-velocity potential. →
󳨀𝑢 (0) = P 𝑠 →
󳨀

푉𝑚 𝑢 0 ,
From→ 󳨀V = ∇ Ψ and the relation 󰜚
푓 𝜕 →

푥
󳨀
푓

푡 V = −∇
푥 𝑝, in Ω ×]0, 𝑇],

푚

between the fluid velocity and the pressure in this model (see 𝑝
푚 (0) = P
푉𝑓 𝑝0
[3], pg. 634), we get 𝑝 = −󰜚
푓 𝜕
푡 Ψ. 𝑚
(24)
𝑢̇
푚 (0) = P
푉𝑚𝑠 →
→
󳨀 󳨀V ,
0
3.4. Definition of Weak Solution and Well Posedness of the
Mathematical Model. Based on the considerations made in 𝑝
푚̇ (0) = P
푉𝑓 𝑞0 ,
𝑚
the previous section, we shall say that a weak solution
to Problems (1a)-(1j) is a pair (→ 󳨀
𝑢 , 𝑝) ∈ D耠 (0, 𝑇; 𝑉
푠 ) × where P
푉𝑚𝑠 and P
푉𝑓 are projection operators onto 𝑉
푚
푠 and 𝑉
푚
푓 ,
耠
푓 𝑚
D (0, 𝑇; 𝑉 ) satisfying (11a)-(11b) and respectively. This is a system of second-order ODEs, which
→
󳨀 has a unique solution. Indeed, (23a)-(23b) can be written in
𝑢 ∈ 𝐶 ([0, 𝑇] ; 𝑉
푠 ) ∩ 𝐶1 ([0, 𝑇] ; 𝐻
푠 ) , the form
with →
󳨀
𝑢 (0) = →
󳨀
𝑢 0 , 𝜕
푡→
󳨀
𝑢 (0) = →
󳨀V ,
0 M 0 𝑐
푚̈ C
푠 0 𝑐
푚̇
[ ][ ] + [ ] [ ]

푡

푓 󰜚
푓 Q⊤ S 𝑑
푚̈ 0 C
푓 𝑑
푚̇
∫ 𝑝 (𝑠) 𝑑𝑠 ∈ 𝐶 ([0, 𝑇] ; 𝑉 ) , (25)
0
(22) K −Q 𝑐
푚 f
푠
+[ ][ ] = [ ],
𝑝 ∈ 𝐶 ([0, 𝑇] ; 𝐻
푓 ) , with 𝑝 (0) = 𝑝0 , 0 H 𝑑
푚 0

𝜕
푡 𝑝 ∈ 𝐶 ([0, 𝑇] ; 𝐻−1 (Ω
푓 )) where we have used the notation 𝑦̇ = 𝑑𝑦/𝑑𝑡 and

with 𝜕
푡 𝑝 (0) = 𝑞0 in 𝐻−1 (Ω
푓 ) . M
푖
푗 fl M (→

󳨀
𝜑
푖 ,→
󳨀
𝜑
푗 ) ,

The main result concerning the well posedness of the mathe- C
푠
푖
푗 fl C
푠
푖
푗 (→
󳨀
𝜑
푖 ,→
󳨀
𝜑
푗 ) ,
matical models (1a)-(1j) is as follows.
K
푖
푗 fl K (→
󳨀
𝜑
푖 ,→
󳨀
𝜑
푗 )
Theorem 3. Let Ω
푠 , Ω
푓 ⊂ R3 be locally Lipschitz bounded
domains with the common boundary Γ1 , as described in 1
푓 →
Q
푖
푗 fl Q (󳨀
𝜑
푖 , 𝜓
푗 ) ,
Section 2. Assume that → 󳨀𝑎
푠 ∈ 𝐿1 (0, 𝑇; 𝐿2 (Ω
푠 )3 ), 𝑝 ∈ 𝑉
푓 ,
0 󰜚
푓
→󳨀 →
󳨀
𝑞0 ∈ 𝐻 , and 𝑢 0 , V 0 ∈ 𝑉
푠 . Then there exists a unique solution

푓 (26)
(→
󳨀
𝑢 , 𝑝) to Problems (11a)-(11b). S
푖
푗 fl S (𝜓
푖 , 𝜓
푗 ) ,

푓
Theorem 3 will be proved along several steps in the next C
푖
푗 fl C
푓 (𝜓
푖 , 𝜓
푗 ) ,
subsections. The a priori estimates (19a)-(19c) for (→
󳨀𝑢 , 𝑃) will
be used to construct a weak solution for the problem, starting H
푖
푗 fl H (𝜓
푖 , 𝜓
푗 ) ,
from a generalization of the classical Galerkin method for
second-order hyperbolic problems (see [19, 20]). The method f
푗
푠 fl F
푠 (→
󳨀
𝜑
푗 ) .
used for studying the transmission problem in [16] might also
be adapted to the equations satisfied by (→
󳨀
𝑢 , 𝑃). The matrices M and S are symmetric, positive definite and
since
3.5. Galerkin Approximations. Let {𝜓
푘 }
푘∈N and {→ 󳨀𝜑
푘 }
푘∈N be M 0

푓
푠
basis of smooth functions for 𝑉 and 𝑉 , respectively. Let det ([
푓 ⊤ ]) = det (M) det (S) ≠ 0 (27)
󰜚 Q S
𝑉
푚
푓 = span{𝜓1 , . . . , 𝜓
푚 } and 𝑉
푚
푠 = span{→ 󳨀
𝜑 1 , . . . ,→
󳨀
𝜑
푚 }.
We seek approximating solutions in the form 𝑢
푚 (𝑥, 𝑡) = →
󳨀 M 0
→
󳨀 the matrix M fl [ ‫ 𝑓ܚ‬Q⊤ S ] is nonsingular.
∑
푚
푙
푚
푙

푙=1 𝑐
푚 (𝑡) 𝜑
푙 (𝑥) and 𝑝
푚 (𝑥, 𝑡) = ∑
푙=1 𝑑
푚 (𝑡)𝜓
푙 (𝑥) with the Let
coefficients 𝑐
푚
푙 (𝑡) and 𝑑
푚
푙
(𝑡) (0 ≤ 𝑡 ≤ 𝑇, 𝑙 = 1 : 𝑚) obtained
푡
from
푙
푙
𝐷
푚 (𝑡) fl ∫ 𝑑
푚 (𝜏) 𝑑𝜏,
0
𝑢
푚 ,→
M (𝜕
푡
푡→
󳨀 󳨀 𝑢
푚 ,→
𝜑
푘 ) + C
푠 (𝜕
푡→
󳨀 󳨀
𝜑
푘 ) − Q
푠 (𝑝
푚 ,→
󳨀
𝜑
푘 )
푡
(28)
(23a) 𝑃
푚 (𝑥, 𝑡) fl ∫ 𝑝
푚 (𝑥, 𝜏) 𝑑𝜏,
𝑢
푚 ,→
+ K (→
󳨀 󳨀
𝜑
푘 ) = F
푠 (→
󳨀
𝜑
푘 ) , 0
Advances in Mathematical Physics 7

푡
so that 𝑝
푚 (𝑥, 𝑡) = 𝜕
푡 𝑃
푚 (𝑥, 𝑡) and 𝑃
푚 (𝑥, 0) = 0. Then the ∫ (∇
푥 𝑃 (𝜏) , ∇
푥 𝜓)Ω𝑓 𝜒 (𝜏) 𝑑𝜏
approximating functions (→ 󳨀
𝑢
푚 , 𝑃
푚 ) satisfy 0

󳨀
M (𝜕 𝑢 , 𝜑 ) + C (𝜕 →
→
󳨀 → 𝑢 ,→
󳨀 󳨀

푠
푡
𝜑 ) 1 1

푡
푡
푚
푘
푡
푚
푘 −∫ [ 2
(𝜕
푡 𝑃 (𝜏) , 𝜓)Ω𝑓 + (𝜕
푡 𝑃 (𝜏) , 𝜓)Γ3 ]
(29a) 0 𝑐 𝑔
− 󰜚
푓 Q
푠 (𝜕
푡 𝑃
푚 ,→
󳨀
𝜑
푘 ) + K (→
󳨀
𝑢
푚 ,→
󳨀
𝜑
푘 ) = F
푠 (→
󳨀
𝜑
푘 ) ,
푡
1
⋅ 𝜒̇ (𝜏) 𝑑𝜏 − ∫ [ (𝑃 (𝜏) , 𝜓)Γ4
S (𝜕
푡
푡 𝑃
푚 , 𝜓
푘 ) + C
푓 (𝜕
푡 𝑃
푚 , 𝜓
푘 ) + Q
푓 (𝜕
푡→
󳨀
𝑢
푚 , 𝜓
푘 ) 0 𝑐
(32b)
+ H (𝑃
푚 , 𝜓
푘 ) = S (P
푉𝑓 𝑞0 , 𝜓
푘 ) + 󰜚 (→
󳨀 󳨀𝑛
푠 , 𝜓) ] 𝜒̇ (𝜏) 𝑑𝜏 = [ 1 (𝑝
𝑢 (𝜏) ⋅ →

푓
(29b)
𝑚 Γ1 𝑐2 0
+ C
푓 (P
푉𝑓 𝑝0 , 𝜓
푘 ) + Q
푓 (P
푉𝑚𝑠 →
󳨀V , 𝜓 ) .
0
푘 1
𝑚 + 𝑞0 , 𝜓)Ω𝑓 𝜒 (0) + (𝑝 + 𝑞0 , 𝜓)Γ3
𝑔 0
Now we multiply (29a) by 𝑐
푚
푘̇ (𝑡) and (29b) by 𝐷̇
푚
푘
(𝑡)/󰜚
푓 and
sum from 𝑘 = 1 to 𝑘 = 𝑚 to conclude that 𝑃
푚 and 󳨀 →𝑢
푚 satisfy
+ 󰜚
푓 ((→
󳨀 󳨀V ) ⋅ →
𝑢 0 +→ 0
󳨀𝑛
푠 , 𝜓) ] 𝜒 (0) ,
relations similar to (14) and (17) in the spaces 𝑉
푚 and 𝑉
푚
푓 and,

푠 Γ 1

consequently, the a priori estimates (19a)-(19c).

for all →
󳨀
𝜑 ∈ 𝑉
푠 and 𝜓 ∈ 𝑉
푓 and all 𝜙, 𝜒 ∈ 𝐶1 ([0, 𝑇]) null in 𝑇.
Lemma 4. For the Galerkin approximations (→
󳨀𝑢 , 𝑃 ) it holds
that

푚
푚
We now take → 󳨀𝜑 ∈ 𝐻01 (Ω
푠 )3 and 𝜓 ∈ 𝐻01 (Ω
푓 ) in (32a)-(32b)
and obtain
{𝜕
푡 𝑃
푚 } 𝑖𝑠 𝑏𝑜𝑢𝑛𝑑𝑒𝑑 𝑖𝑛 𝐿∞ (0, 𝑇; 𝐻
푓 ) ,
󰜚
푠 ⟨𝜕
푡
푡→
󳨀
𝑢 ,→
󳨀
𝜑 ⟩Ω𝑠 = −𝑐
푎 (𝜕
푡→
󳨀
𝑢 ,→
󳨀
𝜑 )Ω𝑠
{𝜕
푡→
󳨀
𝑢
푚 } 𝑖𝑠 𝑏𝑜𝑢𝑛𝑑𝑒𝑑 𝑖𝑛 𝐿∞ (0, 𝑇; 𝐻
푠 ) ,
− 𝜇 (𝜀 (→
󳨀𝑢 ) , 𝜀 (→
󳨀
𝜑 ))Ω𝑠
{𝑃
푚 } 𝑖𝑠 𝑏𝑜𝑢𝑛𝑑𝑒𝑑 𝑖𝑛 𝐿∞ (0, 𝑇; 𝑉
푓 ) , (30)
− 𝜆 (∇
푥 ⋅ →
󳨀
𝑢 , ∇
푥 ⋅ →
󳨀
𝜑 )Ω𝑠
{→
󳨀
𝑢
푚 } 𝑖𝑠 𝑏𝑜𝑢𝑛𝑑𝑒𝑑 𝑖𝑛 𝐿∞ (0, 𝑇; 𝑉
푠 ) ,
→󳨀 󳨀
󵄨
{ 𝜕
푡 𝑃
푚 󵄨󵄨󵄨Γ4 } 𝑖𝑠 𝑏𝑜𝑢𝑛𝑑𝑒𝑑 𝑖𝑛 𝐿2 (0, 𝑇; 𝐿2 (Γ4 )) , + 󰜚
푠 (→
󳨀
𝑔 + 𝑎
푠 ,→
𝜑) , (33)
Ω𝑠

∀→
󳨀
The uniform bounds collected in Lemma 4 allow us 3
𝜑 ∈ 𝐻01 (Ω
푠 ) ,
to obtain a pair (→󳨀𝑢 , 𝑃) as a weak limit of the Galerkin
→
󳨀
approximations ( 𝑢
푚 , 𝑃
푚 ) such that 1 1
⟨𝜕 𝑃, 𝜓⟩Ω𝑓 = − (∇
푥 𝑃, ∇
푥 𝜓)Ω𝑓 + 2 (𝑞0 , 𝜓)Ω𝑓 ,
∞
푓 𝑐2
푡
푡 𝑐
𝑃 ∈ 𝐿 (0, 𝑇; 𝑉 ) ,
∀𝜓 ∈ 𝐻01 (Ω
푓 ) .
𝜕
푡 𝑃 ∈ 𝐿∞ (0, 𝑇; 𝐻
푓 ) ,
󵄨
𝜕
푡 𝑃󵄨󵄨󵄨Γ4 ∈ 𝐿2 (0, 𝑇; 𝐿2 (Γ4 )) , (31) From these two relations, it follows 𝜕
푡
푡→
󳨀𝑢 ∈
∞ −1
푠 3
𝐿 (0, 𝑇; 𝐻 (Ω ) ) and 𝜕
푡
푡 𝑃 ∈ 𝐿 (0, 𝑇; 𝐻−1 (Ω
푓 ).
∞
→
󳨀
𝑢 ∈ 𝐿∞ (0, 𝑇; 𝑉
푠 ) , By Lemma 1 and the well-known density result

푑
𝐿2 (Ω) 󳨅→ 𝐻−1 (Ω), we obtain the following.
𝜕
푡→
󳨀
𝑢 ∈ 𝐿∞ (0, 𝑇; 𝐻
푠 ) ,
together with the identities Lemma 5. → 󳨀
𝑢 ∈ 𝐶
푤 ([0, 𝑇]; 𝑉
푠 ), 𝜕
푡→
󳨀
𝑢 ∈ 𝐶
푤 ([0, 𝑇]; 𝐻
푠 ) and 𝑃 ∈

푇 𝐶
푤 ([0, 𝑇]; 𝑉 ), 𝜕
푡 𝑃 ∈ 𝐶
푤 ([0, 𝑇]; 𝐻
푓 ).

푓

− ∫ [󰜚
푠 (𝜕
푡→
󳨀 𝜑 )Ω𝑠 + 𝑐
푎 (→
𝑢 (𝜏) ,→
󳨀 󳨀
𝑢 (𝜏) ,→
󳨀
𝜑 )Ω𝑠
0
3.6. Strong Continuity, Energy Equality and Uniqueness. Let
+ (𝑃 (𝜏) ,→
󳨀
𝜑 ⋅→
󳨀𝑛
푠 ) ] 𝜙 ̇ (𝜏) 𝑑𝜏
Γ
(→
󳨀𝑢 , 𝑃) have the weak continuity properties established in the
1
previous section and consider the energy function defined by

푇
+ 2𝜇 ∫ [(𝜀 (→
󳨀
𝑢 ) (𝜏) , 𝜀 (→
󳨀
𝜑 ))Ω𝑠
1 󵄩 󵄩2
E (→
󳨀
0 (32a) 𝑢 , 𝑃) (𝑡) fl
푓 󵄩󵄩󵄩∇
푥 𝑃 (𝑡)󵄩󵄩󵄩2,Ω𝑓
2󰜚
+ 𝜆 (∇
푥 ⋅ →
󳨀
𝑢 (𝜏) , ∇
푥 ⋅ →
󳨀
𝜑 )Ω𝑠 ] 𝜙 (𝜏) 𝑑𝜏
1 󵄩󵄩 󵄩2
󳨀V ,→
= [󰜚
푠 (→ 󳨀 →
󳨀 → 󳨀 + 󵄩𝜕 𝑃 (𝑡)󵄩󵄩󵄩2,Ω𝑓
0 𝜑 )Ω𝑠 + 𝑐
푎 ( 𝑢 0 , 𝜑 )Ω𝑠 ] 𝜙 (0) 2󰜚
푓 𝑐2 󵄩
푡

푇 →󳨀 1 󵄩󵄩 󵄩2
+ 󰜚
푠 ∫ (→𝑔 + 𝑎
푠 (𝜏) ,→
󳨀 󳨀
𝜑 ) 𝜙 (𝜏) 𝑑𝜏, + 󵄩𝜕 𝑃 (𝑡)󵄩󵄩󵄩2,Γ3
0 Ω𝑠 2󰜚
푓 𝑔 󵄩
푡
8 Advances in Mathematical Physics

󵄩 󳨀 󵄩2 󰜚
푠 󵄩 󳨀 󵄩󵄩2 1 1 1
+ 𝜇 󵄩󵄩󵄩󵄩𝜀 (→
𝑢 ) (𝑡)󵄩󵄩󵄩󵄩2,Ω𝑠 + 󵄩󵄩󵄩󵄩𝜕
푡→
𝑢 (𝑡)󵄩󵄩󵄩2,Ω𝑠 (𝜕 𝑃
훿 , 𝜓)Ω𝑓 + (𝜕
푡
푡 𝑃
훿 , 𝜓)Γ + (𝜕
푡 𝑃
훿 , 𝜓)Γ
2 𝑐2
푡
푡 𝑔 3 𝑐 4

𝜆 󵄩󵄩 󵄩2
𝑢 ⋅→
+ 󰜚
푓 (𝜕
푡→
󳨀
훿 󳨀
푠
+ 󵄩󵄩∇ ⋅ →
󳨀
𝑢 (𝑡)󵄩󵄩󵄩󵄩2,Ω𝑠 𝑛 , 𝜓)
2 󵄩
푥 Γ1

(a.a. 𝑡 ∈ [0, 𝑇]) . 1
훿 1

+ (∇
푥 𝑃
훿 , ∇
푥 𝜓)Ω𝑓 = (𝑞 , 𝜓)Ω𝑓 + (𝑞0
훿 , 𝜓)Γ (37b)
(34) 𝑐2 0 𝑔 3

1
훿 󳨀
푠
훿
󳨀V ⋅ →
+ (𝑝0 , 𝜓)Γ + 󰜚
푓 ((→ 0 𝑛 ) , 𝜓) ,
We shall prove that E(→ 󳨀𝑢 , 𝑃) ∈ 𝐶([0, 𝑇]) and use this result to 𝑐 4 Γ1
prove the strong continuity of (→ 󳨀𝑢 , 𝑃). To accomplish this, we
will resort to the following intermediate result. ∀𝜓 ∈ 𝑉
푓 .

Lemma 6. If (→ 󳨀
Now, since 𝜕
푡→
󳨀
훿
𝑢 , 𝑃) is the weak solution obtained in the 𝑢 ∈ 𝐶0∞ (𝑡0 , 𝑇∗ ; 𝑉
푠 ) and 𝜕
푡 𝑃
훿 ∈ 𝐶0∞ (𝑡0 , 𝑇∗ ; 𝑉
푓 ),
we can take →
󳨀
𝜑 = 𝜕→ 󳨀
previous section then
훿
𝑢 (𝑡) and 𝜓 = 𝜕 𝑃
훿 (𝑡) in (37a)-(37b), thus

푡
푡
obtaining
𝑑 1 󵄩 󵄩2 󵄩 󳨀 󵄩󵄩2
E (→
󳨀 𝑢 󵄩󵄩󵄩2,Ω𝑠 = 󰜚
푠 (→
𝑢 , 𝑃) +
푓 󵄩󵄩󵄩𝜕
푡 𝑃󵄩󵄩󵄩2,Γ4 + 𝑐
푎 󵄩󵄩󵄩󵄩𝜕
푡→ 󳨀
𝑔 󵄩󵄩 󳨀
훿 󵄩󵄩2
𝑑𝑡 󰜚 𝑐 𝑑 1 󵄩 󵄩2
E (→
󳨀
𝑢 , 𝑃
훿 ) +
푓 󵄩󵄩󵄩󵄩𝜕
푡 𝑃
훿 󵄩󵄩󵄩󵄩2,Γ + 𝑐
푎 󵄩󵄩󵄩𝜕
푡→

훿
𝑢 󵄩󵄩󵄩 𝑠
𝑑𝑡 󰜚 𝑐 4 󵄩 󵄩2,Ω
→󳨀 󳨀 1 1 1
+ 𝑎
푠 , 𝜕
푡→
𝑢) + [ (𝑞 , 𝜕 𝑃) 𝑓 + (𝑞0 , 𝜕
푡 𝑃)Γ3 (35)
󰜚
푓 𝑐2 0
푡 Ω 𝑔
훿 →
󳨀
훿 󳨀
훿 1 1
훿
= 󰜚
푠 (→
󳨀
𝑔 + 𝑎
푠 , 𝜕
푡→
Ω𝑠
𝑢 ) + [ (𝑞 , 𝜕 𝑃
훿 )Ω𝑓
󰜚
푓 𝑐2 0
푡
1 󳨀V ⋅ →
󳨀
푠 Ω𝑠
+ (𝑝 , 𝜕 𝑃) ] + (→ 0 𝑛 , 𝜕
푡 𝑃)Γ
(38)
𝑐 0
푡 Γ4 1 1 1
+ (𝑞0
훿 , 𝜕
푡 𝑃
훿 )Γ + (𝑝0
훿 , 𝜕
푡 𝑃
훿 )Γ ]
𝑔 3 𝑐 4

and E(→ 󳨀
𝑢 , 𝑃) : [0, 𝑇] 󳨀→ R is an absolutely continuous
+ ((→ 󳨀
푠
훿
󳨀V ⋅ →
훿
function. 0 𝑛 ) , 𝜕
푡 𝑃 ) .
Γ1

Proof. Once (35) is proved, since

We let 𝛿 󳨀→ 0 and use the convergence properties (3)-(4)
to obtain (35) in D耠 (𝑡0 , 𝑇∗ ). Since 𝑡0 , 𝑇∗ are arbitrary, we
󵄩󵄩 󵄩󵄩2 󵄩 󳨀 󵄩󵄩2
󵄩󵄩𝜕
푡 𝑃󵄩󵄩2,Γ4 + 󵄩󵄩󵄩󵄩𝜕
푡→
𝑢 󵄩󵄩󵄩2,Ω𝑠 ∈ 𝐿1 (0, 𝑇) , conclude that (35) holds a.e. in (0, 𝑇).

→󳨀 󳨀 1 1 The result of Lemma 5 can be improved to

󰜚
푠 (→
󳨀
𝑔 + 𝑎
푠 , 𝜕
푡→
𝑢) + [ (𝑞 , 𝜕 𝑃) 𝑓
Ω𝑠 󰜚
푓 𝑐2 0
푡 Ω Lemma 7. The pair (→
󳨀
𝑢 , 𝑃) constructed in the previous subsec-
(36)
1 1 tion satisfies
+ (𝑞 , 𝜕 𝑃) + (𝑝 , 𝜕 𝑃) ] + (→
󳨀V
𝑔 0
푡 Γ3 𝑐 0
푡 Γ4 0
𝑃 ∈ 𝐶 ([0, 𝑇] ; 𝑉
푓 ) ∩ 𝐶1 ([0, 𝑇] ; 𝐻
푓 ) ,
⋅→
󳨀𝑛
푠 , 𝜕 𝑃) ∈ 𝐿1 (0, 𝑇) ,

푡 Γ 1 𝑃|Γ3 ∈ 𝐶1 ([0, 𝑇] ; 𝐿2 (Γ3 )) , (39)

we immediately conclude that E(→ 󳨀𝑢 , 𝑃) is an absolutely →

󳨀𝑢 ∈ 𝐶 ([0, 𝑇] ; 𝑉
푠 ) ∩ 𝐶1 ([0, 𝑇] ; 𝐻
푠 )
continuous function.
To prove (35), we use the mollification method, as and the energy equality (18).
described in Section 3.1. Consider 0 < 𝑡0 < 𝑇∗ < 𝑇, some
fixed 𝛿0 > 0 with 𝛿0 < 𝑡0 , 𝛿0 < 𝑇 − 𝑇∗ , and let 0 < 𝛿 < 𝛿0 . Proof. Using the weak continuity properties of (→
󳨀
𝑢 , 𝑃) and the
Inserting the special test functions 𝜙
훿 and 𝜒
훿 , with 𝜙, 𝜒 ∈ →
󳨀
훿
훿
continuity of E( 𝑢 , 𝑃 ) from Lemma 6, we find that
𝐶0∞ (𝑡0 , 𝑇∗ ), in (32a)-(32b) and a simple calculation yields
1 󵄩󵄩 󵄩2

푓
󵄩󵄩∇
푥 𝑃 (𝑡) − ∇
푥 𝑃 (𝑡0 )󵄩󵄩󵄩2,Ω𝑓
2󰜚
𝑢 ,→
󰜚
푠 (𝜕
푡
푡→
󳨀
훿 󳨀
𝑢 ,→
+ 𝑐
푎 (𝜕
푡→
󳨀
훿 󳨀
𝜑) 𝜑) 󳨀𝑛
푠 ,→
− (𝜕
푡 𝑃
훿→ 󳨀
𝜑 )Γ
Ω𝑠 Ω𝑠 1
1 󵄩󵄩 󵄩2
+ 󵄩𝜕 𝑃 (𝑡) − 𝜕
푡 𝑃 (𝑡0 )󵄩󵄩󵄩2,Ω𝑓
+ 2𝜇 (𝜀 ( 𝑢 ) (𝜏) , 𝜀 (→
→
󳨀
훿 󳨀
𝜑 )) 2󰜚
푓 𝑐2 󵄩
푡
Ω𝑠
(37a) 1 󵄩󵄩 󵄩2

훿 →
󳨀
훿 󳨀 + 󵄩𝜕 𝑃 (𝑡) − 𝜕
푡 𝑃 (𝑡0 )󵄩󵄩󵄩2,Γ3
+ 𝜆 (∇
푥 ⋅ →
󳨀
𝑢 , ∇
푥 ⋅ →
󳨀 = 󰜚 (→
󳨀
𝑔 + 𝑎
푠 ,→ 2󰜚
푓 𝑔 󵄩
푡

훿
푠
𝜑) 𝜑) ,
Ω𝑠
Ω𝑠
󰜚
푠 󵄩󵄩 → 󵄩󵄩2
∀→
󳨀 + 󵄩󵄩𝜕
푡 󳨀 →󳨀 󵄩
𝜑 ∈ 𝑉
푠 , 2 󵄩 𝑢 (𝑡) − 𝜕
푡 𝑢 (𝑡0 )󵄩󵄩2,Ω𝑠
Advances in Mathematical Physics 9

󵄩 󳨀 󵄩2
+ 𝜇 󵄩󵄩󵄩󵄩𝜀 (→
𝑢 ) (𝑡) − 𝜀 (→
󳨀
𝑢 ) (𝑡0 )󵄩󵄩󵄩󵄩2,Ω𝑠 is nonsingular, as already seen in the previous section. It will
be convenient to introduce additional notation
𝜆 󵄩󵄩 󵄩2
+ 󵄩󵄩∇
푥 ⋅ →
󳨀
𝑢 (𝑡) − ∇
푥 ⋅ →
󳨀
𝑢 (𝑡0 )󵄩󵄩󵄩󵄩2,Ω𝑠 = E (→
󳨀
𝑢 , 𝑃) (𝑡) C
푠 0
2 󵄩 C fl [ ],
0 C
푓
1
+ E (→
󳨀
𝑢 , 𝑃) (𝑡0 ) −
푓 (∇
푥 𝑃 (𝑡) , ∇
푥 𝑃 (𝑡0 ))Ω𝑓
󰜚 K −Q
K fl [ ],
1 0 H
+
푓 2 (𝜕
푡 𝑃 (𝑡) , 𝜕
푡 𝑃 (𝑡0 ))Ω𝑓 (43)
󰜚 𝑐 u∗
y fl [ ],
1 p∗
+ (𝜕 𝑃 (𝑡) , 𝜕
푡 𝑃 (𝑡0 ))Γ3
󰜚
푓 𝑔
푡
f
+ 󰜚
푠 (𝜕
푡→
󳨀
𝑢 (𝑡) , 𝜕
푡→
󳨀
𝑢 (𝑡0 ))Ω𝑠 f fl [ ]
0
+ 2𝜇 (𝜀 (→
󳨀
𝑢 ) (𝑡) , 𝜀 (→
󳨀
𝑢 ) (𝑡0 ))Ω𝑠 so that (41) can be written as
Mÿ + Cẏ + Ky = f.
+ 𝜆 (∇
푥 ⋅ →
󳨀
𝑢 (𝑡) , ∇
푥 ⋅ →
󳨀 (44)
𝑢 (𝑡0 ))Ω𝑠 󳨀→ 0 as 𝑡 󳨀→ 𝑡0 .
(40) 4.2. State-Space Representation of the Problem. We proceed
with the so-called state-space representation of Problem (41)
with subsequent diagonalization, so we end up with a system
Uniqueness for the constructed (→ 󳨀𝑢 , 𝑃) is consequence of of independent first-order linear differential equations for the
the linearity of the problem (14) and (17) and the validity of discretized problem. As already mentioned in the Introduc-
the energy equality (18). tion, this is preferred over a direct numerical time stepping
Concerning uniqueness for the original (→ 󳨀𝑢 , 𝑝), it suffices scheme applied to (44) because it allows for a comparison
to show that the only weak solution (in the sense of the with the measured natural frequencies for different values of
definition that supports Theorem 3) with null data is (→ 󳨀 water level.
𝑢 , 𝑝) =
→
󳨀 Starting from (44) and following [21, 22], a new variable
( 0 , 0). Again, the idea is to introduce a primitive 𝑃 ∈ z = ẏ is introduced so that (44) is equivalent to
D耠 (0, 𝑇; 𝑉
푓 ) for the original pressure 𝑝 = 𝜕
푡 𝑃. Since 𝑃 = 0
and another primitive of 𝑝 will differ from 𝑃 by a constant, it Mż + Cẏ + Ky = f
(45)
follows that 𝑝 = 0. ẏ = z

4. Discretization of the Equations which in turn can be written as

C M ẏ K 0 y f
4.1. Semidiscretization in Space. Once a mesh and the corre- [ ][ ] + [ ][ ] = [ ]. (46)
−I 0 ż 0 I z 0
sponding basis functions, 𝜓
푖 : Ω
푓 󳨀→ R and → 󳨀
𝜑
푖 : Ω
푠 󳨀→
R3 , have been generated, → 󳨀𝑢 and 𝑝 are approximated by The matrix
𝑢 (𝑥, 𝑡) ≈ 𝑢 (𝑡, 𝑥) := ∑
푖=1 𝑢
푖 (𝑡)󳨀
→
󳨀 →󳨀 ∗ 3
푁𝑠 ∗ →
𝜑
푖 (𝑥), 𝑝(𝑥, 𝑡) ≈ 𝑝∗ (𝑡, 𝑥) fl C
푠 0
M 0
𝑓
∑
푁 ∗ [ ]

푖=1 𝑝
푖 (𝑡)𝜓
푖 (𝑥), and Problems (11a)-(11b) are approximated

푓
푓 ⊤
C M [ 0 C 󰜚 Q S]
by (see (26)) B fl [ [
]=[ ] (47)
−I 0 ]
[−I 0 0 0]
M 0 u∗̈ C
푠 0 u∗̇
[ ][ ]+[ ] [ ] [ 0 −I 0 0]

푓
󰜚 Q ⊤
S p∗̈ 0 C
푓 p∗̇ is nonsingular because det(B) = det(M) ≠ 0. Defining
(41)
∗
K −Q u f K −Q 0 0
+[ ] [ ∗] = [ ] . [ ]
0 H p 0 K 0 [ 0 H 0 0]
A fl [ ]=[
[ 0 0 I 0]
]
Matrices M, C
푠 , and K are (3𝑁
푠 ) × (3𝑁
푠 ) dimensional, S, 0 I [ ]
C
푓 , and H are 𝑁
푓 × 𝑁
푓 dimensional, Q is (3𝑁
푠 ) × 𝑁
푓 [0 0 0 I]
dimensional, and f and s are, respectively, 𝑁
푠 - and 𝑁
푓 - (48)
dimensional vectors. The matrices M and S are symmetric, f
positive definite and the matrix [ ]
f [0]
F fl [ ] = [ ]
[0]
M 0 0 [ ]
M fl [
푓 ⊤ ] (42)
󰜚 Q S [0]
10
and putting k fl u,̇ q fl p,̇ x fl [y z]⊤ = [u p k q]⊤ , we can
write (41) as
u̇ u
[ ]
[ ]
[ṗ ] [p ]
B[ ] [ ]
[ k̇ ] = A [ k ] + F
[ ] [ ]
[q̇ ] [q]
From now on, we consider the system of ODEs
Bẋ (𝑡) = Ax (𝑡) + F (𝑡) ,
x (0) = x0
with the obvious identifications as concerns x0 . In order to
solve (50), the solution of the homogeneous equation Bx(𝑡) ̇ =
Ax(𝑡), 𝑡 ∈]0, 𝑇] can be sought in the form x(𝑡) = b𝑒
휆
푡 , yielding
the generalized eigenvalue problem Ab − 𝜆Bb = 0. Let 𝑚 fl
dim(A) = dim(B) = 2(3𝑁
푠 + 𝑁
푓 ). If all the eigenvalues
are distinct, the matrix pencil (A, B) is diagonalizable; that
is, there exist 𝑚 × 𝑚 matrices W and V such that
W∗ AV = Ω
퐴 ≡ diag (𝛼1 , . . . , 𝛼
푚 ) ,
W∗ BV = Ω
퐵 ≡ diag (𝛽1 , . . . , 𝛽
푚 ) .
The matrices W and V contain the left and right eigenvectors
of (A, B), that is 𝛼
푖 = w
푖∗ Ak
푖 and 𝛽
푖 = w
푖∗ Bk
푖 . In
practice, since the response of the mechanical system is
typically dominated by a relatively small number of the lowest
modes, it may not be necessary to compute the complete
eigensystem of (A, B). Other approaches [12, 13], based on
pseudo-symmetric techniques, can be used for computing the
mode shapes and natural frequencies of this fluid-structure
model.
Using the transformation for modal coordinates x = Vz
and multiplying both sides of (50) by W∗ , yields the modal
system
Ω
퐵 ż = Ω
퐴z + C in ]0, 𝑇]
z (0) = z0
with C = W∗ F and z0 solution of the linear system Vz0 = x0 .
From (51) we get z0 = Ω−1 ∗ −1 ∗

퐴 W Ax0 or z0 = Ω
퐵 W Bx0
so that the calculation of z0 only requires direct matrix
multiplications. Now we have a system of 𝑚 independent
first-order linear ordinary differential equations, which is
easy to solve. Each equation of the system is of the form
𝛽
푖 𝑧̇
푖 (𝑡) = 𝛼
푖 𝑧
푖 (𝑡) + 𝑐
푖 (𝑡) ,
𝑡 ∈ ]0, 𝑇] , with 𝑧
푖 (0) = 𝑧
푖,0
where 𝑐
푖 is the 𝑖-th component of the vector C, which in
practice is only known in a set of discrete points.
4.3. Time Discretization of the Equations. For the discretiza-
tion in time of the diagonalized system (52), we take the
solution of each equation (53) in its integral form

푡

휆 𝑖
푡
휆 𝑖
푡
𝑧
푖 (𝑡) = 𝑒 𝑧
푖 (0) + 𝑒 ∫ 𝑒
0 system was developed, using meshes composed of serendipity
Advances in Mathematical Physics
where 𝜆
푖 fl 𝛼
푖 /𝛽
푖 are the state eigenvalues and 𝑐̃
푖 = 𝑐
푖 /𝛽
푖 ,
𝑖 = 1 : 𝑚. Each function 𝑐
푖 is known in discrete instants of
time 0 < 𝑡1 < ⋅ ⋅ ⋅ < 𝑡
푛−1 < 𝑡
푛 = 𝑇 with Δ𝑡 fl 𝑡
푘+1 − 𝑡
푘 ,
𝑘 = 1 : 𝑛 − 1, and therefore will approximate 𝑐
푖 by a constant
in each interval [𝑡
푘 , 𝑡
푘+1 ]:
(49)
𝑐
푖 (𝑡) ≃ 𝑐
푖 (𝑡
푘 ) (𝑡 ∈ [𝑡
푘 , 𝑡
푘+1 ]) . (55)
From the relation

푡𝑘+1
𝑡 ∈ ]0, 𝑇] 𝑧
푖 (𝑡
푘+1 ) = 𝑒
휆 𝑖 Δ
푡 𝑧
푖 (𝑡
푘 ) + 𝑒
휆 𝑖
푡𝑘+1 ∫ 𝑒−
휆 𝑖
휏 𝑐̃
푖 (𝜏) 𝑑𝜏 (56)
(50)
푡𝑘
and the approximation (55), we get

푡𝑘+1
𝑧
푖 (𝑡
푘+1 ) ≃ 𝑒
휆 𝑖 Δ
푡 𝑧
푖 (𝑡
푘 ) + 𝑒
휆 𝑖
푡𝑘+1 𝑐̃ (𝑡
푘 ) ∫ 𝑒−
휆 𝑖
휏 𝑑𝜏

푡𝑘
(57)

휆 𝑖 Δ
푡 𝑐 (𝑡
푘 )
휆 𝑖 Δ
푡
=𝑒 𝑧
푖 (𝑡
푘 ) + (𝑒 − 1) .
𝜆
푖
In vector form, the algorithm for computing z reads
(51)
z (𝑡
푘+1 ) ≃ 𝑒ΛΔ
푡 z (𝑡
푘 ) + Λ−1 C (𝑡
푘 ) (𝑒ΛΔ
푡 − I) (58)
with Λ = diag(𝜆 1 , . . . , 𝜆
푚 ). Once z is known,
x (𝑡
푘 ) = Vz (𝑡
푘 ) , 𝑘=1:𝑛 (59)
will provide the solution for the original variable.
5. Numerical Simulations
The objective of this section is to illustrate how the
displacement-pressure model can be used for studying the
dynamic behavior of a dam. For this purpose, several
numerical simulations of Cabril dam (shown in Figure 1(a))
were performed, considering different water levels. This is
(52) a common procedure [23–25] as the natural frequencies of
the system depend on the reservoir water level. Cabril dam,
built in 1954, is the highest Portuguese arch dam. It presents
some cracking near the crest which was considered in the
computational geometry of the model (these are horizontal
cracks, developed since the first filling of the reservoir due to
a particular design shape of the dam crest). The numerical
results will be compared with available experimental data
from a permanent monitoring vibration system installed by
LNEC in 2008. More precisely, the computed main natural
(53) frequencies (module of the state eigenvalues 𝜆
푖 introduced
in Section 4) and the corresponding modal configurations
(obtained from the components u
푖 of the state eigenvectors
k
푖 ) will be compared with experimental data identified
from acceleration records. A fast Fourier transform (FFT)
technique was used in [26] to obtain the natural frequencies
from these acceleration records.
5.1. Material Parameters and Finite Element Modeling of Cabril
−
휆 𝑖
휏 Arch Dam. A 3D finite element model of the dam-water
𝑐̃
푖 (𝜏) 𝑑𝜏 (54)
Advances in Mathematical Physics
(a) Located at Zêzere river, 132 m high
Figure 1: Cabril dam and computational domain.
isoparametric finite elements of cubic type, with 20 nodes in
each element (see [27]) in the structure and fluid domains.
Concerning the concrete and foundations parameters,
Young’s modulus is 𝐸 = 32.5GPa, the Poisson’s ratio is ] =
0.2, and the concrete specific mass is 2.45tn/m3 . Although
the damping effect on the structure is well defined through
the drag operator C
푠 , the so-called Rayleigh’s hypothesis was
used to model material damping. This assumes C
푠 as a linear
combination of the mass and stiffness matrices, C
푠 = 𝛼M +
𝛽K, and we took, as in previous studies [26, 28], 𝛼 = 0.5 and
𝛽 = 0.0025.
The finite element algorithm described in Section 4, based
on the geometry shown in Figure 5(b), was implemented
in a Matlab program which allows to compute the modal
characteristics of the whole system (static analysis) and
the corresponding displacements and stress histories under
prescribed forces (dynamic analysis), in particular, under
seismic loads.
5.2. Comparison between Numerical and Experimental Results
of Cabril Dam. Natural Frequencies and Modal Configura-
tions. We recall that the approximation process to simulate
the behavior over time of the structure and the fluid ended
up in the form
𝑐
푖 (𝑡)
𝑧̇
푖 (𝑡) = 𝜆
푖 𝑧
푖 (𝑡) + ,
𝛽
푖
𝑡 ∈ ]0, 𝑇] , with 𝑧
푖 (0) = 𝑧
푖,0 , 𝑖 = 1 : 𝑚,
the original variable x being recovered from a linear combina-
tion of the right eigenvectors of the state-space pencil (A, B)

푚
x = Vz = ∑𝑧
푖 k
푖 ,

푖=1
To each vibration mode, i.e., each eigenvector k
푖 , 𝑖 ∈
{1, . . . , 𝑚}, there corresponds an eigenmode of the state-space
system which is characterized by the way of responding to
external forces, i.e., by vibrating with a certain frequency
𝑓
푖 = |𝜆
푖 |, called a natural frequency of the system, and a certain
relative damping coefficient 𝜉
푖 = −Re(𝜆
푖 )/𝑓
푖 .
Mode shapes and natural frequencies for full reservoir
are presented in Figure 2. In the simulations, 202 serendipity
11
(b) Finite-element meshes
finite elements with 20 nodes each were used at the dam
body and foundations and 424 elements of the same type
were defined at the fluid domain. Here only the displacement
of the structure is shown for the first 8 modes. Concerning
the graphical representation of the computed mode shapes,
it should be noted that each mode 𝑖 corresponds to an
oscillatory motion at the frequency |𝜆
푖 | which is represented
by waves at the nodal points (three waves at each point)
whose amplitude and phase are calculated from the complex
components of the eigenvector u
푖 .
In Figure 3 we present a comparison between numerical
and experimental values of the main natural frequencies of
the dam-reservoir system for different values of the water
level. The experimental values of the natural frequencies
were obtained during the year 2014 by the analysis of several
acceleration records from a permanent vibration monitoring
system installed in Cabril dam [26]. The monitoring system
includes 16 acceleration sensors (uniaxial sensors for measur-
ing radial accelerations) located in the dam body, near the top:
9 at the crest gallery and 7 in a gallery under the cracked zone.
Acceleration files were recorded every hour, during several
months, in order to experimentally characterize the influence
of the water level on the main natural frequencies of the
system. The experimental natural frequencies were identified
as abscises of spectral peaks obtained using a FFT technique
to analyze the time series of measured accelerations.
(60) The results presented in Figure 3 show a good agreement
between identified and computed natural frequencies for
different values of water level, meaning that the implemented
numerical model is reliable and well calibrated. Note that,
for each mode, the natural frequency increases as the water
level decreases, as usually calculated with classical approaches
based on the simplified model of associated water masses
(61) proposed by Westergaard [23] for the simulation of hydro-
dynamic pressure on the upstream dam face.
5.3. Simulation of Seismic Behavior of Cabril Dam Using
the Calibrated Model. We used the accelerogram of the 921
Earthquake, also known as Chichi Earthquake, as input data
for ground motions →󳨀𝑎
푠 . This accelerogram presents a peak
ground acceleration of 0.1g (=1 m/ s2 ), as shown in Figure 4.
Such intensity corresponds to a strong perceived shaking
12 Advances in Mathematical Physics

Mode 1 ＠1 = 2.08 Hz Mode 2 ＠2 = 2.23 Hz

Mode 3 ＠3 = 2.98 Hz Mode 4 ＠4 = 3.36 Hz

Mode 5 ＠5 = 3.70 Hz Mode 6 ＠6 = 4.11 Hz

Mode 7 ＠7 = 4.22 Hz Mode 8 ＠8 = 5.00 Hz

Figure 2: Main natural frequencies and mode shapes, computed for full reservoir.

1ＭＮ mode 2Ｈ＞ mode 3Ｌ＞ mode 4ＮＢ mode 5ＮＢ mode

296 m
(m) 297
Water level (m)

290 Experimental

Numerical FEM
266 m 270

1 2 3 4
Identified and computed natural frequencies (Hz)
250

Figure 3: Variation of the main natural frequencies with the water level. Comparison between numerical and experimental values.

1
Accel. (m/Ｍ2 )

−1
0 10 20 30
Time (s)

Figure 4: Seismic accelerogram: magnitude of →

󳨀𝑎 as a function of time.
푠
Advances in Mathematical Physics
15
Disp. (mm)
0
−15
0 10 20
Time (s)
(a) Radial displacement history at the top (crest central point)
Figure 5: Computed seismic response of Cabril dam for a peak ground acceleration of 0.1g: displacement field.
when the earthquake occurs [29] and is in agreement with the
characteristics of the seismic zone where the dam is located
[30].
The corresponding time response was computed consid-
ering the diagonalization technique for the time integration
described in Section 4, which lead to algorithms (60)-(61). A
reduced number of eigenvectors, more precisely 160 vibration
modes, were fixed after several numerical tests have been
made with a larger number of modes and response values
with the same precision were obtained. Figure 5 shows the
radial displacement response over time at the top of the
central cantilever, with a maximum displacement at the top
of the central section of about 15 mm.
The maximum stress values occur at the crest and are of
about 1.6 MPa (tension and compression). In conclusion,
Cabril dam withstands against the applied earthquake.
6. Conclusions
We studied a fluid-structure interaction model in pressure-
displacement formulation which was derived from the Euler
and Navier equations, as explained in [3].
The difficulty in the theoretical study of the variational
formulation of the coupled equations was caused by the
unbalanced order of the time derivatives of the pressure
and the structure’s displacement in the interface conditions.
We introduced a pressure potential so that the classical
techniques based on Galerkin approximations for second-
order hyperbolic equations could be applied to investigate the
well posedness of the fluid-structure interaction problem.
Concerning the discretization of the problem, the appli-
cation of the Finite Element Method (FEM) to the space
variables was followed by a discretization in time associated
with a state space approach appropriate for the dynamic
analysis of arch dams. This numerical formulation was tested
through a comparative study of the dynamic behavior of
Cabril dam, namely, the simulation of dam’s natural frequen-
cies and vibration mode shapes, for different reservoir water
levels. The good agreement observed between modal iden-
tification outputs (from real measurements) and numerical
results shows the reliability of the state space formulation
implemented in the developed 3-D FEM program.
13
(mm)
15.81
12.65
9.49
6.33
3.16
30 0
(b) Displacement field at the instant when the maximum displace-
ment towards upstream occurs (approx. 15 mm)
Data Availability
The data used to support the findings of this study are
available from the corresponding author upon request.
Conflicts of Interest
The authors declare that they have no conflicts of interest.
Acknowledgments
Ana L. Silvestre would like to acknowledge the partial
financial support received from CEMAT through Project
FCT-UID/Multi/04621/2013.
References
[1] B. Sevim, A. Bayraktar, and A. C. Altunişik, “Finite element
model calibration of berke arch dam using operational modal
testing,” Journal of Vibration and Control, vol. 17, no. 7, pp. 1065–
1079, 2011.
[2] B. Sevim, A. C. Altunişik, and A. Bayraktar, “Earthquake
behavior of berke arch dam using ambient vibration test results,”
Journal of Performance of Constructed Facilities, vol. 26, no. 6,
pp. 780–792, 2012.
[3] O. C. Zienkiewicz, R. L. Taylor, and J. Z. Zhu, The Finite Ele-
ment Method: Its Basis & Fundamentals, Elsevier, Butterworth-
Heinemann, 6th edition, 2006.
[4] T. Nguyen-Thoi, P. Phung-Van, T. Rabczuk, H. Nguyen-Xuan,
and C. Le-Van, “An application of the ES-FEM in solid domain
for dynamic analysis of 2D fluid-solid interaction problems,”
International Journal of Computational Methods, vol. 10, no. 1,
2013.
[5] T. Nguyen-Thoi, P. Phung-Van, S. Nguyen-Hoang, and Q. Lieu-
Xuan, “A coupled alpha-FEM for dynamic analyses of 2D
fluid-solid interaction problems,” Journal of Computational and
Applied Mathematics, vol. 271, pp. 130–149, 2014.
[6] T. Nguyen-Thoi, P. Phung-Van, S. Nguyen-Hoang, and Q.
Lieu-Xuan, “A Smoothed Coupled NS/nES-FEM for Dynamic
Analysis of 2D Fluid–Solid Interaction Problems,” Applied
Mathematics and Computation, vol. 232, pp. 324–346, 2014.
[7] A. Ghobarah, “Performance-based design in earthquake engi-
neering: state of development,” Engineering Structures, vol. 23,
no. 8, pp. 878–884, 2001.
14 Advances in Mathematical Physics

[8] S. Oliveira, A.-M. Toader, and P. Vieira, “Damage identification [26] S. Oliveira, M. Espada, and M. Câmara, “Long term dynamic
in a concrete dam by fitting measured modal parameters,” monitoring of arch dams. The case of Cabril dam,” in In Lisboa:
Nonlinear Analysis: Real World Applications, vol. 13, no. 6, pp. 15 WCEE, In Lisboa: 15 WCEE, Portugal, 2012.
2888–2899, 2012. [27] D. N. Arnold and G. Awanou, “The serendipity family of
[9] L. Tanchev, Dams and Appurtenant Hydraulic Structures. Second finite elements,” Foundations of Computational Mathematics.
Edition, CRC Press, 2nd, 2014. The Journal of the Society for the Foundations of Computational
[10] M. Wieland, “Features of seismic hazard in large dam projects Mathematics, vol. 11, no. 3, pp. 337–344, 2011.
and strong motion monitoring of large dams,” Frontiers of [28] S. Oliveira and P. Mendes, Development of a Cabril Dam Finite
Architecture and Civil Engineering in China, vol. 4, no. 1, pp. 56– Element Model for Dynamic Analysis Using Ambient Vibration
64, 2010. Tests Results, Springer, Netherlands, Dordrecht, 2006.
[11] B. Tiliouine and A. Seghir, “Fluid-structure models for dynamic [29] J. R. Murphy and L. J. O’Brien, “The correlation of peak ground
studies of dam-water systems,” in Proceedings of the Earthquake acceleration amplitude with seismic intensity and other physical
Engineering: Proceedings of the 11th European Conference, A. parameters,” Bulletin of the Seismological Society of America, vol.
Pecke and P. Bisch, Eds., Paris, 1998. 67, no. 3, pp. 877–915, 1977.
[12] V. Lotfi, “Dynamic analysis of concrete gravity dams by decou- [30] V. Silva, H. Crowley, H. Varum, and R. Pinho, “Seismic risk
pled modal approach in the time domain,” in Proceedings of the assessment for mainland Portugal,” Bulletin of Earthquake
13th World Conference on Earthquake Engineering, Vancouver, Engineering, vol. 13, no. 2, pp. 429–457, 2014.
Vancouver, 2004.
[13] S. A. Arjmandi and V. Lotfi, “Comparison of three efficient
methods for computing mode shapes of fluid-structure interac-
tion systems,” Arabian Journal for Science and Engineering, vol.
38, no. 4, pp. 787–803, 2013.
[14] G. C. Hsiao, “On the boundary-field equation methods for
fluid-structure interactions,” in Problems and methods in math-
ematical physics (Chemnitz, 1993), vol. 134 of Teubner-Texte
Math., pp. 79–88, Teubner, Stuttgart, 1994.
[15] G. C. Hsiao, R. E. Kleinman, and G. F. Roach, “Weak solutions of
fluid-solid interaction problems,” Mathematische Nachrichten,
vol. 218, pp. 139–163, 2000.
[16] G. C. Hsiao and N. Nigam, “A transmission problem for fluid-
structure interaction in the exterior of a thin domain,” Advances
in Differential Equations, vol. 8, no. 11, pp. 1281–1318, 2003.
[17] G. C. Hsiao, F.-J. Sayas, and R. J. Weinacht, “Time-dependent
fluid-structure interaction,” Mathematical Methods in the
Applied Sciences, vol. 40, no. 2, pp. 486–500, 2017.
[18] A. Pereira and G. Beer, “Fluid-structure interaction by a
Duhamel-BEM/FEM coupling,” in Recent advances in boundary
element methods, pp. 339–354, Springer, New York, NY, USA,
2009.
[19] M. Renardy and R. C. Rogers, An introduction to partial
differential equations, vol. 13 of Texts in Applied Mathematics,
Springer-Verlag, New York, Second edition, 2004.
[20] R. Dautray and J.-L. Lions, Analyse Mathématique et Calcul
Numérique, Tome 3, Masson, Paris, 1985.
[21] F. Tisseur and K. Meerbergen, “The quadratic eigenvalue
problem,” SIAM Review, vol. 43, no. 2, pp. 235–286, 2001.
[22] Y. Saad, Numerical Methods for Large Eigenvalue Problems,
Classics in Applied Mathematics, Society for Industrial and
Applied Mathematics (SIAM), Philadelphia, PA, USA, 2nd
edition, 2011.
[23] H. M. Westergaard, “Water pressure on dams during earth-
quakes,” Transactions of the American Society of Civil Engineers,
vol. 98, pp. 418–433, 1933.
[24] J. Proulx, P. Paultre, J. Rheault, and Y. Robert, “An experimental
investigation of water level effects on the dynamic behaviour
of a large arch dam,” Earthquake Engineering & Structural
Dynamics, vol. 30, no. 8, pp. 1147–1166, 2001.
[25] B. Sevim, A. Bayraktar, A. C. Altunişik, S. Adanur, and M.
Akköse, “Determination of water level effects on the dynamic
characteristics of a prototype arch dam model using ambient
vibration testing,” Experimental Techniques, vol. 36, no. 1, pp.
72–82, 2012.
Advances in Advances in Journal of The Scientific Journal of
Operations Research
Hindawi
Decision Sciences
Hindawi
Applied Mathematics
Hindawi
World Journal
Hindawi Publishing Corporation
Probability and Statistics
Hindawi
www.hindawi.com Volume 2018 www.hindawi.com Volume 2018 www.hindawi.com Volume 2018 http://www.hindawi.com
www.hindawi.com Volume 2018
2013 www.hindawi.com Volume 2018

International
Journal of
Mathematics and
Mathematical
Sciences

Journal of

Hindawi
Optimization
Hindawi
www.hindawi.com Volume 2018 www.hindawi.com Volume 2018

Submit your manuscripts at

www.hindawi.com

International Journal of
Engineering International Journal of
Mathematics
Hindawi
Analysis
Hindawi
www.hindawi.com Volume 2018 www.hindawi.com Volume 2018

Journal of Advances in Mathematical Problems International Journal of Discrete Dynamics in

Complex Analysis
Hindawi
Numerical Analysis
Hindawi
in Engineering
Hindawi
Differential Equations
Hindawi
Nature and Society
Hindawi
www.hindawi.com Volume 2018 www.hindawi.com Volume 2018 www.hindawi.com Volume 2018 www.hindawi.com Volume 2018 www.hindawi.com Volume 2018

International Journal of Journal of Journal of Abstract and Advances in

Stochastic Analysis
Hindawi
Mathematics
Hindawi
Function Spaces
Hindawi
Applied Analysis
Hindawi
Mathematical Physics
Hindawi
www.hindawi.com Volume 2018 www.hindawi.com Volume 2018 www.hindawi.com Volume 2018 www.hindawi.com Volume 2018 www.hindawi.com Volume 2018