Diffusive Terms Applied in Smoothed Particle Hydrodynamics Simulations of Incompressible and Isothermal Newtonian Fluid Flows

BMSE-D-21-00211_linear.
pdf
Feb 2, 2021
14855 words / 66288 characters
Diffusive Terms Applied in Smoothed Particle Hydrodynamics Simu…
Sources Overview
27%
OVERALL SIMILARITY
Guarnieri, Joseph W.. "Viral and Lipid Regulation of Hepatocyte Signaling Pathways", Drexel University, 2020
1
PUBLICATION
9%
periodicos.unb.br
2
INTERNET
2%
Carlos Alberto Dutra Fraga Filho. "Smoothed Particle Hydrodynamics", Springer Science and Business Media LLC, 2019
3
CROSSREF
2%
Carlos Alberto Dutra Fraga Filho. "A Lagrangian analysis of the gravity-inertial oil spreading on the calm sea using the reWective oil-water interfac…
4
CROSSREF
<1%
upcommons.upc.edu
5
INTERNET
<1%
Jon Orlando Cleary, Nishanth Sivarasan, Christian Burd, Steve EJ Connor. "Head and neck manifestations of granulomatosis with polyangiitis: a p…
6
CROSSREF
<1%
www.drjohnlatham.com
7
INTERNET
<1%
engineering.purdue.edu
8
INTERNET
<1%
Carlos Alberto Dutra Fraga Filho. "On the boundary conditions in Lagrangian particle methods and the physical foundations of continuum mecha…
9
CROSSREF
<1%
aip.scitation.org
10
INTERNET
<1%
www.scielo.br
11
INTERNET
<1%
Joachim C. Manning, Gabriel García Caballero, Anna‐Kristin Ludwig, Herbert Kaltner, Fred Sinowatz, Hans‐Joachim Gabius. "Glycobiology of deve…
12
CROSSREF
<1%
M. Antuono, A. Colagrossi, S. Marrone. "Numerical diffusive terms in weakly-compressible SPH schemes", Computer Physics Communications, 2…
13
CROSSREF
<1%
Gallmeyer, Thomas G.. "The Effects of Heat Treatment on Structure-Property Relationships of Additively Manufactured Inconel 718.", Colorado Sc…
14
PUBLICATION
<1%
Hossein Abdolrahimzadeh Fard, Salahaddin Mahmudi-Azer, Sepideh Seddbakht, Pooya Iranpour et al. "Evaluation of Chest CT scan as a screenin…
15
CROSSREF POSTED CONTENT
<1%
C. A. D. Fraga Filho, L. L. Schuina, B. S. Porto. "An Investigation into Neighbouring Search Techniques in Meshfree Particle Methods: An Evaluatio…
16
CROSSREF
<1%
ssl4799.websiteseguro.com
17
INTERNET
<1%
scholarcommons.usf.edu
18
INTERNET
<1%
hdl.handle.net
19
INTERNET
<1%
Wei, Wei, and Zhaolin Gu. "Electridcation of particulate entrained Wuid Wows—Mechanisms, applications, and numerical methodology", Physics R…
20
CROSSREF
<1%
C. A. D. Fraga Filho, J. T. A. Chacaltana, W. J. N. Pinto. "Meshless Lagrangian SPH method applied to isothermal lid-driven cavity Wow at low-Re n…
21
CROSSREF
<1%
journals.ametsoc.org
22
INTERNET
<1%
Chiara Biscarini. "Computational Wuid dynamics modelling of landslide generated water waves", Landslides, 01/23/2010
23
CROSSREF
<1%
Xiaofei Hou, Joaquim Rigola, Oriol Lehmkuhl, Assensi Oliva. "Simulation of the two-Wuid model on incompressible Wow with Fractional Step meth…
24
CROSSREF
<1%
Donatella Bruno, Francesca De Serio, Michele Mossa. "The FUNWAVE model application and its validation using laboratory data", Coastal Engine…
25
CROSSREF
<1%
Marcela Cruchaga. "A surface remeshing technique for a Lagrangian description of 3D two-Wuid Wow problems", International Journal for Numeri…
26
CROSSREF
<1%
Z. D. Skoula, A. G. L. Borthwick, C. I. Moutzouris. "Godunov-type solution of the shallow water equations on adaptive unstructured triangular grids…
27
CROSSREF
<1%
epdf.pub
28
INTERNET
<1%
Sun, Pengnan, Furen Ming, and Aman Zhang. "Numerical simulation of interactions between free surface and rigid body using a robust SPH meth…
29
CROSSREF
<1%
Junca, Eduardo, Thomaz Augusto Guisard Restivo, José Roberto de Oliveira, Denise Crocce Romano Espinosa, and Jorge Alberto Soares Tenório…
30
CROSSREF
<1%
Barreiro, A., A.J.C. Crespo, J.M. Domínguez, and M. Gómez-Gesteira. "Smoothed Particle Hydrodynamics for coastal engineering problems", Com…
31
CROSSREF
<1%
iwaponline.com
32
INTERNET
<1%
publikationen.bibliothek.kit.edu
33
INTERNET
<1%
royalsocietypublishing.org
34
INTERNET
<1%
etheses.whiterose.ac.uk
35
INTERNET
<1%
theses.gla.ac.uk
36
INTERNET
<1%
Scheer, Alexandra M.. "An Elcient Computational Method for Calculating Properties of Face-Centered Cubic High Entropy Alloys.", New Mexico I…
37
PUBLICATION
<1%
pdxscholar.library.pdx.edu
38
INTERNET
<1%
Abbas Yeganeh-Bakhtiary, Hamid Houshangi, Fatemeh Hajivalie, Soroush Abolfathi. "A Numerical Study on Hydrodynamics of Standing Waves in …
39
CROSSREF
<1%
hal-cea.archives-ouvertes.fr
40
INTERNET
<1%
hrcak.srce.hr
41
INTERNET
<1%
pureportal.coventry.ac.uk
42
INTERNET
<1%
Chiara Biscarini, Silvia Di Francesco, Matteo Mencattini. "Application of the lattice Boltzmann method for large‐scale hydraulic problems", Interna…
43
CROSSREF
<1%
Kenneth K. Cho, Anne E. Cust, Megan Y. Foo, Guy D. Eslick. "Melanomas and stress patterns on the foot: a systematic review and meta-analysis", …
44
CROSSREF
<1%
Roshan Suminda Ranasinghe. "Numerical modelling of wave transformation over submerged breakwaters using Boussinesq-type models", Journ…
45
CROSSREF
<1%
wiki.manchester.ac.uk
46
INTERNET
<1%
Hu, Gusheng. "Towards large eddy simulation of dispersed gas-liquid two-phase turbulent Wows", Proquest, 20111109
47
PUBLICATION
<1%
Jian-Yu Chen, Chong Peng, Fue-Sang Lien, Eugene Yee, Xiao-Hua Zhao. "Simulations for the explosion in a water-dlled tube including cavitation u…
48
CROSSREF
<1%
Vacondio, R., B.D. Rogers, P.K. Stansby, P. Mignosa, and J. Feldman. "Variable resolution for SPH: A dynamic particle coalescing and splitting sch…
49
CROSSREF
<1%
repository.unhas.ac.id
50
INTERNET
<1%
D'Alessandro, F.. "The BCI criterion for the initiation of breaking process in Boussinesq-type equations wave models", Coastal Engineering, 200812
51
CROSSREF
<1%
Gotoh, Hitoshi. "Particle Method", Advanced Series on Ocean Engineering, 2013.
52
CROSSREF
<1%
Hongjie Wen, Bing Ren. "A non-reWective spectral wave maker for SPH modeling of nonlinear wave motion", Wave Motion, 2018
53
CROSSREF
<1%
sigaa.ufrrj.br
54
INTERNET
<1%
www.repositorio.ufop.br
55
INTERNET
<1%
Douillet-Grellier, Thomas, Bruce D. Jones, Ranjan Pramanik, Kai Pan, Abdulaziz Albaiz, and John R. Williams. "Mixed-mode fracture modeling with…
56
CROSSREF
<1%
Environmental Science and Engineering, 2015.
57
CROSSREF
<1%
Marcela A. Cruchaga, Diego J. Celentano, Tayfun E. Tezduyar. "Computational Modeling of the Collapse of a Liquid Column Over an Obstacle and…
58
CROSSREF
<1%
Sangmin Lee, Kwonhwan Ko, Jung-Wuk Hong. "Comparative study on the breaking waves by a piston-type wavemaker in experiments and SPH si…
59
CROSSREF
<1%
eprints.whiterose.ac.uk
60
INTERNET
<1%
iopscience.iop.org
61
INTERNET
<1%
Environmental Science and Engineering, 2014.
62
CROSSREF
<1%
Matt Frain, David P. Schmidt, Yong-Le Pan, Richard K. Chang. "Selective DeWection and Localization of Flowing Aerosols onto a Substrate", Aeros…
63
CROSSREF
<1%
Ryszard Staroszczyk. "Incompressible SPH Model for Simulating Violent Free-Surface Fluid Flows", Archives of Hydro-Engineering and Environm…
64
CROSSREF
<1%
archiv.ub.uni-heidelberg.de
65
INTERNET
<1%
openaccess.city.ac.uk
66
INTERNET
<1%
udspace.udel.edu
67
INTERNET
<1%
www.xsystem.hu
68
INTERNET
<1%
"MARE-WINT", Springer Science and Business Media LLC, 2016
69
CROSSREF
<1%
Dalrymple, R.A.. "Numerical modeling of water waves with the SPH method", Coastal Engineering, 200602
70
CROSSREF
<1%
Dong, G.. "Harbor resonance induced by subaerial landslide-generated impact waves", Ocean Engineering, 201007
71
CROSSREF
<1%
Huabin Shi, Pengfei Si, Ping Dong, Xiping Yu. "A two-phase SPH model for massive sediment motion in free surface Wows", Advances in Water Re…
72
CROSSREF
<1%
Moncho Gomez-Gesteira. "State-of-the-art of classical SPH for free-surface Wows", Journal of Hydraulic Research, 2010
73
CROSSREF
<1%
Prabhu Ramachandran, Kunal Puri. "Entropically damped artidcial compressibility for SPH", Computers & Fluids, 2019
74
CROSSREF
<1%
Rogers, Justin Scott. "Physical Oceanography in Coral Reef Environments: Wave and Mean Flow Dynamics at Small and Large Scales, and Result…
75
PUBLICATION
<1%
S. James, J. Zhu, M. S. Anand. "Large-Eddy Simulations as a Design Tool for Gas Turbine Combustion Systems", AIAA Journal, 2006
76
CROSSREF
<1%
Shrey Joshi, Jean Pierre Franc, Giovanni Ghigliotti, Marc Fivel. "An axisymmetric solid SPH solver with consistent treatment of particles close to t…
77
CROSSREF
<1%
escholarship.org
78
INTERNET
<1%
id.scribd.com
79
INTERNET
<1%
repository.uel.ac.uk
80
INTERNET
<1%
www.abcm.org.br
81
INTERNET
<1%
www.hindawi.com
82
INTERNET
<1%
www.nature.com
83
INTERNET
<1%
Ai, C.. "A multi-layer non-hydrostatic model for wave breaking and run-up", Coastal Engineering, 201204
84
CROSSREF
<1%
Carlos Alberto Dutra Fraga Filho, Chong Peng, Md Rushdie Ibne Islam, Christopher McCabe et al. "Implementation of three-dimensional physical …
85
CROSSREF
<1%
Jian, Wei, Shawn Y. Sim, Zhenhua Huang, and Edmond Yat-Man Lo. "Modelling of Solitary Wave Run-up on an Onshore Coastal Cliff by Smoothed…
86
CROSSREF
<1%
S. J. Lind, P. K. Stansby, B. D. Rogers. "Fixed and moored bodies in steep and breaking waves using SPH with the Froude–Krylov approximation", …
87
CROSSREF
<1%
portal.research.lu.se
88
INTERNET
<1%
Carlos Alberto Dutra Fraga Filho. "Chapter 3 Smoothed Particle Hydrodynamics Method", Springer Science and Business Media LLC, 2019
89
CROSSREF
<1%
Carlos Alberto Dutra Fraga Filho. "Development of a computational instrument using a lagrangian particle method for physics teaching in the are…
90
CROSSREF
<1%
Georgios Fourtakas, Renato Vacondio, Benedict D. Rogers. "On the approximate zeroth and drst-order consistency in the presence of 2-D irregula…
91
CROSSREF
<1%
Ni, Xingye, W. B. Feng, and Di Wu. "Numerical simulations of wave interactions with vertical wave barriers using the SPH method : NUMERICAL SI…
92
CROSSREF
<1%
Nyavor, Yvonne Elikplim Akua. "Association of a High Fat Diet, Gut Microbiota and Host Interactions with Enteric Neuropathy and Dysmotility in a …
93
PUBLICATION
<1%
Oger, G., S. Marrone, D. Le Touzé, and M. de Leffe. "SPH accuracy improvement through the combination of a quasi-Lagrangian shifting transport…
94
CROSSREF
<1%
Excluded search repositories:

None
Excluded from Similarity Report:

Bibliography
Excluded sources:
None
42
42
Manuscript Click here to access/download;Manuscript;ARTICLE
FILE_JOURNAL_OF_BRAZILIAN_SOCIETY.pdf
42
Click here to view linked References
Diffusive Terms Applied in Smoothed Particle Hydrodynamics Simulations

1
2 of Incompressible and Isothermal Newtonian Fluid Flows
3
81
4 Carlos Alberto Dutra Fraga Filho1,2,3*, Fábio Pavan Piccoli1,3
5
6 10
1
7 Federal Institute of Education, Science and Technology of Espírito Santo (IFES) –
8 Av. Vitória, 1729 – Jucutuquara – 29040-780 – Vitória – ES – Brazil
9
10 48
2
11 ESS Engineering Software Steyr GmbH - Berggasse 35 - 4400 - Steyr - Austria
12 32
13 3
Laboratory of Simulation of Free Surface Flows – Federal University of Espírito
14
15 Santo (UFES) – Av. Fernando Ferrari, 514 – Goiabeiras – 29075-910 – Vitória – ES –
16 Brazil
17
18 *Corresponding author’s email:cadff1@gmail.com
19
20
21
22
23 Abstract. Currently, there is not an extensive literature dedicated to the presentation
24
25 of the diffusive modelling applied to the momentum equation of incompressible and
26 64
27 isothermal Newtonian fluid flows solved by the Smoothed Particle Hydrodynamics
28
29 (SPH) method. This paper aims to present the most common viscosity modellings and
30 47 47
31 the LES (Large Eddy Simulation) method applied in the solution of Navier-Stokes
32
33 (N-S) equations. A comparative study showing different modellings of the diffusive
34
35 terms has been performed. Two incompressible free-surface flows were simulated:
36 71 24
37 the generation and propagation of waves on a flat beach and the collapse of a water
38
39 column. In the first case study, the SPH results were compared with those results
40
41 provided by the Eulerian modelling - Boussinesq-type nonlinear wave equations
42
43 solved by the finite difference method and validated from laboratory data. It was
44
45
verified that the laminar shear stresses modelling is the most adequate in the period
46
47
and wave period simulated, although great divergences have not been noticed in
48 3
relation to the other models used. In the collapse of the water column study, the SPH
49
50 results obtained after implementation of different approaches for the diffusive terms
51
52 of N-S equations presented good agreement with experimental or literature data.
53
54
55
56 Keywords: SPH method, turbulence, Large Eddy Simulation, SGS shear stresses,
57 artificial viscosity
58
59
60
61
62
63 1
64
65
1. Introduction
1
2 23
3 In principle, Navier-Stokes equations can be used to simulate both laminar and
4
5 turbulent flows without averaging or approximations other than the necessary
6
7 numerical discretisations [2]. Theoretically, with the employment of an adequate
8
9 numerical model, the conservation equations that describe physical phenomena
10
11 involved in the sciences/ engineering problems could be solved.
12
13 There is scarce literature dedicated to the study and implementation of
14
15 approximations for the diffusive viscous terms and turbulence in the momentum
16
17 equation (when solved using the incompressible SPH Lagrangian method). [43] is
18
19 recent and rare literature that deals with diffusive viscous terms in particle methods,
20
21 which can be cited.
22
23
24 There are other references dedicated to the study of diffusive terms (numerical) in
25
26 the SPH method, as [1] and [21], for example. However, in this case, the objective is
27 13
28 to avoid the generation of spurious numerical oscillations in the pressure and density
29 94
30 fields in weakly-compressible SPH schemes. Literature proposed the introduction of
31 13
32 an additional diffusive term in the continuity equation, to reduce the numerical noise
33
34 inside the density field obtained from the SPH interpolations. Since the pressure field
35 13
36 is evaluated from the density field through a state equation, a smoother density field
37
38 corresponds to a smoother pressure field which improves the accuracy of the pressure
39
40 field.
41 36
42 This paper is dedicated to the presentation of common modellings for the diffusive
43
44 viscous terms in the momentum equation and turbulence used in particle simulations
45
46 of incompressible and isothermal Newtonian fluid flows, with analysis of the results.
47
48 Besides, a brief literature review on the turbulence models implemented in the
49
50 momentum equation in SPH simulations is presented. Particular attention is given to
51 63
52 the Large Eddy Simulation (LES) turbulence model - utilised in this work with the
53 19
54 application of the Sub-Grid Scale (SGS) model.
55
56
57
In this work, four different approaches have been applied in the modelling of the
58
59
viscous diffusive terms/ turbulence in the momentum equation: (a) artificial viscosity,
60
61
62
63 2
64
65
(b) laminar shear stress, (c) laminar shear stresses + turbulence model (LES) and (d)
1
2 laminar shear stresses + artificial viscosity. Simulations have been performed for the
3 24
4 generation and propagation of waves on a flat beach and the collapse of a water
5
6 column.
7
8 In the first case studied, Lagrangian and Eulerian numerical models have been
9
10 applied in simulations. Numerical codes based on Lagrangian SPH method
11
12 (SPHysics) and Boussinesq-type nonlinear wave equations (FUNWAVE2D) have
13
14 been used. A detailed description of the SPHysics and FUNWAVE2D can ben found
15
16 in [19] and [23, 24], respectively.
17
18
19 In the collapse of a water column, two Lagrangian codes have been used in the
20
21 numerical simulations: SPHysics and the computational tool presented in [13,15].
22
23 SPHysics code has been validated for use in this specific problem from laboratory
24
25 experiments [8, 26]. [20] brings some simulation results provided by SPHysics to this
26
27 problem, using different viscosity treatments. The computational tool presented in
28
29 [13,15] employed a modelling of viscosity (laminar shear stresses + artificial
30
31 viscosity) unavailable in SPHysics and validated from experimental data.
32
33 41
34 The remainder of this paper is organised as follows. In Section 2, the physical-
35 3
36 mathematical modelling and the SPH method are presented. A brief literature review
37
38 of turbulence models applied in SPH Method is provided to the reader in Section 3.
39
40 Section 4 presents the viscosity and turbulence modellings employed in the
41 40
42
simulations of this work. The numerical simulations performed and discussions are in
43
44
Section 5. Finally, conclusions are presented in Section 6.
45
46
47 2. Physical-Mathematical Modelling
48
49
50 This section brings the physical-mathematical modelling used in this paper as well
51
52 as a brief presentation of the Smoothed Particle Hydrodynamics method.
53
54
55
56
57
58
59
60
61
62
63 3
64
65
2.1. Lagrangian Particle Modelling and SPH Method
1
2
3 The physical conservation equations (mass and momentum) employed in this paper
4
5 to a Newtonian, incompressible and isothermal fluid flow are presented in Table 1
6
7 together with the respective SPH approximations.
8
9 5
10
SPH method is employed in the solution of the partial differential equations used to
11
12
express mathematically the physical conservation equations.
13 21
14 According to the Lagrangian approach, the continuum domain is discretised into a
15
16 finite number of particles to obtain the physical properties from interpolations of the
17 3
18 physical properties of the neighbouring particles (inside a domain of influence with a
19
20 support radius equal to kh ). These interpolations employ smoothing functions (or
21
22 kernels) that have to satisfy some properties: convergence, smoothness, positivity,
23
24 symmetry, normalisation within the support domain and compact support.
25
26
27
28 Table 1 – Physical Conservation Equations and the SPH Approximations
29
30 11
31 Partial Differential Equations SPH Approximations
32
33 (continuum domain) (domain discretised by particles)
34
35 Continuity equation: 5
36 d a n
37
b
m v aa
bb
v . W (x x , h) (1)
d
38 .v dt b 1
39 dt
40
41 Momentum equation:
42
43 dv a n ab
PP
44 dv P mb W(xa xb , h)
45
2
vg dt b 1
22
ab
a
(2)
46 dt
47 _________________ g
48
49
50 where
5
51 a and b are superscripts that refer to the fixed and neighbouring particles,
52
53 respectively
54 is the density
55 16
56 v is the velocity
57
58 W is the smoothing function (kernel)
59 is the vector differential mathematical operator
60
61
62
63 4
64
65
t is the time
1 16
2 m is the mass
3 n is the number of neighbouring particles
4
5 x is the position
6
7 h is the smoothing length
2
8 is the kinematic fluid viscosity
9
10 P is the pressure
11 g is the gravity
12
13 d
14
is the material derivative
dt
15
16 a _ is the diffusive term of the fixed particle related to the viscosity and turbulence effects
17
18
19 11
20
Figure 1 shows the disposition of the particles within the domain of influence. The
21
22
fixed particle, located in the centre of the local domain has its physical properties
23 4
24
obtained from the interpolation of the properties of the neighbouring particles. The
25 kernel guarantees the greatest contribution of the nearest neighbouring particles to the
26
27 value of the physical quantity in the reference particle.
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47 4
48 Figure 1. Graphical representation of the domain of influence. The reference particle
49 has as neighbours all particles within the domain defined by the circumference with
50 support radius equal to kh .
51
52 Reprinted by permission from Springer: Springer Nature Switzerland AG.
53 Smoothed Particle Hydrodynamics: Fundamentals and Basic Applications in
54 Continuum Mechanics by Carlos Alberto Dutra Fraga Filho. Copyright 2019.
55
56
57 77
58 The cubic spline kernel, presented below, has been used in this work for 2-D
59
60 simulations. Other kernels used in SPH interpolations could be found in [17].
61
62
63 5
64
65
23
x x'
33 x x' 29
1 1 , __
se 0 x x' h
2 24
hh
3
3
4 10 1 x x'
5 W (x x', h) _ _ 2
_ 2- , _ se _ h x x' kh (3)
6 74hh
7 3
8 0, _ in_the_other_case.
9
10
11 where
12
16
13 x x' is the distance between the position of a fixed and a variable point at the
14
15 domain
16
17 h is the smoothing length
18 21 56
19 k is a scale factor that depends on the kernel employed in the interpolations (it is
20
21 equal to 2 for the cubic spline kernel)
22
23 5
24 SPH provides approximations to the physical properties (as density and
25
26 temperature), gradients (as the pressure gradients), divergents (as the velocity
27
28 divergents) and Laplacians (as the temperature and velocities Laplacians), with a
29
30 second order error [17].
31
32 In this method, in simulations for incompressible fluids, a state equation is used in
33
3
34 the prediction of the pressure field. In this paper, the Tait equation has been used:
35
36
37
a
38
39
a
PB
dyn 0
1 , (4)
40
41
42
43 c2 0 (5)
B
44
45
46 3
47 where
48
49 a
Pdyn is the dynamic pressure acting on the fixed particle
50
51 0 2
is the density of the fluid in rest
52
53 = 7, for fluid simulations
54
55 B is the term related to the fluctuations of density
56
57 c is the sound velocity in the fluid
58
59
60
61
62
63 6
64
65
3
1
So that the Tait equation can be applied in the prediction of the pressure field of
2 the fluid, the maximum value of the Mach number must be 0.10. Thus, c 10 v max ,
3
4 10
where v max is the maximum velocity of the fluid in the simulation.
5
6
7 The next section brings a brief explanation of turbulence concepts and especially
8
9 of the LES turbulence model.
10
11
12 3. Brief Literature Review of Turbulence Models applied in SPH Method
13
14
15 The first and more intuitive attempt to capture the effects of turbulence in fluid
16 19
17 flows is using the direct numerical simulation method (DNS). In this approach, the
18
19 whole spectrum of turbulence scales from solving Navier-Stokes equations (no other
20 43
21 modelling is required) is resolved. The discretisation of the domain should capture all
22
23 of the kinetic energy dissipation from the solution of the conservation equations in a
24 20
25 high number of spatial points defined in the domain (from the smallest dissipative
26
27 scales - Kolmogorov microscales -, up to the integral scale, associated with the
28 3
29 motions containing most of the kinetic energy). The resolution of a wide range of
30
31 scales of time and space requires an extremely refined spatial and temporal
32
33 discretisation and a great computational effort (which means that DNS cannot be used
34 3
35 in most simulations). Turbulent flows have high degrees of freedom, which
36
37 correspond to the number of linear equations discretised to be solved at each spatial
38
39 point for good flow characterization. Due to these high degrees of freedom, DNS
40
41 leads to good results for low Reynolds numbers, which is a small fraction of the
42
43
problems in the universe of fluid dynamics [17]. The literature presents applications
44 3
45
of DNS in the particle simulation of decaying turbulence in a non-slip square box [36]
46
47
and laminar periodic hill flow [41] with good results.
48 88
49 In most realistic turbulent flows, that present from moderate to high Reynolds
50
51 numbers, it is necessary to carry out the implementation of a turbulence model. The
52 10
53 traditional closure turbulence models to the Reynolds-averaged-Navier-Stokes
54
55 (RANS) equations - using the Boussinesq eddy viscosity assumption, the mixing-
56 82
57 length Lm model, the two-equations ε model and the Explicit Algebraic
58
59 Reynolds Stress (EARSM) models – as well as the Large Eddy Simulation (LES)
60
61
62
63 7
64
65
4
model are presented in [41], when applied to the Smoothed Particle Hydrodynamics
1
2 (SPH) method. Validation tests have been performed to the following cases: (a) open-
3
4 channel turbulent steady flow ( Lm , ε and EARSM), (b) 2-D collapse of a water
5
6 column ( Lm , ε and EARSM) (c) 3-D turbulent open-channel flow (LES) and
7
8 3
9
(d) 3-D collapse of a water column (LES).
10
11 Large Eddy Simulation is a technique for turbulence modelling employed in
12
13 various area of fluid dynamics. Some important applications are presented in the
14
15 literature: gas turbines [29]; atmospheric boundary layer [35]; wind tunnels
16
17 simulations [3]; aeronautics, aeroacoustics and wall layer modelling [34, 44];
18 86
19 simulations of closed turbulent channel flow [28]; simulation of a near-shore solitary
20
21 wave mechanics [27]; wave interaction with a floating breakwater [37]; breaking of
22
23 plunging waves (Shao and Ji 2006) [38], among others.
24
8
25 LES is used in the simulation of the large scales of the turbulent motions. Using a
26
27 spatial filter, LES divides turbulence flow into large-scale and small-scale (grid-scale)
28
19
29 motions. The large eddies contain most of the turbulent kinetic energy and are
30
31 retained, and solved for directly from the averaged equations. The smaller scales of
32
22
33 the turbulent motion extract the energy and dissipate energy from the larger scales,
34
35 are removed from LES using a low-pass filter [10, 19, 25, 34, 40, 44]. As a
36
37 consequence, in the resulting filtered equations appear additional terms, called sub-
38
39 grid scale (SGS) shear stresses, that will be solved by a closure turbulence model.
40
8
41 When the grid size is sufficiently small, the impact from the sub-grid scale models on
42
43 the flow motion will be small.
44
45
46 Favre-averaged filter is commonly used in the process of separation of the
47
48 turbulence scales [10, 19, 34]. For a physical property f , the Favre-averaging is
49
50 ff, where denotes an spatial filtering.
51
52
49
53 After passing the filter, the mass and momentum conservation equations for a
54
55 Newtonian and incompressible fluid flow can be written (in Einstein notation) as
56
57 follows:
58
59
60
61
62
63 8
64
65
1
(6)
2
3
4
5
6
7
(7)
8
9
10 where
11 * 34
12 is the sub-grid scale (SGS) shear stress tensor which has the following elements:
13
14 (8)
15
76
16 is the strain rate tensor (9)
17
18
19 is the absolute fluid viscosity
20 38
21 ij
is the Kronecker delta ( ij
1,_ if_ i j; _ ij
0,_if _ i j)
22
72
23 ij
, (1, 2,3) represent the Cartesian directions
24
25
26 22
27
The effects of the unresolved scales are contained in the SGS shear stress tensor
28 46
29
and represent the motion that occurs on a scale smaller than the grid spacing ∆. In the
30
31
SPH method, the SGS stress tensor represents the turbulent eddies smaller than the
32 39
33
particle size [19]. There are different SGS models and most of them employ an eddy-
34 viscosity assumption (based on the Boussinesq’s hypothesis) to model the shear stress
35
36 10
37
tensor [7, 25, 44]. The formulation presented below for the components of the SGS
38 stress tensor - according to [20] - has been used in the simulations performed in this
39
40 work:
41
42
43
44
45
(10)
46
47
48
49 (11)
50
51
52
53
54
55 where
56
57
are the elements of the sub-grid scale stress tensor
52
58 is the eddy viscosity
t
59
60 k is the turbulence kinetic energy
61
62
63 9
64
65
Cl is equal to 0.00066
1 79
2 Sij is the strain rate tensor, i and j are subindices that refer to the Cartesian
3
4 directions
5 ui is the fluid velocity and xi the spatial position
6
7
8
8 Eddy-viscosity models are used to determine . Using the equilibrium assumption
t
9
10
11 (that the turbulence in a grid with small scale eddies is in equilibrium, and the
12
13
dissipated energy is implemented from the large scale eddies), the Smagorinsky
14
15
model [39] can be used to calculate the eddy-viscosity:
16
17 (12)
18
19
20 where
21 Cs is the Smagorinsky constant (0.1–0.2) [18]. In the simulation of this work, used as
22 70
23 0.12.
24
25 is the initial particle spacing
26
27
28
29 4. Viscosity and Turbulence Models
30
31 Some different proposals are presented in literature to model the diffusive terms in
32
62
33 the SPH approximation of the momentum conservation equation. The approaches to
34
35 the diffusive terms ( ) used in this work are presented below.
a
36
37
38 73
39 4.1 Artificial Viscosity ( ab )
40
41
42
The formulation used in artificial viscosity modelling, proposed in [30], is used
43
44
instead of the diffusive viscous terms in the momentum equation. It is very applied
45 36
46
because of its simplicity. However, it is important to note that employed in this way,
47 the artificial viscosity becomes only a computational solution tool. An adequate
48
49 choice of some coefficients is necessary to achieve satisfactory simulation results. The
50
51 form proposed in this article is shown in Eq. (13):
52
53
54 ab cab 5
55 , __if_ v a vb . xa xb 0,
56 ab (13)
57 ab
___ 0, _____if_ v a vb . xa xb 0.
58
59
60
61
62
63 10
64
65
90
a
hab vvb
. XX
a b
c aac bb a b h h
1 ab ab
, __ch __ ab
_,, _ __ ab 11 .
2 ab
2
2 2 2 2
XX 0, 01h
3
4
5
6 where
7
8 ab
is the artificial viscosity
9 a and b are the densities of the fixed and the neighbouring particle, respectively
10 11
11 ca and cb are the sound velocities in the fixed and neighbouring particles,
12
13 respectively
14 11
15 is a coefficient used in the calculation of the artificial viscosity
16 17
ha and hb are the support radii of the fixed and neighbouring particles, respectively
17
18 2
is a factor that prevents numerical differences when two particles approach one
19
20 another
21
22
23
In the equations above, the factor is set to 0.01 hab 2 .
24
25 The diffusive terms of the fixed particle modelled using the artificial viscosity is
26
27 presented below.
28
29 n
30 a
mb ab
a
W ( xxb
, h) (14)
31 b 1
32
33
34
35 4.2 Laminar Shear Stresses
36
37
38 The diffusive terms a is the laminar shear stresses of the fixed particle
39 2
v , that can be obtained from Eqs. (15) and (16), presented in [19] and
40 aa LAMINAR
41 [15], respectively:
42
43
44 ab 5
45 2
n
b 4 xx
46
va m a
(x a xb , ) v a v b
. Wh
a a
LAMINAR
b 1
2
abab
xx (15)
47
48
49
50 n
2
ab
xx
2
51 va mb a
(x a xb , ) v a v b
. Wh
52
a a LAMINAR 5b 1
b ab
x x
2 (16)
53
54 9
55 where a is the kinematic viscosity of the fluid particle
56
57
58 The first proposal for laminar shear stresses - Eq. (15) - is used in the SPHysics
59
60 code [19] and the second one in the computational code presented in [13, 15] utilised
61
62
63 11
64
65
58
in the simulation of the collapse of a water column.
1
2
3 61
4 4.3 Laminar Shear Stresses + Sub-grid Scale (SGS) turbulence model
5
6
7
In this model, the diffusive terms a
is composed of two parcels, related to laminar
8
9 shear stresses (Eqs. (15)-(16)) and sub-grid scale (SGS) shear stresses, as shown in
10
11 Eq. (17):
12
13 **
n ab
14 2 (17)
15 a a
v
a
mb 22
a
W (xx
. b
, h)
LAMINAR ab
16 b 1
17
18
* 34 29
19 where is the sub-grid scale (SGS) shear stress tensor - defined in Eq. (7).
20
21
22 The last term on the right hand side of Eq. (17) is presented in [10, 19, 20, 27].
23
24
25
26
4.4 Laminar Shear Stresses + Artificial Viscosity
27
28 In this approach, proposed in [14], the term related to laminar shear stresses is
29
30 implemented as well as the term related to the artificial viscosity as shows Eq.(18):
31
32
n
33 2
34 a a va mb ab
a
W (xxb
, h) (18)
b57
LAMINAR
35 1
36
37 The last term on the right hand side of Eq. (18) has been presented in Eq. (14) and
38 5
39 works as a numerical correction to avoid numerical instabilities and the
40
41 interpenetration between particles.
42
43 89
44 From the physical point of view, in problems involving mainly shock waves there
45 11
46
is conversion of kinetic energy into heat. That energy transformation can be
47
48
represented as a form of viscous dissipation and needs to be measured; which is
49
50
carried out with the artificial viscosity. It should be noted that in [14] the objective of
51 the authors when applying the artificial viscosity was different from that presented in
52
53 subsection 4.1 (in which the artificial viscosity directly replaced the viscous terms in
54
55 the momentum equation).
56
57
58
59
60
61
62
63 12
64
65
This modelling was implemented by the authors of this paper. It was utilised, and
1 26
2 the results analysed, in the second case studied (presented in Section 5.2: the collapse
3
4 of a water column).
5
6
7 5. Numerical Simulations
8
9 In this paper, simulations have been performed for the generation and propagation
10
11 of waves on a flat beach and collapse of a water column, presented below.
12
13
14 5.1 Generation and propagation of waves on a flat beach
15
16
17 28
18
Three zones can be delimited in a beach: wave breaking zone, surf zone and swash
19 zone. In the breaking zone the dissipation of the wave energy occurs. The surf zone is
20
21 the region in which the wave propagates after the break. The swash zone, located
22
23 immediately, is important because a substantial part of the total litoral sediment
24
25 transport occurs in this region. On the coast line the movement of the wave is
26
27 forward, climbing the beach (run-up), and backward, descending that (run-down),
28
84
29 delimiting of swash zone [33]. The action of breaking waves and run-up results in a
30
31 highly complex movement, comprising medium and orbital movements and
32
33 fluctuations (turbulence).
34
69
35 The Lagrangian simulations were performed with and without the use of a
36
37 turbulence model.
38
39
40 5.1.1 Softwares
41
42
43 The computational tools employed in the wave simulations were the SPHysics and
44
45 FUNWAVE2D non-commercial softwares.
46
47
48 The first software is an open computer code developed in FORTRAN
49 60
50 programming language and based on Smoothed Particle Hydrodynamics (SPH) for
51 31
52 the study of free surface flows [19, 20]. It is the result of collaboration by researchers
53
54
from Johns Hopkins University (USA), the University of Vigo (Spain) and the
55
56
University of Manchester (UK). SPHysics presents the options to use artificial
57
58
viscosity, laminar shear stresses or laminar shear stress + SGS shear stresses (for
59
60
turbulence modelling) in simulations.
61
62
63 13
64
65
2
FUNWAVE2D is a non-commercial FORTRAN software, produced by the Center
1
2 for Applied Coastal Research (CACR, Delaware, USA). This code employs the
3
4 numerical method of finite differences to solve the Boussinesq-type nonlinear wave
5
6 equations and to simulate surface waves in coastal areas including internal and
7
8 external surf zones. The wave breaking is treated using a formulation of artificial
9 50
10 turbulent viscosity. A simple eddy viscosity-type formulation is used to calculate the
11
12 wave energy dissipation caused by the wave breaking [22]. A validation study of this
13 25
14 software (from laboratory data) applied to the simulation of wave propagation in deep
15
16 waters and shoaling zones up to the breaking region is in [4].
17
18
19 5.1.2 Domain simulated, initial and boundary conditions
20
21 2
22
It was simulated a beach domain with 2.75m extension in the inclined region (with
23 2
24
an angle of 4.2364°). The amplitude and period of the wave were 0.01 m and 1.40s,
25
26
respectively, and the water level 0.18m. The water was considered a viscous,
80
27
28
incompressible and isothermal fluid. The kinematic viscosity was 1.0 x 10-6 m2/s.
29 4
30 In the SPHysics simulations, the Lagrangian SPH method has been employed to
31
32 solve the conservation equations of mass and momentum, Eqs. (1) and (2).
33 3
34 Considering a isothermal flow of water, the energy conservation equation does not
35
36 need to be solved. In the discretisation of the domain were employed 4,221 water
37 2
38 particles from 0.00 to 3.75 m. The lateral distance between the centres of mass of two
39
40 neighbouring particles was 0.01 m. The support radius was defined as 0.013 m
41
42 (approximately 1.30 times the initial spacing between the centres of mass of the
43
44 particles). Dynamic boundary particles - described in [19] - have been disposed on the
45
46 contours (fixed bottom of the beach and paddle). An amount of 387 fixed particles
47
48 was used to represent the bottom of the beach and 31 mobile particles were fixed
49 2
50 initially at the position 0.13m of the domain to represent the motion of the flap-type
51
52 wavemaker. The wave amplitude was 0.02m. In Tait equation, the parameter B was
53
54 7.175 x 105 kg/ms2 and 0 was 1,000 kg/m3. Shepard filter has been used in the
55
56 density renormalization every 30 time steps. Kernel correction or kernel gradient
57 83
58 correction [19] were not applied. A time step of 4.5x10-5 s was used to simulate the
59
60 generation and propagation of the waves over a time simulation of 30.00s. Figure 2
61
62
63 14
64
65
shows the geometry, the particles disposed inside the domain (in blue), on the paddle
1
2 (mobile, in red) and on the bottom of the channel (immobile, in black).
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17 92
18 Figure 2. Simulated geometry and initial setup of the fluid and boundary particles.
19
20
21 In the Eulerian simulations carried out aiming to verify the SPH results, the
22 2
23 FUNWAVE2D software has been used. A domain mesh with 156 nodes in the
24
25 horizontal direction and 21 nodes in the vertical direction was used. The spacing
26
27 between dots in the horizontal direction was 0.05 m with a total horizontal distance of
28 25
29 7.75 m. Vertically, the spacing varied with the total depth, given by H h , where
30
31 h is the mean sea level and η is the surface elevation, defining a sigma-type mesh.
32
33
34 The dimensions of the Eulerian domain ranged from -4.00 m to 3.75 m. The
35 2
36 simulated period was 30.0s, with a time increment of 1.0 x10-3 s during the
37
38 simulations. The wave breaking was treated by an artificial turbulent viscosity
39 51
40 formulation, which can promote a more realistic description of the beginning and the
41
42 end of the wave breaking in quantitative terms (in terms of the dissipated energy) [4,
43 2
44 5, 6]. For the time integration, in both Eulerian and Lagrangian simulations, the
45 2
46 predictor-corrector method was used and the CFL number assumed a value of 0.20.
47
48
49 5.1.3 Results and Discussions
50
51
52 SPH simulations have been performed for three different treatments of the
53
54 diffusive terms present in Eq. (2): (a) laminar shear stresses; (b) laminar shear stresses
55
56 + SGS shear stresses and (3) artificial viscosity, with = 0.05, 0.20 and 0.30.
57
58
59
60
61
62
63 15
64
65
1
The period used was 1.40s and the wave amplitude was 0.01m in all cases. The
2 simulation results are plotted and the wave elevations can be shown in Figs. 3-6. The
3
4 reaching of the wave over the beach and its elevation are showed in the time instants
5
6 10.0s, 15.0s, 20.0s and 28.0s, for different modellings of viscosity and turbulence.
7
8
9 The results are present in Tables 2-9 and Figs 3-6. They can be divided into two
10
11 groups. The first is composed by the results provided by the physical modellings for
12
13
the diffusive terms, that is, laminar shear stresses and laminar shear stresses + SGS
14
15
shear stresses (in the first two lines of Figs. 3-6). The second group has as elements
16 the SPH results achieved when the artificial viscosity was solitary employed as a
17
18 purely computational modelling to the diffusive terms present in Eq. (2) - in the third
19
20 and fourth lines of Figs. 3-6.
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63 16
64
65
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54 Figure 3. SPH and Eulerian (red line) simulation results at t = 10.0 s. From the top of
55 the page: laminar shear stresses + SGS shear stresses, laminar shear stresses and
56
57 artificial viscosity modellings (with = 0.05, 0.20 and 0.30, respectively).
58
59
60
61
62
63 17
64
65
1
Table 2 - Free Surface Elevation at t = 10.0 s (in meters)
2
3
Modelling Applied
4 Position at
5 the domain 1* 2 3 4 5 6
6
7
(m)
8 0.25 0.174 0.193 0.186 0.193 0.186 0.192
9 0.50 0.179 0.196 0.194 0.196 0.193 0.196
10
11 0.75 0.192 0.201 0.199 0.199 0.197 0.196
12 1.00 0.191 0.187 0.194 0.199 0.197 0.196
13 1.25 0.182 0.182 0.190 0.194 0.193 0.191
14
15 1.50 0.173 0.182 0.186 0.190 0.188 0.186
16 1.75 0.173 0.186 0.186 0.193 0.188 0.186
17 2.00 0.177 0.186 0.186 0.189 0.184 0.190
18
19 2.25 0.190 0.186 0.182 0.189 0.188 0.190
20 2.50 0.190 0.185 0.177 0.184 0.188 0.190
21
22 2.75 0.169 0.178 0.177 0.179 0.180 0.185
23 3.00 0.176 0.171 0.169 0.178 0.179 0.184
24
25 * 1- Boussinesq-type nonlinear wave equations; 2- Laminar + SGS Shear Stresses; 3-
26 Laminar Shear Stresses; 4- Artificial Viscosity (with = 0.05); 5- Artificial
27
28 Viscosity (with = 0.20); 6- Artificial Viscosity (with = 0.30)
29
30 From the results presented in Table 2, the percentual relative errors between the
31 modellings 2-6 and 1 (defined as standard) have been calculated, as shows Table 3.
32
33 The last line in Table presents the average of the relative errors along the points in the
34
35
domain where the free surface elevations were measured.
36
37 Table 3 - Percentual relative error between the results achieved from the modellings
38 2-6 and Boussinesq-type nonlinear wave equations
39
40 Position at the
41
2 3 4 5 6
domain (m)
42
43
0.25 10.92% 6.90% 10.92% 6.90% 10.34%
44 0.50 9.50% 8.38% 9.50% 7.82% 9.50%
45 0.75 4.69% 3.65% 3.65% 2.60% 2.08%
46
47 1.00 2.09% 1.57% 4.19% 3.14% 2.62%
48 1.25 0.00% 4.40% 6.59% 6.04% 4.95%
49 1.50 5.20% 7.51% 9.83% 8.67% 7.51%
50
51 1.75 7.51% 7.51% 11.56% 8.67% 7.51%
52 2.00 5.08% 5.08% 6.78% 3.95% 7.34%
53 2.25 2.11% 4.21% 0.53% 1.05% 0.00%
54
55 2.50 2.63% 6.84% 3.16% 1.05% 0.00%
56 2.75 5.33% 4.73% 5.92% 6.51% 9.47%
57
58 3.00 2.84% 3.98% 1.14% 1.70% 4.55%
59 Average of the
60 4.89% 4.91% 6.26% 5.00% 6.14%
relative errors
61
62
63 18
64
65
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
55 the page: laminar shear stresses + SGS shear stresses, laminar shear stresses and
56
58
59
60
61
62
63 19
64
65
27
1
2
3
Modelling Applied
4 Position at
5 the domain 1* 2 3 4 5 6
6
7
(m)
8 0.25 0.197 0.193 0.191 0.189 0.189 0.197
9 0.50 0.189 0.184 0.182 0.189 0.190 0.187
10
11 0.75 0.176 0.188 0.186 0.189 0.185 0.187
12 1.00 0.171 0.183 0.181 0.184 0.185 0.187
13 1.25 0.175 0.183 0.185 0.189 0.185 0.187
14
15 1.50 0.184 0.191 0.194 0.193 0.190 0.191
16 1.75 0.197 0.191 0.189 0.189 0.190 0.190
17 2.00 0.189 0.178 0.180 0.189 0.190 0.186
18
19 2.25 0.171 0.182 0.180 0.188 0.185 0.180
20 2.50 0.176 0.181 0.180 0.184 0.185 0.185
21
22 2.75 0.184 0.185 0.188 0.184 0.185 0.189
23 3.00 0.197 0.189 0.188 0.193 0.180 0.185
24
25 *1- Boussinesq-type nonlinear wave equations; 2- Laminar + SGS Shear Stresses; 3-
28
30
32
34 domain where the free surface elevations were measured.
35
37
39
40 Position at the
2 3 4 5 6
41 domain (m)
42 0.25 2.03% 3.05% 4.06% 4.06% 0.00%
43
44 0.50 2.65% 3.70% 0.00% 0.53% 1.06%
45 0.75 6.82% 5.68% 7.39% 5.11% 6.25%
46 1.00 7.02% 5.85% 7.60% 8.19% 9.36%
47
48 1.25 4.57% 5.71% 8.00% 5.71% 6.86%
49 1.50 3.80% 5.43% 4.89% 3.26% 3.80%
50
51
1.75 3.05% 4.06% 4.06% 3.55% 3.55%
52 2.00 5.82% 4.76% 0.00% 0.53% 1.59%
53 2.25 6.43% 5.26% 9.94% 8.19% 5.26%
54
55 2.50 2.84% 2.27% 4.55% 5.11% 5.11%
56 2.75 0.54% 2.17% 0.00% 0.54% 2.72%
57 3.00 4.06% 4.57% 2.03% 8.63% 6.09%
58
59 Average of the
3.93% 4.67% 4.30% 4.59% 4.46%
60 relative errors
61
62
63 20
64
65
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
54
55
the page: laminar shear stresses + SGS shear stresses, laminar shear stresses and
57
58
59
60
61
62
63 21
64
65
27
1
2
3
Modelling Applied
4 Position at the
5
1* 2 3 4 5 6
domain (m)
6
7
0.25 0.175 0.179 0.177 0.177 0.183 0.186
8 0.50 0.175 0.184 0.182 0.186 0.187 0.191
9 0.75 0.184 0.188 0.186 0.191 0.192 0.195
10
11 1.00 0.191 0.185 0.191 0.191 0.191 0.200
12 1.25 0.191 0.188 0.191 0.191 0.191 0.200
13 1.50 0.179 0.188 0.182 0.191 0.191 0.195
14
15 1.75 0.175 0.184 0.178 0.186 0.186 0.191
16 2.00 0.175 0.184 0.182 0.186 0.186 0.191
17
18
2.25 0.184 0.188 0.182 0.186 0.190 0.191
19 2.50 0.197 0.193 0.191 0.191 0.194 0.195
20 2.75 0.179 0.188 0.182 0.195 0.189 0.195
21
22 3.00 0.179 0.184 0.182 0.186 0.184 0.195
23
24
*1- Boussinesq-type nonlinear wave equations; 2- Laminar + SGS Shear Stresses; 3-
27
28
31
34
35
38
39 x(m) 2 3 4 5 6
40 0.25 2.29% 1.14% 1.14% 4.57% 6.29%
41
42 0.50 5.14% 4.00% 6.29% 6.86% 9.14%
43 0.75 2.17% 1.09% 3.80% 4.35% 5.98%
44
45
1.00 3.14% 0.00% 0.00% 0.00% 4.71%
46 1.25 1.57% 0.00% 0.00% 0.00% 4.71%
47 1.50 5.03% 1.68% 6.70% 6.70% 8.94%
48
49 1.75 5.14% 1.71% 6.29% 6.29% 9.14%
50 2.00 5.14% 4.00% 6.29% 6.29% 9.14%
51 2.25 2.17% 1.09% 1.09% 3.26% 3.80%
52
53 2.50 2.03% 3.05% 3.05% 1.52% 1.02%
54 2.75 5.03% 1.68% 8.94% 5.59% 8.94%
55 3.00 2.79% 1.68% 3.91% 2.79% 8.94%
56
57 Average of the
58
2.97% 1.68% 3.86% 4.46% 7.61%
relative errors
59
60
61
62
63 22
64
65
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
54 the page: laminar shear tresses + SGS shear stresses, laminar shear stresses and
55
57
58
59
60
61
62
63 23
64
65
27
1
2
3
Modelling Applied
4 Position at
5 the domain 1* 2 3 4 5 6
6
7
(m)
8 0.25 0.179 0.189 0.188 0.191 0.193 0.197
9 0.50 0.192 0.200 0.187 0.199 0.197 0.201
10
11 0.75 0.195 0.200 0.191 0.195 0.197 0.197
12 1.00 0.186 0.192 0.182 0.191 0.193 0.193
13 1.25 0.174 0.188 0.178 0.186 0.189 0.189
14
15 1.50 0.169 0.192 0.182 0.191 0.189 0.189
16 1.75 0.173 0.195 0.181 0.191 0.189 0.189
17 2.00 0.185 0.199 0.189 0.199 0.193 0.193
18
19 2.25 0.197 0.203 0.189 0.191 0.189 0.182
20 2.50 0.176 0.195 0.181 0.186 0.189 0.189
21
22 2.75 0.172 0.190 0.180 0.186 0.189 0.189
23 3.00 0.176 0.202 0.188 0.191 0.185 0.189
24 *1- Boussinesq-type nonlinear wave equations; 2- Laminar + SGS Shear Stresses; 3-
25
28
30
32
35
36
39
40 Position at the
41
2 3 4 5 6
domain (m)
42
43
0.25 5.59% 5.03% 6.70% 7.82% 10.06%
44 0.50 4.17% 2.60% 3.65% 2.60% 4.69%
45 0.75 2.56% 2.05% 0.00% 1.03% 1.03%
46
47 1.00 3.23% 2.15% 2.69% 3.76% 3.76%
48 1.25 8.05% 2.30% 6.90% 8.62% 8.62%
49 1.50 13.61% 7.69% 13.02% 11.83% 11.83%
50
51 1.75 12.72% 4.62% 10.40% 9.25% 9.25%
52 2.00 7.57% 2.16% 7.57% 4.32% 4.32%
53 2.25 3.05% 4.06% 3.05% 4.06% 7.61%
54
55 2.50 10.80% 2.84% 5.68% 7.39% 7.39%
56 2.75 10.47% 4.65% 8.14% 9.88% 9.88%
57
58 3.00 14.77% 6.82% 8.52% 5.11% 7.39%
59 Average of the
60 7.81% 3.45% 6.80% 6.25% 7.50%
relative errors
61
62
63 24
64
65
1
Analysing the free surface elevation results, in the first group was verified that the
2 laminar shear stresses modelling resulted in the wave elevations more adjusted to
3
4 those provided by the Eulerian modelling - the Boussinesq-type nonlinear wave
5
6 equations, used as standard of comparison and presented in the Appendix -, mainly
7
8 from t = 20.0 s. However, the results obtained from the laminar shear stresses + SGS
9
10 shear stresses approach showed, in general, a good agreement with the Eulerian
11
12 modelling, according to previous results shown in [12]. In particular, at t = 28.0s,
13
14 higher differences in wave elevations have been noted, when comparing this last
15
16 modelling with the reference Eulerian results. In conclusion, in the physical
17
18 modelling of the diffusive terms, the effects of the turbulence were not significant -
19
20 due to the low Reynolds number in the period and wave amplitude simulated.
21
22 The second group of results (in three last lines of Figs. 3-6 and modellings 4 to 6 in
23
24 Tables 2-9) showed that the application of the artificial viscosity in the momentum
25
26 equation can lead to reasonable numerical results (depending on the best choice of the
27
28 parameter , which is a function of the problem studied) although a purely
29
30 computational term is applied in the substitution of the diffusive terms .
a
31
32
33 Regarding the maximum wave elevations (at the end of the wet region, on the
34
35 sloping area of the beach), the Boussinesq-type nonlinear wave equations provided the
36 68
37
highest values in the simulations (18.8 cm, 19.0 cm, 19.4 cm and 19.6 cm in the time
38 instants 0.10 s, 0.15 s, 0.20 s and 0.28 s, respectively). The maximum differences
39
40 between the wave elevations provided by the Eulerian and Lagrangian models were:
41
42 +0.8 cm at t = 10.0s (comparing 1st and 4th modellings), +0.5 cm at t = 15.0 s
43
44 (comparing 1st and 5th modellings), +1.4 cm at t = 20.0 s (comparing 1st and 5th
45
46 modellings) and +1.4 cm at t = 28.0 s (comparing 1st and 4th modellings).
47
48
49 [32] is a study that presents an analysis of the effects of the artificial viscosity in
50
51 simulations of propagation of regular waves.
52
53
54
55
56
57
58
59
60
61
62
63 25
64
65
1 5.2 Collapse of a water column
2
3
4 5.2.1 SPHysics simulations (using artificial viscosity, laminar shear stresses and
5
6 laminar + SGS shear stresses)
7
8 66
9 SPHysics software has been used to simulate the collapse of the water column.
10
11 Laboratory experiments shown in [8, 26] have been used in the validation of this
12
13 computational tool for utilisation in this case. The software has been used in
14
15 simulations presented in this subsection.
16
17
18 5.2.1.1 Domain simulated, initial and boundary conditions
19
20
21 The dimensions of the tank were 4.00m x 4.00m and the water column had a width
22 9
23 of 1.00m and a initial height of 2.00 m. The particles were arranged with an initial
24
25 separation between their centres of mass of 3.00 cm. The time step employed was
26
27 started as 1.00 x 10-4s, being variable and calculated by the criterion presented in [19].
28
3 3
29 The continuity and momentum equations have been solved. Tait equation was used
30
31 to predict the particle pressure field. The parameter B was equal to 2.803 x 105
32
33 kg/ms2 and was 1,000 kg/m3. The treatment of boundaries occurred through the
0
34
35 10
36
dynamic contour particles. In order to avoid the tensile instability and the
37
38
interpenetration between particles, SPHysics applied the artificial pressure proposed in
39 85
40
[31]. The density correction has been done with the Shepard filter every 30 time steps.
41 Kernel correction or kernel gradient correction were not utilised. The XSPH method
42
43 has been used for a more ordered movement of the particles in the time. A description
44
45 of these numerical corrections are in [19].
46
47
48
49 5.2.1.2 Results and discussions
50
51
52
53
Simulations have been performed and the results compared for different modelling
54
55
of viscosity and turbulence. Figure 7 shows the graphical results obtained using (a)
56
57
artificial viscosity ( = 0.10), (b) laminar shear stresses and (c) laminar shear
58
59 stresses + SGS shear stresses. The modelling (a), using artificial viscosity and =
60
61
62
63 26
64
65
0.10, whose results were presented in [20], was taken as reference in the comparison
1
2 of the results.
3
4 0.38s 0.86s 1.30s
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41 Figure 7. SPH simulation results. (a) Artificial viscosity, with = 0.10 (according to
42
43 the results validated from laboratory data), presented in [20], (b) Laminar viscosity +
44 SGS shear stresses modelling and (c) Laminar shear stresses modelling.
45
46
47
48 Differences have been seen between the simulation results. The behaviour of the
49
50 waves has shown agreement at the time instant 0.38s - the vertical line in the Fig. 7
51
52 shows that the wave fronts presented abscissas near 2.20 m in all simulations.
53 10
54 In t = 0.86 s, the heights of the waves over the tank wall were different: (a) 1.46m;
55
56 (b) 1.69m and (c) 1.58m. The separation among particles, after the impact against the
57
58 tank wall, were observed. The final heights of 1.92m and 1.86 m, were achieved by
59
60
61
62
63 27
64
65
the detached particles in the simulations (b) and (c), respectively. In t = 1.30 s, the
1
2 detachment of particles was more visible in the modellings (b) and (c).
3
4 The model that employed the laminar shear stresses treatment for the viscosity (c)
5
6 presented a result with a reasonable agreement with the validated simulation (a) - that
7
8 used the artificial viscosity, with = 0.10. The waveforms were qualitatively similar
9
10
11 in all modellings applied during the simulated time.
12
13
14 5.2.2 (Laminar shear stresses + artificial viscosity) SPH Simulations
15
16
17 The literature [11] presented a model in which the viscous terms have been
18
19 modelled using laminar shear stresses + artificial viscosity. According to this
20
21 approach, the physical viscosity is implemented in conjunction with the artificial
22 5
23 viscosity (which in this case works as a numerical correction to avoid numerical
24
25 instabilities and the interpenetration between particles). A complete presentation of
26
27 the simulations performed is in [11]. The computational code utilised is presented in
28
29 [13, 15].
30
31
32 5.2.2.1 Results and discussions
33
34 3
35 Figs. 8 and 9 present the comparison of the positions of the wave fronts provided
36
37 by the SPH simulations and experimental data provided by [9].
38
39 35
40 In the simulation using laminar shear stresses + artificial viscosity ( = 0.20), the
41
42 differences between the numerical results and experimental data were 13.64% and
43
44 2.78% at the time instants 0.10 and 0.20, respectively. When using laminar shear
45
46 stresses + artificial viscosity ( = 0.30), those differences were 17.86% and 2.33%.
47
48 53
49
From the comparison between the numerical results and experimental data it can
50
51
conclude that the formulation using laminar shear stresses + artificial viscosity and
52 reflective boundary conditions [16] provides a consistent solution to the collapse of
53
54 the wave column.
55
56
57
58
59
60
61
62
63 28
64
65
26
t = 0.10 s t = 0.20 s
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26 Figure 8. Simulations employing (Laminar shear stresses + artificial viscosity, with
27
28 = 0.20). At the first line, the experimental results are presented. The vertical
29 dashed line is used to allow the comparison between the positions of the wave fronts.
30 26
31 t = 0.10 s t = 0.20 s
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56 Figure 9. Simulations employing (laminar shear stresses + artificial viscosity, with
57 87
58 = 0.30). At the first line, the experimental results are presented. The vertical
59 dashed line is used to allow the comparison between the positions of the wave fronts.
60
61
62
63 29
64
65
5.2.3 SPHysics (artificial viscosity) vs. (Laminar shear stresses + artificial
1
2 viscosity) SPH simulations
3
4
5 The literature [14] presents a comparison between simulations performed by
6
7 SPHysics and numerical code described in [13, 15].
8
9
10
In that study, SPHysics used the artificial viscosity in substitution to the diffuse
11 9
12
terms a
, dynamic particles were applied in the boundary treatment, and corrections
13
14 to the pressure gradients near the contours have not been realised.
15
9
16 (Laminar shear stresses + artificial viscosity) modelling and reflective boundary
17
18 conditions [16] have been employed in the second computational simulation tool [13,
19
2
20 15]. A coefficient of restitution of kinetic energy equal to 1.00 (elastic collisions of
21
22 the water particles against the tank walls) and density and pressure gradient
23
24 corrections (CSPM) have been employed in this numerical model.
25
26
27 5
28 5.2.3.1 Domain simulated, initial and boundary conditions
29
30
31 The simulated geometry was a tank whose dimensions were 4.00m x 4.00m. The
32 9
33 water column had width of 1.00m and height of 2.00 m.
34
35 In the SPHysics, the particles were arranged with an initial separation between
36
37 their centres of mass of 3.00 cm. In Tait equation, the parameter B was 2.803 x 105
38
39 kg/ms2 and 0 was 1,000 kg/m3. Shepard filter was used in the density correction
40
41 every 30 time steps. Kernel correction and kernel gradient correction were not
42
43 applied. Dynamic boundary particles were employed in the treatment of contours.
44
45 The time step employed was started as 1.00 x 10-4s, being variable and calculated by
46
47 the criterion presented in [19].
48
49
50 In the simulations performed using laminar shear stresses and artificial viscosity -
51
52 in the code presented in [13, 15] -, the discretisation of the fluid volume employed
53 2
54 2,556 particles with an initial lateral separation between the centres of mass of 2.86
55
56 cm. A cubic spline kernel has been used in SPH interpolations. In Tait equation, the
57 2
58 parameter B was 0.85 x 105 Pa and 0 was 1,000 kg/m3. The time step was 1.0 x 10-4
59 3
60 s (kept constant during the simulations). The coefficient of restitution of kinetic
61
62
63 30
64
65
energy - in the implementation of the reflective boundary conditions [16] - was equal
1 91
2 to 1.00. The density correction was applied by the Shepard filter every 30 time steps.
3 2
4 The pressure gradients obtained by the SPH interpolations were corrected (using the
5
6 CSPM method) at each numerical iteration.
7
8 2
9
Figure 10 presents, graphically, the simulation results at the time instants 0.40s and
10 0.80 s. The simulations have been performed employing = 0.30.
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39 Figure 10. At the top: graphical results of the SPHysics simulations - using artificial
40
41 viscosity, = 0.30. At the second line: simulation results perfomed by the code
42 presented in [8,9], with laminar shear stresses + artificial viscosity ( = 0.30).
43
44 Fraga Filho, CAD, Chacaltana, JTA. Revista Interdiscisplinar de Pesquisa em
45 67
46
Engenharia, 2(11), 2016; licensed under a Creative Commons Attribution (CC BY)
47 license.
48
49
50
51 From the results found, it is possible to verify an agreement between the positions
52
53 of the wave fronts and waves heights in both simulations. Additionally, it is possible
54
55
to state that is possible to achieve the physical result through the employment of
56
57 different modellings of a , different boundary treatments and numerical corrections
58
59 in the simulations. The results provided by the second modelling was physically
60
61
62
63 31
64
65
consistent (using laminar shear stresses) and the numerical corrections (necessary due
1 3
2 to the errors of the SPH method) were performed by the artificial viscosity
3
4 implementation.
5
6 The achievement of consistent results depends, therefore, on the adjustment of
7
8 parameters in the simulations performed by different numerical codes.
9
10
11
12
6. Conclusions
13
14 54
15
This work presents an analysis of the the modelling of the diffusive terms in the
16
17
momentum equation of an incompressible and isothermal Newtonian fluid, according
18
19
to the Lagrangian modelling and solution provided by the SPH method. Two cases
20 24
21
were studied: propagation of waves on the beach and collapse of a water column.
22
23 In the first case study - that used SPHysics software in simulations -, it was
24
25 verified that the physical modelling of the diffusive viscous terms and turbulence,
26
27 employing the laminar shear stresses provided the more adjusted wave elevations
28
29 when compared to the Boussinesq-type nonlinear wave equations (Eulerian results
30
31 provided by FUNWAVE code, whose validation is in [4]) in the period and wave
32
74
33 amplitude simulated. The implementation of the artificial viscosity in the momentum
34
35 equation led to reasonable results for the wave elevations (due to the insignificance of
36
37 the turbulence effects in the problem simulated). The accuracy of its application
38
39 depends on the best choice of the parameter , which is a function of the problem
40
9
41 studied. Although it is widely used in the SPH simulations, it is a purely
42
43 59
computational term employed in the modelling of the diffusive viscous terms in the
44
45 momentum equation, with no physical meaning.
46
47
48 In the collapse of the water column, four modellings to the diffusive terms have
49
50 been tested. In the first simulations (using SPHysics and presented in 5.2.1), artificial
51
52 viscosity treatment, laminar viscosity + SGS shear stresses and laminar shear stresses
53
54 have been applied. The simulations that employed the laminar shear stresses
55
56 presented results with reasonable agreement with the validated results (obtained from
57
58 the comparison with experimental data and presented in [20], obtained when the
59
60 artificial viscosity, with = 0.10, was used). Separation among particles, mainly in
61
62
63 32
64
65
the modelling using laminar + SGS shear stresses, has been seen from t = 0.86 s. The
1
2 waveforms were similar in all modellings applied, during the simulated time.
3
4 A comparison between the results of the SPHysics simulations (using artificial
5
6
7
viscosity treatment, with = 0.30, and dynamic boundary particles in the boundary
8
9 treatment) and those provided by the computational code, presented in [13, 15], that
10
11 employed laminar shear stresses + artificial viscosity, with = 0.30, and reflective
12 28
13 boundary conditions, has been presented in 5.2.3. Good agreement has been achieved
14
15 between the results of the simulations. In Tait equation, B was 2.803 x 105 in
16
17 SPHysics and equal to 0.85 x 105 Pa in the second computational code. It can be
18
19 concluded that it is possible to obtain results according to experimental data or
20
21 literature through both computational tools, depending on the adjustment of different
22
23 parameters utilised in the simulations.
24
25 An important point that should finally be commented is that the implementation of
26
27 the artificial (or fictitious) viscosity term added to the laminar shear stresses promotes
28
29 a correction to maintain the stability of the simulations and avoid interpenetration
30
31 between particles is being performed. When the artificial viscosity is applied in direct
32 10
33 substitution of the diffusive terms of the Navier-Stokes equations (often used in SPH
34
35 simulations) it is a purely computational model for the fluid viscosity without
36
37 physical meaning.
38
39
40
41 Declarations
42
43 Funding
44
45
46 Not applicable.
47
48 15
49 Conflict of Interest
50
51 The authors certify that they have no affiliations with or involvement in any
52
53 organization or entity with any financial interest or non-financial interest in the
54
55 subject matter or materials discussed in this manuscript.
56
57
58
59
60
61
62
63 33
64
65
Availability of data and material
1
2 65
3 The data and material that supports the findings of this study are available within the
4
5 article.
6
7
8
9 References
10
11
12 [1] Antuono M, Colagrossi A, Marrone S (2012) Numerical diffusive terms in weakly-
13
14 compressible SPH schemes. Comput. Phys. Commun. 183(12):2570-2580.
15
16 https://doi.org/10.1016/j.cpc.2012.07.006
17
18 [2] Biscarini C, Di Francesco S, Manciola P (2010) CFD modelling approach for
19
20 dam break flow studies. Hydrol. Earth Syst. Sci. 14:705–718.
21
22 https://doi.org/10.5194/hess-14-705-2010
23
24
25 [3] Bossuyt J, Meneveau C, Meyers J (2018). Large Eddy Simulation of a wind
26
27
tunnel wind farm experiment with one hundred static turbine models. Journal of
28
29
Physics: Conf. Series, 1037, Issue 6. https://doi.org/10.1088/1742-
30
31
6596/1037/6/062006
32
33 [4] Bruno D, De Serio F, Mossa M (2009) The FUNWAVE model application and its
34
35 validation using laboratory data. Coastal Engineering, 56(7): 773–787.
36
37 https://doi.org/10.1016/j.coastaleng.2009.02.001
38
39
40
[5] Chen Q, Kirby JT, Dalrymple RA, Kennedy AB, Chawla A (2000) Boussinesq
41 Modeling of Wave Transformation, Breaking, and Runup. II: 2D. Journal of
42
43 Waterway, Port, Coastal, and Ocean Engineering, 126(1).
44
45 https://doi.org/10.1061/(ASCE)0733-950X(2000)126:1(48)
46
47
48 [6] Chen Q, Kirby JT, Dalrymple RA, Wei F, Thornton EB (2003) Boussinesq
49
50 Modeling of Longshore Currents. Journal of Geophysical Research, 108 (C11).
51
52 https://doi.org/10.1029/2002JC001308
53
54 [7] Chow FK, Street RL (2009) Evaluation of Turbulence Closure Models for Large-
55
56 Eddy Simulation over Complex Terrain: Flow over Askervein Hill. Journal of
57
58 Applied Meteorology and Climatology, 48: 1050-1065.
59
60
61
62
63 34
64
65
https://doi.org/10.1175/2008JAMC1862.1
1
2
3
[8] Crespo AC, Dominguez JM, Barreiro A, Gómez-Gesteira M, Rogers BD (2006)
4 GPUs, a new tool of acceleration in CFD: efficiency and reliability on
5
6 Smoothed Particle Hydrodynamics methods, PLoS ONE 6 (6): e20685.
7
8 https://doi.org/10.1371/journal.pone.0020685
9
10
11 [9] Cruchaga MA, Celentano DJ, Tezduyar TE (2007) Collapse of a liquid column:
12
13 numerical simulation and experimental validation. Comput. Mech., 39(4):453-476.
14
15 https://doi.org/10.1007/s00466-006-0043-z
16
17 [10] Dalrymple RA, Rogers BD (2006) Numerical modeling of water waves with the
18
19 SPH method. Coast. Eng., 53: 141–147.
20
21 https://doi.org/10.1016/j.coastaleng.2005.10.004
22
23
24 [11] Fraga Filho CAD, Chacaltana JTA (2015) Study of Fluid Flows using Smoothed
25
26 Particle Hydrodynamics: the modified Pressure Concept Applied to Quiescent Fluid
27
28 and Dam Breaking. In: Proceedings of the XXXVI Ibero-Latin American Congress
29
30 of Computational Method in Engineering – CILAMCE 2015, November 22-25, Rio
31
32 de Janeiro, Brazil. https://doi.org/10.20906/CPS/CILAMCE2015-0071
33
34 [12] Fraga Filho CAD, Piccoli FP, Barbosa DA, Chacaltana JTA (2015b) Numerical
35
36 Study of the Propagation of Waves on Flat Beaches: An Application in Engineering
37
38 using SPHysics and FUNWAVE Models. In: Proceedings of the 23rd ABCM
39
40 International Congress of Mechanical Engineering, December 6-11, Rio de Janeiro,
41
42 RJ, Brazil. https://doi.org/10.20906/CPS/COB-2015-0561
43
44
45 [13] Fraga Filho CAD (2016) Development of a Computer Code using the
46
47 Lagrangian Smoothed Particle Hydrodynamics (SPH) Method for Solution of
48
49 Problems in Fluid Dynamics and Heat Transfer. In: Proceedings of the XXXVII
50
51 Ibero-Latin American Congress of Computational Method in Engineering –
52
53 CILAMCE 2016, November 6-9, Brasília, DF, Brazil.
54
55 Available at http://periodicos.unb.br/index.php/ripe/article/view/14444/12755,
56
57 accessed on June 30, 2020.
58
59
60
61
62
63 35
64
65
[14] Fraga Filho CAD, Chacaltana JTA (2016b) Boundary Treatment Techniques in
1
2 Smoothed Particle Hydrodynamics: Implementations in Fluid and Thermal Sciences
3
4 and Results Analysis, In: Proceedings of the XXXVII Ibero-Latin American
5
6 Congress of Computational Method in Engineering – CILAMCE 2016, Brasília, DF,
7
8 Brazil. Revista Interdiscisplinar de Pesquisa em Engenharia, 2(11). Available at
9
10 https://periodicos.unb.br/index.php/ripe/article/view/21270, accessed on June 25,
11
12 2020.
13
14 [15] Fraga Filho CAD (2017) Development of a computational instrument using a
15
16 lagrangian particle method for physics teaching in the areas of fluid dynamics and
17
18 transport phenomena. Rev. Bras. Ensino Fís. 39(4), e4401.
19
20 https://doi.org/10.1590/1806-9126-rbef-2016-0289
21
22
23 [16] Fraga Filho CAD (2017b) An algorithmic implementation of physical reflective
24
25 boundary conditions in particle methods: Collision detection and response. Physics of
26
27 Fluids, 29, 113602. https://doi.org/10.1063/1.4997054
28
29 [17] Fraga Filho CAD (2019) Smoothed Particle Hydrodynamics, Fundamentals and
30
31 Basic Applications in Continuum Mechanics. Springer Nature, Switzerland.
32
33
34 [18] Gabreil E, Tait SJ, Shao S, Nichols A (2018) SPHysics simulation of laboratory
35
36 shallow free surface turbulent flows over a rough bed. Journal of Hydraulic Research,
37
38 56(5):727-747. https://doi.org /10.1080/00221686.2017.1410732
39
40 [19] Gesteira MG, Rogers BD, Dalrymple RA, Crespo AJC, Narayanaswamy M
41
42 (2010) User Guide for SPHysics Code. University of Manchester, UK. Available at
43
44 https://wiki.manchester.ac.uk/sphysics/images/SPHysics_v2.2.000_GUIDE.pdf,
45
46 accessed on June 07, 2020.
47
48
49 [20] Gomez-Gesteira M, Rogers BD, Crespo AJC, Dalrymple RA, Narayanaswamy
50
51 M, Dominguez JM (2012) SPHysics - development of a free surface fluid solver -
52
53 Part 1: Theory and formulations. Computers & Geosciences, 48:289-299.
54
55 https://doi.org/10.1016/j.cageo.2012.02.029
56
57
58
59
60
61
62
63 36
64
65
[21] Green MD, Vacondio R, Peiró J (2019) A smoothed particle hydrodynamics
1
2 numerical scheme with a consistent diffusion term for the continuity equation.
3
4 Computers & Fluids 179:632-644. https://doi.org/10.1016/j.compfluid.2018.11.020
5
6 [22] Kennedy AB, Chen Q, Kirby JT, Dalrymple RA (2000) Boussinesq Modeling of
7
8 Wave Transformation, Breaking, and Runup. I: 1D. Journal of Waterway, Port,
9
10 Coastal, and Ocean Engineering, 126(1).
11
12 https://doi.org/10.1061/(ASCE)0733-950X(2000)126:1(39)
13
14
15 [23] Kirby JT, Lon W, Shi F (2005) FUNWAVE 2.0: Fully Nonlinear Boussinesq
16
17 Wave Model on Curvilinear Coordinates. Part I. Model Formulations. Delaware, US:
18
19 University of Delaware, Center for Applied Coastal Research, Dept. of Civil &
20
21 Environmental Engineering.
22
23 [24] Kirby JT, Lon W, Shi F (2005b) FUNWAVE 2.0: Fully Nonlinear Boussinesq
24
25 Wave Model on Curvilinear Coordinates. Part II. User’s Manual. Delaware, US:
26
27 Center for Applied Coastal Research, Dept. Of Civil & Environmental Engineering,
28
29 2005.
30
31
32 [25] Kirkil G, Mirocha J (2012) Implementation and Evaluation of Dynamic
33
34 Subfilter-Scale Stress Models for Large-Eddy Simulation Using WRF*. Monthly
35
36 Weather Review, 140:266-284. https://doi.org/10.1175/MWR-D-11-00037.1
37
38 [26] Kleefsman KMT, Fekken G, Veldman AEP, Iwanowski B, Buchner B (2005) A
39
40 Volume-of-Fluid based simulation method for wave impact problems. J. Comput.
41
42 Phys., 206:363-393. https://doi.org/10.1016/j.jcp.2004.12.007
43
44
45 [27] Lo EYM, Shao S (2002) Simulation of Near-Shore Solitary Wave Mechanics by
46
47 an Incompressible SPH Method, Appl. Ocean Res., 24(5):275-286.
48
49 https://doi.org/10.1016/S0141-1187(03)00002-6
50
51 [28] Mayrhofer A, Laurence D, Rogers BD, Violeau D (2015) DNS and LES of 3-D
52
53 wall-bounded turbulence using smoothed particle hydrodynamics. Comput. Fluids,
54
55 115: 86–97. https://doi.org/10.1016/j.compfluid.2015.03.029
56
57
58 [29] Menzies K (2009) Large eddy simulation applications in gas turbines. Phil.
59
60 Trans. R. Soc. A, 367: 2827–2838, 2009. https://doi.org/10.1098/rsta.2009.0064
61
62
63 37
64
65
[30] Monaghan JJ (1992) Smoothed Particle Hydrodynamics. Annual Rev. Astron.
1
2 Appl., 30: 543-574. https://doi.org/10.1146/annurev.aa.30.090192.002551
3
4 [31] Monaghan JJ (2000) SPH without tensile instability. J. Comput. Phys., 159(2):
5
6 290-311. https://doi.org/10.1006/jcph.2000.6439
7
8
9 [32] Padova D, Dalrymple RA, Mossa M (2014) Analysis of the artificial viscosity in
10
11 the smoothed particle hydrodynamics modelling of regular waves. Journal of
12
13 Hydraulic Research, 52(6), 836-848. https://doi.org/10.1080/00221686.2014.932853
14
15 [33] Peregrini DH (1998). Surf Zone Currents. Theoret. Comput. Fluid Dynamics,
16
17 10(1-4): 295-309. https://doi.org/10.1007/s001620050065
18
19
20 [34] Piomelli U (1999) Large-eddy simulation: achievements and challenges.
21
22 Progress in Aerospace Sciences, 35:335-362. https://doi.org/10.1016/S0376-
23
24 0421(98)00014-1
25
26 [35] Porte-Agel F, Meneveau C, Parlangé MB (2000) A scale-dependent dynamic
27
28 model for large-eddy simulation: application to a neutral atmospheric boundary layer.
29
30 J. Fluid Mech., 415:261-284. https://doi.org/10.1017/S0022112000008776
31
32
33 [36] Robinson M, Monaghan JJ (2012) Direct numerical simulation of decaying two-
34
35 dimensional turbulence in a no-slip square box using smoothed particle
36
37 hydrodynamics. Inte. J. Numer. Methods Fluids, 70:37–55.
38
39 https://doi.org/10.1002/fld.2677
40
41 [37] Shao SD, Gotoh H (2004) Simulating Coupled Motion of Progressive Wave and
42
43 Floating Curtain-Wall by SPH-LES Model, Coast. Eng. J. 46(2):171-202.
44
45 https://doi.org/10.1142/S0578563404001026
46
47
48 [38] Shao S, Ji C (2006) SPH Computation of Plunging Waves Using a 2-D Sub-
49
50 Particle Scale (SPS) Turbulence Model. Int. J. Numer. Methods Fluids, 51:913-936.
51
52 https://doi.org/10.1002/fld.1165
53
54 [39] Smagorinsky J (1963) General circulation experiments with the primitive
55
56 equations. I. The basic experiment. Mon. Wea. Rev., 91(3):99–164.
57
58 https://doi.org/10.1175/1520-0493(1963)091<0099:GCEWTP>2.3.CO;2
59
60
61
62
63 38
64
65
[40] Su M, Chen Q, Chiang C.-M (2001) Comparison of different subgrid-scale
1
2 models of large eddy simulation for indoor airflow modelling. Journal of Fluids
3
4 Engineering, 123:628-639. https://doi.org/10.1115/1.1378294
5
6 [41] Violeau D, Issa R (2007) Numerical modelling of complex turbulent free-surface
7
8 flows with the SPH method: an overview. Inte. J. Numer. Methods Fluids, 53:277-
9
10 304. https://doi.org/10.1002/fld.1292
11
12
13 [42] Wei G, Kirby JT, Grilli ST, Subramanya R (1995) A Fully Nonlinear Boussinesq
14
15 Model Equations for Surface Waves. Part 1. Highly Nonlinear, Unsteady Waves, J.
16
17 Fluid Mech., 294:71-92. https://doi.org/10.1017/S0022112095002813
18
19 [43] Zheng X, Ma Q, Shao S (2018) Study on SPH Viscosity Term Formulations.
20
21 Appl. Sci., 8(2), 249. https://doi.org/10.3390/app8020249
22
23
24 [44] Zhiyin Y (2015) Large-eddy simulation: Past, present and the future. Chinese
25
26 Journal of Aeronautics, 28(1): 11–24. https://doi.org/10.1016/j.cja.2014.12.007
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63 39
64
65
APPENDIX
1
2
3 Boussinesq-type nonlinear wave equations
4
5 2
6 These equations are obtained after integration of mass and momentum conservation
7
8 equations in the vertical direction [42].
9
10 Continuity equation:
11
12 η Mi
13 + =0 (A1)
14 tx i
15
16 where the flux M i is:
17
18
19 11 uj
20 = h + δη u +2 z 2
Mii h2 hδη+δ2η2 +
26 xx
ij
21
22 (A2)
23
24 1 hu j
24
z h δη +O
25 2 xx
ij
26
27
28 Momentum equation:
29
30
31 uu η
32
ii
+δ uj + + J24
+δ
i
2 =O
K i (A3)
t x jix
33
34
35 where the non-linear terms Ji and Ki are given as:
36
37
38 1 uu
jj
39 Ji = z 2 +z h
2 xi xj t xi x j t
40
41
42 (A4)
43 1 2 uu
jj
44 δη +δη h
45 xi 2 xj t xj t
46
47
48
49
hu j 1 uj
Ki = z δη u j + z2 δ 2η2 u j +
50 xi xj xj 2 xj xj
51
52
53 (A5)
54 2
55 hu j uj
56
1
+ + δη
57 2 xi xj xj
58
59
60
61
62
63 40
64
65
where the level z is the reference depth for calculating the velocities, which may be
1 given as:
2
3
4
z 0.531h 0 (A6)
5
6 where:
7 45
8
is the free surface elevation
9 h is the total water depth
10
11 t is the time
12
13 a0 h0 and k0 h0 are scales of nonlinearity and dispersion, respectively
14
15 , h0 , are, in sequence, the typical wave amplitude, the depth of water at rest and
ak
00
16
17 the number of waves
18
75
19 ui (u,v) is the velocity at depth in the coordinated zu
: i , being the
20 xi
21
22 velocity potential.
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63 41
64
65
6
6
Figure 1 Click here to access/download;Figure;Figure1.pdf
Permission to publish the figure 1 Click here to
17/08/2020 access/download;Attachment;PERMISSION_FIGURE_SPRIN
RightsLink Printable License
37
SPRINGER NATURE LICENSE
TERMS AND CONDITIONS
Aug 17, 2020
This Agreement between Federal Institute of Education, Science and Technology of Espírito
14
Santo -- Carlos Alberto Fraga Filho ("You") and Springer Nature ("Springer Nature")
consists of your license details and the terms and conditions provided by Springer Nature
and Copyright Clearance Center.
License Number 4891591305773
License date Aug 17, 2020
Licensed Content Publisher Springer Nature
Licensed Content Publication Springer eBook
Licensed Content Title Smoothed Particle Hydrodynamics Method
Licensed Content Author Carlos Alberto Dutra Fraga Filho
93
Licensed Content Date Jan 1, 2019
12
Type of Use Journal/Magazine
Requestor type academic/university or research institute
Is this reuse sponsored by or

associated with a pharmaceutical or a no
medical products company?
Format print and electronic
Portion figures/tables/illustrations
Number of figures/tables/illustrations 1
18
https://s100.copyright.com/AppDispatchServlet 1/6
17/08/2020 RightsLink Printable License
Will you be translating? no
Circulation/distribution 10000 - 19999
Author of this Springer Nature

yes
content
Diffusive Terms Applied in Smoothed Particle

Title of new article Hydrodynamics Simulations of Incompressible and
Isothermal Newtonian Fluid Flows
3
Lead author Carlos Alberto Dutra Fraga Filho
Title of targeted journal Physics of Fluids
44
Publisher American Institute of Physics
Expected publication date Sep 2020
Order reference number 1
Portions Figure 3.1 on page 18
30
Federal Institute of Education, Science and
Technology of Espírito Santo
Av Vitória, 1729 - Jucutuquara
Requestor Location
Vitória, Espírito Santo 29040-780
Brazil
Attn: Federal Institute of Education, Science and
Technology of Espírito Santo
1
Total 0.00 USD
Terms and Conditions
Springer Nature Customer Service Centre GmbH

Terms and Conditions
This agreement sets out the terms and conditions of the licence (the Licence) between you
and Springer Nature Customer Service Centre GmbH (the Licensor). By clicking
'accept' and completing the transaction for the material (Licensed Material), you also
confirm your acceptance of these terms and conditions.
35
1. Grant of License
1. 1. The Licensor grants you a personal, non-exclusive, non-transferable, world-wide

licence to reproduce the Licensed Material for the purpose specified in your order
only. Licences are granted for the specific use requested in the order and for no other
use, subject to the conditions below.
1. 2. The Licensor warrants that it has, to the best of its knowledge, the rights to
license reuse of the Licensed Material. However, you should ensure that the material
you are requesting is original to the Licensor and does not carry the copyright of
another entity (as credited in the published version).
1. 3. If the credit line on any part of the material you have requested indicates that it
was reprinted or adapted with permission from another source, then you should also
seek permission from that source to reuse the material.
2. Scope of Licence
2. 1. You may only use the Licensed Content in the manner and to the extent permitted
by these Ts&Cs and any applicable laws.
2. 2. A separate licence may be required for any additional use of the Licensed
Material, e.g. where a licence has been purchased for print only use, separate
permission must be obtained for electronic re-use. Similarly, a licence is only valid in
the language selected and does not apply for editions in other languages unless
additional translation rights have been granted separately in the licence. Any content
owned by third parties are expressly excluded from the licence.
2. 3. Similarly, rights for additional components such as custom editions and

derivatives require additional permission and may be subject to an additional fee.
Please apply to
Journalpermissions@springernature.com/bookpermissions@springernature.com for
these rights.
2. 4. Where permission has been granted free of charge for material in print,
permission may also be granted for any electronic version of that work, provided that
the material is incidental to your work as a whole and that the electronic version is
essentially equivalent to, or substitutes for, the print version.
2. 5. An alternative scope of licence may apply to signatories of the STM Permissions

Guidelines, as amended from time to time.
3. Duration of Licence
3. 1. A licence for is valid from the date of purchase ('Licence Date') at the end of the
relevant period in the below table:
Scope of Licence Duration of Licence

Post on a website 12 months
Presentations 12 months
Books and journals Lifetime of the edition in the language purchased
4. Acknowledgement
4. 1. The Licensor's permission must be acknowledged next to the Licenced Material in

print. In electronic form, this acknowledgement must be visible at the same time as the
figures/tables/illustrations or abstract, and must be hyperlinked to the journal/book's
homepage. Our required acknowledgement format is in the Appendix below.
5. Restrictions on use
5. 1. Use of the Licensed Material may be permitted for incidental promotional use and
minor editing privileges e.g. minor adaptations of single figures, changes of format,
colour and/or style where the adaptation is credited as set out in Appendix 1 below. Any
other changes including but not limited to, cropping, adapting, omitting material that
affect the meaning, intention or moral rights of the author are strictly prohibited.
5. 2. You must not use any Licensed Material as part of any design or trademark.
5. 3. Licensed Material may be used in Open Access Publications (OAP) before

publication by Springer Nature, but any Licensed Material must be removed from OAP
sites prior to final publication.
6. Ownership of Rights
6. 1. Licensed Material remains the property of either Licensor or the relevant third party
and any rights not explicitly granted herein are expressly reserved.
7. Warranty
IN NO EVENT SHALL LICENSOR BE LIABLE TO YOU OR ANY OTHER PARTY OR

ANY OTHER PERSON OR FOR ANY SPECIAL, CONSEQUENTIAL, INCIDENTAL
OR INDIRECT DAMAGES, HOWEVER CAUSED, ARISING OUT OF OR IN
CONNECTION WITH THE DOWNLOADING, VIEWING OR USE OF THE
MATERIALS REGARDLESS OF THE FORM OF ACTION, WHETHER FOR BREACH
OF CONTRACT, BREACH OF WARRANTY, TORT, NEGLIGENCE, INFRINGEMENT
OR OTHERWISE (INCLUDING, WITHOUT LIMITATION, DAMAGES BASED ON
LOSS OF PROFITS, DATA, FILES, USE, BUSINESS OPPORTUNITY OR CLAIMS OF
THIRD PARTIES), AND
WHETHER OR NOT THE PARTY HAS BEEN ADVISED OF THE POSSIBILITY OF
SUCH DAMAGES. THIS LIMITATION SHALL APPLY NOTWITHSTANDING ANY
FAILURE OF ESSENTIAL PURPOSE OF ANY LIMITED REMEDY PROVIDED
HEREIN.
8. Limitations
8. 1. BOOKS ONLY:Where 'reuse in a dissertation/thesis' has been selected the

following terms apply: Print rights of the final author's accepted manuscript (for clarity,
NOT the published version) for up to 100 copies, electronic rights for use only on a
personal website or institutional repository as defined by the Sherpa guideline
(www.sherpa.ac.uk/romeo/).
18
9. Termination and Cancellation
9. 1. Licences will expire after the period shown in Clause 3 (above).
9. 2. Licensee reserves the right to terminate the Licence in the event that payment is not
received in full or if there has been a breach of this agreement by you.
Appendix 1 — Acknowledgements:
For Journal Content:

Reprinted by permission from [the Licensor]: [Journal Publisher (e.g.
Nature/Springer/Palgrave)] [JOURNAL NAME] [REFERENCE CITATION
(Article name, Author(s) Name), [COPYRIGHT] (year of publication)
For Advance Online Publication papers:

Reprinted by permission from [the Licensor]: [Journal Publisher (e.g.
(Article name, Author(s) Name), [COPYRIGHT] (year of publication), advance
online publication, day month year (doi: 10.1038/sj.[JOURNAL ACRONYM].)
For Adaptations/Translations:
Adapted/Translated by permission from [the Licensor]: [Journal Publisher (e.g.
(Article name, Author(s) Name), [COPYRIGHT] (year of publication)
Note: For any republication from the British Journal of Cancer, the following
credit line style applies:
Reprinted/adapted/translated by permission from [the Licensor]: on behalf of Cancer

Research UK: : [Journal Publisher (e.g. Nature/Springer/Palgrave)] [JOURNAL
NAME] [REFERENCE CITATION (Article name, Author(s) Name),
[COPYRIGHT] (year of publication)
For Advance Online Publication papers:

Reprinted by permission from The [the Licensor]: on behalf of Cancer Research UK:
[Journal Publisher (e.g. Nature/Springer/Palgrave)] [JOURNAL NAME]
[REFERENCE CITATION (Article name, Author(s) Name), [COPYRIGHT] (year
of publication), advance online publication, day month year (doi: 10.1038/sj.
[JOURNAL ACRONYM])
For Book content:

Reprinted/adapted by permission from [the Licensor]: [Book Publisher (e.g.
Palgrave Macmillan, Springer etc) [Book Title] by [Book author(s)]
[COPYRIGHT] (year of publication)
Other Conditions:
Version 1.2
Questions? customercare@copyright.com or +1-855-239-3415 (toll free in the US) or

+1-978-646-2777.
18
18
78
78
Permission to publish the figure 10 Click here to
access/download;Attachment;PERMISSION_FIGURE_PUBLI
SSN 2447-6102
th
24 August, 2020.
To whom it may concern

33
All works published in Revista de Pesquisa Interdisciplinar em Engenharia (RIPE) are licensed under a Creative
7
Commons Attribution-No Derivatives 4.0 International License. According to CC BY-ND 4.0 authors are free to share
copy and redistribute the material in any medium or format for any purpose, even commercially. Under the
following terms:
Attribution — You must give appropriate credit, provide a link to the license, and indicate if changes were made.
You may do so in any reasonable manner, but not in any way that suggests the licensor endorses you or your use.
No Derivatives — If you remix, transform, or build upon the material, you may not distribute the modified
material.
2
As aforementioned the work "Boundary Treatment Techniques in Smoothed Particle Hydrodynamics:
Implementations in Fluid and Thermal Sciences and Results Analysis” authored by Carlos Alberto Dutra Fraga Filho
and Julio Tomás Aquije Chacaltana, can have their results partially reproduced since it follows the CC BY-ND 4.0
rules.
Kind regards,
Prof. Carla Anflor, Dr. Eng.

Editor-in-Chief
55
Revista de Pesquisa Interdisciplinar em Engenharia
RIPE 2020 https://www.periodicos.unb.br/index.php/ripe

Diffusive Terms Applied in Smoothed Particle Hydrodynamics Simulations of Incompressible and Isothermal Newtonian Fluid Flows

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Diffusive Terms Applied in Smoothed Particle Hydrodynamics Simulations of Incompressible and Isothermal Newtonian Fluid Flows

Uploaded by

Copyright:

Available Formats

BMSE-D-21-00211_linear.

Diffusive Terms Applied in Smoothed Particle Hydrodynamics Simu…

Excluded search repositories:

Excluded from Similarity Report:

Diffusive Terms Applied in Smoothed Particle Hydrodynamics Simulations

Aug 17, 2020

License Number 4891591305773

License date Aug 17, 2020

Licensed Content Publisher Springer Nature

Licensed Content Publication Springer eBook

Licensed Content Title Smoothed Particle Hydrodynamics Method

Licensed Content Author Carlos Alberto Dutra Fraga Filho

Requestor type academic/university or research institute

Is this reuse sponsored by or

Format print and electronic

Will you be translating? no

Circulation/distribution 10000 - 19999

Author of this Springer Nature

Diffusive Terms Applied in Smoothed Particle

Title of targeted journal Physics of Fluids

Expected publication date Sep 2020

Order reference number 1

Portions Figure 3.1 on page 18

Terms and Conditions

Springer Nature Customer Service Centre GmbH

1. 1. The Licensor grants you a personal, non-exclusive, non-transferable, world-wide

2. 3. Similarly, rights for additional components such as custom editions and

2. 5. An alternative scope of licence may apply to signatories of the STM Permissions

Scope of Licence Duration of Licence

4. 1. The Licensor's permission must be acknowledged next to the Licenced Material in

5. 3. Licensed Material may be used in Open Access Publications (OAP) before

IN NO EVENT SHALL LICENSOR BE LIABLE TO YOU OR ANY OTHER PARTY OR

8. 1. BOOKS ONLY:Where 'reuse in a dissertation/thesis' has been selected the

9. Termination and Cancellation

9. 1. Licences will expire after the period shown in Clause 3 (above).

For Journal Content:

For Advance Online Publication papers:

Reprinted/adapted/translated by permission from [the Licensor]: on behalf of Cancer

For Advance Online Publication papers:

For Book content:

Questions? customercare@copyright.com or +1-855-239-3415 (toll free in the US) or

To whom it may concern

Prof. Carla Anflor, Dr. Eng.

RIPE 2020 https://www.periodicos.unb.br/index.php/ripe

You might also like