You do not have permission to edit this page, for the following reason:

You are not allowed to execute the action you have requested.


You can view and copy the source of this page.

x
 
1
<!-- metadata commented in wiki content
2
3
4
==A river bed hydrokinetic turbine - a laminated composite material rotor design==
5
6
SERGIO A. OLLER ARAMAYO, Professor, ''INIQUI – CONICET –Faculty of Engineering, National University of Salta, Salta, 4400, Argentina''
7
8
''Email: ''[mailto:sergio.oller@conicet.gov.ar ''sergio.oller@conicet.gov.ar'']'' ''
9
10
LIZ G. NALLIM, Professor,'' INIQUI – CONICET –Faculty of Engineering, National University of Salta, Salta, 4400, Argentina ''
11
12
''Email: ''[mailto:lnallim@unsa.edu.ar ''lnallim@unsa.edu.ar'']'' ''
13
14
SERGIO OLLER, Professor,'' UPC-CIMNE – International Center for Method in Engineering, UPC Technical University of Catalonia (Barcelona Tech), Barcelona 08034, Spain''
15
16
''Email: ''[mailto:sergio.oller@upc.edu ''sergio.oller@upc.edu'']'' (corresponding author)''
17
18
XAVIER MARTINEZ, Professor,'' UPC-CIMNE – International Center for Method in Engineering, UPC Technical University of Catalonia (Barcelona Tech), Barcelona 08034, Spain''
19
20
''Email: ''[mailto:x.martinez@upc.edu ''x.martinez@upc.edu'']
21
22
<span id='_GoBack'></span>
23
24
Keywords: Composite materials; Fiber reinforced laminates; Mixing theory; Constitutive models; Finite element method (FEM); Damage models; Computational Mechanics; Water current turbines (WCT); Rotor turbine design and analysis; Riverbed operation.
25
-->
26
27
==ABSTRACT==
28
29
This chapter presents the composite materials applied to Water Current Turbine (WCT) hydrokinetic turbines. Here will be briefly described the features of these turbines, the fluid-dynamic behavior of the rotor, and its structure formed into a composite material. From the structural viewpoint an advanced composite material formulation that allows an appropriate structural design is introduced. The generalized composite formulations here introduced take into account the nonlinear mechanical behavior of the component materials (matrix and fiber), as the local behavior of plasticity and damage, its anisotropy, the fiber matrix debonding, its material composition via a general mixing theory, and also the homogenized structural damage index definition.
30
31
Hydrokinetic turbines bring newer advantages and greater possibilities for green hydroelectric power generation. For this reason, achieving a very high lift blade rotor to take the maximum kinetic energy advantage for rivers with a slow velocity flow is very important. A very low inertia rotor permits a self-starting effect for the axial water flow turbine to take the maximum advantage of the river kinetic energy which is very important in this kind of devices. A turbine rotor hydrofoil made in composite material can be designed for this purpose.
32
33
One of the most commonly used composite material analysis formulation is herein introduced. Specifically, a particular Serial/Parallel (S/P) Mixing Theory with a better relation between model accuracy vs. computational cost is provided. In front to other formulation, the S/P Mixing Theory not increasing the degrees of freedom of the problem because is a constitutive formulation.
34
35
A brief introduction to fluid-dynamic concept involving in the analysis of a rotor of this type of turbines is presented. This allows seeing the origin of the actions applied to the rotor of this type of turbines.
36
37
In addition, two simple examples that show the potentiality of the model are presented in this chapter. Then, an application to the design of a rotor blade of a passing turbine, made of carbon fiber-reinforced matrix composite material, is shown.
38
39
==1 <br/>General Introduction==
40
41
According to United Nations, 20% of the global population does not have access to electricity, and a further 14% lack reliable access [54]. The use of an axial flow rotor turbine in remote area was claimed to have for pumping irrigation and electrical power generation. Hydrokinetic turbines bring newer, greater possibilities and advantages for hydroelectric power generation. There are applications in water currents of 0.5 m/s in more [53]. Development of renewal energy production in rivers and channels still preserve a very   interesting   power   production potential, being not subjected to the classical hydraulic power exploitation. This solution avoids the construction of expensive dams and reduces considerably the environmental impact produced by classical hydropower generation [52]. Low speed flux and lack of depth are the main obstacles in hydrokinetic operation. For this reason, achieving a very high lift rotor to take the maximum advantage of the kinetic energy of a slow velocity water flow, which belongs to a lowland river type, is a very important topic. The use of a high lift aerodynamic/hydrodynamic profile and composite material for the blades serve to accomplish the task.
42
43
The main purpose of this chapter is to describe a general procedure to achieve a very low inertia rotor minimizing the start-stop effect for the axial water flow turbine, in which is important to take the maximum advantage of the kinetic energy. The composite hydrofoil of the turbine rotor can be designed using reinforced laminate composites, to obtain the maximum strength and lower rotational-inertia. The mechanical and geometrical parameters involved in the design of this fiber-reinforced composite material are the fiber orientation, number of layers, stacking sequence and laminate thickness.
44
45
For this reason, it will be briefly described the features of hydrokinetic turbines (WCT - Water Current Turbine) for river use, their basic design requirements and the response by using matrix-reinforced composite structures. Design requirements for these turbines need a numerical process simulation of the fluid dynamic problem coupled with the behavior of the structure made of composite materials. From the structural viewpoint it is necessary the use of an advanced composite material formulation that allows an appropriate structural design. For this purpose, a "mixing theory" [1,2,7,8,27,29,30,31,32,38,39,42,50] and / or "homogenization theory" [3,4,5,9,43,47,48,49,51] of simple substances are used, with a mapping spaces formulation [6] that allow considering the anisotropy of the constituent and composite materials in the most general possible way, and a fiber matrix debonding formulation [2,29,31,39]. Moreover, within these general formulations, it is also taken into account the nonlinear mechanical behavior of the component materials (matrix and fiber), which allows to know precisely the limits of participation of each one of them into the composite.
46
47
The study of composite materials has been one of the major objectives of computational mechanics in the last decade. The numerical simulation of orthotropic composite materials has been done by means the average properties of their constituents, but this approximation, no model has been found able to work beyond the constituents elastic limit state. Thus, these procedures are limited to the numerical computation to elastic cases. Different theories have been proposed to solve this problem, taking into account the internal configuration of the composite to predict its behavior. The two most commonly used are herein remarked.
48
49
:* Homogenization theory: This method deals with the global problem of composite material in a two-scale context. The macroscopic scale uses the composite materials to obtain the global response of the structure; composites are considered as homogeneous materials in this scale. The microscopic scale corresponds to an elemental characteristic volume in which the microscopic fields inside the composite are obtained; this scale deals with the component materials. Homogenization theory assumes a periodical configuration of the composite material to relate these two scales [3,4,5,9,43,47,48,49,51].
50
51
:* Mixing theory''':''' The first formulation of the mixing theory corresponds to Trusdell and Toupin [7] and it is based in two main hypothesis: 1. All composite constituents have the same strains. 2. Each constituent collaborates to the composite behavior according to its volumetric participation. The main problem of the mixing theory is the iso-strain condition, which forces a parallel distribution of the constituents in the composite. Some improvements to the original formulation can be found in [1,2,7,27,29,30,31,32,38,39,42,50].
52
53
In this chapter a brief introduction to the “Serial / Parallel theory of mixtures”, a more advanced formulation than the classic one, is presented. The election of the mixing theory instead of a homogenization theory is based in the better relation between model accuracy vs. computational cost provided by the former one [5]. A homogenization theory requires a micro-model for each point of the structure that becomes non-linear. Despite the advances made in strategies to reduce the amount of micro-models solved [9,44,46,48,51], the resolution of a real structure with this procedure generates such a big amount of degrees of freedom that the calculation is beyond the computation capabilities of nowadays personal computers. On the other hand, the mixing theory does not increase the degrees of freedom of the problem, as it is only present in the constitutive section of the finite element code.
54
55
==2 Hydrokinetic Turbines - Introduction ==
56
57
This section provides an overview of a "Water Current Turbine" (WCT) allowing understand the hydrodynamic basis for their design and its requirements for the structural function [10,11,12,13,14,15,16,15]. Then, Section 5 presents the basis for the analysis of its structure made up in a reinforced composite material, and a simple application in the examples section are shown too.
58
59
Rivers kinetic energy for electric power generation is a very valuable alternative source. This emerging class of renewable energy technology, the hydrokinetic conversion device (HCD), offers ways to capture the energy of flowing water without the impoundment or diversion of the conventional hydroelectric facilities based on dams and penstocks. Hydrokinetic technologies are designed for deployment in natural streams, like rivers, tidal estuaries, ocean currents, and in some constructed waterways [10,11,12,13,14,15,16,15]. As opposed to the rigid, expensive, and environmentally aggressive construction of tidal barrages, the modularity and scalability of hydrokinetic devices are attractive features [11].
60
61
River streams and other artificial channels have potential for generating electric power through several hydrokinetic energy technologies. This nascent class of renewable energy technology is being strongly considered as an exclusive and unconventional solution falling within the area of both in-land water resource and marine energy [12]. Conventional large or small hydroelectric systems use reservoirs and penstocks to create an artificial water head and extract the potential energy of downwardly falling water through suitable turbo-machinery. In contrast, a river turbine, which could be built as a free-rotor or part of a channel augmented system, may provide an effective alternative mean for generating power. Such systems would potentially require little or no civil work, causing less environmental impact [13,14].
62
63
Khan, Iqbal and Quaicoe [13] showed values that indicate the possibility of higher energy capacity through a river turbine when compared to an equally sized wind energy converter.  Wind turbines are usually designed to operate with rated wind speed of 11–13 m/s while, in contrast, river turbines with augmentation channels could be designed for low effective water velocities of 1.75–2.25 m/s or even higher, depending on site resources.
64
65
Unlike wind energy, the size of these engines is a limitation for this type of energy generation and must be reduced according to the river depth. Another drawback is the low flow velocity, and it requires a set of blades and rotor with a specific design to generate the greater amount of kinetic energy as possible from the water flow [15].
66
67
This chapter describes a general procedure for an efficient fluid-mechanical design of the rotor´s blades. The use of high lift airfoils, and composite materials structural design for low rotational inertia, guarantees the hydrodynamics efficiency. Thus, the chapter is structured taking into account the analysis of this axial hydrokinetic river turbine as the fluid dynamic design of the rotor turbine; and the structural design of the rotor by composite materials. These areas converge in a multidisciplinary methodology depicted in Figures 5.
68
69
==3 Rotor hydrokinetic turbine design. Fluid-Solid interaction==
70
71
Hydrokinetic turbines, unlike conventional hydraulic turbines utilize the kinetic energy of river/channels water for power generation. The performance of these turbines depends of the number of blades, tip speed ratio, type of airfoil, blade pitch, chord length and twist and its distribution along the blade span [55].
72
73
Knowing the inlet and outlet pressure in the micro-scale volume control (VC1), a procedure for the rotor design of a hydrokinetic turbine for riverbed operation is described in this section. The study is focused on the conditions of a standard large-medium sized lowland river. The structural analysis of this rotor engineered in composite materials with reduced inertia and better functionality for low speed currents fluvial beds, is described in Sections 4 and 5. The results of the numerical simulation of the composite rotor structure can also be found in the Section of examples.
74
75
===3.1 Hydrofoil profile and rotor===
76
77
Inside of the micro-scale control volume (VC2) a composite material rotor turbine is placed. A brief hydrofoil design of its profile is here presented.
78
79
The supplied turbine power  <math display="inline">W\mbox{ }</math> is directly proportional to the machine’s operating angular speed  <math display="inline">\omega \mbox{ }</math> and its torque  <math display="inline">T\mbox{ }</math> produced at that specific speed,
80
81
{| class="formulaSCP" style="width: 100%; text-align: center;" 
82
|-
83
| 
84
{| style="text-align: center; margin:auto;" 
85
|-
86
| <math display="inline">W=T.\omega </math>
87
|}
88
| style="width: 5px;text-align: right;white-space: nowrap;" | (1)
89
|}
90
91
92
<div class="center" style="font-size: 75%;">
93
[[File:Draft_Samper_893761715_4505_Fig1.png]]
94
95
'''Figure 1'''. Hydrofoil S1223 profile.
96
</div>
97
98
99
If more lift is obtained by one blade, more torque and angular velocity will be obtained by the turbine. This commitment is achieved by selecting the S1223 foil [16], which belongs to the high lift low Reynolds profiles class (see Figure 1). Initially designed as an airfoil for air working conditions, the S1223 profile has also been tested as a hydrofoil under water conditions operation, showing very good operational qualities [17]. A rotor with S1223 hydrofoil profile keeps the proper balance between lift and drag and maintains an attached flow in the hydrofoil neighborhood. In consequence, this rotor has a better pressure distribution and presents hydrodynamic stability, preventing interference with the rest of the hydrofoils forming the rotor.
100
101
===3.2 Simplified hydrofoil analytical pre-design===
102
103
104
For turbine application, hydrofoil must be designed starting from the premise that it has to maintain fluid mechanics parameters (such as angle of attack, homogeneous pressure distribution, etc.) along the whole wingspan, despite the fact that rotary operating conditions produce different linear velocity of rotation ( <math display="inline">\dot{u}</math> ) along the blade axis (which gets higher the nearer the point is from the wingtip). Working with this condition involves the variation of the blade geometry parameters (like camber angle, airfoil chord, etc.) in relation with the wingspan axis. Figure 2 shows the notation for angles and velocities on the blade profile, where <math display="inline">v</math> is the absolute flow velocity in the micro-scale VC2 volume control,  <math display="inline">\dot{u}</math> represents the blade’s linear rotational speed and <math display="inline">c</math> is the relative flow velocity.
105
106
107
[[Image:Draft_Samper_893761715-picture-Grupo 243718.svg|center|600px]]
108
109
110
The angle of attack &#x03b1; is an aero-hydrodynamic angle defined between <math display="inline">c</math> and the airfoil chord, and depends on the airfoil profile and its camber angle ''&#x03b2;''. Instead, camber angle ''&#x03b2;  ''represents a mechanical angle, defined between the hydrofoil chord and its plane of rotation. By combining hydrodynamics and mechanical angles, the sustentation angle ( <math display="inline">\theta</math> ) appears which is very useful to obtain the variation camber angle in a rotating blade.
111
112
Parameters involving the use of a S1223 profile working as a non-twisted, non-rotatory and unturbined designed hydrofoil are explained below [17]. The suitable angle of attack occurs at the optimum angle of attack  <math display="inline">{\alpha }_{0}=</math><math>{10}^{\circ }</math>, which is considered as a starting parameter of the design sequence; it involves lift coefficient <math display="inline">{c}_{y}=</math><math>2.2</math> and drag coefficient <math display="inline">{c}_{x}=0.046</math>. Lift coefficient can raise until it reaches its maximum at <math display="inline">{\alpha }_{max}=</math><math>{15}^{\circ }</math>, but beyond that angle, detachment of the boundary layer will happen, dropping lift coefficient and increasing drag coefficient enormously [18].
113
114
The Tip Speed Ratio (TSR or <math display="inline">\lambda</math> ) is a non-dimensional parameter that is defined by taking the relationship between the absolute river flow axial flow velocity <math display="inline">c</math> and the blade speed turbine rotor  <math display="inline">\dot{u}</math> , and it is given by
115
116
{| class="formulaSCP" style="width: 100%; text-align: center;" 
117
|-
118
| 
119
{| style="text-align: center; margin:auto;" 
120
|-
121
| <math display="inline">\lambda =\left( \frac{\omega \cdot R}{v}\right)</math>
122
|}
123
| style="width: 5px;text-align: right;white-space: nowrap;" | (2)
124
|}
125
126
127
where  <math display="inline">R</math> is the rotor radius. According to Betz´s law [19], turbine mechanical power  <math display="inline">W</math> specified for axial turbines depends on the flux density  &#x03c1;  and flow speed  <math display="inline">v</math> in VC2 volume control; both values are fixed by the river flow, and so these parameters are fixed as initial conditions and will not be modified during the process of the rotor design. According to this, rotor nominal power can be established, and is computed from
128
129
{| class="formulaSCP" style="width: 100%; text-align: center;" 
130
|-
131
| 
132
{| style="text-align: center; margin:auto;" 
133
|-
134
| <math display="inline">W=\frac{\left( 8\cdot \rho \cdot A\cdot {v}^{3}\right) }{27}</math>
135
|}
136
| style="width: 5px;text-align: right;white-space: nowrap;" | (3)
137
|}
138
139
140
The swept area ( <math display="inline">A</math>) is the unique variable in Equation (3), and it depends  <math display="inline">R</math> (radius of the rotor).
141
142
Despite the rotating condition, it is necessary to maintain the angle of attack along the wingspan; this scenario permits to keep the rotor’s fluid dynamic stability. These commitments are accomplished by varying the geometry parameters of the hydrofoil chord size  <math display="inline">L</math> and camber angle  <math display="inline">\beta </math> , along the wingspan. To achieve this goal the Blade Element Theory can be used; according to Froude [20,21], the airfoil´s total length ( <math display="inline">X</math>) is split in several segments, and each one is designed individually as <math display="inline">x</math> (Figure 3). Sustentation angle  <math display="inline">\theta </math> (Figure 2) is obtained by means of Equation (4), as follows:
143
144
{| class="formulaSCP" style="width: 100%; text-align: center;" 
145
|-
146
| 
147
{| style="text-align: center; margin:auto;" 
148
|-
149
| <math display="inline">\theta =arccot\left( \frac{\left( \omega \cdot x\right) }{v}\right)</math>
150
|}
151
| style="width: 5px;text-align: right;white-space: nowrap;" | (4)
152
|}
153
154
155
The chord size of the airfoil is therefore computed for each segment  <math display="inline">L_x</math> by
156
157
{| class="formulaSCP" style="width: 100%; text-align: center;" 
158
|-
159
| 
160
{| style="text-align: center; margin:auto;" 
161
|-
162
| <math display="inline">L_x=\frac{SP_x\mbox{ }x}{c_y\mbox{ }n_{\mbox{wing}}}</math>
163
|}
164
| style="width: 5px;text-align: right;white-space: nowrap;" | (5)
165
|}
166
167
168
where  <math display="inline">n_{\mbox{wing}}</math> is the actual number of airfoils in the rotor,  <math display="inline">c_y</math> is the lift coefficient corresponding to a defined profile section at a certain radius <math display="inline">\, x</math>, and the airfoil shape factor <math display="inline">{SP}_{x}</math> can be computed by a curve approximation given by
169
170
{| class="formulaSCP" style="width: 100%; text-align: center;" 
171
|-
172
| 
173
{| style="text-align: center; margin:auto;" 
174
|-
175
| <math display="inline">{SP}_{x}=2.2762\cdot \left( {{SR}_{x}}^{-1.323}\right)</math>
176
|}
177
| style="width: 5px;text-align: right;white-space: nowrap;" | (6)
178
|}
179
180
181
<span id='OLE_LINK18'></span><span id='OLE_LINK19'></span><span id='OLE_LINK20'></span><span id='OLE_LINK15'></span><span id='OLE_LINK16'></span><span id='OLE_LINK17'></span>In Equation (6) the non-dimensional parameter <math display="inline">{SR}_{x}</math> is given by
182
183
{| class="formulaSCP" style="width: 100%; text-align: center;" 
184
|-
185
| 
186
{| style="text-align: center; margin:auto;" 
187
|-
188
| <math display="inline">{SR}_{x}=\frac{\left( TSR\cdot x\right) }{X}</math>
189
|}
190
| style="width: 5px;text-align: right;white-space: nowrap;" | (7)
191
|}
192
193
194
195
[[Image:Draft_Samper_893761715-picture-Grupo 243720.svg|center|600px]]
196
197
As a result of the chord modification during the process by Equation (5), the initial attack angle <math display="inline">{\alpha }_{0}</math>  has to be recalculated too through Equation (9), obtaining a new angle of attack <math display="inline">{\alpha }_{n}</math> for each chord <math display="inline">{L}_{x}</math> in each segment <math display="inline">x</math>. For this recalculation, the <math display="inline">KL\,</math> parameter, which represents a relationship between the wingspan and the average of the chord, <math display="inline">{L}_{avg}</math> is necessary,
198
199
{| class="formulaSCP" style="width: 100%; text-align: center;" 
200
|-
201
| 
202
{| style="text-align: center; margin:auto;" 
203
|-
204
| <math display="inline">KL=\frac{R}{{L}_{avg}}</math>
205
|}
206
| style="width: 5px;text-align: right;white-space: nowrap;" | (8)
207
|}
208
209
210
As the length camber and the chord angles have been modified for each wingspan segment, the angle of attack must be verified for each section through the following expression:
211
212
{| class="formulaSCP" style="width: 100%; text-align: center;" 
213
|-
214
| 
215
{| style="text-align: center; margin:auto;" 
216
|-
217
| <math display="inline">{\alpha }_{n}=-{\alpha }_{0}+\left( \frac{{c}_{y}}{0.11}\right) \cdot \left( 1+\right. </math><math>\left. \left( \frac{3}{KL}\right) \right)</math>
218
|}
219
| style="width: 5px;text-align: right;white-space: nowrap;" | (9)
220
|}
221
222
223
Solving from Equation (2) to Equation (9), the airfoil parameters can be obtained. These parameters allow the definitive design of the turbine blade. Blade parameters are strictly germane with the rotor composition through Equation (10), which represents the ideal number of blades  <math display="inline">n_{\mbox{wing}}</math> included in the rotor according to flow and geometry parameters. A higher torque on turbine will be obtained if a higher number of blades will be included in the rotor. This condition also simplifies the starting of the turbine, ergo is a good design requirement to have the more possible number of hydrofoils in the rotor
224
225
{| class="formulaSCP" style="width: 100%; text-align: center;" 
226
|-
227
| 
228
{| style="text-align: center; margin:auto;" 
229
|-
230
| <math display="inline">\frac{{n}_{wing}=\left( {SP}_{x}\cdot x\right) }{\left( {L}_{x}\cdot {c}_{y}\right) }</math>
231
|}
232
| style="width: 5px;text-align: right;white-space: nowrap;" | (10)
233
|}
234
235
===3.3 Brief comment about the 3-D fluid-dynamic numerical simulation of the hydrofoil blade.===
236
237
An ideal representation of the working turbine operation can be performed by a numerical simulation. The model is a confined fluid domain ( <math display="inline">d_x\times d_y\times d_z</math> ) rendering the underwater operation without free surface (VC2). The hydrofoil rotor is located inside that control volume that is made from a box of appropriate measures, as shown in Figure 4. The dimensions are chosen so as a steady flow is needed at the boundaries of the box. Flow with   <math display="inline">x</math> direction will cross from the inlet surface to the outlet surface. The rest of domain surfaces have wall condition.
238
239
240
<div class="center" style="font-size: 75%;">
241
[[File:Draft_Samper_893761715_2091_Fig4.png]]
242
243
'''Figure 4'''. VC2 Confined fluid domain control volume.
244
</div>
245
246
247
Fluid mechanics governing equations for incompressible flows ( <math display="inline">\rho =</math><math>cte</math>) involves mass conservation condition (Equation (4)),
248
249
{| class="formulaSCP" style="width: 100%; text-align: center;" 
250
|-
251
| 
252
{| style="text-align: center; margin:auto;" 
253
|-
254
| <math display="inline">\nabla \cdot v=0</math>
255
|}
256
| style="width: 5px;text-align: right;white-space: nowrap;" | (11a)
257
|}
258
259
and Navier-Stokes equation  [18,29],
260
261
{| class="formulaSCP" style="width: 100%; text-align: center;" 
262
|-
263
| 
264
{| style="text-align: center; margin:auto;" 
265
|-
266
| <math>\frac{\partial }{\partial t}\left(\rho \mbox{ }v_j\right)+\frac{\partial }{\partial x_i}\left(\rho \mbox{ }v_i\mbox{ }v_j\right)\mbox{ }\mbox{ }\mbox{ }=\mbox{ }\mbox{ }\mbox{ }-\frac{\partial p}{\partial x_j}+\frac{\partial }{\partial x_j}\mu \left(\frac{\partial v_i}{\partial x_j}+ \frac{\partial v_j}{\partial x_i}\right)+\rho g_j+F_j</math>
267
|}
268
| style="width: 5px;text-align: right;white-space: nowrap;" | (11b)
269
|}
270
271
where <math display="inline">v</math> is the velocity field inside the VC2 control volume.
272
273
Notice that the governing equations system is constituted by four equations and four unknowns, which are pressure and the three vector components of the field velocity; so numerical techniques are necessary for this treatment. A Finite Element Variational Multiscale Simulation method (FEVMS) [22,23] can be applied as the resolution method.
274
275
==4 The rotor structural design of a hydrokinetic turbine. A composite material structure as a solution==
276
277
Multi-laminated composite structures are an ever-increasingly important topic in the fields of fabrication of mechanical, aerospace, marine, and machinery industries due to their advantages such as durability (no corrosion – lower maintenance cost), survivability (fire resistance, crash energy absorption), excellent resistance against cyclic loading (low fatigue), reparability (restoration and repair), etc. [24]. Multi-layered fiber-reinforced material systems can offer versatility in composite design because the stacking sequence of each orthotropic layer can take full advantage of the superior mechanical properties in terms of its strength, stiffness, and total weight. One of the goals in design of multi-layered composite structure is to increase its strength while lowering its weight/rotational inertia with a given set of fibrous materials. Laminate of fiber-reinforced composites are very useful when low weight/rotational inertia together high strength/stiffness are required, like the case of axial water turbines. As an additional advantage, it is possible to fit the weight without downgrade the efficiency through the design of the fiber orientation, fraction reinforcement volume, choice of large or short fibers, layer thickness and stacking sequence.
278
279
A compact hydrokinetic turbine design, its condition of axial flow and its low rotational inertia due to the composite material rotors, confer the functionality at low speed fluvial beds, avoiding the requirement of great earthworks and expensive civil constructions.
280
281
Next section describes the structural design and analysis of a turbine rotor made-up of a laminate of fiber-reinforced composites material using the serial/parallel mixing theory [1,39]. This formulation manages several linear and non-linear constitutive models simultaneously including damage and plasticity behavior and provide the homogenized damage composite index, taking into account the orthotropic fiber-matrix reinforced and its debonding effect. The composite material used is a laminate composed by epoxy matrix reinforced with long carbon fibers, allowing obtaining better values of stiffness and strength with a smaller weight and rotational-inertia.
282
283
For the structural analysis is necessary to consider a “composition of several single anisotropic material”, together with “fluid-dynamic interaction” by means a “staggered technique approach”. These three aspects are depicted in Figure 5; the proper coupling allows the fiber-reinforced composites rotor design taking into account the successive geometric and mechanical changes of each component materials that forming the composite. Section 3 describe the procedure for the rotor fluid-dynamic analysis and obtain the state of pressures and speeds will be applied on each point of the rotor blades. Thus, the fluid pressures and speeds distribution in the axial camera of the turbine are input data for the rotor structural analysis, employing composite materials. These operations can be carried out through a “staggered” procedure [25,38], solving each problem per time as shown in Figure 5.
284
285
286
<div class="center" style="width: auto; margin-left: auto; margin-right: auto;font-size:75%;">
287
[[File:Draft_Samper_893761715_2949_Fig5.png|center|600px]]
288
289
'''Figure 5''': Flow diagram of solid-fluid interaction of numeric simulation.
290
</div>
291
292
<span id='_Ref508791362'></span><span id='_Ref508888539'></span>
293
294
<span id='_Toc277318150'></span>
295
296
==5 Numerical model for the analysis of composite material rotor==
297
298
This section presents the fundamental concepts for the structural design of turbine rotor hydrofoil in fiber reinforced composite material [39]. For this purpose several constitutive damage models [2,39,40] managed by an orthotropic Serial/Parallel mixing theory [1,39,41, 42,39], anisotropy mapping space [6,39,40], and fiber debonding strategy [39]  are explained. Assembling these numerical models provide a powerful formulation and allows the evaluation of a global homogenized laminate damage index, which come from the damage index provided by the local constitutive damage model.
299
300
This section contains the presentation and explanation of the following formulations: Orthotropic Serial/Parallel mixing theory for the laminate composition material, Constitutive plastic-damage model for a single material, Constitutive damage model for a single material, Some numerical strategies, and  Homogenized laminate damage index definition.
301
302
===5.1 Classical mixing theory===
303
304
The classical rule of mixtures, originally developed by Trusdell and Toupin  ([6,7,8,24,31,39]), uses a phenomenological approach based on macro-scale continuum mechanics for the composite mechanical analysis, and is suitable for describing the mechanical behavior of each point of a composite solid. This formulation is based on the interaction principle of compounding substances of the composite material. The following basic assumptions are considered:
305
306
:1. A set of compounding substances participate in each infinitesimal volume of the composite.
307
308
:2.  Each component contributes in the composite behavior proportionally to their volumetric participation.
309
310
:3. All the components have the same strain (compatibility or closure equation).
311
312
:4.  Each component volume is much smaller than the total composite volume.
313
314
The second hypothesis involves a homogenous distribution of all the component substances at each point of the composite. The different component substance interaction and their corresponding constitutive law determine the composite material’s behavior and it depends on the volume percentage participation of each component and its distribution in the composite. Materials with different behaviors can be combined (elastic, damage, elastoplastic, etc.); each one representing an evolutionary behavior governed by its own law. The third hypothesis says that in the absence of atomic diffusion the following compatibility condition is satisfied for each of the composite material phases:
315
316
{| class="formulaSCP" style="width: 100%; text-align: center;" 
317
|-
318
| 
319
{| style="text-align: center; margin:auto;" 
320
|-
321
| <math display="inline">\boldsymbol{\epsilon }={\boldsymbol{\epsilon }}_{}=</math><math>{\boldsymbol{\epsilon }}_{}=...={\boldsymbol{\epsilon }}_{}</math>
322
|}
323
| style="width: 5px;text-align: right;white-space: nowrap;" | (12)
324
|}
325
326
327
where  <math display="inline">\boldsymbol{\epsilon }</math> and  <math display="inline">{\boldsymbol{\epsilon }}_{}</math> ( <math display="inline">i=1,...n</math> ) represent the composite material strain and the  <math display="inline">ith</math> component strain of such composite material, respectively.
328
329
The composite material’s free energy [33,34,35,36,37,39] is given by the additive composition of the free energy of each of the component materials considered as a function of its volumetric participation, thus
330
331
{| class="formulaSCP" style="width: 100%; text-align: center;" 
332
|-
333
| 
334
{| style="text-align: center; margin:auto;" 
335
|-
336
| <math>m\Psi \left({\boldsymbol{\epsilon }}^e,{\bf q}\right)=\sum_{k=1}^nk_km_k{\Psi }_k\left({\boldsymbol{\epsilon }}^e,{\bf q}_k\right)\mbox{ }\Rightarrow \mbox{ }\Psi =</math><math>\sum_{k=1}^nk_k\frac{m_k}{m}{\Psi }_k</math>
337
|}
338
| style="width: 5px;text-align: right;white-space: nowrap;" | (13)
339
|}
340
341
where  <math display="inline">m</math> is the mass density of the composite material,  <math display="inline">m_k</math> is the mass density of the  <math display="inline">k^{th}</math> component,  <math display="inline">k_k</math> is the volumetric participation coefficient of the  <math display="inline">kth</math> component,  <math>{\Psi }_k</math> is the free energy of the  <math display="inline">kth</math> component,  <math>{\bf q}_k</math> are the internal variables of each compounding models defining the non-linear behavior of any generic component [33,46,47,48,49].
342
343
The weighting factor or volumetric participation coefficient  <math display="inline">k_k</math> gives the contribution of each phase and is obtained by the volumetric participation of each of the component materials with respect to the total volume.
344
345
{| class="formulaSCP" style="width: 100%; text-align: center;" 
346
|-
347
| 
348
{| style="text-align: center; margin:auto;" 
349
|-
350
| <math>k_k=\frac{dV_k}{dV_0}</math>
351
|}
352
| style="width: 5px;text-align: right;white-space: nowrap;" | (14)
353
|}
354
355
356
where  <math>V_k</math> represents the material  <math display="inline">kth</math> component volume and  <math>V_0</math> is the total volume of the composite material. The volumetric participation coefficients of the different components of the composite material must satisfy the following condition:
357
358
{| class="formulaSCP" style="width: 100%; text-align: center;" 
359
|-
360
| 
361
{| style="text-align: center; margin:auto;" 
362
|-
363
| <math>\sum_{k=1}^nk_k=1</math>
364
|}
365
| style="width: 5px;text-align: right;white-space: nowrap;" | (15)
366
|}
367
368
369
in which the free energy can be recovered for a single component materials and the mass conservation can be guaranteed. Following a similar procedure using simple materials based on the Clausius-Duhem inequality and applying Coleman’s method [32,33,34,35,39] in which a positive dissipation is guaranteed, the stress and its constitutive equation can be obtained:
370
371
{| class="formulaSCP" style="width: 100%; text-align: center;" 
372
|-
373
| 
374
{| style="text-align: center; margin:auto;" 
375
|-
376
| <math display="inline">\boldsymbol{\sigma }=m\cdot \frac{\partial \Psi \left({\boldsymbol{\epsilon }}^e,{\boldsymbol{\alpha }}_k\right)}{\partial \boldsymbol{\epsilon }}=</math><math>\sum_{k=1}^nk_km_k\cdot \frac{\partial {\Psi }_k\left({\boldsymbol{\epsilon }}^e,{\boldsymbol{\alpha }}_k\right)}{\partial \boldsymbol{\epsilon }}=</math><math>\sum_{k=1}^nk_k\cdot {\boldsymbol{\sigma }}_k</math>
377
|}
378
| style="width: 5px;text-align: right;white-space: nowrap;" | (16)
379
|}
380
381
382
The secant constitutive equation for composite material (Equation (16)) results:
383
384
{| class="formulaSCP" style="width: 100%; text-align: center;" 
385
|-
386
| 
387
{| style="text-align: center; margin:auto;" 
388
|-
389
| <math display="inline">\boldsymbol{\sigma }={\mathbb{C}}_k^S\cdot {\boldsymbol{\epsilon }}^e=</math><math>\sum_{k=1}^nk_k\cdot {\boldsymbol{\sigma }}_k=\sum_{k=1}^nk_k\cdot \mbox{ }{\mathbb{C}}_k^S\mbox{:}\mbox{ }\mbox{ }{\boldsymbol{\epsilon }}_c^e=</math><math>\sum_{k=1}^nk_k\cdot \mbox{ }{\mathbb{C}}_k^S\mbox{:}\mbox{ }\mbox{ }{\boldsymbol{\epsilon }}_c^{}</math>
390
|}
391
| style="width: 5px;text-align: right;white-space: nowrap;" | (17)
392
|}
393
394
395
The tangent constitutive tensor is obtained by the stress variation with respect to the strains and is given by:
396
397
{| class="formulaSCP" style="width: 100%; text-align: center;" 
398
|-
399
| 
400
{| style="text-align: center; margin:auto;" 
401
|-
402
| <math display="inline">{\mathbb{C}}^T=\frac{\partial \boldsymbol{\sigma }}{\partial \boldsymbol{\epsilon }}=</math><math>m\frac{{\partial }^2{\Psi }_k\left({\boldsymbol{\epsilon }}^e,{\boldsymbol{\alpha }}_k\right)}{\partial {\boldsymbol{\epsilon }}^2}=</math><math>\sum_{k=1}^nk_k\cdot \mbox{ }{\mathbb{C}}_k^T</math>
403
|}
404
| style="width: 5px;text-align: right;white-space: nowrap;" | (18)
405
|}
406
407
408
The classical mixing theory, based on the hypothesis that the strain tensor is exactly the same for all the composite components, is strictly valid only if it is applied to material components working in parallel. These materials are characterized by the fact that their stress state is the results of the sum of the stresses of each component, which are weighted proportionally to their volume in each phase with respect to the total. In order to solve this problem there are two alternatives: to define another closure equation (Equation (12)), suitable for the material phenomena simulation, or to carry out a modification of each one of the component’s properties and keep the strain equality hypothesis in each one of the composite components. Therefore, the main problem of classical mixing theory is the poor ability to represent the serial behavior of the components in the composite (Figure 6, iso-stress case).
409
410
===5.2 Serial/Parallel mixing theory for one layer===
411
412
Due to the classic mixing theory limitation, various modifications have been proposed (see [32]). These allow representing the composite component’s behavior participating in a combination of serial-parallel behaviors. This involves an automatic adjustment of the composite properties taking into consideration each component’s topology and distribution. Thus, each point of a solid can have a different strain.
413
414
<div class="center" style="width: auto; margin-left: auto; margin-right: auto;">
415
[[Image:Draft_Samper_893761715-image60.png|600px]] </div>
416
417
<div class="center" style="width: auto; margin-left: auto; margin-right: auto;font-size:75%;">
418
419
[[Image:Draft_Samper_893761715-picture-Text Box 17.svg|center|600px]]
420
421
'''Figure 6''': Serial/Parallel mixing theory- Composition scheme</div>
422
423
424
Here, a short presentation of the serial/parallel mixing theory is made [1,27,39]. This rule of mixtures improves the classical mixing theory by modifying its closure equation, replacing the iso-strain hypothesis for an iso-strain condition in the fiber direction and an iso-stress condition in the transversal one (Equation  (21)). The modeling of all the components distribution in the composite are shown in Figure 6. This formulation is an alternative to the homogenization technique, based on the multiple scale study [3,4,47,48,51].
425
426
The Serial/Parallel (SP) formulation [1,27,39] considers that the component materials of the composite act in parallel along a certain direction and in serial in the remaining directions. The main hypothesis in which the numerical model of the Serial/Parallel mixing theory are based on:
427
428
:1. The composite is composed by two component materials: fiber and matrix.
429
430
:2. The component materials have the same strain in parallel (fiber) direction.
431
432
:3. The component materials have the same stress in the serial direction.
433
434
:4. The composite material response is in direct relation with the volume fractions of the compounding materials.
435
436
:5. The homogeneous distribution of phases is considered in the composite.
437
438
:6. Perfect bounding between components is assumed.
439
440
Consequently, it is necessary to define and separate the serial and parallel components of the strain and stress tensors. Defining  <math display="inline">\boldsymbol{e}_1\mbox{ }</math> as the director vector that determines the parallel behavior in the fiber direction, the parallel projector tensor  <math display="inline">\boldsymbol{N}_P\mbox{ }</math> can be defined as
441
442
<div style="text-align: right; direction: ltr; margin-left: 1em;">
443
<span style="text-align: center; font-size: 75%;"> <math display="inline">\boldsymbol{N}_P=\boldsymbol{e}_1\otimes \boldsymbol{e}_1</math> (19)</span></div>
444
445
Using  <math display="inline">\boldsymbol{N}_P\mbox{ }</math> , the 4th-order parallel projector tensor  <math display="inline">\boldsymbol{P}_P\mbox{ }</math> , and the complementary serial projector tensor  <math display="inline">\boldsymbol{P}_S\mbox{ }</math> , are defined as:
446
447
{| class="formulaSCP" style="width: 100%; text-align: center;" 
448
|-
449
| 
450
{| style="text-align: center; margin:auto;" 
451
|-
452
| <span style="text-align: center; font-size: 75%;"> <math display="inline">\boldsymbol{P}_P=\boldsymbol{N}_P\otimes \boldsymbol{N}_P\mbox{ }\mbox{ }\mbox{ }\mbox{ }\mbox{ }\mbox{ };\mbox{ }\mbox{ }\mbox{ }\mbox{ }\mbox{ }\mbox{ }\mbox{ }\mbox{ }\boldsymbol{P}_S=</math><math>\boldsymbol{I}-\boldsymbol{N}_P</math> </span>
453
|}
454
| style="width: 5px;text-align: right;white-space: nowrap;" | (20)
455
|}
456
457
458
Both tensors are used to find the parallel and serial part of the strain tensor  <math display="inline">{\boldsymbol{\epsilon }}_P\mbox{ }</math> and  <math display="inline">{\boldsymbol{\epsilon }}_S\mbox{ }</math> respectively,
459
460
{| class="formulaSCP" style="width: 100%; text-align: center;" 
461
|-
462
| 
463
{| style="text-align: center; margin:auto;" 
464
|-
465
| <math display="inline">{\boldsymbol{\epsilon }}_P=\boldsymbol{P}_P:\boldsymbol{\epsilon }</math> ;    <math display="inline">{\boldsymbol{\epsilon }}_S=\boldsymbol{P}_S:\boldsymbol{\epsilon }</math>
466
|}
467
| style="width: 5px;text-align: right;white-space: nowrap;" | (21)
468
|}
469
470
471
Hence, the strain and stress tensors are split into its parallel and serial parts
472
473
{| class="formulaSCP" style="width: 100%; text-align: center;" 
474
|-
475
| 
476
{| style="text-align: center; margin:auto;" 
477
|-
478
| <math display="inline">\boldsymbol{\epsilon }={\boldsymbol{\epsilon }}_P+</math><math>{\boldsymbol{\epsilon }}_S;\mbox{ }\boldsymbol{\sigma }=</math><math>{\boldsymbol{\sigma }}_P+{\boldsymbol{\sigma }}_S;\mbox{ }\mbox{where }\mbox{ }{\boldsymbol{\sigma }}_P=</math><math>\boldsymbol{P}_p\mbox{:}\mbox{ }\boldsymbol{\sigma }\mbox{   and  }{\boldsymbol{\sigma }}_S=</math><math>\boldsymbol{P}_S\mbox{:}\mbox{ }\boldsymbol{\sigma }</math>
479
|}
480
| style="width: 5px;text-align: right;white-space: nowrap;" | (22)
481
|}
482
483
484
The equations that define the stress equilibrium and establish the strain compatibility between components are obtained from the analysis of the model hypothesis. Thus,
485
486
{| class="formulaSCP" style="width: 100%; text-align: center;" 
487
|-
488
| 
489
{| style="text-align: center; margin:auto;" 
490
|-
491
| '''Parallel behavior: ''' <math display="inline">\begin{array}{c}
492
{\boldsymbol{\epsilon }}_{}{}_P={\boldsymbol{\epsilon }}_{}{}_P={\boldsymbol{\epsilon }}_{}{}_P+{\boldsymbol{\epsilon }}_S\\
493
{\boldsymbol{\sigma }}_{}{}_P=k_{}{\boldsymbol{\sigma }}_{}{}_P+k_{}{\boldsymbol{\sigma }}_{}{}_P
494
\end{array}</math>
495
|}
496
| style="width: 5px;text-align: right;white-space: nowrap;" | (23)
497
|}
498
499
{| class="formulaSCP" style="width: 100%; text-align: center;" 
500
|-
501
| 
502
{| style="text-align: center; margin:auto;" 
503
|-
504
| '''Serial behavior: ''' <math display="inline">\begin{array}{c}
505
{\boldsymbol{\epsilon }}_{}{}_S=k_{}{\boldsymbol{\epsilon }}_{}{}_S+k_{}{\boldsymbol{\epsilon }}_{}{}_S\\
506
{\boldsymbol{\sigma }}_{}{}_S={\boldsymbol{\sigma }}_{}{}_S={\boldsymbol{\sigma }}_{}{}_S
507
\end{array}</math>
508
|}
509
| style="width: 5px;text-align: right;white-space: nowrap;" | (24)
510
|}
511
512
513
where,  <math display="inline">{\boldsymbol{\epsilon }}_P\mbox{ }</math> and  <math display="inline">{\boldsymbol{\epsilon }}_S\mbox{ }</math> are the parallel and serial components of the stress tensor respectively,  <math display="inline">{\boldsymbol{\sigma }}_P\mbox{ }</math> and  <math display="inline">{\boldsymbol{\sigma }}_S\mbox{ }</math> are the parallel and serial components of the strain superscripts,  <math display="inline">c\mbox{ }</math> ,  <math display="inline">m\mbox{ }</math> and  <math display="inline">f\mbox{ }</math> denote the composite, matrix and fiber materials, and  <math display="inline">k_{}\mbox{ }</math> and  <math display="inline">k_{}\mbox{ }</math> are the volumetric participation of fiber and matrix in the composite, respectively.
514
515
The serial/parallel mixing theory can use any constitutive equation to describe the behavior of each component material. The constitutive equations chosen can be different for each component (for example, an elastic law to describe the fiber behavior and a damage formulation to describe the matrix behavior). The constitutive equations for the matrix and the fiber can be expressed in the following form:
516
517
{| class="formulaSCP" style="width: 100%; text-align: center;" 
518
|-
519
| 
520
{| style="text-align: center; margin:auto;" 
521
|-
522
| <span style="text-align: center; font-size: 75%;"> <math display="inline">{\boldsymbol{\sigma }}_{}={\mathbb{C}}_{}{}^S:{\boldsymbol{\epsilon }}_{}</math> </span>where <span style="text-align: center; font-size: 75%;"> <math>\left[\begin{array}{c}
523
{\boldsymbol{\sigma }}_{}{}_P\\
524
{\boldsymbol{\sigma }}_{}{}_S
525
\end{array}\right]=\left[\begin{array}{cc}
526
{\mathbb{C}}_{}{}_{PP}^S & {\mathbb{C}}_{}{}_{PS}^S\\
527
{\mathbb{C}}_{}{}_{SP}^S & {\mathbb{C}}_{}{}_{SS}^S
528
\end{array}\right]:\left[\begin{array}{c}
529
{\boldsymbol{\epsilon }}_{}{}_P\\
530
{\boldsymbol{\epsilon }}_{}{}_S
531
\end{array}\right]\mbox{ }</math> </span>
532
|}
533
| style="width: 5px;text-align: right;white-space: nowrap;" | (25) 
534
|}
535
536
537
where  <math display="inline">{\boldsymbol{\sigma }}_{}</math> is the stress tensor of the  <math display="inline">k\mbox{th}\mbox{ }</math> component material, <math display="inline">{\boldsymbol{\epsilon }}_{}</math> is the total strain tensors,  <math display="inline">{\mathbb{C}}_{}{}^S\mbox{ }</math> is the respective damaged secant constitutive tensor and its elements are:  <math display="inline">{\mathbb{C}}_{}{}_{PP}=\boldsymbol{P}_P:{\mathbb{C}}_{}:\boldsymbol{P}_P</math> ;  <math display="inline">{\mathbb{C}}_{}{}_{PS}=\boldsymbol{P}_P:{\mathbb{C}}_{}:\boldsymbol{P}_S</math> ;  <math display="inline">{\mathbb{C}}_{}{}_{SP}=\boldsymbol{P}_S:{\mathbb{C}}_{}:\boldsymbol{P}_P</math> ;  <math display="inline">{\mathbb{C}}_{}{}_{SS}=\boldsymbol{P}_S:{\mathbb{C}}_{}:\boldsymbol{P}_S</math> .
538
539
The schematic S/P (or Generalized) Mixing Theory flow diagram of a numerical implementation, is shown in Figure 7.
540
541
===5.3 Serial/Parallel mixing theory for a stacking layers composites===
542
543
Laminate composites are formed by different layers with different fiber orientations. The orientation of the fiber can be defined by the engineer or automatically by an optimization process in order to obtain the better performance of the composite according to its application. The S/P- Rule of Mixtures formulation can be applied to each layer of the composite and, afterwards, the composite behavior is computed by combining the performance of each constituent layer. The classical mixing theory is applied to each layer to obtain the composite laminate behavior.
544
545
Applying the classical mixing theory onto the different layers of the laminate composite implies the assumption that all laminate are undergoing the same strain. This is a simplified approach, as the performance of a laminate in the direction perpendicular to the layers is in serial. However, as it is stated in Ref. [27], the loss of accuracy is minimal compared to the improvement in the computational effort. This is because the different layers of the laminate usually have fiber orientation distributions disposed in such a way that provide the laminate with an in-plane homogeneous stiffness, and loads are rarely applied perpendicular to the laminate.
546
547
548
[[Image:Draft_Samper_893761715-picture-Group 31.svg|center|600px]]
549
550
5.4 Additional formulations used by S/P mixing theory to simulate reinforcement composite materials
551
552
Having defined the main frame of the formulation to deals with composite materials, there are other special formulations considered to obtain a better performance of the numerical simulation and approximation of the mechanical behavior of composite material structures. In this section, a brief description of them is considered.
553
554
<span id='_Toc160540553'></span>
555
556
===5.4.1  Anisotropy using a mapping space theory===
557
558
The mapping space theory permits take into account the anisotropy of each single component material [6,31,39]. It is based on the transportation of all the constitutive parameters and the stress and strain states of the structure, from a ''real anisotropic space'', to a ''fictitious isotropic space''. Once all variables are in the fictitious isotropic space, an isotropic constitutive model to obtain the new structure behavior can be used. This theory allows considering materials with high anisotropy, such as single or composite materials, using all the techniques and procedures already developed for isotropic materials.
559
560
<div class="center" style="width: auto; margin-left: auto; margin-right: auto;">
561
562
[[Image:Draft_Samper_893761715-image95.png|center|600px]]
563
</div>
564
565
All the anisotropy information is contained in two fourth order tensors. One of them, <math display="inline">A_{}^{\sigma }\mbox{ }</math> , relates the stresses in the fictitious isotropic space ( <math display="inline">{\overline{\sigma }}_{}\mbox{ }</math> ) with the stresses in the real anisotropic space ( <math display="inline">{\sigma }_{}\mbox{ }</math> ) and the other one,  <math display="inline">A_{}^{\epsilon }\mbox{ }</math> , does the same with the strains. The relation of both spaces for the strains and the stresses is exposed in Equation (26).
566
567
{| class="formulaSCP" style="width: 70%;border-collapse: collapse;width: 100%;text-align: center;" 
568
|-
569
| 
570
{| style="text-align: center; margin:auto;width: 100%;"
571
|-
572
|  <math>{\overline{\sigma }}_{}=A_{}^{\sigma }:{\sigma }_{}\mbox{ }\mbox{ }\mbox{ }\mbox{ }\mbox{ }\mbox{ }\mbox{ }\mbox{ }\mbox{ }\mbox{ }\mbox{ };\mbox{ }\mbox{ }\mbox{ }\mbox{ }\mbox{ }\mbox{ }\mbox{ }\mbox{ }\mbox{ }\mbox{ }\mbox{ }\mbox{ }\mbox{ }\mbox{ }\mbox{ }{\overline{\epsilon }}_{}=</math><math>A_{}^{\epsilon }:{\epsilon }_{}</math> 
573
|}
574
|  style="width: 5px;text-align: right;white-space: nowrap;"|(26)
575
|}
576
577
578
Once the stresses and strains are transferred to the respective isotropic spaces, the proposed constitutive equation is integrated, and its results are back to the anisotropic (real) spaces by a simple transportation operator ( <math>A_{}^{\sigma }\mbox{ }\mbox{ };\mbox{ }\mbox{ }A_{}^{\epsilon }</math> ) (Equation (26)).
579
580
In Figure 8 a representation of these transformations is displayed. A more detailed description of this methodology, the extension to large strains and its numerical implementation can be followed in references [8,31,39].
581
582
<span id='_Toc160540554'></span>
583
584
===5.4.2 Fiber-matrix debonding (FDM)===
585
586
Matrix Reinforced Composite materials have a complex nonlinear behavior due to the reinforcement displacement because of the loss of adherence between the matrix and the reinforcement. This relative movement between the reinforcement and the matrix causes a loss of stiffness in the whole set and a decrease of the composite mechanical parameters without fractures in the reinforcement phase is observed.
587
588
The formulation introduced in previous section is based on the mechanics of the continuum medium to deal with the anisotropy and the mixing theory. It involves introducing an irrecoverable inelastic behavior in the constitutive equation to represent an approximation of the relative rigid movement of the body produced between the fiber and the matrix. The incorporation of the FDM into the constitutive equation must take into consideration two main characteristics: a) the global loss of stiffness due to the decrease of the fiber collaboration in the matrix and b) the irrecoverable relative displacement between the fiber and the matrix.
589
590
The fiber reinforced composite materials subjected to tension do not satisfy the kinematic condition imposed by the basic theory of basic substances. A direct consequence of this phenomenon is the matrix limitation to transfer the stresses to the fiber. In other words, the fiber cannot increase its tensional state as a result of the limited adherence in the fiber-matrix interface zone.
591
592
The Fiber-matrix debonding constitutive model is based on the assumption that the loading transfer from the matrix to the fiber varies when the matrix is under plastic strains [6,8,31,39]. The relative movement between the fiber and the matrix can be represented by the mechanics of the continuum medium through an irrecoverable inelastic strain in the fiber. The starting point of this phenomenon is determined through a threshold condition of maximum strength, which compares the effective stress on a point with respect to the fiber strength. That is, the fiber participation in the composite depends on its own strength and on the stress transfer capacity of the fiber-matrix interface. Therefore, its strength is influenced by the medium containing it and its constitutive treatment might involve a non-local formulation. Then, the fiber strength contained in a matrix is defined as:
593
594
{| class="formulaSCP" style="width: 80%;border-collapse: collapse;width: 100%;text-align: center;" 
595
|-
596
| 
597
{| style="text-align: center; margin:auto;width: 100%;"
598
|-
599
|  <math>{\left(f^R\right)}_{fib}=min\left\{\mbox{ }{\left(f^N\right)}_{fib};{\left(f^N\right)}_{mat};\left[2\cdot {\left(f^N\right)}_{fib-mat}/r_{fib}\right]\mbox{ }\right\}</math> 
600
|}
601
|  style="width: 5px;text-align: right;white-space: nowrap;"|(27)
602
|}
603
604
605
in which  <math>r_{fib}</math> represents the radius of the long fiber of the transversal section,  <math>{\left(f^R\right)}_{fib}\mbox{ }</math> is the new fiber strength,  <math>{\left(f^N\right)}_{fib}\mbox{ }</math> is the nominal fiber strength,  <math>{\left(f^N\right)}_{mat}\mbox{ }</math> is the matrix nominal strength and  <math>{\left(f^N\right)}_{fib-mat}\mbox{ }</math> is the fiber-matrix interface nominal strength. Equation (27) shows that the debonding happens when one of the composite constituents reaches its nominal strength (considering the fiber-matrix interface as a constituent). The numerical implementation of this phenomenon is described in the reference [6,8,39].
606
607
<span id='_Toc160540558'></span>
608
609
===5.4.3 Tangent constitutive stiffness tensor===
610
611
Depending on the constitutive equation used in a composite constituent material, the tangent constitutive tensor cannot be analytically obtained. One solution is to use, in these materials, the initial stiffness matrix, which will lead to the equilibrium state but will require a large amount of structural iterations. Thus, in order to obtain a fast and reliable algorithm, the expression of the tangent constitutive tensor is required. To obtain it, when no analytical expression exists, a numerical derivation using a perturbation method is performed [29,39]. The definition of the tangent constitutive tensor is:
612
613
{| class="formulaSCP" style="width: 68%;border-collapse: collapse;width: 100%;text-align: center;" 
614
|-
615
| 
616
{| style="text-align: center; margin:auto;width: 100%;"
617
|-
618
|  <math>\boldsymbol{\dot{\sigma }}={\mathbb{C}}^t:\boldsymbol{\dot{\epsilon }}</math> 
619
|}
620
|  style="width: 5px;text-align: right;white-space: nowrap;"|(28)
621
|}
622
623
624
where,
625
626
{| class="formulaSCP" style="width: 77%;border-collapse: collapse;width: 100%;text-align: center;" 
627
|-
628
| 
629
{| style="text-align: center; margin:auto;width: 100%;"
630
|-
631
|  <math>\begin{array}{c}
632
\boldsymbol{\dot{\sigma }}=\left[\begin{array}{cccc}
633
{\dot{\sigma }}_1 & {\dot{\sigma }}_2 & \ldots  & {\dot{\sigma }}_n
634
\end{array}\right]\mbox{ };\mbox{ }\boldsymbol{\dot{\epsilon }}=\left[\begin{array}{cccc}
635
{\dot{\epsilon }}_1 & {\dot{\epsilon }}_2 & \ldots  & {\dot{\epsilon }}_n
636
\end{array}\right]\mbox{ }\mbox{and}\\
637
{\mathbb{C}}^t=\left[\begin{array}{ccc}
638
c_{11}^t & \ldots  & c_{1n}^t\\
639
\vdots  & \ddots  & \vdots \\
640
c_{n1}^t & \ldots  & c_{nn}^t
641
\end{array}\right]
642
\end{array}</math> 
643
|}
644
|  style="width: 5px;text-align: right;white-space: nowrap;"| (29)
645
|}
646
647
648
The definition of the tangent constitutive tensor, Equation (28), shows that the variation of stresses due to an increment in the value of the  <math display="inline">j</math> element of  <math display="inline">\boldsymbol{\dot{\epsilon }}\mbox{}</math> depends on the values of the  <math display="inline">j</math> column of  <math display="inline">{\mathbb{C}}^t</math> . Thus, writing the  <math display="inline">j</math> column of  <math>{\mathbb{C}}^t</math> as,
649
650
{| class="formulaSCP" style="width: 72%;border-collapse: collapse;width: 100%;text-align: center;" 
651
|-
652
| 
653
{| style="text-align: center; margin:auto;width: 100%;"
654
|-
655
|  <math>c_j^t={\left[\begin{array}{cccc}
656
c_{1j}^t & c_{2j}^t & \ldots  & c_{nj}^t
657
\end{array}\right]}^T</math> 
658
|}
659
|  style="text-align: right;width: 5px;text-align: right;white-space: nowrap;"|(30)
660
|}
661
662
663
the stress variation is:
664
665
{| class="formulaSCP" style="width: 63%;border-collapse: collapse;width: 100%;text-align: center;" 
666
|-
667
| 
668
{| style="text-align: center; margin:auto;width: 100%;"
669
|-
670
|  <math>{\dot{\sigma }}^j=c_j^t\cdot {\dot{\epsilon }}_j</math> 
671
|}
672
|  style="vertical-align: top;width: 5px;text-align: right;white-space: nowrap;"|(31)
673
|}
674
675
676
being  <math>c_j^t\mbox{ }</math> the unknown.
677
678
The perturbation method consists of applying a small perturbation on to the strain vector and, using the constitutive equation of the material, determines the variation that will be obtained in the stress tensor due to this perturbation. At this point, the  <math display="inline">j</math> column of the tangent constitutive tensor can be computed as:
679
680
{| class="formulaSCP" style="width: 65%;border-collapse: collapse;width: 100%;text-align: center;" 
681
|-
682
| 
683
{| style="text-align: center; margin:auto;width: 100%;"
684
|-
685
|  <math>c_j^t=\frac{{\dot{\sigma }}^j}{{\dot{\epsilon }}_j}</math> 
686
|}
687
|  style="width: 5px;text-align: right;white-space: nowrap;"|(32)
688
|}
689
690
691
For the smaller applied value to the perturbation, the better approximation is obtained for the tangent constitutive tensor. Having defined a perturbation value  <math display="inline">{\dot{\epsilon }}_j\mbox{ }</math> , the perturbed stress is computed using the constitutive equation of the material applying the following input strain:
692
693
{| class="formulaSCP" style="width: 72%;border-collapse: collapse;width: 100%;text-align: center;" 
694
|-
695
| 
696
{| style="text-align: center; margin:auto;width: 100%;"
697
|-
698
|  <math>\boldsymbol{\epsilon }={\left[\begin{array}{ccccc}
699
{\epsilon }_1 & \ldots  & {\epsilon }_j+{\dot{\epsilon }}_j & \ldots  & {\epsilon }_n
700
\end{array}\right]}^T\mbox{ }\rightarrow \mbox{ }\boldsymbol{\overset{\mbox{ˆ}}{\sigma }}</math> 
701
|}
702
|  style="width: 5px;text-align: right;white-space: nowrap;"|(33)
703
|}
704
705
706
And the stress variation due to the perturbation is obtained subtracting the original converged stress from the computed one:
707
708
{| class="formulaSCP" style="width: 56%;border-collapse: collapse;width: 100%;text-align: center;" 
709
|-
710
| 
711
{| style="text-align: center; margin:auto;width: 100%;"
712
|-
713
|  <math>{\boldsymbol{\dot{\sigma }}}^\boldsymbol{j}\boldsymbol{=</math><math>\overset{\mbox{ˆ}}{\sigma }-\sigma }</math> 
714
|}
715
|  style="width: 5px;text-align: right;white-space: nowrap;"|(34)
716
|}
717
718
719
This procedure must be repeated for all strain components in order to obtain the complete expression of the tangent constitutive tensor. Hence, the numerical cost of using a perturbation method is rather high. However, this procedure allows obtaining an accurate approximation to this tensor for any constitutive equation used, ensuring the Newton-Raphson convergence of the numerical process in few steps.
720
721
===5.5 Local plastic-damage model for a component material===
722
723
The theory of plasticity provides a suitable physical-mathematical framework to formulate the behavior of metallic materials subjected to loading. From the extension of its main basic principles and the reinterpretation of its main variables, the ''Plastic Damage Model'' has emerged as one of the more general plastic constitutive model [36,37,40]. This plastic model is based on the ''plastic damage variable''  <math>d^p\mbox{ }</math> formulated as an internal variable representing a unit normalized dissipated plastic energy ranging from  <math>0\leq d^p\leq 1\mbox{ }</math> . For  <math>d^p=0\mbox{ }\mbox{ }</math> there is no plastic damage and for  <math>d^p=1\mbox{ }</math> the limit of total plastic damage in a solid point is defined. The latter state can be interpreted as a total loss of strength in a point of the solid produced for the accumulated plastic effect. A brief presentation of this model will be given in this section. For further details, check references [36,37,40]
724
725
In short, for a plastic mechanical process with no stiffness degradation, the plastic damage model uses the following set of plastic internal variables <math display="inline">q^p=\lbrace {\epsilon }^p,{\alpha }^p\rbrace =\lbrace {\epsilon }^p,d^p,c\rbrace \mbox{ }</math> , and its evolution laws will be presented as part of the main equations governing the model.
726
727
* A '''deformation splits''' into an elastic and a plastic part,
728
729
{| class="formulaSCP" style="width: 100%;width: 100%;text-align: center;" 
730
|-
731
| 
732
{| style="text-align: center; margin:auto;width: 100%;"
733
|-
734
|  <math display="inline">\epsilon ={\epsilon }^e+{\epsilon }^P={\mathbb{C}}^{-1}:\sigma +</math><math>{\epsilon }^P</math> 
735
|}
736
|  style="text-align: right;width: 5px;text-align: right;white-space: nowrap;"|(35)
737
|}
738
739
740
where  <math display="inline">\mathbb{C}</math> is the constitutive elastic tensor.
741
742
* A '''plastic yield''' '''and potential plastic criteria''' are defined by the following two equations,
743
744
{| class="formulaSCP" style="width: 100%;width: 100%;text-align: center;" 
745
|-
746
| 
747
{| style="text-align: center; margin:auto;width: 100%;"
748
|-
749
|  <math>F^P\left(\sigma ;q^p\right)=f^P\left(\sigma \right)-</math><math>c{\left(d^p\right)}^{}\leq 0</math> 
750
|}
751
|  style="text-align: right;width: 5px;text-align: right;white-space: nowrap;"|(36)
752
|-
753
| 
754
{| style="text-align: center; margin:auto;width: 100%;"
755
|-
756
|  <math>G^P\left(\sigma ;q^p\right)=g^P\left(\sigma \right)-</math><math>cte^{}=0</math> 
757
|}
758
|  style="text-align: right;width: 5px;text-align: right;white-space: nowrap;"|(37)
759
|}
760
761
762
being  <math>F^P\left(\sigma ;q^p\right)\mbox{ }</math> the plastic damage threshold function,  <math>G^P\left(\sigma ;q^p\right)\mbox{ }</math> the plastic potential,  <math>c\left(d^p\right)</math> the cohesion or uniaxial strength evolution, depending on internal plastic damage variable  <math>d^p</math> ,  <math>f^P\left(\sigma \right)\mbox{ }</math> and  <math>g^P\left(\sigma \right)\mbox{ }</math> are two scalar functions of tensorial arguments called yield function and plastic potential respectively, and can be represented by any classical limit criteria (von Mises, Mohr Coulomb, Drucker Prager, etc.)  [26,34,40]
763
764
The uniaxial strength evolution  <math>c\left(d^P\right)\mbox{ }</math> is assumed as a scaled magnitude regarding an ultimate strength to uniaxial compression of the composite  <math>{\sigma }_C^{}\mbox{ }</math> (stress discontinuity threshold), that is the stress level for which the volumetric deformation  <math display="inline">{\epsilon }_V\mbox{ }</math> reaches its maximum value. Therefore, the initial uniaxial strength, is defined as  <math>c^0\propto {\sigma }_C^{}\mbox{ }</math> for  <math>d^p=0\mbox{ }</math> , setting the initial position of the yield criterion and the final uniaxial strength of the material totally deteriorated as,  <math>c^u=0\mbox{ }</math> for  <math>d^p=1\mbox{ }</math> , defining the final position of the yield criterion (see Figure 9).
765
766
Unlike the classic plasticity formulation with isotropic hardening, the cohesion or uniaxial strength in this case is not a simple function of the plastic hardening variable <math display="inline">c(d^p)\mbox{ }</math> , but is an internal variable that depends on the evolution of the elasto-plastic process governed by its evolution equation.
767
768
* A '''plastic strain, plastic damage, and cohesion internal variables,'''
769
770
{| class="formulaSCP" style="width: 100%;width: 100%;text-align: center;" 
771
|-
772
| 
773
{| style="text-align: center; margin:auto;width: 100%;"
774
|-
775
|  <math>\boldsymbol{q}=\left\{\begin{array}{c}
776
{\epsilon }^P\\
777
{\alpha }^p
778
\end{array}\right\}=\left\{\mbox{ }\begin{array}{c}
779
{\epsilon }^P\\
780
d^P\\
781
c
782
\end{array}\right\}\mbox{ }\mbox{ }\mbox{ }\mbox{ }\mbox{ }\mbox{ }\mbox{ }\mbox{ }\mbox{ }\mbox{ }\mbox{ }\mbox{ }\begin{array}{c}
783
{\mbox{   Plastic}}_{}\mbox{Strain      }\\
784
\mbox{ }\mbox{ }\mbox{ }\mbox{ }\mbox{ }{\mbox{Plastic}}_{}{\mbox{Damage}}_{}\mbox{Variable}\\
785
{\mbox{Cohesion}}^{}\mbox{Variable }
786
\end{array}</math> 
787
|}
788
|  style="text-align: right;width: 5px;text-align: right;white-space: nowrap;"|(38)
789
|}
790
791
792
all of them defined by the following evolution equations,
793
794
{| class="formulaSCP" style="width: 100%;width: 100%;text-align: center;" 
795
|-
796
| 
797
{| style="text-align: center; margin:auto;width: 100%;"
798
|-
799
|  <math display="inline">\frac{dq^p}{dt}={\dot{q}}^p=\left\{\begin{array}{c}
800
{\dot{\epsilon }}^P\\
801
{\dot{d}}^p\\
802
\dot{c}
803
\end{array}\right\}\equiv \dot{\lambda }\mbox{ }\cdot \boldsymbol{H}=</math><math>\dot{\lambda }\cdot \left\{\begin{array}{c}
804
\underset{}{\frac{\partial G^P}{\partial \sigma }}\\
805
\underset{}{\boldsymbol{h}_p\boldsymbol{:}\frac{\partial G^P}{\partial \sigma }}\\
806
h_c\cdot \boldsymbol{h}_p\boldsymbol{:}\frac{\partial G^P}{\partial \sigma }
807
\end{array}\right\}\equiv \left\{\begin{array}{c}
808
\underset{}{{\dot{\epsilon }}^P}\\
809
\underset{}{\boldsymbol{h}_p\boldsymbol{:}{\dot{\epsilon }}^P}\\
810
h_c\cdot \boldsymbol{h}_p\boldsymbol{:}{\dot{\epsilon }}^P
811
\end{array}\right\}</math> 
812
|}
813
|  style="text-align: right;width: 5px;text-align: right;white-space: nowrap;"|(39)
814
|}
815
816
817
where  <math display="inline">\boldsymbol{h}_p</math> and  [[Image:Draft_Samper_893761715-image150.png|12px]] are a second-degree tensor and a scalar function respectively that will be defined later, and which depend on the current state of the free variable  <math display="inline">{\epsilon }^e</math> and the rest of the internal variables  <math display="inline">q^p</math> . As observed in this equation, the main internal variable is the plastic deformation  <math display="inline">{\epsilon }^p</math> , and the others are obtained from it. The plastic consistency factor  [[Image:Draft_Samper_893761715-image154.png|12px]] is obtained from the consistency condition of the plastic yield function [36,37,40].
818
819
* '''A secant and tangent constitutive equation, '''defined as the classic plasticity,
820
821
{| class="formulaSCP" style="width: 100%;width: 100%;text-align: center;" 
822
|-
823
| 
824
{| style="text-align: center; margin:auto;width: 100%;"
825
|-
826
|  <math display="inline">\begin{array}{c}
827
\mbox{ }\mbox{ }\mbox{ }\mbox{ }\mbox{ }\mbox{ }\mbox{ }\mbox{ }\mbox{ }\mbox{ }\mbox{ }\mbox{ }\mbox{ }\mbox{ }\mbox{ }\mbox{ }\mbox{ }\mbox{ }\mbox{ }\mbox{ }\mbox{ }\mbox{ }\mbox{ }\mbox{ }\mbox{ }\mbox{ }\mbox{ }\mbox{ }\mbox{ }\mbox{ }\mbox{ }\mbox{ }\mbox{ }\mbox{ }\mbox{ }\mbox{ }\mbox{ }\mbox{ }\mbox{ }\mbox{ }\mbox{ }\mbox{ }\mbox{ }\mbox{ }\mbox{ }\mbox{ }\mbox{ }\mbox{ }\mbox{ }\mbox{ }\mbox{ }\mbox{ }\mbox{ }\mbox{ }\mbox{ }\mbox{ }\mbox{ }\mbox{ }\mbox{ }\mbox{ }\mbox{ }\mbox{ }\mbox{ }\mbox{ }\mbox{ }\mbox{ }\mbox{ }\mbox{ }\mbox{ }\mbox{ }\mbox{ }\mbox{ }\mbox{ }\mbox{ }\sigma =\mathbb{C}\boldsymbol{:}\left(\epsilon -{\epsilon }^p\right)\\
828
\\
829
\dot{\sigma }=\left\{\mathbb{C}-\frac{\left[\mathbb{C}\boldsymbol{:}\frac{\partial G^P}{\partial \sigma }\right]\otimes \left[\frac{\partial F^P}{\partial \sigma }\boldsymbol{:}\mathbb{C}\right]}{-c_k\cdot \frac{\partial F^P}{\partial \eta }\boldsymbol{:}\frac{\partial G^P}{\partial \sigma }+h_c\mbox{ }\boldsymbol{h}_p\boldsymbol{:}\frac{\partial G^P}{\partial \sigma }+\left(\frac{\partial F^P}{\partial \sigma }\boldsymbol{:}\mathbb{C}\boldsymbol{:}\frac{\partial G^P}{\partial \sigma }\right)}\right\}\boldsymbol{:}\dot{\epsilon }\mbox{ }\mbox{ }\mbox{ }\mbox{ }\mbox{ }\mbox{ }\Rightarrow \mbox{ }\mbox{ }\mbox{ }\mbox{ }\mbox{ }\mbox{ }\dot{\sigma }={\mathbb{C}}^T\boldsymbol{:}\dot{\epsilon }
830
\end{array}</math> 
831
|}
832
|  style="text-align: center;width: 5px;text-align: right;white-space: nowrap;"|(40)
833
|}
834
835
836
where  <math display="inline">\eta </math> is the kinematic plastic flow orientation [40]
837
838
The constitutive model resulting from these basic definitions shows a very good response during the general behavior composite process. To sum up, the model has the following characteristics:
839
840
:* It defines a constitutive law depending on the internal variables of cohesion and plastic damage to represent non-radial complex loading situations.
841
842
:* It deals with complex states of multiaxial stress in a unified way.
843
844
:* It admits that materials have different limits of maximum strength and ultimate deformation, depending on the mechanical process in progress.
845
846
:* It admits different plastic yield and potential criterion. Although this is not a characteristic of the model, it can be pre-established as one of its variables.
847
848
<span id='_Toc488140689'></span><span id='_Toc488142097'></span>:* It can obtain all the information related to the point deterioration through the point mechanical information post processing.
849
850
===5.5.1 Definition of the plastic damage variable===
851
852
The classical plasticity theory establishes a hardening variable as a function of the effective plastic strain  <math display="inline">{\overline{\epsilon }}^p</math> , or also as a function of the specific plastic work  <math display="inline">{\omega }^p=\overline{\sigma }\mbox{ }{\overline{\epsilon }}^p=</math><math>\sigma :{\epsilon }^p</math> [34,35,40]. These definitions are suitable for materials whose final deformation is equal in tension as in compression as metal behavior. However, this assumption is not true for many materials as composites one. Therefore, it is necessary to establish an internal variable defining a unit-normalized dissipation, which is the relation between the density of the dissipated energy at a specific time of the process and the maximum dissipation of the point of the solid. Hence, it is said that ''the plastic damage'' ''variable ''is a unit-normalized measure of the dissipated energy during the plastic process.
853
854
Generally, for a generic loading process, the plastic damage variable is defined for a multiaxial mechanical process as,
855
856
{| class="formulaSCP" style="width: 100%;width: 100%;text-align: center;" 
857
|-
858
| 
859
{| style="text-align: center; margin:auto;width: 100%;"
860
|-
861
|  <math display="inline">{\dot{d}}^p=h_p:{\boldsymbol{\dot{\epsilon }}}^p</math> 
862
|}
863
|  style="text-align: right;width: 5px;text-align: right;white-space: nowrap;"|(41)
864
|}
865
866
867
where  <math display="inline">h_p</math> is a second-degree tensor that, for uniaxial tension and compression processes, leads to a plastic damage depending on the loading process. To recover the plastic hardening variable of the classic plasticity theory, this tensor becomes equal to the stress tensor  <math display="inline">\boldsymbol{h}_p=\sigma </math> and in the general case it can be defined as a normalized dissipation to the unit for isotropic materials.
868
869
{| class="formulaSCP" style="width: 100%;width: 100%;text-align: center;" 
870
|-
871
| 
872
{| style="text-align: center; margin:auto;width: 100%;"
873
|-
874
|  <math display="inline">{\dot{d}}^p=\boldsymbol{h}_p\boldsymbol{:}{\dot{\epsilon }}^p=</math><math>\left[\frac{r(\sigma )}{g_f^p}+\frac{1-r(\sigma )}{g_C^p}\right]\cdot {\Xi }_m</math> 
875
|}
876
|  style="text-align: right;width: 5px;text-align: right;white-space: nowrap;"|<span id='_Ref511298171'></span>(42)
877
|}
878
879
880
where  <math display="inline">{\Xi }_m=\sigma :{\dot{\epsilon }}^p</math> is the plastic dissipation energy,  <math display="inline">g_f^p\mbox{ }=G_f^p/l_f</math> and  <math display="inline">g_c^p\mbox{ }=G_c^p/l_f</math> are the fracture and crushing energy per unit of volume respectively,  <math display="inline">G_f^p</math> and  <math display="inline">G_c^p</math> are the mechanical fracture and crushing energy parameter,  <math display="inline">l_f</math> is a regularization parameter called characteristic length related with the finite element size,  <math display="inline">r(\sigma )={\sum }_{I=1}^3\langle {\sigma }_I\rangle /{\sum }_{I=1}^3\vert {\sigma }_I\vert </math> is a scalar function defining the tension-compression behavior states in each point as a function of the stress state, and  [[Image:Draft_Samper_893761715-image170.png|102px]] is the Macaulay bracket. Note the following particular cases,  <math display="inline">r(\sigma )=1</math> for pure tension problems,  <math display="inline">r(\sigma )=0</math> for pure compression and  <math display="inline">r(\sigma )=0,5</math> for a pure shear state. Consequently, the plastic dissipation will always be normalized with respect to the maximum energy of the process at every moment.
881
882
Thus, the plastic damage variable is objective and evolves within the same limits regardless of the mechanical process. Thus, the total plastic damage in a point is reached when  <math display="inline">d^p=1</math> , but the dissipated energy will be  [[Image:Draft_Samper_893761715-image175.png|18px]] in a pure tension process, and  [[Image:Draft_Samper_893761715-image176.png|18px]] in a pure compression process.
883
884
===5.5.2 Definition of the cohesion or uniaxial strength evolution law  <math>c-d^p</math> ===
885
886
This plastic damage model assumes that micro-plastic damage in most materials is due to the loss of strength. Due to this failure mechanism, a softening in the strain-stress behavior can only be observed as a macroscopic effect (phenomenological model) caused by the average behavior of a set of points.
887
888
The plastic damage constitutive model carries out the numerical analysis on a finite domain (integration point of the constitutive equation) through the finite element functional approximation technique. Therefore, every point under analysis represents infinite material points contained in its area of influence. Thus, at macroscopic level, the softening phenomenon by deformation can be considered as a material property and, in such a case, a plastic hardening function must be defined taking into account this phenomenon. This hardening function is represented by the cohesion, written as an internal variable to make it more general and its evolution equation for any quasi-static loading process is defined as,
889
890
{| class="formulaSCP" style="width: 100%;width: 100%;text-align: center;" 
891
|-
892
| 
893
{| style="text-align: center; margin:auto;width: 100%;"
894
|-
895
|  <math>\dot{c}=h_c\cdot {\dot{d}}^p=h_c\cdot \boldsymbol{h}_p\boldsymbol{:}{\dot{\epsilon }}^p</math> 
896
|}
897
|  style="width: 5px;text-align: right;white-space: nowrap;"|  (43)
898
|}
899
900
901
where  <math display="inline">h_c(\sigma ,d^p,c)</math> is a scalar function of the current state of the stress-free variable  <math display="inline">\sigma </math> and of the internal variables  <math display="inline">d^p</math> and  <math display="inline">c</math> . The expression used for the evolution law of the internal variable of cohesion, or uniaxial strength, is obtained from the following expression for  <math display="inline">h_c\mbox{ }</math> ,
902
903
{| class="formulaSCP" style="width: 100%;width: 100%;text-align: center;" 
904
|-
905
| 
906
{| style="text-align: center; margin:auto;width: 100%;"
907
|-
908
|  <math display="inline">h_c=c\cdot \mbox{ }\mbox{ }\left[\frac{r(\sigma )}{c_T^{}}\mbox{ }\mbox{ }\frac{dc_T^{}}{d\mbox{ }d^p}+\right. </math><math>\left. \frac{1-r(\sigma )}{c_C^{}}\mbox{ }\mbox{ }\frac{dc_C^{}}{d\mbox{ }d^p}\right]</math> 
909
|}
910
|  style="width: 5px;text-align: right;white-space: nowrap;"|(44)
911
|}
912
913
914
where  <math display="inline">r(\sigma )</math> is the aforementioned function which sets the type of behavior (tension or compression or tension-compression), developed during the mechanical process in each point of a solid. The cohesion function  <math display="inline">c_C(d^p)</math> [13,40] is obtained in explicit form and it represents the cohesion evolution during a uniaxial simple compression test. The relation between cohesion and uniaxial stress of compression is given by the following expression,
915
916
{| class="formulaSCP" style="width: 100%;width: 100%;text-align: center;" 
917
|-
918
| 
919
{| style="text-align: center; margin:auto;width: 100%;"
920
|-
921
|  <math display="inline">c_C({\kappa }^p)=\frac{1}{\aleph }{\sigma }_C(d^p)</math> 
922
|}
923
|  style="width: 5px;text-align: right;white-space: nowrap;"|(45)
924
|}
925
926
927
such that  <math display="inline">\aleph \mbox{ }</math> is a coefficient depending on the criterion of the discontinuity threshold and represents a scalar factor between cohesion and the uniaxial stress of compression [36,37,40]. For example, for Tresca and von-Mises its value is  <math display="inline">\aleph =1\mbox{ }</math> , for Mohr Coulomb  <math display="inline">\aleph =2\mbox{ }\sqrt{R_0}\mbox{ }</math> where  <math display="inline">R^0=\left[f_C^0/f_T^0\right]={\left[{\sigma }_C(d^p=0)/{\sigma }_T(d^p=0)\right]}^{}</math> is the relation between compression-tension uniaxial strengths, for Drucker-Prager inscribed in the surface of Mohr-Coulomb  <math display="inline">\aleph =6\mbox{ }cos\varphi /(sin\varphi -3)\mbox{ }</math> and for Drucker-Prager circumscribed in the surface of Mohr-Coulomb  <math display="inline">\aleph =6\mbox{ }cos\varphi /(3\mbox{ }sin\varphi -</math><math>3)\mbox{ }</math> . Thus, for any discontinuity threshold criterion, this coefficient must be defined.
928
929
The function  <math display="inline">c_T(d^p)</math> (see Figure 9) can be obtained explicitly and represents the cohesion evolution during a uniaxial simple tension test. The relation between cohesion and uniaxial tension stress is given by the following expression,
930
931
{| class="formulaSCP" style="width: 100%;width: 100%;text-align: center;" 
932
|-
933
| 
934
{| style="text-align: center; margin:auto;width: 100%;"
935
|-
936
|  <math display="inline">c_T(d^p)=\frac{R^0}{\aleph }{\sigma }_T(d^p)</math> 
937
|}
938
|  style="width: 5px;text-align: right;white-space: nowrap;"|(46)
939
|}
940
941
942
For Tresca and von-Mises its value is  <math display="inline">R^0/\aleph =1\mbox{ }</math> , for Mohr Coulomb  <math display="inline">R^0/\aleph =\sqrt{R_0}/2\mbox{ }</math> , for Drucker-Prager inscribed in the Mohr-Coulomb surface  <math display="inline">R_0/\aleph =(3+3\mbox{ }sin\varphi )/6\mbox{ }cos\varphi \mbox{ }</math> and for Drucker-Prager circumscribed in the Mohr-Coulomb surface  <math display="inline">R/\aleph =(3+sin\varphi )/6\mbox{ }cos\varphi \mbox{ }</math> .
943
944
Some materials strength curves in simple tension and compression, obtained in uniaxial experimental tests, have similar shapes, in other words, it can be stated that the scale relationship between them is a constant during the whole quasi-static process and is given by
945
946
{| class="formulaSCP" style="width: 100%;width: 100%;text-align: center;" 
947
|-
948
| 
949
{| style="text-align: center; margin:auto;width: 100%;"
950
|-
951
|  <math display="inline">R(d^p)=\frac{{\sigma }_C(d^p)}{{\sigma }_T(d^p)}=cte.\mbox{ }\mbox{ }\mbox{ }\Rightarrow \mbox{ }\mbox{ }R(d^p=</math><math>0)=R^0\mbox{ }</math> 
952
|}
953
|  style="width: 5px;text-align: right;white-space: nowrap;"|(47)
954
|}
955
956
957
In such case the explicit functions of uniaxial tension and compression cohesion coincide.
958
959
 [[Image:Draft_Samper_893761715-image201.png|600px]]
960
961
5.6 Local damage (elastic degradation) model for a component material
962
963
The local damage constitutive model [28,36,40] used to set the threshold for the initiation of non-linear elastic modulus degradation process in each point of component material of the composite one and its subsequent evolution is presented in this subsection. This concept allows the new definition of a global homogenized threshold criterion of damage for the entire laminate, resulting from the composition of the local damage index over all involved material in the laminate composite.
964
965
Material degradation -or damage- in a simple continuum material component due to a dissipative process can be simulated by means a local damage formulation [16,36,37,40,49]. This model is used at each simple matrix material embedded in the composite, inducing a stiffness degradation and strength reduction in the entire laminate.
966
967
The isotropic damage formulation is based on a scalar internal variable  <math display="inline">d^d</math> that represents the level of elastic degradation at each simple component material. This variable is bounded between 0 and 1, being zero for an undamaged and one for a completely damaged state of a single component material. The local damage variable  <math display="inline">d^d\mbox{ }</math> is used to link the real stress tensor <math display="inline">\sigma </math> with the effective undamaged stress tensor  <math display="inline">{\sigma }_0</math> . Therefore, the relation between the damaged stress and the strain in the matrix component included in each layer depends on the internal damage variable  <math display="inline">d^d</math> and the elastic constitutive tensor  <math>{\mathbb{C}}_0\mbox{ }</math> ,
968
969
{| class="formulaSCP" style="width: 100%;width: 100%;text-align: center;" 
970
|-
971
|  style="width: 5px;text-align: right;white-space: nowrap;"|<span style="text-align: center; font-size: 75%;">(48)</span>
972
|}
973
974
975
In the same form of the plastic formulation, the stress condition at which damage starts and the evolution of the damage variable can be described by the following threshold function:
976
977
{| class="formulaSCP" style="width: 100%;width: 100%;text-align: center;" 
978
|-
979
|  style="width: 5px;text-align: right;white-space: nowrap;"|<span style="text-align: center; font-size: 75%;">(49)</span>
980
|}
981
982
983
being  <math>F^d\left({\sigma }_0;q^d\right)</math> the damage threshold function,  <math>f^d\left({\sigma }_0\right)</math> the scalar equivalent stress function and  <math>c\left(d^d\right)</math> the uniaxial strength evolution depending of the internal damage variable  <math display="inline">d^d</math> .
984
985
In the same form of the plastic damage model, this formulation allows the damage onset and evolution using any isotropic limit criteria already defined in literature (von Mises, Mohr Coulomb, Drucker Prager, etc.)  [26,48], and the anisotropic behavior is included by means the mapping spaces theory previously defined in section 5.4.
986
987
The norm of the principal stresses with a different degradation path for tension and compression loads is also here defined for the widest range of structures in composite materials. That is,
988
989
{| class="formulaSCP" style="width: 100%;width: 100%;text-align: center;" 
990
|-
991
|  style="width: 5px;text-align: right;white-space: nowrap;"|<span style="text-align: center; font-size: 75%;">(50)</span>
992
|}
993
994
995
being  <math display="inline">{\sigma }_c</math> and  <math>{\sigma }_t</math> the uniaxial ultimate strength of the material in compression and tension, respectively,  <math>{\sigma }_I</math> the principal stress tensor, and  <math>r(\sigma )</math> is the same a scalar function defined in the plastic damage model, that take into account the tension-compression behavior states in each point as a function of the stress state,
996
997
{| class="formulaSCP" style="width: 100%;width: 100%;text-align: center;" 
998
|-
999
|  style="width: 5px;text-align: right;white-space: nowrap;"|<span style="text-align: center; font-size: 75%;">(51)</span>
1000
|}
1001
1002
1003
and  <math display="inline">\langle \bullet \rangle \mbox{ }</math> is the Macaulay bracket already defined.
1004
1005
The mechanical evolution of the damage inner variable  <math display="inline">d^d</math> or damage index for a simple component material is obtained by means the damage consistency equation [40] and the Kuhn–Tucker load/unload conditions [28,48]. The uniaxial strength function is defined as,
1006
1007
{| class="formulaSCP" style="width: 100%;width: 100%;text-align: center;" 
1008
|-
1009
|  style="width: 5px;text-align: right;white-space: nowrap;"|<span style="text-align: center; font-size: 75%;">(52)</span>
1010
|}
1011
1012
1013
The function  <math display="inline">G^d</math> defines the softening evolution of the material. The behavior evolution of the damage material in the present work uses an explicit exponential softening, which is defined as,
1014
1015
{| class="formulaSCP" style="width: 100%;width: 100%;text-align: center;" 
1016
|-
1017
| 
1018
{| style="text-align: center; margin:auto;width: 100%;"
1019
|-
1020
| <span style="text-align: center; font-size: 75%;"> <math>d^d=G^d\left[f^d\left({\sigma }_0\right)\right]\mbox{ }=</math><math>1-\frac{{\sigma }_c}{f^d\left({\sigma }_0\right)}\mbox{ }\mbox{ }\mbox{ }\mbox{e}^{A\left(1-\mbox{ }\frac{f^d\left({\sigma }_0\right)}{{\sigma }_c}\right)}\mbox{ }\mbox{ }\mbox{ }\mbox{ }\mbox{ }\mbox{ }\mbox{ }\mbox{ };\mbox{ }\mbox{ }\mbox{ }\mbox{ }\mbox{ }\mbox{with}\mbox{ }\mbox{ }\mbox{ }\mbox{ }\mbox{ }\mbox{ }\mbox{ }\mbox{ }{\sigma }_c=</math><math>c^{max}</math> </span>
1021
|}
1022
|  style="width: 5px;text-align: right;white-space: nowrap;"|<span style="text-align: center; font-size: 75%;">(53)</span>
1023
|}
1024
1025
1026
where  <math display="inline">A</math> is a parameter that depends of the fracture energy of each simple material. This parameter can be obtained for an exponential softening material as
1027
1028
{| class="formulaSCP" style="width: 100%;width: 100%;text-align: center;" 
1029
|-
1030
| 
1031
{| style="text-align: center; margin:auto;width: 100%;"
1032
|-
1033
| <span style="text-align: center; font-size: 75%;"> <math>A={\left(\frac{G_c^d\cdot \mbox{ }C_0}{l_f\cdot {\sigma }_c^2}-\frac{1}{2}\right)}^{-1}</math> </span>
1034
|}
1035
|  style="width: 5px;text-align: right;white-space: nowrap;"|<span style="text-align: center; font-size: 75%;">(54)</span>
1036
|}
1037
1038
1039
being  <math display="inline">C_0</math> the uniaxial Young modulus of the material,  <math display="inline">G_c^d</math> the damage compression energy of the material and  <math display="inline">l_f</math> the geometrically regularization parameter, called ''characteristic fracture length'' related with the characteristically size of the finite element used. The introduction of this fracture length in the formulation makes the degradation process mesh independent [28,48].
1040
1041
===5.7 Global composite homogenized laminate damage index===
1042
1043
In this section, the global damage index [38] to ensure that the composite laminate is found within the elastic range is defined.
1044
1045
According to the Serial/Parallel mixing theory previously introduced, fibers only collaborate to the composite strengthening in longitudinal direction of the fibers. Thus, the damage on the composite material is mainly concentrated in the matrix but not in the reinforcement fibers. Thereby, when stress in matrix reach its maximum elastic value (damage threshold), the material component falls according to the “damage constitutive law” or “plastic damage constitutive low” previously presented. At this time, the energy fracture dissipation begins in each component, and the material point cannot support any more the stresses level, the stiffness contribution disappears, and starts a crack evolution producing a delamination phenomenon. In addition, the lack of strength in the matrix in all directions is produced, except in the longitudinal fibers (because fibers do not reach the damaged threshold). Hence, this mechanism induces localized fracture (delamination) at constitutive level without the computational cost of breaking the mesh and re-meshing the new delaminated area.
1046
1047
In the case of laminates, the global composite damage index  <math display="inline">d^L</math> is obtained by the homogenization of de local damage variable  <math display="inline">d^p</math> or  <math display="inline">d^d</math> of each simple component materials (see previous sections). This definition of homogenized laminar damage can also be understood as a safety laminate output warning. This new structural damage index  <math display="inline">d^L</math> is also bounded between 0 and 1 as local damage variable and it is defined as,
1048
1049
{| class="formulaSCP" style="width: 100%;width: 100%;text-align: center;" 
1050
|-
1051
| 
1052
{| style="text-align: center; margin:auto;width: 100%;"
1053
|-
1054
| <span style="text-align: center; font-size: 75%;"> <math>d^L={\left(\sum_{i=1}^{n_{GP}}V_i\right)}^{-1}\cdot \sum_{i=1}^{n_{GP}}V_i\mbox{ }d_i^{p,d}</math> </span>
1055
|}
1056
|  style="width: 5px;text-align: right;white-space: nowrap;"|<span style="text-align: center; font-size: 75%;">(55)</span>
1057
|}
1058
1059
1060
where  <math>d_i^{p,d}</math> are the local damage for a plastic or degradation process, and  <math display="inline">V_i</math> are the local damage variable and its Gauss point volume for each single material,  <math display="inline">n_{GP}</math> is the number of Gauss points involved in all materials included in the layers participating in the finite element.
1061
1062
The safety laminate factor can be used, for instance, in optimization process to obtain the composite material design that accomplishes with the structural and functional design parameters [38].
1063
1064
The delamination phenomenon stops when a damaged point can provide enough shear strength to equilibrate the shear stresses that appears in the inter-laminar zone.
1065
1066
A potentiality application of this structural composite formulation, including a Water Current Turbine (WCT) rotor are introduced in next section.
1067
1068
==6 “Micromodel” vs. “mixing theory”: Conceptual comparative behavior example==
1069
1070
First, a simple and conceptual example intent shows the capabilities of the formulation here presented. Here a little sample of a fiber reinforced matrix is introduced, and shows the capacities of the “mixing theory with anisotropy in large strains” comparing the results obtained with a micromodel where each of the component materials is individualized.
1071
1072
The example consists in subjecting a unit size cube of a composite material under tension in which the reinforcement fiber and matrix component are discretized. Then the results obtained by this micromodel are compared to the results obtained by the macro module proposed work. In Figure 10 the unit size piece is shown, where both phases of the composite material have been discretized. The boundary conditions imposed can also be observed. The finite element mesh is set up by 5701 triangular finite elements of 3 nodes and 2940 nodes.
1073
1074
As an alternative to the mesh described before for the numerical simulation through the model previously described, the same unit size piece modeled by only one single finite element of 4 nodes and 2 x 2 points of integration is analyzed. The sliding phenomenon between the fiber and matrix - ''Fiber-matrix debonding (FDM)''-, will not be considered in this example.
1075
1076
The mechanical properties of the materials making up the composite are shown in Table 1, and Table 2.
1077
1078
{| style="width: 100%;border-collapse: collapse;" 
1079
|-
1080
|  style="border: 2pt solid black;vertical-align: top;"|<span id='_Ref477109003'></span>Isotropic Young modulus
1081
|  style="border: 2pt solid black;vertical-align: top;"|13,00 ''MPa''
1082
|-
1083
|  style="border: 2pt solid black;vertical-align: top;"|Poisson coefficient
1084
|  style="border: 2pt solid black;vertical-align: top;"|0,325
1085
|-
1086
|  style="border: 2pt solid black;vertical-align: top;"|Yield or threshold stress
1087
|  style="border: 2pt solid black;vertical-align: top;"|43,323 ''MPa''
1088
|-
1089
|  style="border: 2pt solid black;vertical-align: top;"|Damage-Plastic Post-yield behavior law
1090
|  style="border: 2pt solid black;vertical-align: top;"|Exponential with softening 
1091
|-
1092
|  style="border: 2pt solid black;vertical-align: top;"|Fracture energy
1093
|  style="border: 2pt solid black;vertical-align: top;"|10 ''N/m''
1094
|-
1095
|  style="border: 2pt solid black;vertical-align: top;"|Volume participation'' V<sub>m</sub>''
1096
|  style="border: 2pt solid black;vertical-align: top;"|76%
1097
|}
1098
1099
1100
<span id='_Ref32658576'></span><span id='_Ref467833878'></span> Table.1 – Epoxy resin properties for the macromodel and micromodel.
1101
1102
{| style="width: 100%;border-collapse: collapse;" 
1103
|-
1104
|  style="border: 2pt solid black;vertical-align: top;"|Axial fiber Young modulus
1105
|  style="border: 2pt solid black;vertical-align: top;"|239,551 ''MPa''
1106
|-
1107
|  style="border: 2pt solid black;vertical-align: top;"|Transversal fiber  Young modulus <sup>(*)</sup>
1108
|  style="border: 2pt solid black;vertical-align: top;"|13,00 ''MPa''
1109
|-
1110
|  style="border: 2pt solid black;vertical-align: top;"|Poisson coefficient 
1111
|  style="border: 2pt solid black;vertical-align: top;"|0,0
1112
|-
1113
|  style="border: 2pt solid black;vertical-align: top;"|Yield or threshold stress
1114
|  style="border: 2pt solid black;vertical-align: top;"|3000 ''MPa''
1115
|-
1116
|  style="border: 2pt solid black;vertical-align: top;"|Damage Post-yield behavior law
1117
|  style="border: 2pt solid black;vertical-align: top;"|Linear with hardening
1118
|-
1119
|  style="border: 2pt solid black;vertical-align: top;"|Volume participation'' V<sub>f</sub>''
1120
|  style="border: 2pt solid black;vertical-align: top;"|24%
1121
|}
1122
1123
1124
<sup>(*)</sup>Note: Adopted equal to matrix elastic modulus
1125
1126
<div id="_Ref16306933" class="center" style="width: auto; margin-left: auto; margin-right: auto;">
1127
Table 2 – Fiber carbon properties for the macromodel and micromodel.</div>
1128
1129
The micromodel is made up of three materials: reinforced fiber, epoxy matrix and fiber-matrix interface. Epoxy matrix and interface material are considered isotropic and homogenous and have mechanical properties that coincide with the mechanical properties of the macromodel components. In Figure 11 the micromodel materials distribution are shown schematically.
1130
1131
<div class="center" style="width: auto; margin-left: auto; margin-right: auto;">
1132
 [[Image:Draft_Samper_893761715-image236.png|360px]] </div>
1133
1134
<div class="center" style="width: auto; margin-left: auto; margin-right: auto;">
1135
 [[Image:Draft_Samper_893761715-image237.png|414px]] </div>
1136
1137
The numerical test consists in imposing displacements on the upper part of the unit size structure producing tension. This stress state on the specimen leads to the fiber reinforcement alignment with the load direction. In Figure 12 the deformed shape in its final state is shown. It can also be observed that the fibers have aligned themselves with the direction of the applied stress. This alignment of the reinforcement phase with the stress direction makes it necessary to introduce the theory of large strains in the constitutive model.
1138
1139
1140
[[Image:Draft_Samper_893761715-image238.png|center|600px]]
1141
1142
The advantage of using a micromodel is that a detailed analysis of the mechanical processes can be done during the load application. Figure 13 shows the shear stress on the material for different loading cases. Figure13-1 shows the stress states in a loading phase in which the stresses above the elastic limits of the component materials are not verified (see the plasticity internal variable in Figure 13). It can also be observed in the same figure that the matrix zone among the fibers is the one presenting a higher tensional state. As the displacements increase (Figure 13-2, Figure 13-3 and Figure 13-4) a homogenization of the matrix stress state is observed.
1143
1144
1145
[[Image:Draft_Samper_893761715-picture-Grupo 243753.svg|center|600px]]
1146
1147
Figure 14 shows the stresses in the micromodel in the direction of the imposed displacements. Figure 14-1 corresponds to a stress state in a loading step in which the composite materials stresses above the elastic limit are not verified (see Figure 15). Figure 14-2, Figure 14-3 and Figure 14-4 show the stress state in the direction of the imposed displacement as displacements increase. It can also be observed that in the first loading steps the matrix has a homogenous stress state in the direction of the applied stresses. Figure 14-2 shows that the reinforcement increases considerably as it aligns itself with the direction of the stress applied.
1148
1149
Figure 14 shows the plasticity contours in each composite component. It also shows that as the displacement increases, the irreversible strains in the matrix are verified in the areas between reinforcements (see Figure 14-2 and Figure 14-3). In Figure 14-4 it can be observed that the elastic limit has been exceeded, consequently leading to irreversible strains.
1150
1151
1152
[[Image:Draft_Samper_893761715-image241.jpeg|center|600px]]
1153
1154
1155
[[Image:Draft_Samper_893761715-picture-x0000_s1082.svg|center|600px]]
1156
1157
.
1158
1159
1160
[[Image:Draft_Samper_893761715-image243.jpeg|center|600px]]
1161
1162
Figure 16 shows the micro and macro models loading-displacement response. Different values of the transversal module of the reinforcement phase are considered. The same figure shows that the value of the transversal elastic module of this phase plays a fundamental role in the macromodel response. When the shear modulus is zero, it is observed that the matrix reaches its limit of proportionality while the tension in the composite decreases until the fibers coincide with the direction of applied stress. Beyond this point, the reinforcement phase provides stiffness to the system. The response corresponding to the small strains assumption can also be observed. In this case, once the matrix’s elastic limit is achieved, the material response decreases and the fibers do not participate in the response. This is because according to the small strain hypothesis the geometry is not updated and consequently the fibers cannot align themselves with the applied stress direction.
1163
1164
1165
[[Image:Draft_Samper_893761715-picture-161 Grupo.svg|center|600px]]
1166
1167
==7 “Micro model” vs. “mixing theory”: Conceptual “Fiber Matrix Displacement” (debonding) behavior example==
1168
1169
This example show a simple validation and comparison between the fibers sliding effect into matrix by means a "an isotropic mixing constitutive model formulation" and an "explicit finite element micro model".
1170
1171
An example of the formulation application combining the mixing theory, the anisotropic model in large strains and the theory that includes the FMD phenomenon analysis (Fiber-Matrix Displacement) is described below. This example compares the numerical simulation of the composite material specimen (reinforced concrete) with a central notch subjected to traction where the reinforced and matrix phases have been discretized (micro model), with a similar specimen in which only a composite material made up by a reinforced phase and the matrix (macro model) exists.
1172
1173
<span id='_Ref17032636'></span>
1174
1175
1176
[[Image:Draft_Samper_893761715-picture-Grupo 243761.svg|center|600px]]
1177
1178
1179
{|
1180
|-
1181
| [[Image:Draft_Samper_893761715-picture-Cuadro de texto 184.svg|center|48px]]
1182
| [[Image:Draft_Samper_893761715-image249.png|center|600px]]
1183
|}
1184
1185
1186
1187
[[Image:Draft_Samper_893761715-picture-Grupo 187.svg|center|600px]]
1188
1189
The numerical simulations have been carried out using a linear rectangular finite element mesh of 4 nodes with a total of 343 elements, 392 nodes and 766 degrees of freedom for the micro model and 291 elements, 336 nodes and 644 degrees of freedom for the macro model. Figure 17 shows the geometry, material assignment, meshes and boundary conditions used for each case.
1190
1191
{| style="width: 100%;border-collapse: collapse;" 
1192
|-
1193
|  style="border: 1pt solid black;vertical-align: top;"|
1194
|  style="border: 1pt solid black;text-align: center;vertical-align: top;"|<span style="text-align: center; font-size: 75%;">'''Material 1 Matrix of the'''</span>
1195
1196
<span style="text-align: center; font-size: 75%;">'''Concrete'''</span>
1197
|  style="border: 1pt solid black;text-align: center;vertical-align: top;"|<span style="text-align: center; font-size: 75%;">'''Material 2'''</span>
1198
1199
<span style="text-align: center; font-size: 75%;">'''Steel'''</span>
1200
1201
<span style="text-align: center; font-size: 75%;">'''Reinforcement'''</span>
1202
|  style="border: 1pt solid black;text-align: center;vertical-align: top;"|<span style="text-align: center; font-size: 75%;">'''Material 3'''</span>
1203
1204
<span style="text-align: center; font-size: 75%;">'''Matrix-Reinforcement'''</span>
1205
1206
<span style="text-align: center; font-size: 75%;">'''Interface'''</span>
1207
|-
1208
|  style="border: 1pt solid black;"|'''Type of material behavior'''
1209
|  style="border: 1pt solid black;text-align: center;"|<span style="text-align: center; font-size: 75%;">Mohr-Coulomb</span>
1210
1211
<span style="text-align: center; font-size: 75%;">Isotropic Elasto-plastic</span>
1212
1213
<span style="text-align: center; font-size: 75%;">Model</span>
1214
|  style="border: 1pt solid black;text-align: center;"|<span style="text-align: center; font-size: 75%;">Isotropic  Elastic</span>
1215
1216
<span style="text-align: center; font-size: 75%;">Model</span>
1217
|  style="border: 1pt solid black;text-align: center;"|<span style="text-align: center; font-size: 75%;">Kachanov  Damage</span>
1218
1219
<span style="text-align: center; font-size: 75%;">Model</span>
1220
|-
1221
|  style="border: 1pt solid black;"|'''Young Modulus [kp/cm<sup>2</sup>]'''
1222
|  style="border: 1pt solid black;text-align: center;"|<span style="text-align: center; font-size: 75%;">3,5×10<sup>5</sup></span>
1223
1224
1225
|  style="border: 1pt solid black;text-align: center;"|<span style="text-align: center; font-size: 75%;">2,1×10<sup>6</sup></span>
1226
1227
1228
|  style="border: 1pt solid black;text-align: center;"|<span style="text-align: center; font-size: 75%;">3,5×10<sup>5</sup></span>
1229
1230
1231
|-
1232
|  style="border: 1pt solid black;"|'''Poisson Coefficient'''
1233
|  style="border: 1pt solid black;text-align: center;"|<span style="text-align: center; font-size: 75%;">0,2</span>
1234
|  style="border: 1pt solid black;text-align: center;"|<span style="text-align: center; font-size: 75%;">0,0</span>
1235
|  style="border: 1pt solid black;text-align: center;"|<span style="text-align: center; font-size: 75%;">0,0</span>
1236
|-
1237
|  style="border: 1pt solid black;"|'''Internal Friction'''
1238
|  style="border: 1pt solid black;text-align: center;"|<span style="text-align: center; font-size: 75%;">30º</span>
1239
|  style="border: 1pt solid black;text-align: center;"|<span style="text-align: center; font-size: 75%;">-</span>
1240
|  style="border: 1pt solid black;text-align: center;"|<span style="text-align: center; font-size: 75%;">30º</span>
1241
|-
1242
|  style="border: 1pt solid black;"|'''Compression Strength [kp/cm<sup>2</sup>]'''
1243
|  style="border: 1pt solid black;text-align: center;"|<span style="text-align: center; font-size: 75%;">200</span>
1244
|  style="border: 1pt solid black;text-align: center;"|2000
1245
|  style="border: 1pt solid black;text-align: center;"|20
1246
|-
1247
|  style="border: 1pt solid black;"|'''Tension Strength [kp/cm<sup>2</sup>]'''
1248
|  style="border: 1pt solid black;text-align: center;"|<span style="text-align: center; font-size: 75%;">20</span>
1249
|  style="border: 1pt solid black;text-align: center;"|<span style="text-align: center; font-size: 75%;">2000</span>
1250
|  style="border: 1pt solid black;text-align: center;"|<span style="text-align: center; font-size: 75%;">20</span>
1251
|-
1252
|  style="border: 1pt solid black;"|'''G<sub>f</sub> , G<sub>c </sub>[kp/cm]'''
1253
|  style="border: 1pt solid black;text-align: center;"|<span style="text-align: center; font-size: 75%;">0,25  ,   26,0</span>
1254
|  style="border: 1pt solid black;text-align: center;"|
1255
<span style="text-align: center; font-size: 75%;">* </span>
1256
|  style="border: 1pt solid black;text-align: center;"|<span style="text-align: center; font-size: 75%;">2,0  ,  2,0</span>
1257
|-
1258
|  style="border: 1pt solid black;"|'''Behavior law after the '''
1259
1260
'''Yield point'''
1261
|  style="border: 1pt solid black;text-align: center;"|<span style="text-align: center; font-size: 75%;">Line function </span>
1262
1263
<span style="text-align: center; font-size: 75%;">with softening</span>
1264
|  style="border: 1pt solid black;text-align: center;"|
1265
<span style="text-align: center; font-size: 75%;">* </span>
1266
|  style="border: 1pt solid black;text-align: center;"|<span style="text-align: center; font-size: 75%;">Exponential function</span>
1267
1268
<span style="text-align: center; font-size: 75%;">with softening</span>
1269
|}
1270
1271
1272
<div class="center" style="width: auto; margin-left: auto; margin-right: auto;">
1273
Table 3 – Mechanical properties used in the micro and macro model.</div>
1274
1275
The micro model consists of three materials: matrix, fiber-matrix interface zone and reinforcement. The macro model is made up of one composite material consisting of three component materials: reinforced fiber, matrix, and fiber-matrix interface. Table 3 shows the mechanical properties of the materials used in the micro model. The mechanical properties of the composite material phases in the macro model are identical to the corresponding ones in the matrix and reinforcement of the micro model.
1276
1277
1278
[[Image:Draft_Samper_893761715-picture-149 Grupo.svg|center|600px]]
1279
1280
The purpose of this example is to show the loading transfer phenomenon from the matrix to the reinforcement phase. This is achieved first by comparing the “load-displacement” curve obtained by the micro model and then by the macro model made up of composite material, the components of which are not possible to identify physically (mixing theory) Figures 18 and 21  show the shear stress evolution in the fiber-matrix interface zone for different loading increments. The change of sign of the stresses, which are mainly due to the presence of the notch, can be observed in the central zone. Figure 20 and 21 show the longitudinal stress evolution in the reinforcement phase for different loading increments. It can be observed that for the first loading increment the maximum shear stresses are obtained at the reinforcement end zone, while the longitudinal stresses increase from zero at the end zone to a constant value along the reinforcement. Moreover, a variation of the longitudinal stress in the central zone due to the presence of the notch is observed. Additionally, a decrease of the stress transfer capacity from the matrix to the fibers is observed. This phenomenon also causes a modification of the stress state and, as observed, the stress distribution curve along the reinforcement is no longer constant. Figure 22 shows the interface zones exceeding the material proportionality limit for different loading stages. It can be noted that the fiber-matrix relative sliding starts at the fiber’s end zone and moves towards the specimen’s center.
1281
1282
1283
[[Image:Draft_Samper_893761715-picture-150 Grupo.svg|center|600px]]
1284
.
1285
1286
<span id='_Ref17032897'></span>
1287
1288
1289
[[Image:Draft_Samper_893761715-picture-151 Grupo.svg|center|600px]]
1290
1291
1292
[[Image:Draft_Samper_893761715-picture-152 Grupo.svg|center|600px]]
1293
1294
1295
[[Image:Draft_Samper_893761715-image260.jpeg|center|600px]]
1296
1297
1298
[[Image:Draft_Samper_893761715-picture-27 Grupo.svg|center|600px]]
1299
1300
1301
{|
1302
|-
1303
| [[Image:Draft_Samper_893761715-picture-Cuadro de texto 111.svg|center|167px]]
1304
| [[Image:Draft_Samper_893761715-picture-Cuadro de texto 245249.svg|center|167px]]
1305
|}
1306
1307
1308
1309
{|
1310
|-
1311
| [[Image:Draft_Samper_893761715-picture-Cuadro de texto 245250.svg|center|167px]]
1312
| [[Image:Draft_Samper_893761715-picture-Cuadro de texto 245251.svg|center|167px]]
1313
|}
1314
1315
1316
Figure 23 shows the displacement contours in the first and final converged loading increments for the micro and macro models. As observed, in this last increment the displacements are basically in the specimen’s central zone and along the central reinforcement. A displacement is observed between the fiber and the reinforcement at the specimen’s ends.
1317
1318
1319
[[Image:Draft_Samper_893761715-picture-153 Grupo.svg|center|600px]]
1320
1321
1322
{|
1323
|-
1324
| [[Image:Draft_Samper_893761715-picture-Cuadro de texto 245253.svg|center|60px]]
1325
| [[Image:Draft_Samper_893761715-picture-Cuadro de texto 245252.svg|center|60px]]
1326
|}
1327
1328
1329
Figure 24 shows the total forces’ response for micro and macro models. It can be noted that the micro model’s results match satisfactorily those of the macro model. It is important to highlight that the micro model cannot carry out the simulation of the relative movements between different phases but it can carry out the reinforcement simulation. However, the characterization of the latter would involve a considerable high computational cost of analysis due to the carbon fibers small dimension.
1330
1331
==8 Numerical simulation of a structural analysis of a “composite material rotor-hydrofoil” of a Water Current Turbine (WCT)==
1332
1333
The constitutive model described in this chapter is used in the structural numerical simulation of the composite turbine rotor. Reduction of rotational inertia of the WCT rotor is one of the principal aims of the fiber reinforced composite material application to this kind of structure. This will lower resistance to rotation in front of the river speed changes, allowing more flexibility in starting and stopping of the turbine rotation.
1334
1335
The numerical simulation of the multilayered composite structure design of the flow axial turbine rotor by means of finite elements method is presented in this section. A comparative study considering the structural response of the steel turbine rotor vs. fibers-reinforced composite material is carried out [38]. The composite material analysis is developed employing the orthotropic mixing theory previously presented, while an isotropic constitutive model is used for the steel rotor.
1336
1337
===Geometry, boundary conditions and finite element mesh===
1338
1339
The rotor is placed under an axial water flow described in section 3 that causes a distribution of pressures on the hydrofoils. These flow pressures are obtained by CFD finite element code and, particularly, in the leading edge of the hydrofoils of the rotor. The 8 rotor blades have a hydrodynamic profile with 15º of attack angle (see Figures 2, 3 and Table 4).
1340
1341
{| style="width: 100%;margin: 1em auto 0.1em auto;border-collapse: collapse;" 
1342
|-
1343
|  style="border: 1pt solid black;text-align: center;"| <math>X\, (m)</math>
1344
|  style="border-top: 1pt solid black;border-bottom: 1pt solid black;border-right: 1pt solid black;text-align: center;"|0.175
1345
|  style="border-top: 1pt solid black;border-bottom: 1pt solid black;border-right: 1pt solid black;text-align: center;"|0.35
1346
|  style="border-top: 1pt solid black;border-bottom: 1pt solid black;border-right: 1pt solid black;text-align: center;"|0.525
1347
|  style="border-top: 1pt solid black;border-bottom: 1pt solid black;border-right: 1pt solid black;text-align: center;"|0.7
1348
|  style="border-top: 1pt solid black;border-bottom: 1pt solid black;border-right: 1pt solid black;text-align: center;"|0.875
1349
|  style="border-top: 1pt solid black;border-bottom: 1pt solid black;border-right: 1pt solid black;text-align: center;"|1.05
1350
|  style="border-top: 1pt solid black;border-bottom: 1pt solid black;border-right: 1pt solid black;text-align: center;"|1.225
1351
|  style="border-top: 1pt solid black;border-bottom: 1pt solid black;border-right: 1pt solid black;text-align: center;"|1.4
1352
|  style="border-top: 1pt solid black;border-bottom: 1pt solid black;border-right: 1pt solid black;text-align: center;"|1.575
1353
|  style="border-top: 1pt solid black;border-bottom: 1pt solid black;border-right: 1pt solid black;text-align: center;"|1.75
1354
|-
1355
|  style="border-left: 1pt solid black;border-bottom: 1pt solid black;border-right: 1pt solid black;text-align: center;"| <math>L\, (m)</math>
1356
|  style="border-bottom: 1pt solid black;border-right: 1pt solid black;text-align: center;"|0.182
1357
|  style="border-bottom: 1pt solid black;border-right: 1pt solid black;text-align: center;"|0.145
1358
|  style="border-bottom: 1pt solid black;border-right: 1pt solid black;text-align: center;"|0.128
1359
|  style="border-bottom: 1pt solid black;border-right: 1pt solid black;text-align: center;"|0.116
1360
|  style="border-bottom: 1pt solid black;border-right: 1pt solid black;text-align: center;"|0.108
1361
|  style="border-bottom: 1pt solid black;border-right: 1pt solid black;text-align: center;"|0.102
1362
|  style="border-bottom: 1pt solid black;border-right: 1pt solid black;text-align: center;"|0.104
1363
|  style="border-bottom: 1pt solid black;border-right: 1pt solid black;text-align: center;"|0.093
1364
|  style="border-bottom: 1pt solid black;border-right: 1pt solid black;text-align: center;"|0.089
1365
|  style="border-bottom: 1pt solid black;border-right: 1pt solid black;text-align: center;"|0.086
1366
|-
1367
|  style="border-left: 1pt solid black;border-bottom: 1pt solid black;border-right: 1pt solid black;text-align: center;"| <math>\beta \, (\mbox{°})</math>
1368
|  style="border-bottom: 1pt solid black;border-right: 1pt solid black;text-align: center;"|63.40
1369
|  style="border-bottom: 1pt solid black;border-right: 1pt solid black;text-align: center;"|52.23
1370
|  style="border-bottom: 1pt solid black;border-right: 1pt solid black;text-align: center;"|45.99
1371
|  style="border-bottom: 1pt solid black;border-right: 1pt solid black;text-align: center;"|38.84
1372
|  style="border-bottom: 1pt solid black;border-right: 1pt solid black;text-align: center;"|32.83
1373
|  style="border-bottom: 1pt solid black;border-right: 1pt solid black;text-align: center;"|27.81
1374
|  style="border-bottom: 1pt solid black;border-right: 1pt solid black;text-align: center;"|23.59
1375
|  style="border-bottom: 1pt solid black;border-right: 1pt solid black;text-align: center;"|19.52
1376
|  style="border-bottom: 1pt solid black;border-right: 1pt solid black;text-align: center;"|17.04
1377
|  style="border-bottom: 1pt solid black;border-right: 1pt solid black;text-align: center;"|14.59
1378
|-
1379
|  style="border-left: 1pt solid black;border-bottom: 1pt solid black;border-right: 1pt solid black;text-align: center;"| <math>\alpha \, (\mbox{°})</math>
1380
|  style="border-bottom: 1pt solid black;border-right: 1pt solid black;text-align: center;"|16.23
1381
|  style="border-bottom: 1pt solid black;border-right: 1pt solid black;text-align: center;"|15.61
1382
|  style="border-bottom: 1pt solid black;border-right: 1pt solid black;text-align: center;"|15.2
1383
|  style="border-bottom: 1pt solid black;border-right: 1pt solid black;text-align: center;"|14.89
1384
|  style="border-bottom: 1pt solid black;border-right: 1pt solid black;text-align: center;"|14.65
1385
|  style="border-bottom: 1pt solid black;border-right: 1pt solid black;text-align: center;"|14.46
1386
|  style="border-bottom: 1pt solid black;border-right: 1pt solid black;text-align: center;"|14.33
1387
|  style="border-bottom: 1pt solid black;border-right: 1pt solid black;text-align: center;"|14.79
1388
|  style="border-bottom: 1pt solid black;border-right: 1pt solid black;text-align: center;"|14.17
1389
|  style="border-bottom: 1pt solid black;border-right: 1pt solid black;text-align: center;"|14.05
1390
|}
1391
1392
1393
From the geometry of the rotor a mesh of 4100 linear shell triangular finite elements is generated (rotational-free shell triangle, [22]), with 2012 nodes (Figure 25).
1394
1395
These shells structure are analyzed firstly made of steel material and then made of laminate composites material with layers of epoxy matrix reinforced by unidirectional carbon fibers. In both cases, the properties of the materials are detailed in tables 4 and 5.
1396
1397
===Action on the hydrofoil’s rotor===
1398
1399
The water pressures obtained by the Section 3 formulation on the blade of the turbine rotor in the axial axis are applied. These pressures are obtained from CFD code for the fluvial flow at low speed river (see section 3)  [15,29]. This pressures cause two kinds of loads in the rotor:
1400
1401
:* '''Load 1''': Rotation loads on the surfaces of the hydrofoils produced by the differential pressures between the up and down surfaces of the wing. This load is obtained by the CFD finite element code (Computational Fluid Dynamics) to obtain the speed, correct attack angle of hydrofoils, diagrams of pressures on the wing areas, etc.
1402
1403
:* '''Load 2:''' Axial loads caused by the directly applied pressures over the attack edge of the hydrofoil, that cause its deformation and the tensional state of the rotor, trending to break it in the perpendicular direction of the plane of the rotor. This reaction forces are studied and analyzed in this example through the previously composite-formulation included in a structural finite element program (Figure 26).
1404
1405
1406
[[Image:Draft_Samper_893761715-image264-c.jpeg|center|600px]]
1407
1408
1409
1410
{|
1411
|-
1412
| [[Image:Draft_Samper_893761715-picture-Flecha derecha 242.svg|center|74px]]
1413
| [[Image:Draft_Samper_893761715-image265-c.jpeg|center|600px]]
1414
|}
1415
1416
1417
Kinematic pressures are obtained by CFD finite element procedure which is available on the reference Oller et al. [15,29].  Thus, using this pressure has been obtained an applied load of  
1418
{|
1419
|-
1420
| <math>F_{\mbox{Rotor}}=\mbox{ }</math>
1421
| <math>672N\mbox{ }</math>
1422
|}
1423
at the leading edges surfaces of the  <math display="inline">n_{\mbox{wing}}\mbox{ }</math> hydrofoils. This load is distributed over all nodes of the blades, and is applied in one time step, in the rotor at time [[Image:Draft_Samper_893761715-image269.png|84px]] .  A more complete fluid dynamic numerical simulation using the section 2 a 3 formulation can be found in reference [15,29].
1424
1425
The restrictions of movement are applied to the nodes corresponding to the turbine shaft, representing the sharing points between the rotor and the axis of the turbine.
1426
1427
===Numerical simulations of the rotor made of steel and composite material.===
1428
1429
The details of the structural behavior of each rotor conformed in steel and composite laminate are described below.  In the structural analysis of the composite laminate, the previously general formulation is used. An additional parametric comparison is also carried out in this case chosen from three pairs of fiber directions and stacking sequences.
1430
1431
Numerical simulations involves the turbine steel rotor with the following mechanical parameters: density <math>m=7850kg/m^3\mbox{ }</math> , Young modulus <math>E=210.0\times {10}^6\mbox{ }Pa\mbox{ }\mbox{ }</math> , Poisson ratio  <math>\nu =0.3\mbox{ }</math> , and thickness <math>t=1.2\mbox{ }mm\mbox{ }</math> ; and the turbine rotor made by composite six layers, each one with <math>e=0.30mm\mbox{ }</math> thickness, and three different composite layups: <math>\pm {45}^0\mbox{ }\mbox{ }</math> , 0º-90º and 0º-45º, (see mechanical properties of components in Table 5), are used in the analysis comparison.
1432
1433
It can be observed after the ''1s'' applied load, that the minimum stress,  <math>\sigma =10622\, Pa\mbox{ }\mbox{ }</math> (Table 6 and Figure 26b), corresponds to the non-orthogonal 0º-45º layup configuration, which occurs in the blades near to the shaft junction.
1434
1435
{| style="width: 61%;border-collapse: collapse;" 
1436
|-
1437
|  style="border: 1pt solid black;"|'''Composite'''
1438
|  style="border: 1pt solid black;width: 35%;"| <math display="inline">e=6\times 0.3mm=1.8mm</math> 
1439
|-
1440
|  rowspan='5' style="border: 1pt solid black;text-align: center;"|<span style="text-align: center; font-size: 75%;">Matrix</span>
1441
|  style="border: 1pt solid black;width: 35%;"| <math display="inline">\rho =1200kg/m^3</math> 
1442
|-
1443
|  style="border: 1pt solid black;width: 35%;"| <math display="inline">E=4.0\times {10}^9\mbox{ }Pa</math> 
1444
|-
1445
|  style="border: 1pt solid black;width: 35%;"| <math display="inline">\nu =0.353</math> 
1446
|-
1447
|  style="border: 1pt solid black;width: 35%;"| <math display="inline">f_T=610MPa</math> 
1448
|-
1449
|  style="border: 1pt solid black;"|60% of volume fraction
1450
|-
1451
|  rowspan='4' style="border: 1pt solid black;text-align: center;"|<span style="text-align: center; font-size: 75%;">Fiber</span>
1452
1453
<span style="text-align: center; font-size: 75%;">reinforcement</span>
1454
|  style="border: 1pt solid black;width: 35%;"| <math display="inline">\rho =1800kg/m^3</math> 
1455
|-
1456
|  style="border: 1pt solid black;width: 35%;"| <math display="inline">E=242.1\times {10}^9\mbox{ }Pa</math> 
1457
|-
1458
|  style="border: 1pt solid black;width: 35%;"| <math display="inline">\nu =0.22</math> 
1459
|-
1460
|  style="border: 1pt solid black;width: 35%;"| <math display="inline">f_T^{long}=3800MPa</math> 
1461
|}
1462
1463
1464
<div class="center" style="width: auto; margin-left: auto; margin-right: auto;">
1465
Table 5 – Mechanical properties of matrix and fibers of the composite material.</div>
1466
1467
{| style="width: 100%;border-collapse: collapse;" 
1468
|-
1469
|  style="border: 1pt solid black;text-align: center;"|'''Type of rotor'''
1470
|  style="border-top: 1pt solid black;border-bottom: 1pt solid black;border-right: 1pt solid black;text-align: center;"|'''T                   [mm]'''
1471
|  style="border-top: 1pt solid black;border-bottom: 1pt solid black;border-right: 1pt solid black;text-align: center;"|'''Required Starting torque  [Nm]'''
1472
|  style="border-top: 1pt solid black;border-bottom: 1pt solid black;border-right: 1pt solid black;text-align: center;"|'''Maximum stress            [Pa]'''
1473
|  style="border-top: 1pt solid black;border-bottom: 1pt solid black;border-right: 1pt solid black;text-align: center;"|'''Maximum displacement [m]'''
1474
|-
1475
|  style="border-left: 1pt solid black;border-bottom: 1pt solid black;border-right: 1pt solid black;vertical-align: bottom;"|'''Steel'''
1476
|  style="border-bottom: 1pt solid black;border-right: 1pt solid black;text-align: center;vertical-align: bottom;"|<span style="text-align: center; font-size: 75%;">1.8</span>
1477
|  style="border-bottom: 1pt solid black;border-right: 1pt solid black;text-align: center;vertical-align: bottom;"|<span style="text-align: center; font-size: 75%;">217</span>
1478
|  style="border-bottom: 1pt solid black;border-right: 1pt solid black;text-align: center;vertical-align: bottom;"|<span style="text-align: center; font-size: 75%;">25.324,00</span>
1479
|  style="border-bottom: 1pt solid black;border-right: 1pt solid black;text-align: center;vertical-align: bottom;"|<span style="text-align: center; font-size: 75%;">2,21E-03</span>
1480
|-
1481
|  style="border-left: 1pt solid black;border-bottom: 1pt solid black;border-right: 1pt solid black;vertical-align: bottom;"|'''Comp. ±45º'''
1482
|  style="border-bottom: 1pt solid black;border-right: 1pt solid black;text-align: center;vertical-align: bottom;"|<span style="text-align: center; font-size: 75%;">6×0.3=1.8</span>
1483
|  style="border-bottom: 1pt solid black;border-right: 1pt solid black;text-align: center;vertical-align: bottom;"|<span style="text-align: center; font-size: 75%;">40</span>
1484
|  style="border-bottom: 1pt solid black;border-right: 1pt solid black;text-align: center;vertical-align: bottom;"|<span style="text-align: center; font-size: 75%;">13.397,00</span>
1485
|  style="border-bottom: 1pt solid black;border-right: 1pt solid black;text-align: center;vertical-align: bottom;"|<span style="text-align: center; font-size: 75%;">1,65E-03</span>
1486
|-
1487
|  style="border-left: 1pt solid black;border-bottom: 1pt solid black;border-right: 1pt solid black;vertical-align: bottom;"|'''Comp. 0º/45º'''
1488
|  style="border-bottom: 1pt solid black;border-right: 1pt solid black;text-align: center;vertical-align: bottom;"|<span style="text-align: center; font-size: 75%;">6×0.3=1.8</span>
1489
|  style="border-bottom: 1pt solid black;border-right: 1pt solid black;text-align: center;vertical-align: bottom;"|<span style="text-align: center; font-size: 75%;">40</span>
1490
|  style="border-bottom: 1pt solid black;border-right: 1pt solid black;text-align: center;vertical-align: bottom;"|<span style="text-align: center; font-size: 75%;">10.622,00</span>
1491
|  style="border-bottom: 1pt solid black;border-right: 1pt solid black;text-align: center;vertical-align: bottom;"|<span style="text-align: center; font-size: 75%;">2,73E-03</span>
1492
|-
1493
|  style="border-left: 1pt solid black;border-bottom: 1pt solid black;border-right: 1pt solid black;vertical-align: bottom;"|'''Comp. 0º/90º'''
1494
|  style="border-bottom: 1pt solid black;border-right: 1pt solid black;text-align: center;vertical-align: bottom;"|<span style="text-align: center; font-size: 75%;">6×0.3=1.8</span>
1495
|  style="border-bottom: 1pt solid black;border-right: 1pt solid black;text-align: center;vertical-align: bottom;"|<span style="text-align: center; font-size: 75%;">40</span>
1496
|  style="border-bottom: 1pt solid black;border-right: 1pt solid black;text-align: center;vertical-align: bottom;"|<span style="text-align: center; font-size: 75%;">12.342,00</span>
1497
|  style="border-bottom: 1pt solid black;border-right: 1pt solid black;text-align: center;vertical-align: bottom;"|<span style="text-align: center; font-size: 75%;">2,42E-03</span>
1498
|}
1499
1500
1501
<div class="center" style="width: auto; margin-left: auto; margin-right: auto;">
1502
Table 6 – Comparison between the four most significant rotors.</div>
1503
1504
1505
[[Image:Draft_Samper_893761715-picture-Grupo 21.svg|center|600px]]
1506
1507
Figure 26b and 26d shows that the "0º/45º” laminate composite rotor works with nearly 40% level of stresses of the steel rotor. The "±45º" composite laminate rotor and the "0º/90º" composite laminate also work at less stresses than steel rotor (Table 6).
1508
1509
All composite laminate rotors have less stiffness than the steel rotor, but particularly the composite with fibers oriented to "±45º has a high stiffness and near to the steel value (see its relative comparison, Figure 26d). However, the composite laminate of "0º/45º” has a much lower stiffness than the other (Figure 26d), but is enough for this machine requirements, as its maximum displacement is tolerable in these work functions (see its relative comparison, Figure 26c).
1510
1511
The reduced starting torque of composite laminate rotor is a big advantage during the operating work of the water turbine, since composites have 5.5 times less starting torque than the steel rotor (Figure 26a). It means a machine with better performances at low water flux velocities, easier to ship, handle, repair, start, etc.
1512
1513
Concluding, the composite laminated rotor of fibers oriented to "±45º" is the best suited material for this function, since it has a very low rotational inertia (18.3% of the steel rotor. See Figure 26a), a maximum working stress a 47% lower than the steel rotor (Figure 26b), and finally has a good stiffness (53% of the value corresponding to the steel rotor. (see Figure 26d)
1514
1515
==9 Conclusions==
1516
1517
Riverbed hydroelectric development is a concept that is currently being successfully explored by several researchers, and the use of composite materials in the design of these turbines adds a significant technical and economic improvement over the use of classical materials as shown in this work.
1518
1519
This book chapter’s presents a numerical formulation for the design and analysis of composite material structures to be used within the energy sector of renewable energies.
1520
1521
The numerical formulation is based on the Serial/Parallel mixing theory as manager of the different constituent models for the composite material components. Also the local damage constitutive model is employed to set the threshold for the initiation of non-linear damage process (initiation damage) in each point of the composite material. These concepts allow a threshold criterion of damage resulting from the composition of the behaviors of all material involved in the theory of mixtures. The homogenization of local damages obtained in each one of the composite constituents to measure the damage in a shell structure is also presented.
1522
1523
Taking advantage of the composite materials compared to the classic ones, a very good performance of the Water Current Turbines (WCT) immersed in the rivers is obtained, improving its performance thanks to its smaller rotational inertia.
1524
1525
The numerical model for the treatment of the Composites of Water Current Turbines (WCT), and its fluid dynamics, shows a way forward and establishes a work base that offers an important tool for the design and analysis of these turbines.
1526
1527
==10 Future works==
1528
1529
Two research sub-lines are proposed below to future research in the electric-power exploitation using hydrokinetic turbines.
1530
1531
A) From the turbine point of view and its location in each particular place, the research progress is currently made on fluid-dynamic design to optimize the rotor and the rest of the machine. In this order of things are proposed new devices and infrastructures of hydraulic flow control, which will allow a speed increase at the entrance to the turbine to improve the electrical performance. Another aspect that is also being studied is the design of sanders and separation of material in suspension to avoid impact of objects on the rotor. All this type of studies is related to the particular behavior of each kind of river.
1532
1533
B) From the point of view of the composite material constituting the turbine, the central theme of this chapter, future research is aimed at obtaining new fabric-reinforced compounds, which are more resistant, rigid and appropriate to withstand impacts, than compounds Reinforced with unidirectional fibers. To this end, the formulation of the S / P Theory of Mixtures presented here is currently being reformulated and generalized to fit these new materials. This, along with the great deformations and the incorporation of the misalignment of the fibers of the fabric, will allow to properly considering the interaction between orthogonal fibers located in the same plane, bases for the numerical simulation of the fabrics-reinforced composites.
1534
1535
==11 Nomenclature==
1536
1537
''A          Swept area''
1538
1539
<math display="inline">{\alpha }_{0}</math>       Design attack angle (º)
1540
1541
<math display="inline">\alpha</math>       Real angle of attack (º)
1542
1543
<math display="inline">{\alpha }_{max}</math>     Maximum aerodynamic profile’s <math display="inline">\alpha</math>  (º)
1544
1545
''&#x03b2;''       Hydrofoil’s camber angle (º)
1546
1547
<math display="inline">C</math>       Airfoil chord (''m'')
1548
1549
<math display="inline">c</math>             Relative flow velocity
1550
1551
<math display="inline">{c}_{y}</math>       Lift coefficient
1552
1553
<math display="inline">{c}_{x}</math>       Drag coefficient
1554
1555
 <math>d_x,d_y,d_z</math> VC2 dimensions
1556
1557
<math display="inline">KL</math>      Ratio between <math display="inline">\, X</math> and  <math display="inline">{L}_{avg}</math>
1558
1559
<math display="inline">{L}_{avg}</math>     Chord average
1560
1561
<math display="inline">{L}_{x}</math>      Airfoil’s chord for each <math display="inline">x</math> (''m'')
1562
1563
 <math>p\mbox{ }</math> Pressure (''Pa'')
1564
1565
 <math>g</math> Gravity acceleration ( <math display="inline">m/s^2</math> )
1566
1567
<math display="inline">\rho</math>        Fluid density ( <math display="inline">\frac{Kg}{{m}^{3}}</math>)
1568
1569
 <math display="inline">R</math> Radius of the rotor
1570
1571
<math display="inline">SP</math>         Shape factor
1572
1573
<math display="inline">{SP}_{x}</math>        Shape factor for each <math display="inline">x</math>
1574
1575
<math display="inline">T</math>        Torque (''Nm'')
1576
1577
<math display="inline">t</math>            Time (''s'')
1578
1579
''&#x03b8;''        Hydrofoil’s sustentation angle (º)
1580
1581
<math display="inline">u</math>        Blade linear rotational speed (''m/s'')
1582
1583
<math display="inline">v</math>        Absolute flow velocity (''m/s'')
1584
1585
<math display="inline">\lambda</math> ''           '' <math display="inline">TSR</math>'':  ''Ratio between <math display="inline">c</math> and <math display="inline">u</math>
1586
1587
<math display="inline">W</math>       Power (w)
1588
1589
<math display="inline">\omega</math>       Angular speed (''rad/s'')
1590
1591
<math display="inline">x</math>       Length of each hydrofoil segment (m)
1592
1593
<math display="inline">X</math>            Wingspan (m)
1594
1595
 <math display="inline">n_{\mbox{wing}}</math> Number of blades of the rotor
1596
1597
 <math display="inline">F</math> Load (''N'')
1598
1599
 <math>E</math> Young modulus ( [[Image:Draft_Samper_893761715-image298.png|48px]] )
1600
1601
 [[Image:Draft_Samper_893761715-image299.png|12px]] Poisson ratio
1602
1603
 [[Image:Draft_Samper_893761715-image300.png|18px]] Tension strength ( [[Image:Draft_Samper_893761715-image298.png|48px]] )
1604
1605
''e ''           Thickness (''m'')
1606
1607
 [[Image:Draft_Samper_893761715-image301.png|12px]] Material strain
1608
1609
 <math display="inline">{\boldsymbol{\epsilon }}^P</math> Plastic material strain
1610
1611
 <math>{}^i\boldsymbol{\epsilon }</math> i-Component material strain
1612
1613
 <math display="inline">{\epsilon }_V\mbox{ }</math> Volumetric deformation
1614
1615
 <math display="inline">k_i</math> Volumetric participation coefficient of the  <math display="inline">i^{th}</math> component
1616
1617
 <math>{\Psi }_i</math> Free energy of the  <math display="inline">i^{th}</math> component
1618
1619
 <math display="inline">\sigma </math> Composite material stress ( [[Image:Draft_Samper_893761715-image298.png|48px]] )
1620
1621
 <math>{\sigma }_i</math> i-Component material stress ( [[Image:Draft_Samper_893761715-image298.png|48px]] )
1622
1623
 <math display="inline">{\mathbb{C}}_i^S\, ,\mbox{ }\mbox{ }{\mathbb{C}}_i^T</math> i-Secant and i-Tangent Constitutive tensors
1624
1625
 <math display="inline">{\boldsymbol{\epsilon }}_s\mbox{ },\mbox{ }\mbox{ }{\boldsymbol{\epsilon }}_p</math> Serial and Parallel strains
1626
1627
 <math display="inline">{\sigma }_s\mbox{ },\mbox{ }\mbox{ }{\sigma }_p</math> Serial and Parallel stresses ( [[Image:Draft_Samper_893761715-image298.png|48px]] )
1628
1629
 <math display="inline">\boldsymbol{N}_P\mbox{ }</math> 4th-order parallel projector tensor
1630
1631
 <math display="inline">\boldsymbol{P}_P\mbox{ }</math> 4th-order complementary serial projector tensor
1632
1633
 <math display="inline">A_{}^{\sigma }\mbox{ }</math> 4th-order stress mapping tensor
1634
1635
 <math display="inline">A_{}^{\epsilon }\mbox{ }</math> 4th-order strain mapping tensor
1636
1637
 <math display="inline">{\overline{\sigma }}_{}\mbox{ }</math> Isotropic fictitious stress ( [[Image:Draft_Samper_893761715-image298.png|48px]] )
1638
1639
 <math display="inline">{\overline{\epsilon }}_{}\mbox{ }</math> Isotropic fictitious strain
1640
1641
 <math display="inline">\eta </math> Kinematic plastic flow orientation
1642
1643
 <math display="inline">{\Xi }_m</math> Plastic dissipation energy
1644
1645
 <math>{\left(f^R\right)}_{fib}\mbox{ }</math> Fiber strength ( [[Image:Draft_Samper_893761715-image298.png|48px]] )
1646
1647
 <math>{\left(f^N\right)}_{fib}\mbox{ }</math> Nominal fiber strength ( [[Image:Draft_Samper_893761715-image298.png|48px]] )
1648
1649
 <math>{\left(f^N\right)}_{mat}\mbox{ }</math> Nominal matrix strength ( [[Image:Draft_Samper_893761715-image298.png|48px]] )
1650
1651
 <math>{\left(f^N\right)}_{fib-mat}\mbox{ }</math> Nominal fiber-matrix strength ( [[Image:Draft_Samper_893761715-image298.png|48px]] )
1652
1653
 <math>F^P\left(\sigma ;q^p\right)\mbox{ }\mbox{ },\mbox{ }\mbox{ }\mbox{ }F^d\left({\sigma }_0;q^d\right)\mbox{ }\mbox{ }</math> Plastic and Damage thresholds
1654
1655
 <math>q^p\mbox{ },\mbox{ }\mbox{ }\mbox{ }q^d</math> Group of the Plastic and Damage internal variables
1656
1657
 <math>G^P\left(\sigma ;q^p\right)\mbox{ }\mbox{ }</math> Plastic potential
1658
1659
 <math display="inline">G^d</math> Function of the damage strength softening evolution
1660
1661
 <math display="inline">G_c^d</math> Damage compression energy of the material
1662
1663
 <math>d^p\mbox{ },\mbox{ }\mbox{ }\mbox{ }d^d</math> Plastic and Damage internal variables
1664
1665
 <math display="inline">d^L</math> Structural damage index
1666
1667
==12 References==
1668
1669
[1] F. Rastellini, S. Oller, O. Salomón, E. Oñate, Composite material non-linear modelling for long fibre-reinforced laminates. Continuum basis, computational aspects and validations. Computers and Structures 86 (2008) 879-896.
1670
1671
[2] X. Martinez, S. Oller, E. Barbero, Study of delamination in composites by using the serial/parallel mixing theory and a damage formulation, Composites 2007, U. Porto. 12th to 14th Septemer/2007.
1672
1673
[3] E.S. Sánchez-Palencia, Boundary layers and edge effects in composites, in E. Sánchez Palencia, A. Zaoui (Eds), Homogenization Techniques for Composite Media, Springer-Verlag, Berlin, 1987, 121-192.
1674
1675
[4] S. Oller, J. Miquel, F. Zalamea, Composite material behavior using a homogenization double scale method, Journal of Engineering Mechanics131 (2005) 65-79.
1676
1677
[5] F. Otero, S. Oller, X. Martinez, O. Salomón, Numerical homogenization for composite materials analysis. Comparison with other micro mechanical formulations. Composite Structures 122 (2015) 405-416.
1678
1679
[6] S. Oller, E. Car, J. Lubliner, Definition of a general implicit orthotropic yield criterion.  Computer Methods, Applied Mechanics and Engineering 192 (2003) 895-912.
1680
1681
[7] C. Trusdell, R. Toupin, The classical field theories, Handbuch der physic, iii/I ed., Springer-Verlag, Berlin,1960.
1682
1683
[8] E. Car, S. Oller, E. Oñate, An anisotropic elasto-plastic constitutive model for large strain analysis of fiber reinforced composite materials, Computer Methods in Applied Mechanics and Engineering 185 (2000) 245-277.
1684
1685
[9] F. Otero, X. Martinez, S. Oller, O. Salomón, An efficient multi-scale method for non-linear analysis of composite structures, Composite Structures 131 (2015) 707-719.
1686
1687
[10] L.I. Lago, F.L. Ponta, L. Chen, Advances and trends in hydrokinetic turbine systems. Energy for Sustainable Development 14 (2010) 287–296.
1688
1689
[11] S. Ben Elghali, M. Benbouzid, J. Charpentier, Marine tidal current electric power generation technology: state of the art and current status, IEEE International Electric Machines & Drives Conference 2 (2007) IEMDC'07.
1690
1691
[12] M.A.R. Shafei, D.K. Ibrahim, A.M. Ali, M.A.A. Younes, E.A. El-Zahab, Novel approach for hydrokinetic turbine applications, Energy for Sustainable Development 27 (2015) 120–126.
1692
1693
[13] M. Khan, M.M Iqbal, J. Quaicoe, River current energy conversion systems: Progress, prospects and challenges, Renewable and Sustainable Energy Reviews 12 (2008) 2177–2193.
1694
1695
[14] M. Khan, G. Bhuyan, M. Iqbal, J. Quaicoe, Hydrokinetic energy conversion systems and assessment of horizontal and vertical axis turbines for river and tidal applications: A technology status review, Applied Energy 86 (2009) 1823-1835.
1696
1697
[15] S.A. Oller Aramayo, L.G. Nallim, S. Oller, Fluid dynamic design of an axial rotor for hydrokinetic riverbed turbine-improvement introduced by a high lift foil profile, Environmental Progress & Sustainable Energy 35 (2016) 1198–1206.
1698
1699
[16] M.S. Selig, J.J. Guglielmo, High-Lift Low Reynolds Number Airfoil Design, Journal of aircraft 34 (1997) 72-79.
1700
1701
[17] S.A. Oller Aramayo, L.G. Nallim, S. Oller, The usability of the Selig S1223 profile airfoil as a high lift hydrofoil for hydrokinetic application, Journal of Applied Fluid Mechanics 9 (2016) 537-542.
1702
1703
[18] J.N. Goundar, M. Rafiuddin Ahmed, Y. Lee, Numerical and experimental studies on hydrofoils for marine current turbines, Renewable Energy 42 (2012) 173-179.
1704
1705
[19] A. Betz, Introduction to the Theory of Flow Machines, Pergamon Press, Oxford, New York, 1966.
1706
1707
[20] W. Froude, On the elementary relation between pitch, slip and propulsive efficiency, Transactions of the Royal Institute of Naval Architects (1878).
1708
1709
[21] J.S. Carlton, Marine propellers and propulsion,2th ed., Elsevier, Oxford, 2012.
1710
1711
[22] T. Hughes, Multiscale phenomena, Green’s functions, the Dirichletto-Neumann formulation, subgrid-scale models, bubbles and the origin of stabilized methods, Computer Methods in Applied Mechanics and Engineering 127 (1995) 387–401.
1712
1713
[23] J. Guermond, Stabilization of Galerkin approximations of transport equations by subgrid modeling, M2AN 33 (1999) 1293–1316.
1714
1715
[24] M. van Tooren, Response 1 - Airbus composite aircraft fuselages – next or never, in: C. Vermeeren (Ed), Around Glare a New Aircraft Material in Context, Springer, Netherlands, 2007, pp. 145-158.
1716
1717
[25] J. Ferziger, M. Perić. Computational methods for fluid dynamics, Springer-Verlag, Berlin Heidelberg, New York, 2002.
1718
1719
[26] O.C. Zienkiewicz, L.R. Taylor. The finite element method, McGraw-Hill, London, England, 1991.
1720
1721
[27] X. Martinez, F.  Rastellini, F. Flores, S. Oller, E. Oñate, Computationally optimized formulation for the simulation of Composite materials and delamination failures, Composites Part B: Engineering  Part B 42 (2011) 134–144.
1722
1723
[28] J. Oliver, M. Cervera, S. Oller, J. Lubliner, A Simple Damage Model For Concrete, Including Long Term Effects, Second International Conference on Computer Aided Analysis and Design of Concrete Structures, Viena,  2 (1990) 945-958.
1724
1725
[29] X. Martinez, S. Oller, F. Rastellini, A. Barbat, A numerical procedure simulating RC structures reinforced with FRP using the serial/parallel mixing theory, Computers and Structures 86 (2008) 1604 – 1618.
1726
1727
[30] X. Martinez, S. Oller, Numerical Simulation of Matrix Reinforced Composite Materials Subjected to Compression Loads, Arch Comput Methods 16 (2009) 357-397.
1728
1729
[31] E. Car, S. Oller, E. Oñate, A Large Strain Plasticity for Anisotropic Materials – Composite Material Application, International Journal of Plasticity 17 (2001) 1437-1463.
1730
1731
[32] S. Oller, E. Oñate, A Hygro-Thermo-Mechanical constitutive model for multiphase composite materials, Int J Solids and Structures 33 (1996) 3179-3186.
1732
1733
[33] J. Lubliner, Thermomechanics of Deformable Bodies. Department of Civil Engineering, University of California, Lecture Notes, Berkeley, U.S.A, 1985.
1734
1735
[34] J. Lubliner, Plasticity theory. MacMillan, New York, 1990.
1736
1737
[35] G.A. Maugin, The thermomechanics of plasticity and fracture, University Press, Cambridge, 1992.
1738
1739
[36] J. Lubliner, J. Oliver, S. Oller, E. Oñate, A plastic-damage model for concrete, International Journal of Solids and Structures 25  (1989) 299 - 326.
1740
1741
[37] S. Oller, E. Oñate, J. Oliver, J. Lubliner, Finite element non-linear analysis of concrete structures using a plastic-damage model, Engineering Fracture Mechanics 35 (1990) 219-231.
1742
1743
[38] S.A. Oller Aramayo, L.G. Nallim, S. Oller, An Integrated procedure for the structural design of a composite rotor-hydrofoil of a water current turbine (WTC), Applied Composite Materials 20 (2013) 1273-1288.
1744
1745
[39] S. Oller, Numerical Simulation of Mechanical Behavior of Composite Materials. CIMNE-Springer, Barcelona, Spain, 2014. ISBN 978-3-319-04932-8.
1746
1747
[40] S. Oller, Nonlinear dynamics of structures. CIMNE-Springer, Barcelona, Spain, 2014. ISBN 978-3-319-05193-2.
1748
1749
[41] J. A. Paredes, A H. Barbat and S. Oller, A compression-tension concrete damage model, applied to a wind turbine reinforced concrete tower. Engineering Structures 33 (2011), pp. 3559-3569. ISSN: 0141-0296. DOI information: 10.1016/j.engstruct.2011.07.020.
1750
1751
[42] E. Comellas, S. Ivvan Valdez, S. Oller, S. Botello, Optimization method for the determination of material parameters in damaged composite structures. Composite Structures 122 (2015), pp. 417–424. ISSN: 0263-8223, doi:10.1016/j.compstruct.2014.12.014.
1752
1753
[43] F. Otero, X. Martínez, S. Oller, O. Salomón, Study and prediction of the mechanical performance of a nanotube-reinforced composite. Composite Structures. Vol. 94, pp.2920–2930. 2012. ISSN: 0263-8223
1754
1755
[44] R. F. Rango, L. G. Nallim, S. Oller, Static and dynamic analysis of thick laminated plates using enriched macroelements. Composite Structures, Volume 101, 2013, Pages 94–103. ISSN: 0263-8223.
1756
1757
[45] M.A. Pérez, X. Martínez, S. Oller, L. Gil, F. Rastellini, F. Flores, Impact damage prediction in carbon fiber-reinforced laminated composite using the matrix-reinforced mixing theory. Composite Structures. Volume 104, issue, year 2013, pp. 239 - 248. ISSN: 0263-8223.
1758
1759
[46] R. F. Rango, L. G. Nallim, S. Oller, An Enriched Macro Finite Element for the Static Analysis of Thick General Quadrilateral Laminated Composite Plates. Mechanics of Advanced Materials and Structures (2015).  ISSN 1537-6494 (Print), 1537-6532 (Online).  DOI: 10.1080/15376494.2015.1068400
1760
1761
[47] M. Petracca, L. Pelà, R. Rossi, S. Oller, G. Camata. E. Spacone, Regularization of first order computational homogenization for multiscale analysis of masonry structures.  Computational Mechanics, 57(2), 257-276, (2016). DOI 10.1007/s00466-015-1230-6. Print ISSN: 178-7675, Online ISSN: 0178-7675.
1762
1763
[48] F. Otero, X. Martínez, S. Oller, O. Salomón, An efficient multi-scale method for non-linear analysis of composite structures. Composite Structures, Volume 131, Pages 707-719. 2015. ISSN: 0263-8223, doi:10.1016/j.compstruct.2015.06.006.
1764
1765
[49] J. Paredes, S. Oller, A. Barbat (2016). New Tension-Compression Damage Model for Complex Analysis of Concrete Structures. Journal of Engineering Mechanics – ASCE. 04016072. Jun./2016. ISSN: 0733-9399. doi: 10.1061/(ASCE)EM.1943-7889.0001130
1766
1767
[50] C. Escudero, S. Oller, X. Martinez, A. Barbat, A laminated structural finite element for the behavior of large non-linear reinforced concrete structures. Finite Elements in Analysis and Design. Vol. 119, 15 October 2016, pp 78–94. ISSN: 0168-874X. doi:10.1016/j.finel.2016.06.001
1768
1769
[51] M. Petracca, L. Pela, R. Rossi, S. Oller, G. Camata, E. Spacone, Multiscale computational first order homogenization of thick shells for the analysis of out-of-plane loaded masonry walls. Computer Methods in Applied Mechanics and Engineering. 315 (2017), pp 273–301. ISSN: 0045-7825. DOI: [http://dx.doi.org/10.1016/j.cma.2016.10.046. http://dx.doi.org/10.1016/j.cma.2016.10.046.]
1770
1771
[52] A. Botto, P. Claps, D. Ganora, F. Laio, Regional-scale assessment of energy potential from hydrokinetic turbines used in irrigation channels. 4th International Conference on sustainability & Environmental Protection (SEEP2010) Conference Proceedings, June 29th–July 2nd, Bari, Italy, 2010.
1772
1773
[53] M. Anyi, B. Kirke, Review: Evaluation of small axial flow hydrokinetic turbines for remote communities. Energy for Sustainable Development. Volume 14, Issue 2, June 2010, Pages 110–116.
1774
1775
[54] K. A. R. Ismail, T. P. Batalha. A comparative study on river hydrokinetic turbines blade profile. Int. Journal of Engineering Research and Applications, Vol. 5, Issue 5, (Part 1) May 2015, pp.01-10.
1776
1777
[55] N. Kolekar, Z. Hu, A. Banerjee, X. Du, Hydrodynamic Design and Optimization of Hydro‐kinetic Turbines using a Robust Design Method. Proceedings of the 1st Marine Energy Technology Symposium METS13 April 10‐11, 2013, Washington, D.C.
1778

Return to Oller-Aramayo et al 2017a.

Back to Top

Document information

Published on 01/01/2017

Licence: CC BY-NC-SA license

Document Score

0

Views 80
Recommendations 0

Share this document

claim authorship

Are you one of the authors of this document?