You are on page 1of 11

MBS 7041 No.

of Pages 11, Model 5G


25 March 2010
ARTICLE IN PRESS

Mathematical Biosciences xxx (2010) xxx–xxx


1

Contents lists available at ScienceDirect

Mathematical Biosciences
journal homepage: www.elsevier.com/locate/mbs

2 Review

Theoretical modeling of fluid flow in cellular biological media: An overview

F
3

4 George E. Kapellos a,*, Terpsichori S. Alexiou a,b, Alkiviades C. Payatakes a,c,1

OO
5 a
Department of Chemical Engineering, University of Patras, GR-26504 Patras, Greece
6 b
Institute of Chemical Engineering and High Temperature Chemical Processes (ICEHT), GR-26504 Patras, Greece
7 Q2 c
Foundation of Research and Technology, Hellas (FORTH), GR-70013 Heraklion, Crete, Greece

PR
a r t i c l e i n f o a b s t r a c t
1 4
2 0
11 Article history: Fluid–structure interactions strongly affect, in multiple ways, the structure and function of cellular bio- 25
12 Received 30 August 2009 logical media, such as tissues, biofilms, and cell-entrapping gels. Mathematical models and computer 26
13 Received in revised form 10 March 2010 simulation are important tools in advancing our understanding of these interactions, interpreting exper- 27
14 Accepted 12 March 2010
imental observations, and designing novel processes and biomaterials. In this paper, we present a com- 28
15 Available online xxxx
prehensive survey and highlight promising directions of future research on theoretical modeling of 29

16
17
Keywords:
Fluid–structure interactions
D
momentum transport in cellular biological media with focus on the formulation of governing equations
and the calculation of material properties both theoretically and experimentally. With regard to the gov-
erning equations, significant work has been done with single-scale approaches (e.g. mixture theory),
30
31
32
18 Hierarchical multiscale modeling
TE
19 Material properties whereas traditional upscaling methods (e.g. homogenization, volume averaging) or multiscale equa- 33
20 Hydraulic permeability tion-free approaches have received limited attention. The underlying concepts, strengths, and limitations 34
21 Biofilm of each approach, as well as examples of use in the field of biomaterials are presented. The current status 35
22 Tissue of knowledge regarding the dependence of macroscopic material properties on the volume fractions, 36
23
geometry, and intrinsic material properties of the constituent phases (cells, extracellular matrix and 37
fluid) is also presented. The observation of conformational changes that occur at finer levels of the struc- 38
EC

tural hierarchy during momentum transport, the correlation of macro-properties with geometrical and 39
topological features of materials with heterogeneous and anisotropic microstructure, as well as the deter- 40
mination of dynamic material properties are among important challenges for future research. 41
Ó 2010 Published by Elsevier Inc. 42

43
RR

44
45 1. Introduction lar fluid and the deformable solid matrix (cells plus extracellular 61
matrix) plays multiple important roles: (a) affects the internal 62
46 Fluid–structure interactions in cellular biological media, such as architecture (spatial arrangement of cells and EPS) and the external 63
47 tissues, artificial cell-entrapping gels, microbial flocs, and biofilms, morphology (overall size and shape) of the cellular biological med- 64
are of great importance in quite diverse fields of science and tech- ium (e.g. [6]), (b) enhances the mass transfer rate of chemical spe-
CO

48 65
49 nology, including medicine, biology, biotechnology, tissue and cies (nutrients, wastes, chemical signaling molecules, etc.) within 66
50 environmental engineering. In particular, the theoretical analysis the cellular biological medium [1,7–10], and (c) regulates the func- 67
51 of momentum transfer during the flow of the extracellular fluid tion of cells through the action of mechanical stresses, which are 68
52 is necessary in many technological applications and natural pro- either applied directly on the outer surface of the biological cell, 69
53 cesses, such as the in vitro construction of artificial tissues from hu- or transmitted indirectly through the extracellular polymeric ma- 70
UN

54 man stem cells in order to replace damaged tissues [1], the trix [11–15]. 71
55 biodegradation of organic contaminants by microbial biofilms in Among the first contributions highlighting the role of fluid flow 72
56 porous media [2], the settling and consolidation of suspended in enhancing solute transport through convection in tissues is the 73
57 microbial flocs in solid–liquid separation processes [3], the forma- work of Swabb et al. [7], who showed that the relative importance 74
58 tion of microbial cakes on membrane filtration units [4], and the of convective versus diffusive mass transfer depends strongly on 75
59 delivery of chemotherapeutic agents to malignant tumors [5]. In the glycosaminoglycan (GAG) content of the tissue and the size 76
60 all these processes, the interaction between the flowing extracellu- of solute molecules. Larger solutes and lower GAG content result 77
in convection dominated mass transfer. In a similar vein, Piekarski 78

* Correspondence to: G.E. Kapellos, Department of Chemical Engineering, and Munro [16] suggested that the cyclic mechanical loading in 79
University of Patras, Karatheodori Str. 1, GR-26504 Patras, Achaia, Greece. Tel.: bone causes significant pulsatile flow in the canalicular network, 80
+30 2610 996 219; fax: +30 2610 990 328. which in turn results in enhanced mass transfer. Further informa- 81
E-mail addresses: gek222@chemeng.upatras.gr (G.E. Kapellos), xalexiou@chem
tion on the flow enhanced mass transfer in the interstitial space of 82
eng.upatras.gr (T.S. Alexiou).
1
Deceased.
tissues is given in [17]. 83

0025-5564/$ - see front matter Ó 2010 Published by Elsevier Inc.


doi:10.1016/j.mbs.2010.03.003

Please cite this article in press as: G.E. Kapellos et al., Theoretical modeling of fluid flow in cellular biological media: An overview, Math. Biosci. (2010),
doi:10.1016/j.mbs.2010.03.003
MBS 7041 No. of Pages 11, Model 5G
25 March 2010
ARTICLE IN PRESS

2 G.E. Kapellos et al. / Mathematical Biosciences xxx (2010) xxx–xxx

84 In addition, over the last 30 years, significant evidence has accu- scribe the mechanical behavior of cells and tissues, in [30] for com- 150
85 mulated to support the role of interstitial flow in stimulating tissue putational modeling of cell spreading and tissue growth, in [31] for 151
86 growth and regeneration. Perhaps, the first paradigm of this role is flow within the vascular network. As pointed out in the stimulating 152
87 the increase of bone mass under electromechanical stimulation, review by Humphrey [28] and also advocated here much work re- 153
88 where the coupling between electrical and mechanical forces is ef- mains to be done. 154
89 fected by streaming potentials developed during the flow of the io- The purpose of this paper is to present a comprehensive survey 155
90 nic aqueous solution in the canalicular network of bone [18,19]. of the literature pertinent to the theoretical analysis of momentum 156
91 Another interesting example is the work by Sikavitsas et al. [20] transfer in cellular biological media with focus on the formulation 157
92 who clearly showed that fluid shear affects strongly the osteogenic of governing equations and the calculation of material properties. 158
93 activity of osteoblasts growing in artificial scaffolds within perfu- Promising directions of future research are also highlighted. This 159
94 sion bioreactors. Specifically, by increasing the viscosity and keep- work was necessitated by the ever-growing interest on modeling 160
95 ing constant the flow rate of the feed solution, they achieved to fluid–structure interactions in cellular biological media over the 161

F
96 increase the fluid shear while maintaining practically constant last decade and the lack of a relevant review in the literature. 162
97 the mass transfer fluxes in the bioreactor. It was observed that

OO
98 the increased shear resulted in increased deposition and more uni-
99 form spatial distribution of mineralized matrix in the scaffold. Re- 2. Cellular biological media 163
100 cent advances on the flow-induced regulation of cellular functions
101 and mechanotransduction (i.e., translation of a physical force to The analysis of transport processes requires knowledge of the 164
102 the corresponding intracellular signal) within various tissues are detailed composition, structure, and even function of the cellular 165
103 reviewed in [12–15]. biological medium under consideration. A biofilm is a microbial 166

PR
104 With regard to microbial cellular biological media, the analysis consortium embedded in a three-dimensional meshwork of extra- 167
105 of interstitial flow has been of prime interest in filtration and sep- cellular polymeric substances (EPS), which is swollen in an aque- 168
106 aration processes. Notably, the extracellular flow affects strongly ous environment, and is usually immobilized on a solid surface 169
107 the settling velocity of suspended microbial flocs, as well as the fil- [32,33]. The EPS hydrogel is composed mainly of polysaccharides 170
108 tration efficiency of membranes covered by microbial cakes and (e.g. alginates, xanthan, cellulose) and proteins, which accumulate 171
109 biofilms [3,4]. On the other hand, the effects of extracellular flow D in the extracellular space, primarily by active secretion from the 172
110 and mechanical stresses on cell functions and biofilm growth have microbial cells and as residues from cell lysis [34]. A microbial floc 173
111 received limited attention. For example, the interesting work of is quite similar to a biofilm, but it differs in that it is found sus- 174
112 Fowler and Richardson [11], who showed that forced extracellular pended in an aquatic environment rather than attached on a solid 175
TE
113 flow increases the metabolic activity of Escherichia coli cells, has surface. Finally, a tissue is an organized population of eukaryotic 176
114 been ignored in the literature. cells all of which perform the same functions. The confined space 177
115 The analysis of the mechanics and fluid–structure interactions between tissue cells is known as the interstitial space (or interstit- 178
116 in cellular biological media requires a multidisciplinary approach ium) and contains fluid and an extracellular meshwork of proteins 179
117 integrating knowledge from mathematics, biology, biophysics, bio- (collagen, fibrin, elastin) and GAG (also referred to as EPS hydro- 180
EC

118 chemistry along with engineering principles. Our understanding of gel). Many tissues also contain a network of blood microvessels 181
119 the complex physical and biological mechanisms has grown by and a network of lymphatic microvessels. Here, we confine our 182
120 using more sophisticated and more accurate experimental tools attention to the extravascular space, which contains the cellular 183
121 (e.g. magnetic resonance imaging, computed micro-tomography, and interstitial spaces. 184
122 confocal laser microscopy), which allow one to study these pro- Despite the existence of significant differences in the detailed 185
RR

123 cesses with a narrow observation window in space-time and, thus, composition, structure and function of different types of cellular 186
124 track phenomena and behavior associated even with single cells or biological media, they all exhibit a hierarchical structure, which 187
125 single biomacromolecules. For example, Liu et al. [21] combined evolves in time and presents a high degree of spatial organization 188
126 atomic force with fluorescence microscopy to study the mechani- at every characteristic length-scale. There exist at least three fun- 189
127 cal properties of individual fibrin fibers. Interestingly, they found damental scales of observation with characteristic lengths defined 190
128 that fibrin fibers can be stretched to nearly three times their nor- by the diameter of the extracellular polymeric fibers (several nm, 191
CO

129 mal length without loosing elasticity and up to six times before P-scale), the size of a single biological cell (a few lm, K-scale) 192
130 rupturing. and the size of the overall cellular biological medium (from 193
131 Measurable progress has been achieved in the development of tenths of lm to a few cm, B-scale). For instance, Fig. 1 schemat- 194
132 mathematical models and computer-aided simulators, which are ically illustrates these three fundamental scales of the structural 195
133 valuable tools for the elucidation of the underlying mechanisms hierarchy of a microbial biofilm. Note that we use the term scale 196
UN

134 and the prediction of the process under consideration. The first to denote the spatial dimension of an object or process character- 197
135 models accounting for the coupling between fluid flow and matrix ized by both extent and resolution [35, p. 29]. The term extent de- 198
136 deformation in tissues appeared in the early 80s and were based on notes the overall size of the observable area or volume. The term 199
137 Biot’s poroelasticity [19] and mixture theory [22]. Over the last resolution denotes the size of the smallest discernible element of 200
138 three decades, these theories have been further developed and re- area (pixel) or volume (voxel). Both extent and resolution depend 201
139 fined to account for finite deformations, electrokinetics, and other on the instrument which is used to make the observation. One 202
140 features of the physical systems [23–25]. Recently, upscaling very important feature of the physical systems under consider- 203
141 methods, such as volume averaging and homogenization, were ation is the large separation of characteristic length-scales. Thus, 204
142 used to connect the macroscale description to microscale structure it is feasible to define the dimensions of a voxel, which is associ- 205
143 and function of cellular biological media (e.g. [26]). These works ated with a point on a coarse spatial scale, to be sufficiently large 206
144 were focused mainly to verify and define the validity domain of to include all the phases present at the finer spatial scale, and 207
145 the equations used in poroelasticity and mixture theories, but also much smaller than the extent of the coarse spatial scale. This fea- 208
146 to provide means for the theoretical calculation of constitutive ture allows one to implement standard upscaling methods (vol- 209
147 parameters. Comprehensive discussions on various aspects of ume averaging, homogenization, etc.) in order to derive a 210
148 modeling momentum transport in cellular biological media are macroscale description of the process from the corresponding 211
149 provided in [27–29] with regard to constitutive models used to de- governing equations at finer spatial scales. 212

Please cite this article in press as: G.E. Kapellos et al., Theoretical modeling of fluid flow in cellular biological media: An overview, Math. Biosci. (2010),
doi:10.1016/j.mbs.2010.03.003
MBS 7041 No. of Pages 11, Model 5G
25 March 2010
ARTICLE IN PRESS

G.E. Kapellos et al. / Mathematical Biosciences xxx (2010) xxx–xxx 3

F
OO
PR
D
TE
Fig. 1. Schematic representation of the hierarchical structure of a microbial biofilm.

213 3. Formulation of the conservation equations and constitutive into account the effects of anisotropy, viscoelasticity for the solid 242
EC

214 relations matrix, and inertia for the fluid (a summary of his work can be 243
found in [37]). Biot’s theory of poroelasticity has been used to 244
215 An integrated theoretical analysis of any such process requires predict the mechanics of bone [19,38], brain [39], solid tumors 245
216 dealing with two fundamental issues. The first issue regards the [40–42], and polymeric hydrogels [43,44]. In passing, we mention 246
217 formulation of the continuity and momentum conservation equa- that other phenomenological theoretical approaches for deform- 247
RR

218 tions as well as the necessary constitutive relations at the cellular able porous media have been developed for the process of cake 248
219 biological medium scale of observation (B-scale). The second issue filtration (e.g. [45,46]). 249
220 regards the proper connection of the parameters, which appear in The theory of mixtures was developed for the mathematical 250
221 the constitutive equations, with the geometrical parameters and description of momentum, mass, and energy transport in multi- 251
222 the physicochemical parameters that pertain to the phenomena component, multiphase systems at a scale of observation with res- 252
223 taking place at the molecular- and cell-scales of observation. With olution much larger than the characteristic length of the individual 253
CO

224 regard to the formulation of the momentum conservation equa- constituents and phases (the structure of the system at finer spatial 254
225 tions and the necessary constitutive relations at the cellular biolog- scales is disregarded completely). The mathematical foundation of 255
226 ical medium scale (B-scale), we distinguish three fundamentally the theory was originally set by Truesdell [47], and relies upon the 256
227 different routes of theoretical analysis, which are presented below. concept that the constituents of the mixture can be modeled as 257
superimposed, interacting continua. In this way, a material point 258
UN

is assigned for each constituent at every point in space, which is 259


228 3.1. Single-scale approaches occupied by the mixture. The governing conservation laws for each 260
constituent of the mixture contain the usual terms (that appear in 261
229 One commonly used approach is to disregard the finer spatial formulations for a single-phase), and an additional term which ac- 262
230 scales of the physical system and formulate these equations di- counts for the interaction between the reference constituent and 263
231 rectly on the B-scale. In this direction, the theoretical analysis of the other constituents of the mixture. Exhaustive reviews on the 264
232 momentum transfer in cellular biological media and biological theory of mixtures are given in [48,49]. Mixture theory was intro- 265
233 hydrogels has been based mainly on Biot’s theory of poroelasticity, duced in the field of biomaterials by Mow et al. [22], in their sem- 266
234 as well as on the theory of mixtures combined with a thermody- inal work on the mechanical behavior of cartilage. They showed 267
235 namic approach. Motivated by the problem of soil subsidence, Biot that a biphasic mixture theory (linearly elastic solid, Newtonian 268
236 [36] formulated a phenomenological theory on the consolidation of fluid) captures very well the experimentally observed behavior of 269
237 deformable porous media saturated with a fluid. In that pioneering articular cartilage under confined compression. In order to achieve 270
238 work, he considered the porous medium as homogeneous and iso- this agreement between theory and experiment it was necessary to 271
239 tropic at the macroscale, and treated the solid matrix as a linearly express the permeability of the cartilage as a function of the solid 272
240 elastic solid and the fluid, which occupies the pore space, as incom- dilatation. Another key finding of their work is that tissue-scale 273
241 pressible and Newtonian. Later, he extended his theory by taking viscoelastic behavior (creep, stress relaxation) can be observed as 274

Please cite this article in press as: G.E. Kapellos et al., Theoretical modeling of fluid flow in cellular biological media: An overview, Math. Biosci. (2010),
doi:10.1016/j.mbs.2010.03.003
MBS 7041 No. of Pages 11, Model 5G
25 March 2010
ARTICLE IN PRESS

4 G.E. Kapellos et al. / Mathematical Biosciences xxx (2010) xxx–xxx

275 a result of the relative motion between the two constituents, even The formulation of the conservation laws and constitutive equa- 334
276 if none of the constituents is viscoelastic. Thereafter, the theory has tions directly on the B-scale is of limited value because of the ab- 335
277 been extended and used widely to predict the mechanics of carti- sence of cross-scale consistency and, in many cases, because of 336
278 lage [23–25], endothelial glycocalyx [50], solid tumors [51,52], the ambiguous correspondence between theoretical dependent 337
279 hydrogels [53], soft tissues in general [54–59], and filter cakes variables and experimentally measured quantities. In addition, sin- 338
280 [60]. In the simplest case, a cellular biological medium can be con- gle-scale approaches do not provide a consistent theoretical frame- 339
281 sidered as a biphasic mixture which consists of a linearly elastic so- work for the calculation of the constitutive parameters and, 340
282 lid constituent (cells plus EPS meshwork) and a Newtonian fluid usually, leave this task for the experimenter. 341
283 constituent (extracellular aqueous solution). In the framework of
284 mixture theory, significant effort has been made to incorporate 3.2. Multiscale bottom-up approaches 342
285 electrokinetic phenomena (for example, the triphasic mixture the-
286 ory of Lai et al. [23]), as well as non-linear phenomena, including An approach that is more versatile and robust than the single- 343

F
287 the dependence of the permeability on solid dilatation [25], finite scale ones is to develop the governing equations on the B-scale 344
288 solid deformation [25,57], and viscoelastic behaviour [53,56]. A starting from ‘‘well-established” first principles on the K- or even 345

OO
289 comprehensive presentation of the equations that describe fluid the P-scale. For this purpose, one can implement mathematical 346
290 flow and matrix deformation in a cellular biological medium, in procedures, including the spatial homogenization method (e.g. 347
291 the context of Biot’s theory of poroelasticity and the theory of mix- [26,71]) and volume averaging methods (e.g. [72–77]), which were 348
292 tures, is given in Appendix A (this also facilitates the comparison initially applied to the analysis of transport phenomena and 349
293 with the equations established using the spatial averaging method mechanics of porous or composite materials, in the context of con- 350
294 in a companion paper [61]). tinuum mechanics. 351

PR
295 In many cases, it is possible to decouple extracellular fluid flow These upscaling methods offer two significant advantages over 352
296 and solid matrix deformation and, thus, model the two processes in single-scale approaches. First, one can obtain precise correspon- 353
297 a sequential fashion. This decoupling hypothesis is applicable, for dence between theoretical variables defined at different spatial 354
298 example, if the solid matrix behaves as an elastic solid (that is, re- scales, as well as between theoretical variables and the respective 355
299 sponds instantaneously to the applied stress), or if the flow-in- experimentally measured quantities (via the use of appropriate 356
300 duced deformation of the solid matrix is negligible. Along this D weight functions in the averaging procedure). Second, a framework 357
301 direction, the theory of Newtonian fluid flow through rigid porous is provided for the linkage of constitutive parameters defined at 358
302 media has been adopted for the description of flow through sus- the coarse spatial scale with the structure of and the phenomena 359
303 pended microbial aggregates (e.g. [62–64]), biofilms [65], and the occurring at finer spatial scales. Thereby, a direct comparison be- 360
TE
304 extravascular space of tissues (e.g. [66–68]). tween theory and experiment can be accomplished without the 361
305 Finally, many single-scale models incorporate (usually, use of adjustable parameters. In any case, the derived upscaled 362
306 ‘‘silently”) the perfect coupling hypothesis, according to which all momentum balance equations (inclusive of the constitutive rela- 363
307 the constituent phases of the system move with the same average tions) should meet two prerequisites. First, these must contain 364
308 velocity and, thus, a monophasic description is employed. Under this only measurable average variables (density, pressure, velocity, dis- 365
EC

309 hypothesis, momentum transfer in a cellular biological medium, at placement) and a minimum number of physically meaningful con- 366
310 the B-scale of observation, follows the formulation of conservation stitutive parameters (elastic moduli, hydraulic permeability, etc.). 367
311 laws and constitutive equations for single-phase materials. In the Second, the velocity (or displacement) field in the cellular biologi- 368
312 tissue literature, it is common to adopt the theory of elasticity or vis- cal medium, which is obtained from the solution of the upscaled 369
313 coelasticity for the description of tissue mechanics [28,29]. Analo- momentum equations, should be identical to the field that would 370
RR

314 gous examples from the biofilm literature include the work by be obtained if a complete description at the molecular scale was 371
315 Dupin et al. [69], where the biofilm is modelled as a highly viscous feasible. This prerequisite is usually verified through comparison 372
316 Newtonian fluid, and the work by Towler et al. [70], where the bio- with data from well-controlled experiments. 373
317 film is modelled as a Burgers viscoelastic fluid. Limited work has been made up to now in the analysis of trans- 374
318 Using order-of-magnitude analysis in the context of the spatial port processes in cellular biological media using multiscale ap- 375
319 averaging method, we have obtained the following conditions proaches. To our knowledge, the first decent efforts in this 376
CO

320 which delimit the domain of validity of the biphasic and monopha- direction were made in the field of bone mechanics by Crolet 377
321 sic (elastic or viscoelastic) models for the macroscopic description et al. [78] and Hollister et al. [79], both of which applied the spatial 378
322 of momentum transport in a fluid–solid system (linearly elastic so- homogenization method to derive macroscopic equations and cor- 379
323 lid, Newtonian fluid) [61]. relate the apparent elastic moduli of the bone with the microstruc- 380
uf lf ture. The collagen and hydroxyapatite, which compose the bone, 381
IF ¼ O½1 THEN monophasic viscoelastic
UN

u s ls ss were assumed to behave as linearly elastic solids. Reasonable 382


! agreement was observed between theory and experiment, in terms 383
kp s2s;c uf lf of the apparent elastic moduli. However, the importance of inter- 384
IF þ 1 << << 1 THEN monophasic elastic
uf ‘2p s2s u s ls ss stitial fluid flow was not fully appreciated at that time and, thus, 385
!
u f lf keff s 2 flow effects were not included in the models. Significant work 386
s;c
IF < þ1 THEN biphasic model has been made by Wood and Whitaker [80–82] who used a volume 387
us ð2ls þ ks Þss uf ‘2p s 2
s
325 averaging method to derive the macroscopic equations that de- 388
qffiffiffiffiffiffiffiffiffiffiffiffiffi
2
ffi scribe solute diffusion and reaction, cell growth, and EPS produc- 389
326 with: ss;c ¼ ð2lqssþk
‘p

. tion in biofilms at the B-scale of observation, starting with a 390
327 Here, uf is the volume fraction of fluid, us is the volume fraction continuum-based formulation at the cell-scale. They developed a 391
328 of solid, lf is the dynamic viscosity of the fluid, ls ; ks are the Lamé closure scheme for the calculation of the diffusion coefficient ten- 392
329 parameters for the solid, qs is the density of the solid, kp is the sor, and observed good agreement between theoretical predictions 393
330 hydraulic permeability of the fluid–solid system, ss is the charac- and experimentally determined values. However, an empirical ap- 394
331 teristic time for changes in the displacement of the solid (e.g. per- proach was introduced for the rate of accumulation of cellular 395
332 iod of harmonic oscillation), and ‘p is the characteristic length of mass that is caused by convection in order to circumvent the com- 396
333 the system on the macroscopic scale. plex problem of momentum transport. The spatial homogenization 397

Please cite this article in press as: G.E. Kapellos et al., Theoretical modeling of fluid flow in cellular biological media: An overview, Math. Biosci. (2010),
doi:10.1016/j.mbs.2010.03.003
MBS 7041 No. of Pages 11, Model 5G
25 March 2010
ARTICLE IN PRESS

G.E. Kapellos et al. / Mathematical Biosciences xxx (2010) xxx–xxx 5

398 method was used by Moyne and Murad [26] to derive the steady- expected to become useful tools in the study of systems with 462
399 state equations that govern flow and deformation in cartilage, at non-linear material behavior. 463
400 the macroscale. The aqueous solution that occupies the pore space
401 of the cartilage was considered as Newtonian, and the fiber net-
4. Calculation of the constitutive parameters 464
402 work as a linearly elastic solid. It should be noted that the presence
403 of ions in the aqueous solution and surface charge on the fibers
The calculation of the parameters that appear in the constitutive 465
404 were taken into account. The upscaled equations obtained via the
equations, which describe the behavior of the material at the B-scale 466
405 homogenization method are similar in form with those formulated
of observation, can be performed either experimentally or theoreti- 467
406 in the triphasic mixture theory of Lai et al. [23]. Nonetheless, Moy-
cally. For the studied process, constitutive parameters of interest ap- 468
407 ne and Murad did not proceed with the theoretical calculation of
pear in the expressions for the stress tensors (viscosity of the 469
408 the constitutive parameters which appear in their model. Recently,
extracellular fluid, elastic moduli of the solid matrix), and the inter- 470
409 we used the spatial averaging method via a weight function to

F
action force between the extracellular fluid and the EPS-plus-cells 471
410 establish the equation which describes solute conservation at the
matrix (hydraulic permeability). These material properties depend 472
411 B-scale, starting with a continuum-based formulation of solute

OO
on the phenomena and the structure observed at finer spatial scales. 473
412 transport at finer spatial scales [83]. An effective-medium model
For the problem at hand, the uncovering of the correlation between 474
413 was developed for the calculation of the diffusion coefficient, in
microstructure and macro-properties is a challenging task, both in 475
414 consistence with the averaging method. The model predicted the
theory and experiment, because conformational changes occur at 476
415 qualitative trend as well as the quantitative variability of a large
every level of the hierarchical structure during momentum 477
416 number of published experimental data on the diffusion coefficient
transport. 478
417 of oxygen in cell-entrapping gels, microbial flocs, biofilms, and

PR
418 mammalian tissues. In a companion paper, we use the spatial aver-
419 aging method to establish the equations governing momentum 4.1. Experimental determination of constitutive parameters 479
420 transfer in cellular biological media, and develop theoretical mod-
421 els for the calculation of the hydraulic permeability [61]. The experimental determination of constitutive parameters can 480
be achieved by fitting the outcome of a theoretical model to the 481
422 3.3. Multiscale computational, constitutive-equation-free approaches corresponding measurements from well-controlled experiments. 482

423 In many cases of practical interest the behavior of a cellular bio-


D The suitable experimental method usually differs for different
types of cellular biological media. For instance, the hydraulic per-
483
484
424 logical medium cannot be captured satisfactorily over a wide range meability has been estimated with infusion [87–89], compression 485
TE
425 of working conditions using well-established, simple constitutive [90,91], or perfusion experiments [88,92,93] for tissues, with set- 486
426 equations. Single-scale methods usually rely on heuristic ap- tling experiments for suspended microbial flocs [9,94], and with 487
427 proaches, which stem from the modeller’s experience along with filtration techniques for microbial cakes and biofilms formed on 488
428 trial-and-error, so as to extend the validity range of existing consti- the surface of porous membranes [4,95–100]. Note that in the tis- 489
429 tutive models. Further, the implementation of standard upscaling sue literature, the term infusion refers to localized delivery of fluid 490
EC

430 methods, such as homogenization and volume averaging, becomes to a tissue from a point source (i.e., the tip of a needle or a catheter), 491
431 a formidable task if there exist material components which exhibit while the term perfusion refers to the delivery of fluid from a sur- 492
432 non-linear mechanical behaviour at the finer spatial scale or, if face source with dimensions comparable to the tissue dimensions. 493
433 there does not exist a large separation of characteristic length- An issue of crucial importance is to ensure that the determined 494
434 scales (which is necessary in order to perform significant simplifi- constitutive parameters are independent of the specific procedure, 495
RR

435 cations in the calculus). Over the last decade, a new line of analysis measuring device, and specimen size used in the experiments. This 496
436 has been developed for the study of complex systems with hierar- issue is frequently overlooked. The influence of these factors 497
437 chical structure: the so-called equation-free approach [84]. The should be minimized by choosing an appropriate experimental 498
438 main idea of the equation-free approach is to circumvent the for- method and, further, taken into account explicitly in the formula- 499
439 mulation of macroscopic constitutive equations and calculate the tion of the theoretical model for the process. The influence of the 500
440 quantities of interest at the macroscale by performing appropri- measuring process can be represented mathematically using 501
CO

441 ately initialized computational simulations at the microscale. A weight functions, which allow one to obtain a precise correspon- 502
442 representative elementary volume (REV), which contains a geo- dence between theoretical dependent variables and the respective 503
443 metric representation of the structure at the microscale, is assigned experimentally measured quantities, as well as to relate experi- 504
444 to every macroscale point or, in practice, to every element of the mental measurements obtained with different techniques [101]. 505
445 macroscale computational mesh. Regarding the conceptual frame- Another robust approach to account for the effect of the measuring 506
UN

446 work of the equation-free approach, the establishment of direct process and sample size is the computer-aided simulation of the 507
447 handshaking for bidirectional information transfer between two experimental test (virtual experiment). Some noteworthy exam- 508
448 scales is of crucial importance. There is no golden rule for the suc- ples in this direction include the simulated compression test on a 509
449 cessful cross-scale linkage. Local information from the macroscale tissue specimen by van Rietbergen et al. [102] for the calculation 510
450 must be properly projected to the microscale and used to define of the elastic moduli, the simulated perfusion experiment by John- 511
451 appropriate boundary and initial conditions (downscaling). Then, son and Deen [103] for the estimation of the hydraulic permeabil- 512
452 computational simulation of the process is performed, in the con- ity of hydrogels, and the simulated infusion experiment by 513
453 text of the REV, for a given period of time and the distributions of Milosevic et al. [89] for the estimation of the hydraulic permeabil- 514
454 microscale quantities are properly averaged to obtain the values of ity of tissues in vivo. 515
455 the corresponding macroscale quantities at that point or element The endmost goal is to obtain a correlation between the constitu- 516
456 (upscaling). The upscaling step is achieved using theorems estab- tive parameter of interest and other measurable properties of the 517
457 lished in the context of the homogenization or volume averaging system. In principle, this correlation should involve only features 518
458 methods. In the field of biomechanics, equation-free approaches of the structure and composition at finer spatial scales. However, 519
459 have been developed to predict the mechanical response of tissue in practice it is very difficult (if feasible) to follow-up the conforma- 520
460 constructs [85] and collagen networks [86] under compression, tional changes that occur at every level of the hierarchical structure 521
461 omitting the flow of the extracellular fluid. Such approaches are during the experimental test and, thus, it is customary to correlate 522

Please cite this article in press as: G.E. Kapellos et al., Theoretical modeling of fluid flow in cellular biological media: An overview, Math. Biosci. (2010),
doi:10.1016/j.mbs.2010.03.003
MBS 7041 No. of Pages 11, Model 5G
25 March 2010
ARTICLE IN PRESS

6 G.E. Kapellos et al. / Mathematical Biosciences xxx (2010) xxx–xxx

523 the constitutive parameters with other macroscopic quantities, such crease of the elastic and shear moduli [78,106] and in a decrease of 587
524 as the apparent density of the cellular biological medium [104], the the hydraulic permeability [105,107]. 588
525 dilatation/strain of the solid matrix [22–24,93], and the pressure Equally important is the effect of the intrinsic material proper- 589
526 gradient of the extracellular fluid [4]. Relatively few experimental ties of the constituent phases which, however, are usually difficult 590
527 studies have presented data regarding the quantitative correlation to determine from independent experiments. An issue that should 591
528 between the mechanical properties (elastic moduli and hydraulic be examined carefully concerns the effect of the spatial variability 592
529 permeability), the composition, and the microstructure of a cellular in the material properties of the constituent phases. Using a com- 593
530 biological medium. Most of those studies have demonstrated that putational homogenization approach Sengers et al. [112] found 594
531 the volume fractions of the constituent phases (cells, EPS, water) that the hydraulic permeability and the elastic modulus of an arti- 595
532 strongly affect the mechanical properties [90,91,98,100,105]. More- ficial tissue construct are to a large extent insensitive to the EPS 596
533 over, comprehensive studies on fluid flow through beds of biological distribution at the cell-scale of observation. Only in the case of very 597
534 cells have shown that the hydraulic permeability depends markedly localized pericellular deposition did the elastic modulus take sig- 598

F
535 on the size, shape, and orientation of cells [95,98–100], as well as on nificantly lower values than those in the case of homogeneous dis- 599
536 the gelatinous sheath that covers the cells [97]. tribution of the same amount of EPS in the extracellular space. The 600

OO
hydraulic permeability was not affected even though appreciably 601
537 4.2. Theoretical calculation of constitutive parameters different flow patterns were obtained for each case (Fig. 4 in 602
[112]). Recently, we developed a theoretical model for diffusive 603
538 The theoretical calculation of a constitutive parameter requires mass transfer in cellular biological media [83] and showed that 604
539 the development of: (1) a structural model, which provides the the diffusion coefficient obtains substantially different values for 605
540 internal geometry and topology of the cellular biological medium the two extreme cases of homogeneous distribution and pericellu- 606

PR
541 with sufficiently high spatial resolution, (2) a process model, which lar deposition of the EPS. The difference between the two bounding 607
542 mathematically describes the momentum transport in the context values depends on the volume fraction of extracellular space, the 608
543 of the structural model, and (3) a closure scheme, which links math- porosity of the EPS hydrogel, and the size of diffusing species and 609
544 ematically the value of the constitutive parameter with the struc- EPS fibers. In a companion paper [61], we elaborate further on this 610
545 tural and process models used for the finer spatial scale. issue with respect to the hydraulic permeability. 611
546 With regard to the structural model we distinguish three differ- D The characteristic dimensions, shape and spatial arrangement 612
547 ent approaches. The simplest approach is to consider the cellular of the component material elements (cells, EPS fibers, and pores) 613
548 biological medium as an ensemble of (geometric) unit cells also affect the mechanical properties of a cellular biological med- 614
549 [78,79,83,106–111]. Each unit cell represents the structure in the ium significantly. For instance, Beno et al. [109] presented a theo- 615
TE
550 vicinity of a few biological cells. A more realistic approach is to rep- retical model for the hydraulic permeability of cortical bone, and 616
551 resent the cellular biological medium as a regular or random array concluded that the canalicular and osteocyte process dimensions, 617
552 of biological cells [112]. Finally, the most promising approach is to the geometry of the GAG matrix (interfiber spacing and fiber ra- 618
553 use digitized images of real cellular biological media, which can be dius) that partially occupies the canaliculi, as well as the lacuna–la- 619
554 obtained using three-dimensional reconstruction from two-dimen- cuna distance have a strong effect on the permeability. Also, they 620
EC

555 sional serial sections, confocal laser scanning microscopy (CLSM), suggested that the osteocyte lacuna shape and size, along with 621
556 X-ray computed micro-tomography (micro-CT), or any other suit- the three-dimensional distribution of canaliculi, determine the de- 622
557 able visualization technique. Up to date, this approach has been gree of anisotropy of the permeability. 623
558 used for the prediction of bone mechanical properties [102,113] The architecture of the EPS matrix is a key determinant of the 624
559 but it is expected to become very popular and useful for all types permeability of the cellular biological medium. The problem of 625
RR

560 of cellular biological media in the near future. flow through fibrous materials is well-studied and there exist 626
561 With regard to the process model we distinguish two different several models that can be used to obtain estimates for the 627
562 approaches. In the first approach, a constituent phase is treated as intrinsic permeability of the EPS hydrogel as a function of fiber 628
563 a continuum at the finer spatial scale and momentum transport volume fraction and arrangement (see Table 1; for further read- 629
564 (fluid flow and solid matrix deformation) is described with a set ing see the reviews [118–120]). A simple and useful correlation 630
565 of partial differential equations, which can be solved analytically has been proposed by Jackson and James [118]. Specifically, they 631
CO

566 or numerically depending on the complexity of the adopted struc- considered that for randomly oriented fibers the resistance ex- 632
567 tural model. In the second approach, a constituent phase is treated erted by the solid matrix on fluid flow can be estimated by 633
568 as a population of interacting agents (fluid and solid particles, adding the resistance resulting from fibers oriented with their 634
569 mass-and-spring elements, molecules, etc.) and momentum trans- axis normal to the macroscopic flow direction to the resistance 635
570 port is described using computer-aided simulations, such as spring resulting from fibers oriented with their axis parallel to the mac- 636
UN

571 network analysis, molecular dynamics and other, in the context of roscopic flow direction. The possibility that a fiber axis is ori- 637
572 statistical mechanics. With regard to the closure scheme, several ented normal to the flow is twice that of being parallel to the 638
573 different approaches have been developed in the literature, flow. Therefore, the total resistance equals 2/3 of the resistance 639
574 including heuristic REV analysis [106–109], effective-medium resulting from normal orientation plus 1/3 of the resistance 640
575 approaches [83,110,114], asymptotic homogenization [78,79,111– resulting from parallel orientation 641
576 113], and volume averaging [83,115].
1 2 1
577 Considerable work has been done for the prediction of the elas- ¼ þ
kp 3kp? 3kp== 643
578 tic properties of tissues, especially of bone and cartilage (e.g.
579 [78,79,102,106,110,111,113,114,116]). On the other hand, limited where kp? is the permeability for fibers oriented normal to the mac- 644
580 work has been done for the hydraulic permeability [107– roscopic flow direction and kp== is the permeability for fibers ori- 645
581 109,112,117]. In general, the mechanical properties (elastic moduli ented parallel to the macroscopic flow direction. Jackson and 646
582 and hydraulic permeability) of a cellular biological medium de- James used the results of [121] for the calculation of kp? and kp== . 647
583 pend on the volume fractions, geometry, topology, and intrinsic The final expression is 648
584 material properties of the constituent phases at finer spatial scales.
kp 3
585 In particular, the effect of the volume fractions is prominent. An in- ¼ ½ ln us  0:931
586 crease in the volume fraction of solids (EPS, cells) results in an in- r2fiber 20us 650

Please cite this article in press as: G.E. Kapellos et al., Theoretical modeling of fluid flow in cellular biological media: An overview, Math. Biosci. (2010),
doi:10.1016/j.mbs.2010.03.003
MBS 7041 No. of Pages 11, Model 5G
25 March 2010
ARTICLE IN PRESS

G.E. Kapellos et al. / Mathematical Biosciences xxx (2010) xxx–xxx 7

Table 1
Theoretical models for the permeability of fibrous porous materials.

Source Equation Comments


Jackson and kp
¼ 3
½Inur  0:931 Combination of analytical solutions
r 2p 20ur
James [118]
Davies [124] kp
¼ 1 Fit to a large number of data for the flow of air through fibrous materials
r 2p 3=2
ð1þ56u3r Þ
16ur

Johnston [125] h i2 Fit to experimental data for flow through disordered fibrous materials
kp
r 2p
¼ 0:01064 ð1uruÞð2ur Þ
r
Koponen kp 5:55
¼ expð10:1 Fit to data from the simulation of flow in disordered fibrous structures with the lattice-
r 2p u r Þ1
et at. [126] Boltzmann method
Clague kp
1 pffiffiffiffiffiffiffiffiffiffiffiffiffi 2 Fit to data from flow simulations in periodic arrays with the LB method
r 2p
¼ c1 2 p=ur  1 expðc2 ur Þ
et al. [123] c1 ¼ 0:50941; c2 ¼ 1:8042 (square) c1 ¼ 0:71407; c2 ¼ 0:51854 (random)
h i
Happel [127] u2r Cylinder in cell model. Analytical solution (\ denotes flow normal to the fiber axis)

F
kp?
r 2p
¼ 8u1  ln ur  1
1þu2r
r
h i
Happel [127] kp==
¼ 4u1  ln ur  32 þ 2ur  2r
u2 Cylinder in cell model. Analytical solution (// denotes flow parallel to the fiber axis)
r2p r
 

OO
Kuwabara [128] kp?
¼ 8u1  ln ur  32 þ 2ur Cylinder in cell model. Analytical solution
r 2p r
 
Sangani and kp?
¼ 8u1  ln ur  1:476 þ 2ur  1:774u2r þ 4:076u3r Periodic square array
r2
p r
Acrivos [129]
 
Sangani and kp?
¼ 8u1  ln ur  1:490 þ 2ur  1:5u2r Periodic hexagonal array
r 2p r
Acrivos [129]
h i
Drummond and kp== u2
¼ 4u1  ln ur þ K DT þ 2ur  2r KDT = 1.476 (square array)
r2p r
Tahir [121]

PR
KDT = 1.498 (trigonal array)
KDT = 1.354 (hexagonal array)
 
Drummond and kp?
¼ 8u1 lnur  1:476 þ 2ur  1:774u2r þ 4:078u3r Periodic square array
r 2p r
Tahir [121]
pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
Gebart [130] kp?
¼ ð16=9pc1 Þ½ ur: max =ur  15=2 Analytical solution for periodic arrays based on lubrication approximation
r 2p pffiffiffi pffiffiffi pffiffiffiffiffiffi
ðc1 ; ur;max Þ ¼ ð 2; p=4Þ (square) ðc1 ; ur;max Þ ¼ ð 6; p= 12Þ (hexagonal)
Gebart [130] kp==
¼ c82 ð1uu2r Þ
3
c2 ¼ 57 square arrangement
r2p r D
c2 ¼ 53 hexagonal arrangement
TE
651 where r fiber is the EPS fiber radius, and us is the fiber volume frac- macroscopic material properties and heterogeneous, anisotropic 688
652 tion. The above equation provides satisfactory agreement with a microstructures characterized by wide distributions of geometrical 689
653 large number of experimental data for flow through fibrous materi- (size and shape), topological (e.g. connectivity, orientation), and 690
654 als (including hydrogels of acrylamide, hyaluronic acid, collagen other properties (e.g. electrical charge) at the cell- or EPS-scale. 691
655 [118]). Further, a comparison the data from numerical simulations
EC

656 of [122,123], as well as the experimental data of [103] showed that


657 the model of Jackson and James provides reliable estimates of the 5. Perspectives for future work 692
658 permeability in the range 0:10 6 us 6 0:20, whereas underesti-
659 mates the permeability for very small fiber volume fractions. From the literature survey becomes obvious that significant 693
660 The mechanical behavior at the surfaces of separation between work has been done, but much remains to be done in the future. 694
RR

661 different contiguous phases is another important factor. Schwartz Multiscale approaches, which take into account the process and 695
662 et al. [110] developed a theoretical model for the elastic moduli structure at finer spatial scales of the hierarchal structure of cellu- 696
663 of articular cartilage and showed that the shear modulus depends lar biological media, create a premise for improved modeling of 697
664 strongly on the extent of slipping between collagen fibers and the momentum transfer and fluid–structure interactions at the macro- 698
665 surrounding proteoglycan matrix (which might be caused by scopic B-scale (which is the scale of practical interest). Some prom- 699
666 imperfect bonding between the two materials). ising directions for future research are given below: 700
CO

667 Finally, the presence of charged species at material interfaces


668 and within bulk phases has also a noticeable effect. Eisenberg  Derivation of the governing equations using traditional upscal- 701
669 and coworkers [107,108] developed a theoretical model for the cal- ing methods, such as spatial averaging and homogenization. 702
670 culation of the hydraulic permeability of the extracellular matrix The coupling of hydrodynamics with electrokinetics, osmotic 703
671 (EPS hydrogel) taking into account the electrohydrodynamic cou- and other phenomena should be addressed. 704
pling caused by the interaction between mobile charged species  Incorporation of advanced non-equilibrium thermodynamic
UN

672 705
673 dissolved in the flowing aqueous solution and immobile charged frameworks, which account to a certain extent for the micro- 706
674 species located at the surface of the EPS fibers. They showed that structure (e.g. via a conformation tensor), in single-scale 707
675 the hydraulic permeability depends on the surface charge density, approaches (‘‘microstructured mixture theory”). 708
676 as well as on the salt concentration in the bulk aqueous solution.  Development of multiscale equation-free approaches for sys- 709
677 The degree of dependence is substantial for the open-circuit per- tems with non-linear mechanical behavior at the microscale. 710
678 meability (zero net electric current) and minor for the short-circuit Usually, such approaches should be tailored to a specific cellular 711
679 permeability (zero gradient of electrical potential) in consistence biological medium. Among other issues, the inclusion of hydro- 712
680 with experimental observations. Furthermore, Mattern and Deen dynamics, and comparison of different schemes for cross-scale 713
681 [117] showed that the distribution of surface charge density has handshaking should be addressed. 714
682 a marked effect on the open-circuit permeability.  Comparison between different theoretical approaches in a fair 715
683 Admittedly, theoretical modeling has shed significant light on context, that is for a given system under a common set of 716
684 the dependence of macroscopic constitutive parameters on micro- assumptions and boundary conditions. For instance, under what 717
685 structural features. However, some important issues have not been conditions are the equations obtained via homogenization 718
686 addressed yet. For instance, it deserves to use more realistic and equivalent to a mixture theory formulation, for a cellular bio- 719
687 complex structural models to examine the correlation between logical medium with macroscopic heterogeneities? 720

Please cite this article in press as: G.E. Kapellos et al., Theoretical modeling of fluid flow in cellular biological media: An overview, Math. Biosci. (2010),
doi:10.1016/j.mbs.2010.03.003
MBS 7041 No. of Pages 11, Model 5G
25 March 2010
ARTICLE IN PRESS

8 G.E. Kapellos et al. / Mathematical Biosciences xxx (2010) xxx–xxx

721  Conduction of well-controlled experiments that will provide the where reff is the ‘‘effective” stress tensor for the entire material 786
722 material properties (if possible, all the components of the ten- (fluid plus solid), u is the displacement of the material, P is the fluid 787
723 sors) along with information about the composition and pressure, e ¼ r  u is the material dilatation, v is the fluid velocity, 788
724 microstructure, under quasi-steady-state conditions. The deter- leff and keff are the Lamé parameters, lf is the fluid viscosity, keff 789
725 mination of the material properties alone is of little or no use. is the permeability of the material, h is the variation of the fluid 790
726  Determination of dynamic material properties under transient volume fraction (‘‘variation in water content”), and a; M are phenom- 791
727 conditions both experimentally and theoretically. These are enological coefficients (also known as Biot or Biot–Willis coeffi- 792
728 expected to be substantially different from quasi-steady state cients). Substituting the constitutive Eq. (A2) in the momentum 793
729 values. This task will require the use or development of appro- balance (A1), and combining with the Eqs. A3, A4 and A5 we obtain 794
730 priate imaging techniques and computer simulation methods  
731 for the determination of the conformational changes that occur leff r2 u þ leff þ keff re  arP ¼ 0 ðA6Þ
732 at finer levels of the structural hierarchy during momentum @e 1 @P keff 2

F
733 transport. The definition of proper measures for the quantifica- a þ  r P¼0 ðA7Þ
@t M @t lf 796
734 tion of geometrical and topological features of the structure of

OO
735 cellular biological media at finer characteristic spatial scales In [131] a worthy effort is made to attribute a clear physical 797
736 (cell-, EPS-) should also be addressed. meaning to the coefficients a and M, and methods for their exper- 798
737  Consideration of the influence of specimen geometry (size, imental determination are proposed. For example, the coefficient a 799
738 shape) and experimental measuring device via the use of expresses the ratio of the variation of pore volume to the dilatation 800
739 weight functions and/or virtual experiments. of the solid, under constant total volume conditions. Usually, the 801
740  Use of more realistic and complex structural models to examine fluid and the solid matrix are considered to be intrinsically incom- 802

PR
741 theoretically the dependence of the macroscopic material prop- pressible and, thus, the values a ¼ 1 and 1=M ¼ 0 are used [39–42]. 803
742 erties and their degree of anisotropy on the statistical and spa- In this case we have 804
743 tial distributions and correlations of geometrical (e.g. cell size 805
 
744 and shape), topological (e.g. interstitial connectivity, cell orien- leff r2 u þ leff þ keff re  rP ¼ 0 ðA8Þ
745 tation), and other properties (e.g. EPS permeability, electrical @e keff 2
 r P¼0 ðA9Þ
746 charge) of the structure at the cell- or EPS-scale. @t lf 807
747
D By taking the divergence of (A8) we obtain [39] 808
748 In a companion paper we present work in some of the directions
749 mentioned above. In closing, it should be mentioned that although  
2leff þ keff r2 e ¼ r2 P ðA10Þ
TE
810
750 transport phenomena in cellular biological media can be treated in
751 a unified manner, specific solutions should be tailored to specific Therefore we can equivalently express Eq. (A9) as follows: 811
752 applications with different types of cellular biological media. The 812
753 processes of cell growth and matrix production were beyond the @e keff  
¼ 2leff þ keff r2 e ðA11Þ
754 scope of this work and definitely deserve to be treated in detail @t lf 814
EC

755 separately.
It is important to note that the equations formulated by Biot di- 815
rectly on the macroscopic scale, are similar to those obtained by 816
756 Appendix A. Elements of Biot’s theory of poroelasticity and applying the method of spatial homogenization [71] and the meth- 817
757 mixture theory od of spatial averaging [75] to the equations that hold at the scale 818
of pores and grains.
RR

819
758 A.1. Biot’s phenomenological theory for poroelastic materials

A.2. Mixture theory for a biphasic fluid–solid system 820


759 A cellular biological medium can be considered as a poroelastic
760 material, which consists of an elastic solid matrix (cells plus EPS
In the simplest case, a cellular biological medium can be consid- 821
761 meshwork) and well inter-connected pores that are occupied by
CO

ered as a biphasic mixture which consists of a solid constituent 822


762 the extracellular fluid solution. In the context of the initial Biot the-
(cells plus EPS network) and a fluid constituent (extracellular 823
763 ory for poroelastic materials, the equations that govern momen-
fluid). We consider that the fluid (f-constituent) behaves as a New- 824
764 tum transfer in the cellular biological medium, on the B-scale of
tonian fluid, and the solid (s-constituent) behaves as a linearly 825
765 observation, are [36]
elastic isotropic solid. For the systems under consideration, it is 826
766 Momentum balance in the cellular biological medium
767 reasonable to neglect the convective momentum transfer and mass 827
UN

769 r  reff ¼ 0 ðA1Þ production–consumption (for time scales of observation much 828
smaller than the characteristic time of cell division). Under these 829
770 Constitutive equation for the ‘‘effective” stress tensor
771 h i assumptions, in the context of biphasic mixture theory, the equa- 830
T tions that govern momentum transfer in the cellular biological
773 reff ¼ leff ru þ ðruÞ þ keff eI  aPI ðA2Þ 831
medium, on the B-scale of observation can be written as follows 832
774 Darcy’s ‘‘law” for the fluid motion [50,53,55]: 833
775
Mass balance for the ath constituent ða ¼ f ; sÞ 834
keff
v¼ rP ðA3Þ
777 lf @ qa ua
þ r  ðqa ua v a Þ ¼ 0 ðA12Þ 836
778 Mass balance for the fluid @t
779
@h Momentum balance for the ath constituent ða ¼ f ; sÞ 837
781 ¼ r  v ðA4Þ
@t @va X
782 Constitutive equation for the ‘‘variation in water content”
qa ua ¼ r  ra þ Fx!a þ qa ua ba ðA13Þ
783
@t x–a 839
785 h ¼ ae þ P=M ðA5Þ
Constitutive equation for the stress tensor ra ða ¼ f ; sÞ 840

Please cite this article in press as: G.E. Kapellos et al., Theoretical modeling of fluid flow in cellular biological media: An overview, Math. Biosci. (2010),
doi:10.1016/j.mbs.2010.03.003
MBS 7041 No. of Pages 11, Model 5G
25 March 2010
ARTICLE IN PRESS

G.E. Kapellos et al. / Mathematical Biosciences xxx (2010) xxx–xxx 9

h i
rs ¼ us PI þ ks ðr  us ÞI þ ls rus þ ðrus ÞT ðA14Þ References 895

2 h i
[1] G.N. Bancroft, V.I. Sikavitsas, J. van den Dolder, T.L. Sheffield, C.G. Ambrose, 896
842 rf ¼ uf PI  lf ðr  vf ÞI þ lf rv f þ ðrvf ÞT ðA15Þ J.A. Jansen, A.G. Mikos, Fluid flow increases mineralized matrix deposition in 897
3
3D perfusion culture of marrow stromal osteoblasts in a dose-dependent 898
843 Constitutive equation for the interaction force F x!a manner, Proc. Natl. Acad. Sci. USA 99 (20) (2002) 12600. 899
[2] E.J. Bouwer, H.H.M. Rijnaarts, A.B. Cunningham, R. Gerlach, Biofilms in porous 900
845 Ff!s ¼ Fs!f ¼ K ðv f  v s Þ þ Prus ðA16Þ media, in: J.D. Bryers (Ed.), Biofilms II: Process Analysis and Applications, 901
Willey-Liss, 2000, p. 123. 902
846 Saturation condition [3] T.G. Ellis, B. Eliosov, C.G. Schmit, K. Jahan, K.Y. Park, Activated sludge and 903
847 other aerobic suspended culture processes, Water Environ. Res. 74 (4) (2002) 904
849 uf þ us ¼ 1 ðA17Þ 385. 905
[4] G. Foley, A review of factors affecting filter cake properties in dead-end 906
850 Equation for the time evolution of volume fraction [54] microfiltration of microbial suspensions, J. Membr. Sci. 274 (2006) 38. 907
[5] R.K. Jain, Delivery of molecular and cellular medicine to solid tumors, J. 908

F
852 us ¼ u0s þ u0s r  us ðA18Þ Control. Release 53 (1998) 49. 909
[6] C.P. Ng, M.A. Swartz, Mechanisms of interstitial flow-induced remodeling of 910
853 Relationship between velocity–displacement for the ath constituent fibroblast-collagen cultures, Ann. Biomed. Eng. 34 (3) (2006) 446. 911

OO
[7] E.A. Swabb, J. Wei, P.M. Gullino, Diffusion and convection in normal and 912
854 ða ¼ f ; sÞ 913
855 neoplastic tissues, Cancer Res. 34 (1974) 2815.
[8] R.H. Bobo, D.W. Laske, A. Akbasak, P.F. Morrison, R.L. Dedrick, E.H. Oldfield, 914
@ua
857 va ¼ ðA19Þ Convection-enhanced delivery of macromolecules in the brain, Proc. Natl. 915
@t Acad. Sci. USA 91 (6) (1994) 2076. 916
[9] X.-Y. Li, Y. Yuan, H.-W. Wang, Hydrodynamics of biological aggregates of 917
858 where ua is the volume fraction (the exponent ‘0’ denotes the initial different sludge ages: an insight into the mass transport mechanisms of 918
859 value), qa is the true density, v a is the velocity, ua is the displace- 919

PR
bioaggregates, Environ. Sci. Technol. 37 (2003) 292.
860 ment, ra is the ‘‘partial” stress tensor of the ath phase, P is the fluid [10] R.C. Evans, T.M. Quinn, Solute convection in dynamically compressed 920
cartilage, J. Biomech. 39 (2006) 1048. 921
861 pressure, Fx!a is the interaction force (per unit volume) between [11] J.D. Fowler, C.R. Robertson, Metabolic behavior of immobilized aggregates of 922
862 the ath and xth constituents of the mixture, ba is a body force Escherichia coli under conditions of varying mechanical stress, Appl. Environ. 923
863 (e.g. in the case of gravity ba is set equal to the gravity acceleration Microbiol. 57 (1) (1991) 93. 924
[12] M.V. Hillsley, J.A. Frangos, Bone tissue engineering: the role of interstitial 925
864 constant, g), ls and ks are the Lamé parameters for the solid, lf is 926
fluid flow, Biotechnol. Bioeng. 43 (1994) 573.
865 the fluid viscosity, and K is a phenomenological coefficient that D [13] V.I. Sikavitsas, J.S. Temenoff, A.G. Mikos, Biomaterials and bone 927
866 can be determined experimentally. It is convenient to set mechanotransduction, Biomaterials 22 (2001) 2581. 928
[14] J.M. Tarbell, S. Weinbaum, R.D. Kamm, Cellular fluid mechanics and mechano- 929
lf transduction, Ann. Biomed. Eng. 33 (12) (2005) 1719. 930
868
K¼ u2f ðA20Þ [15] J.M. Rutkowski, M.A. Swartz, A driving force for change: interstitial flow as a 931
keff
TE
morphoregulator, Trends Cell Biol. 17 (1) (2006) 44. 932
869 where keff is the permeability of the material. Further, we can con- [16] K. Piekarski, M. Munro, Transport mechanism operating between blood 933
supply and osteocytes in long bones, Nature 269 (1977) 80. 934
870 sider each constituent to be intrinsically incompressible, and disre- [17] M.A. Swartz, M.E. Fleury, Interstitial flow and its effects in soft tissues, Annu. 935
871 gard the local acceleration, the body force, and the viscous Rev. Biomed. Eng. 9 (2007) 229. 936
872 contributions in the fluid stress tensor. Under these considerations, [18] J.C. Anderson, C. Eriksson, Piezoelectric properties of dry and wet bone, 937
EC

Nature 227 (1970) 491. 938


873 the mass and momentum balances for the fluid and solid constitu- [19] M.W. Johnson, D.A. Chakkalakal, R.A. Harper, J.L. Katz, S.W. Rouhana, Fluid 939
874 ents become flow in bone in vitro, J. Biomech. 15 (11) (1982) 881. 940
875 [20] V.I. Sikavitsas, G.N. Bancroft, H.L. Holtorf, J.A. Jansen, A.G. Mikos, Mineralized 941
@ us matrix deposition by marrow stromal osteoblasts in 3D perfusion culture 942
þ r  ðus v s Þ ¼ 0 ðA21Þ 943
@t increases with increasing fluid shear forces, Proc. Natl. Acad. Sci. USA 100 (25)
(2003) 14683. 944
@ uf
RR

þ r  ðuf v f Þ ¼ 0 ðA22Þ [21] W. Liu, L.M. Jawerth, E.A. Sparks, M.R. Falvo, R.R. Hantgan, R. Superfine, S.T. 945
@t Lord, M. Guthold, Fibrin fibers have extraordinary extensibility and elasticity, 946
2 lt Science 313 (2006) 634. 947
0 ¼ us rP þ ls r us þ ðls þ ks Þrðr  us Þ þ u2t ðv f  v s Þ ðA23Þ [22] V.C. Mow, S.C. Kuei, W. M Lai, C.G. Armstrong, Biphasic creep and stress 948
k
949
lf relaxation of articular cartilage in compression: theory and experiments, J.
877
0 ¼ uf rP  u2f ðv f  v s Þ ðA24Þ Biomech. Eng. 102 (1980) 73. 950
keff [23] W.M. Lai, J.S. Hou, V.C. Mow, A triphasic theory for the swelling and 951
CO

deformation behaviors of articular cartilage, J. Biomech. Eng. 113 (1991) 245. 952
878 By adding Eqs. (A23) and (A24) we obtain [24] W.Y. Gu, W.M. Lai, V.C. Mow, A mixture theory for charged-hydrated soft 953
879
tissues containing multi-electrolytes: passive transport and swelling 954
881 0 ¼ ls r2 us þ ðls þ ks Þrðr  us Þ  rP ðA25Þ behaviors, J. Biomech. Eng. 120 (1998) 169. 955
[25] E.S. Almeida, R.L. Spilker, Finite element formulations for hyperelastic 956
882 By taking the divergence of Eq. (A25) it follows that [39]: transversely isotropic biphasic soft tissues, Comput. Methods Appl. Mech. 957
883 Eng. 151 (1998) 513. 958
885 ð2leff þ keff Þr2 e ¼ r2 P ðA26Þ 959
UN

[26] C. Moyne, M.A. Murad, Macroscale modeling of cartilage: mixture theory


versus homogenization, Biorheology 41 (2004) 215. 960
886 where e ¼ r  us is the solid dilatation. By adding the Eqs. (A21) and [27] S.C. Cowin, Bone poroelasticity, J. Biomech. 32 (1999) 217. 961
[28] J.D. Humphrey, Continuum biomechanics of soft biological tissues, Proc. R. 962
887 (A22), and taking into account the saturation condition (A17) and 963
Soc. Lond. A 459 (2003) 3.
888 Eqs. (A19), (A24) and (A26) it follows that: [29] C. Verdier, Rheological properties of living materials. From cells to tissues, J. 964
889 965
Theor. Med. 5 (2) (2003) 67.
r  ðus v s þ uf v f Þ ¼ 0 [30] B.G. Sengers, M. Taylor, C.P. Please, R.O.C. Oreffo, Computational modelling of 966
cell spreading and tissue regeneration in porous scaffolds, Biomaterials 28 967
() r  ½v s þ uf ðv f  v s Þ ¼ 0 968
(2007) 1926.
@e keff 2 [31] M. Sugihara-Seki, B.M. Fu, Blood flow and permeability in microvessels, Fluid 969
()  r P¼0 Dyn. Res. 37 (2005) 82. 970
@t lf [32] J.W. Costerton, Z. Lewandowski, D. de Beer, D. Caldwell, D. Korber, G. James, 971
@e keff   Biofilms: the customized microniche, J. Bacteriol. 176 (8) (1994) 2137. 972
() ¼ 2leff þ keff r2 e ðA27Þ [33] R.M. Donlan, Biofilms: microbial life on surfaces, Emerg. Infect. Dis. 8 (9) 973
891 @t lf (2002) 881. 974
[34] J. Wingender, T.R. Neu, H.-C. Flemming, What are bacterial extracellular 975
892 We observe that Eqs. (A25) and (A27) have the same form with 976
polymeric substances?, in: J. Wingender, T.R. Neu, H.-C. Flemming (Eds.),
893 Eqs. (A8) and (A11), which were obtained previously in the context Microbial Extracellular Polymeric Substances: Characterization, Structure 977
894 of Biot’s theory (under the same set of hypotheses). and Function, Springer, 1999, p. 1. 978

Please cite this article in press as: G.E. Kapellos et al., Theoretical modeling of fluid flow in cellular biological media: An overview, Math. Biosci. (2010),
doi:10.1016/j.mbs.2010.03.003
MBS 7041 No. of Pages 11, Model 5G
25 March 2010
ARTICLE IN PRESS

10 G.E. Kapellos et al. / Mathematical Biosciences xxx (2010) xxx–xxx

979 [35] M.G. Turner, R.H. Gardner, R.V. O’Neill, Landscape Ecology in Theory and [70] B.W. Towler, A. Cunningham, P. Stoodley, L. McKittrick, A model of fluid– 1063
980 Practice: Pattern and Process, Springer, New York, 2001. biofilm interaction using a Burger material law, Biotechnol. Bioeng. 96 (2) 1064
981 [36] M.A. Biot, General theory of three-dimensional consolidation, J. Appl. Phys. 12 (2007) 259. 1065
982 (1941) 155. [71] R. Burridge, J.B. Keller, Poroelasticity equations derived from microstructure, 1066
983 [37] M.A. Biot, Mechanics of deformation and acoustic propagation in porous J. Acoust. Soc. Am. 70 (4) (1981) 1140. 1067
984 media, J. Appl. Phys. 33 (4) (1962) 1482. [72] C.M. Marle, Écoulements monophasiques en milieu poreux, Revue de l’ 1068
985 [38] J.L. Nowinski, C.F. Davis, A model of the human skull as a poroelastic spherical Institut Français du Pétrole XXII (10) (1967) 1471. 1069
986 shell subjected to a quasistatic load, Math. Biosci. 8 (1970) 397. [73] C.M. Marle, On macroscopic equations governing multiphase flow with 1070
987 [39] P. Basser, Interstitial pressure, volume, and flow during infusion into brain diffusion and chemical reactions in porous media, Int. J. Eng. Sci. 20 (5) (1982) 1071
988 tissue, Microvasc. Res. 44 (1992) 143. 643. 1072
989 [40] P.A. Netti, L.T. Baxter, Y. Boucher, R. Skalak, R.K. Jain, Macro- and microscopic [74] V. de la Cruz, T.J.T. Spanos, Seismic wave propagation in a porous medium, 1073
990 fluid transport in living tissues: application to solid tumors, AIChE J. 43 (3) Geophysics 50 (10) (1985) 1556. 1074
991 (1997) 818. [75] S. Whitaker, Flow in porous media III: deformable media, Transp. Porous 1075
992 [41] T. Roose, P.A. Netti, L.L. Munn, Y. Boucher, R.K. Jain, Solid stress generated by Media 1 (1986) 127. 1076
993 spheroid growth estimated using a linear poroelasticity model, Microvasc. [76] S. Whitaker, The Method of Volume Averaging, Kluwer Academic Publishers., 1077

F
994 Res. 66 (2003) 205. The Netherlands, 1999. 1078
995 [42] J.H. Smith, J.A.C. Humphrey, Interstitial transport and transvascular fluid [77] S. Whitaker, Mechanics of composite solids, J. Eng. Mech. 128 (8) (2002) 823. 1079
996 exchange during infusion into brain and tumor tissue, Microvasc. Res. 73 [78] J.M. Crolet, B. Aoubiza, A. Meunier, Compact bone: numerical simulation of 1080

OO
997 (2007) 58. mechanical characteristics, J. Biomech. 26 (6) (1993) 677. 1081
998 [43] D.L. Johnson, Elastodynamics of gels, J. Chem. Phys. 77 (3) (1982) 1531. [79] S.J. Hollister, D.P. Fyhrie, K.J. Jepsen, S.A. Goldstein, Application of 1082
999 [44] C.-Y. Hui, V. Muralidharan, Gel mechanics: a comparison of the theories of homogenization theory to the study of trabecular bone mechanics, J. 1083
1000 Biot and Tanaka, Hocker, and Benedek, J. Chem. Phys. 123 (2005) 154905. Biomech. 24 (9) (1991) 825. 1084
1001 [45] D.J. Lee, C.H. Wang, Theories of cake filtration and consolidation and [80] B.D. Wood, S. Whitaker, Diffusion and reaction in biofilms, Chem. Eng. Sci. 53 1085
1002 implications to sludge dewatering, Water Res. 34 (1) (2000) 1. (3) (1998) 397; 1086
1003 [46] R. Bai, C. Tien, Further work on cake filtration analysis, Chem. Eng. Sci. 60 B.D. Wood, S. Whitaker, Erratum, Chem. Eng. Sci. 55 (2000) 2349. 1087
1004 (2005) 301. [81] B.D. Wood, S. Whitaker, Cellular growth in biofilms, Biothechnol. Bioeng. 64 1088

PR
1005 [47] C.T. Truesdell, On the foundations of mechanics and energetic, Continuum (6) (1999) 656. 1089
1006 Mechanics: The Rational Mechanics of Materials, vol. II, Gordon & Breach, NY, [82] B.D. Wood, S. Whitaker, Multi-species diffusion and reaction in biofilms and 1090
1007 1965, p. 293. cellular media, Chem. Eng. Sci. 55 (2000) 3397. 1091
1008 [48] A. Bedford, D.S. Drummheller, Theory of immiscible and structured mixtures, [83] G.E. Kapellos, T.S. Alexiou, A.C. Payatakes, A multiscale theoretical model for 1092
1009 Int. J. Eng. Sci. 21 (1983) 863. diffusive mass transfer in cellular biological media, Math. Biosci. 210 (1) 1093
1010 [49] K.R. Rajagopal, L. Tao, Mechanics of Mixtures, World Scientific, Singapore, (2007) 177. 1094
1011 1995. [84] I.G. Kevrekidis, C.W. Gear, G. Hummer, Equation-free: the computer-aided 1095
1012 [50] E.R. Damiano, B.R. Duling, K. Ley, T.C. Skalak, Axisymmetric pressure-driven analysis of complex multiscale systems, AIChE J. 50 (7) (2004) 1346. 1096
1013 flow of rigid pellets through a cylindrical tube lined with a deformable porous [85] R.G.M. Breuls, B.G. Sengers, C.W.J. Oomens, C.V.C. Bouten, F.P.T. Baaijens, 1097
1014 wall layer, J. Fluid Mech. 314 (1996) 163.
D Predicting local cell deformations in engineered tissue constructs: a 1098
1015 [51] P.A. Netti, L.T. Baxter, Y. Boucher, R. Skalak, R.K. Jain, Time-dependent multilevel finite element approach, J. Biomech. Eng. 124 (2002) 198. 1099
1016 behavior of interstitial fluid pressure in solid tumors: implications for drug [86] T. Stylianopoulos, V.H. Barocas, Volume-averaging theory for the study of the 1100
TE
1017 delivery, Cancer Res. 55 (1995) 5451. mechanics of collagen networks, Comput. Methods Appl. Mech. Eng. 196 1101
1018 [52] H. Byrne, L. Preziosi, Modelling solid tumor growth using the theory of (2007) 2981. 1102
1019 mixtures, Math. Med. Biol. 20 (2003) 341. [87] Y. Boucher, C. Brekken, P.A. Netti, L.T. Baxter, R.K. Jain, Intratumoral infusion 1103
1020 [53] P.A. Netti, F. Travascio, R.K. Jain, Coupled macromolecular transport and gel of fluid: estimation of hydraulic conductivity and implications for the 1104
1021 mechanics: poroviscoelastic approach, AIChE J. 49 (6) (2003) 1580. delivery of therapeutic agents, Br. J. Cancer 78 (1998) 1442. 1105
1022 [54] S.I. Barry, G.K. Aldis, Comparison of models for flow induced deformation of [88] X.-Y. Zhang, J. Luck, M.W. Dewhirst, F. Yuan, Interstitial hydraulic 1106
1023 1107
EC

soft biological tissue, J. Biomech. 23 (7) (1990) 647. conductivity in a fibrosarcoma, Am. J. Physiol. Heart Circ. Physiol. 279
1024 [55] S.I. Barry, G.K. Aldis, Flow-induced deformation from pressurized cavities in (2000) H2726. 1108
1025 absorbing porous tissues, Bull. Math. Biol. 54 (6) (1992) 977. [89] M. Milosevic, S.J. Lunt, E. Leung, J. Skliarenko, P. Shaw, A. Fyles, R.P. Hill, 1109
1026 [56] V.H. Barocas, R.T. Tranquillo, An anisotropic biphasic theory of tissue- Interstitial permeability and elasticity in human cervix cancer, Microvasc. 1110
1027 equivalent mechanics: the interplay among cell traction, fibrillar network Res. 75 (2008) 381. 1111
1028 deformation, fibril alignment, and cell contact guidance, J. Biomech. Eng. 119 [90] G. Vunjak-Novakovic, I. Martin, B. Obradovic, S. Treppo, A.J. Grodzinsky, R. 1112
1029 (1997) 137. Langer, L.E. Freed, Bioreactor cultivation conditions modulate the 1113
RR

1030 [57] B.G. Sengers, C.W.J. Oomens, F.P.T. Baaijens, An integrated finite-element composition and mechanical properties of tissue-engineered cartilage, J. 1114
1031 approach to mechanics, transport and biosynthesis in tissue engineering, J. Orthop. Res. 17 (1999) 130. 1115
1032 Biomech. Eng. 126 (2004) 82. [91] W.Y. Gu, H. Yao, Effects of hydration and fixed charge density on fluid 1116
1033 [58] B. Loret, F.M.F. Simões, A framework for deformation, generalized diffusion, transport in charged hydrated soft tissues, Ann. Biomed. Eng. 31 (2003) 1162. 1117
1034 mass transfer and growth in multi-species multi-phase biological tissues, Eur. [92] B. Reynaud, T.M. Quinn, Anisotropic hydraulic permeability in compressed 1118
1035 J. Mech. A Solids 24 (2005) 757. articular cartilage, J. Biomech. 39 (2006) 131. 1119
1036 [59] G. Lemon, J.R. King, H.M. Byrne, O.E. Jensen, K.M. Shakesheff, Mathematical [93] P. Heneghan, P.E. Riches, Determination of the strain-dependent hydraulic 1120
1037 modeling of engineered tissue growth using a multiphase porous flow 1121
CO

permeability of the compressed bovine nucleus pulposus, J. Biomech. 41


1038 mixture theory, J. Math. Biol. 52 (2006) 571. (2008) 903. 1122
1039 [60] R. Burger, Phenomenological foundation and mathematical theory of [94] R.M. Wu, G.W. Tsou, D.J. Lee, Estimate of sludge floc permeability, Chem. Eng. 1123
1040 sedimentation–consolidation processes, Chem. Eng. J. 80 (2000) 177. J. 80 (2000) 37. 1124
1041 [61] G.E. Kapellos, T.S. Alexiou A.C. Payatakes, A multiscale theoretical model for [95] K. Nakanishi, T. Tadokoro, R. Matsuno, On the specific resistance of cakes of 1125
1042 Q3 fluid flow in cellular biological media, in preparation. microorganisms, Chem. Eng. Commun. 62 (1987) 187. 1126
1043 [62] B.E. Logan, J.R. Hunt, Bioflocculation as a microbial response to substrate [96] J.D. Fowler, C.R. Robertson, Hydraulic permeability of immobilized bacterial 1127
1044 limitations, Biotechnol. Bioeng. 31 (2) (1988) 91. cell aggregates, Appl. Environ. Microbiol. 57 (1) (1991) 102. 1128
UN

1045 [63] G. Stephanopoulos, K. Tsiveriotis, The effect of intraparticle convection on [97] P.H. Hodgson, G.L. Leslie, R.P. Schneider, A.G. Fane, C.J.D. Fell, K.C. Marshall, 1129
1046 nutrient transport in porous biological pellets, Chem. Eng. Sci. 44 (9) (1989) Cake resistance and solute rejection in bacterial microfiltration: the role of 1130
1047 2031. the extracellular matrix, J. Membr. Sci. 79 (1993) 35. 1131
1048 [64] X.-Y. Li, Y. Yuan, Settling velocities and permeabilities of microbial [98] T. Oolman, T.-C. Liu, Filtration properties of mycelial microbial broths, 1132
1049 aggregates, Water Res. 36 (2002) 3110. Biotechnol. Prog. 7 (1991) 535. 1133
1050 [65] G.E. Kapellos, T.S. Alexiou, A.C. Payatakes, Hierarchical simulator of biofilm [99] T. Tanaka, K.-I. Abe, K. Nakanishi, Shear-induced arrangement of cells in cake 1134
1051 growth and dynamics in granular porous materials, Adv. Water Resour. 30 during crossflow filtration of Escherichia coli cells, Biotechnol. Tech. 8 (1) 1135
1052 (2007) 1648. (1994) 57. 1136
1053 [66] L.T. Baxter, R.K. Jain, Transport of fluid and macromolecules in tumors. I. Role [100] A.A. McCarthy, D.G. O’Shea, N.T. Murray, P.K. Walsh, G. Foley, Effect of cell 1137
1054 of interstitial pressure and convection, Microvasc. Res. 37 (1989) 77. morphology on dead-end filtration of the dimorphic yeast Kluyveromyces 1138
1055 [67] C. Pozrikidis, D.A. Farrow, A model of fluid flow in solid tumors, Ann. Biomed. marxianus var. marxianus NRRLy2415, Biotechnol. Prog. 14 (1998) 279. 1139
1056 Eng. 31 (2003) 181. [101] R. Beckie, A comparison of methods to determine measurement support 1140
1057 [68] J. Stachowska-Pietka, J. Waniewski, M.F. Flessner, B. Lindholm, A distributed volumes, Water Resour. Res. 37 (4) (2001) 925. 1141
1058 model of bidirectional protein transport during peritoneal fluid adsorption, [102] B. van Rietbergen, H. Weinans, R. Huiskes, A. Odgaard, A new method to 1142
1059 Adv. Perit. Dial. 23 (2007) 23. determine trabecular bone elastic properties and loading using 1143
1060 [69] H.J. Dupin, P.K. Kitanidis, P.L. McCarty, Pore-scale modeling of biological micromechanical finite-element models, J. Biomech. 28 (1) (1995) 69. 1144
1061 clogging due to aggregate expansion: a material mechanics approach, Water [103] E.M. Johnson, W.M. Deen, Hydraulic permeability of agarose gels, AIChE J. 42 1145
1062 Resour. Res. 37 (12) (2001) 2965. (5) (1996) 1220. 1146

Please cite this article in press as: G.E. Kapellos et al., Theoretical modeling of fluid flow in cellular biological media: An overview, Math. Biosci. (2010),
doi:10.1016/j.mbs.2010.03.003
MBS 7041 No. of Pages 11, Model 5G
25 March 2010
ARTICLE IN PRESS

G.E. Kapellos et al. / Mathematical Biosciences xxx (2010) xxx–xxx 11

1147 [104] B. Hellagson, E. Perilli, E. Schileo, F. Taddei, S. Brynjólfsson, M. Viceconti, [117] K.J. Mattern, W.M. Deen, ‘‘Mixing rules” for estimating the hydraulic 1180
1148 Mathematical relationships between bone density and mechanical permeability of fiber mixtures, AIChE J. 54 (1) (2008) 32. 1181
1149 properties: a literature review, Clin. Biomech. 23 (2008) 135. [118] G.W. Jackson, D.F. James, The permeability of fibrous porous media, Can. J. 1182
1150 [105] A.L. Zydney, W.M. Saltzman, C.K. Colton, Hydraulic resistance of red cell beds Chem. Eng. 64 (1986) 365. 1183
1151 in an unstirred filtration cell, Chem. Eng. Sci. 44 (1) (1989) 147. [119] J.R. Levick, Flow through the interstitium and other fibrous matrices, Q. J. Exp. 1184
1152 [106] L.J. Gibson, The mechanical behaviour of cancellous bone, J. Biomech. 18 (5) Physiol. 72 (1987) 409. 1185
1153 (1985) 317. [120] L. Skartsis, J.L. Kardos, B. Khomami, Resin flow through fiber beds during 1186
1154 [107] S.R. Eisenberg, A.J. Grodzinsky, Electrokinetic micromodel of extracellular composite manufacturing processes. Part I: review of Newtonian flow 1187
1155 matrix and other polyelectrolyte networks, Physicochem. Hydrodyn. 10 (4) through fiber beds, Polym. Eng. Sci. 32 (4) (1992) 221. 1188
1156 (1988) 517. [121] J.E. Drummond, M.I. Tahir, Laminar viscous flow through regular arrays of 1189
1157 [108] P. Chammas, W.J. Federspiel, S.R. Eisenberg, A microcontinuum model of parallel solid cylinders, Int. J. Multiphase Flow 10 (1984) 515. 1190
1158 electrokinetic coupling in the extracellular matrix: perturbation formulation [122] J.J.L. Higdon, G.D. Ford, Permeability of three-dimensional models of fibrous 1191
1159 and solution, J. Colloid Interface Sci. 168 (1994) 526. porous media, J. Fluid Mech. 308 (1996) 341. 1192
1160 [109] T. Beno, Y.-J. Yoon, S.C. Cowin, S.P. Fritton, Estimation of bone permeability [123] D.S. Clague, B.D. Kandhai, R. Zhang, P.M.A. Sloot, Hydraulic permeability of 1193
1161 using accurate microstructural measurements, J. Biomech. 39 (2006) 2378. (un)bounded fibrous media using the lattice Boltzmann method, Phys. Rev. E 1194

F
1162 [110] M.H. Schwartz, P.H. Leo, J.L. Lewis, A microstructural model for the elastic 61 (1) (2000) 616. 1195
1163 response of articular cartilage, J. Biomech. 27 (7) (1994) 865. [124] C.N. Davies, The separation of airborne dust and particles, Proc. Inst. Mech. 1196
1164 [111] T.W. Secomb, A.W. El-Kareh, A theoretical model for the elastic properties of Eng. Lond. B1 (1952) 185. 1197

OO
1165 very soft tissues, Biorheology 38 (2001) 305. [125] P.R. Johnston, Revisiting the most probable pore-size distribution in filter 1198
1166 [112] B.G. Sengers, C.C. van Donkelaar, C.W.J. Oomens, F.P.T. Baaijens, The local media: the Gamma distribution, Filtr. Sep. 35 (3) (1998) 287. 1199
1167 matrix distribution and the functional development of tissue engineered [126] A. Koponen, D. Kandhai, E. Hellén, M. Alava, A. Hoekstra, M. Kataja, K. 1200
1168 cartilage, a finite element study, Ann. Biomed. Eng. 32 (12) (2004) 1718. Niskanen, P. Sloot, J. Timonen, Permeability of three-dimensional random 1201
1169 [113] S.J. Hollister, J.M. Brennan, N. Kikuchi, A homogenization sampling procedure fiber webs, Phys. Rev. Lett. 80 (4) (1998) 716. 1202
1170 for calculating trabecular bone effective stiffness and tissue level stress, J. [127] J. Happel, Viscous flow relative to arrays of cylinders, AIChE J. 5 (1959) 175. 1203
1171 Biomech. 27 (4) (1994) 433. [128] S. Kuwabara, The forces experienced by randomly distributed parallel 1204
1172 [114] A. Fritsch, C. Hellmich, ‘Universal’ microstructural patterns in cortical circular cylinders or spheres in a viscous flow at small Reynolds numbers, 1205

PR
1173 and trabecular, extracellular and extravascular bone materials: J. Phys. Soc. Jpn. 14 (1959) 527. 1206
1174 micromechanics-based prediction of anisotropic elasticity, J. Theor. [129] A.S. Sangani, A. Acrivos, Slow flow past periodic arrays of cylinders with 1207
1175 Biol. 244 (2007) 597. application to heat transfer, Int. J. Multiphase Flow 8 (1982) 193. 1208
1176 [115] B.D. Wood, M. Quintard, S. Whitaker, Calculation of effective diffusivities for [130] B.R. Gebart, Permeability of unidirectional reinforcements for RTM, J. 1209
1177 biofilms and tissues, Biotechnol. Bioeng. 77 (5) (2002) 495. Compos. Mater. 26 (8) (1992) 1100. 1210
1178 [116] P.K. Zysset, A review of morphology–elasticity relationships in human [131] M.A. Biot, D.G. Willis, The elastic coefficients of the theory of consolidation, J. 1211
1179 trabecular bone: theories and experiments, J. Biomech. 36 (2003) 1469. Appl. Mech. 24 (1957) 595. 1212

D 1213
TE
EC
RR
CO
UN

Please cite this article in press as: G.E. Kapellos et al., Theoretical modeling of fluid flow in cellular biological media: An overview, Math. Biosci. (2010),
doi:10.1016/j.mbs.2010.03.003

You might also like