Figures
Abstract
Background
The finite element method has complimented research in the field of network mechanics in the past years in numerous studies about various materials. Numerical predictions and the planning efficiency of experimental procedures are two of the motivational aspects for these numerical studies. The widespread availability of high performance computing facilities has been the enabler for the simulation of sufficiently large systems.
Objectives and Motivation
In the present study, finite element models were built for sintered, metallic fibre networks and validated by previously published experimental stiffness measurements. The validated models were the basis for predictions about so far unknown properties.
Materials and Methods
The finite element models were built by transferring previously published skeletons of fibre networks into finite element models. Beam theory was applied as simplification method.
Results and Conclusions
The obtained material stiffness isn’t a constant but rather a function of variables such as sample size and boundary conditions. Beam theory offers an efficient finite element method for the simulated fibre networks. The experimental results can be approximated by the simulated systems. Two worthwhile aspects for future work will be the influence of size and shape and the mechanical interaction with matrix materials.
Citation: Bosbach WA (2015) The Elastic Behaviour of Sintered Metallic Fibre Networks: A Finite Element Study by Beam Theory. PLoS ONE 10(11): e0143011. https://doi.org/10.1371/journal.pone.0143011
Editor: Paul J. Atzberger, UC Santa Barbara, UNITED STATES
Received: August 17, 2015; Accepted: October 29, 2015; Published: November 16, 2015
Copyright: © 2015 Wolfram A. Bosbach. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited
Data Availability: All relevant data are within the paper.
Funding: This research was supported by the European Research Council Grant No 240446 (http://erc.europa.eu/).
Competing interests: The authors have declared that no competing interests exist.
Introduction
Background
The mechanics of metallic [1] and non-metallic [2] fibre networks have since long been the subject of research studies. One possible way to categorize the available studies is by the specific material which is investigated. Early and recent studies about cellulose material and paper can be found in [2–6]. Numerous studies exist about the mechanics of polymeric networks [7–11]. The modelling of polymeric non-woven fabrics and their particularly complex behaviour [12] has been in the focus of several research studies [13–17]. In the field of biomaterials, a big number of studies exists about the mechanics of actin networks and cytoskeletons [18–22]. A range of studies refers to theoretical networks whose geometries have been generated by computer code [23–31].
The mechanics of sintered, metallic fibre networks as used for the present study have been investigated by [1, 30–35]. In [36], an architectural characterization of the six network samples of the present study was published, together with experimentally obtained mechanical properties.
For finite element (FE) analyses of structures whose dimensions are dominated by their extension along only one axis, such as screws or fibres, beam theory can offer an efficient simulation method [37]. The foundations of beam theory were laid in [38–40] and are available in today’s textbooks [41]. For the application of boundary conditions (BC) to random fibre networks and for the determination of the representative volume element (RVE), [29, 42] have provided a much referred to concept [43–46]. By this concept, the RVE is obtained for a defined relative error for each physical property. Image acquisition by computed tomography (CT) scanning which uses the principle of Röntgen radiation [47] is a commonly used method for measuring the dimensions of metallic three-dimensional (3D) structures [48].
Motivation and scope of the present study
The motivation of the present study was to predict previously unknown mechanical properties for metallic, sintered fibre networks. For that purpose, FE models were developed. The input geometries for the FE models were based on CT scans, acquired from real network samples in [36]. Experimental values for the network Young’s modulus from that same study were used for a validation of the FE models. The properties of these six network samples define the scope and at the same time also the limitations of the present study. Other metallic, sintered networks with similar but different properties will require a re-evaluation of the proposed FE models.
Mathematical notation
Throughout the present study, the following notation is used: x for scalars, for vectors, and
for 2nd rank tensors. Cube faces are written X. The vector product is given as “×” and dot product as “⋅”. Relations which are greater-than and approximately equal are written “
”. If the relation is greater-than or equal “≥” is used.
Materials and Methods
Network samples and meshing step
The present study uses as geometry input for the FE models the dimensions of six AISI (American Iron and Steel Institute) 316L network samples. The architectural network values have been published in [36] (see Table 1). Each scanned sample section is a cube of volume V = 43 mm3 (see Fig 1a). Following the two-phase model in [29], V is split into fibre volume and void volume:
(1)
(a) Dimensions of fibre network cube V = 43 mm3 (fibre network details in [36]) and (b) six quadratic cube faces (Fx0, Fx1, Fy0, Fy1, Fz0, and Fz1) forming S.
Equally, the volume boundary ∂V is defined to consist of a section of fibre boundary and a section of void boundary on the cube surface S:
(2)
Six quadratic cube faces form S (see Fig 1b), two of them perpendicular each to one of the three axes:
(3)
Two samples each are manufactured with a fibre volume fraction f of 10, 15, and 20%. In [36], the fibre production process by bundle drawing [49], the network manufacturing at N.V. Bekaert S.A. (Belgium), the CT scan acquisition at General Electric (resolution 7.75 μm), and the applied skeletonisation algorithm [50–52] are discussed in detail. The skeletonisation algorithm reduces the 3D fibre bodies to their medial axes which are strings of voxels (see Fig 2). The medial axis models obtained for [36] are transferred in the present study into beam assemblies and run as FE models. The approximately hexagonal fibre cross-section is simulated as a round cross-section of radius R = 20 μm.
FE models
For the present study, two FE models are implemented (see Table 2). Model A and Model B are based on five specifications:
- Linear elasticity and static equilibrium,
- Euler-Bernoulli or Timoshenko beam elements as AISI 316L fibres,
- Rigid joints or torsional springs as inter-fibre joint models at contact points between fibre paths (see Fig 2),
- BC as defined in Eqs (8) and (11), and
- the new introduced model parameter of BC-depth hBC.
Linear elasticity and static equilibrium conditions.
For mechanical FE analyses, the simulated system is transferred into a representation by the global stiffness matrix ; the most common assembly method being the direct stiffness method [53].
links the global load vector
and the global displacement vector
:
(4)
The meshed geometries and material specifications define the variables of . In the present study, the material is assumed to behave linearly elastic (i.e. fulfil Hook’s law that stress and strain are linked by the Young’s modulus: σ = Eϵ [54]). In this case, the variables of
become constants. The AISI 316L material stiffness is simulated as Efibre = 200GPa with a Poisson’s ratio ν = 0.3 which expresses the material’s lateral contraction as fraction of the axial extension. For plastic or non-elastic simulations, the variables of
are functions of e.g. force or displacement.
BC are imposed in FE by prescribed values for entries of and
. The non-prescribed entries of
and
are determined by the FE solver (present study: Abaqus 6.13 [55]) with solutions respecting the equilibrium conditions of forces in Eq (5) and of moments in Eq (6) [56]. Both equations adopt the Lagrangian reference frame [57]. (A comparison to the Eulerian reference frame and its advantages for the modelling of fluids is available in [58].)
(5)
(6)
The surface traction vector is obtained as the product of the Cauchy stress tensor
and the unit outward normal
. (A detailed discussion of
is available in today’s textbooks [59].) In the mechanical simulations of the present study, body force per volume
is neglected; gravitational forces or magnetic forces being typical examples for
. The point vector
specifies the location of a point relative to the origin; i.e. also the equilibrium of moments in Eq (6) being taken about the origin.
Beam elements.
One Euler-Bernoulli beam element (B33) and two Timoshenko beam elements, one with linear interpolation (B31) and one with quadratic interpolation (B32), are implemented for the present study. Table 3 contains the complete list of implemented Abaqus elements. The neglected shear strain of the Euler-Bernoulli beam and the advanced Timoshenko beam are further discussed in [41] for the simplified 2D case. Literature recommends in general for beam assemblies which include short beam elements the Timoshenko beam [28, 60–62]. For this case, [62] has documented an overestimation of the structural stiffness by Euler-Bernoulli beam elements.
Inter-fibre joint models.
Model A is simulated for rigid inter-fibre joints. Torsional spring elements (CONN3D2) are inserted into the medial axis models in Model B between intersecting fibre paths (see Fig 2). This concept for adaptive joint strength of Model B is proposed for comparable cases in publications such as [63]. In the present study, the inter-fibre joint stiffness Kjoint is defined:
(7)
The scaling factor s [m] in Eq (7) allows the directed variation of Kjoint for an approximation of the experimental values of [36]. The cross-sectional area Afibre and Efibre are included as they relate the modelled value of Kjoint to the fibre dimensions and to the fibre stiffness. Alternative model variables could be chosen too.
Kinematic uniform BC and load cases.
In [42], a simulation set of kinematic uniform BC (KUBC), static uniform BC (SUBC), and periodic BC (PBC) was proposed for a random heterogeneous composite. KUBC are applied in the present study as defined in Eq (8). In the case of KUBC, a macroscopic strain tensor imposes the displacement vector
on all
located on ∂V:
(8)
Six independent load cases are implemented in the present study by KUBC for the network samples: i = 1 to 3 for the simulation of tensile tests along the axes x, y, z and i = 4 to 6 for the corresponding shear tests. The combination of all six load cases leads to the symmetrical stiffness matrix , where the six entries Cii on the main diagonal stand for the Young’s moduli E and Shear moduli G [64, 65]:
(9)
The general 4th rank stiffness tensor contains in total 81 components. The simplification to only 21 independent variables in is achieved by energy considerations and symmetry [66].
Mixed BC.
The application of further BC to random fibre networks requires additional considerations [29]. In the case of SUBC, a macroscopic traction vector is imposed on ∂V. PBC add a periodic fluctuation term
. In [29], it is discussed for random fibre networks that considering:
(10)
can’t be prescribed on ∂Vvoid, only on ∂Vfibre. Due to the randomness of the fibre network geometry, the equilibrium conditions in Eqs (5) and (6) are not fulfilled when
is imposed. For overcoming this, mixed BC (MBC) are proposed by [29] and adopted in the present study for the simulation of tensile testing along the axes x and y by one independent load case each (load case i = 1 and 2, definition see above). A macroscopic strain tensor
imposes the primary displacement
along the axis of tension only:
(11)
Along the two non-prescribed axes, a secondary displacement is caused as a consequence of the imposed
. The total deformation of the sample is obtained under MBC as:
.
The results obtained in [29] for the elastic modulus of computer generated random fibre networks without preferred orientation direction predict greater stiffness for KUBC: CMBC < CKUBC.
BC-depth.
The present study extends the BC model of Eqs (8) and (11) for the investigated material by the variable of hBC. The value of hBC prescribes to which depth into the material along the inward normal BC are imposed on ∂V. The reason for the model modification is the obtained structural response which will be discussed further in the following section by Model A in Figs 3 and 4.
Obtained ET values of samples 316L-10%-No.1, and 316L-20%-No.1 depending on hBC.
Obtained ux [μm] for sample 316L-20%-No.1 with hBC = 7.75 or 77.50 μm in load case i = 1 (tensile-x).
Representative volume element determination
For the determination of the RVE size VRVE, the present study relied on an algorithm proposed by [42]. This algorithm has been applied since its publication to random fibre networks in [29, 32] and other materials [43–46].
In [42], VRVE is obtained as a probability estimate. The RVE is understood as the minimum cube volume VRVE for which a particular physical property Z after n realisations can be determined within the margin of the relative error ϵrel. Due to this definition, VRVE is in each case a function of these three variables (see Eq (20)).
In order to acquire the input for the RVE algorithm, the available material sample is split into sets of sub-samples of Vk (with ) of nk realisations. The size function of the RVE is then determined by a regression analysis for the standard deviation DZ(Vk) of Z over Vk. In [45], the mean for Vk of a physical property Z(Vk) and the respective variation
are calculated as follows:
(12)
(13)
Using the expressions gained in Eqs (12) and (13), the absolute sampling error ϵabs and ϵrel are defined for this algorithm, following [29, 42, 67]:
(14)
(15)
V is linked to the variation and the point variance
[29, 42, 68]:
(16)
(17)
Zfibre takes in the present study the value of the respective physical property in Vfibre, Zvoid of the property in Vvoid (with Evoid = 0). Eq (16) requires from the sample volume V [29]:
(18)
The term integral range A3 [69–72] provides a mathematical measure for the characterisation of random structures.
Inserting Eq (14) into Eq (16) yields Eq (19) which provides for Z the required relation between V, ϵabs, and n [29, 42]:
(19)
Eq (19) can be transformed into the final RVE-formula. Now, V stands for the sought variable VRVE [29, 42]:
(20)
In Eq (20), only two unknown variables remain whose values need to be determined for the calculation of VRVE, i.e. A3 and the exponential RVE-factor α. Linear regression (y = ax + b) of the log-transformed relation given in Eq (16) produces these two values [29, 42]:
(21)
Results and Discussion
FE modelling of experimental in-plane Young’s modulus
The transverse Young’s modulus ET () is obtained through the in-plane load cases i = 1 and 2 by Model A and B. The experimental results of [36] are approximated by Model B.
Model A—Rigid joint model.
The value of hBC has a non-negligible influence on the obtained ET in particular for hBC < 77.50 μm (see Fig 3 for the exemplary values of sample 316L-10%-No.1, and 316L-20%-No.1). Whether hBC still changes ET for hBC ≥ 77.50 μm depends on the imposed BC type. The deformation plots demonstrate for MBC that for low hBC the applied BC cause a mechanical response almost exclusively in the close proximity to the two constrained cube faces Fx0 and Fx1 (see Fig 4). For greater hBC, a nearly linear displacement increase through the sample along the axis of tension can be observed in analogy to the increased ET in Fig 3.
The change from linear to quadratic interpolation decreases the ET obtained by Timoshenko beams only marginally (see Tables 4 and 5). ET increases considerably when Euler-Bernoulli beams are simulated:
(22)
The reduced number of prescribed degrees of freedom (DOF) in the case of MBC reduces also the value of ET when compared to KUBC:
(23)
The findings in Eqs (22) and (23) hold true irrespectively of the hBC value and confirm the prediction of [62] and [29] respectively.
Model B—Spring joint model.
ET is obtained by Model B for values of s defined in Table 2 with hBC = 77.50 μm. This hBC value is chosen because of the obtained ET magnitude under MBC (similar to the experimental values of [36]) and because of the considerably reduced dET/dhBC at this value. As required, results for Model B tend towards Model A for greater s independently of the applied BC (see Fig 5):
(24)
Obtained ET values of Model A and Model B depending on s.
Under MBC with only two constrained cube faces, ET results match better the experimentally obtained stiffness values of in-plane tensile testing of [36]. The best match of Model B under MBC is obtained for s = 5 μm (see Fig 6) which is used for the following analyses.
Obtained ET values of Model B depending on f and experimental in-plane tensile testing results of [36].
Stiffness matrix and transverse isotropy
The dominantly in-plane fibre orientation of the network samples (see [36, 73] for detailed analyses) influences the mechanical behaviour to a great extent when the remaining load cases i = 3 to 6 (see Fig 7 for sample 316L-20%-No.2) are obtained. It can be described as transversely isotropic.
Obtained u [μm] for sample 316L-20%-No.2 in load cases i = 3 to 6.
Eqs (25)–(27) present the elastic moduli together with the confidence intervals [Cij ± 2DCij] under KUBC for B32 of each f. All non-bold entries would be expected to equal zero. However, they only nearly vanish which is caused by the numerical artefacts. C11 deviates from C22 by a maximum error of 6.13% for f = 10%. The conditions that C44 ≈ C55 and C13 ≈ C23 show the biggest errors for f = 15% (5.67% and 4.80% respectively). The obtained error for varies between 3.20% for f = 10% and 11.27% for f = 20%.
(25)
(26)
(27)
Size effect and prediction of RVE size
The sample set of Vk and respective nk in Table 6 with k between 1 and 6 was extracted randomly for each f from the network samples. nk replicates the scheme found in [45]. The original publication [29] relates to the measurement error which would be rather difficult to quantify for the presented meshed fibres network systems.
ET shows for the two BC types two different size effects, decreasing for KUBC and increasing for MBC (see Fig 8). These two patterns can be found similarly in [29], there in particular for the bulk modulus under MBC, and for KUBC in [32].
Obtained ET values of Model B depending on V1/3 (error bar = ±DET(V1/3)).
The application of the regression algorithm in Eq (21) (with Z = ET) leads to the values of Table 7. Greater f reduces VRVE,ET. The influence of the BC type on the material is found to be as predicted by [29]:
(28)
For f = 10% under MBC, the condition of Eq (18) [29] isn’t fulfilled any more. The precision of the values for this particular simulation point could be investigated further in future work by greater sample sets.
Prediction of fibre deformation mechanism
The fibre segment deformation inside the network sample by deflection w* (perpendicular to the fibre segment axis) is compared to the elongation ΔΛ (along the axis) by the median of their ratio under KUBC (see Fig 9). For greater simplicity, load cases i = 1 and 2 are summarised to ϵT, i = 4 and 5 to the out-of-plane shear deformation ϵshear-z. Load case i = 3 stands for ϵtensile-z and i = 6 for ϵshear-xy.
Obtained median of [w*/ΔΛ] depending on f.
Increased f has been shown to reduce the fibre segment length λ (see Table 1 and [36]) which reduces simultaneously in the present study the obtained median [w*/ΔΛ]. The obtained median ratio is in each case greater than 2, increasing for shear and out-of-plane deformation:
(29)
Conclusions and Future Work
The present study simulates the elastic mechanics of metallic fibre networks by two FE models (rigid inter-fibre joints or torsional spring connectors) and validates the results by experimentally obtained values from [36]. While [29, 32] used solid volume meshes, beam theory is shown to offer a valid simplification method for the investigated material.
BC types MBC and KUBC are simulated and a new model parameter hBC is introduced for the depth to which BC are prescribed into the material along on ∂V. MBC with fewer prescribed DOF provide a greater match to the experimental values of in-plane tensile testing than KUBC do. The results for rigid inter-fibre joints and for torsional spring connectors converge incrementally. During this process, the influence of the spring constant on the overall structural stiffness decreases. This implies for the manufacturing process that the stiffness of inter-fibre joints becomes increasingly negligible as soon as a minimum joint strength has been achieved in the sintering process.
In [29], an isotropic material behaviour was obtained for computer generated fibre networks without preferred fibre orientation direction. The material of the present study exhibits a dominantly in-plane fibre orientation [36] and an in-plane transversely isotropic mechanical behaviour. Depending on the applied BC type, a decreasing or increasing size effect is observed. Fibre segment deflection dominates the deformation inside the network over fibre segment elongation. These three findings imply for future work and future design studies that size and shape of the material are of importance, as well as the manufactured fibre orientation.
Of interest for future work is also the achievable strain magnitude in a matrix material located in the network’s void phase and the material’s suitability for mechanical bone growth stimulation [74].
Acknowledgments
Dr Garth Wells’ advice about Timoshenko beam types, joints, and boundary conditions is acknowledged. Dr Arul Britto has to be mentioned for sharing his vast experience with Abaqus. This research was supported by the European Research Council Grant No 240446.
Author Contributions
Conceived and designed the experiments: WAB. Performed the experiments: WAB. Analyzed the data: WAB. Contributed reagents/materials/analysis tools: WAB. Wrote the paper: WAB.
References
- 1. Ducheyne P, Aernoudt E, De Meester P. The mechanical behaviour of porous austenitic stainless steel fibre structures. Journal of Materials Science. 1978;13(12):2650–2658.
- 2. Cox H. The elasticity and strength of paper and other fibrous materials. British Journal of Applied Physics. 1951;3(3):72–79.
- 3. Rigdahl M, Andersson H, Westerlind B, Hollmark H. Elastic Behaviour of Low Density Paper Described by Network Mechanics. Fibre Science and Technology. 1983;19(2):127–144.
- 4. Rigdahl M, Westerlind B, Hollmark H. Analysis of cellulose networks by the finite element method. Journal of Materials Science. 1984;19(12):3945–3952.
- 5. Alava M, Niskanen K. The physics of paper. Reports on Progress in Physics. 2006 mar;69(3):669–723.
- 6. Alkhagen M, Toll S. The Effect of Fiber Diameter Distribution on the Elasticity of a Fiber Mass. Journal of Applied Mechanics. 2009;76(4):041014.
- 7. Head DA, Levine AJ, MacKintosh FC. Mechanical response of semiflexible networks to localized perturbations. Physical Review E. 2005 dec;72(6):061914.
- 8. Heussinger C, Frey E. Role of architecture in the elastic response of semiflexible polymer and fiber networks. Physical Review E. 2007 jan;75(1):011917.
- 9. Bai M, Missel AR, Levine AJ, Klug WS. On the role of the filament length distribution in the mechanics of semiflexible networks. Acta biomaterialia. 2011 may;7(5):2109–2118. pmid:21187172
- 10. Kurniawan NA, Enemark S, Rajagopalan R. The role of structure in the nonlinear mechanics of cross-linked semiflexible polymer networks. The Journal of Chemical Physics. 2012 feb;136(6):065101. pmid:22360221
- 11. Carrillo JMY, MacKintosh FC, Dobrynin AV. Nonlinear Elasticity: From Single Chain to Networks and Gels. Macromolecules. 2013 may;46(9):3679–3692.
- 12.
Mao N, Russell SJ, Pourdeyhimi B. 9 Characterisation, testing and modelling of nonwoven fabrics. In: Russell SJ, editor. Handbook of nonwovens. Cambridge(UK); 2007. p. 401–502.
- 13. Mannarino MM, Rutledge GC. Mechanical and tribological properties of electrospun PA 6(3)T fiber mats. Polymer. 2012;53(14):3017–3025.
- 14. Ridruejo A, González C, LLorca J. A constitutive model for the in-plane mechanical behavior of nonwoven fabrics. International Journal of Solids and Structures. 2012 sep;49(17):2215–2229.
- 15. Silberstein MN, Pai CL, Rutledge GC, Boyce MC. Elastic-plastic behavior of non-woven fibrous mats. Journal of the Mechanics and Physics of Solids. 2012 feb;60(2):295–318.
- 16. Farukh F, Demirci E, Sabuncuoglu B, Acar M, Pourdeyhimi B, Silberschmidt VV. Numerical analysis of progressive damage in nonwoven fibrous networks under tension. International Journal of Solids and Structures. 2014;51(9):1670–1685.
- 17. Farukh F, Demirci E, Sabuncuoglu B, Acar M, Pourdeyhimi B, Silberschmidt VV. Mechanical analysis of bi-component-fibre nonwovens: Finite-element strategy. Composites Part B: Engineering. 2015;68:327–335.
- 18. Kwon RY, Lew AJ, Jacobs CR. A microstructurally informed model for the mechanical response of three-dimensional actin networks. Computer Methods in Biomechanics and Biomedical Engineering. 2008 aug;11(4):407–418. pmid:18568835
- 19. Lieleg O, Schmoller KM, Claessens MMAE, Bausch AR. Cytoskeletal polymer networks: viscoelastic properties are determined by the microscopic interaction potential of cross-links. Biophysical journal. 2009 jun;96(11):4725–4732. pmid:19486695
- 20. Fletcher DA, Mullins RD. Cell mechanics and the cytoskeleton. Nature. 2010 jan;463(7280):485–492. pmid:20110992
- 21. Stricker J, Falzone T, Gardel ML. Mechanics of the F-actin cytoskeleton. Journal of Biomechanics. 2010 jan;43(1):9–14. pmid:19913792
- 22. Yamaoka H, Matsushita S, Shimada Y, Adachi T. Multiscale modeling and mechanics of filamentous actin cytoskeleton. Biomechanics and modeling in mechanobiology. 2012 mar;11(3):291–302. pmid:21614531
- 23. Astrom JA, Makinen JP, Hirvonen H, Timonen J. Stiffness of compressed fiber mats. Journal of Applied Physics. 2000;88(9):5056–5061.
- 24. Head DA, MacKintosh FC, Levine AJ. Nonuniversality of elastic exponents in random bond-bending networks. Physical Review E. 2003 aug;68(2):025101(R).
- 25. DiDonna BA, Lubensky TC. Nonaffine correlations in random elastic media. Physical Review E. 2005 dec;72(6):066619.
- 26. Hatami-Marbini H, Picu RC. Heterogeneous long-range correlated deformation of semiflexible random fiber networks. Physical Review E. 2009 oct;80(4):046703.
- 27. Broedersz CP, Sheinman M, MacKintosh FC. Filament-Length-Controlled Elasticity in 3D Fiber Networks. Physical Review Letters. 2012 feb;108(7):078102. pmid:22401259
- 28. Shahsavari AS, Picu RC. Size effect on mechanical behavior of random fiber networks. International Journal of Solids and Structures. 2013 oct;50(20-21):3332–3338.
- 29. Dirrenberger J, Forest S, Jeulin D. Towards gigantic RVE sizes for 3D stochastic fibrous networks. International Journal of Solids and Structures. 2014 jan;51(2):359–376.
- 30. Jin MZ, Chen CQ, Lu TJ. The mechanical behavior of porous metal fiber sintered sheets. Journal of the Mechanics and Physics of Solids. 2013 jan;61(1):161–174.
- 31. Zhao TF, Jin MZ, Chen CQ. A phenomenological elastoplastic model for porous metal fiber sintered sheets. Materials Science and Engineering: A. 2013 oct;582:188–193.
- 32.
Tsarouchas D. Fibre Network Materials: Architecture & Effective Linear Elastic Properties. University of Cambridge; 2012.
- 33. Yi P, Peng L, Liu N, Lai X, Ni J. A micromechanics elastic-plastic constitutive model for sintered stainless steel fiber felt. Materials & Design. 2013 oct;51:876–885.
- 34.
Rodney D, Gadot B, Martinez OR, du Roscoat SR, Orgéas L. Reversible dilatancy in entangled single-wire materials. Nature Materials. 2015;(September (advance online publication)). Available from: http://www.nature.com/doifinder/10.1038/nmat4429.
- 35.
Bosbach W. The mechanical and magnetic behaviour of sintered fibre networks and their suitability for a therapeutic, biomedical application. University of Cambridge (viva passed on 20 March 2015); 2015.
- 36. Neelakantan S, Bosbach W, Woodhouse J, Markaki AE. Characterization and deformation response of orthotropic fibre networks with auxetic out-of-plane behaviour. Acta Materialia. 2014 mar;66:326–339.
- 37. Wieding J, Souffrant R, Fritsche A, Mittelmeier W, Bader R. Finite Element Analysis of Osteosynthesis Screw Fixation in the Bone Stock: An Appropriate Method for Automatic Screw Modelling. PloS ONE. 2012 jan;7(3):1–10.
- 38.
Eulero L. Methodus inveniendi lineas curvas maximi minimive proprietate gaudentes, sive solutio problematis isoperimetrici latissimo sensu accepti. Lausanne (Switzerland); 1744.
- 39. Timoshenko S. On the Correction for Shear of the Differential Equation for Transverse Vibrations of Prismatic Bars. Philosophical Magazine Series 6. 1921;41(245):744–746.
- 40. Timoshenko S. On the transverse vibrations of bars of uniform cross-section. Philosophical Magazine Series 6. 1922;43(253):125–131.
- 41.
Reddy JN. An Introduction to the Finite Element Method. 3rd ed. College Station, Texas (USA); 2006.
- 42. Kanit T, Forest S, Galliet I, Mounoury V, Jeulin D. Determination of the size of the representative volume element for random composites: statistical and numerical approach. International Journal of Solids and Structures. 2003 jun;40(13-14):3647–3679.
- 43. Kanit T, N’Guyen F, Forest S, Jeulin D, Reed M, Singleton S. Apparent and effective physical properties of heterogeneous materials: Representativity of samples of two materials from food industry. Computer Methods in Applied Mechanics and Engineering. 2006 jul;195(33-36):3960–3982.
- 44. Brun E, Vicente J, Topin F, Occelli R, Clifton MJ. Microstructure and Transport Properties of Cellular Materials: Representative Volume Element. Advanced Engineering Materials. 2009 oct;11(10):805–810.
- 45. Grimal Q, Raum K, Gerisch A, Laugier P. A determination of the minimum sizes of representative volume elements for the prediction of cortical bone elastic properties. Biomechanics and modeling in mechanobiology. 2011 dec;10(6):925–937. pmid:21267625
- 46. Kanit T, Forest S, Jeulin D, N’Guyen F, Singleton S. Virtual improvement of ice cream properties by computational homogenization of microstructures. Mechanics Research Communications. 2011 mar;38(2):136–140.
- 47.
Röntgen WC. Ueber eine neue Art von Strahlen (Vorläufige Mittheilung). In: Sonderabbdruck der Sitzungsberichte der Würzburger Physik.-medic. Gesellschaft. Würzburg (Germany); 1895.
- 48.
Buzug M. Computed Tomography—From Photon Statistics to Modern Cone-Beam CT. Luebeck (Germany); 2008.
- 49.
Francois R. US-Patent 5,071,713; 1991.
- 50. Lee TC, Kashyap RL, Chong-Nam C. Building Skeleton Models via 3-D Medial Surface/Axis Thinning Algorithms. Graphical Models and Image Processing. 1994;56(6):462–478.
- 51.
Lindquist WB. 3DMA General Users Manual. Stony Brook (USA); 1999.
- 52.
Yang H. A Geometric and Statistical Analysis of Fibrous Materials from Three-Dimensional High Resolution Images. State University of New York at Stony Brook; 2001.
- 53.
Desai CS, Abel JF. Introduction to the Finite Element Method—A Numerical Method for Engineering Analysis. New York (USA); 1972.
- 54.
Hooke R. Lectures de Potentia Restitutiva, or of Spring—Explaining the Power of Springing Bodies. London (UK); 1678.
- 55.
Dassault Systèmes. Abaqus 6.13 Online Documentation; 2013. Available from: http://129.97.46.200:2080/v6.13/.
- 56.
Dassault Systèmes. Abaqus Theory Guide 1.5.1 Equilibrium and virtual work; 2013. Available from: http://129.97.46.200:2080/v6.13/books/stm/default.htm?startat=ch01s05ath08.html.
- 57.
Logan JD. Applied Mathematics. 6th ed. Lincoln, Nebraska (USA); 2006.
- 58. Atzberger PJ. Stochastic Eulerian Lagrangian methods for fluid-structure interactions with thermal fluctuations. Journal of Computational Physics. 2011;230(8):2821–2837.
- 59.
Gross D, Hauger W, Schröder J, Wall WA, Bonet J. Engineering Mechanics 2: Mechanics of Materials. Heidelberg (Germany); 2011.
- 60. Shahsavari A, Picu RC. Model selection for athermal cross-linked fiber networks. Physical Review E. 2012 jul;86(1):011923.
- 61.
Dassault Systèmes. Abaqus Analysis User’s Guide 29.3.3 Choosing a beam element; 2013. Available from: http://129.97.46.200:2080/v6.13/books/usb/default.htm?startat=pt06ch29s03alm08.html.
- 62. Wang CW, Berhan L, Sastry AM. Structure, Mechanics and Failure of Stochastic Fibrous Networks—Part I Microscale Considerations. Journal of Engineering Materials and Technology. 2000;122(4):450–459.
- 63. Berhan L, Yi YB, Sastry AM, Munoz E, Selvidge M, Baughman R. Mechanical properties of nanotube sheets: Alterations in joint morphology and achievable moduli in manufacturable materials. Journal of Applied Physics. 2004;95(8):4335–4345.
- 64.
Kaw AK. Mechanics of Composite Materials. 2nd ed. Boca Raton, Florida (USA); 2006.
- 65.
Zohdi TI, Wriggers P. An Introduction to Computational Micromechanics. In: Pfeiffer F, Wriggers P, editors. Lecture Notes in Applied and Computational Mechanics. 1st ed. Hannover (Germany); 2005.
- 66.
Christensen RM. 1.1 Elasticity Theory Results. In: Mechanics of Composite Materials. revised ed. Mineola, New York (USA); 2005.
- 67.
Matheron G. Estimating and Choosing. Paris (France); 1989.
- 68.
Jeulin D. Variance scaling of Boolean random varieties. HAL: hal-00618967, version 1. 2011;p. 1–16.
- 69.
Matheron G. The Theory of Regionalized Variables and its Applications. In: Les Cahiers du Centre de Morphologie Mathematique de Fontainebleau. vol. 5. Paris (France); 1971. p. 1–211.
- 70.
Matheron G. Random Sets and Integral Geometry. Paris (France); 1975.
- 71. Lantuejoul C. Ergodicity and integral range. Journal of Microscopy. 1991;161(3):387–403.
- 72. Cailletaud G, Jeulin D, Rolland P. Size Effect on Elastic Properties of Random Composites. Engineering Computations. 1994;11(2):99–110.
- 73. Spear RL, Srigengan B, Neelakantan S, Bosbach W, Brooks RA, Markaki AE. Physical and Biological Characterization of Ferromagnetic Fiber Networks: Effect of Fibrin Deposition on Short-Term In Vitro Responses of Human Osteoblasts. Tissue Engineering Part A. 2015;21(3-4):463–474. pmid:25145466
- 74.
Bosbach W. Finite element modelling of bone growth stimulation and its suitability for therapeutic, biomedical applications. In: von-Behring-Röntgen-Symposium. Giessen (Germany); 2015. p. 1–2. Available from: https://www.repository.cam.ac.uk/handle/1810/251137?show=full.