Abstract
The surface morphology of faults controls the spatial anisotropy of their frictional properties and hence their mechanical stability. Such anisotropy is only rarely studied in seismology models of fault slip, although it might be paramount to understand the seismic rupture in particular areas, notably where slip occurs in a direction different from that of the main striations of the fault. To quantify how the anisotropy of fault surfaces affects the friction coefficient during sliding, we sheared synthetic fault planes made of plaster of Paris. These fault planes were produced by 3D-printing real striated fault surfaces whose 3D roughness was measured in the field at spatial scales from millimeters to meters. Here, we show how the 3D-printing technology can help for the study of frictional slip. The results show that fault anisotropy controls the coefficient of static friction, with , the friction coefficient along the striations being three to four times smaller than , the friction coefficient along the orientation perpendicular to the striations. This is true both at the meter and the millimeter scales. The anisotropy in friction and the average coefficient of static friction are also shown to decrease with the normal stress applied to the faults, as a result of the increased surface wear under increased loading.
1 Introduction
Faults in the Earth’s crust are complex systems along which earthquakes nucleate and propagate [e.g., Wibberley et al. (2008)]. Faults hold structures and heterogeneities at all scales [, , ]. While they are often simplified to their simplest two-dimensional description (i.e., the fault plane), increasing complexity is now added to faults models [e.g., ]. It is indeed considered that, to fully understand seismicity in various areas, it is paramount to account for some disorder in the faults frictional properties such as secondary faulting, off-fault damage or roughness of the fault plane [, , , , Tesei et al. (2014), , , ]. For instance, the volume of damaged rocks, during the activation of a fault, depends on the initial contact roughness [] and, thus, a fault with a stronger roughness presents a different energy budget than a flat fault, as more energy is converted into surface area energy. In particular, roughness encourages the triggering of local events, but is believed to prevent the propagation of large-slip earthquakes []. Additionally, large scale roughness tends to inhibit the propagation of any rupture faster than the shear wave velocity of surrounding rocks [].
Another degree of complexity is more rarely considered when modeling geological contacts and fault slip; the possible anisotropy in their frictional properties. Morphological anisotropy is a known feature of faults, notably impacting the seismic waves velocity in their vicinity [, ; ] or the mobility of natural and injected fluids [Vadacca et al. (2018)] in the subsurface. Frictional anisotropy, interestingly, is also regularly studied in other fields than seismology, for instance the tribology of rubber tires [, Tiwari et al. (2016)], the strength of advanced adhesives , or the mitigation of water condensation . It is also considered to play a major role in nature , for instance in the motion of numerous animals [Zheng et al. (2007), , ] and the hydration of some plants [, Wang F. et al. (2017)]. In most cases, frictional anisotropy derives from the existence of preferential topographical orientations on, at least, one of the contact surfaces [Yu and Wang (2012), Tapia et al. (2016)]. The length scale for such structural directivity can be as small as micrometer [Stupkiewicz et al. (2014)] to nanometer [, ].
In seismic faults, such preferential orientations in their topography are observed at all scales [, , ] and originate from several processes. At the molecular level, rock forming crystals may display some frictional anisotropy. It is notably the case for antigorite, a mineral abundant in the Earth’s upper mantle []. At the mesoscopic scale, the shear strength of foliated rocks is known to be anisotropic, due to the oriented planes in their constitutive mineralogy [, ]. Fault zones in sedimentary basins are initiated by early fractures that often propagate in layered sediments. It can result in an anisotropic ramp-flat morphology of these fracture surfaces []. For more mature faults having accumulated enough displacement, and above a given length scale [], the topography of the fault planes is also marked by slip induced wear, with striations and grooves of various wavelengths and amplitudes oriented along the main direction of slip [; ]. If such morphological anisotropy of fault surfaces is well-known, its effect on the anisotropy of the frictional properties remains to be characterized. Such a characterization of frictional anisotropy could also be of interest for other types of rock contacts than strictly seismic faults, in particular for shallow rock joints and fractures, whose three-dimensional geometry is key in geotechnical engineering and for the structural stability of many man-made constructions [, , , ]. Here, we study how the morphology of faults controls the static coefficient of friction and the anisotropy of friction with regards to the main stress orientation during slip. To reach this goal, we produce 3D-prints of actual faults surfaces whose topography was measured in the field []. We perform friction experiments with plaster of Paris casts of these 3D-printed faults. Results show that the coefficient of static friction along faults is highly anisotropic, a property that should henceforward be considered in numerical models of slip on seismic faults. We also show that this anisotropy is stress-dependent and should decrease with depth [e.g., ].
2 3D Printing and Plaster Casting of Fault Planes
The actual morphology of natural faults can be difficult to assess, even if their long wavelength structures can be inferred by surface or subsurface imaging techniques [, , ]. Yet, some fault planes are accessible to direct high resolution measurements, notably as they were exhumed by erosion and tectonic processes. For this study, we have used a series of digital fault surfaces. These fault roughness data were acquired with Light Detection and Ranging (LiDAR), laboratory laser profilometry, or white light interferometry techniques []. These data are available on an online public database [] and in a repository with a DOI number []. Should the reader hold some similar data, these authors welcome additions to this database. We have specifically selected fault roughness measurements performed on the Corona Heights fault [] that outcrops near the Peixotto playground in San Francisco, California. These data cover surface areas with spatial scales in the range of millimeters to meters. Figures 1, 2 show the fault surface at two spatial scales, one surface at the meter scale, defined on a 5 mm × 5 mm grid, and one surface at the millimeter scale, defined on a 2 µm × 2 µm grid. We will further refer to these two surfaces as, respectively, and . Already, one can notice some preferential orientations in these topographies, and that the amplitude of fault roughness is, relatively to their size, somewhat larger at smaller scales () than at larger scales () [].
FIGURE 1
FIGURE 2
For our tests, we chose to limit these anisotropic surfaces to a circular sample geometry. We also applied a mild running-window median filter to smooth out spikes in the measured surfaces that could be associated to measurement noise. The window length of the filter was 10 space steps, accounting for 5 cm for and for 20 µm for . In order to run the friction experiments, we generated some opposing surfaces to the ones presented in Figures 1, 2. These opposing surfaces could not be measured, as the actual fault walls that were facing and are now eroded. To reconstruct them, we have applied the following transformation to the 3D coordinates of and :where , , and are the coordinates of the generated opposing surfaces and give the map location of a given surface point of elevation Z, as represented in Figure 1. We have thus assumed that the missing fault walls are complementary to the measured ones, so that, when pressed together before the friction tests, they form a bulk with negligible aperture between the two blocks. Such assumption for natural faults would only be partly verified. When having accumulated enough slip, a granular layer of gouge material may there have formed, and the two opposing sliding surfaces may not always perfectly match. However, our assumption is relevant for the youngest faults with a small amount of slip. We have also assumed that erosion did not significantly alter the fault plane, such that the measured topography is representative of the one of an actual buried fault. For the Corona Heights fault, this assumption is valid because the fault offsets silica-rich chert rocks with a high resistance to weathering.
After having obtained the surfaces, we isotropically (i.e., with the same factor in all directions) down- or up-scaled and to fit a standard 4 cm diameter disk that matches the clamp size of our shear deformation apparatus. We also regridded the surfaces to match the lateral resolution of our 3D printer {Ultimaker2 Extended+ (Ultimaker2 Extended+, 2015)} that has a nozzle size of 250 µm. The four surfaces (two fault surfaces and two opposing surfaces) were then 3D-printed into polylactic acid (PLA) material, as shown in Figure 3. It should be noted that, even when designed to be flat, printed objects can present a natural roughness [], at a scale however smaller that the grooves observed on the printed faults. These intrinsic imperfections shall be comparable to 60 µm, the elementary thickness of the PLA layers deposited by our 3D printer. In comparison, the 2D standard deviation of the elevation in our printed objects topography are 0.66 mm for and 1.7 mm for . The maximal elevation of these objects is, respectively, 3.7 and 8.3 mm. We thus consider that the small scale roughness () from the printer’s limit in resolution has a second order effect on the frictional properties of the surfaces.
Although we could have performed the friction experiments with the plastic pieces produced with the 3D printer, we have rather produced samples of plaster of Paris (gypsum) blocks molded from the plastic faults. Plaster is known to be a reasonable model of porous brittle materials [Vekinis et al. (1993)], and the main goal of these casts was to work with a rock-like material, notably because plaster may wear and deform differently than plastic under shear. The fragile nature of plaster, and the potential friction-induced wear that the plaster was subjected to in our experiments, made us use new casts for each experimental realization. The casts were generated with the following protocol: five volumes of water and eight volumes of powder of plaster of Paris were mixed and poured over the plastic molds, then let to dry during one and half hour. The molds, an example of which is shown in Figure 3, were sprayed before each cast with a thin layer of silicon grease to avoid some of the fine plaster details to stick to the plastic during the mold release. The last step in the casts preparation was to dry them in an oven at a temperature of for 1 h. As a result, we produced fault planes in blocks made of plaster of Paris, as shown in Figure 4.
FIGURE 3
FIGURE 4
3 Experimental Set-Up and Experimental Conditions
The shear apparatus used to perform the friction tests is shown in Figure 5. The two complementary surfaces are pressed together and mounted one on top of the other between the clamps of the shear apparatus. A normal force is applied on the top surface by using adjustable weights. In addition, a spring system of stiffness 625 N m−1 allows, if desired, to compensate for the machine empty weight of 13.7 N (i.e., the normal weight transmitted to the friction surfaces by the machine top clamp and structure and the top cast when no extra mass is used). A tangential driving shear force is then applied to the top fault wall in a given direction of the plane. The amplitude of the force is measured by a Sauter® force gauge (). The shear orientation is defined by the angle from the orientation of the main grooves on the fault surface, as defined in Figures 1, 2. A horizontal mechanical slider makes sure that the friction is evaluated in the direction of interest only, and a vertical slider allows upward or downward displacement of the top surface. While these sliders would ideally be perfectly lubricated, we have estimated their frictional resistance at the sliding velocity of our experiments, by performing a friction test with no fault installed in the machine (that is, with only air between the two clamps represented in Figure 5).
FIGURE 5
The target speed of the test bench (that is, the demanded slip velocity) was fixed to a constant and equal to 1.3 mm s−1. Of course, such a velocity may be orders of magnitude above that of typical tectonic solicitations (for instance, an ultra-fast oceanic ridge may reach an opening rate of 10–20 cm yr−1 []. Here, we define the static friction in the experiments as the peak shear stress reached before sliding occurs divided by the normal stress.
We characterize the anisotropy of this laboratory static friction coefficient for both the and surfaces by performing a series of experiments where we vary the angle of loading with respect to the grooves. We ran friction tests every on both plaster faults. At each angle, the experiment was repeated at least three times in order to ensure that the results are reproducible (with new casts each time, to avoid any wear-related deviation). The standard deviation computed on these multiple measurements (typically 5–10 N) was used to compute the error bars on our characteristics coefficients of static friction. was sheared under a normal stress , while the tests performed on the rougher surface were performed under . A total of 76 experiments were performed, 37 using and 39 using . At the onset of slip, the laboratory static friction coefficient is defined using a standard Coulomb’s law []:where is the coefficient of static friction and is the (local) maximum tangential force applied at the onset of slip. In the following, we will also consider the mean driving and normal stresses, denoted and , where is the radius of the cast. Figure 6 shows a typical measurement of a friction test, from which and hence are calculated.
FIGURE 6
4 Results
4.1 Friction Anisotropy
The results are presented in Figure 7. The derived coefficient of frictions are larger than 1, which does not come as a surprise due to the nonnegligible roughness of our fault samples. Indeed, a large part of the contact area is bound to be perpendicular to the demanded slip (in any direction), inducing a strong resistance to motion.
FIGURE 7
On both fault samples, one can observe the strong anisotropy of the coefficient of static friction, with the maximum value of being about four times larger than its minimum for and about three times larger for . In both sets of experiments, the minimal friction is obtained along the main groove orientation (i.e., at or ), and the resistance to shear is larger perpendicularly to this orientation. The maximum of value is however never obtained exactly at or but rather along a neighboring direction. Local maxima are indeed obtained for or when shearing and for or when shearing .
4.2 Damage and Stress Dependence
Most of our experiments were destructive, with visible wear on the plaster samples after the shearing tests. This wear was the main reason calling for the production of new plaster casts for each experimental realization, as we verified that repeating a same experiment with a previously used cast led to a significant (and here unwanted) drop in friction. The observed damage consists either in the formation of plaster powder (gouge) or in the rupture of topographic highs of the fault surfaces. Part of it might have initiated at the onset of the fault displacement (and hence be related to the static friction), while some of it has rather been induced by the subsequent sliding. Figure 8 shows some examples of these damage types.
FIGURE 8
Wear of seismic faults has been studied [e.g., Wang and Scholz (1994); ] to, in particular, better understand the energy budget of the deformation, but also because this process may lubricate faults during slip [e.g., , ] or modify the fault permeability to fluid flow [e.g., , ]. The present study focuses on the measurement of the coefficient of static friction and on its anisotropy, but we suggest that our 3D-print–based set-up could also enable the quantitative characterization of damage during sliding along analog fault surfaces. We here keep to a qualitative assessment of which parts of the surfaces were mainly worn during each experiment. It seems that most of the shear resistance of the Corona Heights fault, at the millimeter scale (), arises from its grooves. By contrast, the friction of (representing a metric scale) is dominated by one chip in its field-scanned morphology (i.e., the main topographic low in Figure 1). This chip being on the edge of the 3D-printed surface, but not on the edge of the real word fault, a finite size effect (inducing an artificial asymmetry of the fault wall) is certainly at play in the results reported in Figure 7 (top), notably explaining the strong asymmetry in for the opposite directions and . Indeed, due to the finite size of our tested pieces, the interlocking of the chip is relatively free to unlock laterally along the direction, but is strongly locked by the surrounding plaster along the direction (see Figure 9).
FIGURE 9
Because the overall friction is likely to be affected by the surface wear, and because this wear is likely stress-dependent, we have performed some friction tests on under various loads , at a given angle . The results are shown in Figure 10. At the highest tested stresses, seems, to an extent, stress-independent, with its mean variations lying within the measured error bars. While this result is compatible with the classical Coulomb theory [e.g., ], one can observe, over a wider range of normal stresses, a consistent decrease in the friction coefficient with a higher normal stress. It could, in part, emanate from some limitations in our experimental set-up. For instance, at lower , the internal friction of the device () accounts for a more significant portion of the total measured tangential force, and the friction characterization could thus be less accurate. We have however run similar tests on flat plaster surfaces, showing no significant variations in over the same stress range (see Figure 10). The drop in friction coefficient with is then likely related to the increased damage under higher normal loading (see Figure 8), reducing the overall shear resistance as asperities are easier to break. It might also result from the change in effective contact area with a higher load, with proportionally more pressure being borne by the surfaces parallel to the demanded motion, causing proportionally less of a resistance to slip. Such a drop of with the normal stress has already been reported for tilted contacts [], as the static coefficient of friction is not an absolute material constant [, ].
FIGURE 10
We have therefore assessed the effect of the normal stress on the anisotropy of the static friction coefficient. We performed frictional tests on the four poles of () at a higher load () than the load used before (i.e., , as reported in Figure 7). The newly measured coefficients of static friction are shown in Figure 11. One can notice the reduced friction anisotropy at high . The ratio between the maximum and the minimum value of indeed drops from 3.1 at to 1.5 at . This result suggests that the frictional anisotropy of faults is smaller at depth. This concept is naturally linked with the seminal ’s law, stating that the roughness of fault planes (and the type of their constitutive rocks) has less effect on their maximum static frictional properties at larger depths.
FIGURE 11
5 Discussion and Conclusion
Here, we have shown how the multiscale anisotropy of fault plane topography leads to an anisotropy in the frictional properties. Results confirm that seismic faults are prone to slide along some preferential orientations. The orientation that is the most likely is the one that faults have previously slid along and which has shaped some guiding grooves in their morphology. Yet, displacements following other orientations are possible. Predicting the rupture direction of the next earthquake on a fault is thus not only dependent on assessing the main regional stress. The question should rather be along which orientation a rupture criterion [e.g., ] will first be exceeded. Such a subtlety might be of little importance for mature faults for which the stress principle orientations have not changed with time, because in this case, the main stress is likely to act along the lowest coefficient of friction anyway. Yet, it could be paramount for faults under a changing geological load, where this alignment is not verified, or for immature faults, where the slip could be mainly governed by the anisotropy of early surfaces (i.e., where the slip does not coincide with the stress principal orientation, but is nonassociated). Examples of slickensides (i.e., fault planes) commonly exist with several overlapping striations orientations [e.g., ], with rake and striations oblique to the actual orientation of the fault plane [e.g., ]. These observations indicate that the original slip direction (if assumed to initiate the following Andersonian criterion) [] does not completely determine the direction of the next episode. Earthquakes occurring along abnormal directions (i.e., not in agreement with the local stress state) have been observed [e.g., , , ], and their understanding might be eased by accounting for the possible frictional anisotropy of their surfaces [, ].
Note that frictional anisotropy should not only be considered at the full fault scale, but this property may spatially vary along the fault walls. Analytical solutions demonstrate that the stress around a fault is perturbed by its roughness, and a local slip can occur much before the entire fault is under yielding conditions in a given direction []. While we have here only measured the static coefficient of friction, we suggest that similar studies could be performed to characterize the anisotropy of the coefficient of dynamic friction (i.e., by analyzing the evolution of the resistance to motion, after the plaster faults start moving, as a function of the sliding orientation). Hence, not only the initial slip direction of an earthquake could be impacted by frictional anisotropy, but the complete slip trajectory [Tapia et al. (2016)]. Changes in the slip direction within single earthquake rupture events are indeed sometimes observed, notably from bent grooves on postmortem fault walls []. We have, additionally, measured how the anisotropy in friction becomes less significant when the normal stress acting on a fault increases (i.e., with the fault depth), in general agreement with Byerlee’s law []. Such an effect likely derives from the stress-related changes in rupture rheology and in damage type. The transition from a highly anisotropic to a relatively isotropic regime should typically occur when local stresses on the fault reach the yield strength of the material, MPa in the case of plaster [Vekinis et al. (1993)]. This is about two orders of magnitude above the transition at which we observed a strong reduction in anisotropy (see Figure 11), but our computed is an average value which does not account for the potential strong stress concentrations at play in our faults. Considering that the strength of rocks [e.g., ] is about two orders of magnitude (100 times) larger than that of plaster, fault frictional anisotropy could thus be only at play at pressures less than about . This would correspond to the shallowest faults, at depths less than , where ρ is the volumetric mass of rocks () and g the gravity acceleration. Some care should however be taken when deriving such a conclusion by analyzing resized samples as ours ( diameter samples representing meter or millimeter topographies), as the way matter breaks is length-scale-dependent [e.g., ]. Note also that other fault geometries than the one we have here studied may induce lesser stress concentration, so that a significant damage only occurs at a mean stress level directly comparable to the yield stress of rocks. Thus, the possibility of frictional anisotropy should not be overlooked when studying fault buried up to . The heterogeneity of fault planes, and thus the anisotropy in this heterogeneity, may also still play a role under high stress, as roughness does not only encourage local yield, but also helps to suppress large slip events on moving faults [].
Additionally to the assessment of the stability of (at least) shallow seismic faults, the characterization of the frictional anisotropy of rock surfaces may be of importance in geotechnical engineering, for instance, for the stability of tunnels and foundations. There, the intrinsic strength anisotropy of foliated rocks is well-studied [, ]. Our work shows how one can also characterize the mechanical anisotropy of rough rock contacts, for instance, along joints [, , , ] and fractures [, ] between or inside rock formations.
A main point of this manuscript is, finally, to illustrate how the 3D-printing technology can help with new experimental designs in Earth Sciences, and this technology is getting a growing attention from the community [Wang L. et al. (2017), , Wang et al. (2018), , ], including the study of the frictional properties of 3D-printed fault analogs []. A direct continuation of the present work, for instance, could be to 3D-print and to test some faults surfaces beforehand filtered with various band-pass filters, in order to understand how the various wavelengths of the topography contribute to the global static friction coefficient, to the dynamical friction coefficient and to analyze the spatial distribution of the fault wear produced under various stresses and amounts of slip.
Statements
Data availability statement
The original contributions presented in the study are included in the article/Supplementary Material; further inquiries can be directed to the corresponding authors.
Author contributions
RT proposed the guidelines of this work, AS built the test machine and, with TV-D, printed the faults and performed the friction experiments. FR advised on the mechanics of seismic faults. TV-D wrote the first version of this manuscript and all authors contributed to the writing of its final version.
Funding
We acknowledge the support of the University of Strasbourg, of the IRP France–Norway D-FFRACT, and of SFF Porelab (project number 262644 of the Research Council of Norway). Readers are welcome to comment and correspondence should be addressed to tom.vincent-dospital@fys.uio.no. or renaud.toussaint@unistra.fr.
Acknowledgments
We are grateful for the early experimental explorations performed by Marine-Sophie Jacob, Céline Fliedner, Aldo Mellado Aguilar, Gaëtan Leca, and Laifa Rahmi, students from the EOST/IPGS faculty at the University of Strasbourg. We also thank Amir Sagy from the Geological Survey of Israel for fruitful discussions. We thank the Strasbourg AV.Lab association for letting us use their 3D printer.
Conflict of interest
The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.
Supplementary material
The Supplementary Material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/feart.2021.627823/full#supplementary-material
References
1
AllamA. A.KrollK. A.MillinerC. W. D.Richards‐DingerK. B. (2019). Effects of fault roughness on coseismic slip and earthquake locations. J. Geophys. Res. Solid Earth124 (11), 11336–11349. 10.1029/2018JB016216
2
AndoR.TakedaN.YamashitaT. (2012). Propagation dynamics of seismic and aseismic slip governed by fault heterogeneity and Newtonian rheology. J. Geophys. Res.117 (B11), 11308. 10.1029/2012JB009532
3
AochiH.IdeS. (2009). Complexity in earthquake sequences controlled by multiscale heterogeneity in fault fracture energy. J. Geophys. Res.114 (B3). 10.1029/2008JB006034
4
AvouacJ.-P.AyoubF.WeiS.AmpueroJ.-P.MengL.LeprinceS.et al (2014). The 2013, Mw 7.7 Balochistan earthquake, energetic strike-slip reactivation of a thrust fault. Earth Planet. Sci. Lett.391, 128–134. 10.1016/j.epsl.2014.01.036
5
BartonN.ChoubeyV. (1977). The shear strength of rock joints in theory and practice. Rock Mech.10, 1–54. 10.1007/BF01261801
6
Ben-DavidO.FinebergJ. (2011). Static friction coefficient is not a material constant. Phys. Rev. Lett.106, 254301. 10.1103/PhysRevLett.106.254301
7
BlakelyR. J.WellsR. E.WeaverC. S.JohnsonS. Y. (2002). Location, structure, and seismicity of the Seattle fault zone, Washington: evidence from aeromagnetic anomalies, geologic mapping, and seismic-reflection data. GSA Bull.114 (2), 169–177. 10.1130/0016-7606(2002)114¡0169:LSASOT¿2.0.CO;2
8
BlanpiedM. L.LocknerD. A.ByerleeJ. D. (1992). An earthquake mechanism based on rapid sealing of faults. Nature358, 574–576. 10.1038/358574a0
9
BottM. H. P. (1959). The mechanics of oblique slip faulting. Geol. Mag.96 (2), 109–117. 10.1017/S0016756800059987
10
BouchonM.KarabulutH.BouinM.-P.SchmittbuhlJ.ValléeM.ArchuletaR.et al (2010). Faulting characteristics of supershear earthquakes. Tectonophysics493 (3), 244–253. 10.1016/j.tecto.2010.06.011
11
BowdenF. P.TaborD. (1951). The friction and lubrication of solids. Am. J. Phys.19 (7), 428–429. 10.1119/1.1933017
12
BradyB. T.DuvallW. I.HorinoF. G. (1993). An experimental determination of the true uniaxial stress-strain behavior of brittle rock. Rock Mech.5, 107–120. 10.1007/BF01240161
13
BraunP.TzortzopoulosG.StefanouI. (2020). Design of sand-based, 3d-printed analogue faults with controlled frictional properties. J. Geophys. Res. [preprint]. 10.1002/essoar.10503633.1
14
BrodskyE. E.KirkpatrickJ. D.CandelaT. (2016). Constraints from fault roughness on the scale-dependent strength of rocks. Geology44 (1), 19–22. 10.1130/G37206.1
15
BrownS. R.ScholzC. H. (1985). Broad bandwidth study of the topography of natural rock surfaces. J. Geophys. Res.90 (B14), 12575–12582. 10.1029/JB090iB14p12575
16
BruhatL.KlingerY.VallageA.DunhamE. M. (2019). Influence of fault roughness on surface displacement: from numerical simulations to coseismic slip distributions. Geophys. J. Int.220 (3), 1857–1877. 10.1093/gji/ggz545
17
ByerleeJ. (1978). Friction of rocks. Pure Appl. Geophys.116 (4), 615–626. 10.1007/BF00876528
18
CampioneM.CapitaniG. C. (2013). Subduction-zone earthquake complexity related to frictional anisotropy in antigorite. Nat. Geosci.6, 847–851. 10.1038/ngeo1905
19
CandelaT.BrodskyE. E. (2016). The minimum scale of grooving on faults. Geology44 (8), 603–606. 10.1130/G37934.1
20
CandelaT.RenardF. (2012). Fault morphology database. Available at: https://www.isterre.fr/french/recherche-observation/equipes/mecanique-des-failles/moyens-et-outils/article/donnees.html (Accessed 2020).
21
CandelaT.RenardF.SchmittbuhlJ.BouchonM.BrodskyE. E. (2011). Fault slip distribution and fault roughness. Geophys. J. Int.187 (2), 959–968. 10.1111/j.1365-246X.2011.05189.x
22
CandelaT.RenardF.KlingerY.MairK.SchmittbuhlJ.BrodskyE. E. (2012). Roughness of fault surfaces over nine decades of length scales. J. Geophys. Res.117 (B8), B08409. 10.1029/2011JB009041
23
CarboneG.LorenzB.PerssonB. N. J.WohlersA. (2009). Contact mechanics and rubber friction for randomly rough surfaces with anisotropic statistical properties. Eur. Phys. J.29 (3), 275–284. 10.1140/epje/i2009-10484-8
24
CashmanP. H.EllisM. A. (1994). Fault interaction may generate multiple slip vectors on a single fault surface. Geology22 (12), 1123–1126. 10.1130/0091-7613(1994)022¡1123:FIMGMS¿2.3.CO;2
25
CélérierB. (2008). Seeking Anderson's faulting in seismicity: a centennial celebration. Rev. Geophys.46 (4). 10.1029/2007RG000240
26
CochranE. S.VidaleJ. E.LiY.-G. (2003). Near-fault anisotropy following the Hector Mine earthquake. J. Geophys. Res.108 (B9) 2436. 10.1029/2002JB002352
27
Di ToroG.HanR.HiroseT.De PaolaN.NielsenS.MizoguchiK.et al (2011). Fault lubrication during earthquakes. Nature471, 494–498. 10.1038/nature09838
28
EdwardsJ. H.KluesnerJ. W.SilverE. A.BrodskyE. E.BrothersD. S.BangsN. L.et al (2018). Corrugated megathrust revealed offshore from Costa Rica. Nat. Geosci.11, 197–202. 10.1038/s41561-018-0061-4
29
EngelderJ. T. (1974). Microscopic wear grooves on slickensides: indicators of paleoseismicity. J. Geophys. Res.79 (29), 4387–4392. 10.1029/JB079i029p04387
30
EvansR. (1984). Anisotropy: a pervasive feature of fault zones?Geophys. J. Int.76 (1), 157–163. 10.1111/j.1365-246X.1984.tb05031.x
31
FengX.-T.GongY.-H.ZhouY.-Y.LiZ.-W.LiuX.-F. (2019). The 3D-printing technology of geological models using rock-like materials. Rock Mech. Rock Eng.52, 2261–2277. 10.1007/s00603-018-1703-y
32
FesslerG.SadeghiA.GlatzelT.GoedeckerS.MeyerE. (2019). Atomic friction: anisotropy and asymmetry effects. Tribol. Lett.67 (2), 59. 10.1007/s11249-019-1172-9
33
FilippovA.GorbS. N. (2013). Frictional-anisotropy-based systems in biology: structural diversity and numerical model. Sci. Rep.3, 1240. 10.1038/srep01240
34
FilippovA. E.WesthoffG.KovalevA.GorbS. N. (2018). Numerical model of the slithering snake locomotion based on the friction anisotropy of the ventral skin. Tribol. Lett.66. 10.1007/s11249-018-1072-4
35
GongH.ZhuP.SiL.ZhangX.XieG. (2018). “M-shape” nanoscale friction anisotropy of phosphorene. Comput. Mater. Sci.150, 364–368. 10.1016/j.commatsci.2018.04.013
36
GrasselliG.EggerP. (2003). Constitutive law for the shear strength of rock joints based on three-dimensional surface parameters. Int. J. Rock Mech. Mining Sci.40 (1), 25–40. 10.1016/S1365-1609(02)00101-6
37
HanaorD. A. H.GanY.EinavI. (2016). Static friction at fractal interfaces. Tribol. Int.93, 229–238. 10.1016/j.triboint.2015.09.016
38
JinK.TianY.EricksonJ. S.PuthoffJ.AutumnK.PesikaN. S. (2012). Design and fabrication of gecko-inspired adhesives. Langmuir28, 5737–5742. 10.1021/la204040p
39
KimY.-S.PeacockD. C. P.SandersonD. J. (2004). Fault damage zones. J. Struct. Geology26 (3), 503–517. 10.1016/j.jsg.2003.08.002
40
LiZ.PengZ.Ben‐ZionY.VernonF. L. (2015). Spatial variations of shear wave anisotropy near the San Jacinto fault zone in Southern California. J. Geophys. Res. Solid Earth120 (12), 8334–8347. 10.1002/2015JB012483
41
LiY.TangC. A.LiD.WuC. (2020). A new Shear strength criterion of three-dimensional rock joints. Rock Mech. Rock Eng.53, 1477–1483. 10.1007/s00603-019-01976-5
42
MassonnetD.RossiM.CarmonaC.AdragnaF.PeltzerG.FeiglK.et al (1993). The displacement field of the Landers earthquake mapped by radar interferometry. Nature364, 138–142. 10.1038/364138a0
43
MitchellT. M.FaulknerD. R. (2008). Experimental measurements of permeability evolution during triaxial compression of initially intact crystalline rocks and implications for fluid flow in fault zones. J. Geophys. Res.113 (B11). 10.1029/2008JB005588
44
OtsuboM.ShigematsuN.ImanishiK.AndoR.TakahashiM.AzumaT. (2013). Temporal slip change based on curved slickenlines on fault scarps along Itozawa fault caused by 2011 Iwaki earthquake, Northeast Japan. Tectonophysics608, 970–979. 10.1016/j.tecto.2013.07.022
45
PeacockD. C. P.SandersonD. J. (1992). Effects of layering and anisotropy on fault geometry. J. Geol. Soc.149 (5), 793–802. 10.1144/gsjgs.149.5.0793
46
PechmannJ. C.KanamoriH. (1982). Waveforms and spectra of preshocks and aftershocks of the 1979 imperial valley, california, earthquake: evidence for fault heterogeneity?J. Geophys. Res.87 (B13), 10579–10597. 10.1029/JB087iB13p10579
47
PionnierN.VeraJ.ContrairesE.BenayounS.BergerR.ValetteS. (2018). The effect of the orientation and the height of periodic sub-micrometric texturing on dropwise condensation. J. Colloid Interf. Sci.526, 184–193. 10.1016/j.jcis.2018.04.043
48
PollardD. D.SaltzerS. D.RubinA. M. (1993). Stress inversion methods: are they based on faulty assumptions?J. Struct. Geol.15 (8), 1045–1054. 10.1016/0191-8141(93)90176-B
49
PonsonL.BonamyD.AuradouH.MourotG.MorelS.BouchaudE.et al (2006a). Anisotropic self-affine properties of experimental fracture surfaces. Int. J. Fract.140, 27–37. 10.1007/s10704-005-3059-z
50
PonsonL.BonamyD.BouchaudE. (2006b). Two-dimensional scaling properties of experimental fracture surfaces. Phys. Rev. Lett.96, 035506. 10.1103/PhysRevLett.96.035506
51
QueenerC. A.SmithT. C.MitchellW. L. (1965). Transient wear of machine parts. Wear8 (5), 391–400. 10.1016/0043-1648(65)90170-5
52
QuinsatY.LartigueC.BrownC. A.HattaliL. (2018). Characterization of surface topography of 3d printed parts by multi-scale analysis. Int. J. Interact Des. Manuf.12 (3), 1007–1014. 10.1007/s12008-017-0433-9
53
RawlingG. C.BaudP.WongT.-F. (2002). Dilatancy, brittle strength, and anisotropy of foliated rocks: experimental deformation and micromechanical modeling. J. Geophys. Res.107 (B10), 8–1. 10.1029/2001JB000472
54
RempelA. W.RiceJ. R. (2006). Thermal pressurization and onset of melting in fault zones. J. Geophys. Res.111 (B9). 10.1029/2006JB004314
55
RenardV.HekinianR.FrancheteauJ.BallardR. D.BackerH. (1985). Submersible observations at the axis of the ultra-fast-spreading East Pacific rise (17°30′ to 21°30′S). Earth Planet. Sci. Lett.75 (4), 339–353. 10.1016/0012-821X(85)90178-5
56
RenardF.VoisinC.MarsanD.SchmittbuhlJ. (2006). High resolution 3d laser scanner measurements of a strike-slip fault quantify its morphological anisotropy at all scales. Geophys. Res. Lett.33 (4). 10.1029/2005GL025038
57
RenardF. (2020). Data From: Frictional anisotropy of 3D-printed fault surfaces. Fault roughness data. https://dx.doi.org/10.3389/feart.2021.627823.
58
RiceJ. R.Ben-ZionY. (1996). Slip complexity in earthquake fault models. Proc. Natl. Acad. Sci.93 (9), 3811–3818. 10.1073/pnas.93.9.3811
59
Roth-NebelsickA.EbnerM.MirandaT.GottschalkV.VoigtD.GorbS.et al (2012). Leaf surface structures enable the endemic Namib desert grass Stipagrostis sabulicola to irrigate itself with fog water. J. R. Soc. Interf.9 (73), 1965–1974. 10.1098/rsif.2011.0847
60
SagyA.HamielY. (2017). Three-dimensional structure and evolution of an asymmetric pull-apart basin. Int. J. Earth Sci. (Geol. Rundsch)106, 1619–1630. 10.1007/s00531-016-1374-4
61
SagyA.LyakhovskyV. (2019). Stress patterns and failure around rough interlocked fault surface. J. Geophys. Res. Solid Earth124 (7), 7138–7154. 10.1029/2018JB017006
62
SatakeK.KanamoriH. (1991). Abnormal tsunamis caused by the June 13, 1984, Torishima, Japan, earthquake. J. Geophys. Res.96 (B12), 19933–19939. 10.1029/91JB01903
63
Sauter FH-S. (2018). Technical information. Technical report, Sauter. Available at: https://dok.kern-sohn.com/manuals/files/English/FH-S-BA-e-1819.pdf (Accessed 2020).
64
SeidelJ. P.HaberfieldC. M. (1995). The application of energy principles to the determination of the sliding resistance of rock joints. Rock Mech. Rock Eng.28, 211–226. 10.1007/BF01020227
65
SheaW. T.KronenbergA. K. (1993). Strength and anisotropy of foliated rocks with varied mica contents. J. Struct. Geol.15 (9), 1097–1121. 10.1016/0191-8141(93)90158-7
66
SquelchA. (2017). 3D printing rocks for geo-educational, technical, and hobbyist pursuits. Geosphere14 (1), 360–366. 10.1130/GES01364.1
67
StupkiewiczS.LewandowskiM. J.LengiewiczJ. (2014). Micromechanical analysis of friction anisotropy in rough elastic contacts. Int. J. Sol. Struct.51 (23), 3931–3943. 10.1016/j.ijsolstr.2014.07.013
68
TapiaF.TourneauD. L.GéminardJ.-C. (2016). Anisotropic friction: assessment of force components and resulting trajectories. EPJ Techn. Instrum.3, 1. 10.1140/epjti/s40485-016-0029-y
69
TeseiT.CollettiniC.BarchiM. R.CarpenterB. M.Di StefanoG. (2014). Heterogeneous strength and fault zone complexity of carbonate-bearing thrusts with possible implications for seismicity. Earth Planet. Sci. Lett.408, 307–318. 10.1016/j.epsl.2014.10.021
70
TiwariA.DoroginL.SteenwykB.WarhadpandeA.MotamediM.FortunatoG.et al (2016). Rubber friction directional asymmetry. Epl116 (6), 66002. 10.1209/0295-5075/116/66002
71
Ultimaker2 Extended+. (2015). Technical information. Technical report, Ultimaker. Available at: https://ultimaker.com/download/7386/UserManual-UM2Extended-v2.1.pdf (Accessed 2020).
72
VadaccaL.ColciagoC. M.MichelettiS.ScottiA. (2018). Effects of the anisotropy of the fault zone permeability on the timing of triggered earthquakes: insights from 3d-coupled fluid flow and geomechanical deformation modeling. Pure Appl. Geophys.175 (12), 4131–4144. 10.1007/s00024-018-1936-4
73
VekinisG.AshbyM. F.BeaumontP. W. R. (1993). Plaster of Paris as a model material for brittle porous solids. J. Mater. Sci.28, 3221–3227. 10.1007/BF00354239
74
WangW.ScholzC. H. (1994). Wear processes during frictional sliding of rock: a theoretical and experimental study. J. Geophys. Res.99 (B4), 6789–6799. 10.1029/93JB02875
75
WangF.WangL.WuH.PangJ.GuD.LiS. (2017a). A lotus-leaf-like sio2 superhydrophobic bamboo surface based on soft lithography. Colloids Surf. A Physicochem. Eng. Asp.520, 834–840. 10.1016/j.colsurfa.2017.02.043
76
WangL.JuY.XieH.MaG.MaoL.HeK. (2017b). The mechanical and photoelastic properties of 3D printable stress-visualized materials. Sci. Rep.7. 10.1038/s41598-017-11433-4
77
WangJ.StewartR. R.DyaurN. I. (2018). Seismic response analysis of a 3D-printed dual-porosity physical model: Marine case. SEG, 301–305. 10.1190/segam2018-2997200.1
78
WibberleyC. A. J.YieldingG.Di ToroG. (2008). Recent advances in the understanding of fault zone internal structure: a review. Geol. Soc. Lond. Spec. Publ.299 (1), 5–33. 10.1144/SP299.2
79
YuC.WangQ. J. (2012). Friction anisotropy with respect to topographic orientation. Sci. Rep.2, 988. 10.1038/srep00988
80
ZhengY.GaoX.JiangL. (2007). Directional adhesion of superhydrophobic butterfly wings. Soft Matter3, 178–182. 10.1039/B612667G
Summary
Keywords
friction, anisotropy, seismic faults, 3D printing, plaster 3D model, frictional damages
Citation
Vincent-Dospital T, Steyer A, Renard F and Toussaint R (2021) Frictional Anisotropy of 3D-Printed Fault Surfaces. Front. Earth Sci. 9:627823. doi: 10.3389/feart.2021.627823
Received
10 November 2020
Accepted
05 January 2021
Published
07 May 2021
Volume
9 - 2021
Edited by
Takashi Nakagawa, Kobe University, Japan
Reviewed by
Shengji Wei, Nanyang Technological University, Singapore
Amir Sagy, Geological Survey of Israel, Israel
Updates
Copyright
© 2021 Vincent-Dospital, Steyer, Renard and Toussaint.
This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.
*Correspondence: Tom Vincent-Dospital, tom.vincent-dospital@fys.uio.no; Renaud Toussaint, renaud.toussaint@unistra.fr
This article was submitted to Solid Earth Geophysics, a section of the journal Frontiers in Earth Science
Disclaimer
All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article or claim that may be made by its manufacturer is not guaranteed or endorsed by the publisher.