Visualization and Quantification of the Penetration Behavior of 1 Bentonite Suspensions into the Pore Network of non-cohesive 2 Media by using μ-CT Imaging

13 Bentonite suspensions are an essential tool for different construction techniques in horizontal 14 and vertical drilling, in diaphragm and bored pile walls as well as in pipe jacking and 15 tunneling. One of the main tasks of the suspension is to prevent the surrounding ground from 16 collapsing during the excavation process of trenches, drill holes or tunnels. In order to 17 maintain the soil stability close to the excavation, the bentonite suspension has to counteract 18 against the earth and water pressure. Therefore, the pressure acting in the suspension has 19 to counter the groundwater pressure and to be transferred into an effective stress to support 20 the soil skeleton. 21 The creation of a pressure transfer mechanism can be achieved in two ways. A direct 22 relation exists between the mechanism of the pressure transfer and the penetration behavior 23 of the bentonite suspension in the subsoil. The relation of the size of the bentonite particles 24 in the suspension and the size of the pores in soft soil is decisive. In addition, the yield 25 strength of the bentonite suspension is a determining factor. 26 Concerning the penetration behavior two theoretical models exist actually: formation of a 27 filter cake and entire penetration into the pore space. If the pore space is smaller than the 28 size of the bentonite particles, a filtration process takes place. Here, the bentonite particles 29 agglomerate gradually at the entrance of the pore space and create a thin nearly 30 impermeable layer. This membrane is named filter cake. If the pore space is larger than the 31 size of the bentonite particles, the suspension penetrates into the subsoil up to a certain 32 depth. 33 These models have a more theoretical character due to missing visual evidence concerning 34 the interaction of the bentonite suspension in the pore space. Here, the micro CT technique 35 delivers a valuable contribution to this research. 36 37


Introduction
Bentonite suspensions are an and vertical drilling, in diaphragm and bored pile walls tunneling.One of the main tasks of the suspension is to prevent the collapsing during the excavation process of trenches, drill holes or tunnels.maintain the soil stability close to the excavation against the earth and water pressure.to counteract the groundwater pressure and support the soil skeleton.
Currently, the dominant theory in tunneling practice is technology [Müller-Kirchenbauer, 1977German Standard [DIN 4126, 2004 of effective stress can be achieved in two ways: with a limited penetration zone The membrane, named filter cake the size of the suspended bentonite particles filtered at the entrance of the pore soil.By gradual agglomeration of more bentonite particle build (Figure 1 (left)).Here, the area in terms of effective stress In case the pore size of the soil exceeds the size of th suspension penetrates completely into the pore space of the ground up to a certain depth [Walz, 2001].Due to the yield transferred along the surface of the soil penetration process stagnates in a certain depth, when the shear stress and groundwater pressure are balanced [Mueller-Kirchenbauer 1977] describe pressure transfer and the penetration behavior of the bentonite suspension in the subsoil in reference to the pore size in detail. 2 an essential tool for different construction techniques in horizontal and vertical drilling, in diaphragm and bored pile walls as well as in pipe jacking and One of the main tasks of the suspension is to prevent the surrounding excavation process of trenches, drill holes or tunnels.maintain the soil stability close to the excavation, the bentonite suspension has to counteract against the earth and water pressure.Therefore, the pressure acting in the suspension the groundwater pressure and has to be transferred into effective stress to the dominant theory in tunneling practice is adopted from diaphragm wall Kirchenbauer, 1977] and summarized in DIN 4126 (2004)., 2004] the formation of a pressure transfer mechanism can be achieved in two ways: (a) with a thin and flexible membrane penetration zone in the soil.
filter cake (a), develops when the pore size of the the size of the suspended bentonite particles [Walz, 2001].Here, the bentonite particles are d at the entrance of the pore space and the remaining filtrate water drains through the gradual agglomeration of more bentonite particles, a thin, impermeable membrane Here, the suspension pressure is transferred through effective stress to the soil skeleton.the pore size of the soil exceeds the size of the suspended bentonite particles suspension penetrates completely into the pore space of the ground up to a certain depth .Due to the yield point of the bentonite suspension, shear stresses are e surface of the soil particles within the penetration zone (b penetration process stagnates in a certain depth, when the suspension pressure undwater pressure are balanced (Figure 2 (left)).
] describes the direct relation between the mechanism of the pressure transfer and the penetration behavior of the bentonite suspension in the subsoil in in detail.
rinciple of support pressure transfer in the soil due to formation of a filter cake (left) and experimental result of filter cake on macroscale [I for different construction techniques in horizontal in pipe jacking and surrounding ground from excavation process of trenches, drill holes or tunnels.In order to the bentonite suspension has to counteract , the pressure acting in the suspension has ed into effective stress to from diaphragm wall and summarized in DIN 4126 (2004).According to of a pressure transfer mechanism in terms ) with a thin and flexible membrane or (b) of the soil is smaller than Here, the bentonite particles are er drains through the impermeable membrane is ed through the membrane e suspended bentonite particles, the suspension penetrates completely into the pore space of the ground up to a certain depth ion, shear stresses are within the penetration zone (b).The suspension pressure, transfer of he direct relation between the mechanism of the pressure transfer and the penetration behavior of the bentonite suspension in the subsoil in support pressure transfer in the soil due to formation of a filter cake experimental result of filter cake on macroscale [Imerys, 1998] Solid Earth Discuss., doi:10.5194/se-2016-42,2016 Manuscript under review for journal Solid Earth Published: March 2016 c Author(s) 2016.CC-BY 3.0 License.Based on long-term practical experienc [Anagnostou & Kovári 1994, Krause 1987, Boere 2001Boere 2015] ] and in addition proven by Min et.at. 2013, Heinz 2006, A In this study the penetration process is CT for the first time.The analysis provides detailed information concerning the interaction between the bentonite suspension and the non Furthermore, the µ-CT imaging provides the determination of size, pore size distribution a between the fluid and the glass beads, the bentonite suspension fluid" with a contact angle <90°.Both principles and analyzed using µ-CT imaging so that these phenomena are validated As an important result, the suspension can be demonstrated in detail.Beside the particles (glass beads), air and the penetrated bentonite particles in the filter cake and the filtrate

Materials
Bentonite is a natural clay.M Montmorillonite.A single Montmorillonite crystal consi Between these elementary layers case of Na + ions the bentonite is called Sodium bentonite, in case of Ca bentonite.
Preparation of a bentonite sus suspended in water (1) and dispersed by introducing high shear forces separate the single crystal layer suspension.Due to an additional between the elementary layers adsorbed at the cations and at the surface of the single l 3 Theoretical principle of support pressure transfer in the soil due to formation of a penetration ] (left) and experimental result of a penetration zone on macroscale practical experience, the theoretical principles are [Anagnostou & Kovári 1994, Krause 1987, Boere 2001, Bezuijen 2001 and in addition proven by several experimental expertise on the macroscale Arwanitaki 2009] (Figure 1 (right), Figure 2 (right)).
the penetration process is visualized and analyzed on the microscale analysis provides detailed information concerning the interaction bentonite suspension and the non-cohesive media within the pore space.
imaging provides the determination of parameters e.g.porosity, pore and particle size distribution.By analyzing the contact angle tween the fluid and the glass beads, the bentonite suspension is identified with a contact angle <90°.Both principles -filter cake and penetration imaging so that these phenomena are validated important result, the single phenomena of the filtration process suspension can be demonstrated in detail.Beside the "standard" identification of solid particles (glass beads), air and the penetrated bentonite suspension, the particles in the filter cake and the filtrated suspension water are detected.

Materials & Methods
Bentonite is a natural clay.Main component of bentonite is the plate like clay mineral .A single Montmorillonite crystal consists of 15-20 elementary layers.Between these elementary layers different cations (e.g.Na + , Ca 2+ , Mg 2+  ions the bentonite is called Sodium bentonite, in case of Ca 2+ reparation of a bentonite suspension consists of three steps: The powdery bentonite is and dispersed by introducing high shear forces (2 single crystal layer mechanically and distribute them homogeneously in the n additional swelling process (3), water molecules are embedded between the elementary layers of the Montmorillonite crystal.These water molecules are adsorbed at the cations and at the surface of the single layers as well.Hence, the distance on the microscale using µanalysis provides detailed information concerning the interaction cohesive media within the pore space.ters e.g.porosity, pore nd particle size distribution.By analyzing the contact angle identified as a "wetting filter cake and penetration -are identified on the microscale.
process of the bentonite identification of solid the filtered bentonite plate like clay mineral 20 elementary layers.2+ ) are adsorbed.In ions it is a Calcium The powdery bentonite is 2).The shear forces homogeneously in the , water molecules are embedded These water molecules are .Hence, the distance between the layers increases and the volume of the dispersed/suspended solids changes.This break-up of the layer corpuses suspension to develop.The required 16 hours.Afterwards the particle size of the suspended be determined (Appendix 1).solid contents were employed: Ca % by weight.
Glass beads with particle size of 2 used to ensure the reproducibility of the performed combinations of bentonite suspensi and non-cohesive media.The surface s SEM (Figure 3).Here, small parts of unevenness were detected.The penetration tests were conducted in test tubes made of length 160 mm) and silica glass suitable material for the µ-CT the type and concentration of bentonite suspension shows the combinations for scanning cake or the penetration process between the layers increases and the volume of the dispersed/suspended solids changes.up of the layer corpuses is essential for the rheological properties of th required swelling time of different bentonites 16 hours.Afterwards the particle size of the suspended Na-and Ca-bentonite particles .In the experimental study bentonite suspension with varying employed: Ca-bentonite in 25 % by weight, Na-bentonite in 8 Glass beads with particle size of 2 mm and 600 µm and a mean density of 2600 kg/m³ o ensure the reproducibility of the performed combinations of bentonite suspensi The surface structure of the glass beads was determined using small parts of unevenness were detected.
Image of the surface condition of the 2 mm glass beads using ration tests were conducted in test tubes made of acrylic glass glass (Ø 21 mm and length 200 mm) in order to provide the most scans.The label of each sample describes the container type, concentration of bentonite suspension and the size of the glass beads.combinations for scanning with µ-CT, which provide the performance of a filter process.
ombinations of test tube material, bentonite suspensions and glass beads size

Preliminary laboratory experiments
In preparation of the µ-CT scans, fundamental tests were performed in the bentonite laboratory at Ruhr-University Bochum to identify and determine the influence of different parameters on the penetration behavior of the bentonite suspensions into glass beads.Furthermore, the penetration depth of the suspension has to be limited to some extend in order to ensure a high quality of the µ-CT scans due to smaller areas of interest.This is a challenging task because bentonite suspensions are non-Newtonian fluids with a yield point and exhibit thixotropic behavior [Luckham & Rossi 1999, Maxey 2007, API RP 13B] Here, the bentonite suspensions were prepared with varying solid contents of Ca 2+ in 20 % and 25 % by weight and Na + in 8 % and 13 % by weight.All suspensions are combined with the glass beads of 2 mm, 600 µm and a combination of 2 mm + 600 µm.After swelling times of 24, 48, 72, 96, 120 and 192 hours the penetration tests were performed using test tubes of acrylic glass and silica glass and the penetration depth was measured.
The suspension made of Ca 25 % penetrates into the glass beads size of 2 mm, 600 µm and the combination 2 mm + 600 µm.The highest penetration depth is reached in the coarse material of 2 mm; the penetration depths in 600 µm and the combination of 2 mm + 600 µm are comparable.Comparison of the penetration depth measured in the acrylic glass and silica glass tubes shows slight differences of the absolute values for the same swelling times (Figures 4 and 5).
The suspensions made of Ca 2+ 20 % and Na + 8 % show high values of penetration depth within glass beads of 600 µm and the combination of 2 mm + 600 µm (Figures 4 and 5).This area is too large for a µ-CT scan of high quality.Therefore, the solid content of Na + suspension was increased to 13 % and the Ca 25 % suspension was chosen for further testing.Following general observations can be made: -The penetration depth in the glass beads decreases with increasing swelling time of the bentonite suspension.Here, Ca 2+ bentonite shows a distinct response in terms if a reduced penetration depth in comparison to Na + bentonite.-The general performance of the penetration behavior is irrespective of the material of the test tube within a swelling time up to 120 hours.Slight differences of absolute values of penetration depth are detected.Bentonite suspensions with swelling times beyond 120 hours should be refused for the µ-CT scans.-The penetration depth of the same bentonite suspension in glass beads of 600 µm and the combination of 2 mm + 600 µm are equal.The penetration depth in glass beads of 2 mm is higher by trend.

µ-CT Imaging
X-ray computed tomography was used for the 3D imaging of the samples, using a nanotom S 180 µ-CT (tube characteristics: 180 kV, 500 mA) device, of the Leibniz Institute for Applied Geophysics (Tab.2).The nanotom is a compact CT system for pore scale imag i.e. for high resolution imaging within the micrometer (typically 1 range (about 700 nm for very small samples), featuring high image sharpness due to a significantly reduced penumbra effect [Brunke et al. 2008] CT imaging and 3D image reconstruction is given by Buzug (2010).The 3D image data were processed with the AVIZO Fire software suite (Visualization Sciences Group).Due to the low image noise and due to the fact that only segmentation processing has been performed by the fast and robust "automatic threshold selection method" described by Otsu (1979).
Development of penetration depth (cm) of bentonite suspension in glass beads using test made of acrylic glass for increasing swelling times (h) Development of penetration depth (cm) of bentonite suspension in glass beads using test made of quartz glass for increasing swelling times (h) ray computed tomography was used for the 3D imaging of the samples, using a nanotom CT (tube characteristics: 180 kV, 500 mA) device, of the Leibniz Institute for Applied Geophysics (Tab.2).The nanotom is a compact CT system for pore scale imag i.e. for high resolution imaging within the micrometer (typically 1-2 µm) to sub range (about 700 nm for very small samples), featuring high image sharpness due to a cantly reduced penumbra effect [Brunke et al. 2008].A comprehensive overview of µ CT imaging and 3D image reconstruction is given by Buzug (2010).The 3D image data were processed with the AVIZO Fire software suite (Visualization Sciences Group).Due to the low image noise and due to the fact that only few phases exist for segmentation, has been performed by the fast and robust "automatic threshold selection method" described by Otsu (1979).

Development of penetration depth (cm) of bentonite suspension in glass beads using test
Development of penetration depth (cm) of bentonite suspension in glass beads using test ray computed tomography was used for the 3D imaging of the samples, using a nanotom CT (tube characteristics: 180 kV, 500 mA) device, of the Leibniz Institute for Applied Geophysics (Tab.2).The nanotom is a compact CT system for pore scale imaging purposes, 2 µm) to sub-micrometer range (about 700 nm for very small samples), featuring high image sharpness due to a prehensive overview of µ-CT imaging and 3D image reconstruction is given by Buzug (2010).The 3D image data were processed with the AVIZO Fire software suite (Visualization Sciences Group).Due to the low segmentation, phase has been performed by the fast and robust "automatic threshold Solid Earth Discuss., doi:10.5194/se-2016-42,2016 Manuscript under review for journal Solid Earth Published: 8 March 2016 c Author(s) 2016.CC-BY 3.0 License.

Image Processing
The preparation and execution of the µ-CT scans followed a standardized procedure: 1. Sample preparation; 2. Sample installation and µ-CT setup; 3. µ-CT scan; 4. 3D data reconstruction; 5. Image processing and analysis.
Here, the glass beads are filled in the test tube using a funnel equipment with a specific height to provide a constant dense packing or density.Afterwards, the bentonite suspension was added and the penetration process took place.The time period for finishing this process is short (< 2 min).The test tube is installed and the µ-CT device, i.e. detector and tube setup, are prepared individually (Table 2).All samples have been positioned such way, that the interfacial surface area between the suspension -glass bead -air filled void space has been investigated as a region of interest.After data reconstruction, the 3D data sets have been processed with the AVIZO Fire software suite.Image processing has been performed for all samples as followed: 1. Data filtering (either using a median or a non local means filter [Ohser & Schladitz, 2009]); 2. Phase segmentation (void space, suspension, matrix); 3. Pore space analysis (pore separation, pore size distribution, porosity); 4. Grain size analysis (grain separation, grain size distribution); 5. Suspension distribution analysis (spatial extent, air inclusions and contact angles).
Table 2: Parameters of detector and tube setup 3 Results

Porosity and pore size distribution of non-cohesive media
The theoretical model of the penetration process of a bentonite suspension into noncohesive media depends -amongst others -on the porosity and permeability.In a soil mechanical sense, porosity is a dimensionless number that quantifies the existing pore space in reference to the whole soil volume.Here, the pores may be completely or partly filled with air, bentonite suspension or water.In addition, permeability is a measure for the connectivity of the single pores.a high porosity a high permeability cannot be deduced.A high porosity may be based on a huge amount of small pores or on the contrary on a small amount of big pores [ 1951].In general, the porosity is determined by the shape, grading and compactness of the non-cohesive media / soil particles.
Basically, porosity is identical for samples with the same volume, particle shape, particle grading and compactness/density.For determination of porosity samples were chosen: Sample 1 25 % -600 µm, Sample 3: silica glass volume was identical; the non concerning the resulting porosity, p The porosity obtained from µlatter were determined by comparison of the weight of materials under dry and water saturated conditions.Here, distilled water with a density of Comparing the porosity values in Table 3 shows evidence for the general laws of soil mechanics: glass beads with only one particle size of values of porosity of 39.60 % ( same bulk density.Thereby, the a higher median value of the pore size diameter, while the higher pore quantity due to smaller median value In contrast, the particle size of the glass beads of Sample 3 contain 2 mm / 600 µm.Porosity is smaller due to the fact into the pore space of the bigger glass beads of reduces to 34.59 %. connectivity of the single pores.A high permeability is associated to a high porosity, but from gh permeability cannot be deduced.A high porosity may be based on a huge amount of small pores or on the contrary on a small amount of big pores [ he porosity is determined by the shape, grading and compactness of the / soil particles.
porosity is identical for samples with the same volume, particle shape, particle grading and compactness/density.For determination of porosity using µ Sample 1: silica glass -Ca 25 % -2 mm, Sample 2 ilica glass -Ca 25 % -2 mm + 60µm (Figure ; the non-local-means filter was applied.Table 3 presents information porosity, pore size and quantity of pores. illustrations of pore space of glass beads of Sample 1 (2 mm -left), Sample 2 ( 600 µm -right).
-CT imaging fits well the experimental determined values.The by comparison of the weight of materials under dry and water saturated conditions.Here, distilled water with a density of 1000 kg/m³ was used.porosity and pore size of glass beads 2 mm, 600 µm Comparing the porosity values in Table 3 shows evidence for the general laws of soil mechanics: glass beads with only one particle size of 2 mm or 600 µm show % (2 mm) and 39.28 % (600 µm) due to the same volume and the Thereby, the 2 mm glass beads provide a smaller quantity of pore of the pore size diameter, while the 600 µm glass beads provide a smaller median value of the pore size diameter.
the particle size of the glass beads of Sample 3 contains a ratio of 50 orosity is smaller due to the fact, that the smaller glass beads of 600 into the pore space of the bigger glass beads of 2 mm.Therefore, the porosity of this mixture A high permeability is associated to a high porosity, but from gh permeability cannot be deduced.A high porosity may be based on a huge amount of small pores or on the contrary on a small amount of big pores [Engelhard he porosity is determined by the shape, grading and compactness of the porosity is identical for samples with the same volume, particle shape, particle µ-CT imaging, three Sample 2: silica glass -Ca (Figure 6).The analyzed Table 3 presents  Comparing the porosity values in Table 3 shows evidence for the general laws of soil show nearly identical ) due to the same volume and the glass beads provide a smaller quantity of pores due to glass beads provide a diameter.a ratio of 50 % / 50 % of that the smaller glass beads of 600 µm fit the porosity of this mixture The characterization of the pore space are evident for the analysis of the penetration Therefore, the histograms of pore size distribution are transferred into a diagram showing the size of the pores in reference to the proportion of the pores for glass beads of 1), 600 µm (Sample 2) and combination of

Sample 1: Ca 25 %
Sample 1 contains a viscous bent determined optically and by using an even penetration performance, bo depth of the suspension into the pore space of the visualization of the sample shows artefacts below the penetration zone in the area of dry glass beads.From the optical point of view an explicit bounda space filled with bentonite suspension and the air filled pores.penetration effect took place.be demonstrated that the suspension stagnates at the depth of 15 mm The characterization of the pore space concerning pore size, porosity and thus permeability for the analysis of the penetration behavior of the bentonite suspension.

Segmentation of penetration depth and filter cake -silica glass -2 mm
Sample 1 contains a viscous bentonite suspension Ca 25 %.The penetration depth was using imaging with application of module "Measurement" an even penetration performance, both results show the same value of 15 depth of the suspension into the pore space of the 2 mm glass beads packing visualization of the sample shows artefacts below the penetration zone in the area of dry From the optical point of view an explicit boundary exist between the space filled with bentonite suspension and the air filled pores.Due to the coarse material penetration effect took place.Following the visualization and analysis of the 3D data, it can he suspension flows as a homogenous fluid into the pores and mm.Sample 2 contains the same viscous bentonite suspension as Sample 1. Due to an uneven penetration process, the visible penetration depth of 5 mm (0.5 cm) is smaller than the etration depth determined by applying the module "measurement" using the suspension penetrates deeper in the middle of the test tube into in a histogram (Figure 9).This Histogram represents a vertical line through the sample and shows different grey values of the detected media.Here, the glass beads have the highest density and therefore show the highest grey value.The lower the density of the medium, the lower the grey value in the Analyzing the grey value using the module "Line Probe" shows three different within the pore space (Figure 9).The .This feature indicates that constant within the pore space, bentonite suspension and air.
Sample 2 contains the same viscous bentonite suspension as Sample 1. Due to an uneven is smaller than the etration depth determined by applying the module "measurement" using imaging of 9 mm in the middle of the test tube into

Sample 3: Ca 25 %
Sample 3 shows the penetration process of viscous bentonite suspension Ca combination of 2 mm + 600 µm visible penetration depth of 8 determined by applying the mod cm).The mean value of the penetration depth is comparable to the value of Sample 2 with glass beads of 600 µm.The visualization of zone in the area of dry glass beads be buoyant too in the viscous fluid.homogeneous fluid into the pore space of the between the pore space filled with bentonite suspension and the air filled pores.
Analyzing the grey value using the module "Line Probe" shows again three different values of glass beads, air and bentonite suspension within the pore space (Figure suspension remains at nearly the same representative value that means the concentration or solid content of the suspension is constant within the pore space (three phases).
than at the visible edge (Figure 10, left).The small glass beads tend to be buoyant in the viscous fluid.Despite that, the bentonite suspension penetrates as a homogeneous fluid into the glass beads.Again, an explicit boundary exist space filled with bentonite suspension and the air filled pores.
Analyzing the grey value using the module "Line Probe" shows again three different values of glass beads, air and bentonite suspension within the pore space (Figure suspension remains at nearly the same representative value that means the concentration or solid content of the suspension is constant within the pore space (three phases).
Identification of three phases: glass beads, bentonite suspension and air -silica glass -2 mm + 600 µm shows the penetration process of viscous bentonite suspension Ca 600 µm.The uneven penetration process leads to the deviation of the visible penetration depth of 8 mm (0.8 cm) in comparison to the determined by applying the module "measurement" between 6 mm (0.6 cm) and 8 The mean value of the penetration depth is comparable to the value of Sample 2 with .The visualization of the sample shows artifacts below the penetration zone in the area of dry glass beads (Figure 11, left).Some of the small glass beads tend to in the viscous fluid.In general, the bentonite suspension penetrates as a pore space of the glass beads.Again, an explicit boundary exist between the pore space filled with bentonite suspension and the air filled pores.
Analyzing the grey value using the module "Line Probe" shows again three different values of , air and bentonite suspension within the pore space (Figure suspension remains at nearly the same representative value that means the concentration or solid content of the suspension is constant within the pore space (three phases).
The small glass beads tend to be buoyant in the viscous fluid.Despite that, the bentonite suspension penetrates as a s between the pore Analyzing the grey value using the module "Line Probe" shows again three different values of glass beads, air and bentonite suspension within the pore space (Figure 10, right).The ue that means the concentration or solid content of the suspension is constant within the pore space (three phases).phases: glass beads, bentonite suspension and air.
shows the penetration process of viscous bentonite suspension Ca 25 % into the .The uneven penetration process leads to the deviation of the in comparison to the penetration depth cm) and 8 mm (0.8 The mean value of the penetration depth is comparable to the value of Sample 2 with facts below the penetration glass beads tend to the bentonite suspension penetrates as a glass beads.Again, an explicit boundary exists between the pore space filled with bentonite suspension and the air filled pores. Analyzing the grey value using the module "Line Probe" shows again three different values of , air and bentonite suspension within the pore space (Figure 11, right).The suspension remains at nearly the same representative value that means the concentration or solid content of the suspension is constant within the pore space (three phases).600 µm and results in a This value is quite high in comparison to the 600 µm glass beads.On suspension is the more viscous fluid due to the higher solid entonite particle size (Appendix 1) particles.Therefore, the pore space of the same (left) it is visible that a nalyzing the grey value using the module "Line bentonite suspension and water are The amount of water is very low and concentrates on a very limited area within the test tube.As it can be seen from Figure 12, the shows a nearly constant value solid content of the suspension phases: glass beads, bentonite suspension, water and air.The visualization of Sample 5 glass beads (Figure 13).Some of the small glass beads tend to viscous fluid.In general, the bentonite suspension penetrates as a homogeneous fluid into the pore space of the glass beads.Again, an explicit boundary exist space filled with bentonite suspension and the air fille The analysis of the grey value using the module "Line Probe" shows an additional phase of water below the end of the penetration zone of the suspension 13).The water does not appear at the walls of the test tubes, it beads.
The µ-CT scanning is based on the detection of media with different densit difference in density, the easier the single medi value of water 1000 kg/m³ is close to the density of the left side of the histogram (Figure placed above the area of glass beads (sample length approx.0 space of the glass beads (sample length approx.5000 bentonite suspension is slightly higher.value close to the original bentonite suspension.This gi took place within the pore space.Here, the suspension water is separated in a small amount from the bentonite particles.The particles remain in the pore space, the filtrate water drains into the pore space below.The so pores increases gently, in place where some suspension water is filtrated Sample 5 shows the penetration process of viscous bentonite suspension Ca whereas the test tube is made of acrylic glass.The even penetration process leads to the visible penetration depth of 15 mm (1.5 cm) and to the penetration determined by applying the module "Measurement" This value matches the penetration depth of Sample 1 using Ca 25 % and glass beads of in a test tube made of silica glass (Figure 8).It provides evidence that the material of the the penetration depth.
5 shows artifacts below the penetration zone in the a Some of the small glass beads tend to be buoyant too in the the bentonite suspension penetrates as a homogeneous fluid into the pore space of the glass beads.Again, an explicit boundary exists space filled with bentonite suspension and the air filled pores.
The analysis of the grey value using the module "Line Probe" shows an additional phase of water below the end of the penetration zone of the suspension within the pore space (Figure The water does not appear at the walls of the test tubes, it is located inside the glass scanning is based on the detection of media with different densit difference in density, the easier the single media can be identified.However, the density kg/m³ is close to the density of the Ca 25 % suspension 1 istogram (Figure 13) shows the grey value of the bentonite suspension placed above the area of glass beads (sample length approx.0 -2500 µm) (sample length approx.5000 -16000 µm), the grey value of the bentonite suspension is slightly higher.In addition, the filtrated water shows a lower grey value close to the original bentonite suspension.This gives evidence that a filter process took place within the pore space.Here, the suspension water is separated in a small amount from the bentonite particles.The particles remain in the pore space, the filtrate water drains The solid content/density of the bentonite suspension within the pores increases gently, in place where some suspension water is filtrated.dentification of four phases: glass beads, bentonite suspension, water and air Sample 5 shows the penetration process of viscous bentonite suspension Ca 25 % into 2 of acrylic glass.The even penetration m) and to the penetration determined by applying the module "Measurement" (Figure 13).and glass beads of 2 .It provides evidence that the material of the facts below the penetration zone in the area of dry be buoyant too in the the bentonite suspension penetrates as a homogeneous fluid into between the pore The analysis of the grey value using the module "Line Probe" shows an additional phase of within the pore space (Figure is located inside the glass scanning is based on the detection of media with different densities.The larger the can be identified.However, the density suspension 1025 kg/m³.The grey value of the bentonite suspension that is 2500 µm).Within the pore the grey value of the water shows a lower grey ves evidence that a filter process took place within the pore space.Here, the suspension water is separated in a small amount from the bentonite particles.The particles remain in the pore space, the filtrate water drains lid content/density of the bentonite suspension within the phases: glass beads, bentonite suspension, water and air Solid Earth Discuss., doi:10.5194/se-2016-42,2016 Manuscript under review for journal Solid Earth Published: 8 March 2016 c Author(s) 2016.CC-BY 3.0 License.

Sample 6: Na 13% -silica glass
In Sample 6 the penetration process of bentonite suspension Na 13% performs into the combination of 2 mm + 600 µm leads to a filter cake performance bentonite particles are filtered from the suspension, the bentonite particles attach the pore access and the filtrate water of the suspension drains thro (Figure 14).Therefore, the visualization of Sample 6 shows analysis using the module "measurement".The Histogram in Figure 15 shows the increase of the density of the bentonite suspension within a short area/depth followed by the detection of the filtrate water bentonite suspension above the level of glass beads is shown.Within the pore space of the glass beads the grey value of the bentonit water shows a lower grey value bentonite particles took place the bentonite particles.The particles remain in the pore space, the filtrate water drains into the pore space below.The solid content/density of the bentonite suspension within the pores increases, which is verified by increasing grey values.In Sample 6 the penetration process of bentonite suspension Na 13% performs into the 600 µm glass beads in a test tube of silica glass.Here, the performance at the entrance of the pore space of glass beads.The bentonite particles are filtered from the suspension, the bentonite particles attach the pore access and the filtrate water of the suspension drains through the glass beads the visualization of Sample 6 shows no penetration depth within the easurement".
Sample 6: 3D embodiment of the area of filter cake as an assembly of solid material with a (left) and apparent accumulation of bentonite particles at the pore access (right) shows the increase of the density of the bentonite suspension followed by the detection of the filtrate water.The grey value of the bentonite suspension above the level of glass beads is shown.Within the pore space of the glass beads the grey value of the bentonite suspension is slightly higher than the filtrate ater shows a lower grey value.This gives evidence that the filtration process of the took place at the pore access.The suspension water is separated from particles.The particles remain in the pore space, the filtrate water drains into the pore space below.The solid content/density of the bentonite suspension within the pores verified by increasing grey values.
dentification of four phases: glass beads, bentonite, filtrate water and air especially the density of the bentonite suspension varies.
In Sample 6 the penetration process of bentonite suspension Na 13% performs into the Here, the infiltration at the entrance of the pore space of glass beads.The bentonite particles are filtered from the suspension, the bentonite particles attach gradually at ugh the glass beads no penetration depth within the as an assembly of solid material with a (left) and apparent accumulation of bentonite particles at the pore access (right).
shows the increase of the density of the bentonite suspension he grey value of the bentonite suspension above the level of glass beads is shown.Within the pore space of the e suspension is slightly higher than the filtrate the filtration process of the he suspension water is separated from particles.The particles remain in the pore space, the filtrate water drains into the pore space below.The solid content/density of the bentonite suspension within the pores phases: glass beads, bentonite, filtrate water and air, Solid Earth Discuss., doi:10.5194/se-2016-42,2016 Manuscript under review for journal Solid Earth Published: 8 March 2016 c Author(s) 2016.CC-BY 3.0 License.

Contact angle of bentonite suspensions and glass beads
In general, the contact angle determined using µ-CT imaging helps to classify the manner of interaction between a fluid and the surface of a solid, e.g.type of fluid, material and surface roughness of solid.A contact angle of 90° is the limit between a wetting and non-wetting fluid.Contact angles over are typical for Mercury, contact angles less than 90° present the behavior of Water (Figure 16).For the bentonite suspension Ca 25 % the contact angles are determined using the module "angular measurement" for the different glass beads 2 mm (Sample 1), 600 µm (Sample 2) and 2 mm + 600 µm (Sample 3).Table 4 shows the results of contact angles.In general, the contact angle for bentonite suspension Ca 25 % is smaller than 90°.The suspension is classified as a wetting fluid.

Discussion
In this study, two different bentonites Ca 2+ and Na + were used.Samples 1, 2, 3 and 5 contain Ca-bentonite with solid content of 25 %.In Sample 1 and 5 the bentonite suspension penetrates into glass beads 2 mm and both tests perform identical values for penetration depth of 15 mm.It can be shown, that the material of the test tube -silica glass or acrylic glass -does not influence the result of the penetration test (Figure 8 and 12).
Using the smaller glass beads of 600 µm and the combination of 2 mm + 600 µm, the penetration depth reduces considerably (Figure 10 and 11).As shown in Figure 7, the distribution of the pore size is nearly identical for glass beads of 600 µm and the combination of 2 mm + 600 µm.In general, the pore size ranges between 0.3 mm -1.5 mm with a peak around 0.7 mm.The range of pore size for the 2 mm glass beads is between 0.6 mm -2.1 mm with a peak around 0.9 mm.As a consequence, the same bentonite suspension Ca 25 % with a particle size of approximately 10 µm (Appendix 1) penetrates deeper into the glass beads 2 mm with larger pore size (Figure 17).The Histogram of Samples bentonite suspension and air within the pore space.penetrates as a homogeneous fluid into the pores of the coarse material.Within this stu this is defined as the "standard penetration behavior" and can be linked directly to the experimentally performed penetration zones in first time, this penetration behavior on microscale using µ-CT.
According to Appendix 1, Napenetration depth of Na 8 % into glass beads of 600 µm value of the conducted test se particles of Ca-bentonite (approximately 10 µm) the smaller the penetration depth into the same size of glass beads of Sample 2 and Exceptionally, four different phases are detected within the beads, bentonite suspension depth.In addition, the histogram of Sample 5 identifies the same phases: bentonite suspension, air and filtrated water.Compared water within Sample 4 is quite large.This may be due to the lower solid content of Na 8 % (Sample 4) compared to Ca 25 % (Sample 5).
Sample 6 using Na 13 % and glass the filter cake within this test series.The bentonite particles remain at the "entrance" of the pore space of the glass beads and the filtrated suspension water flows deeper into the tube.The filter cake is illustrated in the 3D embodiment of Figure 1 shows a varying density of the bentonite suspension within the limited filtration area at the pore space "entrance".For the first time suspension is visualized on microscale using µ From the gained results of the six performance of a penetration zone and reasons of the filtration of water within the penetration effect as seen in Sample 4 and 5 using different types of bentonite work is needed.
Average diameter of pore space [mm] within glass beads 2 mm, 600 µm and combination 2 mm + 600 µm in reference to the determined penetration depth [cm].
1, 2, and 3 identifies three different phases: glass beads, bentonite suspension and air within the pore space.Here, the bentonite suspension penetrates as a homogeneous fluid into the pores of the coarse material.Within this stu this is defined as the "standard penetration behavior" and can be linked directly to the experimentally performed penetration zones in Figure 2 (right) on the macroscale.For the first time, this penetration behavior of bentonite suspension into coarse material -bentonites show a size of approximately 3 µm.Therefore, the penetration depth of Na 8 % into glass beads of 600 µm in Sample 4 reaches the largest value of the conducted test series 23 mm (see also Figure 5  , four different phases are detected within the histogram of Sample 4: beads, bentonite suspension, air and filtrated water at the lower area of the penetration istogram of Sample 5 identifies the same phases: and filtrated water.Compared to Sample 5, the area of filtrated water within Sample 4 is quite large.This may be due to the lower solid content of Na 8 % (Sample 4) compared to Ca 25 % (Sample 5).
Sample 6 using Na 13 % and glass beads of combination 2 mm + 600 µm performs uniquely the filter cake within this test series.The bentonite particles remain at the "entrance" of the pore space of the glass beads and the filtrated suspension water flows deeper into the tube.cake is illustrated in the 3D embodiment of Figure 14.Furthermore, the Histogram shows a varying density of the bentonite suspension within the limited filtration area at the pore space "entrance".For the first time, the performance of a filter cake on microscale using µ-CT.
From the gained results of the six samples, a general classification concerning the performance of a penetration zone and of a filter cake cannot be derived.Furthermore, the reasons of the filtration of water within the penetration effect as seen in Sample 4 and 5 using different types of bentonite cannot be explained satisfyingly.Here, further research Average diameter of pore space [mm] within glass beads 2 mm, 600 µm and combination 1, 2, and 3 identifies three different phases: glass beads, the bentonite suspension penetrates as a homogeneous fluid into the pores of the coarse material.Within this study, this is defined as the "standard penetration behavior" and can be linked directly to the on the macroscale.For the material is visualized show a size of approximately 3 µm.Therefore, the in Sample 4 reaches the largest and Figure 6).The larger the bentonite (approximately 10 µm) the smaller the penetration depth into the istogram of Sample 4: glass and filtrated water at the lower area of the penetration istogram of Sample 5 identifies the same phases: glass beads, the area of filtrated water within Sample 4 is quite large.This may be due to the lower solid content of Na 8 % beads of combination 2 mm + 600 µm performs uniquely the filter cake within this test series.The bentonite particles remain at the "entrance" of the pore space of the glass beads and the filtrated suspension water flows deeper into the tube.urthermore, the Histogram shows a varying density of the bentonite suspension within the limited filtration area at the the performance of a filter cake of bentonite a general classification concerning the a filter cake cannot be derived.Furthermore, the reasons of the filtration of water within the penetration effect as seen in Sample 4 and 5 cannot be explained satisfyingly.Here, further research

Conclusions
The aim of this study is the visualization and analysis of the penetration behavior of bentonite suspensions in non-cohesive granular material on microscale using µ-CT scanning.
The widely accepted scenarios of filter cake formation and entire penetration of the suspension into the pore space were conducted experimentally in test tubes using different combinations of bentonite suspension and granular material (glass beads).These phenomena were scanned with high-resolution µ-CT technique.The 3D embodiment of the different samples were analyzed concerning soil mechanical aspects, e.g.particle size distribution, pore size distribution, porosity of the granular material, and concerning the interaction of the bentonite suspension within the pore space, e.g.contact angle, penetration depth and filter cake thickness.
These effects are verified by investigating the different phase situation in a histogram.The Histogram represents a vertical line through the sample and shows different grey values of the detected media.The lower the density of the medium, the lower the grey value in the histogram.Here, the histogram of each sample offers the identification of single phases: glass beads, bentonite suspension and air as well as the filter cake at the "entrance" of the pore space of the glass beads or the variation of density of a bentonite suspension filtrating within the penetration depth.
Sample 1, 2 and 3 show the penetration behavior of bentonite suspension into coarse material is visualized on microscale using µ-CT.Three phases -glass beads, bentonite suspension and air -are detected in the Histogram.An additional phase -water filtrated from the suspension -is identified in Sample 4 and 5.The performance of a filter cake of bentonite suspension is visualized in Sample 6 on microscale using µ-CT.The agglomeration of the bentonite particles at the entrance of the pore space results in the development of a filter cake.This area is identified as a distinct media.Furthermore, the histogram of Sample 6 shows a varying density of the bentonite suspension within the limited filtration area at the pore space "entrance".
In summary, the µ-CT technique delivers a valuable contribution for the research on the interaction of bentonite suspensions penetration the pore space of non-cohesive media.This study shows the missing visual evidence concerning the theoretical interaction models of the bentonite suspension in the pore space on microscale.

Figure 2 :
Figure 2: Theoretical principle of support pressure transfer in the soil due to formation of a penetration zone [Zizka & Thewes, 2016] (left) and experimental result of a penetration zone on macroscale [Imerys, 1998] (right).
Theoretical principle of support pressure transfer in the soil due to formation of a penetration ] (left) and experimental result of a penetration zone on macroscale the theoretical principles are widely accepted 2001, Zizka & Thewes experimental expertise on the macroscale [IBECO, (Figure 1 (right), Figure 2 (right)).

Figure 3 :
Figure 3: Image of the surface condition of the 2 Microscopy (SEM)

Figure 4 :Figure 5 :
Figure 4: Development of penetration depth (cm) of bentonite suspension in glass beads using test tubes made of acrylic glass for increasing swelling times (h)
the histograms of pore size distribution are transferred into a diagram showing the size of the pores in reference to the proportion of the pores for glass beads of 2) and combination of 2 mm + 600 µm (Sample 3) in Figure distribution of glass beads 2 mm (Sample 1), 600 µm (Sample 2) concerning pore size, porosity and thus permeability of the bentonite suspension.the histograms of pore size distribution are transferred into a diagram showing the size of the pores in reference to the proportion of the pores for glass beads of 2 mm (Sample (Sample 3) in Figure 7. (Sample 2) and combination .The penetration depth was easurement".Due to sults show the same value of 15 mm penetration glass beads packing (Figure 8).The visualization of the sample shows artefacts below the penetration zone in the area of dry ry exist between the pore Due to the coarse material the Following the visualization and analysis of the 3D data, it can flows as a homogenous fluid into the pores and Solid Earth Discuss., doi:10.5194/se-2016-42,2016 Manuscript under review for journal Solid Earth Published: 8 March 2016 c Author(s) 2016.CC-BY 3.0 License.

Figure 8 :
Figure 8: Sample 1, 3D embodiment of penetration depth This effect was revised by investigating the three This Histogram represents a vertical line through the sample and shows different grey values of the detected media.Here, the glass beads have the highest density and therefore show the highest grey value.The lower the density of the medium, the lower the histogram.Analyzing the grey value using the module "Line Probe" shows three different values of glass beads, bentonite suspension and suspension is identified at nearly the same representative the concentration/solid content of the suspension remains the suspension stays as a homogenous fluid within the pores.

Solid
Earth Discuss., doi:10.5194/se-2016-42,2016 Manuscript under review for journal Solid Earth Published: 8 March 2016 c Author(s) 2016.CC-BY 3.0 License. the pore space than at the visible edge buoyant in the viscous fluid.Despite that, the bentonite suspension penetrates as a homogeneous fluid into the glass beads.Again, an explicit boundary exist space filled with bentonite suspension and the Analyzing the grey value using the module "Line Probe" shows again three different values of glass beads, air and bentonite suspension within the pore space (Figure suspension remains at nearly the same representative val solid content of the suspension is constant within the pore space (three phases).

Solid
Earth Discuss., doi:10.5194/se-2016-42,2016 Manuscript under review for journal Solid Earth Published: 8 March 2016 c Author(s) 2016.CC-BY 3.0 License. the penetration process of viscous bentonite suspension Ca mm glass beads whereas the test tube is process leads to the visible penetration depth of 15 depth of 15 mm (1.5 cm) determined by applying the module "Measurement" This value matches the penetration depth of Sample 1 using Ca mm in a test tube made of silica glass test tube has no influence on the

Figure 14 :
Figure 14: Sample 6: 3D embodiment of closed surface (left) and apparent accumulation of bentonite particles at the pore access (right)

Figure 15 :
Figure 15: Sample 6, identification of especially the density of the bentonite suspension varies

Figure 17 :
Figure 17: Average diameter of pore space [mm] within glass beads 2 mm, 600 µm and combination 2 mm + 600 µm in reference to the determined penetration depth [cm] and Figure bentonite (approximately 10 µm) the smaller the penetration depth into the same size of glass beads of Sample 2 and Sample 3.

Table 1 :
Combinations of test tube material

Table 4 :
Results of determination of contact angles between bentonite suspension and glass beads inSolid Earth Discuss., doi:10.5194/se-2016-42,2016 Manuscript under review for journal Solid Earth Published: 8 March 2016 c Author(s) 2016.CC-BY 3.0 License.