evaluation of crack generation formed by local
EVALUATION OF CRACK GENERATION FORMED BY LOCAL SUBSIDENCE USING THE LARGE-SCALE BENTONITE-MIXED SOIL LAYER MODEL SADAHIKO USAMI*, KENJI SHIBATA**, JOJI HINOBAYASHI*** * Yachiyo Engineering CO., LTD., Asakusabashi 5-20-8 Taito-ku Tokyo,JAPAN ** Obayashi CO., LTD., Konan 2-15-2 Minato-ku Tokyo, JAPAN *** Dainippon Plastics CO., LTD., minorudai 5-1-1 Matsudo City Chiba, JAPA
SUMMARY: In Japan, bentonite mixed soil (BMS) is widely used as landfill leachate barrier system. By local subsidence of the foundation ground, there is a risk that a crack is generated in BMS’s deformation. Therefore, it is important to understand the changes in the water barrier capability due to local subsidence of BMS. Therefore, we carried out an experiment that simulates the occurrence of local subsidence by the life-size BMS layer, and evaluated the behavior by using the elastic-plastic FEM analysis. Model BMS layer is thickness 50 cm, width 300 cm, and depth 150cm. And lower center portion of BMS layer, the groove was provided that is depth of 20cm, width of 120cm. This groove was backfilled with sand to scrape the sand for generated subsidence. Crack is not generated for BMS layer until the groove width 90cm. When the groove width reaches 105 ~ 120cm, crack was progress from BMS layer underside to 40 ~ 45cm. BMS was constructed divided into three layers. In initial stage of subsidence, three-layer BMS had behaved as a unit. However, as the subsidence progresses, delamination each layer occurs, and found to lead to destruction. And when laying Geo-net under the bottom of the BMS layer, it was found that an effect of suppressing shear deformation of the BMS.
1. INTRODUCTION In Japan, bentonite mixed soil (BMS) which is made by mixing the bentonite into sand about 10 to 15 % in dry mass in order to have a hydraulic conductivity of less than 1x10-8 m/s is widely used as soil barrier layer of landfill1). The BMS layer may experience local subsidence of the base ground. In case of large local subsidence, the BMS could crack to lose the integrity of the barrier system. The authors constructed five large models of BMS layer in the field which had a thickness of 0.5 m, a width of 1.5 m and a length of 3.0 m. Among them, four models were underlaid with a geo-net having a lattice width of 10 mm. As results it is found that the crack appeared on the bottom surface of the BMS layer at the edge of the groove and progressed upward at an angle of about 60 degree to the horizon. The models underlaid with geo-net had also cracked but did not progress to the top surface while those without geo-net had a crack reaching the top surface.In this paper, we report to reproduce the results using the elasticplastic FEM analysis.
Proceedings Sardinia 2017 / Sixteenth International Waste Management and Landfill Symposium/ 2 - 6 October 2017 S. Margherita di Pula, Cagliari, Italy / © 2017 by CISA Publisher, Italy
Sardinia 2017 / Sixteenth International Waste Management and Landfill Symposium / 2 - 6 October 2017
2. USING MATERIALS 2.1 Bentonite mixed soil (BMS) As the base material of the BMS, the maximum particle diameter of 4.75mm, gravel content of 13.4%, sand content of 84.0%, the fine particle fraction of 2.5% of crushed stone sand was used. It was prepared BMS was added to a 10% Na-type bentonite on dry weight ratio thereto. 2.2 Geo-net (GN) GN are made from high-density polyethylene, were used two types. The first type is referred to as N-24, the lattice width 10 mm, shield 49%, a tensile strength 9,120 N/m. The second type is referred to as the N-248, it has 85% tensile strength of the N-24.
3. STRUCTURE OF MEDEL BMS LAYER Model BMS layer has a thickness of 0.5 m, width 3.0 m, a width 1.5 m. As shown in Figure-1, in the central part under the BMS layer, it was provided with a groove depth of 10cm or 20cm in width 120cm. The excavated width is the same length as the excavated groove width of the groundwater drain pipe installed under the actual water barrier layer, in which to be able to simulate the subsidence of excavation groove portion. It was installed BMS layer this part backfilled in a steel tube and sand. Experimental cases was shown in Table-1. Table-1 Experimental case of model BMS layer2) Conten ts Geonet
None20
None10
N24820
N24810
N2420
N2410
none
none
N248
N248
N24
N24
20cm
10cm
Protective cover soil 50cm BMS layer 50cm
Sand
Steel pipe
Base ground stabilized with cement
Partition plate Geo-net(if necessary) Depth 20cm
120cm 300cm
Groove depth
20cm
10cm
20cm
Figure-1 Structure of the model of the BMS layer
10cm
Sardinia 2017 / Sixteenth International Waste Management and Landfill Symposium / 2 - 6 October 2017
4. The RESULTS OF BMS LAYER LOCAL SUBSUDENCE EXPERIMENT As shown in Figure-2, in all experiments case, at the time of the groove width 90cm did not crack occurs in the BMS layer. When the groove width reaches 105 ~ 120cm, crack was progress from BMS layer from underside to 40 ~ 45cm. Further during 44kN loading, in other cases of laying the GN of N248, cracks were penetrated to the BMS layer top surface. In addition, BMS layer was created separately three layers from the bottom 15cm, 15cm, and 20cm. Therefore, as shown in Figure-3 (a), in the groove 20cm case without GN, while removing the sand groove width is transferred to 120cm from 105cm, detached with BMS layers occurs. As shown in Figure-3 (b), in the case of the groove 20cm was with N-248, detached with the groove width 120cm at BMS layers occurs.
44k N Load ing
Figure-2 Progress of crack vs. Groove width
(a) GN-None、Groove 20cm
(b) N24、Groove 20cm
Figure-3 Situation of delamination
5. FEM ANALYSIS 5.1 Analysis Model BMS layer subsidence experiment has been carried out in the plane strain state. Therefore, FEM analysis was also carried out in a two-dimensional plane strain conditions. Subsidence was expressed by removing the element of the groove (green in Figure-4). It shows an input constants of FEM analysis in Table-2. Input constants of BMS were determined from uniaxial compression test and direct shear test. BMS analysis using Drucker-Prager model, others were modelled as an elastic materials.
Sardinia 2017 / Sixteenth International Waste Management and Landfill Symposium / 2 - 6 October 2017
Table-2 Input constants of FEM analysis Layer Protective soil BMS BMS weak layer Geo-net Base ground
Wet Density ρt (g/cm3) 1.90 1.95 1.95
Modulus of Adhesive deformation force (MN/m2) c(kN/m2) 14.7 14 21.5 14 2.15
0.95 1.90
Internal friction angle φ(°) 32.7 3.7
112 24.5
Poisson's ratio ν 0.4 0.4 0.4 0.4 0.35
5.2 Analysis case The analysis cases were shown in Table-3. In the experiment, BMS layer was detached off at the boundary of the compacted layer. Therefore, as shown in Figure-4, it was provided with a weak layer of strength 15cm and 30cm position from the bottom of the BMS, and it was modelled as a state in which overlapping three layers. In addition, in the analysis of the model that the model BMS layer of the bottom 15cm were considered to subsidence (Single Layer Model), or 30cm BMS layer was considered to subsidence (Two-Layer Model), the load of BMS layer and cover soil layer over the analysis layer, It was considered to loading by a uniformly distributed load. The strength of the weak layer of strength is referred to as the "BMS inter-strength", the C and φ were set to 1/10 of the BMS strength. Table-3 Analysis cases of FEM Case Case-1 Case-2 Case-3 Case-4
Analysis Model BMS 3layers+Protective soil BMS lower single layer BMS lower two layers BMS lower single layer
Geo-net Protective Soil 50cm
None 3rd BMS layer 20cm
placed
2nd BMS layer 15cm 1st BMS layer 15cm Base ground
←weak layer ←weak layer
Groove depth 20cm
Figure-4 Structure model for analysis 5.3 Results of analysis In Case-1 was modelled as a state in which the BMS layer are overlapped three-layer, even in the groove width under the BMS layer is 120cm, subsidence amount of BMS layer was up to 3.2mm. This results have not been able to reproduce the experimental results. Figure-5 shows the subsidence amount of bottom-center of BMS layer by experiment and analysis. Analysis results shown in Figure-5 were Case-2 and Case-3 without GN, and Case-4 with GN. Subsidence of the groove width 135cm in Case-3 is a diagram showing a BMS subsidence of the second layer bottom after removing an element of the BMS first layer component.
Sardinia 2017 / Sixteenth International Waste Management and Landfill Symposium / 2 - 6 October 2017
Analysis results in Case-2 intended for the BMS 1st layer, subsidence amount of up to groove width 60cm is a few millimeters, and subsidence amount of 2.15cm when it comes to the groove width 75cm has occurred, and the subsidence of 10cm in the groove width 90cm going on. This subsidence amount is greater than the subsidence amount immediately before the first layer BMS was collapse in the experimental. The maximum shear stress is generated at both ends of the groove (show Figure-6). In the experiment, 4.6 cm subsidence occurs in the groove width 105cm in cases without GN. As the groove width is increased, the delamination of BMS layers is in progress, BMS layer began to change in the simple overlapped state of the three-layer or double-layer. Analysis results are considered to represent such a phenomenon. In the analysis results of the BMS double-layers (Case-3) is a subsidence of about 1.3cm be groove width reaches 120cm. In the stage where the bottom first layer and the second layer formed thereon were behaving as a unit, it can be estimated that does not reach the state as leading to destruction. However, after the first layer has collapsed, the second layer has been collapse in a short time, it shows a similar behavior with the experimental situation in the analysis. To compare the difference of subsidence depending on with and without of GN. Until the groove width 60cm, there is no difference in the subsidence amount depending on with and without of GN. When it comes to the groove width 75cm, subsidence amount in the Case-2 of without GN is 2.2cm, and subsidence amount in the Case-4 of with GN is 1.3cm, the difference in subsidence amount depending on with or without of GN occurs. In the groove width 90cm, subsidence amount in the case of without GN is 9.8cm, it is about twice of the subsidence amount in the case of with GN (subsidence amount = 5.4cm). In previous studies3), the authors ware modelled as a fixed-fixed beam experimental results produced delamination, and estimates the amount of deflection and the generated stress. As a results, in the liable to occur delamination state (normal construction), It has been estimated that the allowable shear stress of BMS layer of 15cm layer thickness is 30kN/m2, the allowable bending stress is 200kN/m2, allowable deflection rate is about 0.4%. Based on this allowable shear stress, and seek a groove width BMS layer reaches the allowable value from Figure-7 and Figure-8, It is assumed that is the groove width 75 ~ 90cm in without GN of Case-2, the groove width 90 ~ 120cm is with GN of Case-4. When laying the GN, groove width of up to shear stress reaches the allowable value tend to be l arger than without GN, the effect of suppressing is look at the shear deformation.
44k N Loa ding
Sardinia 2017 / Sixteenth International Waste Management and Landfill Symposium / 2 - 6 October 2017
Figure-5 Relation of Groove width and subsidence
Figure-6 Relation of Groove width and subsidence on Case-2
Figure-7 Relation of Groove width and maximum shear stress Figure-8 Maximum shear stress at collapse
44 kN Lo ading
Sardinia 2017 / Sixteenth International Waste Management and Landfill Symposium / 2 - 6 October 2017
6. CONCLUSION They were compared with BMS layer local subsidence experiment and FEM reproducible analysis. As a results, it was found that the following. a) BMS three layers had behaved as an integral initially, the groove width becomes large, which led to the change destroys the state of the mere superposition of three-layer or two-layer b) When GN laying, the groove width to reach the allowable shear stress become large, GN has an effect of suppressing shear deformation
AKNOWLEDGEMENTS The authors are also very thankful for their valuable advice from a member of Landfill System & Technologies Research Association of Japan, NPO.
REFERENCES Landfill system and technology association: Handbook for The Leachate Barrier System of Landfill, pp.38-45, pp.111-115, 2008 Imaizumi S.,Shibata K,.Usami S.,Hinobayashi J.,Kudo K.,Nonoda M.: Crack Evaluation of the bentonite mixed soil layer undergoing local settlement (part 2), Proceedings of 25th Japan Waste Resource Recycling Society Symposium, pp.389-390, 2014 Usami S.,Shibata K,. Hinobayashi J.: Crack Evaluation of the bentonite mixed soil layer undergoing local settlement (part 3), Proceedings of 26th Japan Waste Resource Recycling Society Symposium, pp.405-406, 2015
SUMMARY: In Japan, bentonite mixed soil (BMS) is widely used as landfill leachate barrier system. By local subsidence of the foundation ground, there is a risk that a crack is generated in BMS’s deformation. Therefore, it is important to understand the changes in the water barrier capability due to local subsidence of BMS. Therefore, we carried out an experiment that simulates the occurrence of local subsidence by the life-size BMS layer, and evaluated the behavior by using the elastic-plastic FEM analysis. Model BMS layer is thickness 50 cm, width 300 cm, and depth 150cm. And lower center portion of BMS layer, the groove was provided that is depth of 20cm, width of 120cm. This groove was backfilled with sand to scrape the sand for generated subsidence. Crack is not generated for BMS layer until the groove width 90cm. When the groove width reaches 105 ~ 120cm, crack was progress from BMS layer underside to 40 ~ 45cm. BMS was constructed divided into three layers. In initial stage of subsidence, three-layer BMS had behaved as a unit. However, as the subsidence progresses, delamination each layer occurs, and found to lead to destruction. And when laying Geo-net under the bottom of the BMS layer, it was found that an effect of suppressing shear deformation of the BMS.
1. INTRODUCTION In Japan, bentonite mixed soil (BMS) which is made by mixing the bentonite into sand about 10 to 15 % in dry mass in order to have a hydraulic conductivity of less than 1x10-8 m/s is widely used as soil barrier layer of landfill1). The BMS layer may experience local subsidence of the base ground. In case of large local subsidence, the BMS could crack to lose the integrity of the barrier system. The authors constructed five large models of BMS layer in the field which had a thickness of 0.5 m, a width of 1.5 m and a length of 3.0 m. Among them, four models were underlaid with a geo-net having a lattice width of 10 mm. As results it is found that the crack appeared on the bottom surface of the BMS layer at the edge of the groove and progressed upward at an angle of about 60 degree to the horizon. The models underlaid with geo-net had also cracked but did not progress to the top surface while those without geo-net had a crack reaching the top surface.In this paper, we report to reproduce the results using the elasticplastic FEM analysis.
Proceedings Sardinia 2017 / Sixteenth International Waste Management and Landfill Symposium/ 2 - 6 October 2017 S. Margherita di Pula, Cagliari, Italy / © 2017 by CISA Publisher, Italy
Sardinia 2017 / Sixteenth International Waste Management and Landfill Symposium / 2 - 6 October 2017
2. USING MATERIALS 2.1 Bentonite mixed soil (BMS) As the base material of the BMS, the maximum particle diameter of 4.75mm, gravel content of 13.4%, sand content of 84.0%, the fine particle fraction of 2.5% of crushed stone sand was used. It was prepared BMS was added to a 10% Na-type bentonite on dry weight ratio thereto. 2.2 Geo-net (GN) GN are made from high-density polyethylene, were used two types. The first type is referred to as N-24, the lattice width 10 mm, shield 49%, a tensile strength 9,120 N/m. The second type is referred to as the N-248, it has 85% tensile strength of the N-24.
3. STRUCTURE OF MEDEL BMS LAYER Model BMS layer has a thickness of 0.5 m, width 3.0 m, a width 1.5 m. As shown in Figure-1, in the central part under the BMS layer, it was provided with a groove depth of 10cm or 20cm in width 120cm. The excavated width is the same length as the excavated groove width of the groundwater drain pipe installed under the actual water barrier layer, in which to be able to simulate the subsidence of excavation groove portion. It was installed BMS layer this part backfilled in a steel tube and sand. Experimental cases was shown in Table-1. Table-1 Experimental case of model BMS layer2) Conten ts Geonet
None20
None10
N24820
N24810
N2420
N2410
none
none
N248
N248
N24
N24
20cm
10cm
Protective cover soil 50cm BMS layer 50cm
Sand
Steel pipe
Base ground stabilized with cement
Partition plate Geo-net(if necessary) Depth 20cm
120cm 300cm
Groove depth
20cm
10cm
20cm
Figure-1 Structure of the model of the BMS layer
10cm
Sardinia 2017 / Sixteenth International Waste Management and Landfill Symposium / 2 - 6 October 2017
4. The RESULTS OF BMS LAYER LOCAL SUBSUDENCE EXPERIMENT As shown in Figure-2, in all experiments case, at the time of the groove width 90cm did not crack occurs in the BMS layer. When the groove width reaches 105 ~ 120cm, crack was progress from BMS layer from underside to 40 ~ 45cm. Further during 44kN loading, in other cases of laying the GN of N248, cracks were penetrated to the BMS layer top surface. In addition, BMS layer was created separately three layers from the bottom 15cm, 15cm, and 20cm. Therefore, as shown in Figure-3 (a), in the groove 20cm case without GN, while removing the sand groove width is transferred to 120cm from 105cm, detached with BMS layers occurs. As shown in Figure-3 (b), in the case of the groove 20cm was with N-248, detached with the groove width 120cm at BMS layers occurs.
44k N Load ing
Figure-2 Progress of crack vs. Groove width
(a) GN-None、Groove 20cm
(b) N24、Groove 20cm
Figure-3 Situation of delamination
5. FEM ANALYSIS 5.1 Analysis Model BMS layer subsidence experiment has been carried out in the plane strain state. Therefore, FEM analysis was also carried out in a two-dimensional plane strain conditions. Subsidence was expressed by removing the element of the groove (green in Figure-4). It shows an input constants of FEM analysis in Table-2. Input constants of BMS were determined from uniaxial compression test and direct shear test. BMS analysis using Drucker-Prager model, others were modelled as an elastic materials.
Sardinia 2017 / Sixteenth International Waste Management and Landfill Symposium / 2 - 6 October 2017
Table-2 Input constants of FEM analysis Layer Protective soil BMS BMS weak layer Geo-net Base ground
Wet Density ρt (g/cm3) 1.90 1.95 1.95
Modulus of Adhesive deformation force (MN/m2) c(kN/m2) 14.7 14 21.5 14 2.15
0.95 1.90
Internal friction angle φ(°) 32.7 3.7
112 24.5
Poisson's ratio ν 0.4 0.4 0.4 0.4 0.35
5.2 Analysis case The analysis cases were shown in Table-3. In the experiment, BMS layer was detached off at the boundary of the compacted layer. Therefore, as shown in Figure-4, it was provided with a weak layer of strength 15cm and 30cm position from the bottom of the BMS, and it was modelled as a state in which overlapping three layers. In addition, in the analysis of the model that the model BMS layer of the bottom 15cm were considered to subsidence (Single Layer Model), or 30cm BMS layer was considered to subsidence (Two-Layer Model), the load of BMS layer and cover soil layer over the analysis layer, It was considered to loading by a uniformly distributed load. The strength of the weak layer of strength is referred to as the "BMS inter-strength", the C and φ were set to 1/10 of the BMS strength. Table-3 Analysis cases of FEM Case Case-1 Case-2 Case-3 Case-4
Analysis Model BMS 3layers+Protective soil BMS lower single layer BMS lower two layers BMS lower single layer
Geo-net Protective Soil 50cm
None 3rd BMS layer 20cm
placed
2nd BMS layer 15cm 1st BMS layer 15cm Base ground
←weak layer ←weak layer
Groove depth 20cm
Figure-4 Structure model for analysis 5.3 Results of analysis In Case-1 was modelled as a state in which the BMS layer are overlapped three-layer, even in the groove width under the BMS layer is 120cm, subsidence amount of BMS layer was up to 3.2mm. This results have not been able to reproduce the experimental results. Figure-5 shows the subsidence amount of bottom-center of BMS layer by experiment and analysis. Analysis results shown in Figure-5 were Case-2 and Case-3 without GN, and Case-4 with GN. Subsidence of the groove width 135cm in Case-3 is a diagram showing a BMS subsidence of the second layer bottom after removing an element of the BMS first layer component.
Sardinia 2017 / Sixteenth International Waste Management and Landfill Symposium / 2 - 6 October 2017
Analysis results in Case-2 intended for the BMS 1st layer, subsidence amount of up to groove width 60cm is a few millimeters, and subsidence amount of 2.15cm when it comes to the groove width 75cm has occurred, and the subsidence of 10cm in the groove width 90cm going on. This subsidence amount is greater than the subsidence amount immediately before the first layer BMS was collapse in the experimental. The maximum shear stress is generated at both ends of the groove (show Figure-6). In the experiment, 4.6 cm subsidence occurs in the groove width 105cm in cases without GN. As the groove width is increased, the delamination of BMS layers is in progress, BMS layer began to change in the simple overlapped state of the three-layer or double-layer. Analysis results are considered to represent such a phenomenon. In the analysis results of the BMS double-layers (Case-3) is a subsidence of about 1.3cm be groove width reaches 120cm. In the stage where the bottom first layer and the second layer formed thereon were behaving as a unit, it can be estimated that does not reach the state as leading to destruction. However, after the first layer has collapsed, the second layer has been collapse in a short time, it shows a similar behavior with the experimental situation in the analysis. To compare the difference of subsidence depending on with and without of GN. Until the groove width 60cm, there is no difference in the subsidence amount depending on with and without of GN. When it comes to the groove width 75cm, subsidence amount in the Case-2 of without GN is 2.2cm, and subsidence amount in the Case-4 of with GN is 1.3cm, the difference in subsidence amount depending on with or without of GN occurs. In the groove width 90cm, subsidence amount in the case of without GN is 9.8cm, it is about twice of the subsidence amount in the case of with GN (subsidence amount = 5.4cm). In previous studies3), the authors ware modelled as a fixed-fixed beam experimental results produced delamination, and estimates the amount of deflection and the generated stress. As a results, in the liable to occur delamination state (normal construction), It has been estimated that the allowable shear stress of BMS layer of 15cm layer thickness is 30kN/m2, the allowable bending stress is 200kN/m2, allowable deflection rate is about 0.4%. Based on this allowable shear stress, and seek a groove width BMS layer reaches the allowable value from Figure-7 and Figure-8, It is assumed that is the groove width 75 ~ 90cm in without GN of Case-2, the groove width 90 ~ 120cm is with GN of Case-4. When laying the GN, groove width of up to shear stress reaches the allowable value tend to be l arger than without GN, the effect of suppressing is look at the shear deformation.
44k N Loa ding
Sardinia 2017 / Sixteenth International Waste Management and Landfill Symposium / 2 - 6 October 2017
Figure-5 Relation of Groove width and subsidence
Figure-6 Relation of Groove width and subsidence on Case-2
Figure-7 Relation of Groove width and maximum shear stress Figure-8 Maximum shear stress at collapse
44 kN Lo ading
Sardinia 2017 / Sixteenth International Waste Management and Landfill Symposium / 2 - 6 October 2017
6. CONCLUSION They were compared with BMS layer local subsidence experiment and FEM reproducible analysis. As a results, it was found that the following. a) BMS three layers had behaved as an integral initially, the groove width becomes large, which led to the change destroys the state of the mere superposition of three-layer or two-layer b) When GN laying, the groove width to reach the allowable shear stress become large, GN has an effect of suppressing shear deformation
AKNOWLEDGEMENTS The authors are also very thankful for their valuable advice from a member of Landfill System & Technologies Research Association of Japan, NPO.
REFERENCES Landfill system and technology association: Handbook for The Leachate Barrier System of Landfill, pp.38-45, pp.111-115, 2008 Imaizumi S.,Shibata K,.Usami S.,Hinobayashi J.,Kudo K.,Nonoda M.: Crack Evaluation of the bentonite mixed soil layer undergoing local settlement (part 2), Proceedings of 25th Japan Waste Resource Recycling Society Symposium, pp.389-390, 2014 Usami S.,Shibata K,. Hinobayashi J.: Crack Evaluation of the bentonite mixed soil layer undergoing local settlement (part 3), Proceedings of 26th Japan Waste Resource Recycling Society Symposium, pp.405-406, 2015
