Summary of International Workshop

on Seismic Performance of Geosynthetic-Reinforced Soil Structures

(From Kobe to Ji-Ji and Izmit Earthquakes)

 

Hoe I. Ling, Columbia University

 

1. About Columbia Workshop

The International Workshop on Seismic Design of Reinforced Soil Structures (hereafter known as Columbia Workshop) was part of a project funded by the Geotechnical Program of the National Science Foundation, entitled “Seismic Performance and Design of Reinforced Soil Structures with Reference to Lessons Learned from the 1999 Earthquakes of Taiwan and Turkey” (P.I.s: Hoe Ling and Dov Leshchinsky; Grant No. CMS-0084449). The Workshop was a 1.5-day event with the speakers and panelists from leading institutions in the United States, Canada, Japan, Taiwan and Turkey.  About 40 researchers and engineers attended the Workshop.

 

The study on the seismic performance of reinforced soil structures has received much attention following 1995 Kobe Earthquake, and especially with the failures reported during Ji-Ji Earthquake in Taiwan. In the past years, NSF and some state DOTs have awarded a few research projects on this subject. In Japan, works were initiated by different groups of researchers at the institutions and government agencies. There have been rather limited discussions in progress among the researchers working on this subject. Therefore, there was a need to create an opportunity to learn from past experience, to be aware of the on-going achievement in this field, and to allow for intensive discussions by identifying the problems and seeking rooms to advance our knowledge.

 

The Workshop was held at the Faculty House of Columbia University on October 30 and 31, 2000. It was opened by Prof. Rimas Vaiacitis, Chair of the Department of Civil Engineering and Engineering Mechanics, Columbia University. The first day consisted of 14 invited presentations. The information is also available at the webpage http://www.civil.columbia.edu/~ling/nsf. The one-page abstract with a representative publication related to individual presentation is included in this Volume. If a full paper has been submitted, it is mentioned under representative publication. A proceedings with collection of already-published papers of the speakers were distributed. During the second day, three discussion sessions were organized that focused on the Design Issues, Testing and Experimental Studies, and Numerical Method and Modeling.

 

It is emphasized that the case histories help to advance our state-of-the-art, and they are by no means to discourage our use of existing innovative technologies or developing new technologies. The experience learned from this Workshop enabled a special technical session to be organized for the International Conference on Geosynthetics in 2002 at Nice, France.

2. Presentations

Seismic Performance of Reinforced Soil Retaining Walls

High Seismic Performance of Prestressed Geosynthetic-Reinforced Soil Structure Equipped with a Ratchet System

Seismic Design of Geosynthetic Reinforced Slopes

Observation of Segmental Retaining Walls in Chi-Chi Earthquake

The Reinforced Slope Failure of National Chi-Nan University During Chi-Chi Earthquake in Taiwan

Probabilistic Characterization of the Chi-Chi Earthquake Ground Motion

Model Tests on Seismic Stability of Several Types of Retaining Walls

Applications of Centrifuge Modeling to Dynamic Response of Reinforced Soil Walls

Seismic Response of Reinforced Slopes

Reinforced Slopes Subjected to Seismic Loads

Uncertainty in Soil Properties and Its Effects on Some Geotechnical Engineering Problems

Centrifuge and Full Scale Models of Geotextile Reinforced Lime Stabilized Cohesive Soil Walls and Two Case Studies of Segmental Retaining Walls in Turkey

Geosynthetics for Earthquake Hazard Mitigation

Centrifuge Modeling of the Seismic Performance of Slopes Reinforced with Geotextile Strips

 

3. Discussion Sessions

Design Issues (Session leaders: Dov Leshchinsky and Jerry DiMaggio)

Testing and Experimental Studies (Session Leader: Fumio Tatsuoka, University of Tokyo)

Numerical Method and Modeling (Session leader: Andrew Whittle, M.I.T.)

 

Seismic Performance of Reinforced Soil Retaining Walls


Richard J. Bathurst, Royal Military College, Canada



The paper is focused on experimental and numerical work recently completed or underway at RMCC that investigates the seismic performance of metallic and geosynthetic reinforced soil retaining walls. The experimental work involves the construction of 1m-high reduced scale reinforced retaining walls constructed on a shaking table and subjected to staged base excitation. The walls were built with different footing conditions, wall batter and reinforcement configurations. Examples of FLAC modeling used to simulate the experimental results are also presented. A summary of recent numerical investigations of the seismic response of reinforced soil walls will be summarized. These investigations have systematically investigated the effects of fundamental frequency, wall geometry, reinforcement configuration, and ground motion history on dynamic response of walls. Implications of the research at RMCC on seismic design and performance of reinforced soil walls are identified.

 

Representative publication:

Bathurst, R.J. and Simac, M.R. (1997). “Design and performance of the facing column for geosynthetic reinforced segmental walls.” Mechanically Stabilized Backfill, Wu, ed., Balkema, Rotterdam, 193-208.

 

High Seismic Performance of Prestressed Geosynthetic-Reinforced

Soil Structure Equipped with a Ratchet System


Tatsuoka, F., Department of Civil Engineering,

University of Tokyo



Geosynthetic-reinforced soil structures used as critical structures, such bridge abutments and piers, can become very stable even against very high seismic loads, such as those the GRS walls were subjected to during the 1995 Kobe Earthquake and the 1999 ChiChi Earthquake, by means of prestressing. The prestressing is applied vertically by using vertical tie rods extending from below to above the backfill of GRS structure. By using a ratchet system for locking the top of tie rods, the prestress can be kept nearly constant even when the height of backfill tends to decrease due to cyclic straining, while the height of backfill is kept constant when the height of backfill tends to increase by, for example, bending deformation and dilatancy of the backfill. By the latter function, the backfill behaves under essentially constant volume conditions, which make the ultimate strength of compacted backfill very high. A series of model shaking model tests and cyclic loading model tests were performed to validate the above. A protoype ratchet system has been developed for use with full-scale structures.

 

Representative Publication:

Tatsuoka, F., Koseki, J., Tateyama, M., Munaf, Y., and Horii, K. (1998). “Seismic stability against high seismic loads of geosynthetic-reinforced soil retaining structures.” Proceedings of Sixth International Conference on Geosynthetics, 103-142.

 

Seismic Design of Geosynthetic Reinforced Slopes

 

Holtz, R.D., Kramer, S.L., and Taylor, T.A., Department of Civil Engineering,

University of Washington, Seattle

 

The objectives of the study were to increase our fundamental understanding of the seismic behavior of geosynthetic-reinforced slopes and to develop realistic displacement-based models for the practical seismic design of these structures. Model tests were conducted in a geotechnical centrifuge and on a large shaking table, and the results were modeled using FLAC. These analyses enabled the development of physically-reasonable simplified models of the seismic behavior of reinforced soil slopes that can be used for design.

Twenty model tests of wrapped-face reinforced steep slopes at 1/30 scale of a 6 m high prototype were subjected to in-flight shaking in the RPI geotechnical centrifuge. These tests indicated that horizontal deflections increased and the yield acceleration at mid-height of the slope decreased with decreasing reinforcement tensile strength and increasing reinforcement vertical spacing. The mode of failure appeared to be a sliding block type failure.

Similar results were found in a series of 1-g model test of steep slopes 1.2 m high conducted on the UW shaking table. In one series of 10 tests, a constant strength geotextile just strong enough to provide a static FS @ 1 was used, and the L/H, reinforcement spacing and lift thickness were varied. In the second series of 12 shaking table tests, variables included the reinforcement spacing, embedment length and tensile strength and stiffness. Instrumentation showed that the failure surface was in general a bilinear sliding block type failure with its exact shape of a function of the strength, spacing, and length of the reinforcement. Failure began at the top slope of the slope, just behind the reinforced zone, and progressed downward at an angle until heights of h/2 to h/4 above the base. Then the surface entered the reinforced zone at a flatter angle and exited the face of the slope between the first and third layers of reinforcement. The soil above the failure surface appeared to move almost as a rigid body.

Yield accelerations increased with increasing slope FS, specifically with increasing reinforcement strength and stiffness and decreasing reinforcement spacing. With all other factors held constant, the shaking table tests indicated that displacement per cycle increased with (a) increasing base acceleration amplitude, (b) decreasing reinforcement length, (c) increasing reinforcement spacing, and (d) decreasing reinforcement strength and stiffness. With increasing reinforcement spacing strength and stiffness, the reinforced soil slope also tended to amplify the input base motion; the effect was most pronounced at the crest of the slope.

The FLAC results were generally consistent with those observed in the model tests – a consistent pattern of deformation develops in geosynthetic-reinforced slopes subjected to seismic loading. Failure surfaces were clearly bilinear, and the soil (and reinforcement) above the failure surface appeared to behave as a rigid block. Thus a simplified Newmark-type model was developed to predict the general level of deformation of a reinforced slope. Procedures for the identification of model input parameters from material and geometric properties. The new model runs on a personal computer in a matter of seconds and predicts behavior consistent with that observed in the model tests produced by FLAC analyses (which can take considerable longer run periods). Thus the method is a useful design tool for the seismic stability of geosynthetic-reinforced slopes.

 
Observations of Damages of Segmental Retaining Walls in Chi-Chi Earthquake

 

Lee, S S-H., National Chi-Nan University, Taiwan

 

 

Some segmental retaining walls (SRWs) have been found fail in the 921 Chi-Chi earthquake. Many criticism focused on the weakness of shear pins between vertically adjacent units were raised. Although most fail walls are moderate- to high-height SRWs, one can see a typical MCU geogrid-reinforced wall, with heights less than three meters, budged out at middle height. The vertical spacing of geogrid layers in this wall was found as much as 80 centimeters. So, the writer regards that too large vertical distance between reinforcements are the major reason for failures of SRWs.

 

For low- to moderate-height reinforced walls, the depth dimension of modular concrete unit as well as vertical connection mechanism between MCUs can provide significant contribution to the stability of a SRW. However, the lateral earth pressure of natural soils in a wall higher than four meters will be much larger than the strength of regular shear pins. A SRW needs zero earth pressure, which is due to the small vertical spacing of reinforcements, at the facing of a geogrid-reinforced wall in this situation. An important thumb-of-rule is that the smaller the vertical spacing, the higher the apparent cohesion in a mechanically stabilized soil.

             

Fortunately, there are many other SRWs in central Taiwan area successfully survived after 921 Chi-Chi earthquake. They are mostly constructed with vertical spacing in 40 cm typically. That is why they are still alive.

 

Representative publication:

Lee, H.H., (1997) “Recent Development And Research Of Geosynthetics In Taiwan.” Proceedings of Second Conference on Geotechnical Problems of Rapidly Developing Areas in China, Guangzhou, China, 77-83.

 

Note: This abstract was edited as a short form of the writing submitted to the Workshop.

Failure Investigation of The Reinforced Slope at The National Chi-Nan University

During The 1999 Chi-Chi Earthquake

 

Chou, N.N.S. and Fann C-C., Genesis Group/Taiwan, Taipei, Taiwan

 

 

The reinforced slope, which is 60 to 70 meters high, located at the entrance of the National Chi-Nan University in Puli collapsed during the 1999 Chi-chi earthquake. Although the Chi-chi earthquake is the most severe earthquake ever for the past 100 years in Taiwan, geologic condition at the site and some design aspects may also played a role in the failure of the reinforced slope. Failure investigation and dynamic simulation of the reinforced slope were conducted. The result shows that the slip surface took place along a thin layer of clayey material. The reinforced slope with low ratio of reinforcement length to height of the slope is critical to the stability of the slope. In addition, the recompacted backfill on the upslope of the reinforced slope may also decrease the overall stability of the slope.

 

Representative publication:

Full paper included in Reinforced Soil Engineering: Advances in Research and Practice

Probabilistic Characterization of The Ch-Chi Earthquake Ground Motion

 

Smyth, A.W., Department of Civil Engineering and Engineering Mechanics,

Columbia University

 

 

A previously developed procedure to condense nonstationary random excitation data to perform analytical random vibration response studies is used to investigate the 1999 Chi-chi earthquake recorded ground motions. An ensemble of free-field ground motion records from the main earthquake event collected from locations near the Chelungpu fault was used to create the second order statistics of the earthquake excitation. Using the compaction procedure, the covariance matrix of the excitation process was spectrally decomposed by the Karhunen-Loeve expansion. The dominant eigenvectors, i.e., those with the largest eigen values, represent the dominant features of the earthquake process. Second order descriptions of the transient dynamic response of discrete systems to the compact form of the earthquake process are obtained. This type of result can be used to facilitate improved design standards for civil structures, and to perform reliability studies. A comparison is made of the preliminary results of this study and those obtained from a similar analysis performed on an ensemble of ground motions from the 1994 Northridge earthquake.

 

Representative publication:

Full paper included in Reinforced Soil Engineering: Advances in Research and Practice

 

Model Tests on Seismic Stability of Several Types of Retaining Walls

 

Koseki, J., Institute of Industrial Science,

University of Tokyo



A series of irregular shaking tests was conducted on retaining wall models consisting of six different types. In some tests, after the first failure plane was formed in the backfill, the second failure place was formed at higher seismic loads. This can be explained by considering the effects of strain localization in the backfill soil and associated post-peak reduction in the shear resistance from peak to residual values along a previously formed failure plane. Such behavior has not been observed in the tilting tests and the sinusoidal shaking tests that were conducted on the same models in the previous study. In these tests, reinforced-soil retaining wall models with a rigid full-height facing exhibited a ductile behavior compared to conventional type retaining wall models such as gravity-type, leaning-type and cantilever-type ones. When the model walls started to tilt, concentration of subgrade reactions at the toe of conventional type retaining walls resulted in local failure due to loss of bearing capacity. On the other hand, under similar conditions, tensile force in the reinforcements of the reinforced-soil retaining walls could be mobilized effectively to resist against the tilting displacement.

 

Representative publication:

Full paper included in Reinforced Soil Engineering: Advances in Research and Practice

 

Applications of Centrifuge Modeling to Dynamic Response of Reinforced Soil Walls

Fragaszy, R.J., Division of Civil & Mechanical Systems,

National Science Foundation



This presentation describes the use of centrifuge modeling to investigate the behavior of reinforced soil walls subjected to explosive loading. Ten nominally identical centrifuge tests were conduced at 1/30 scale to determine the reliability of test data. Statistical analyses were performed to determine mean, standard deviation and coefficient of variation on outward wall displacement, wave speed, peak soil/wall interface pressure, and peak wall acceleration. Additional tests were then performed in which modifications were made to the wall, including varying soil density, number of reinforcement layers and length of reinforcement. The purpose of these tests was to determine which parameters had a significant influence on wall behavior. Soil density was found to have by far the most influence on dynamic wall response. The advantages of first obtaining a statistical data base are described. Also described is the use of centrifuge testing to support questionable full-scale field data.

Representative publication:

Fragaszy, R.J., Purcell, M., Olen, K.L., and Brownell, K.C. (1994). “Assessing the reliability of results in a dynamic centrifuge tests.” Centrifuge 94, Leung, Lee and Tan., eds., Balkema, Rotterdam, 313-318.

 

Seismic Response of Reinforced Soil Slopes

 

Sitar, N., Department of Civil and Environmental Engineering,

University of California-Berkeley

 

 

Documented case histories of seismic field performance of reinforced soil structures showed that reinforced soil slopes and walls tend to perform well under earthquake loading. However, field reports point out a lack of monitoring in practice, making it difficult to validate seismic design assumptions. In general, the structures tend to be flexible and deform without reaching catastrophic failure. A review of previous experimental studies also shows the inherent flexibility of reinforced soil under dynamic loading. In fact, in most of the studies, the walls were able to maintain their integrity even under severe shaking. Thus, both field and experimental data show that deformations need to be considered in the seismic design of these structures. Centrifuge tests were used to study the dynamic behavior of soil slopes reinforced with geosynthetics. The main objectives were to determine the failure mechanism and amount of deformations under seismic loading, and to identify the main parameters controlling seismically-induced deformations. Geosynthetically reinforced slopes (2V:1H) were subjected to earthquake motions with maximum foundation accelerations of up to 0.86g. The experimental results show that slope movement can occur under relatively small base accelerations, and significant lateral and vertical deformations can occur within the reinforced soil mass under strong shaking. However, no distinct failure surfaces tend to occur, and the magnitude of deformations is directly related to the backfill density and reinforcement stiffness.

 

Representative publication:

Nova-Roessig, L. and Sitar, N. (1999). “Centrifuge model studies of the seismic response of reinforced soil slopes.” Proceedings of Second International Conference on Earthquake Geotechnical Engineering, Seco and Pinto, eds., Balkema, Rotterdam.

Reinforced Slopes Subjected to Seismic Loads


Michalowski, R.L., Department of Civil and Environmental Engineering,

University of Michigan

 


The kinematic theorem of limit analysis can be easily used to evaluate the amount of reinforcement necessary to prevent collapse of slopes. The analysis based on this approach entails global equilibrium of the soil mass with seismic influence included as quasi-static forces. The computational results will be presented in form of charts. While this approach is routinely used in practice, it does not reflect the earthquake shaking process, and it does not provide any information about permanent displacements that may have occurred during that process. The interest in analyses of slopes subjected to seismic loads was renewed after recent earthquakes. This presentation will focus on displacements calculations of reinforced slopes. Design of reinforced slopes using quasi-static approach may lead to unrealistically long reinforcement for large ground accelerations. If a slope is allowed to move by even a small displacement, then the reinforcement length can be significantly reduced. Two mechanisms of failure of reinforced slopes subjected to seismic conditions are considered: rotational collapse, and sliding directly over the bottom layer of reinforcement. Calculations are separated into those related to the record of seismic acceleration and those dependent on the structure itself. An example will be shown to illustrate the results, and charts will be presented to indicate convenient application of the results in design.

 

Representative publication:

Michalowski, R.L. and You, L. (2000). “Displacements of reinforced slopes subjected to seismic loads.” Journal of Geotechnical and Geoenvironmental Engineering, ASCE, 126(8), 685-694.

Uncertainty in Soil Properties and Its Effect on Some Geotechnical Engineering Problems

 

Deodatis, G., Department of Civil Engineering,

Princeton University



Although soil properties are known to exhibit very significant spatial variability, there has been relatively little effort until now to establish probabilistic models describing this variability and to quantify its effect on various geotechnical problems. This study presents some models that have been developed recently to model the stochastic spatial variability of certain soil properties based on in-situ measurements of natural and artificial deposits. Specifically, these models include marginal probability distribution functions for the uncertain properties and correlation functions to describe their variability in space. These models are then used in some traditional geotechnical earthquake engineering problems including slope stability analysis and risk assessment of interacting soil-structure systems due to liquefaction. The results are provided in terms of fragility curves indicating the probability of exceeding a certain damage level as a function of the intensity of the seismic event. Studying these results, it becomes immediately obvious that the random spatial variability of soil properties has a very significant effect on these problems. In addition, it is shown that it is impossible to capture certain attributes of the system behavior by assuming homogeneous soil properties.

 

Representative publication:

Popescu, R., Prevost, J.H., and Deodatis, G. (1997). “Effects of spatial variability on soil liquefaction: some design recommendations.” Geotechnique, 47(5), 1019-1036.

Centrifuge and Full Scale Models of Geotextile Reinforced Lime Stabilized Cohesive Soil Walls and Two Case Studies of Segmental Retaining Walls in Turkey

 

Guler, E., Department of Civil Engineering,

Bogazici University, Istanbul, Turkey

 


Thirteen centrifuge models of lime stabilized kaolin geotextile reinforced soil retaining walls were tested to failure by increased self-weight. Three failure models were identified depending on the length of the geotextile reinforcement. The addition of lime substantially improved the stability even when the geotextile length was equal to only one half of the wall height.


Based on these findings a full height reinforced soil retaining structure was constructed. Lime stabilized cohesive clay was used as the backfill material and the wall was instrumented to determine stress distribution in the soil and the reinforcement. The Factor of Safety was chosen as unity in the design and surcharge loads were applied in order to bring the wall to failure. However it was not possible to fail the wall.


Two geotextile reinforced segmental retaining walls were constructed for the Turkish Highway Authority. The wall in Istanbul is at its maximum point 10 m high and its reinforcement was instrumented with strain gauges. The other wall is constructed in Antalya and is a tiered wall with a total height of 6 m. The walls are in the first and second-degree earthquake zones respectively.

 

Representative publication:

Full paper included in Reinforced Soil Engineering: Advances in Research and Practice

Geosynthetics for Earthquake Hazard Mitigation

Yegian, M.K., Department of Civil and Environmental Engineering,

Northeastern University



Under dynamic excitations, slip deformations occur along smooth geosynthetic interfaces, thus reducing the energy transmitted through the interfaces. Geosynthetic liners placed under foundations can absorb seismic energy, and hence transmit smaller ground motions to an overlying structure. This concept of using geosynthetics for Foundation Isolation is similar to mechanical isolators used in structural engineering. In addition, using geosynthetic liners at different depths within new slopes, embankments or hydraulic fills - known to be extremely vulnerable to liquefaction - can reduce the soil amplification effect on the ground motion and the liquefaction potential of loose fills. This concept is referred to as Soil Isolation.


A research program, funded by the National Science Foundation, is investigating the technical feasibility, practicality, and limitations of using geosynthetics for foundation and soil isolation to mitigate earthquake hazard.


The scope of this research includes shaking table tests on various geosynthetics that show the best promise for foundation and soil isolation. Constitutive models are being developed that describe the dynamic response of geosynthetic interfaces. Model tests on buildings and soil profiles are conducted using the shaking table, and the results are utilized to develop procedures for analysis of the dynamic response of structures and soil profiles that are isolated using geosynthetic liners.


The innovative concept of using geosynthetic liners to reduce earthquake ground motion intensity, can be a very cost effective and simpler alternative to conventional earthquake hazard mitigation measures.

 

Representative publication:

Yegian, M.K., Kadakal, U., and Catan, M. (1999). “Geosynthetics for earthquake hazard mitigation.” Proceedings of Geosynthetics ‘99, 87-100.

 

Centrifuge Modeling of the Seismic Performance of Slopes Reinforced

with Geotextile Strips

Pamuk, A., and Zimmie, T.F., Department of Civil and Environmental Engineering,

Rensselaer Polytechnic Institute


This project utilizes dynamic centrifuge modeling. The research focuses on the feasibility, performance and testing of an innovative reinforcement method to be applied to natural slopes and earth structures (e.g., slopes and embankments) which contain low quality soft soils, located in seismic regions. The proposed method will be applied to the structures in a manner similar to conventional soil nailing, and will increase the stability of marginally stable slopes in two ways; that is, by introducing non-woven geotextile strips which act as reinforcement, and also serve as a drainage media, thus decreasing pore pressure build-up.


Cohesive soil models and geotextiles can be tested using centrifuges by considering proper scaling laws. However, in general it is difficult to simulate soil reinforcing processes (i.e., the model has to be constructed prior to testing) as it actually progresses in the field. Therefore, a remotely controlled miniature mandrel driver capable of inserting geotextile strips horizontally or inclined into the model clay slopes is used, while the centrifuge was spinning. After the strips are inserted at a predetermined acceleration, the reinforced soil models are subjected to an input earthquake motion utilizing an in-flight shaker, until the slope fails. The input motions use increasing amplitudes or constant amplitudes with increasing number of cycles, and are applied to slopes reinforced with various numbers of geotextile strips.


Stability is dependent on the type of input motion, length, spacing and orientation of the geotextile strips. Deformations of the reinforced slopes were less than unreinforced slopes, showing that the method can be effective in increasing the factor of safety of marginally stable slopes subject to earthquake loads. The centrifuge tests revealed realistic data about the seismic performance of the new type of mechanically stabilized earth system, and offered an inexpensive and practical alternative to large-scale field tests. The results obtained can be useful for the design of actual prototypes.

 

Representative publication:

Mahmud, M.B. and Zimmie, T.F. (1997). “Centrifuge modeling of a rapidly installed mechanically stabilized earth system using geotextile strips.” Proceedings of Fourteenth International Conference on Soil Mechanics and Foundation Engineering, Hamburg, Germany, 1765-1768.

Presentation and Discussion Session Chairs

Prof. Ling chaired the morning presentations

Prof. Leshchinsky chaired the afternoon presentations

Prof. Tatsuoka chaired one of the discussion sessions

Prof. Whittle chaired one of the discussion sessions

Prof. Bathurst was a panelist in all three discussion sessions

Mr. DiMaggio chaired one of the discussion sessions

 

 

Speakers

(in the order of presentation)

Prof. Richard Bathurst

Prof. Fumio Tatsuoka

Prof. Robert Holtz

Prof. Shannon Lee

Dr. Nelson Chou

Prof. Andrew Smyth

Prof. Junichi Koseki

Dr. Richard Fragaszy

Prof. Nick Sitar

Prof. Radoslaw Michalowski

Prof. George Deodatis

Prof. Erol Guler

Prof. Mishac Yegian

Prof. Thomas Zimmie

Mr. Mike Adams