ISRM International Symposium - 6th Asian Rock Mechanics Symposium

Abstract:

The stability of slopes can be assessed by a factor of safety, which is usually determined by a limit equilibrium method. In order to evaluate the factor of safety, we need such strength parameters as cohesion and the internal friction angle. In the monitoring of slope stability, displacements are commonly measured. Therefore, the strength parameters should be evaluated from the measured displacements. Sakurai and Nakayama (1999) developed a back analysis procedure for determining the strength parameters from the measured displacements. The back analysis procedure was successfully developed by introducing both an “anisotropic parameter” and “critical shear strain”. Recently, on the basis of this back analysis procedure, a computer code called the “Universal Back Analysis Program for Slope Stability” (UBAPSS) has been completed. In this paper, UBAPSS is briefly described, and a case study is shown to demonstrate its applicability for assessing the slope stability at an open pit coal mine. The surface displacements were measured in the mine by GPS and the data were collected until failure occurred. The strength parameters were determined using the data obtained just one day before the failure occurred. The results of the back analysis indicate that the factor of safety becomes approximately 1.0. This means that the computer code UBAPSS is well applicable to the assessment of the stability of slopes.



1. INTRODUCTION

The stability of slopes can be assessed by a factor of safety, which is usually determined by a limit equilibrium method. In order to evaluate the factor of safety, we need such strength parameters as cohesion and the internal friction angle. However, determining the strength parameters is not an easy task, because soil and rock have numerous geological and geotechnical uncertainties. However, the question of how to evaluate the strength parameters from the measured displacements is raised. To answer this question, Sakurai (1992) developed a back analysis procedure which can determine (1) the mechanism of the slope deformation, whether it is sliding or toppling, (2) the location of sliding planes, and (3) the strength parameters, such as cohesion and the internal friction angle, from the measured displacements, even only from the surface displacements. The back analysis procedure was successfully developed by introducing both an “anisotropic parameter” and “critical shear strain”. Recently, on the basis of this back analysis procedure, a computer code called the “Universal Back Analysis Program for Slope Stability (UBAPSS) has been completed. In this paper, UBAPSS is briefly described, and a case study is shown to demonstrate the applicability of the proposed back analysis procedure for assessing the slope stability at an open pit coal mine, where the surface displacements were measured by GPS and the data were collected until failure occurred. The strength parameters were determined using the data obtained just one day before the failure occurred. The results of the back analysis indicate that the factor of safety becomes approximately 1.0. This proves that the computer code UBAPSS is well applicable to the assessment of the stability of slopes.

Abstract:

Although application of concrete faced rock fill dams (CFRD) have a rather long history, these have begun to be widespread lately in Turkey. These dam types are preferred generally due to being secure, comply with land conditions, their construction being practical and economic. The geological situation of the dam location is a significant factor for rock fill dams. Concrete faced rock fill dams are constructed up to now successfully at great heights. Since the whole of the concrete faced rock fill dams are dry, the earthquake do not form cavity water pressure in the rock fill cavities. Because of this property, concrete faced rock fill dams is resistant to earthquake. In this study the properties of concrete faced rock fill dams are briefly explained and Dim Dam (Antalya) is given as an example. Dim Dam is one of the most important concrete faced rock fill dams which construction continues. The dam is located on the Dim creek 13 km northeast of Antalya province Alanya town. The dam is being constructed for the purpose of energy, irrigation and fresh water and its height from the foundation is 134.50 m and crest length is 365 m. Bahçeli formation formed by Upper Permian aged schist with limestone blocks and Quaternary aged alluvium are outcropped in the dam location. Upstream face slope of the dam is 1.40 horizontal / 1 vertical, downstream face slope is 1.50 horizontal / 1 vertical. The impermeability of the dam is provided by the concrete coating on the upstream face.



1. ADVANTAGES OF CONCRETE FACED ROCK FILL DAM

Dams with concrete faced have been started to be constructed intensively in the entire world in recent years. The main reasons for preferring them are summarized in articles hereunder.

- Since the filling slopes are hillsideer than clay - core fillings, the body filling volume is less and it is possible to shorten the length of their discharge installations (spillway and derivation structures).

- Since zones numbered 1 (clay) and 2 (semi-permanent) do not exist, it is possible to continue work during rainy days and no damage is caused on agricultural lands - In case any problems that may increase leakage occurs on the front face, its repair is easy.

- The reservoir water load on the front face is safer since it is transferred to the base rock in the downstream of the dam axis.

- Under the reservoir water load + earthquake loads, high safety numbers are obtained since the material parameters do not change.

Furthermore, since there is no water movement within the dam, no space water pressure increases occur under earthquake vibrations and no decrease tendency is observed at sliding strength.

- Injection works can be performed in parallel with the filling construction. Base improvement requirement and cost cover the region only under the heel plate, therefore it is less than the other dam types.

- It is possible to construct ramps in any direction within the dam body. This minimizes the access ways in the dam.

SYNOPSIS:

As is well known from rock mechanics literature, the mechanical parameters of a rock mass cannot be obtained by conventional laboratory tests due to the difficulties encountered in the preparation of cores from a rock mass containing discontinuities. To overcome these difficulties, researchers have focused on developing empirical equations for predicting stress-strain behavior of a rock mass, including based on measurements of the discontinuity patterns. However, the UCS value of rock mass (UCSRM) can be predicted by decreasing UCSi based on quality of rock mass such as Rock Mass Rating (RMR), Geological Strength Index (GSI), Q value, etc. For this reason, an empirical equation in a unique reducing curve form has limited application in generalizing on the prediction of UCSRM from particularly soft rock mass to hard rock mass. In this study, a new empirical approach is developed to be used for predicting of strength of rock masses from soft to hard rock masses. In addition, a new procedure for defining of the disturbance effect on the strength of rock mass is introduced to the new empirical approach in conjunction with Hoek and Brown failure criterion.



1. INTRODUCTION

Determination of overall strength of jointed rock masses have been attractive research topic in rock mechanics community due to the difficulties encountered during the preparation of representative undisturbed samples for laboratory studies. The sophisticated and large scaled laboratory equipments are needed to perform experiments on undisturbed jointed rock mass even if the cores can be obtained. Therefore empirical equations have been used as practical tools during pre-design stages of engineering application to be constructed in/on rock masses. To overcome this difficulty, numerous empirical equations have been proposed in the literature and as given in Figure 1 (Yudhbir et al., 1983; Ramamurthy, 1986; Kalamaris and Bieniawski, 1995; Sheorey, 1997; Aydan and Dalgic, 1998; Heok and Brown, 1997 etc.). By using most of the existing empirical approaches, the UCS value of rock mass (UCSRM) can be predicted by reducing UCSi based on quality of rock mass such as Rock Mass Rating (RMR), Geological Strength Index (GSI), Q value, etc (Bieniawski 1989, Hoek and Brown, 1997 and Barton 2002). At this point, the question of “which one is the best for prediction of the strength of a rock mass?” cannot be definitively answered, as each of them tried to represent their original database. Most of the empirical equations consider just RMR, GSI or Joint Parameter-JP as an input parameter. But these characterization schemes may not explicitly include the strength and deformability of rock material may play some important role on the strength behavior of particularly soft rock mass. As can be followed from literature, the lowest strength of rock materials used for the calibration of Hoek and Brown empirical failure criterion is about 39.9 MPa (Hoek and Brown, 1980). Therefore, it is difficult said that the Hoek-Brown failure criterion with its current form has sufficiently a capable of prediction of strength of rock masses composed of particularly softer rock materials.

ABSTRACT:

In tunneling a reliable prediction of advance rates is essential for calculating budget and construction time. Estimation of penetration rate is expressed as the basis for predicting advance rate in underground excavation using Tunnel Boring Machine. In this study the obtained data from 10KM of excavated Zagros tunnel project in Iran were subjected to statistical analyses using MATLAB for ANN modeling. When the neural network has been successfully trained, its performance is tested on a separate testing data set. Finally, the penetration rate was predicted by the trained neural network. The results show that the developed ANN method is efficient for predicting the PR in Zagros tunnel. The ANN model for next 0.5 KM which is recently excavated is well compatible with the real calculated PR of TBM in the field with 79% confident level. Result of sensibility analysis on the effect of Thrust and Torque on the PR shows that the maximum PR in given ground condition occurred in the optimum limits of Thrust and Torque.



1 INTRODUCTION

Since the first Tunnel Boring Machine (TBM) was built, the performance analysis and the development of accurate prediction models of the machines have been the ultimate goals of many research works [1-9]. A reliable prediction of TBM performance is needed for time planning and budget control of projects. Both penetration rate (PR) and advance rate (AR) are estimated in performance prediction of TBM. Penetration rate is defined as the distance excavated divided by the operating time during a continuous excavation phase, while advance rate is the actual distance mined and supported divided by the total time [10]. The AR includes downtimes for TBM maintenance, machine breakdown and tunnel failure. Even in stable rock, the rate of advance is considerably lower than the net rate of penetration and utilization coefficients (U%=(AR/PR)×100) is estimated about 30-50% mainly due to a TBM daily maintenance. When low quality rocks are excavated, the penetration rate could potentially be very high. However it demands a strong support pattern and face to rock jams and gripper bearing failure which results a relatively low advance rate with utilization coefficients about 5- 10% [11]. According to the literature, important parameters in TBM performance could be categorized in two major parts as follows:

1. Ground Condition: It includes characteristic parameters of intact rock and rock mass properties. Mostly reported important properties of intact rock in TBM performance are Uniaxial Compressive Strength (UCS), Brazilian Tensile Strength (BTS) and Point load index (Is50). In terms of Rock Mass properties, discontinuity spacing (Js), Rock Quality Designation (RQD), the angle between the tunnel axis and the planes of weakness (Discontinuity Dip and Dip Direction), rock mass quality using classification system such as RMR, Q and GSI have significant impact on TBM performance.

2. TBM operational parameters such as value of thrust, torque, Round Per Minute (RPM) and disc specifications have great influence on TBM performance. Tarkoy (1973) presented a model to predict penetration rate of TBM using only total hardness of intact rocks.

Abstract:

The Balaghat mine have all service roadways as well as shaft driven in hanging wall, having a weak layer of phyllitic zone in contact with the orebody. On exposure its bulging in the stopes is observed. In few cross-cuts, hangwall collapse is also observed near this contact. To adjudge its stability and corresponding changes in stope parameters when level interval is increased from 30 m to 45 m between 12th level (12L) to 13.5th level (13.5L), numerical modeling analysis was performed. 2D elastic models were solved for stope back stability and crown pillar design, while 3D elastic models were solved for post pillar design and cross-cut stability analysis. Simulation was performed by using FLAC3D software. It was found that, stope can be worked safely to a maximum span of 15 m without post pillars, beyond which post pillars of size 5 m x 5 m is required at different intervals. Result of detailed analysis is discussed in this paper.



1.0 INTRODUCTION

The orebody dipping approximately at 60° varies in width from 9 m to 20 m. Footwall contact is of Felspathic quartzite. Hangwall contact is of phyllite and its thickness varies from 0 to 10 m. It is thicker at northern and southern end while quite thin near central zone of the orebody. On exposure its bulging in the stopes is observed even though a thin layer of ore, 0.5 m thick, is left against the hanging wall contact. Some cracks in cross-cuts and hangwall drivage were also observed following stope settlement in higher levels. In few cross-cuts, hangwall collapse is also observed near the contact. Stope distance from the haulage roadway is generally 30 m. In the stope, an opening of 3 m is maintained between stope back and fill material. Present practice is to use cable bolt of 12 m length at an interval of 2 m, to stabilize the immediate roof along with rock bolts of 1.5 m length. Up to 12L, which is at a depth of 320 m, 30 m Stope height is maintained with a crown pillar of height 5 m. A representative cross-section showing working levels is shown in fig.1. However, to improve the production and productivity, it was decided to increase the level interval to 45 m. Subsequently, numerical modelling studies was undertaken to adjudge safe stope dimensions along with changes in support pattern.



2.0 INPUT PARAMETERS FOR SIMULATION

A review of all the plans and sections from the mine reveals that the orebody thickness, depth of cover, orebody dip and thickness of phyllite in the hangwall are varying from north side to south side of the deposit. The physico-mechanical properties utilized in numerical models are as shown in table 1. Since measured horizontal in situ stresses are not available at any of the underground manganese mine site, a theoretical equation for estimating the mean horizontal stresses proposed by Sheorey [1] given in equation (1) is used to evaluate mean horizontal stress for Bhalaghat Mine and is given in equation (2).

SYNOPSIS:

Determination of rock strength and its anisotropy is important in rock mechanics and engineering geology investigations. Destructive tests such as measurement of uniaxial compressive strength and/or point load strength index are commonly employed for this purpose. In the present investigation, an attempt is made to explore the possibility of using anisotropy of magnetic susceptibility (AMS) data as a gauge of rock strength anisotropy. For this, 10 blocks of quartzites that do not show mesoscopic foliations were sampled from a quarry in the Ghatshila area (eastern India). AMS analysis was performed on these blocks to identify the magnetic foliation in them. This identified foliation direction was then marked on the blocks and subsequently cores in directions parallel as well as perpendicular to the foliation were extracted. Subsequently the point load strength index in cores perpendicular to the foliation (SP) and parallel to the foliation (SL) was determined on the basis of which the strength anisotropy (RS = SP/SL) was calculated. It is noted that rocks with high SP also have a high SL. It is also noted that in rocks with a degree of magnetic anisotropy (P′) greater than 1.045, RS tends to increase with the increase in P′ value. It is therefore concluded that the P′ value of 1.045 is the threshold value for the quartzites beyond which the RS increases with P. These initial results provide the possibility of using AMS data for gauging rock strength anisotropy and thus hold promise for future studies in rock mechanics.



1. INTRODUCTION

Rocks containing well-developed mesoscopic foliations show strength anisotropy which is crucial in rock engineering (1). Many rocks in the earth are foliated (e.g., granite gneisses, mica schist etc.) and it is well established that rocks with well developed foliations have strong magnetic susceptibility anisotropy too (2, 3, 4). Therefore, in strongly foliated rocks, the strength and magnetic susceptibility anisotropies are expected to be related. Several studies have focused on determination of strength anisotropy of various foliated rocks (5, 6, 7, 8, 9,10,11). But when a rock is devoid of mesoscopic scale foliations, any strength anisotropy in that rock is not immediately apparent. An important usefulness of anisotropy of magnetic susceptibility (AMS) is that weak planar fabric defined by preferred orientation of minerals (magnetic foliation) can be identified even if the rock is devoid of mesoscopic scale foliations. In the past, AMS analysis has been used to carry out petrofabric studies in variety of rocks including those that lack visible foliation (2, 12, 13, 14, 15, 16, 17). Hence, although on the mesoscopic scale a rock may not contain a foliation, fabric elements can be identified from AMS data. As rock strength tests are of destructive nature, this investigation is also meant to explore the feasibility of using AMS data obtained from non-destructive measurements as an index of rock strength anisotropy. The present investigation aims to explore this issue for quartzite rocks (from Ghatshila area, eastern India) that do not contain mesoscopic foliations.

Abstract:

This paper presents geo-mining conditions along with the problems of coal bumps encountered at Chinakuri Mine of ECL (Raniganj coalfield) and Lazy Mine of OKC. Further, the developed and practiced measures at Lazy Mine of OKC to control the coal bump are detailed and analysed to assess their suitability for underground extraction of Dishergarh coal seam of Chinakuri Mine, ECL. It is observed that the conventional techniques, being practiced to release the stress concentrations and to create a network of fissures in the solid accompanying rocks in Indian coalfield, need a complete change. On the basis of this analysis, a suitable method of mining with overlying strata management approach is advised to suit the conditions of the Chinakuri Mine of ECL.



1. INTRODUCTION

Out of different rock mechanics problems of underground coal mining of deep seated deposits, coal bump/rock burst is identified ( CMRI, 1994) as a major hazard during underground coal mining at greater depth. Coal bump/rock burst engage violent and rapid failure of coal/rock in and around an underground excavation. Sudden release of accumulated elastic stain energy from a rock mass in the free face, created due to excavation, is the origin of this phenomenon and is, mainly, related with the geo-mining conditions of the site, characteristics of the coal/rock mass and stress regime of the area. CIMFR undertook an investigation related to this issue (CMRI, 1994) but this investigation remained limited, mainly, to identify different coal seams of the country likely to pose the coal bump/rock burst problems and their causative factors. Some approaches to control the problems of coal bump/rock bursts were also investigated but their field application achieved partial success. However, this study could project characteristics (Table -1) of some the coal seams and found that the Dishergarh coal seam of Chinakuri Mine (Raniganj Coalfield) is one of the most bump/burst susceptible seam in the country. Recently, CIMFR has collaborated with the Institute of Geonics, Ostrava, the Czech Republic for rock mechanics investigations to meet challenges of strata control of deep underground coal mining. During this collaboration, the success of the Czech counterpart in controlling coal bumps/rock bursts during underground visits of Czech mines is experienced. This paper describes the geo-mining conditions of Chinakuri Mine of ECL and the results of investigations taken to characterise the coal/rock mass of the mine.



2. Chinakuri Mine

Chinakuri Colliery 1&2 Pits of ECL is situated in the heart of the Raniganj coalfield on the bank of river Damodar near Asansol city of West Bengal. This is the deepest coal mine in the country, where underground mining is taking place at nearly 700 m depth of cover. Before nationalisation, this mine was owned by M/S. Andrew Yule & Co. and has experienced mining of a number of coal seams by different techniques. However, mining of the Dishergarh coal seam at this mine by bord and pillar and longwall methods of this colliery has always been a problem, mainly due to occurrence of coal bumps.

1. INTRODUCTION

The structural integrity of deep, large underground facilities such as tunnels, mines, pumped storage facilities, and physics laboratories requires the ability to predict rock mass stability under loading to ensure the safety of human occupants and the longevity of the underground space. Deformation occurs over time scales that range from milliseconds to decades and spatial scales that range from millimeters to facility scale. Beginning with design, prediction is typically based on finite element models using available or estimated properties. As with most geotechnical problems, much of the difficulty of prediction lies in the inability to sufficiently characterize the rock properties, especially discontinuities. As a consequence, semi-quantitative measures, such as Rock Mass Rating (RMR) or the Hoek-Brown Geological Structure Index (GSI) [1], are used to characterize the rock mass together with empirical charts for design criteria such as rock bolt spacing for ground control. During and following construction, validating model predictions is necessary to assess their performance. Parameter adjustment, or even the physics incorporated within the model, can be made using back analysis. This monitoring should be a continuous or periodic process over the life of the facility. For civil structures, the post-construction era will be measured in decades. With the inherent uncertainties and high stresses associated with the deep underground environment, the potential for rock failure must always be borne in mind. Mitigating the risk is prudent, but formal cost-benefit analysis may be precluded by the uncertainties. Keeping abreast of the condition of the facility through Structural Health Monitoring (SHM) is gaining acceptance for underground construction [2]. One reason for the growth in research in monitoring is that maturing technologies, like fiber-optic sensors and associated instrumentation, can collect data that were not previously achievable. They are robust and geometrically flexible, possess long-term stability, are cost effective, and extend coverage in spatial extent with improved resolution or provide data at a higher sampling rate. In addition to fiber-optic technology, a host of new technologies with potential for underground geotechnical applications exist, including LIDAR, wireless “smart dust”, piezoelectric sensors, and high resolution electrical and seismic imaging [3; 4; 5; 6; 7]. The subject of this paper is mainly to describe preliminary experiments, future needs, and instrumentation and monitoring plans of the authors' research activities in the 2400-meter Deep Underground Science and Engineering Laboratory (DUSEL) in the Black Hills of South Dakota, USA, where fiber-optic sensors and water-level tiltmeter arrays have been installed.



2. FIBER-OPTICSENSING

Fiber optic sensors consist of two main types: discrete Fiber Bragg Grating (FBG) sensors and distributed strain and temperature (DST) cables. FBG sensors consist of a glass filament core contained in protective cladding. A periodic refractive index grating is written into the core of the fiber. Broad-spectrum light is transmitted down the fiber from an optical laser interrogation box. The grating reflects the wavelength of light that matches the Bragg wavelength while all other wavelengths are transmitted through the fiber. Strain and temperature introduce a shift of the Bragg wavelength [8].

ABSTRACT:

The rock slope failures occur along pre-existing natural rock discontinuities or plane of weakness. In rock slope stability problems, the actual failure or sliding surface depends upon spatial orientation, frequency and distribution of the discontinuities and the inherent shear strength of filling material in the discontinuities. Rock Mass Quality and Rock Defect study by determining the complicated features of rock helps to characterize the rock slope and indicate the causes of slope failure in specific location. The new and modified techniques help to establish the cost and complexity of both design and re-construction of rock slope for a restoration of damaged road or a new road alignment by-passing the existing one as an alternative solution. These remedial measures generally comprise of improving surface and sub-surface drainage, restraining structures, reinforcement techniques, Rock bolts, Shotcrete, Steel fiber reinforced shotcrete(SFRS), etc. A new technique of Rock Engineering investigation and remediation for unstable rock slope is attempted to identify the elements of a high quality investigation in different stretches of National highways and state highways of India.



INTRODUCTION

The slope stability is one of the most common problems in natural hill slopes in India. There are several failure criteria available to describe the deformation of the rock mass. These failure criteria are based on the rock mass properties. Mainly two types of failure are noticed in the rock mass viz. shear failure and tensile failure. The joint controlled rock slopes prone to failure need an immediate attention to find out cost optimum solution in the form of remediation programmes for execution in hilly terrain of India. Unstable blocks of rocks adjacent to roadways can have a significant effect on the safety and economic feasibility of highways. Rock slope failure causes extensive damages to roads, buildings, forest, plantations and agricultural fields and habitats. Cloudbursts, flashfloods and short duration high intensity rainfall often triggers natural disasters in Himalayas. There are a few examples of earthquake induced landslides triggered by seismic waves. However, they do cause lot of damage to infrastructure as well as habitat. The stabilization of hill rock slope is attempted in a process as shown in Venn diagram in Fig.1. The physical evidences of rock slope failures investigated so far indicate that almost all rock slope failures occur along pre-existing natural rock discontinuities or plane of weakness. In rock slope stability problems, the actual failure or sliding surface depends upon spatial orientation, frequency and distribution of the discontinuities and the inherent shear strength of material present in the discontinuities.

Rock Quality Designation (RQD) and core recovery determination has to be carried out during logging of bore holes. Variation in thickness of strata and presence of discontinuities such as joints, faults, shear zones etc. beneath the slope may be interpreted from the data collected from the field using geophysical techniques. Presence of water body located underneath the hill slope may also be demarcated by these techniques. The use of ground penetrating radar is an effective substitution for bore hole investigations.

ABSTRACT:

The long water transfer tunnel of Kerman in the length of 63.5 kilometers is a part of the longterm water supply project of the city of Kerman, which should be conveyed water from Safa and Sarmeshk dam reservoir to the Kerman water treatment plant. Due to the high length and depth of this tunnel, it may face difficulties during excavation. In this paper, a method for probabilistic quantifying these problems is proposed. Using the computational model it is possible to determine the uncertainties of the geological prediction, the distribution of the geotechnical parameters and the rock mass characterization. Each model provides a physical value in which one aspect of the rock mass response, such as displacements, depth of failure zone, level of squeezing, intensity of rock burst and etc is described. In this paper, the rock mass of the tunnel path is divided into calculation segments according to the rock mass types and influencing factors and its parameters are studied in the form of probabilities distribution. By using empirical correlations and computer simulation, the strength and the behavior of the rock mass in tunnel path are assessed in the form of probability, and finally the results are demonstrated in quantities. According to the obtained results, during tunnel excavation, the risk of failure, serious instability, squeezing and rock burst are predicted the position and the quantity probability of the expressed risks at different parts of the tunnel route are determined.



1- INTODUCTION

The necessary information for the design of underground structures, are mainly obtained from local exploration and measurements, but these data are very limited, therefore, a great deal of uncertainty (geological and hydrogeological conditions) is concluded from the construction of underground structures in rock mass. So, design engineers and decision makers carry out parameter optimizations not only in their own desires but also with risk. The risk can be defined as the product of the probability and the eventual impact of a hazard. This risk gets new side in the long tunnels that its rate of danger depends on the type and the amount of geotechnical datas. Ideally, all potential risks are related to the rock mass characteristics and should be covered by a rock mass classification system. With this proposed computational model, it is possible to take into account the uncertainties of the geological prediction and the distribution of the geotechnical parameters. The Monte-Carlo method is now one of the most powerful and commonly used techniques for analyzing complex problems. In this paper, the data of the water transfer tunnel of Kerman are entered to the @Risk software in the form of probabilities distribution. By using empirical correlations and simulation, strength and behavior of the rock mass in tunnel route are obtained in the form of probability, and finally the results are demonstrated in quantities. It should be noted that the obtained results give us only an aspect of tunnel behavior and not more, so for more detailed evaluation, more precise data and general analysis methods are needed.