STUDY ON CUTTING HIGH-SLOPE STABILITY EVALUATION BASED ON FUZZY COMPREHENSIVE EVALUATION METHOD AND NUMERICAL SIMULATION

Original scientific paper The cutting of a high-slope is a complex nonlinear system that is random, fuzzy and variable. Stability evaluation for it from both qualitative and quantitative perspectives is essential in disaster management. This study firstly conducts the engineering experience analogy, giving a qualitative description of the stability of the slope. Then the cutting slope stability evaluation index system and the grade evaluation standard was established. With the weight vector and membership degree in the index evaluation system determined by the fuzzy comprehensive evaluation method, the degree of stability of the slope was comprehensively evaluated according to the membership function group. The key point on the slope was quantitatively evaluated by the numerical simulation method based on qualitative results acquired from fuzzy comprehensive evaluation. This research shows that this comprehensive analysis model can be more accurate in reflecting the various factors affecting the stability of a high slope. Furthermore, the calculation and analysis result is reasonable and of relatively higher precision, so it can provide some guidance for complex slope engineering and its systems.


Introduction
The steadiness of cutting high side slope is subjected to a combination of side slope lithology, including rock mass structure, geological structure, degree of weathering degree, moisture, disposition of bed, structural plane and other relevant factors.If slope structure is developed, the interaction between its bed disposition and slope poses a more significant influence to the side slope steadiness.Hence, the factors that influence the stability of side-slope are numerous, some of which are difficult to be confirmed and estimated.Through the above factors influencing side-slope stability, some scholars conducted a series of research projects [1÷3], and the current evaluation methods for side slope steadiness are divided into two types.The first one is quantitative evaluation, which includes the engineering analogy method, limit equilibrium method and numerical calculation analysis method.Although these methods can analyse the steadiness of side slope to some degree, they still have difficulty in judging and confirming uncertain, random or complicated factors [4÷7].The second type is uncertainty analysis methods such as the fuzzy evaluation method [8], neural network evaluation method [9÷12], grey theory method and genetic algorithm [13], etc.However, it is for this type of method to accurately judge and confirm random, unclear or changeable factors in a complicated side slope's qualitative and quantitative analysis via solely the evaluation method.
Hence, following the research above, this article proposes to combine the uncertainty analysis method and numerical simulation to analyse the steadiness of side slope from quantitative and qualitative perspectives to provide a leading for project construction.The uncertainty analysis method refers to adopting the fuzzy transform principle and maximum membership principle to analyse with consideration of relevant factors.Such a model is adopted with wizard mode and finished in four steps to select quantitative index factors.The confirmation of weight vector can be done by two methods: analytic hierarchy process (AHP) and the expert evaluating method.For quantitative index factors, Technical Gazette 24, 3(2017), 801-807 membership is confirmed by membership function, while for the qualitative index factors, it is confirmed by uncertain computation, as it is a discrete date.The implementation of such a model is shown in Fig. 1.
In the end, the results from uncertain computation are regarded as the degrees of danger and taken back to landslide monitoring and early warning systems, which will be illustrated by figures.

Establishment of the model
Firstly, the evaluation factors were selected, then it should establish the evaluation factors set and determine the forecast evaluation set; that is, ranged target levels of very unstable, unstable, potential unstable, basically stable, and stable.
The evaluation factors set are: The evaluation results set are: In the same way, it can be concluded that the single factor evaluation set under the influence of the factor On the basis of this, the general evaluation matrix is constructed, and is made up of m factors; that is: where R is the evaluation factor set U to the evaluation result set of a fuzzy relation, r ij is membership degree of the factor u i versus v i corresponding.When carrying out a single factor u i slope safety assessment, it is defined as fuzzy sets: ) ,..., , ( The weight vector matrix and fuzzy "synthetic", calculated evaluation set are: ) ,..., , ( Where, (0 < b j ≤ 1), .,..., 1 , 0 , According to the principle of maximum of membership degree, the value of b i0 was calculated as b i0 = max{b j }.

Confirmation of weight vector
The weight vector is confirmed by the analytic hierarchy process (AHP), which is strictly verified in mathematical evidence, so it is suitable for various complicated vector systems to analyse, confirm and check its importance sequence.

Confirmation of r ij membership
The confirmation of membership value of those quantitative predictors can be expressed by trapezoidal membership function, and given μ 1 as the first danger degree (the steady membership); given μ 2 as the second danger degree (the basically steady membership); given μ 3 as the third danger degree (the potential unsteady membership); given μ 4 as the fourth danger degree (the unsteady membership); given μ 5 as the fifth danger degree (the extreme unsteady membership).

General of project
A certain section of highway in Shanxi province is deep-cutting, which spans 455 meters.In the middle of route, the maximum depth of digging reached 38,0 meters, the maximum height of side slope was 46,3 meters.The cutting side slope is composed of slightly-intensive collapsed silt from the Quaternary period (Q3p)l and intensively-moderately weathered sandy shale, sandstone and shale P1X and P1S from the Permian period.The component type of this section's side slope is upperlithologic side slope, so it is soil-stone composite side slope.The research shows that the sliding aspect of the left side incline direction is the same as that of rock mass, presenting an unfavourable combination.Whereas, it is converse with the right slope, it presents a favourable combination.Both sides of the cutting are high side slopes, in which the gob covered a length of 470 meters and a thickness of 4 meters and adopted a short wall mining method.This gob is located 76 ÷ 167 m underground where the ground surface has cracks and sinks.

Comprehensive uncertain analysis for cutting steadiness 3.2.1 Establishment of risk evaluation index system
According to reference [13], the engineering geology, hydrogeology, project design and construction, organization and management and relations between human and environment should all be taken into consideration and selected when considering the identification method of features and risk sources of landslide, collapse and sharp deformation in the side slope of a road bed.Meanwhile, it should be considered that the road bed is located in a gob when the first index is further specified.Hence, the location risks should be considered more seriously when designing the project.From the analysis above, the specific risk evaluation indexes [14÷18] are shown in Fig. 3.
From a similar method, the second weight index value and total weight value can be computed [19,20], which are illustrated in Tab. 2. The number of people 0,2599 0,0257 The construction of environmental impact 0,4126 0,0407 Mechanical quantity 0,3275 0,0323

Fuzzy comprehensive evaluation
After determining the membership degree of each index, the fuzzy operation was used by the fuzzy evaluation matrix (R) and weight set (W).Then, the result of fuzzy operation needs to; that is, the normalized result is the result of fuzzy evaluation.Therefore, the result of fuzzy comprehensive evaluation is: According to the maximum membership principle, the degree of risk can be judged as extreme danger.

Computation of analysis results
The road cutting is 455 meters long, and its largest side slope is 46,3 meters high, which is located on the left side of road.The comprehensive analysis of influencing factors on side slope steadiness is demonstrated as follows.
(1) The tendency of rock mass and slope is consistent, presenting an unfavourable combination.The body of the slope consists of sandy shale or sandy shale and silt, from which the upper rocks are seriously weathered with cracks appearing in joints that are combined with soft and hard rocks.The soil is vertically developed and contains pores so that rainfall water can penetrate into the sandy shale or soil-rock surface, resulting in the sandy shale becoming softer and shaped into a soft sandwich or soft-plastic soil, which may subsequently cause a hard-to-harnessed unsteady side slope and landslide.From the above analysis, the left side slope is unsteady.However, the tendency of rock mass on the right sideslope is the opposite, presenting a favourable combination.The whole body, consisting of sandy shale, soft rocks in particular, is basically steady.Weathered by winds, joints cracks develop and small landslides collapse in the slope surface.Hence, preventative and protective measures should be taken in advance.From the above analysis, the right side slope is unsteady.
(2) The main influencing factors on the side slope include the combined relationships of the both tendency of rock mass and slope, joint cracks and weathering and the erosion effect.

Value analysis and check 4.1 Establishment of the finite element model
Selecting a fracture surface K25+414 in the right side at 25 m, a 2D model is set by Plaxis software.Physics parameters are selected in accordance with a geological investigation report on deep cutting with consideration of digging, the anchorage system and retaining wall.These parameters are shown in Fig. 4. The deformation of A-H points with digging is shown in Fig. 4, and the maximum settlement is shown in Fig. 5.As is shown in Tab. 5, with further construction and excavation, the road cutting settlement reached the maximum value in point A at 24 cm.Hence, the result of calculation and numerical analysis is the same.

Necessary construction measures
According to results from fuzzy comprehensive evaluation and numerical analysis, a sharp deformation appeared in the deep excavation cut.Hence, a landslide and collapse is more likely to occur, and the harnessing of the mined-out area should be given more importance.The measures are divided into a couple of aspects: monitoring and early-warning, engineering protection.

Monitoring and early-warning
The Anchorage system and retaining wall are applied into the side slope, so they should be monitored.As for rock bold, 3÷5 monitoring sites can be arranged in one rock bold, and the protection wall is monitored by a soil pressure meter.The clinometer will be buried in deep road cutting where rainfall meters are installed to monitor rainfall levels and intensity.According to depth and collapse features in the gob area, the full-pressure granting method as well as a supplementary project of granting-retaining wall are recommended to harness this area.

Preventative measures
According to the depth and collapse features in the mined-out area, the full-pressure granting method as well as a supplementary project of granting-retaining wall is recommended to harness this area.Hence, the slope cutting method can be adopted and counter-pressure methods as well as installing a protective wall in the side slope so as to avoid damage from falling stones.

Conclusions
With a qualitative description of high cutting side slope steadiness through the engineering analogy method and the establishment of an evaluation index as well as degree standards and the comprehensive evaluation of west side slope from membership functions, several conclusions can be reached, as follows.
(1) The steadiness index establishes five first-class indexes and sixteen second-class indexes, and the risk evaluation is reasonable and feasible.
(2) To conduct a safety evaluation for high cutting slope and a quantitative analysis, key points can be determined through numerical simulation.It can be known that the high cutting slope safety model has a high accuracy and reasonable classification of computation, which can more easily and quickly analyse and provide a protective scheme for complex highway high side slopes.

Figure 1
Figure 1 Operation flow chart of the fuzzy comprehensive evaluation model 2 Fuzzy comprehensive evaluation method:

Figure 2
Figure 2 Sectional drawing of deep excavation

Figure 3
Figure 3 Map of the indicator system

Figure 4 Figure 5
Figure 4 Maximum displacement nephogram of area

Table 1
First class index weights

Table 2
Weights of evaluation indexes

Table 3
Evaluation indexes weight and membership degree

Table 4
Physical and mechanical parameters of rocks

Table 5
Maximum displacement of A-H point