- Email: [email protected]
Investigation on support pattern of a coal mine roadway within soft rocks — a case study
International Journal of Coal Geology 140 (2015) 31–40
Contents lists available at ScienceDirect
International Journal of Coal Geology journal homepage: www.elsevier.com/locate/ijcoalgeo
Investigation on support pattern of a coal mine roadway within soft rocks — a case study H.P. Kang ⁎, J. Lin, M.J. Fan Mining & Designing Branch, China Coal Research Institute, Beijing, China State Key Laboratory of Coal Mining and Clean Utilization (China Coal Research Institute), Beijing, China
a r t i c l e
i n f o
Article history: Received 28 September 2014 Received in revised form 6 January 2015 Accepted 6 January 2015 Available online 30 January 2015 Keywords: Soft rock Ground reinforcement Rock bolt Discrete element method Coal mine
a b s t r a c t Supporting coal mine roadway within soft rocks is a typical challenge in underground mining practices. As the most widely used support structure, rock bolt system has been successfully used to support coal mine roadways in various complex geological and geotechnical conditions, including roadway in extremely weak rock masses, roadways 1000 m below ground, open-off cut roadways with large section, and roadways along the edges of mined-out areas with thin pillars or even no pillars. However, the effectiveness and applicability of rock bolt system for the reinforcement of soft rock masses has not yet been established. In this paper, we present a case study of rock bolt system as used for ground reinforcement of longwall entries within soft rock masses. The study site was the tailgate of the longwall panel 5-2S at the Hongmiao coal mine in the Pingzhuang coalmining district in China, a typical soft rock coal mine of the region. Since traditional rock bolt system failed to maintain the stability of the tailgate, the effectiveness of rock bolt systems for this application remained doubtful. The reasons for the failure of the rock bolt system were first examined. A discrete element method simulation was then performed to better understand the mechanism of rock bolts in supporting soft rocks. An improved rock bolt system was finally proposed that was used to support the tailgate. The field monitor showed that the improved rock bolt system successfully suppressed cracking and dilation of the tailgate. This case study is useful in enhancing engineering applications of rock bolts to support longwall entries excavated in soft rock masses. © 2015 Elsevier B.V. All rights reserved.
1. Introduction Rock bolt systems have been extensively used for ground reinforcement in underground coal mine roadways around the world (Brown, 1999; Peng and Tang, 1984). In China, more than 70% of roadways in key state-owned coal mines are supported by the system (Kang et al., 2010). As an active support pattern, rock bolts suppress the deformation of and damage to rock masses immediately after installation. Rock bolts are economical, easily installed and reliably supportive. The principal objective of rock bolting is to help rock masses support themselves. Bolts increase the stress and the frictional strength across joints, causing loose blocks or thinly stratified beds to become wedged together and act as a composite beam (Goel et al., 2007; Karanam and Dasyapu, 2005; Mark, 2000). Rock bolts are generally installed with pretension. The high pretension applied during installation has been proven to be a very important component of the supportive effect of rock bolts (Li, 2006). Pre-tensioned rock bolts compress and reinforce the rock mass in their vicinity. This effect spreads to a further section of the rock surface through accessories such as load-bearing plates and screens, creating a confining pressure on the rock surface (Ghazvinian et al., ⁎ Corresponding author. Tel.: +86 84263125. E-mail address: [email protected] (H.P. Kang).
http://dx.doi.org/10.1016/j.coal.2015.01.003 0166-5162/© 2015 Elsevier B.V. All rights reserved.
2012). Active pretension modifies roof behavior by dramatically reducing bed separation and delamination immediately within 0.5–0.8 m of the roof (Frith and Thomas, 1998). In addition, thinly laminated roof beds can be clamped into a thick beam that is more resistant to bending (Peng, 1998). Bolts installed with pretension increase frictional resistance along bedding planes, minimizing roof sag and defection, and decreasing the likelihood of lateral movement due to horizontal stress (Stankus and Peng, 1996). Rock bolt systems have been successfully used in complex geological and geotechnical conditions, such as roadways 1000 m below ground, open-off cut roadways with large section and roadways along the edges of mined-out areas with thin pillars or even no pillars (Kang, 2013; Zhang and Gao, 2004). However, the effectiveness and applicability of rock bolt system for the reinforcement of soft rock masses has not yet been established (Jiao et al., 2013; Wang et al., 2000; Yoshinaka et al., 1998). Soft rock refers to a group of geotechnical materials that has a uniaxial compressive strength between 0.5 and 25 MPa and has similar geotechnical characteristics, including slaking, swelling, compressibility, time dependence, and volume change (Jiao et al., 2013; Wang et al., 2000; Yoshinaka et al., 1998). Roadways caved out of soft rocks exhibit large deformation in the form of severe roof sag, wall convexity, and floor heave. These roadways also have distinct characteristics that include a high rate of rapid deformation, a high sensitivity to
32
H.P. Kang et al. / International Journal of Coal Geology 140 (2015) 31–40
water invasion and stress changes, evident rheological deformation and deformation in all directions of the roadway cross section (Wang et al., 2000). Maintaining the stability of roadways excavated in soft rocks has been a great challenge for underground coal mining engineers. In cases where roadway closure is severe, rehabilitation and re-support of the roadway have to be continually arranged during the roadway's service period to maintain its cross section and guarantee that it remains functional for ventilation and transportation. This causes safety issues, delays of mining activities, and results in dire economic loss. Some researchers suggested that yieldable and flexible support patterns should be used to support soft rock roadways, instead of stiff supports such as the rock bolt system (Jiao et al., 2013; Wang et al., 2000). A common structure for supporting roadways in soft rocks is the installation of yieldable U-shaped steel sets. This support pattern is believed to provide sufficient support to suppress the deformation of soft rocks surrounding a roadway and has been proven to be successful where implemented (Jiao et al., 2013; Wang et al., 2000). In contrast to rock bolting, the yieldable U-shaped steel sets are passive, costly, and time-consuming to install. In underground coal mine practice, yieldable U-shaped steel sets are more suitable to support main roadways with a long-term service period and the absence of mining activities (i.e., retreat longwall mining) where rock bolting is not applicable. In longwall entries intended for the extraction of a panel, yieldable U-shaped steel sets are not effective; they have a detrimental impact on the shearer and must therefore be removed, at high cost. This paper presents a case study on the application of rock bolt systems for ground reinforcement of longwall entries within soft rock masses. The study site was the tailgate of the longwall panel 5-2S at the Hongmiao coal mine, Pingzhuang coal-mining district, China. Traditional rock bolt system failed to support the entries. The failure reasons were first examined. A discrete element method simulation was then performed to investigate the effects of rock bolts on suppressing cracking and dilation in soft rock masses. An improvement on the rock bolt system was finally proposed and used to support the tailgate. Field monitoring of roadway deformation was conducted to evaluate the success of the improved rock bolt system in maintaining the stability of the tailgate.
Thickness (m)
Column
Lithology
Geological descripons
Coal 5-1
light, shiny, half-intact
2.63-4.94 3.79
Sandy mudstone light, shiny, carbonaceous shale laminates
Coal 5-2
3.66-7.00 5.99
Sandy mudstone Coal 5-3
light, shiny, half-intact
Fine-grained sandstone
White, argillaceous cement
2.55-8.04 5.30 0.6-15.0 5.30
Coal 6-1
light, shiny, half-intact
Fine-grained sandstone
White, argillaceous cement
1.67-5.58 3.50
Fig. 2. Lithological descriptions of the rock units at the Hongmiao coal mine.
strength of 6.2 MPa. It was sub-horizontal with a dip angle of 15°. Below the Coal 5-2 was found a sandy mudstone with a thickness of 1.0–2.0 m and a compressive strength of 13.4 MPa. The Coal 5-3 was 2.55–8.04 m thick and was situated immediately below the sandy mudstone. Below the Coal 5-3 was a fine-grained sandstone with a thickness varying between 0.6 and 15 m. The material situated immediately below the finegrained sandstone was the Coal 6-1 with a thickness of 1.67–5.58 m. The Coal 5-1 had been mined out, and current mining activity was extracting the Coal 5-2 using the longwall mining method. The longwall panel 5-2S was the first panel in the Coal 5-2. The depth of this panel varied between 350 and 400 m. The panel 5-2 had a length of 560 m in the strike direction and a width of 156 m in the dip direction of the Coal 5-2, see Fig. 2. In situ tests including in situ stress measurement, borehole strength measurement and borehole optical televiewer imaging had been performed in a vertical borehole in the main roadway close to the panel 5-2S, see Fig. 1(b). The borehole was 56 mm in diameter and 20 m in length. The in situ stress was measured using the hydraulic fracturing method and the results were σH = 14.62 MPa, σh = 7.35 MPa, and σv = 9.68 MPa. The direction of σH was N72oE, leading to an angle of approximately 80° with respect to the direction of the 5-2S tailgate. In borehole strength tests, a rod with a probe attached to the top was
2. Geological and geotechnical conditions The Hongmiao coal mine, Pingzhuang coal-mining district was a typical soft rock mine located in the Inner Mongolia Autonomous Region in China, see Fig. 1(a). It extracted coal seams formed during the Cretaceous period. The coal seams were separated by a few rock layers, as illustrated in Fig. 2. The first coal layer was the Coal 5-1 with a thickness varying between 2.63 and 4.94 m. The material situated immediately below the Coal 5-1 was made up of 0.5–2.0 m of sandy mudstone. Below the sandy mudstone was the Coal 5-2. The thickness of the Coal 5-2 varied between 3.66 and 7.0 m, and it had a compressive
N
Beijing
㚄㔌ᐧ
Hongmiao Coal Mine
285 m
300 m Tailgate Longwall panel 5-2S Headgate
Preliminary rock bolt system Displacement monitoring staon
(a)
Improved rock bolt system In situ tests posion
(b)
Fig. 1. (a) The location of the Hongmiao coal mine, Pingzhuang coal mining district, China. (b) Plan view of the roadways and panel at the study site in the Hongmiao coal mine.
H.P. Kang et al. / International Journal of Coal Geology 140 (2015) 31–40
inserted into the borehole. The probe was connected to a high pressure pump with a pipe. Under the hydraulic pressure generated by the pump, the probe was pushed against the borehole wall until failure, identified by a sudden increase in the displacement of the probe and a constant pressure value. The pressure was recorded and used to estimate the rock strength, according to the relationship between the pressure and the unconfined compressive strength (UCS) of the rock
33
which was established based on laboratory tests on different types of rocks. The borehole strength test results indicated that the roof rocks in the roof were generally of poor to average quality (Hoek and Brown, 1997). Borehole optical televiewer imaging showed that the roof was heavily fractured, see Fig. 3. Furthermore, the rocks were typical soft rock and displayed considerable swelling where they had encountered water.
(a) 1.6 m
(b) 2.3 m
(c) 3.6 m
(d) 4.1 m
(e) 7.2 m
(f) 8.5 m
(g) 15.5 m
(h) 16.1 m
Fig. 3. Borehole optical televiewer image showing a highly fractured roof at the study site.
34
H.P. Kang et al. / International Journal of Coal Geology 140 (2015) 31–40
Fortunately, only small amounts of water had penetrated the roadway, significantly reducing the swelling degree of the roadway's moisturesensitive rocks. The tailgate of the panel 5-2S was initially driven in a round arch cross section and supported by rock bolts. The rock bolts were 2.0 m long with a diameter of 18 mm and installed with a spacing of 0.9 m. As the tailgate was excavating, significant deformation of the roadways occurred tens of meters behind the advancing face, especially when the roadways were driven below the pillar of the above mined-out panel in the Coal 5-1. The coal surrounding the tailgate was heavily fractured due to the severe mining-induced stresses caused by the extraction of the above panel. The roadway cross section was difficult to obtain using the road header. Skin roof falls and rib falls were frequently observed in the advancing face and continued to occur even after rock bolts, cables, and screens were installed, leading to ‘string bags’ as shown in Fig. 4. The excavation of the tailgate was suspended due to severe instability and deformation. It was expected that these issues would significantly deteriorate with the consequent extraction of the panel 5-2S. An improvement of the rock bolt system had to be performed. 3. Discrete element method simulation of rock bolting The improvement of the rock bolt system necessitated a better understanding of the mechanism of rock bolts in suppressing cracking and dilation of soft rock masses around the tailgate. To achieve this, a numerical study was carried out. 3.1. Model configuration The UDEC Trigon approach proposed by Gao and Stead (2014) was adopted in this study due to its intrinsic capability to simulate the initiation, propagation, and coalescence of fractures, as well as the interaction between them and any pre-existing discontinuities. In the UDEC Trigon approach a rock mass is represented as an assembly of triangular blocks cemented at their contacts. A block is represented as a mesh of finite difference elements and is deformable and elastic. The fracturing process is represented by the sliding or opening of the contact, depending on its strength and the stresses it suffers. A tensile crack forms when the normal stress applied on the contact reaches the contact's tensile strength. A shear crack forms when the shear stress reaches the contact's shear strength, which is governed by c + σn tanϕ, where c and ϕ are the cohesion and friction angle of the contact, respectively, and σn is the normal stress applied to the contact. Gao and Stead illustrate that the UDEC Trigon approach can capture many aspects of the mechanical behavior of rocks from laboratory to rock mass scale (Gao
and Stead, 2014). Furthermore, the ability of rock bolts in suppressing rock mass dilation can be successfully simulated using this approach (Gao et al., 2014). By using the UDEC Trigon approach, a numerical model was created to simulate the effects of rock bolting in suppressing fracturing and dilation of soft rocks. Fig. 5 shows the geometry of the model which was based on the lithology of the 5-2S tailgate as illustrated in Fig. 2. In an effort to increase computational efficiency, the UDEC Trigon approach was only adopted for generating triangular blocks in the area of interest surrounding the roadway. These triangular blocks were elastic and rock mass failure was represented as cracking through the contacts between them. The average edge length of these triangular blocks was 0.25 m. This block size had been shown to be sufficiently fine to simulate crack propagation. The remainder of the model was simulated by big blocks which were represented as finite difference elements. These finite difference elements obeyed the Mohr–Coulomb constitutive mode and, as a result, rock mass failure was represented as plastic zones. The intact rock strength properties for the three types of rocks—coal, sandy mudstone, and fine-grained sandstone—were obtained from borehole in situ strength tests (as listed in Table 1). Young's modulus values were estimated as Ec = MR ∗ σc (Feng and Jimenez, 2014). The intact rock strength properties were multiplied by a scale factor of 0.58 to obtain rock mass strength properties (Esterhuizen et al., 2013). The tensile strength of the rock masses was reasonably estimated to be 1/12 of the compressive strength. The mechanical properties used in the model are listed in Table 2. The mechanical behavior of the coal and sandy mudstone, represented as triangular blocks in the model, was mainly governed by their contact properties. These contact properties were calibrated against the rock mass properties listed in Table 1. For the remainder rock masses represented as big blocks which were represented as finite difference elements, the Mohr–Coulomb properties were assigned. Note that bedding planes were incorporated into each layer and their contact properties are given in Table 2. The UDEC built-in “Cable” element was used to simulate rock bolts, and the “Liner” element simulated steel joists and straps. The parameters of the “Cable” and “Liner” elements are presented in Table 2. A detailed description of the two support elements in UDEC can be referenced at Itasca Consulting Group Inc. (2014). The field-measured in situ stresses were applied to the UDEC model to set up the pre-mining stress field. Note that the extraction of the 5-1 coal seam was not accounted for in the model because the mined-out area in 5-1 coal seam above the 5-2S tailgate had stabilized prior to the excavation and had no significant influence on the ground behavior of the 5-2S tailgate. The model was run under two different roadway
Fig. 4. Photographs taken in the tailgate of panel 5-2S at the Hongmiao coal mine showing a heavy roughness of roadway profile due to large dilation and skin falls. (a) Roof, and (b) rib.
H.P. Kang et al. / International Journal of Coal Geology 140 (2015) 31–40
35
Applied vercal stress to compensate for overburden pressure
3.8 m
3.1 m
60 m
60 m
Fig. 5. Configuration of the UDEC model used for simulating rock bolts in suppressing cracking and dilation within soft rock masses around the 5-2S tailgate.
support conditions, one with no support and the second with a rock bolt system which corresponded with the preliminary support parameters previously discussed. 3.2. Modeling results There are two components to the supportive effects of pre-tensioned resin-anchored rock bolts in Coal Measure rock masses (Gao et al., 2014): i) Reinforcement is provided by the rebar that bonded into the rock strata. The rebar provides axial resistance along its length, which restrains the deformation of the rock mass, i.e. the opening of pre-existing cracks. The rebar also provides shear resistance along pre-existing fractures so that sliding is restrained. ii) Surface support is provided by rock bolt accessories, including the load-bearing plate, strap and screen. The pre-tension applied to the bolts, which extends through the plate and strap provides confinement on the roadway surface. Even though this confinement is very low when compared with the rock mass stress, it significantly increases the residual strength of the supported rock mass. In combination, the reinforcement and support greatly restrain crack propagation. These effects were successfully captured in the UDEC model as shown in Fig. 6. When the roadway was not supported, heavy cracks were observed around it. Shear cracks were majorly developed in relatively deep zones while tensile cracks were mainly developed close to the roadway surface, see Fig. 6(a). These tensile cracks caused rock mass dilation and subsequently, led to significant deformation and roadway convergence, see Fig. 7(a). Comparing Fig. 6(b) with Fig. 6(a), it is apparent that the
Table 1 Intact rock properties and scaled rock mass properties of Coal Measures at the Hongmiao coal mine. Coal Measures
Coal Sandy mudstone Fine-grained sandstone
Intact rock σc (MPa)
MR
6.2 13.4 35.3
425 425 333
Ec (MPa) 2 600 5 700 11 800
Scale factor
0.58 0.58 0.58
Rock mass Em (GPa)
σcm (MPa)
σtm (MPa)
2.6 5.7 11.8
3.6 7.8 20.3
0.3 0.6 1.7
*E is Young's Modulus, kn and kn are the normal and shear stiffness, respectively.
installation of the initial rock bolt system significantly suppressed the crack propagation within the roof and ribs where the rock bolts were installed. The predicted depths of shear cracking in the roof and ribs were 2.6 m and 1.6 m, respectively, compared to 3.5 m and 2.2 m with no support. The depth and extent of tensile cracking were also significantly reduced, leading to notable decrease in rock mass dilation and roadway convergence (see Fig. 7(b)). Even though this rock bolt system greatly reduced cracking and dilation in the rock masses, it appears from the numerical model that the initial rock bolt system did not successfully stabilize the roadway. For example, extensive tensile cracks were observed in the roof and ribs (Fig. 6(b)). As a consequence, dilation and large deformation occurred in these areas as shown in Fig. 7(b), which was consistent with the ‘string bags’ observed in the field (see Fig. 4). Moreover, the numerical predicted cracking was beyond the anchored zone (see Fig. 7(b)), which suggests the potential for a sudden roof collapse. An improvement to the initial rock bolt system was subsequently simulated in an attempt to examine if the improvement could successfully control fracturing and dilation of the soft rock masses. There were four aspects to the improvement, as illustrated in Table 3: (1) the tensile strength of the rock bolts was increased from 120 kN to 180 kN which corresponded with 22-mm diameter rebar with a steel tensile strength of 470 MPa; (2) the rock bolt length was increased from 2.0 m to 2.4 m; (3) the pre-tension was increased from 20 kN to 70 kN; and (4) the stiffness and strength of the “Linear” element were doubled to account for more robust surface control accessories such as a welded steel screen and a big W-shaped load-bearing plate. As can be seen in Fig. 6(c), the improvement was successful at suppressing cracking around the roadway. In the roof, the predicted depth of shear cracking was approximately 2.0 m, substantially less than the 2.6 m predicted for the initial rock bolt system. More importantly, cracking was constrained within the anchored zone. Cracking in the ribs was also significantly reduced. Furthermore, tensile cracking barely propagated in the roof and ribs where the improved rock bolt system had been installed. Consequently, there was no noticeable dilation, see Fig. 7(c). An interesting outcome from the numerical modeling was that the cracking in the roadway floor appears to be reduced even though no rock bolts were installed in that area (compare (a), (b) and (c) in Fig. 6). This suggests that competent reinforcement in the roof and ribs may have a positive effect on floor stability.
36
H.P. Kang et al. / International Journal of Coal Geology 140 (2015) 31–40
Table 2 Mechanical properties used in the UDEC Trigon model. Lithology
Coal Sandy mudstone Fine-grained sandstone a
Matrix properties
Contact properties
Density (kg·m−3)
E (GPa)
Cohesion (MPa)
Friction angle
Tensile strength
kn (GPa/m)
ks (GPa/m)
Cohesion (MPa)
Friction (°)
Tensile strength (MPa)
1400 2600 2600
3.0 10.0 15.0
1.0 2.0 10. 5
23 28 33
0.3 0.6 1.7
14.4 18.0 27.0
5.8 7.2 10.8
1.0/0.5a 2.0/1.0a NA/3.0a
23/23a 28/28a NA/33a
0.3/0a 1.0/0a NA/0a
The first value refers to triangular block contacts representing rock mass, while the second refers to persistent contacts representing bedding planes.
It is also interesting to note that a rock bolt system seems to constrain the generation of tensile stresses in the anchored zone, see Fig. 8. Localized tensile stresses significantly prompt tensile crack propagation, as stated by Diederichs et al. (2004) and many others.
designed to be installed with a moment of at least 400 Nm, inducing a pretension greater than 70 kN. The cables were designed to be installed with a pretension of 200 kN. The improved rock bolt system, as shown in Fig. 9, was implemented in the remainder of the 5-2S tailgate at the Hongmiao coal mine.
4. Improvement of the rock bolt system
5. Field monitoring
The author's experience with coal mine roadways supports and the numerical simulations suggest that poor ground reinforcement in the existing stabilization system was likely a result of the following reasons: (1) The installed rock bolts were 18 mm in diameter with a tensile capability of approximately 120 kN, which was too weak for this roadway. Many fractured rock bolts were found in the entries. (2) The thread of the rock bolts was too short, at only 80 mm. The high degree of roadway surface roughness implied that, the contact between the plate and roadway surface was likely loose, resulting in an ineffective support. (3) The installed cables were 15.2 mm in diameter with a fracture force of 260 kN which was weak for the circumstances. Furthermore, the cable was installed in a borehole with a diameter of 28 mm, leading to a resin annulus thickness of 6.4 mm-far beyond the optimum annulus range of 2–4 mm (Hagan, 2003). (4) The screen and the steel joists were too weak to transfer the applied compression on the rock bolts to the roadway surface, leading to a low confining pressure. In addition, the screen was not strong enough to hold detached rocks. (5) The rock bolts were installed with low pretension. An improvement of the preliminary rock bolt system was proposed. The most substantial improvement included: (1) adopting highstrength rock bolts with a diameter of 22 mm, a length of 2.4 m and a thread length of 150 mm. The tensile capability of these bolts was more than 180 kN. The bolts were designed to be fully encapsulated; (2) adopting high-strength cables with a diameter of 22 mm and a capacity of 600 kN; (3) rock bolts were installed with a high strength load-bearing plate, W-shaped steel plates that were 280 mm wide, 450 mm long and 4 mm thick and welded steel screen with a mesh size of 100 × 100 mm for screen control; (4) the rock bolts were
A monitoring station was set up in the 5-2S tailgate to evaluate the effectiveness of the improved rock bolt system, see Fig. 1(b). At the monitoring station, the convergence of the roadway was measured regularly as the roadway advances. Two-point borehole extensometers were installed in the roadway roof to measure roof delamination. The monitoring instruments were installed immediately behind the advancing face, in an attempt to monitor the entire roadway deformation as it advanced. 5.1. Roadway deformation during its developing period The cumulative amounts of the cross section convergence of the 5-2S tailgate versus roadway advance are plotted in Fig. 10. The convergence of the tailgate began right after its excavation and continued to increase with face advance. The wall-to-wall convergence and roof sag stabilized after the face had advanced more than 53 m away from the monitoring station, while the roof-to-floor convergence kept increasing as the face advance, indicating an increasing floor heave. When the excavation of the 5-2S tailgate was finished, the wall-to-wall convergence, roof-to-floor convergence and roof sag were 79 mm, 281 mm and 43 mm, respectively. The wall-to-wall convergence was only 2% of the roadway width, demonstrating that the improved rock bolt system had successfully controlled wall-to-wall convergence. The substantial roof-to-floor convergence was attributed to the floor heave which was 238 mm. It should be noted that significant floor heave was only observed in two short sections (approximately 50 m in total) along the portion of the 5-2S tailgate that was supported by the improved
S/T = 1671/480
S/T = 1050/222
S/T = 775/140
(a)
(b)
(c)
Fig. 6. UDEC model predicted fracture modes of the soft strata around the 5-2S tailgate under different support patterns. (a) No support; (b) preliminary rock bolt system; and (c) improved rock bolt system. Red and blue refer to tensile and shear cracks, respectively, and orange and red refer to rock bolts and steel joist in (b) and steel strap in (c). S/T indicates the ratio of shear crack number to tensile crack number. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)
H.P. Kang et al. / International Journal of Coal Geology 140 (2015) 31–40
Disp.(mm) 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150
(a)
37
Tensile stress (MPa) 0.0 0.2 0.4 0.6 0.8 1.0
(a)
(b)
(b)
(c)
(c) Fig. 7. UDEC model predicted displacement vectors around the 5-2S tailgate under different support patterns. (a) No support; (b) preliminary rock bolt system; and (c) improved rock bolt system.
Table 3 Support element properties used in the UDEC Trigon model. Properties Cable
Elastic modulus (GPa) Tensile capability (kN) Stiffness of the grout (N/m/m) Cohesive capacity of the grout (N/m) Pre-tension (kN) Linear Elastic modulus (GPa) Tensile yield strength (MPa) Compressive yield strength (MPa) Interface normal stiffness (GPa/m) Interface shear stiffness (GPa/m)
Initial support Optimized support 200 120 2e9 4e5 20 100 250 250 5 5
200 180 2e9 4e5 70 200 500 500 10 10
Fig. 8. UDEC model predicted localized tensile stresses around the 5-2S tailgate under different support patterns. (a) No support; (b) preliminary rock bolt system; and (c) improved rock bolt system.
rock bolt system. The monitored roof delamination versus the face advance is plotted in Fig. 11. The roof delamination increased after the installation of the extensometer and stabilized when the face had advanced approximately 100 m. After the tailgate was finished, delamination in the shallow area (the anchored zone) and the deep area was only 14 mm and 22 mm, respectively, indicating a stable roof. Generally, the tailgate maintained its profile while the roadway developed, as seen in Fig. 12. 5.2. Roadway deformation during the longwall mining period The tailgate served the extraction of the 5-2S panel using the retreat longwall mining method. There was significant concern about roadway deformation caused by mining-induced stresses during the retreating period. The monitored roadway convergence of the 5-2S tailgate during
H.P. Kang et al. / International Journal of Coal Geology 140 (2015) 31–40
0.9 m 0.9 m 0.9 m
38
Welded steel screen
W-shaped strap
3.8 m : roof bolt
: cable
(a) Resin anchored cable (22-mm diameter, 4.3 m long, 580 kN capacity)
3.1 m
Resin anchored bolt (22-mm diameter, 2.4 m long, 180 kN capacity)
3.8 m
(b) Fig. 9. Plan view of the improved support design for the 5-2S tailgate. (a) Top view; (b) vertical view.
convergence and roof sag were 186 mm, 124 mm, and 110 mm, respectively. It can be seen from Fig. 14 that the profile of the tailgate was successfully maintained within 10 m of the longwall face. Significantly, the rib of the tailgate close to the longwall face was not even fractured and the roof was only slightly fractured. 6. Conclusions Supporting coal mine roadway within soft rocks is a typical challenge in underground mining practices. This paper presented an
Wall-to-wall Roof-to-floor Roof sag
250 200
Delaminaon (mm)
Convergence (mm)
300
150 100 50
40
Shallow delaminaon
35
Deep delaminaon
30
Delaminaon-All
25 15
0
50
100
150
200
Distance from roadway face (m) Fig. 10. Monitored convergence of the 5-2S tailgate with face advance.
Shallow
10
Roof line
5 0
0
Deep
20
2m
the longwall mining period is shown in Fig. 13. Tailgate convergence began to increase when the distance between the longwall face and the monitoring station was 70–80 m, and it accelerated considerably when this distance was approximately 40 m. The most dramatic increase in convergence was observed when the distance was approximately 20 m. The continuous measurement of convergence was unavailable due to the installation of hydraulic jacks, trapezoidal I-shaped steel, and timber that provided supplemental support, see Fig. 14. The final measurement was carried out when the longwall face was 16 m ahead and the wall-to-wall convergence, roof-to-floor
0
50
100
150
200
Distance from roadway face (m) Fig. 11. Monitored roof delamination of the 5-2S tailgate with face advance.
H.P. Kang et al. / International Journal of Coal Geology 140 (2015) 31–40
39
Fig. 12. Photographs taken in the 5-2S tailgate showing a successful reinforcement of the soft rock masses by using improved rock bolt system. (a) An overall review; (b) a zoom-in review of the roof.
150
Wall-to-wall Roof-to-floor
100
Roof sag
50
Convergence (mm)
200
0 100
80
60
40
20
0
Distance from roadway face (m) Fig. 13. Monitored convergence of the 5-2S tailgate during the longwall panel retreating period.
investigation on the support pattern of a longwall tailgate excavated in soft rock masses. The tailgate served for the 5-2S longwall panel at the Hongmiao coal mine which was a typical soft rock mine in China. Before the tailgate was driven, a rock bolt system was initially implemented to support it. However, the excavation was halted at 285 m because large deformations were observed in the tailgate behind the working face. Ongoing excavation of the tailgate required an improved rock bolt system, which, in turn, necessitated an improved understanding of the effects of rock bolts on suppressing cracking and dilation in soft rock masses. To achieve this understanding, a discrete element method
simulation based on the study site was conducted. The findings from the numerical simulation indicate that a competent rock bolt system is able to constrain the propagation of shear cracking within the anchored zone, and suppress, or even eliminate, tensile cracking in the vicinity of the roadway surface. Based on the author's experience with coal mine roadway supports and the numerical results, an improvement of the rock bolt system was proposed as follows: (1) adopting high-strength rock bolts with a diameter of 22 mm, a length of 2.4 m and a thread length of 150 mm;(2) adopting high-strength cables with a diameter of 22 mm and a capacity of 600 kN; (3) adopting more competent accessories including high strength load-bearing plates and steel-welded screens; and (4) installing the rock bolts and cables with high pretension. The improved rock bolt system was implemented in the remaining 300 m of the 5-2S tailgate after the initial rock bolt system had failed. The resulting roadway convergence and roof delamination measurements demonstrate that, the improved rock bolt system successfully maintained the stability of the 5-2S tailgate during the entry developing period. The monitored maximum wall-to-wall convergence, roof-tofloor convergence and roof sag were 79 mm, 281 mm and 43 mm, respectively. The monitored maximum roof delamination was only 37 mm, indicating quite a stable roof without fractures. During the longwall mining period, the convergence of the tailgate started to increase when the distance between the longwall face and the monitoring station was 60–80 m and considerably accelerated when the distance was 20–30 m. The roadway convergence deteriorated as the longwall
Timber Unfractured rib
Trapezoidal I-shaped steel Hydraulic jack
(a)
(b)
Fig. 14. Photographs taken in the 5-2S tailgate showing (a) supplemental support installed at 10 m ahead of the 5-2S panel working face, and (b) the rib was not fractured even under mining-induced abutment pressure.
40
H.P. Kang et al. / International Journal of Coal Geology 140 (2015) 31–40
face approached. With the combined supplemental support of hydraulic jacks, trapezoidal I-shaped steel and timber, significant contraction in the roadway cross section was successfully prevented. Findings from this case study will serve to strengthen engineer's confidence in rock bolting systems as support for longwall entries excavated in soft rock masses. References Brown, E.T., 1999. The evolution of support and reinforcement philosophy and practice for underground mining excavations. Proceedings of the International Symposium on Rock Support and Reinforcement Practice in Mining. Balkema, pp. 3–18. Diederichs, M., Kaiser, P., Eberhardt, E., 2004. Damage initiation and propagation in hard rock during tunnelling and the influence of near-face stress rotation. Int. J. Rock Mech. Min. Sci. 41, 785–812. http://dx.doi.org/10.1016/j.ijrmms.2004.02.003. Esterhuizen, G.S., Bajpayee, T., Ellenberger, J.L., Murphy, M., 2013. Practical Estimation of Rock Properties for Modeling Bedded Coal Mine Strata Using Coal Mine Roof Rating. San Francisco, USA, p. Paper ID: 13–154. Feng, X., Jimenez, R., 2014. Bayesian prediction of elastic modulus of intact rocks using their uniaxial compressive strength. Eng. Geol. 173, 32–40. http://dx.doi.org/10. 1016/j.enggeo.2014.02.005. Frith, R., Thomas, R., 1998. The pre-tensioning pedigree. Aust. Min. Mon. Undergr. Equip. Technol. 7, 68–72. Gao, F.Q., Stead, D., 2014. The application of a modified Voronoi logic to brittle fracture modelling at the laboratory and field scale. Int. J. Rock Mech. Min. Sci. 68, 1–14. http://dx.doi.org/10.1016/j.ijrmms.2014.02.003. Gao, F., Stead, D., Kang, H., 2014. Simulation of roof shear failure in coal mine roadways using an innovative UDEC Trigon approach. Comput. Geotech. 61, 33–41. http://dx. doi.org/10.1016/j.compgeo.2014.04.009. Ghazvinian, A., Sarfarazi, V., Schubert, W., Blumel, M., 2012. A study of the failure mechanism of planar non-persistent open joints using PFC2D. Rock Mech. Rock. Eng. 45, 677–693. http://dx.doi.org/10.1007/s00603-012-0233-2. Goel, R.K., Swarup, A., Sheorey, P.R., 2007. Bolt length requirement in underground openings. Int. J. Rock Mech. Min. Sci. 44, 802–811. http://dx.doi.org/10.1016/j.ijrmms. 2006.12.001.
Hagan, P.C., 2003. The effect of resin annulus on anchorage performance of fully encapsulated rockbolts. Proceedings of the 10th congress of international society for rock mechanics. South Africa; 14-16 October 2003, pp. 447–450. Hoek, E., Brown, E.T., 1997. Practical estimates of rock mass strength. Int. J. Rock Mech. Min. Sci. 34, 1165–1186. http://dx.doi.org/10.1016/S1365-1609(97)80069-X. Itasca Consulting Group Inc, 2014. UDEC (Universal Distinct Element Code), Version 6.0 (Itasca, Minneapolis). Jiao, Y.-Y., Song, L., Wang, X.-Z., Coffi Adoko, A., 2013. Improvement of the U-shaped steel sets for supporting the roadways in loose thick coal seam. Int. J. Rock Mech. Min. Sci. 60, 19–25. http://dx.doi.org/10.1016/j.ijrmms.2012.12.038. Kang, H., 2013. In-situ Geomechanics Measurements for Coal and Rock Masses and Their Application on Strata Control. 1st ed. Science Press, Beijing. Kang, H., Wang, J., Lin, J., 2010. Case studies of rock bolting in coal mine roadways. Chin. J. Rock Mech. Eng. 29, 649–664. Karanam, U.M.R., Dasyapu, S.K., 2005. Experimental and numerical investigations of stresses in a fully grouted rock bolts. Geotech. Geol. Eng. 23, 297–308. http://dx.doi. org/10.1007/s10706-004-9518-x. Li, C.C., 2006. Rock support design based on the concept of pressure arch. Int. J. Rock Mech. Min. Sci. 43, 1083–1090. http://dx.doi.org/10.1016/j.ijrmms.2006.02.007. Mark, C., 2000. Design of roof bolt system. Proceeding of New Technology for Coal Mine Roof Support. National Institute for Occupational Safety and Health (NIOSH), USA, pp. 111–131 (Pittsburgh). Peng, S., 1998. Roof bolting adds stability to weak strata. Coal Age Mag. 11, 32–38. Peng, S.S., Tang, D.H.Y., 1984. Roof bolting in underground mining: a state-of-the-art review. Int. J. Min. Eng. 2, 1–42. http://dx.doi.org/10.1007/BF00880855. Stankus, J., Peng, S., 1996. A new concept for roof support. Coal Age Mag. 9, 2–6. Wang, C., Wang, Y., Lu, S., 2000. Deformational behaviour of roadways in soft rocks in underground coal mines and principles for stability control. Int. J. Rock Mech. Min. Sci. 37, 937–946. http://dx.doi.org/10.1016/S1365-1609(00)00026-5. Yoshinaka, R., Tran, T.V., Osada, M., 1998. Non-linear, stress- and strain-dependent behavior of soft rocks under cyclic triaxial conditions. Int. J. Rock Mech. Min. Sci. 35, 941–955. http://dx.doi.org/10.1016/S0148-9062(98)00158-2. Zhang, N., Gao, M., 2004. High-strength and pretension bolting support of coal roadway and its application. J. China Univ. Min. Technol. 33, 524–527.
Contents lists available at ScienceDirect
International Journal of Coal Geology journal homepage: www.elsevier.com/locate/ijcoalgeo
Investigation on support pattern of a coal mine roadway within soft rocks — a case study H.P. Kang ⁎, J. Lin, M.J. Fan Mining & Designing Branch, China Coal Research Institute, Beijing, China State Key Laboratory of Coal Mining and Clean Utilization (China Coal Research Institute), Beijing, China
a r t i c l e
i n f o
Article history: Received 28 September 2014 Received in revised form 6 January 2015 Accepted 6 January 2015 Available online 30 January 2015 Keywords: Soft rock Ground reinforcement Rock bolt Discrete element method Coal mine
a b s t r a c t Supporting coal mine roadway within soft rocks is a typical challenge in underground mining practices. As the most widely used support structure, rock bolt system has been successfully used to support coal mine roadways in various complex geological and geotechnical conditions, including roadway in extremely weak rock masses, roadways 1000 m below ground, open-off cut roadways with large section, and roadways along the edges of mined-out areas with thin pillars or even no pillars. However, the effectiveness and applicability of rock bolt system for the reinforcement of soft rock masses has not yet been established. In this paper, we present a case study of rock bolt system as used for ground reinforcement of longwall entries within soft rock masses. The study site was the tailgate of the longwall panel 5-2S at the Hongmiao coal mine in the Pingzhuang coalmining district in China, a typical soft rock coal mine of the region. Since traditional rock bolt system failed to maintain the stability of the tailgate, the effectiveness of rock bolt systems for this application remained doubtful. The reasons for the failure of the rock bolt system were first examined. A discrete element method simulation was then performed to better understand the mechanism of rock bolts in supporting soft rocks. An improved rock bolt system was finally proposed that was used to support the tailgate. The field monitor showed that the improved rock bolt system successfully suppressed cracking and dilation of the tailgate. This case study is useful in enhancing engineering applications of rock bolts to support longwall entries excavated in soft rock masses. © 2015 Elsevier B.V. All rights reserved.
1. Introduction Rock bolt systems have been extensively used for ground reinforcement in underground coal mine roadways around the world (Brown, 1999; Peng and Tang, 1984). In China, more than 70% of roadways in key state-owned coal mines are supported by the system (Kang et al., 2010). As an active support pattern, rock bolts suppress the deformation of and damage to rock masses immediately after installation. Rock bolts are economical, easily installed and reliably supportive. The principal objective of rock bolting is to help rock masses support themselves. Bolts increase the stress and the frictional strength across joints, causing loose blocks or thinly stratified beds to become wedged together and act as a composite beam (Goel et al., 2007; Karanam and Dasyapu, 2005; Mark, 2000). Rock bolts are generally installed with pretension. The high pretension applied during installation has been proven to be a very important component of the supportive effect of rock bolts (Li, 2006). Pre-tensioned rock bolts compress and reinforce the rock mass in their vicinity. This effect spreads to a further section of the rock surface through accessories such as load-bearing plates and screens, creating a confining pressure on the rock surface (Ghazvinian et al., ⁎ Corresponding author. Tel.: +86 84263125. E-mail address: [email protected] (H.P. Kang).
http://dx.doi.org/10.1016/j.coal.2015.01.003 0166-5162/© 2015 Elsevier B.V. All rights reserved.
2012). Active pretension modifies roof behavior by dramatically reducing bed separation and delamination immediately within 0.5–0.8 m of the roof (Frith and Thomas, 1998). In addition, thinly laminated roof beds can be clamped into a thick beam that is more resistant to bending (Peng, 1998). Bolts installed with pretension increase frictional resistance along bedding planes, minimizing roof sag and defection, and decreasing the likelihood of lateral movement due to horizontal stress (Stankus and Peng, 1996). Rock bolt systems have been successfully used in complex geological and geotechnical conditions, such as roadways 1000 m below ground, open-off cut roadways with large section and roadways along the edges of mined-out areas with thin pillars or even no pillars (Kang, 2013; Zhang and Gao, 2004). However, the effectiveness and applicability of rock bolt system for the reinforcement of soft rock masses has not yet been established (Jiao et al., 2013; Wang et al., 2000; Yoshinaka et al., 1998). Soft rock refers to a group of geotechnical materials that has a uniaxial compressive strength between 0.5 and 25 MPa and has similar geotechnical characteristics, including slaking, swelling, compressibility, time dependence, and volume change (Jiao et al., 2013; Wang et al., 2000; Yoshinaka et al., 1998). Roadways caved out of soft rocks exhibit large deformation in the form of severe roof sag, wall convexity, and floor heave. These roadways also have distinct characteristics that include a high rate of rapid deformation, a high sensitivity to
32
H.P. Kang et al. / International Journal of Coal Geology 140 (2015) 31–40
water invasion and stress changes, evident rheological deformation and deformation in all directions of the roadway cross section (Wang et al., 2000). Maintaining the stability of roadways excavated in soft rocks has been a great challenge for underground coal mining engineers. In cases where roadway closure is severe, rehabilitation and re-support of the roadway have to be continually arranged during the roadway's service period to maintain its cross section and guarantee that it remains functional for ventilation and transportation. This causes safety issues, delays of mining activities, and results in dire economic loss. Some researchers suggested that yieldable and flexible support patterns should be used to support soft rock roadways, instead of stiff supports such as the rock bolt system (Jiao et al., 2013; Wang et al., 2000). A common structure for supporting roadways in soft rocks is the installation of yieldable U-shaped steel sets. This support pattern is believed to provide sufficient support to suppress the deformation of soft rocks surrounding a roadway and has been proven to be successful where implemented (Jiao et al., 2013; Wang et al., 2000). In contrast to rock bolting, the yieldable U-shaped steel sets are passive, costly, and time-consuming to install. In underground coal mine practice, yieldable U-shaped steel sets are more suitable to support main roadways with a long-term service period and the absence of mining activities (i.e., retreat longwall mining) where rock bolting is not applicable. In longwall entries intended for the extraction of a panel, yieldable U-shaped steel sets are not effective; they have a detrimental impact on the shearer and must therefore be removed, at high cost. This paper presents a case study on the application of rock bolt systems for ground reinforcement of longwall entries within soft rock masses. The study site was the tailgate of the longwall panel 5-2S at the Hongmiao coal mine, Pingzhuang coal-mining district, China. Traditional rock bolt system failed to support the entries. The failure reasons were first examined. A discrete element method simulation was then performed to investigate the effects of rock bolts on suppressing cracking and dilation in soft rock masses. An improvement on the rock bolt system was finally proposed and used to support the tailgate. Field monitoring of roadway deformation was conducted to evaluate the success of the improved rock bolt system in maintaining the stability of the tailgate.
Thickness (m)
Column
Lithology
Geological descripons
Coal 5-1
light, shiny, half-intact
2.63-4.94 3.79
Sandy mudstone light, shiny, carbonaceous shale laminates
Coal 5-2
3.66-7.00 5.99
Sandy mudstone Coal 5-3
light, shiny, half-intact
Fine-grained sandstone
White, argillaceous cement
2.55-8.04 5.30 0.6-15.0 5.30
Coal 6-1
light, shiny, half-intact
Fine-grained sandstone
White, argillaceous cement
1.67-5.58 3.50
Fig. 2. Lithological descriptions of the rock units at the Hongmiao coal mine.
strength of 6.2 MPa. It was sub-horizontal with a dip angle of 15°. Below the Coal 5-2 was found a sandy mudstone with a thickness of 1.0–2.0 m and a compressive strength of 13.4 MPa. The Coal 5-3 was 2.55–8.04 m thick and was situated immediately below the sandy mudstone. Below the Coal 5-3 was a fine-grained sandstone with a thickness varying between 0.6 and 15 m. The material situated immediately below the finegrained sandstone was the Coal 6-1 with a thickness of 1.67–5.58 m. The Coal 5-1 had been mined out, and current mining activity was extracting the Coal 5-2 using the longwall mining method. The longwall panel 5-2S was the first panel in the Coal 5-2. The depth of this panel varied between 350 and 400 m. The panel 5-2 had a length of 560 m in the strike direction and a width of 156 m in the dip direction of the Coal 5-2, see Fig. 2. In situ tests including in situ stress measurement, borehole strength measurement and borehole optical televiewer imaging had been performed in a vertical borehole in the main roadway close to the panel 5-2S, see Fig. 1(b). The borehole was 56 mm in diameter and 20 m in length. The in situ stress was measured using the hydraulic fracturing method and the results were σH = 14.62 MPa, σh = 7.35 MPa, and σv = 9.68 MPa. The direction of σH was N72oE, leading to an angle of approximately 80° with respect to the direction of the 5-2S tailgate. In borehole strength tests, a rod with a probe attached to the top was
2. Geological and geotechnical conditions The Hongmiao coal mine, Pingzhuang coal-mining district was a typical soft rock mine located in the Inner Mongolia Autonomous Region in China, see Fig. 1(a). It extracted coal seams formed during the Cretaceous period. The coal seams were separated by a few rock layers, as illustrated in Fig. 2. The first coal layer was the Coal 5-1 with a thickness varying between 2.63 and 4.94 m. The material situated immediately below the Coal 5-1 was made up of 0.5–2.0 m of sandy mudstone. Below the sandy mudstone was the Coal 5-2. The thickness of the Coal 5-2 varied between 3.66 and 7.0 m, and it had a compressive
N
Beijing
㚄㔌ᐧ
Hongmiao Coal Mine
285 m
300 m Tailgate Longwall panel 5-2S Headgate
Preliminary rock bolt system Displacement monitoring staon
(a)
Improved rock bolt system In situ tests posion
(b)
Fig. 1. (a) The location of the Hongmiao coal mine, Pingzhuang coal mining district, China. (b) Plan view of the roadways and panel at the study site in the Hongmiao coal mine.
H.P. Kang et al. / International Journal of Coal Geology 140 (2015) 31–40
inserted into the borehole. The probe was connected to a high pressure pump with a pipe. Under the hydraulic pressure generated by the pump, the probe was pushed against the borehole wall until failure, identified by a sudden increase in the displacement of the probe and a constant pressure value. The pressure was recorded and used to estimate the rock strength, according to the relationship between the pressure and the unconfined compressive strength (UCS) of the rock
33
which was established based on laboratory tests on different types of rocks. The borehole strength test results indicated that the roof rocks in the roof were generally of poor to average quality (Hoek and Brown, 1997). Borehole optical televiewer imaging showed that the roof was heavily fractured, see Fig. 3. Furthermore, the rocks were typical soft rock and displayed considerable swelling where they had encountered water.
(a) 1.6 m
(b) 2.3 m
(c) 3.6 m
(d) 4.1 m
(e) 7.2 m
(f) 8.5 m
(g) 15.5 m
(h) 16.1 m
Fig. 3. Borehole optical televiewer image showing a highly fractured roof at the study site.
34
H.P. Kang et al. / International Journal of Coal Geology 140 (2015) 31–40
Fortunately, only small amounts of water had penetrated the roadway, significantly reducing the swelling degree of the roadway's moisturesensitive rocks. The tailgate of the panel 5-2S was initially driven in a round arch cross section and supported by rock bolts. The rock bolts were 2.0 m long with a diameter of 18 mm and installed with a spacing of 0.9 m. As the tailgate was excavating, significant deformation of the roadways occurred tens of meters behind the advancing face, especially when the roadways were driven below the pillar of the above mined-out panel in the Coal 5-1. The coal surrounding the tailgate was heavily fractured due to the severe mining-induced stresses caused by the extraction of the above panel. The roadway cross section was difficult to obtain using the road header. Skin roof falls and rib falls were frequently observed in the advancing face and continued to occur even after rock bolts, cables, and screens were installed, leading to ‘string bags’ as shown in Fig. 4. The excavation of the tailgate was suspended due to severe instability and deformation. It was expected that these issues would significantly deteriorate with the consequent extraction of the panel 5-2S. An improvement of the rock bolt system had to be performed. 3. Discrete element method simulation of rock bolting The improvement of the rock bolt system necessitated a better understanding of the mechanism of rock bolts in suppressing cracking and dilation of soft rock masses around the tailgate. To achieve this, a numerical study was carried out. 3.1. Model configuration The UDEC Trigon approach proposed by Gao and Stead (2014) was adopted in this study due to its intrinsic capability to simulate the initiation, propagation, and coalescence of fractures, as well as the interaction between them and any pre-existing discontinuities. In the UDEC Trigon approach a rock mass is represented as an assembly of triangular blocks cemented at their contacts. A block is represented as a mesh of finite difference elements and is deformable and elastic. The fracturing process is represented by the sliding or opening of the contact, depending on its strength and the stresses it suffers. A tensile crack forms when the normal stress applied on the contact reaches the contact's tensile strength. A shear crack forms when the shear stress reaches the contact's shear strength, which is governed by c + σn tanϕ, where c and ϕ are the cohesion and friction angle of the contact, respectively, and σn is the normal stress applied to the contact. Gao and Stead illustrate that the UDEC Trigon approach can capture many aspects of the mechanical behavior of rocks from laboratory to rock mass scale (Gao
and Stead, 2014). Furthermore, the ability of rock bolts in suppressing rock mass dilation can be successfully simulated using this approach (Gao et al., 2014). By using the UDEC Trigon approach, a numerical model was created to simulate the effects of rock bolting in suppressing fracturing and dilation of soft rocks. Fig. 5 shows the geometry of the model which was based on the lithology of the 5-2S tailgate as illustrated in Fig. 2. In an effort to increase computational efficiency, the UDEC Trigon approach was only adopted for generating triangular blocks in the area of interest surrounding the roadway. These triangular blocks were elastic and rock mass failure was represented as cracking through the contacts between them. The average edge length of these triangular blocks was 0.25 m. This block size had been shown to be sufficiently fine to simulate crack propagation. The remainder of the model was simulated by big blocks which were represented as finite difference elements. These finite difference elements obeyed the Mohr–Coulomb constitutive mode and, as a result, rock mass failure was represented as plastic zones. The intact rock strength properties for the three types of rocks—coal, sandy mudstone, and fine-grained sandstone—were obtained from borehole in situ strength tests (as listed in Table 1). Young's modulus values were estimated as Ec = MR ∗ σc (Feng and Jimenez, 2014). The intact rock strength properties were multiplied by a scale factor of 0.58 to obtain rock mass strength properties (Esterhuizen et al., 2013). The tensile strength of the rock masses was reasonably estimated to be 1/12 of the compressive strength. The mechanical properties used in the model are listed in Table 2. The mechanical behavior of the coal and sandy mudstone, represented as triangular blocks in the model, was mainly governed by their contact properties. These contact properties were calibrated against the rock mass properties listed in Table 1. For the remainder rock masses represented as big blocks which were represented as finite difference elements, the Mohr–Coulomb properties were assigned. Note that bedding planes were incorporated into each layer and their contact properties are given in Table 2. The UDEC built-in “Cable” element was used to simulate rock bolts, and the “Liner” element simulated steel joists and straps. The parameters of the “Cable” and “Liner” elements are presented in Table 2. A detailed description of the two support elements in UDEC can be referenced at Itasca Consulting Group Inc. (2014). The field-measured in situ stresses were applied to the UDEC model to set up the pre-mining stress field. Note that the extraction of the 5-1 coal seam was not accounted for in the model because the mined-out area in 5-1 coal seam above the 5-2S tailgate had stabilized prior to the excavation and had no significant influence on the ground behavior of the 5-2S tailgate. The model was run under two different roadway
Fig. 4. Photographs taken in the tailgate of panel 5-2S at the Hongmiao coal mine showing a heavy roughness of roadway profile due to large dilation and skin falls. (a) Roof, and (b) rib.
H.P. Kang et al. / International Journal of Coal Geology 140 (2015) 31–40
35
Applied vercal stress to compensate for overburden pressure
3.8 m
3.1 m
60 m
60 m
Fig. 5. Configuration of the UDEC model used for simulating rock bolts in suppressing cracking and dilation within soft rock masses around the 5-2S tailgate.
support conditions, one with no support and the second with a rock bolt system which corresponded with the preliminary support parameters previously discussed. 3.2. Modeling results There are two components to the supportive effects of pre-tensioned resin-anchored rock bolts in Coal Measure rock masses (Gao et al., 2014): i) Reinforcement is provided by the rebar that bonded into the rock strata. The rebar provides axial resistance along its length, which restrains the deformation of the rock mass, i.e. the opening of pre-existing cracks. The rebar also provides shear resistance along pre-existing fractures so that sliding is restrained. ii) Surface support is provided by rock bolt accessories, including the load-bearing plate, strap and screen. The pre-tension applied to the bolts, which extends through the plate and strap provides confinement on the roadway surface. Even though this confinement is very low when compared with the rock mass stress, it significantly increases the residual strength of the supported rock mass. In combination, the reinforcement and support greatly restrain crack propagation. These effects were successfully captured in the UDEC model as shown in Fig. 6. When the roadway was not supported, heavy cracks were observed around it. Shear cracks were majorly developed in relatively deep zones while tensile cracks were mainly developed close to the roadway surface, see Fig. 6(a). These tensile cracks caused rock mass dilation and subsequently, led to significant deformation and roadway convergence, see Fig. 7(a). Comparing Fig. 6(b) with Fig. 6(a), it is apparent that the
Table 1 Intact rock properties and scaled rock mass properties of Coal Measures at the Hongmiao coal mine. Coal Measures
Coal Sandy mudstone Fine-grained sandstone
Intact rock σc (MPa)
MR
6.2 13.4 35.3
425 425 333
Ec (MPa) 2 600 5 700 11 800
Scale factor
0.58 0.58 0.58
Rock mass Em (GPa)
σcm (MPa)
σtm (MPa)
2.6 5.7 11.8
3.6 7.8 20.3
0.3 0.6 1.7
*E is Young's Modulus, kn and kn are the normal and shear stiffness, respectively.
installation of the initial rock bolt system significantly suppressed the crack propagation within the roof and ribs where the rock bolts were installed. The predicted depths of shear cracking in the roof and ribs were 2.6 m and 1.6 m, respectively, compared to 3.5 m and 2.2 m with no support. The depth and extent of tensile cracking were also significantly reduced, leading to notable decrease in rock mass dilation and roadway convergence (see Fig. 7(b)). Even though this rock bolt system greatly reduced cracking and dilation in the rock masses, it appears from the numerical model that the initial rock bolt system did not successfully stabilize the roadway. For example, extensive tensile cracks were observed in the roof and ribs (Fig. 6(b)). As a consequence, dilation and large deformation occurred in these areas as shown in Fig. 7(b), which was consistent with the ‘string bags’ observed in the field (see Fig. 4). Moreover, the numerical predicted cracking was beyond the anchored zone (see Fig. 7(b)), which suggests the potential for a sudden roof collapse. An improvement to the initial rock bolt system was subsequently simulated in an attempt to examine if the improvement could successfully control fracturing and dilation of the soft rock masses. There were four aspects to the improvement, as illustrated in Table 3: (1) the tensile strength of the rock bolts was increased from 120 kN to 180 kN which corresponded with 22-mm diameter rebar with a steel tensile strength of 470 MPa; (2) the rock bolt length was increased from 2.0 m to 2.4 m; (3) the pre-tension was increased from 20 kN to 70 kN; and (4) the stiffness and strength of the “Linear” element were doubled to account for more robust surface control accessories such as a welded steel screen and a big W-shaped load-bearing plate. As can be seen in Fig. 6(c), the improvement was successful at suppressing cracking around the roadway. In the roof, the predicted depth of shear cracking was approximately 2.0 m, substantially less than the 2.6 m predicted for the initial rock bolt system. More importantly, cracking was constrained within the anchored zone. Cracking in the ribs was also significantly reduced. Furthermore, tensile cracking barely propagated in the roof and ribs where the improved rock bolt system had been installed. Consequently, there was no noticeable dilation, see Fig. 7(c). An interesting outcome from the numerical modeling was that the cracking in the roadway floor appears to be reduced even though no rock bolts were installed in that area (compare (a), (b) and (c) in Fig. 6). This suggests that competent reinforcement in the roof and ribs may have a positive effect on floor stability.
36
H.P. Kang et al. / International Journal of Coal Geology 140 (2015) 31–40
Table 2 Mechanical properties used in the UDEC Trigon model. Lithology
Coal Sandy mudstone Fine-grained sandstone a
Matrix properties
Contact properties
Density (kg·m−3)
E (GPa)
Cohesion (MPa)
Friction angle
Tensile strength
kn (GPa/m)
ks (GPa/m)
Cohesion (MPa)
Friction (°)
Tensile strength (MPa)
1400 2600 2600
3.0 10.0 15.0
1.0 2.0 10. 5
23 28 33
0.3 0.6 1.7
14.4 18.0 27.0
5.8 7.2 10.8
1.0/0.5a 2.0/1.0a NA/3.0a
23/23a 28/28a NA/33a
0.3/0a 1.0/0a NA/0a
The first value refers to triangular block contacts representing rock mass, while the second refers to persistent contacts representing bedding planes.
It is also interesting to note that a rock bolt system seems to constrain the generation of tensile stresses in the anchored zone, see Fig. 8. Localized tensile stresses significantly prompt tensile crack propagation, as stated by Diederichs et al. (2004) and many others.
designed to be installed with a moment of at least 400 Nm, inducing a pretension greater than 70 kN. The cables were designed to be installed with a pretension of 200 kN. The improved rock bolt system, as shown in Fig. 9, was implemented in the remainder of the 5-2S tailgate at the Hongmiao coal mine.
4. Improvement of the rock bolt system
5. Field monitoring
The author's experience with coal mine roadways supports and the numerical simulations suggest that poor ground reinforcement in the existing stabilization system was likely a result of the following reasons: (1) The installed rock bolts were 18 mm in diameter with a tensile capability of approximately 120 kN, which was too weak for this roadway. Many fractured rock bolts were found in the entries. (2) The thread of the rock bolts was too short, at only 80 mm. The high degree of roadway surface roughness implied that, the contact between the plate and roadway surface was likely loose, resulting in an ineffective support. (3) The installed cables were 15.2 mm in diameter with a fracture force of 260 kN which was weak for the circumstances. Furthermore, the cable was installed in a borehole with a diameter of 28 mm, leading to a resin annulus thickness of 6.4 mm-far beyond the optimum annulus range of 2–4 mm (Hagan, 2003). (4) The screen and the steel joists were too weak to transfer the applied compression on the rock bolts to the roadway surface, leading to a low confining pressure. In addition, the screen was not strong enough to hold detached rocks. (5) The rock bolts were installed with low pretension. An improvement of the preliminary rock bolt system was proposed. The most substantial improvement included: (1) adopting highstrength rock bolts with a diameter of 22 mm, a length of 2.4 m and a thread length of 150 mm. The tensile capability of these bolts was more than 180 kN. The bolts were designed to be fully encapsulated; (2) adopting high-strength cables with a diameter of 22 mm and a capacity of 600 kN; (3) rock bolts were installed with a high strength load-bearing plate, W-shaped steel plates that were 280 mm wide, 450 mm long and 4 mm thick and welded steel screen with a mesh size of 100 × 100 mm for screen control; (4) the rock bolts were
A monitoring station was set up in the 5-2S tailgate to evaluate the effectiveness of the improved rock bolt system, see Fig. 1(b). At the monitoring station, the convergence of the roadway was measured regularly as the roadway advances. Two-point borehole extensometers were installed in the roadway roof to measure roof delamination. The monitoring instruments were installed immediately behind the advancing face, in an attempt to monitor the entire roadway deformation as it advanced. 5.1. Roadway deformation during its developing period The cumulative amounts of the cross section convergence of the 5-2S tailgate versus roadway advance are plotted in Fig. 10. The convergence of the tailgate began right after its excavation and continued to increase with face advance. The wall-to-wall convergence and roof sag stabilized after the face had advanced more than 53 m away from the monitoring station, while the roof-to-floor convergence kept increasing as the face advance, indicating an increasing floor heave. When the excavation of the 5-2S tailgate was finished, the wall-to-wall convergence, roof-to-floor convergence and roof sag were 79 mm, 281 mm and 43 mm, respectively. The wall-to-wall convergence was only 2% of the roadway width, demonstrating that the improved rock bolt system had successfully controlled wall-to-wall convergence. The substantial roof-to-floor convergence was attributed to the floor heave which was 238 mm. It should be noted that significant floor heave was only observed in two short sections (approximately 50 m in total) along the portion of the 5-2S tailgate that was supported by the improved
S/T = 1671/480
S/T = 1050/222
S/T = 775/140
(a)
(b)
(c)
Fig. 6. UDEC model predicted fracture modes of the soft strata around the 5-2S tailgate under different support patterns. (a) No support; (b) preliminary rock bolt system; and (c) improved rock bolt system. Red and blue refer to tensile and shear cracks, respectively, and orange and red refer to rock bolts and steel joist in (b) and steel strap in (c). S/T indicates the ratio of shear crack number to tensile crack number. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)
H.P. Kang et al. / International Journal of Coal Geology 140 (2015) 31–40
Disp.(mm) 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150
(a)
37
Tensile stress (MPa) 0.0 0.2 0.4 0.6 0.8 1.0
(a)
(b)
(b)
(c)
(c) Fig. 7. UDEC model predicted displacement vectors around the 5-2S tailgate under different support patterns. (a) No support; (b) preliminary rock bolt system; and (c) improved rock bolt system.
Table 3 Support element properties used in the UDEC Trigon model. Properties Cable
Elastic modulus (GPa) Tensile capability (kN) Stiffness of the grout (N/m/m) Cohesive capacity of the grout (N/m) Pre-tension (kN) Linear Elastic modulus (GPa) Tensile yield strength (MPa) Compressive yield strength (MPa) Interface normal stiffness (GPa/m) Interface shear stiffness (GPa/m)
Initial support Optimized support 200 120 2e9 4e5 20 100 250 250 5 5
200 180 2e9 4e5 70 200 500 500 10 10
Fig. 8. UDEC model predicted localized tensile stresses around the 5-2S tailgate under different support patterns. (a) No support; (b) preliminary rock bolt system; and (c) improved rock bolt system.
rock bolt system. The monitored roof delamination versus the face advance is plotted in Fig. 11. The roof delamination increased after the installation of the extensometer and stabilized when the face had advanced approximately 100 m. After the tailgate was finished, delamination in the shallow area (the anchored zone) and the deep area was only 14 mm and 22 mm, respectively, indicating a stable roof. Generally, the tailgate maintained its profile while the roadway developed, as seen in Fig. 12. 5.2. Roadway deformation during the longwall mining period The tailgate served the extraction of the 5-2S panel using the retreat longwall mining method. There was significant concern about roadway deformation caused by mining-induced stresses during the retreating period. The monitored roadway convergence of the 5-2S tailgate during
H.P. Kang et al. / International Journal of Coal Geology 140 (2015) 31–40
0.9 m 0.9 m 0.9 m
38
Welded steel screen
W-shaped strap
3.8 m : roof bolt
: cable
(a) Resin anchored cable (22-mm diameter, 4.3 m long, 580 kN capacity)
3.1 m
Resin anchored bolt (22-mm diameter, 2.4 m long, 180 kN capacity)
3.8 m
(b) Fig. 9. Plan view of the improved support design for the 5-2S tailgate. (a) Top view; (b) vertical view.
convergence and roof sag were 186 mm, 124 mm, and 110 mm, respectively. It can be seen from Fig. 14 that the profile of the tailgate was successfully maintained within 10 m of the longwall face. Significantly, the rib of the tailgate close to the longwall face was not even fractured and the roof was only slightly fractured. 6. Conclusions Supporting coal mine roadway within soft rocks is a typical challenge in underground mining practices. This paper presented an
Wall-to-wall Roof-to-floor Roof sag
250 200
Delaminaon (mm)
Convergence (mm)
300
150 100 50
40
Shallow delaminaon
35
Deep delaminaon
30
Delaminaon-All
25 15
0
50
100
150
200
Distance from roadway face (m) Fig. 10. Monitored convergence of the 5-2S tailgate with face advance.
Shallow
10
Roof line
5 0
0
Deep
20
2m
the longwall mining period is shown in Fig. 13. Tailgate convergence began to increase when the distance between the longwall face and the monitoring station was 70–80 m, and it accelerated considerably when this distance was approximately 40 m. The most dramatic increase in convergence was observed when the distance was approximately 20 m. The continuous measurement of convergence was unavailable due to the installation of hydraulic jacks, trapezoidal I-shaped steel, and timber that provided supplemental support, see Fig. 14. The final measurement was carried out when the longwall face was 16 m ahead and the wall-to-wall convergence, roof-to-floor
0
50
100
150
200
Distance from roadway face (m) Fig. 11. Monitored roof delamination of the 5-2S tailgate with face advance.
H.P. Kang et al. / International Journal of Coal Geology 140 (2015) 31–40
39
Fig. 12. Photographs taken in the 5-2S tailgate showing a successful reinforcement of the soft rock masses by using improved rock bolt system. (a) An overall review; (b) a zoom-in review of the roof.
150
Wall-to-wall Roof-to-floor
100
Roof sag
50
Convergence (mm)
200
0 100
80
60
40
20
0
Distance from roadway face (m) Fig. 13. Monitored convergence of the 5-2S tailgate during the longwall panel retreating period.
investigation on the support pattern of a longwall tailgate excavated in soft rock masses. The tailgate served for the 5-2S longwall panel at the Hongmiao coal mine which was a typical soft rock mine in China. Before the tailgate was driven, a rock bolt system was initially implemented to support it. However, the excavation was halted at 285 m because large deformations were observed in the tailgate behind the working face. Ongoing excavation of the tailgate required an improved rock bolt system, which, in turn, necessitated an improved understanding of the effects of rock bolts on suppressing cracking and dilation in soft rock masses. To achieve this understanding, a discrete element method
simulation based on the study site was conducted. The findings from the numerical simulation indicate that a competent rock bolt system is able to constrain the propagation of shear cracking within the anchored zone, and suppress, or even eliminate, tensile cracking in the vicinity of the roadway surface. Based on the author's experience with coal mine roadway supports and the numerical results, an improvement of the rock bolt system was proposed as follows: (1) adopting high-strength rock bolts with a diameter of 22 mm, a length of 2.4 m and a thread length of 150 mm;(2) adopting high-strength cables with a diameter of 22 mm and a capacity of 600 kN; (3) adopting more competent accessories including high strength load-bearing plates and steel-welded screens; and (4) installing the rock bolts and cables with high pretension. The improved rock bolt system was implemented in the remaining 300 m of the 5-2S tailgate after the initial rock bolt system had failed. The resulting roadway convergence and roof delamination measurements demonstrate that, the improved rock bolt system successfully maintained the stability of the 5-2S tailgate during the entry developing period. The monitored maximum wall-to-wall convergence, roof-tofloor convergence and roof sag were 79 mm, 281 mm and 43 mm, respectively. The monitored maximum roof delamination was only 37 mm, indicating quite a stable roof without fractures. During the longwall mining period, the convergence of the tailgate started to increase when the distance between the longwall face and the monitoring station was 60–80 m and considerably accelerated when the distance was 20–30 m. The roadway convergence deteriorated as the longwall
Timber Unfractured rib
Trapezoidal I-shaped steel Hydraulic jack
(a)
(b)
Fig. 14. Photographs taken in the 5-2S tailgate showing (a) supplemental support installed at 10 m ahead of the 5-2S panel working face, and (b) the rib was not fractured even under mining-induced abutment pressure.
40
H.P. Kang et al. / International Journal of Coal Geology 140 (2015) 31–40
face approached. With the combined supplemental support of hydraulic jacks, trapezoidal I-shaped steel and timber, significant contraction in the roadway cross section was successfully prevented. Findings from this case study will serve to strengthen engineer's confidence in rock bolting systems as support for longwall entries excavated in soft rock masses. References Brown, E.T., 1999. The evolution of support and reinforcement philosophy and practice for underground mining excavations. Proceedings of the International Symposium on Rock Support and Reinforcement Practice in Mining. Balkema, pp. 3–18. Diederichs, M., Kaiser, P., Eberhardt, E., 2004. Damage initiation and propagation in hard rock during tunnelling and the influence of near-face stress rotation. Int. J. Rock Mech. Min. Sci. 41, 785–812. http://dx.doi.org/10.1016/j.ijrmms.2004.02.003. Esterhuizen, G.S., Bajpayee, T., Ellenberger, J.L., Murphy, M., 2013. Practical Estimation of Rock Properties for Modeling Bedded Coal Mine Strata Using Coal Mine Roof Rating. San Francisco, USA, p. Paper ID: 13–154. Feng, X., Jimenez, R., 2014. Bayesian prediction of elastic modulus of intact rocks using their uniaxial compressive strength. Eng. Geol. 173, 32–40. http://dx.doi.org/10. 1016/j.enggeo.2014.02.005. Frith, R., Thomas, R., 1998. The pre-tensioning pedigree. Aust. Min. Mon. Undergr. Equip. Technol. 7, 68–72. Gao, F.Q., Stead, D., 2014. The application of a modified Voronoi logic to brittle fracture modelling at the laboratory and field scale. Int. J. Rock Mech. Min. Sci. 68, 1–14. http://dx.doi.org/10.1016/j.ijrmms.2014.02.003. Gao, F., Stead, D., Kang, H., 2014. Simulation of roof shear failure in coal mine roadways using an innovative UDEC Trigon approach. Comput. Geotech. 61, 33–41. http://dx. doi.org/10.1016/j.compgeo.2014.04.009. Ghazvinian, A., Sarfarazi, V., Schubert, W., Blumel, M., 2012. A study of the failure mechanism of planar non-persistent open joints using PFC2D. Rock Mech. Rock. Eng. 45, 677–693. http://dx.doi.org/10.1007/s00603-012-0233-2. Goel, R.K., Swarup, A., Sheorey, P.R., 2007. Bolt length requirement in underground openings. Int. J. Rock Mech. Min. Sci. 44, 802–811. http://dx.doi.org/10.1016/j.ijrmms. 2006.12.001.
Hagan, P.C., 2003. The effect of resin annulus on anchorage performance of fully encapsulated rockbolts. Proceedings of the 10th congress of international society for rock mechanics. South Africa; 14-16 October 2003, pp. 447–450. Hoek, E., Brown, E.T., 1997. Practical estimates of rock mass strength. Int. J. Rock Mech. Min. Sci. 34, 1165–1186. http://dx.doi.org/10.1016/S1365-1609(97)80069-X. Itasca Consulting Group Inc, 2014. UDEC (Universal Distinct Element Code), Version 6.0 (Itasca, Minneapolis). Jiao, Y.-Y., Song, L., Wang, X.-Z., Coffi Adoko, A., 2013. Improvement of the U-shaped steel sets for supporting the roadways in loose thick coal seam. Int. J. Rock Mech. Min. Sci. 60, 19–25. http://dx.doi.org/10.1016/j.ijrmms.2012.12.038. Kang, H., 2013. In-situ Geomechanics Measurements for Coal and Rock Masses and Their Application on Strata Control. 1st ed. Science Press, Beijing. Kang, H., Wang, J., Lin, J., 2010. Case studies of rock bolting in coal mine roadways. Chin. J. Rock Mech. Eng. 29, 649–664. Karanam, U.M.R., Dasyapu, S.K., 2005. Experimental and numerical investigations of stresses in a fully grouted rock bolts. Geotech. Geol. Eng. 23, 297–308. http://dx.doi. org/10.1007/s10706-004-9518-x. Li, C.C., 2006. Rock support design based on the concept of pressure arch. Int. J. Rock Mech. Min. Sci. 43, 1083–1090. http://dx.doi.org/10.1016/j.ijrmms.2006.02.007. Mark, C., 2000. Design of roof bolt system. Proceeding of New Technology for Coal Mine Roof Support. National Institute for Occupational Safety and Health (NIOSH), USA, pp. 111–131 (Pittsburgh). Peng, S., 1998. Roof bolting adds stability to weak strata. Coal Age Mag. 11, 32–38. Peng, S.S., Tang, D.H.Y., 1984. Roof bolting in underground mining: a state-of-the-art review. Int. J. Min. Eng. 2, 1–42. http://dx.doi.org/10.1007/BF00880855. Stankus, J., Peng, S., 1996. A new concept for roof support. Coal Age Mag. 9, 2–6. Wang, C., Wang, Y., Lu, S., 2000. Deformational behaviour of roadways in soft rocks in underground coal mines and principles for stability control. Int. J. Rock Mech. Min. Sci. 37, 937–946. http://dx.doi.org/10.1016/S1365-1609(00)00026-5. Yoshinaka, R., Tran, T.V., Osada, M., 1998. Non-linear, stress- and strain-dependent behavior of soft rocks under cyclic triaxial conditions. Int. J. Rock Mech. Min. Sci. 35, 941–955. http://dx.doi.org/10.1016/S0148-9062(98)00158-2. Zhang, N., Gao, M., 2004. High-strength and pretension bolting support of coal roadway and its application. J. China Univ. Min. Technol. 33, 524–527.