sensors Article
Study on the Optimal Groove Shape and Glue Material for Fiber Bragg Grating Measuring Bolts Yiming Zhao 1,2 , Nong Zhang 1, * 1
2 3
*
ID
, Guangyao Si 1,3, * and Xuehua Li 1
Key Laboratory of Deep Coal Resource Mining of the Ministry of Education, School of Mines, China University of Mining and Technology, Xuzhou 221116, China;
[email protected] (Y.Z.);
[email protected] (X.L.) Shenzhen Oceanpower Co. Ltd., Shenzhen 518040, China School of Minerals and Energy Resources Engineering, University of New South Wales, Sydney, NSW 2052, Australia Correspondence:
[email protected] (N.Z.);
[email protected] (G.S.); Tel.: +86-136-4153-3550 (N.Z.)
Received: 29 March 2018; Accepted: 29 May 2018; Published: 2 June 2018
Abstract: Fiber Bragg grating (FBG) measuring bolts, as a useful tool to evaluate the behaviors of steel bolts in underground engineering, can be manufactured by gluing the FBG sensors inside the grooves, which are usually symmetrical cuts along the steel bolt rod. The selection of the cut shape and the glue types could perceivably affect the final supporting strength of the bolts. Unfortunately, the impact of cut shape and glue type on bolting strength is not yet clear. In this study, based on direct tension tests, full tensile load–displacement curves of rock bolts with different groove shapes were obtained and analyzed. The effects of groove shape on the bolt strength were discussed, and the stress redistribution in the cross-section of a rock bolt with different grooves was simulated using ANSYS. The results indicated that the trapezoidal groove is best for manufacturing the FBG bolt due to its low reduction of supporting strength. Four types of glues commonly used for the FBG sensors were assessed by conducting tensile tests on the mechanical testing and simulation system and the static and dynamic optical interrogators system. Using linear regression analysis, the relationship between the reflected wavelength of FBG sensors and tensile load was obtained. Practical recommendations for glue selection in engineering practice are also provided. Keywords: underground engineering; groove shape; glue material; FBG measuring bolts
1. Introduction Bolt supporting technology has been widely used in a variety of geotechnical engineering applications [1–4]. The axial stress distribution and force transfer mechanism of bolts have been studied by many researchers. To obtain the axial stress distribution and the force transfer mechanism of the bolt, various measurement techniques have been developed, such as strain gauges [5–7], vibrating-wire sensors [8], dynamometer [9], and non-destructive testing technology [10–12]. However, these techniques can only provide information about the short-term loading process due to environmental disturbances, corrosion, and electromagnetic interferences [13]. Given its unique high accuracy, excellent electromagnetic immunity, durability, and distributed measurement characteristics, the fiber Bragg grating (FBG) sensor has been widely applied as a perceived element and transmission medium to study the behavior of rock bolts or anchors by many studies in the field of underground engineering [14]. Each type of FBG sensor has the ability of monitor a specific reflected wavelength induced by strain deformation to calculate the load stress in the rock bolt or anchor. Schmidt-Hattenberge et al. [15] manufactured a glass fiber-reinforced polymer rock bolt for geotechnical strain measurements, whereby a Bragg grating was attached to the Sensors 2018, 18, 1799; doi:10.3390/s18061799
www.mdpi.com/journal/sensors
Sensors 2018, 18, 1799
2 of 12
one mm wide and two mm deep groove of rock bolts by a special two-component epoxy. Lin et al. [16] bonded FBG in the 5 cm long, 0.3 cm wide, and 0.1 cm deep slot of the anchor end and on the surface of a single wire of the cable using “102” glue to obtain the relationship between the FBG wavelength shifts and stress. Chai el al. [17] glued three FBG sensors in the 1 mm deep and 30 mm long groove of a bolt, and used them to record axial stresses and strains along the bolt during a pull-out test. Weng el al. [18] adhered the FBG sensor to the surface of a tie-solder rod with “502” glue and wrapped with four layers of thermal plastic pipe, then monitored the seven strain conditions of the rock bolts during construction. Zhu et al. [19] captured the variations in internal displacement profiles in a model dam using sensing bars, in which FBG sensors were adhered in its grooves and covered with epoxy resin. Schroeck et al. [20] proposed a unique arrangement of FBG sensors to allow up to 20% strain measurements of steel rock bolts. Huang et al. [21] designed a novel distributed self-sensing fiber reinforced polymer (FRP) anchor rod with a built-in optical fiber sensor to predict the mechanical behavior of the anchor rod. Building upon FBG sensors, De Waele et al. [22] developed a load cell to measure the forces in the ground anchors of a quay wall, and concluded that fiber-based instrumentation is highly suitable for long-term monitoring purposes. Kim et al. [23] and Do et al. [24] encapsulated FBG sensors into the central king cable of a 7-wire steel strand to monitor the changes in tensile force and its distribution along the tendons during in-service state. They stated the sensors were embedded into the steel tube by a liquid glue with low viscosity. Xu et al. [25] measured the strain in 100 cm intervals along an anchor rock bolt grouted in the slope of intact rock using FBG strain sensors. Chen et al. [26] proposed five different FBG strain sensors and highlighted the prospective use of FBG in the strain monitoring of cables. For a steel bolt, one common approach for installing the FBG sensors is to glue them inside the symmetrical grooves along the bolt rod, and then refill the groove using silica gel to protect the FBG sensors. Then, the FBG measuring bolt is manufactured for field applications to obtain the stress distribution and its variation along the bolt. However, the depth and shape of grooves cut within the bolt undoubtedly reduce the strength of the bolt itself, which may have a direct impact on the long-term monitoring accuracy of FBG sensors, especially given the complexity of the underground environment. In addition, the application of various types of glue material may also affect the monitoring accuracy of the FBG measuring bolt. Unfortunately, in the abovementioned literature, all factors interfering with the FBG monitoring sensors have not yet been properly investigated. In this paper, the effects of groove shape on bolt strength were analyzed based on tensile experiments on conventional rock bolts. Following that, the stress distribution in an anchor bolt with different groove shapes was numerically analyzed by ANSYS, a continuum mechanics simulator. Furthermore, four commonly used glues and their effects on FGB monitoring sensors were assessed. Based on these results, the optimal groove shape and glue material are proposed to improve the monitoring performance of FGB bolts in underground engineering. Note that although the main engineering background in this research aims at understanding the impact of groove shape and glue material on FBG measuring bolts, the findings are also transferable to other strain gauge-based measuring bolts that also require cutting grooves and glue material. 2. Tensile Tests on Bolts with Different Groove Shapes 2.1. Tensile Test Conditions The tensile tests were performed by using the mechanical testing and simulation (MTS) electrohydraulic servo testing machines, as shown in Figure 1. The basic parameters for the MTS machine were as follows: (1) peak load 1000 kN; (2) maximum tensile space 790 mm; (3) and displacement rate 0.5–90 mm/min.
Sensors 2018, 18, 1799
3 of 12
Sensors 2018, 3, x FOR PEER REVIEW Sensors 2018, 3, x FOR PEER REVIEW
3 of 12 3 of 12
Figure 1. Mechanical testing and simulation (MTS) electrohydraulic servoservo testing machine. Figure 1. Mechanical testing and simulation (MTS) electrohydraulic testing machine. Figure 1. Mechanical testing and simulation (MTS) electrohydraulic servo testing machine.
All specimens were cut from steel bolts that were used in underground coal mines. The length All specimens specimens were were cut cut from steel bolts were in underground coal The length from bolts that that was were22used used underground coal mines. mines. Theinto length of theAll specimens was 250 mm, andsteel the diameter mm.inThe specimens were classified six of the specimens was 250 mm, and the diameter was 22 mm. The specimens were classified into six of the specimens wasgroove 250 mm, and To therule diameter was 22effect mm. The specimens were the classified into six categories based on shape. out the size of various grooves, six specimens categories based on groove shape. To rule out the size effect of various grooves, the six specimens categories based on groovethe shape. rule out thefor size effect of various grooves, the six specimens aimed to achieve roughly sameTo cutting area different shapes with a laser-cutting machine. aimed to to achieve achieve roughly roughly the the same same cutting cutting area area for for different different shapes shapes with with aa laser-cutting laser-cutting machine. machine. aimed Two independent specimens for each type of groove shape were manufactured and numbered to Two independent specimens for each type of groove shape were manufactured and numbered to Two independent specimens for each type ofgroove groovesizes shape manufactured and numbered to reduce sampling bias. The groove shapes and forwere the specimens are shown in Figure 2. reduce sampling bias. The groove shapes and groove sizes for the specimens are shown in Figure 2. reduce sampling bias. The groove shapes and groove sizes for the specimens are shown in Figure 2.
(a) (a)
(b) (b)
(c) (c)
(d) (d)
(e) (e)
(f) (f)
Figure 2. Comparisons between different groove shapes and sizes (mm): (a) no-groove; (b) U-groove; Figure 2. between different groove and sizes (mm): no-groove; (c) U-groove with chamfer; (d) inverted groove; trapezoidal and(b) (f) U-groove; V-groove. Figure 2. Comparisons Comparisons between differenttrapezoidal groove shapes shapes and(e) sizes (mm): (a) (a)groove; no-groove; (b) U-groove; (c) U-groove with chamfer; (d) inverted trapezoidal groove; (e) trapezoidal groove; and (f) V-groove. (c) U-groove with chamfer; (d) inverted trapezoidal groove; (e) trapezoidal groove; and (f) V-groove.
2.2. Test Procedure and Results Analysis 2.2. Test Procedure and Results Analysis 2.2. Test andprocess Resultsused Analysis TheProcedure tensile test closed-loop control positioning. The tensile tests for the specimens The tensile test process used closed-loop control positioning. The tensile tests for the wereThe initiated at a constant loading rate of 5.1 mm/min under normal temperature andspecimens pressure tensile test process used closed-loop control positioning. The tensile tests for the specimens were initiated atstress–strain a constant loading rate of 5.1specimens mm/min were underrecorded normal temperature and pressure conditions. The responses of the during the tensile loading were initiated at a constant loading rate of 5.1 mm/min under normal temperature and pressure conditions. The stress–strain responses of the specimens were recorded during the tensile loading process, andThe the stress–strain frequency of responses data collection 1.0 Hz. Note the actual loading lengthloading of 110 conditions. of thewas specimens were that recorded during the tensile process, and the frequency of data collection was 1.0 two Hz. ends Note held that the actual loading length of 110 mm was determined by deducting the length of the by the machine. The specimens process, and the frequency of data collection was 1.0 Hz. Note that the actual loading length of 110 mm mm was determined by deducting the into length ofpieces. the two ends held of bythe thetensile machine. The weredetermined loaded untilbythe specimen broke The images tests arespecimens shown in was deducting the length of two the two ends held by the machine. The specimens were were loaded until the specimen broke into two pieces. The images of the tensile tests are shown in Figure loaded 3. until the specimen broke into two pieces. The images of the tensile tests are shown in Figure 3. Figure 3.
Sensors 2018, 18, 1799
4 of 12
Sensors 2018, 3, x FOR PEER REVIEW
4 of 12
Sensors 2018, 3, x FOR PEER REVIEW
4 of 12
(a)
(b)
(a) (b) Figure test procedure procedure for for the the specimen: specimen: (a) (a) before before loading loading and and (b) (b) after after loading. loading. Figure 3. 3. Tensile Tensile test Figure 3. Tensile test procedure for the specimen: (a) before loading and (b) after loading.
The tension strength and displacement relationships for all 12 tested specimens are shown in The tension strength and displacement relationships for all 12 tested specimens are shown in Figure 4. The tension strength and displacement relationships for all 12 tested specimens are shown in Figure Figure 4. 4.
Figure 4. Relationship curves of tensile load–displacement for specimens: (a) no-groove; (b) U-groove;
Figure 4. Relationship curves of tensile load–displacement for specimens: (a) no-groove; (b) U-groove; (c) U-groove with chamfer; (d) trapezoidal groove; (e) inverted trapezoidal groove; and (f) V-groove. (c) U-groove with chamfer; (d) trapezoidal groove; (e) inverted trapezoidal groove; and (f) V-groove.
Figure 4. Relationship curves of tensile load–displacement for specimens: (a) no-groove; (b) U-groove; (c) U-groove with chamfer; (d) trapezoidal groove; (e) inverted trapezoidal groove; and (f) V-groove.
Sensors 2018, 18, 1799
5 of 12
The yield strength, ultimate strength, and failure strength for these specimens with different groove shapes are summarized in Table 1. We concluded that the creation of grooves can somewhat reduce the strength of specimens, and the effect of strength reduction varies with groove shape. Table 1. Strength of specimens with different groove sharps. Groove Shape Specimen No. Yield strength (kN) Average yield strength (kN) Reduction (%) Peak strength (kN) Average peak strength (kN) Reduction (%) Failure strength (kN) Average failure strength (kN) Reduction (%)
No Groove 1–1 140
1–2 140
U-Groove 2–1 131
140 215
212
199
206
156
3–2 134
Inverted Trapezoidal Groove 4–1 134
134.5 3.93 205
202.5 5.15 166
166.5 -
3–1 135
133 5.00
213.5 172
2–2 135
U-Groove with Chamfer
151
204
162
203
204
166
6–1 132
205
204
164
204
200 202 5.39
165 165.5 0.60
6–2 128 130 7.14
204.5 4.22 165
162.5 0.60
5–2 136
V-Groove
136 2.85
203.5 4.68 162
162 2.70
5–1 136
133 5.00
204.5 4.22
153.5 7.81
4–2 132
Trapezoidal Groove
164
158 161 3.30
Figure 4 suggests that all 12 specimens followed almost the same trend in the tensile loading test. A linear elastic stage occurred between the tensile load and displacement during the initial stage of the tensile test. The specimens then reached the yield stage where the average yield strength of the no-groove specimens was 140 kN. In comparison with the no-groove type, the average yield strength for the specimens with the trapezoidal groove shape was 136 kN, with a strength reduction of 2.85%. Furthermore, the percentage of yield strength reduction for the V-groove specimen was the largest at 7.14%. After the tensile load passed yield strength, the tensile load continuously increased, which resulted in the realignment of the internal grain structure of the specimens. This widely observed strain hardening behavior reflected the decrease in the specimen’s ability to resist further deformation despite the increasing tensile strength. After a period of yielding, the tensile load reached the peak strength and the internal structure of the specimen broke at around 80% of peak strength, which terminated the experiment. The average peak strength of the no-groove specimen was 213.5 kN. In comparison, the average peak strength dropped by 4.22% for both the U-groove with a chamfer and the trapezoidal groove, which was the lowest among all specimens with grooves. More than a 5% reduction in peak strength was observed in U-groove. By the time tensile load exceeded the peak strength, the lateral size of the specimen shrank dramatically. This necking phenomenon occurred shortly after the specimen broke. The breaking tensile load of the no-groove specimen was 166.5 kN, whereas it was nearly 8% lower for U-groove specimen. The experimental results suggest that bolts with a trapezoidal groove have the least strength deterioration compared with the other groove shapes, which should be recommended for the manufacturing of FBG-based monitoring bolts. 3. Numerical Analysis of Stress Distribution of Bonded Bolts with Different Groove Shapes 3.1. Model Geometry and Boundary Condition To understand the stress distribution in bonded bolts with different groove shapes under tensile loading conditions, the following numerical models were developed in ANSYS. Note that numerical models developed in this section aimed to reproduce fully loaded bolts at actual working conditions, which are bonded to the surrounding rock, rather than a free state boundary, as shown in the tensile tests above. The bolt model was developed using 7400 SOLID, 95 elements, and 32,611 nodes. The material parameters used for this study were: model length 110 mm, diameter 22 mm, Young’s modulus for the bolting steel 20 GPa, and Poisson’s ratio 0.3. The boundary conditions of the bonded bolt were as follows: the x and z displacement were restrained; a fully fixed bottom surface; and the groove and top surfaces were free. The external stress applied to the model top surface
Sensors 2018, 18, 1799
6 of 12
was 119 kN. Model geometries (as shown in Figure 5) for various groove shapes and dimensions Sensors 2018, 3, x FOR PEER REVIEW 6 of 12 exactly followed the experimental designs as introduced in Section 2.1.
(a)
(b)
(c)
(d)
(e)
(f)
Figure 5. Model geometries for various groove shapes: (a) no-groove; (b) U-groove; (c) U-groove with Figure 5. Model geometries for various groove shapes: (a) no-groove; (b) U-groove; (c) U-groove with chamfer; (d) inverted trapezoidal groove; (e) trapezoidal groove; and (f) V-groove. chamfer; (d) inverted trapezoidal groove; (e) trapezoidal groove; and (f) V-groove.
3.2. Simulation Results 3.2. Simulation Results The stress contours of the middle cross-section for different groove shapes are shown in Figure The stress contours of the middle cross-section for different groove shapes are shown in Figure 6. 6. After introducing grooves, stress concentration occurred surrounding each groove. For the VAfter introducing grooves, stress concentration occurred surrounding each groove. For the V-shaped shaped grooves, which had the largest stress disturbance area, the magnitude of the elevated stress grooves, which had the largest stress disturbance area, the magnitude of the elevated stress can be as can be as twice as much as the stress in the undisturbed area. The majority of elevated stress in the twice as much as the stress in the undisturbed area. The majority of elevated stress in the disturbed disturbed zone ranged from 1 to 1.3 times the undisturbed stress. Notably, the stressed measured by zone ranged from 1 to 1.3 times the undisturbed stress. Notably, the stressed measured by FBG (or any FBG (or any other type strain-gauge) is not the true stress state in a no-groove bolt, but an elevated other type strain-gauge) is not the true stress state in a no-groove bolt, but an elevated stress state stress state induced by the specific shape of the grooves. Due to this artificial effect, compared with induced by the specific shape of the grooves. Due to this artificial effect, compared with no-groove no-groove bolts, the measured stress in grooved bolts can be amplified by a factor of 1.3, as shown in bolts, the measured stress in grooved bolts can be amplified by a factor of 1.3, as shown in Figure 6. Figure 6. This needs to be taken into consideration in engineering practice. This needs to be taken into consideration in engineering practice.
Sensors 2018, 18, 1799 Sensors 2018, 3, x FOR PEER REVIEW
7 of 12 7 of 12
Figure Figure6.6.Stress Stressdistribution distributionininthe thecentral centralcross-section cross-sectionfor forbonded bondedbolts boltswith withdifferent differentgroove grooveshapes. shapes.
Themaximum maximum stress values the central cross-section forgroove different groove shapes are The stress values in thein central cross-section for different shapes are summarized summarized in Table 2. The maximum stress value for the no-groove bolt was 179 MPa. Under the in Table 2. The maximum stress value for the no-groove bolt was 179 MPa. Under the same tensile load, same tensile stress load, the maximum stress value for 378 the MPa V-groove bolt wasconcentration 378 MPa andfactor the stress the maximum value for the V-groove bolt was and the stress was concentration factor was 2.11 by comparing with the no-groove case, which is the highest value 2.11 by comparing with the no-groove case, which is the highest value compared with the other types with the other types of grooved bolts. stress Correspondingly, stress valuegroove for the ofcompared grooved bolts. Correspondingly, the maximum value for thethe boltmaximum with the trapezoidal bolt335 with theand trapezoidal was 335factor MPa and stress concentration factor wasall 1.87, which is was MPa the stressgroove concentration was the 1.87, which is the smallest among scenarios. the smallest among all scenarios. Table 2. Maximum stress values in the central cross-section of a bonded bolt. Table 2. Maximum stress values in the central cross-section of a bonded bolt. Groove Type Maximum Stress Value (MPa) Stress Concentration Factor
Groove Type Maximum Stress Value (MPa) Stress Concentration Factor No-groove 179 No-groove 179 U-groove 359 2.01 U-groove U-groove with chamfer 348359 1.942.01 U-groove with chamfer Inverted trapezoidal groove 365348 2.041.94 Trapezoidal groovegroove 335365 1.872.04 Inverted trapezoidal V-groove 378 2.11 Trapezoidal groove 335 1.87 V-groove 378 2.11 The modelling results are consistent with the experiment findings. The creation of a trapezoidal grooveThe hasmodelling the least impact bolt strength, since the stress findings. concentration at the corner points are resultson arethe consistent with the experiment The creation of a trapezoidal the lowest. the V-groove, which may induce twice as much stress concentration, groove hasConversely, the least impact on the bolt strength, since the stress concentration at the cornershould points be are avoided in engineering the lowest. Conversely,practice. the V-groove, which may induce twice as much stress concentration, should be avoided in engineering practice. 4. Glue Material Recommendation 4. Glue Material Recommendation 4.1. Glue Materials 4.1. FBG Glue sensors Materialsare normally glued inside the symmetrical grooves along the bolt axial direction duringFBG the sensors manufacturing of FBGglued measuring The glue material can along not only FBGdirection sensors are normally insidebolts. the symmetrical grooves theensure bolt axial remain the cuts, but of also bemeasuring able to deform conjunction with the thesensors bolts. duringinside the manufacturing FBG bolts.in The glue material can notdeformation only ensure of FBG Four typical glues were used to attach FBG sensors to specimens in this study. Table 3 summarizes remain inside the cuts, but also be able to deform in conjunction with the deformation of the bolts. general information about the to four typeFBG of glues we to used in this research. NoteTable that all specimens Four typical glues were used attach sensors specimens in this study. 3 summarizes studied this sectionabout were applied groove, since it had the least on bolt generalininformation the fourwith typethe of trapezoidal glues we used in this research. Note thatimpact all specimens strength, as found in previous sections. studied in this section were applied with the trapezoidal groove, since it had the least impact on bolt strength, as found in previous sections.
Sensors 2018, 18, 1799
8 of 12
Table 3. General information of four type of glues used in this research. Sensors 2018, 3, x FOR PEER REVIEW No. 001 No. 002 001
003 002 003
004 004
Groove Shape
Table Trapezoidal groove3.
8 of 12 Glue Name
Main Component
Methacrylate and mix of some toughening agent, reinforcing agent, General Modifiedinformation acrylate glue of four type of glues used in this research. and stabilizer
Ethyl α-cyanoacrylate and mix of some toughening agent, reinforcing Glue Name Main Component 502 glue agent, and Trapezoidal groove Modified acrylate glue Methacrylate and mixstabilizer of some toughening agent, reinforcing agent, and stabilizer Trapezoidal groove Epoxy resin A-B glue Two components based ontoughening epoxy resin with high temperature resistance Trapezoidal groove 502 glue Ethyl α-cyanoacrylate and mix of some agent, reinforcing agent, and stabilizer Epoxy resin, agent, methyl Trapezoidal groove Epoxy resin A-B glue Two components basedcuring on epoxy resin with hightetrahydrophthalic temperature resistanceanhydride, Trapezoidal groove H814-Epoxy resin glue aluminum hydroxide, and polyester poly Epoxy resin, curing agent, methyl tetrahydrophthalic anhydride, aluminum hydroxide, and Trapezoidal groove H814-Epoxy resin glue polyester poly Groove Shape Trapezoidal groove
Since sensors are are made made of of glass glass material, material, to to ensure ensure their their integrity, integrity, the gluing process process should should Since FBG FBG sensors the gluing follow follow the the proposed proposed manufacturing manufacturing steps steps during during the the preparation preparation of of the the specimens. specimens. First, First, sandpaper sandpaper was used to grind each groove until its bottom surface was sufficiently smooth. Second, acetone was used to grind each groove until its bottom surface was sufficiently smooth. Second, acetone and and alcohol the glue glue was was spread spread alcohol were were used used to to scrub scrub the the groove groove to to remove remove dust dust and and impurities. impurities. After After that, that, the into into the the grooves grooves while while maintaining maintaining the the bare bare FBG FBG sensors sensors in in aa straight-line straight-line orientation, orientation, which which was was followed by placing the specimen in a fixture to allow the glue to solidify. The required curing followed by placing the specimen in a fixture to allow the glue to solidify. The required curing time time ◦ C was 2 h, for 60 ◦ C was 1 h, for 80 ◦ C was 30 min, and for 100 ◦ C was 5 min. The sample at 25 °C at 25 was 2 h, for 60 °C was 1 h, for 80 °C was 30 min, and for 100 °C was 5 min. The sample was ◦ was then cooled down to room temperature (about Finally,a aprotection protectionplastic plasticthin thin tube tube was then cooled down to room temperature (about 19 19 °C).C). Finally, was placed on the optical fiber except the optical grating part, and the grooves were filled with 704 placed on the optical fiber except the optical grating part, and the grooves were filled with 704 silica silica to The 704 to encapsulate encapsulate and and protect protect FBG FBG sensors. sensors. The 704 silica silica is is aa single-component single-component adhesive adhesive material material with with good adhesion, high strength, and no corrosion, which has excellent electrical insulation, sealing good adhesion, high strength, and no corrosion, which has excellent electrical insulation, sealing performance, characteristics performance, moisture moisture resistance, resistance, and and anti-seismic anti-seismic and and aging aging resistance. resistance. Note Note that that the the characteristics of FBG length 0.3 nm, nm, of FBGs FBGs were were as as follows: follows: FBG length 10 10 mm, mm, fiber fiber type type was was SMF-28C SMF-28C fiber, fiber, FWHM FWHM BW BW ≤ ≤ 0.3 reflectivity ≥ 85%, acrylate recoat, and FC/APC connector. The completely packaged specimens with reflectivity ≥ 85%, acrylate recoat, and FC/APC connector. The completely packaged specimens with jumper jumper connections connections are are shown shown in in Figure Figure 7. 7.
(a)
(b)
Figure 7. Photos of packaged specimens: (a) specimens using different glues and (b) specimen with Figure 7. Photos of packaged specimens: (a) specimens using different glues and (b) specimen with jumper connection. connection. jumper
4.2. Tensile Test and Data Analysis 4.2. Tensile Test and Data Analysis The tensile test of the packaged specimens was performed using an MTS system and static and The tensile test of the packaged specimens was performed using an MTS system and static and dynamic optical interrogators (SDOI) system. The SDOI system is composed of the static and dynamic optical interrogators (SDOI) system. The SDOI system is composed of the static and dynamic dynamic optical interrogators, an industrial personal computer, and corresponding software, which optical interrogators, an industrial personal computer, and corresponding software, which can be can be used to record the specific reflected wavelength of FBG sensors embedded in a bolt with a used to record the specific reflected wavelength of FBG sensors embedded in a bolt with a trapezoidal trapezoidal groove. The general information about the interrogator is shown in Table 4. An MTS groove. The general information about the interrogator is shown in Table 4. An MTS system is used to system is used to record the tensile loading exerting on the packaged specimens. During the tensile record the tensile loading exerting on the packaged specimens. During the tensile loading process, the loading process, the reflected wavelength of the FBG sensors changes accordingly. Then, the tensile reflected wavelength of the FBG sensors changes accordingly. Then, the tensile load and the reflected load and the reflected wavelength of the FBG sensor were recorded during the tensile process, with wavelength of the FBG sensor were recorded during the tensile process, with a 1.0 Hz frequency of a 1.0 Hz frequency of data collection. The test system is shown in Figure 8. data collection. The test system is shown in Figure 8.
S ensors2018, 2018,18, 3, x1799 FOR PEER REVIEW Sensors
99 of of 12 12
Table 4. General specifications of the interrogator used in this study. Table 4. General specifications of the interrogator used in this study.
Part Parameter OpticalPart channels 4 individual channels Parameter Wavelength range 1525–1565 nmchannels Optical channels 4 individual Minimum wavelength spacing 0.4 nm Wavelength range 1525–1565 nm Minimum wavelength spacing 0.4 nm Wavelength precision ±1 pm Sensors 2018,Wavelength 3, x FOR PEER REVIEW 9 of 12 precision ± Scan and report rate 500 Hz1 pm Scan and report rate 500 Hz OpticalTable connector Ferrule Connector/A-Physical 4. General specifications of the interrogator used in this study. Connection Optical connector Ferrule Connector/A-Physical Connection Power 220V220V AC 1A Powersupply supply AC 1A Part Parameter Communication USB 2.0 2.0 Communication USB Optical channels 4 individual channels ◦ Operating temperature 0–40 Wavelength range 1525–1565 Operating temperature 0–40nm °C C Dimensions 95 mm Minimum wavelength spacing nm××400 Dimensions 4800.4 ×480 400 95 × mm Wavelength precision Scan and report rate Optical connector Power supply Communication Operating temperature Dimensions
±1 pm 500 Hz Ferrule Connector/A-Physical Connection 220V AC 1A USB 2.0 0–40 °C 480 × 400 × 95 mm
Figure 8. The mechanical testing and simulation (MTS) system and static and dynamic optical Figure 8. The mechanical testing and simulation (MTS) system and static and dynamic optical interrogators (SDOI) system system used used for for the the ensile ensile loading loading experiment. experiment. interrogators (SDOI)
The relationship curves between the tensile loads and the reflected wavelength of the FBG Figure 8. The mechanical testing and (MTS) system and static and dynamic optical The relationship curves between the simulation tensile loads and the reflected wavelength of the FBG sensors with interrogators different (SDOI) gluessystem are used shown Figure Peak wavelength was calculated from the for thein ensile loading9.experiment. sensors with different glues are shown in Figure 9. Peak wavelength was calculated from the reflection reflection spectrum using a built-in software package in the interrogator purchased for this project. spectrum using a built-in software packagethe in tensile the interrogator purchased for this project. The software curvessearch between loads and the reflected of the spectrum FBG The software The canrelationship automatically for the peak wavelength oncewavelength the reflection was sensors with different glues arewavelength shown in Figure Peak wavelength was calculated from the Before the can automatically search for the peak once9.the reflection spectrum was obtained. obtained. Before the bolts reached the yield strength, the reflected wavelength of the FBG sensors for reflection spectrum using a built-in software package in the interrogator purchased for this project. bolts reached the yield strength, the reflected wavelength of the FBG sensors for all specimens exhibited all specimens exhibited reliable linear relationship the increase tensile load. However, when The software canaautomatically search for the peakwith wavelength once theinreflection spectrum was a reliable linear relationship with the increase in tensile load. However, when the tensile load exceeded Before theyield bolts reached the the yieldmonitoring strength, the reflected wavelength of the FBG sensors for the tensileobtained. load exceeded strength, of the reflected wavelength of specimens No. yield strength, the monitoring of thelinear reflected wavelength of specimens No.However, 002 and No. 003 was all specimens exhibited a reliable relationship with the increase in tensile load. when 002 and No. 003 was interpreted and started recording zero. This indicates that the FBG sensors, theand tensile load exceeded yieldzero. strength, theindicates monitoringthat of thethe reflected of specimens interpreted started recording This FBGwavelength sensors, attached byNo. the 502 glue attached by the 502 glue and the epoxy resin AB, might have detached from the bolt or were partially 002 and No. 003 was interpreted and started recording zero. This indicates that the FBG sensors, and the epoxy resin AB, might have detached from the bolt or were partially damaged. Although damaged.attached Although the wavelength of specimens No. 001 and were still recording by the 502reflected glue and the epoxy resin AB, might have detached from theNo. bolt 004 or were partially the reflected wavelength of specimens No. 001 and No. 004 were still recording positive damaged. the reflected wavelength specimens No.and 001 and No.relationship 004 were still recording positive data, theyAlthough both experienced dramaticoffluctuation, their with tensile data, load they both experienced dramatic fluctuation, and their relationship with tensile load was no longer linear, positive data, they both experienced dramatic fluctuation, and their relationship with tensile load was no longer linear, which means the FBGs either broadened or split. The reflected wavelengths for was no longer linear, which means the or FBGs either broadened or split. The reflectedfor wavelengths forspecimens which means the FBGs either broadened split. The reflected wavelengths these two these two these specimens fell to zero when the tensile load exceeded the peak strength, which suggested two specimens fell to zero when the tensile load exceeded the peak strength, which suggested fell to zero when the tensile load exceeded the peak strength, which suggested the deterioration of the the deterioration of the last specimens. based onresult, this result, the evolution of the deterioration of thetwo last two specimens.Correspondingly, Correspondingly, based on this the evolution of last two specimens. Correspondingly, based on this result, the evolution of the reflected wavelength the reflected wavelength canused be used examine the the damage to the bolt [27,28] if the perfect the reflected wavelength can be to to examine damage to the bolt [27,28] if thegluing perfect gluing can be used to examine damage to the bolt [27,28] if the perfect gluing quality of FBGs is achieved. of FBGs isthe achieved. quality of quality FBGs is achieved. 1555
1555
001 002 003 004
1550
Wavelength /nm
Wavelength /nm
1550 1545 1545 1540 1540
001 002 003 004
1535 1530
1535 1525
1530
0
25
50
75
100
125
150
175
200
Tensile stress /kN
1525 Figure 9.0 Wavelength variation during tensile loading. 25 50 75 100 125 150 175 200
Tensile stress /kN
Sensors 2018, 3, x FOR PEER REVIEW
10 of 12
Sensors 2018, 18, 1799
10 of 12
Figure 9. Wavelength variation during tensile loading.
From the theabove abovemeasurement measurement results, when the tensile load bolt exceeded its yield From results, when the tensile load on theon boltthe exceeded its yield strength, strength, the expected bolt was expected to experience a significant amount tensile displacement. Since FBG the bolt was to experience a significant amount of tensileofdisplacement. Since FBG sensors sensors have relativity limited ability to resist tension deformation, this non-compatible deformation have relativity limited ability to resist tension deformation, this non-compatible deformation that that occurred between theand boltthe and thesensor FBG sensor may to the malfunction these monitoring occurred between the bolt FBG may lead tolead the malfunction of theseof monitoring sensors. sensors. Therefore, FBG sensors can provide accurate the recorded is within Therefore, FBG sensors can provide accurate resultsresults only ifonly the if recorded load load rangerange is within the the yielding strength thebeing bolt being monitored. However, the malfunction FBG sensors can be yielding strength of theofbolt monitored. However, the malfunction of FBGof sensors can be seen as seen as an indicator bolts haveinternal serious damage, internal damage, which mayinbeengineering used in engineering to an indicator that boltsthat have serious which may be used to diagnose diagnose bolt integrity. bolt integrity. The accuracy of using FBG-based sensors to measure tensile load on bolts is highly reliant on the reflected reflected wavelength wavelength to tensile tensile deformation, deformation, which was only observed in the linear response of the the elastic stage before reaching the yielding point. Therefore, for the purpose of bolt tensile tensile load monitoring, a sound and robust linear relationship between tensile load and reflected wavelength is critically important in FBG-based monitoring of bolts. The elastic range of the four tested specimens were linearly linearly fitted fitted using using regression regressionmethod methodas asshown shownin inFigure Figure10. 10. during their tensile tests were
(a) No. 001 specimen
(b) No. 002 specimen
(c) No. 003 specimen
(d) No. 004 specimen
Figure 10. 10. Linear load in in specimens specimens with with Figure Linear relationship relationship between between reflected reflected wavelength wavelength and and tensile tensile load different glues. different glues.
Although a relatively high correlation (R2 > 9.87) was obtained for all four specimens, the Although a relatively high correlation (R2 > 9.87) was obtained for all four specimens, the selection selection of glue material had various impacts on the wavelength–tensile load relationship. The of glue material had various impacts on the wavelength–tensile load relationship. The specimens specimens No. 001–No. 003 all presented a constant and reliable linear relationship with a correlation No. 001–No. 003 all presented a constant and reliable linear relationship with a correlation coefficient coefficient over 0.995. Conversely, the linear curve of specimen No. 004, which used H814-epoxy resin over 0.995. Conversely, the linear curve of specimen No. 004, which used H814-epoxy resin glue, glue, fluctuated over its elastic period, showing the lowest correlation. In addition, the linear slope fluctuated over its elastic period, showing the lowest correlation. In addition, the linear slope of of the reflected wavelength using this glue was the smallest among all samples, which indicates the the reflected wavelength using this glue was the smallest among all samples, which indicates the measured tensile load would be much more sensitive to wavelength change compared with the other measured tensile load would be much more sensitive to wavelength change compared with the other three. A tiny wavelength change may result in a large variance in measured tensile load. Therefore, three. A tiny wavelength change may result in a large variance in measured tensile load. Therefore, in engineering practice, we recommend selecting glues with robust linearity and proper sensitivity in engineering practice, we recommend selecting glues with robust linearity and proper sensitivity response to the reflected wavelength, for instance, glue types No. 001 and No. 002 in this research. response to the reflected wavelength, for instance, glue types No. 001 and No. 002 in this research. The effect of different glues on FBG bolts is mainly determined by glue components and the proportion of each component. For the modified acrylate glue, its main components are methacrylate
Sensors 2018, 18, 1799
11 of 12
The effect of different glues on FBG bolts is mainly determined by glue components and the proportion of each component. For the modified acrylate glue, its main components are methacrylate with a mix of toughening agents, reinforcing agents, and stabilizers. Compared with others, this type of glue can more firmly attach FBG sensors to bolts, which satisfies the harmonic deformation between FBG sensors and bolts before reaching yield strength. Note that due to the limited amount of samples being tested in this research, the variability of sample performance was not fully considered. Although not directly related with this research topic, the four glues tested here were also used in other research projects involving FBG bolts led by the authors and all glues showed a relative consistent performance. Nevertheless, repeatability of the gluing procedure may have an influence on the testing results, and this will be investigated in the future work. 5. Conclusions This research focused on investigating the effect of various groove shapes on the reduction of bolting strength when using FBG measuring bolts. The performance of five different groove types were assessed in both tensile load tests and numerical simulations. Both experimental and numerical results suggested that the trapezoidal groove has the least impact on bolting strength, with only a 0.37% reduction in tensile yield strength. Another interesting finding from the numerical modelling results is the FBG (or any other type strain-gauge) measured stress is not the true stress state in a no-groove bolt, but an elevated stress state induced by the specific shape of grooves. Four different types of glues normally used in practice were also tested to further understand their role in affecting FBG measuring bolts, according to the test results, we recommend that, among four tested glues, the modified acrylate glue is the optimal glue material for manufacturing the FBG measuring bolt. Author Contributions: All the authors contributed equally to this work. Acknowledgments: This study was supported by the China Postdoctoral Science Foundation funded project (Grant No. 2016M601916); the National key research and development program of China (Grant No. 2016YFC0600904); the National Natural Science Foundation of China (Grant No. 51404256) and PAPD. The authors would also like to thank the Open Fund Project of State Key Laboratory of Coal Resources and Safe Mining (Grant No. SKLCRSM16KF02). Conflicts of Interest: The authors declare no conflict of interest.
References 1. 2.
3. 4. 5.
6. 7. 8.
Cai, Y.; Esaki, T.; Jiang, Y.J. An analytical model to predict axial load in grouted rock bolt for soft rock. Tunn. Undergr. Space Technol. 2004, 19, 607–618. [CrossRef] Ghadimi, M.; Shahriar, K.; Jalalifar, H. A new analytical solution for the displacement of fully grouted rock bolt in rock joints and experimental and numerical verifications. Tunn. Undergr. Space Technol. 2015, 50, 143–151. [CrossRef] Kang, H.; Wu, Y.; Gao, F.; Jiang, P.; Cheng, P.; Meng, X.; Li, Z. Mechanical performances and stress states of rock bolts under varying loading conditions. Tunn. Undergr. Space Technol. 2016, 52, 138–146. [CrossRef] Li, C.C.; Kristjansson, G.; Høien, A.H. Critical embedment length and bond strength of fully encapsulated rebar rockbolts. Tunn. Undergr. Space Technol. 2016, 59, 16–23. [CrossRef] Pan, C.; Yamaguchi, T.; Suzuki, Y. Evaluation of the Bolt Axial Force in High Strength Bolted Joints by Using Strain Gauges; Memoirs of the Faculty of Engineering, Osaka City University: Osaka, Japan, 2011; Volume 52, pp. 37–41. Starikova, R.; Schön, J.J. Experimental study on fatigue resistance of composite joints with protruding-head bolts. Compos. Struct. 2002, 55, 1–11. [CrossRef] Mucho, T.P.; Mark, C.; Zelenko, J.C.; Compton, C.S. Roof Support Performance in High Stress Conditions; West Virginia University: Morgantown, WV, USA, 1995. Xue, Y.H.; Chen, Y. Application of Vibrating Wire sensors in Dam Safety. Autom. Water Resour. Hydrol. 2006, 4, 39–42.
Sensors 2018, 18, 1799
9. 10. 11. 12.
13.
14. 15.
16. 17. 18. 19. 20.
21. 22. 23. 24. 25. 26.
27. 28.
12 of 12
Wu, Z.G.; Ju, W.J. Research and application of new dynamometer for anchor or bolt. Coal Sci. Technol. 2007, 11, 36–38. Beard, M.D.; Lowe, M.J.S. Non-destructive testing of rock bolts using guided ultrasonic waves. Int. J. Rock Mech. Min. Sci. 2003, 40, 527–536. [CrossRef] Lee, I.M.; Han, S.I.; Kim, H.; Yu, J.D.; Min, B.K.; Lee, B.K. Evaluation of rock bolt integrity using Fourier and wavelet transforms. Tunn. Undergr. Space Technol. 2012, 28, 304–314. [CrossRef] Si, G.; Durucan, S.; Jamnikar, S.; Lazar, J.; Abraham, K.; Korre, A.; Shi, J.Q.; Zavšek, S.; Mutke, G.; Lurka, A. Seismic monitoring and analysis of excessive gas emissions in heterogeneous coal seams. Int. J. Coal Geol. 2015, 149, 41–54. [CrossRef] Staveley, C. Get smart, go optical: Example uses of optical fibre sensing technology for production optimisation and subsea asset monitoring. In Proceedings of the Fiber Optic Sensors and Applications XI, Baltimore, MD, USA, 18 June 2014; Volume 9098. Zhao, Y.M.; Zhang, N.; Si, G.Y. A Fiber Bragg Grating-Based Monitoring System for Roof Safety Control in Underground Coal Mining. Sensors 2016, 16, 1759. [CrossRef] [PubMed] Schmidt-Hattenberger, C.; Borm, G. Bragg grating extensometer rods (BGX) for geotechnical strain measurements. In Proceedings of the European Workshop on Optical Fibre Sensors, Peebles, UK, 18 June 1998; Volume 3483, pp. 214–217. Lin, C.N.; Liu, Q.S.; Gao, W. Application of fiber optical sensing technology to monitoring axial forces of anchor bolts. Rock Soil Mech. 2008, 11, 3161–3164. Chai, J.; Zhao, W.H.; Li, Y.; Wang, D.C.; Cui, C. Pull out tests of figer Bragg grating sensor fitted bolts. J. China Univ. Min. Technol. 2012, 41, 719–724. Weng, X.L.; Ma, H.H.; Wang, J. Stress Monitoring for Anchor Rods System in Subway Tunnel Using FBG Technology. Adv. Mater. Sci. Eng. 2015, 2015, 1–7. [CrossRef] Zhu, H.H.; Yin, J.H.; Zhang, L.; Jin, W.; Dong, J.H. Monitoring Internal Displacements of a Model Dam Using FBG Sensing Bars. Adv. Struct. Eng. 2010, 13, 249–261. [CrossRef] Schroeck, M.; Ecke, W.; Graupner, A. Strain monitoring in steel rock bolts using FBG sensor arrays. In Proceedings of the International Society for Optical Engineering, Applications of Optical Fiver Sensor, Glasgow, UK, 31 August 2000; Volume 4074, pp. 298–304. Huang, M.; Zhou, Z.; Huang, Y.; Ou, J. A distributed self-sensing FRP anchor rod with built-in optical fiber sensor. Measurement 2013, 46, 1363–1370. [CrossRef] De Waele, W.; Moerman, W.; Taerwe, L.; Degrieck, J.; Himpe, J. Measuring ground anchor forces of a quay wall with Bragg sensors. J. Struct. Eng. 2005, 123, 322–328. Kim, J.M.; Kim, H.W.; Park, Y.H.; Yang, I.H.; Kim, Y.S. FBG Sensors Encapsulated into 7-Wire Steel Strand for Tension Monitoring of a Prestressing Tendon. Adv. Struct. Eng. 2012, 15, 907–917. [CrossRef] Do, T.M.; Kim, Y.S. Prediction of load transfer depth for cost-effective design of ground anchors using FBG sensors embedded tendon and numerical analysis. Geomech. Eng. 2016, 10, 737–755. [CrossRef] Xu, G.Q.; Xiong, D.Y.; Duan, Y.; Cao, X.S. Open pit slope deformation monitoring by fiber Bragg grating sensors. Opt. Eng. 2014, 54, 011003–011008. [CrossRef] Chen, W.M.; Zhang, P.; Wu, J.; Liu, L.H. Embedding technology of Fiber Bragg Grating strain sensor for cable tension monitor. In Proceedings of the International Conference on Optical Instruments and Technology: Optical Sensors and Applications, Beijing, China, 20 December 2013; Volume 9044. Shin, C.S.; Liaw, S.K.; Yan, S.W. Post-Impact fatigue damage monitoring using Fiber Bragg Grating Sensors. Sensors 2014, 14, 4144–4153. [CrossRef] [PubMed] Kim, J.M.; Kim, C.M.; Choi, S.Y.; Lee, B.Y. Enhanced strain measurement range of an FBG sensor embedded in seven-wire steel strands. Sensors 2017, 17, 1654. [CrossRef] [PubMed] © 2018 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).
