CFRP网格加固受损盾构隧道模型试验研究
来源:步遥情感网
Journal of Southeast University(English Edition) Vo1.33,No.2,PP.196—202 June 2017 ISSN 1003--7985 Experimental study on a pre-damaged scaled tunnel model strengthened with CFRP grids Li Wensheng Wu Gang Dong Zhiqiang Wang Shuying ( Electric Power Research Institute of Guangdong Power Grid Co.,Ltd.,Guangzhou 5 10080,China) ( Key Laboratory of Concrete and Prestressed Concrete Structures of Ministry of Education,Southeast University, Nanjing 210096,China) Abstract:The effectiveness of carbon fiber reinforced polymer decks reinforced with carbon fiber reinforced polymer (CFRP)grids as the strengthening materials for a pre—damaged scaled tunnel model is experimentally investigated.First,the bond performances between the CFRP grid and the concrete under different types of adhesive and surface treatment were testde.The most efficient anchonng system was adopted for the subsequent scaled tunnel strengthening.Test results show htat when the epoxy structural adhesive was used as the obnding materila,the failure mode was cFRP鲥ds rupturing, and the anchorage performance was optima1. n the polymer mortar was used as the adhesive,the surface rteatments with anchored bolts and grooves can improve the bond performance,and the failure mode was sliidng failure wiht the polymer mortar peeled off.After strengthening wim CFRP grids.both the stiffness and the load capacity of the pre-damaged scaled tunnel model were improved.Additionally, hte results obtianed by fiber bragg grating(FaG)sensors indicate that the strains across tunnel segments were reduced. nad the over ̄l performance ofthe tunnel was improved. Key words:carbon fiber reinforced polymer(CFRP)grid; double.shear bond test; surface treatment; scaled tunnel model:relative displacement DOI:10.3969/i.issn.1003—7985.2017.02.012 F isbhere rtse,ainfonrdc edp lpaotelys mhearvse( FbRePens )wini dtehley foursm edo fi nb arhts,e strengthening of concrete structures for decades 11.In re. cent years。FRP grids,as another form ofFRP materials. have attracted increasing interest from researchers as both inner tensile reinforcementsI and external bonded strengthening materials 一 ….The properties of concrete beams,slabs and bridge decks that were reinforeed or strengthened with F1王P grids have been extensively inves. tigated.Rahman et a1.1Ol tested the performance of bridge Received 20l6—12.1O. Biographies:Li Wensheng(1984一),male,doctor;Wu Gang(co丌e- sponding author),male,professor,g.wu@seu.edu.cn. Foundafion items:The Science and Technology Project of China South— erll Power Grid Co.,Ltd.(No.GDKJoOOoo03O),the Nafional Key Technology R&D Program of China(No.2叭6YFO07O140o),the Na- tional Natural Science Foundation of China(No.51525801). Citation:Li Wensheng,Wu Gang,Dong Zhiqinag,et a1.Experimentla study on a pre.damaged scaled tunnel model s ̄engthened wiht CFRP grids[J】.Journal of Southeast University(English Edition),2017,33 (2):196—202.DOI:10.3969/j.issn.1003—7985.2O17.O2.012. (CF】 P)grids under service load and ultimate load.It was found that the construct ability of CFRP grids in strengthening bridge decks was good.The deflection and stress were both reduced as indicared by the cyclic loading test results.Researchers also tested the static and seismic performance of concrete columns reinforced or strength. ened with FRP ds… .Wu et a1.㈣conducted an axi. al compression loading test on concrete columns wrapped by n grids;it was concluded that both the strength and ductility of the confined column were clearly improved, nad the external bonded FRP grids can effectively increase the seismic performance of RC concrete structures. WitIl the expanding body of research.the FRP grid is regarded as one item that can be effectively used in tunnel strengthening due to its properties of being lightweight, ofhi曲strength,and good durability.It can not only re- inforce the strength and stiffness of the cross-section of hte tunnel but also improve the longitudinal flexural stiff- ness of it and thus lower the impact of uneven settlement on the tunne1.For a subway tunnel in a city,the strengthening can only be carded out during the suspen・ sion of operation of the subway;the period of the suspen— sion of operation is only approximately three to four hours every day.This limited period requires that the strengthe— ning method must have a high construction emciency.  ̄rhen the FRP grids are used in tunnel strengthening.a light grid can be easily fixed onto the sprayed substrate concrete using a compressed air gun;subsequently,a thin layer of concrete is sprayed as cover.Technically。it is mature.simple and convenient.Thus. grids have a good prospect for application in tunnel strengthening. To further investigate the feasibility of using FRP grids as a strengthening material in a tunne1.CFRP grids are manufactured in the laboratory,and the bond performance Of CFRP grids with concrete under different adhesives and surface treatments are investigated.A1so.a loading test is conducted on a CFRP grid strengthened scaled tunnel model in the laboratory. 1 Materials and Specimen Design 1.1 Materials The CFRP grid used in this paper was manufactured Experimental study on a pre—damaged scaled tunnel model strengthened with CFRP grids l97 using the vacuum.assisted resin transfer molding f VAR。 TM、technique in the laboratory.The carbon fibers were s provided by the Japan Toray Company, and the resin ma— trix of vinyl ester was provided by the Yabang Investment Holding Group Co.,LTD.As shown in Fig.1,there were eight strands of T800 yarn in the longitudinal direction and ive strafnds of T700 yam in the transverse direction.The grid spacing was 30 iilnl×30 mrn.and the cross—section of the single bar was 3 nqn3 fwidth)×2 mill fthickness)in both directions.The tested mechanical properties in the two directions are shown in Tab.1. of yarn f a) strands of yarn T7oo Trall'SV (b) Anchor steel plate Wood separator Fig・1 The form of the manufactured CFRP grid(units:mm) Tab.1 Tested mechanical propeaies of the CFRP grid s :el plate \ 匝 』一T二 ————,A ch0rb0lt 2 、…— — :三 ____一—_J_“删Stel ph ・—- _一 1一 e 匕■ 盈_ I曩■黛圈圈圈啊-一l一 —’ ru“ 250 一I b 0I_80 ̄1_80—1 0LBd … I 300Note:COV denotes the coeficifent of variations ——I l一 一…(c) f-_●●● ● The epoxy structural adhesive used was type TLS-40l, which was provided by the Nanjing Telesun S&T Industry Co..LTD.The performance speciicatifons of TLS-40 l are listed in Tab.2.The polymer mortar was made by mixing dry powder and the liquid matrix at a mass ratio of 3:1.The dry powder was comprised of cement:coarse sand:fine sand at a ratio of 2:2:1.and the 1iquid matrix was comprised of water:acrylic.emulsion:90 l glue at a ratio of 2:0.5:0.9.The tested flexural strength was higher than 8.0 MPa,and the cohesive pulling strength was higher than 1.2 MPa. Tab.2 The performance of the structural adhesive used in this (d) Fig.2 Schematic diagram of the double—shear bond test speci— paper(type TLS-401) mens(unit:mm).(a)Surface treatment with grooves;(b)Cross- section of the anchored end;(c)Surface treatment with anchor bolts; (d)Cross-section of the test section on the research bv Shi et a1. 131The targeted concrete .grade was C50,and the tested 28・day cube strength was 1.2 Specimen design 1.2.1 Double-shear bond specimens 65.7 MPa.The test matix are shown in Tab.3.To im—r prove the bond performance,two types of surface treat— ments.i.e.,grooves and anchor bolts,were adopted,as Fig.2 illustrates the specimens prepared for the double— shear bond tests;the sizes of concrete blocks were based shown in Figs.2(a)and(c). l98 Li Wensheng,Wu Gang,Dong Zhiqiang,and Wang Shuying Tab.3 Test matrix for the double—shear bond performance test Note:Each group contains three identical samples As shown in Fig.2,the CFRP grids were symmetrical— Fig.3 The scaled tunnel mode with the CFRP grid placed in— side in the laboratory lv attached to the two sides of the concrete block.Steel plates were embedded inside the concrete blocks and were connected to the machine grips when testing.The two concrete blocks were separated by a wood separator at the middle section.At one end,the concrete block was clamped by two steel plates,as shown in Fig.2(b), which formed the anchored end.The other end was used as the test section.The tests were conducted on the uni— versal testing machine,and the loading speed was 0.8 mm/min.As shown in Figs.2(a)and(c),strain gauges were attached to the middle CFRP along the pull direc— tion. 1.2.2 Scaled tunnel model According to the analogous theory,a scaled tunnel model was manufactured J41.Tab.4 shows a comparison of the parameters of the practical tunnel segment and the scaled mode1.The adopted scale similarity ratio was 8, and the stress scale similarity ratio was 1. Tab.4 Parameters of the practical tunnel segment and its scaled mode1 Number in the 1O circumferential direction l0 Joints Number in the 16 longitudinal direction Diameter of bolts/mm 32 4 Fig.3 shows the simulation of the tunnel’s surrounding conditions.The surrounding soil was silty clay.and the tested modulus of compression was 3 MPa.The thickness of the soi1 between the ground and the tunnel bottom was 0.3 m,and the tota1 height of the soil was 0.9 m. According to the test set—up shown in Fig.4,the load was applied with an MTS loading system.Uniformly (a) (b) Fig.4 Test set—up.(a)Side view;(b)Top view distributed loads were applied to the middle six segments for the simulation of a local settlement.As shown in Fig. 4(b),for the creation of a uniformly distributed 1oad and also to avoid loca1 shear failure.arc steel plates with a width of 380 mm were placed between the load sensor and tunnel segments.In addition,rubber blankets with a thickness of 1 0 mm were placed between the arc stee1 plates and the tunnel segments.A displacement control mode was adopted as the load method in this experiment. There was a total of l 08 load steps.with a step size of O.5 mm.For the first 70 steps.the 7lth to 90th step. and the 91 th step to the end,the loading rates were 0.5, 1,and 2 mm/min,respectively. As shown in Fig.5(a),the tunnel model was divided into l 6 elements,named E1 to E1 6 one by one.with each element comprised of two half tunnel segments and one circumferential ioint and the l st and the 1 7 th half seg— ment was not considered.Since the damage at the middle eight elements,i.e.,E5 to E1 2。should be much more serious than that at the two end parts,sixteen fiber bragg grating(FBG)sensors were longitudinally attached to the bottom and top concrete walls inside the tunne1.FBG Experimental study on a pre—damaged scaled tunnel model strengthened with CFRP grids 199 sensors could monitor the overal1 longitudinal strains a. Tab.5 Double—shear bond test results 1ong the bottom and top concrete walls.These FBG sen— sots were named F1 to Fl 6 one by one.Regarding the monitoring of the displacements,the displacements at the top,bottom,and lateral walls were tested using linear variable differential transformers f LVDTs).The positions of the LVDTs are shown in Fig.5(a),and Fig.5(b) shows the inside photograph of the tunnel with the LVDTs placed.The top and bottom displacements at the l st,5th,9th,l3th,and 17th segments were testedand .the 1ateral displacements at the 1 st.9th.and 1 7th seg.ments were tested. Fll Fl3 FI5 F9Fl0 Fl2 F14 FI6 ・ 十・ ・ _.一一一・…一一一一一FBG l lll0 jlll Nore: ':Top LVDT D ・():Later LVDT ●:Bottom LVDT D1D D2B D3 DEl ,sE2 E3FAo 熙殴巴 :式 。粤Q F1 F2 F3 F4 F5 F6 F7 F8 E【)-Segments:F()--FBGs;D()()一LVDTs (a) (b) Fig.5 The positions of LVDTs and FBGs.(a)Schematic dia gram;(b)Inside photograph 2 Double.Shear Bond Test Results 2。1 Failure modes and ultimate loads The test results of the l 2 double—shear bond specimens are 1isted in Tab.4.and their typical bond failure modes are shown in Fig.6.As shown in Figs.6(a)and(b),the specimens with epoxy structural adhesive both failed be. cause the CFRP ruptured,and almost no slips were ob— served.With the increase of the load.the structural adhe— sive first cracked at one side,and then small sounds of if. ber rupture were heard.As the load continued to in. crease,the structural adhesive at the other side also cracked and an increasing number of sounds of fiber rup— ture was heard.Finally.the CFRP grids near the wood separator ruptured.For the HC group with grooves off the concrete surface.oblique cracks of 45。on the concrete near the crack of structural adhesive were observed.indi. caring the good bond performance between the structural adhesive and the concrete substrate.The ultimate loads of the H0 and HC groups were similar.indicating that the surface treatment with grooves was unnecessary when the structural adhesive was used. Note:MF denotes material failure;SF denotes sliding failure (a) (d) Fig.6 Typical bond failure modes for the double—shear tests (a)H0;(b)HC;(c)JM;(d)JC As shown in Figs.6(c)and(d),the inner CFRP grids of specimens with polymer mortar as adhesive did not urpture,and the mortar layer peeled off from the concrete substrate.For the JM group,as shown in Fig.6(c),the anchored bolts provided an out..of..plane constraint on the CFRP grids.and the mortar did not obviously fall off. Li Wensheng,Wu Gang,Dong Zhiqiang,and Wang Shuying The local fiber rupture observed on the CFRP grid near he boltts was due to the shear force.However,it did not influence the overall bond performance.For the JC group shown in Fig.6(d),the mortr ain grooves served as shear connectors,and they were sheared.off after the test. Since there were no out.of-plane constraints,the mortar obviously fen off from the concrete substrate.and the CFRP grid was pulled out.Similar to the groups with印- oxy structural adhesive.the ultimate loads of JC and JM group were similar,but they were obviously lower than hose of tthe H0 and HC groups. Posiion of sttra ̄gauges/ram (a) 2.2 Strain distribution at CFIuP grids Fig.7 shows the tested strain values of CFRP grids an— der different load levels;the horizontal axis here repre. sents the distance between the strain gauges and the wood separator(as shown in Fig.2).Figs.7(a)and(b)show the results of the H0 and HC groups wih epoxy sttructural adhesive.The strains close to the pull end were found to be lrge wiah tthe values decreasing rapidly with the in— crease in the distances.The bond strength transfer lengths were approximately 250 nLrn for both the H0 and HC 0'【 — 皇∞ 0H,Ⅱ ∞ groups.By comparing Fig.7(a)and Fig.7(b),the 0'【,Ⅱ— 摹∞ Posiiton of strain gauges/ram (b) 0'I壹∞ CFRP strains of the HC group with grooves at concrete substrate were found to decrease much more rapidly than tllat of the H0 group due to the shear resistance of adhe. sive inside the grooves. Figs.7(c)and(d)show the results of the JM and JC groups with polymer mortar.The strain values were found to be obviously higher than those of the specimens wih structuralt adhesive due to a weaker bond effect.For the group with anchored bolts.the strains decreased Posiiton of strain gauges/mm with the increase of distances as well,but the speed was obviously lower than the specimens with stuctrural adhe. (c) sire.For the JC group wih grooves itn the concrete sub- strate,in the first 15O mill length of the bond area.the strain values were basically the same;with the further in. crease of the length,the strains suddenly decreased. 3 Test resuits of the Scaled Tunnel Model 3.1 Load vs.relative displacement curves Considering the influence of compressive deformations Posiiton of strain gauges/ram of the surrounding soils,the analysis is based on the difference between the displacement tested by the top LVDT and that of the bottom LVDT.The difference va1. ue was defined as the relative displacement.Fig.8 shows the load vs.relative displacement curves of the undam。 aged and strengthened tunnel at the middle span.Note that the loading of the undamaged tunnel was stopped (d) Fig.7 CFRP strain distribution aiong the longitudinal direc— iton.(a)HO;(b)HC;(c)YM;(d)JC wim Cf佩P grids and epoxy structural adhesive.The strengthened tunnel was then tested until the load capacity obviously decreased. As shown in Fig.9,when comparing the theoretical re— when the crack widths reached the value the code pre. scribed.At that point,the load value was 165.68 kN, load path of the pre—damaged tunnel with the strengthened curve,both the stiffness and strength of the strengthened tunnel were found to be significantly improved.Within nd tahe relative displacement was 6.89 mm.Next。the pre—damaged tunnel model was unloaded and strengthened the initial 0 to 2.5 mm relative displacement.the load Experimental study on a pre—damaged scaled tunnel model strengthened with CFRP grids 201 卯 2 ∞ 2 ∞ pl 蚤, 3 ∞ 1 ∞ O 蚕,pg good agreement with the theoretical results based on the elasitc foundation beam heorty.As shown in Fig.10(b)。 for the strains at the top section,the compressive strains mainly appeared at the loading area,i.e.,the F10 to F15 area.The tested strain distributions at the top section met he actuatl loading condition.The observed crack damage was due to pre—loading at he tjoint of tunnel segments that he Flt4 FBG crossed.However.after strengthening witll Cf;l grids.the strains at F4 FBG remained at small va1. ues because tl1e tunnel segments were further connected as 0 10 20 30 40 50 60 a whole by the CFRP grid layer.thereby improving the Relative displacement/ram Fig.8 Load vs.relative displacement curves ofthe scaled tun- nel mode at the middle span Fig.9 Comparison of the load-relative displacement curve of hte undamaged tunnel and the strengthened tunnel within the salne range of relative displacements capacity of the strengthened tunnel was still slightly lower tllan mat under the undamaged condition because the tun. nel was pre.damaged.With the increase of deformation, hte utilization rate of the CFRP grids improved.and the load capacity of the strengthened tunnel was higher than tIlat of the undamaged tunne1.The load capacity of the strengthened tunnel was found to be increased by 19.2% at the loading stopping point of the undamaged tunne1. However,we should also be aware of the excessive posi. rive effect of the epoxy strucrural adhesive on the im. provement of the stifness because the practical adhesive layer will be very thick after multiplying it by the scalde ratio。which is unrealistic. 3.2 Strain distributions tested by FBG s既Isors 111e eight loading grades as listed in Tab.6 were adopt. ed in the analysis of the strain distribution tested by the FBG sensors.The results are shown in Fig.10.硼1e two loading arc plates were located at the E6.E8 and E9.El1 segments,as shown in Fig.5.As shown in Fig.10(a), hte srtain values measured bv F3 and F6(1ocated under hte loading points)were obviously higher than those measured by F4 and F5 due to a higber base sealement. hTe strain distribution curves were saddle—shaped,in overall performance. Tab.6 The eight loading grades adopted in the analysis o.D-0H 月 ∞ 0'I,Ⅱ_fg∞ 加∞∞∞印∞加∞∞∞∞ 抛啪姗 啪啪∞∞∞ Number of FBG (a) 6 Number of髓G (b) iFg.10 Testde strains by the FBGs.(a)Bottom section;(b)Top section 4 Conclusions 1)For the double.shear bond specimens,when也e ad- hesive was e口oxy structural adhesive,the failure mode was the rupturing of the CFRP grids.When the polymer mortar was used as an adhesive,the failure mode was sliding failure.and the CFRP grids did not rupture. 21 The surface treatment with grooves was unnecessary when the structural adhesive was used.For the bond specimens using polymer mortar,the anchored bolts pro- vided an out.of-plnae constraint on me CFI grid.and Li Wensheng,Wu Gang,Dong Zhiqiang,and Wang Shuying the mortar did not obviously fa11 off.In contrast.the 【7】Wu G,Wu Z S,Jiang J B,et a1.A new technology of strengthening concrete structures with FRP grids and its mortar obviously fe11 off from the concrete substrate,and he CFRP grid was removed for specimens witth surface grooves. . application[J】.Const Tech,2007,36(12):98—102. (in Chinese) 【8】Zheng Y Z,Wang W W,Brigham J C。Flexural behav— iour of reinforced concrete beams strengthened wih a t31 For the strengthened tunnel,with the increase of de— formation.the utilization rate of the CFRP grids im— proved,and the load capacity of the strengthened tunnel was higher than that of the undamaged tunne1. 4)The tunnel segments were further connected as a whole by the CFRP grid layer.and the overM1 perform— ance of the tunnel was improved. composite reinforcement layer:BH grid and ECC【J]. Construction and Building Materials,2016,115:424— 437.DOI:10.1016/j.conbuildmat.2016.04.038. 【9】Papanicolaou C,Triantafillou T,Lekka M.Externally bonded grids as strengthening and seismic retrofitting ma- terils of masonry panels【J].Constructiaon and Building Materials,2011,25(2):504—514.DOI:10.1016/j. 1 ferences 【1】Teng J G. P strengthened尺c structures[M].Beijing: conbuildmat.2010.07.018. 【10】Corradi M,Borri A,Castori G,et a1.Shear strengthe— China Building Industry Press,2005.(in Chinese) 『21 Banthia N,A1.Asaly M,Ma S.Behavior of concrete slabs reinforced with fiber—reinforced plastic grid f J 1. ning of wall panels through jacketing with cement mortr areinforced by GFRP grids【J】.Composites Part B:Engi— neering,2014,64:33—42.DOI:10.1016/j.compos— itesb.2014.03.022. Journaf of Materials in Civii Engineering,1995,7(4): 252—257.DOI:10.1061/(asce)0899—1561(1995)7:4 (252). 『3 1 Mattllvs S,Taerwe L.Concrete slabs reinforeed wiⅡl FRP 【11】Wu G,Wu Z S,Luo Y B,et a1.Aseismic performance of circular concrete columns confined wih FRP grtids【J】. Journal f oArchitecture and Civil Engineering,2007,24 (4):39—44.(in Chinese) 【12】Bentayeb F,Tahar K A,Chateauneuf A.New technique for reinforcement of concrete columns confined by embed— grids.I:One—way bending[J】.Journal of Composites r Construction,2ooO,4(3):145—153.DOI:10.1061/ (asce)1090-0268(2000)4:3(145). 『41 Matthys S,Taerwe L.Concrete slabs reinforced with同 ded composite grid[J】.Constuctrion and Building Mate— grids.Ⅱ:Punching resistance[J】.Journal of Compos— ites for Construction,2O00,4(3):154~l61.D0I:10. rials,2008,22(8):1624—1633.DOI:10.1O16/j.con— buildmat.2007.06.015. 1061/(asce)1090-0268(2o00)4:3f 154). 【5】Yost J R,Goodspeed C H,Schmeckpeper E R.Flexural performance of concrete beams reinforced witll FI P grids 【13】Shi J W,Zhu H,Wu Z S,et a1.Experimental study on bond behavior between basalt/hybrid FRP sheets and con— crete substrates[J】.Journal fSoutoheast University:Nat— ural Science脚tion,2010,40(3):554—558.(in Chi— nese) [14】Shen S.Development of structural health monitoring sys— tem for underground tunnel based on the distributed fiber [J】.Journal ofCompositesfor Construction,2oo1,5(1): 18—25.D0I:10.1061/(asce)1090 268(20o1)5:1(18). 【61 Rahman A H,Kingsley C Y,Kobayashi K.Service and ultimate load behavior of bridge deck reiforced winth cal'- bon FRP grid【J】.Journal of Composites for Construc— tion,20o0,4(1):16—23.DOI:10.1O61/(asce)1090— 0268(2000)4:1(16). optic sensing technique【D】.Nanjing:School of Civil Engineering,Southeast University,201 1.(in Chinese) CFRP网格加固受损盾构隧道模型试验研究 李文胜 吴 刚 董志强 王淑莹 ( 广东电网有限责任公司电力科学研究院,广州510080) ( 东南大学混凝土及预应力混凝土结构教育部重点实验室,南京210096) 摘要:对碳纤维增强树脂(CFRP)网格加固受损缩尺盾构隧道模型的有效性进行了试验研究.首先,对CFRP 网格与混凝土表面采用不同粘结材料和表面处理方式时的粘结性能进行了研究.采用最有效的锚固方式对 随后的缩尺隧道进行了加固.研究结果表明,采用环氧结构胶作为粘结材料时,粘结试验均表现为网格材料 破坏,粘结锚固性能最优;而采用聚合物砂浆锚固时,表面锚栓锚固处理和表面开槽处理均能提高粘结性 能,破坏模式为聚合物砂浆剥落的滑移破坏.受损隧道模型采用CFRP网格加固后,刚度和承载力都得到了 明显提升.此外,FBG的测试结果表明,加固后的隧道环缝应变减小,隧道整体性能得到提升. 关键词:CFRP网格;双面剪切试验;表面处理方式;缩尺隧道模型;相对挠度 中图分类号:TU502