BRIDGE INSPECTION STANDARDS IN JAPAN AND US

Transcription

1 BRIDGE INSPECTION STANDARDS IN JAPAN AND US Masahiro Shirato 1 an Takashi Tamakoshi 2 Abstrat Elmnt-lvl brig insption has bn vlop in Japan an th US rsptivly. Convntional maintnan / rhabilitation urgny rating ns a iagnosis givn by qualifi nginrs an is somhow subjtiv. In aition to that, th lmnt ata roring of amag stat rat in an objtiv mannr is a nw trn. Th prsnt papr ompars th historis an onpts of brig insption program in both ountris an highlights som of th lmnt-lvl insption rsults in Japan. Th rsults show that th aumulation an big ata mining of objtiv lmnt-lvl ata has a hug potntial to improv brig managmnt an sign / insption stanars. Introution Brig insption is primarily onut to assss th strutural safty an rlat maintnan urgny for iniviual brigs. Aoringly, brig insption mans a omprhnsiv nginring (or subjtiv) jugmnt for strutural safty an maintnan urgny at th strutural mmbr lvl or omponnt lvl or brig lvl. Howvr, rntly ollting objtiv / sintifi amag rating at lmnt lvl has bn wily apt both in Japan an th US an xut in aition to th onvntional insption stanars. For xampl, th amnmnt of Roa Law was approv in Japan in May 2013, larifying that it is an obligation for all roa aministrators to inspt struturs with onsiration of prvntiv maintnan. It also mpowrs Ministr for Lan, Infrastrutur, Lan, Transport an Tourism to invstigat roa aministrators statuss for highway maintnan for th sak of thnology vlopmnt. In th US, MAP-21 (Moving Aha for Progrss in th 21 st Cntury At) was nat in Whil th onvntional national brig insption stanars with a 0-9 sal rating ontinu to b xut for strutural omponnts of all brigs on publi roas, now Stats ar rquir to ollt lmnt lvl onition stat ata st of brigs on th National Highway Systm. With ths bakgrouns, both Japan an th US hav stablish lmnt-lvl brig insption stanars ths ays, rsptivly. Howvr, th finition of lmnt in brig insption is iffrnt btwn Japan an th US. It is not a problm of whih is bttr or whih is wors. It shoul b important to mak th finition of lmnt an ata roring strutur mt th aims of lmnt lvl insption, whih may vary with brig ownr by brig ownr. Aoringly, th prsnt papr onsirs what shoul b onsir to st out 1 Snior Rsarhr (Brig & Struturs), National Institut for Lan an Infrastrutur Managmnt, Ministry of Lan, Infrastrutur, Transport an Tourism 2 Ha, Brig an Struturs Division, itto

2 th finition of lmnt or unit of onition rating or ata olltion strutur in insption. Th prsnt papr first rviws xpt outputs from big-ata mining in brig insption ata in th ra of sintifi lmnt lvl brig insption. Ministry of Lan, Infrastrutur, Transport an Tourism brig insption protools an US NBIS (National Brig Insption Stanars) ar ompar, pointing out th iffrn in th finition an unit of omponnts an lmnts in brig insption (MLIT, 2004). Sonly, th tails in MLIT s finit lmnt lvl amag rating protool ar rviw, in whih th MLIT protool is ompris of amag apparan ratings for finit lmnts an maintnan urgny ratings for strutural mmbrs., whr th finition of finit lmnt will b xplain latr. Finally, th prsnt papr shows som highlight sintifi fats that ar foun out in th ata of MLIT `finit lmnt lvl brig insption an isusss th potntials of MLIT s finit lmnt lvl brig insption on promoting ata-rivn brig maintnan an R&D in brig nginring. Brif historis of Elmnt-lvl Brig Insption in Japan an th US Brig insption protools pn on th aims of insption. Tabl 1 shows xpt sintifi / nginring ahivmnts an aministrativ ahivmnts in brig insption. Th primary onrn of brig insption is to sur th strutural safty for passngrs moving unr an on brigs. Convntional brig insption programs both in MLIT an th US ar stipulat mainly for valuating th urgny for maintnan or othr ations from th viwpoint of safty. In Japan, a mol brig insption manual vlop in 1988 by Publi Works Rsarh Institut (PWRI) of thn Ministry of Constrution an us to b rommn (lgally non-bining). A singl rating iniator was assss for ah strutural mmbr in ah span, aounting for th xtnt of amag an th urgny of maintnan simultanously. In th US, th Silvr Brig spanning th Ohio Rivr ollaps in 1967 an NBIS was nfor in Insption is manatory for all brigs on publi roas. Insption frquny is two yars in prinipl. Stats hav ollt ovrall onition ata for omponnts at th brig lvl, whr omponnts ar fin as suprstrutur, substrutur, k, an ulvrts. Stats also shall onut monitoring th brig whn a ritial fining is foun uring an insption. NBIS was upat svral tims, larning brig ollapss an major failur vnts. In 1990s-2000s, th importan of prvntiv maintnan was raliz in both Japan an th US, rsptivly, for bttr brig managmnt. Tabl 2 shows th numbr of brigs an vhil shars by ownr in Japan. MLIT has an oprats as many as 20,000 brigs on signat stions of national highways. Th Japans roa ntwork was intnsivly vlop uring th rapi onomi growth of th 1970s, an th numbr of roa brigs has now rah approximatly 700,000 (brig lngth 2 m). As shown in Figur 1, it is prit that th numbr of brigs olr than 50 yars will aount for almost half of brigs in 15 yars. Fatigu in stl pirs an k plats, hlori ingrss in prstrss onrt bams, alkali silia ration (ASR) in onrt struturs, an fatigu failur of onrt in rinfor onrt ks wr wily rport. MLIT rais a prvntiv maintnan initiativ an orr National Highway Offis to implmnt th prsnt on-in-5-yar brig

3 insption protool in 2004 (MLIT, 2004). Maintnan urgny rating for strutural mmbrs is inlu, but amag apparan rating for mor tail units is also ror. Damag apparan ratings ar introu baus of xpting a sintifi ahivmnt of No. 2 in Tabl 1, ata-rivn systmati prvntiv maintnan. Howvr, it is also sign to ahiv ata-rivn upats an improvmnts in sign spifiations an insption protools as wll as th unrstaning of ns in brig prsrvation thnology, orrsponing to th xpt sintifi ahivmnts Nos. 3 an 4 in Tabl 1. Baus prvntiv maintnan is rquir to ru futur maintnan osts, MLIT start a subsiy program in 2007 for loal govrnmnts to stablish long-trm brig maintnan programs. As a rsult, prfturs now hav brig insption programs that follow th MLIT protool an somtims with som hangs. Namly, brigs on artrial routs that arry a larg part of traffi ar now inspt prioially. Howvr, som muniipalitis still o not inspt thir brigs. Som onut but insption protool an quality ar not stanariz. In th US, svral stats start improving thir brig insption program to obtain onition ata for strutural lmnts for brig managmnt systm us. CoR lmnt ata olltion gui was vlop in AASHTO for brig ownrs to aim at bttr brig prsrvation programs an prforman-bas bugting. Most stats start to follow it (AASHTO 2002). MAP-21 was nat in 2012 an now all stats will n to ollt lmnt lvl insption ata for brigs on NHS, following th rporting systm of AASHTO Manual for Brig Elmnt Insption (AASHTO, 2013). FHWA is also irt unr MAP-21 to stuy ost-fftivnss, bnfits an fasibility of ollting lmnt-lvl ata for non-nhs. Summary of Brig Insption Protool in Japan Maintnan Urgny Rating In both Japan an th US, brig insption protools ar ompris of th part for assssing strutural safty an th part for ollting quantitativ an objtiv amag ata for lmnts. In trms of th strutural safty part, thr is no big iffrn in philosophy btwn th MLIT protool an th US NBIS, xpt for th unit of assssmnt. In th MLIT protools, brigs with a span longr than 2 m ar inspt on in vry fiv yars. Hans-on visual insption is rquir. Tools an vis may b us with a limit amount to trmin suh as fatigu rak in stl mmbrs. For ah strutural mmbr in ah span, th onition is translat into ithr of th following maintnan urgny ratings: A B C E1 No rpairs n. No immiat rpairs n. Rpair n Emrgny ation is nssary from th viwpoint of strutural safty

4 E2 M S an stability Emrgny ation is nssary baus of othr fators. Rpairs n in th ours of th rgular maintnan work Furthr tail invstigations n Maintnan urgny ratings ar iagnosis givn by xprin nginrs in a subjtiv mannr, rommning to brig ownrs th ns for ation by th tim of th nxt insption. Enginrs ar rquir to intrprt th maintnan urgny for ah mmbr, taking into aount th amag apparan ratings in finit lmnts in th strutural mmbr, spifi haratristis of th amag suh as th irtion of rak, th loation of amag in th strutural mmbr, th funtion of th strutural mmbr, likly auss / sours of amag, intrations with othr amag at othr strutural mmbrs an omponnts in th brig, arlir rmial work history, th k / slab oating systm, th rainag systm, nvironmntal fators suh as traffi volum, iing salt umping volums, th istan from th sa t an so on. No numri ritria lik rak with an lngth ar spifi for maintnan urgny ratings. In th US, ah brig is ivi into four major parts for assssmnt: k, suprstrutur, substrutur an ulvrt. Insptors giv ithr of 0-9 ratings in a subjtiv mannr of intrprtation for amag. Th rating sal for ah part rangs from 0-9 pning on th svrity of amag an urgny of ation. Namly, four inis srib th xtnt of amag as a whol brig. Figur 2 shows a shmati iagram of trioration urvs. Dtrioration rats hav a hug varity an th varity inrass with inras in th xtnt of amag. Figur 3 shows an xampl of tst rsult for th fatigu of onrt in RC k slabs. Whl loaing tsts wr onut for two spimns, whr a whl subjt to a givn axl loa mov bak an for on th spimn until th spimn ollaps. Th tst spimns wr ut out of two iffrnt xisting brigs an thir rak withs an nsitis in onrt ar vry similar to ah othr. Howvr, Spimn A in Figur 3 i not ollaps at a whl movmnt of 200,000 yls whil Spimn B ollaps at a whl movmnt of 20,000 yls, with 1/10 of th urability ompar to Spimn A. This shoul b baus som raks ha alray pntrat through th ntir pth of k slab bfor th xprimnt as iniat by fflorsn mark with th r irl in Figur 3. As also wll known, othr fators suh as watr oming from th k slab surfa an alrat th volution of fatigu rak in k slab onrt. Figur 4 shows raks in stl mmbrs. Th influn of raks on th safty of brigs pns on rak loations an irtions. Namly th maintnan urgny rating for raks annot b asily stanariz as a funtion of rak withs an lngths an it shoul misla nginrs to st out any numri ritria for maintnan urgny ratings suh as a funtion of rak withs an othr numri paramtrs. Enginr s iagnosis on sit shoul go first. (Finit) Elmnt-Lvl Damag Apparan Rating Prvntiv maintnan has bn promot in both Japan an th US in rnt

5 yars to ru th futur rhabilitation ost. Data-rivn managmnt has bn xpt to hlp brig aministrators sk a prfrabl timing an prvntiv rmial masur to provi for iniviual brigs. Quantitativ istributions of iffrnt xtnts of amag within omponnts ar rquir to stimat futur maintnan osts mor prisly, baus prvntiv maintnan is somtims onut span-by-span or portion-by-portion. In aition to brig managmnt us, th importan in ata objtivity has bn wily rogniz to xamin th long-trm brig prforman an improv brig sign. Howvr, th maintnan urgny rating os not nssarily qual th xtnt of amag apparan. It involvs a subjtiv prition by nginrs rgaring th tim volution in xisting amag an rlat graation in strutural safty, alling for taking into aount various fators suh as th possibl auss of amag. Aoringly, an lmnt-lvl ata olltion with sintifi / objtiv amag onition ratings is rquir to aptur th typ an istribution of istrss in omponnts an monitor th xat xtnt of prsnt amag at ah istrss an at th tim of th insption. Both lmnt-lvl ata olltions in Japan an th US ror objtiv, not subjtiv, stanariz onition stats for spifi fts. Howvr, as illustrat in Figur 5, thr is a notabl iffrn in th finition of lmnt an th ata roring philosophis / struturs btwn Japan an th US. As spifi in th AASHTO manual (AASHTO, 2013), lmnts in th US ar brakowns of omponnts that ar irtly rlat to th loa apaity suh as th group of girrs, th group of olumns, th group of abutmnts, th group of fix barings an th group of movabl barings in a brig. In aition, sonary omponnts suh as prottiv oating systms, waring surfas, an joints ar also st out. Th ata strutur is summariz as follows: Catgory of lmnt Spifi fts Damag ratings: goo, fair, poor, an svr for ah atgory of ft Quantitis of ah atgory of ft in ft, ara, or ah for numrat lmnts for ah atgory of ft an ah amag xtnt. Elmnts fin in th MLIT protools ar subivi portions of iniviual strutural mmbrs at iniviual spans. For xampl, as shown in Figur 5, vry singl girr for ah span is subivi into svral parts at th position of floor bams in a span. Figur 6 illustrats xampls of lmnt atgoris an finit lmnt mshing for amag apparan ratings in th MLIT protool. In Figur 6, a lin from ot to ot or an ara from panl to panl is a finit lmnt. Hraftr, lmnt in Japan will b rfrr to as finit lmnt in this papr in omparison of `lmnt in th US, baus th gomtry of lmnts an th insption ata strutur ar analogous to thos of finit lmnt analyss. Figur 7 illustrats th ata roring strutur, in whih: Iniviual amag ratings for 13 ft atgoris at maximum for ah finit lmnt. For xampl,

6 A finit lmnt of a stl bam has amag apparan ratings for #1 Corrosion, #2 Craking, #3 Loosnss / Falling, #4 Ruptur, an #5 Dtrioration of orrosion-proofing funtion, rsptivly A finit-lmnt of a onrt bam has amag onition ratings for #6 Craking, #7 Pling an xposur of rinforing bars, #8 Lakag an fr lim, #9 Falling out of pla, #10 Damag onrt rinformnt, an #12 Lifting, rsptivly. Damag onitions ar from a bing no amag an bing th worst. Evn th xistn of no amag shall b ror. Furthrmor, in rlation to th amag atgory #6, raking, rak pattrs ar also atgoriz as also shown in Figur 7. As also shown in Figur 7, whn hoosing a span an an lmnt atgory, you an s th layrs of finit lmnt amag rating maps as many as spifi ft atgoris --- big ata prossing frinly. Data objtivity is thought of ruial an a rfrn manual is publish by NILIM, MLIT, to kp th ata objtivity, showing sampl photographs of ah amag atgory of ah amag rating. Insptors ar rqust to ror th xistn an xtnt of amag as prisly as possibl in a igital mannr, a bing no amag an bing th worst. Thy hav to assign th amag apparan rating of a to sort of automatially with no subjtiv translation, omparing sampl photographs an som numri ritria lik rak with on th rfrn with what thy s an masur on sit. It is also worth noting that MLIT, in prati, has awar th maintnan urgny rating insption an amag onition rating insption, sparatly. Both ratings ar ross-hk by brig aministrators as sort of a quality assuran systm. Baus ata objtivity is sur an th istributions, atgoris, an ratings of amag at finit lmnts ar igitiz an lustr, MLIT s finit-lmnt lvl insption is xpt to ovr all aspts list in Tabl 1. Big Data Mining on th MLIT s Finit Elmnt Lvl Brig Insption Databas To xamin th fftivnss of th sign for MLIT s finit lmnt-lvl insption protool, som highlight finings pross from a big ata of MLIT s finit lmnt lvl brig insption ar shown blow. Sin 2003 th finit-lmnt lvl brig insption has bn implmnt for as many as 20,000 brigs unr th jurisition of th national govrnmnt. Most brigs hav bn inspt twi following th sam finit lmnt lvl insption protool an som brigs hav bn inspt thr tims. Damag apparan rating ata obtain in FY ar mainly us blow. Figurs 8 an 9 ar xampls of showing th potntial of th promotion of ata-rivn prvntiv maintnan using th finit lmnt lvl insption. Figur 8 shows th ratio of th numbr of spans with any amag at signat finit lmnts to th numbr of all finit lmnts inspt in trms of stl I-bam brigs. Figur 8 als with orrosion of I-bam. Baus of th finit lmnt insption, a tnny in

7 th istribution of amag xtnt in a girr or a span an b unrstoo. Span-ns an outsi girrs ar susptibl to amag ompar with span-ntrs an innr girrs. This may b attribut to th watr that oms through xpansion joints an stay aroun girr-ns an th supply of hloris from th sa brought by win or iing salt from th roa surfa to outsi girrs. Baus finit-lmnt ata is likly to show th iffrn in th istribution of istrss in a strutural mmbr, w an xpt to grasp th ns of nw an bttr prvntiv maintnan thniqus. For xampl, bas on suh finings, NLIM an MLIT highway offis hav propos a zon painting manual. Figur 9 shows a stohasti transition in orrosion of stl girrs. Using two baths of finit lmnt ata that ovrs 10 yars, th hang in amag onition rating for orrosion of th sam finit lmnts is ount an th transition probabilitis ar alulat from a to a, b,,,, rsptivly, from b to b,,,, rsptivly, from to,,, rsptivly, an so on, rsulting in a stohasti tim volution in orrosion of brig girrs with yars. In this alulation, finit lmnts that wr appli to som rmial work in th past suh as th rfurbishmnt of surfa oating systm wr isar. Diffrnt trioration tnnis within a span an within a girr ar foun out, whih shoul gt involv in brig managmnt systms to limit th xtnt of ovrstimation or unrstimation of trioration in stimating th futur total maintnan osts for brigs. Figur 10 is an xampl of showing th potntial of ata-rivn improvmnt of sign spifiations using finit lmnt-lvl brig insption. Figur 10(a) sribs th numbrs of spans with spifi rak pattrns in post-tnsion PC T-bams. Crak pattrns ar atgoriz into twnty iffrnt intitis in insption as illustrat in Figur 10(), whr only major pattrns ar shown in th illustration. Th rsult shown in th lft-han si of Figur 10(a) is bas on all insption rsults (3,874 spans in total), whil th right-han si of Figur 10(a) is only bas on initial insption rsults out of all insptions (136 spans in total), whr in MLIT nw brigs ar inspt within two yars aftr putting in srvi using th finit lmnt lvl brig insption protool. Th ata larly shows that raks volv with yars. Craks along PC abls suh as pattrns #2 an #20 or along stirrups suh as pattrn #10 may iniat th orrosion of abls an rinformnt. Strutural tails of PC tnons an ovr pths hav bn hang for almost th last twnty yars an w woul lik to follow th insption rsults to figur out th fftivnss of suh hangs. Figur 11(b) shows th omparison in rak pattrns btwn post-tnsion an pr-tnsion PC bams. Craks of Pattrns #2 an #4 spially appar mor in post-tnsion girrs. Mor stringnt onstrution quality ontrols an b rquir for post-tnsion bams. Manwhil, th xistn of raks with pattrn #1 iniats that som flaw oul xist in sign prati. For xampl, th prsnt sign stanars in Japan o not inorporat rsiual strsss aumulat uring onstrution suh as thrmal strsss uring onrt asting into strss alulations, whil usually rsiual strsss u to wling ar involv in stting th strngth urv of stl bam an olumns. Long-trm loas suh as shrinkag an rp of onrt an th rlat rstrain

8 strsss u to th xistn of rinformnt bars may b nssary to b xamin. Craks an giv avrs ffts on th long-trm urability of onrt struturs an violat th prsumptions / thoris to alulat strsss in ross-stions an strngths of prstrss onrt bams. Figur 11 is an xampl of th potntial of vloping a mor logial insption protool by onuting a big ata analysis on th finit lmnt lvl brig insption ata. Figur 11 shows th numbr of stl I-bam spans an box-bam spans with or without any rak by ag an by avrag aily larg vhils. Crak tns to appar ithr whn a brig is olr than yars or whn arrying mor than 10,000 larg vhils a ay. Th rsult iniats that furthr analyss may larify th ns for th introution of spial insption programs by ag an typ of istrss. In onlusion, th rsults highlight abov show th fftivnss of finit lmnt lvl brig insption to aim at all ahivmnts list in Tabl 1. Conluing Rmarks W ar sing th avnt of th us of big ata in brig insption. Both Japan an th US just hav hang laws in trms of brig insption an aopt finit lmnt or lmnt lvl insption. Both ountris ar now onsir to fa a signifiant hallng to us th big ata of brig insption wisly. To onut ata mining, ata olltion protool shoul b arfully sign to mak th ata strutur harmoniz with big-ata mining. Th prsnt papr shows that th MLIT s finit lmnt lvl brig insption protool is likly to o a goo job in this rgar. This papr spially points out potntials that finit-lmnt lvl amag rating an b usful to figur out sintifi fats that bakup ata-rivn prvntiv maintnan, ata-rivn thnology vlopmnt in maintnan, an ata-rivn improvmnt / vlopmnt in sign spifiations an insption stanars, with xampls. Data olltion stratgis an hang with brig aministrators / ownrs ns as summariz in Tabl 1 an thr is no guian for rlvant assssmnt units (finition of lmnt ) an ata olltion struturs. Aoringly, th authors hop to ontinu to xhang an shar with ah othr btwn Japan an th US th information on: Bnfits of tail ata olltion suh as bst pratis in th ata-rivn managmnt of iniviual brigs or th ata-rivn oorination of ntwork-lvl brig prsrvation programs. Exampls of things to b improv in sign an onstrution bas on sintifi fat finings from th ata Exampls of ata-rivn thnology vlopmnt in brig insption an maintnan (.g. Clarifying vlopmnt targts an ns for non-strutiv tsting tools) Rfrns AASHTO (2002). Gui for Commonly Rogniz (CoR) Strutural Elmnts,

9 CORE-1-I1. AASHTO, Washington, DC. AASHTO (2013). Manual for Brig Elmnt Insption, 1 st Eition. AASHTO, Washington, DC. Ministry of Lan, Infrastrutur, Transport an Tourism (2004). Brig Insption Manual. Tabl 1. Rlationship btwn aims of prioial brig insption an Japan an US brig insption stanars Expt Ratings Insption units Expt sintifi / nginring aministrativ ahivmnts ahivmnts 1. Prsrvation of brigs Laing to maintnan or othr ations Sning to tail insption 2. Data rivn systmati invstmnt in prvntiv maintnan at ntwork lvl an brig lvl, rsptivly 3. Data rivn kaizn or ontinuous upats in sign spifiations, insption stanars, rtrofit guilins t 4. Data rivn thnology vlopmnt for maintnan Suring safty for passngrs unr an ovr brigs Bttr funing shm an prforman masurmnt in managmnt Thnology vlopmnt Maintnan urgny rating (Subjtiv) Damag apparan rating (Objtiv) Componnts in NBIS or Mmbrs in MLIT protool Elmnts in AASHTO manual or Finit lmnts in MLIT protool US NBIS MLIT AASHTO Elmntlvl Tabl 2. Brig Invntory in Japan (As of April 2013) National Exprssways (Own by MLIT an oprat by xprssway ompanis) National Highways Dsignat Stions (Own an oprat by MLIT) Numbr of roa brigs* 7,246 (1.1%) 20,763 (3.1%) Avrag 24 hour traffi numbrs** Larg vhil total 9,068 27,884 3,326 16,641 National Highways Othr Stions 1,127 30,200 (4.4%) (Own an oprat by Prfturs) 8,120 Prftural Roas 100,152 (14.7%) 568 Muniipal Roas 521,173 (76.7%) 4,941 Total 679,534 (10) *As of April 2013 ** Bas on th 2010 roa traffi nsus ata

10 9% 27% : Ovr 50 yars : Unr 50 yars Figur 1. Prntags of brigs olr than 50 yars at prsnt, in 10 yars an in 20 yars. Funtion/Strngth Objtiv ata is prfrr for prvntiv maintnan. Subjtiv assssmnt by xprin nginr is nssary Tim Figur 2. Shmati iagram of trioration urvs, showing th variation in trioration rat boms largr with inras in trioration Bfor tst At a moving yl of 200,000 Dfltion (mm) H22 No.1 (a) With pntrating rak H22 No.25 Collaps 貫通ひび割れあり at a moving yl of 20,000 (b) 貫通ひび割れなし Without pntrating rak Not ollaps until a moving yl of 200,000 Bfor tst (a) Without pntrating rak At a moving yl of 20, Numbr of whl loaing yls (thousans) (b) With pntrating rak Figur 3. Exampl of moving whl loaing tsts for fatigus in RC k slabs having similar amag apparan stats

11 (a) Joint of main girr an gusst (b) Sol plat () Cross fram Figur 4. Craks at iffrnt positions in stl suprstruturs Suprstrutur Dk slab Baring Substrutur Girr Dk slab Baring (M) Abutmnt Baring (F) Column.g. vry singl girr is subivi at th position of floor bams in a span Girr 01 Girr 02 Girr 03 Girr 04 Figur 5. Diffrn in finition of lmnt in insption btwn MLIT (Japan) an th US Girr bams Floor bams / Cross frams Lowr latrals Dk slabs Figur 6. Exampls of lmnt atgoris an finit lmnt mshs in th MLIT s finit lmnt lvl brig insption

12 Span N #0202 Catgoris of amag raking amag rating pattrns #6 raking #2 rak pattrn:#2 #0302 raking: Rating rak pattrn: #2 Catgoris of amag raking amag rating pattrns #6 raking #1 Span N raking ata b b b a b b a b Span N ata of amag rating rak pattrn:#1 Span N #0202 raking: Rating rak pattrn: #1 Catgoris of amag #7 pling an xposur of rinforing bars amag rating a lakag an fr lim pling an xposur of rinforing bars ata a a a b b a a a a b a b b a b raking ata #0302 Catgoris of amag #7 pling an xposur of rinforing bars amag rating Span N pling an xposur of rinforing bars a a a a a a a a #8 lakag an fr lim Figur 7. Data struturs in finit lmnt lvl brig insption of MLIT 10 a b 8 Prntags All Outsi Insi Outsi Insi Span ns Cntr Figur 8. Prntags of finit lmnts in trms of orrosion in stl bams of stl I-bam brigs ( a bing no amag an bing th worst)

13 Rat of xtnt of amag 10 75% 5 25% Ag(yars) All finit lmnts b a Rat of xtnt of amag 75% 5 25% Ag(yars) Span ntr finit lmnts of outsi girrs b a Rat of xtnt of amag 75% 5 25% Ag(yars) Span n finit lmnts of outsi girrs b a Rat of xtnt of amag 75% 5 25% b a Rat of xtnt of amag 75% 5 25% b a Ag(yars) Span ntr finit lmnts of insi girrs (Girrs xpt for outmost girrs) Ag(yars) Span n finit lmnts of insi girrs Figur 9. Stohasti tim volution in orrosion with yars on stl bams of stl I-bam brigs (Bas on th insption ata from FY2009 to FY2013)

14 ,874 spans #1 #2 #3 #4 #5 #10 #11 #19 #20 rak pattrn All insption rsults spans #1 #2 #3 #4 #5 #10 #11 #19 #20 rak pattrn Nw brigs inspt within 2 yars aftr putting in srvi (a) Crak pattrns in post tnsion T girrs ,452 spans #1 #2 #3 #4 #5 #10 #11 #19 #20 rak pattrn Post tnsion PC girrs ,291 spans #1 #2 #3 #4 #5 #10 #11 #19 #20 rak pattrn Pr tnsion PC girrs (b) Comparison in rak pattrns btwn post tnsion girrs an pr tnsion girrs #4 #1 #11 #19 #10 #2 #3 #20 #5 () Crak pattrns in MLIT s insption manual Figur 10. Numbr of spans with on or mor finit lmnts having iniviual pattrns of rak rgaring PC bams. Figur 11. Numbr of spans with rak tt in stl I-bams an box-bams by ag an avrag aily larg vhils