Jtam-A4.dvi JOURNAL OF THEORETICAL AND APPLIED MECHANICS 50, 3, pp. 871-883, Warsaw 2012 50th Anniversary of JTAM SAFETY OF BRIDGES AND BREAKWATERS ESTIMATED THROUGH GENERAL PRINCIPLES FOR LIMIT STATE DESIGN BY JAPAN SOCIETY OF CIVIL ENGINEERS Eiichi Watanabe Kyoto University (Professor Emeritus) and Regional Planning Institute of Osaka (Chairperson), Osaka, Japan e-mail: watanabe@rpi.or.jp Although the specifications for Japanese highwaybridges are currently under revision in the format of performance-based design, the degree of safety of the designed bridges is not easy to understand. Thus, this paper attempts to disclose the safety and the underlying reliabi- lity indices of roadway bridges designed in accordance with the Japanese Specifications for Highway Bridges in the light of a proposal of general principles for the limit state design. The reliability indices and the probability of failure are numerically evaluated assuming the Normal distributions both for the resistance and for the load effect based on the Advanced First-order Second-MomentMethod focusing on the relationship between the probability of failure and the factor of safety.Further special reference ismade regarding tsunami breakwa- ters in view of a low frequency but high consequence damages by the Great Eastern Japan Earthquake ofMarch 11, 2011. It was found from the comparison that great discrepancy of the design philosophy and thus the factor of safety exists between the designs of bridges and breakwaters. Key words: factor of safety, probability of failure, reliability index 1. Introduction The Great Eastern Japan Earthquake (Tohoku - Pacific Ocean Earthquake) occurred on March 11, 2011 inflicting victims over 20 thousand dead and missing inclusive. The people realized the terror of a low frequency but high consequence disaster and the inducedmelt-down of nuclear reactors in Fukushima Prefecture afresh. At the same time, most of the Japanese structural engineers must have strongly felt the need for public disclosure of the safety of in- frastructures such as bridges, buildings, dikes and tsunami breakwaters that they have built so far. It is really a pity to know that few Japanese engineers understand in reality to what degree bridges, dikes and breakwaters that they have designed and built are safe although they tend to take for granted thatmost of these infrastructures have been performing quite properly. This anxiety of engineers may be due to so many unknown factors in the design and construction in view of the fact that the Japanese Specifications for Highway Bridges are still in the process toward the complete revision in the format of the performance-based design (JRA 2002). At the present time, bridges are designed to satisfy either the performance requirement or the specifications. The performance requirement is to satisfy the target performance; whereas the specification requirement is to classify the types and kinds of the structural components so as to satisfy each of the specifications accordingly. However, in the case of the specification requirement, it is unfortunate to say that the target performance is not clearly described. Not only the structural performance but also the safety of bridges against the external loads is not perfectly clear either. From this reason, a subcommittee on the common design code on steel and concrete structures was organized by the Japan Society of Civil Engineers, JSCE (Subcom- mittee, 1992) tomake clear the concept for the limit state design of steel and concrete structures 872 E.Watanabe following a design guide for steel structures (JSCE 1987). The present paper is based on the Le- vel I load-factored design method, Level II reliability design method (Melchers, 1987; Stahlbau Handbuch, 1982; Thoft-Christensen and Murotsu, 1986) and the report by this JSCE subcom- mittee in an attempt to findout in general howquantitatively safe the designed bridges could be when designed accordingly. The required performance of designed bridges can be summarized in the following: (1) Reasonable safety in the ultimate limit state (2) Reasonable serviceability and functionality in the serviceability limit state (3) Reasonable fatigue strength (4) Easy inspection andmonitoring/Reasonable maintenance, management and repair (5) Environmentally friendly and aesthetically good. 2. Relability index and probability of failure of bridges What is the safety? One example of the answers to this question may be the “relative freedom from danger, risk, or threat of harm, injury, or loss to personnel and/or property, whether caused deliberately or by accident” (www.businessdictionary.com/definition/safety.html). Then what would be the appropriate index to measure the safety? It may depend on (1) to what degree the safety should be raised and (2) how to improve it? To the first question, although it may at first look desirable to eliminate every possibility of dangerous states and raise the safety to the highest degree, the cost may become astronomically high. Furthermore, people may feel being somehow deprived of their freedom to accomplish the perfect safety. Thus, how to get rid of danger while keeping and enjoying social human life seems to be the problem. Thus, in the human history, people have taken long time to acquire experience to make sure of the safety each time newly developed technology is born. There may be a wide variety of dangers or failures in bridges. Figures 1a and 1b show sym- bolically examples of a fail-safe parallel bridge system and a weakest-link series bridge system (Thoft-Christensen andMurotsu, 1986). People feel at easewith the fail-safe system rather than with the weakest-link system because of the inherent redundancy of the former. It must be ad- ded, however, that even the fail-safe systemmay fail just like theweakest-link systemdue to the aging and deteriorations caused by corrosion, fatigue, excessive natural or artificial forces or by accidents. Once in awhile, itmay fail progressively in a domino fashion due to accelerated dete- rioration of performance. Furthermore, there are also great varieties of basicmaterial properties at the ultimate state. Figure 2a shows the ductile fracture model undergoing plastic flow and Fig. 2b the brittle fracturemodel for example, such as of cable fracture (Thoft-Christensen and Murotsu, 1986). Naturally, the degree of failure may differ considerably depending on whether the material is ductile or brittle. In many cases, the brittle fracture may result in more unfa- vorable results. Once in a while, local fracture of structures or structural elements may develop into more serious global fracture. Fig. 1. (a) Fail-safe bridge system and (b) weakest-link bridge system Safety of bridges and breakwaters estimated... 873 Fig. 2. (a) Ductile fracture and (b) brittle fracture models Figure 3 shows a list of phenomena resulting in the failure of bridges (Tanaka et al., 2010). Large natural forces include such as typhoons, tornados, earthquakes and tsunamis. The causes of structural defects include such as corrosion, fatigue, buckling, plastic deformation, creep, relaxation and erosion or scouring of bridge piers. Fig. 3. Phenomena resulting in the failure of bridges The reliability design methods of structures may be basically classified into 3 groups of Le- vel I, Level II and Level III (Melchers, 1987; Stahlbau Handbuch, 1982; Thoft-Christensen and Murotsu, 1986). Figures 4a and 4b summarize a simple comparison of Level II and Level III re- liability designmethods, respectively. In this paper, the reliability indices and the corresponding probabilities are evaluated by Level II reliability designmethod so that the results are reflected in the more realistic and practical method of Level I: Load factored design. The main interest is the evaluation of values of the reliability indices and the corresponding probabilities of failure when the values of the factor of safety, load factor and the resistant factor are varied. The Level Imethod as shown in Fig. 5 is used for practical design purposes through calibra- tions by either Level II or Level III design methods to confirmwhether the factor of safety and other safety factors such as the load effect factor γF and resistance factors γm namely, material resistance factor and structure factor are appropriate. In this paper, the Level II designmethod is used to evaluate the probability of failure for the sake of simplicity. Naturally, for the safety of structures, the load effect F should not be greater than the resistance f. In reality, both the load effect F and resistance f have stochastic variations. The stochastic value of load effect F can be confirmed only by observations whereas that of f can be confirmed only by experiments. Furthermore, these stochastic values of F and f are in general described in terms of the expected values, Fm and fm, respectively. The central factor of safety νc is then defined by νc = fm/Fm. The characteristic value for the load effect F can be usually chosen to be Fm. Whereas for the characteristic value of resistance f, some kind of a guaranteed lower limit is naturally expected to be specified to reduce the danger of 874 E.Watanabe Pf = ∞ ∫ −∞ fF(η) dη η ∫ −∞ ff(ξ) dξ= ∞ ∫ −∞ fF(η)Ff(η) dη¬Pallowable Ff(η) = η ∫ −∞ ff(ξ) dξ – probability distribution function of resistance Fig. 4. Comparison between the reliability designmethods of Level II and Level III; (a) Level II (normal distribution is assumed on principle), (b) Level III (Pf is evaluated by exact multiple integrals), fF(η) – probability density function of load effect, ff(ξ) – probability density function of resistance Fig. 5. Reliability designmethod Level I. Load factor γF andmaterial resistance factor γm deficient performance and consequently can be often expressed in the casewhere the distribution of the resistance f follows the Normal distribution by fk = fm(1− kVf), where k = 1.64 with the probability of non-exceedance of 5% as shown in this figure where σf refers to the standard deviation of resistance f and Vf(= σf/fm) refers to its coefficient of variation. The design load effect Fd and the design material resistance fd can be determined respectively by Fd = FmγF and fd = fk/γm, where γF and γm are referred to as the load effect factor and the resistance factor, respectively. Then the factor of safety ν is defined by ν = fd/Fd (ν can Safety of bridges and breakwaters estimated... 875 have many values but the value of 1.7 has been used traditionally for bridges) (JSCE 1987). Consequently, the relationship between ν and the central factor of safety νc can be shown by ν = fd/Fd = (fk/γm)/(FmγF) = νc(1−kVf)/(γmγF). The load effect factor and the resistance factor are specified in Table 1 and Table 2, respectively (Subcommittee, 1992). Table 1. Load effect factor γF Limit state Kinds of load effect Load effect factor γF Permanent load 1.0∼ 1.2 or 1.0∼ 0.8 (when smaller value becomes more unfavorable) Ultimate limit Principal variable load 1.1∼ 1.2 Secondary variable load 1.0 Accidental load 1.0 Serviceability limit All loads 1.0 Fatigue limit All loads 1.0 Table 2.Resistance factor γm Kind of material Limit state Resistance factor γm Ultimate limit 1.0∼ 1.05 Steel (m= s) Serviceability limit γs 1.0 Fatigue limit 1.0∼ 1.05 Ultimate limit 1.3 Concrete (m= c) Serviceability limit γc 1.0 Fatigue limit 1.3 According to the Advanced First-Order Second-Moment Method (AFOSM) in the Level II reliability design scheme, assuming the normal distributions for both of the load effect and the resistance, the reliability index β and the probability of failure Pf can be generally expressed by Eq. (2.1) in general or in the case of a low frequency, but a high consequence load effect β= Zm σZ = fm−Fm √ σ2F +σ 2 f = νc−1 √ V 2F +ν 2 cV 2 f Pf =Φ(−β)= 1−Φ(β) (2.1) In the latter case, VF =0and the histogram of the load effect of F in Fig. 5may be represented by a limited number of isolated scattered spectra of very narrow-banded step functions. In the case of a low frequency but a high consequence load effect β= νc−1 νcVf (2.2) where β, Φ(β), σF and VF(= σF/Fm) are referred to as the reliability index, the normalized normal distribution function of β, the standard deviation and the coefficient of variation of the load effect F, respectively. Theprobability of failure and the reliability index are shown inFig. 6 andFig. 7, respectively for different values of the factor of safety ν in two cases of VF =0.1; Vf =0.05 and VF =0.2; 876 E.Watanabe Vf =0.1, respectively. For distributions other than the Normal, references are recommended to be quoted (Thoft-Christensen and Baker, 1982; Thoft-Christensen andMurotsu, 1986). Fig. 6. Probability of failure for different values of the factor of safety ν, when γmγF =1.0 Fig. 7. Reliability Index for different values of the factor of safety ν, when γmγF =1.0 3. Determination of structural performance D. Frangopol andP. Thoft-Christensen showed how the performance of bridgesmay be reduced by aging on the basis of their research works on concrete bridges in UK studies (Frangopol and Das, 1999; Thoft-Christensen, 1999). Figures 8 and 9 show the general decrement of the value of reliability indices or the performance due to the deteriorations of concrete bridge performances. Particularly, the aging shown in Fig. 9 is really surprisingly large. They concluded in the following manner: (1) Reliability index β ranges from 5 to 12 at the initial stage when bridges are newly built (in the case of normal distribution, Pf =3.0 ·10 −7 at β=5 and Pf =0.0 at β=12) (2) Deterioration process starts at a time point tI after the service of bridges started (Figs. 8a and 9) (3) Essential maintenancemust be carried out if β decreases below the target value of 4.6 (in the case of normal distribution, Pf =2.2 ·10 −6) at a certain time point tII (4) Preventivemaintenance is recommended to be carried out economically well in advance at the time point such as at tIII or tIV (Fig. 8b) so that the essential maintenance may not have to be carried out at the time point tII. Safety of bridges and breakwaters estimated... 877 Fig. 8. Deterioration of performance of a concrete bridge; (a) simplified figure (Frangopol and Das, 1999), (b) essential and preventivemaintenance (Thoft-Christensen, 1999) Fig. 9. Time-dependent deterioration process of the reliability index. Note the expected value and stochastic variations – concrete bridge under shear (Frangopol and Das, 1999) 4. Some considerations on the safety of tsunami breakwaters OnMarch11, 2011, a great tragedywas inflictedon theNorthEasternJapan.Thedisaster is now officially named the Great Eastern Japan Earthquake. According to the official announcement, themagnitude of the earthquake wasMomentMagnitude (MM) 9.0, which is an unprecedented magnitude in Japan. The investigation party was dispatched by JSCE and headed by prof. H. Maruyama (http://committees.jsce.or.jp/report/node/39) (in Japanese). He reported that the main initial cause of the failure of such breakwaters might be the overflow of tsunami over the top of breakwaters. Thus, the overflowed stream could be the initial cause of the scouring of the breakwaters and their turnover. Furthermore, prof. T. Hara of Kohchi University explained the scouring might have occurred in such a way that the overflowed stream turned around just behind the breakwater and washed out the ground. The breakwater becomes unstable due to this scouring,with the buoyancy force acting on the breakwater and the risenwater level pushed it from the ocean side to inside resulting in an easier turnover of the breakwater. Figure 10 shows the estimated tsunami travel time forecast map for the 2011 Great Eastern Japan(hereafter referred toasGEJ for convenience)Earthquake fromtheU.S.NOAA.Moreover, Figs. 11a-c show some pictures of a piled-up ship over a breakwater and turnover of breakwaters by the courtesy of Mr. I. Teranishi who surveyed damages at several places badly failed during this earthquake as a member of the Japan Bridge Association. Figure 12 classifies the failures of breakwaters by the Asia-Pacific Center for Coastal Di- saster Research of the Port and Airport Research Institute (www.bousai.go.jp/jishin/chubou/ higashinihon/4/2-2.pdf in Japanese). As already mentioned, JSCE has reported that the main initial cause of the failure of such breakwatersmight be the overflow of tsunami over the top of breakwaters as shown in Fig. 12d: overflowed stream resulting in the scouring of the ground just behind and inside the breakwater and the subsequent turnover. After literature survey, the factor of safety of the breakwater in 878 E.Watanabe Fig. 10. Estimated tsunami travel time forecastmap for the 2011GEJ Earthquake from the U.S. NOAA (http://en.wikipedia.org/wiki/File:2011Sendai-NOAA-TravelTime-Ttvulhvpd9-06.jpg) Fig. 11. (a) Piled-up ship on bank at Kamaishi Harbor, (b) breakwaters turned over counter-clockwise at Yamada-Cho, (c) breakwaters turned over showing bottoms at Yamada-Cho (courtesy of Mr. I. Teranishi) Fig. 12. Classification of the failures of tsunami breakwaters (courtesy of Mr. S. Takahashi of the Port and Airport Research Institute) Japan was found to be only 1.0 and that of the walls against the sliding is only 1.2. These low factors of safety are verymuch surprising to us, structural engineers who are accustomed to the factor of safety as high as about 1.7 (JSCE 1987). It has been found that one of the largest tsunamis in Japan in the past with the resemblance to the 11March 2011 tsunamiwould be the Safety of bridges and breakwaters estimated... 879 one which occurred at Johgan Earthquake of 869 A.D. Thus, the return period of this size of tsunamimay be considered to be once in every 1000 years. Keeping this inmind, the probability of overflow, namely, the probability tsunami exceeding the height of breakwater is evaluated for different values of the factor of safety. Tsunami hazard curves are often used to designate the relationship between the probability of exceedance of occurrence and the height of tsunami. Needless to say, these curves generally depend on many factors such as the topography of the surrounding sea, locations of active faults, epicenter and magnitude of earthquake, starting point of asperity and non-uniformity of the faults. For example, Annaka et al. developed the logic-tree models for local tsunami sources around Japan and for distant tsunami sources along the South American subduction zones. Logic-trees were made for tsunami source zones, size and frequency of tsunamigenic earthquakes, faultmodels, and standarderror of estimated tsunamiheights (Annaka et al., 2007). In this paper, however, since the author is not an expert in this area, the discussions are made presumptuously in a very simple manner on the assumption that tsunami can occur anywhere near the sea zones potentially susceptible to tsunami. A simplified relationship between the expected tsunami height H and the return period of tsunami T may be expressed satisfactorily quitewell byEq. (4.1) andFig. 13b as anapproximation of thehazard curve,Fig. 13abyYoshida et al. (2007) H H0 = logT logT0 = log(1/T) log(1/T0) (4.1) where T0 corresponds to the return period of themaximum tsunami height H0. From evidences from investigations such as geological excavations, the JohganEarthquake is thought to be very close to March 11, 2011 GEJ Earthquake. The Johgan tsunami occurred in AD 869 with the earthquake of an estimated magnitude of M8.3 or more. Thus, it may be conveniently assumed that T0 =800 years and H0 =30m (max tsunami height recorded onMarch 11, 2011) as shown in Table 3. Table 3. Prediction of tsunami height in reference to March 11, 2011 earthquake and Johgan earthquake T [Year] 800 100 50 30 10 H [m] 30 20.66 17.6 15.3 10.3 Fig. 13. Relationship between log(1/T) and tsunami height H; (a) hazard curve by Yoshida (2007), (b) hazard curve, Eq. (4.1) Since T0 refers to the return period, the probability of tsunami occurrence at least once in consecutive Q years Pr may be expressed by Eq. (4.2) Pr =1− ( 1− 1 T0 )Q (4.2) 880 E.Watanabe Thus, the probability Pr of tsunami occurrence can be given as in Fig. 14 and inTable 4. Itmay be observed that naturally, the longer the value of Q, the larger the probability Pr becomes. It would be very important to know fromTable 4 and Fig. 14 that when the period Q is close to the target return period T0, the probability of occurrence of the tsunami is something like 63% while it is about 10% when the period Q is 1/10 of the return period T0. Fig. 14. Probability Pr of tsunami occurrence such asMarch 11, 2011 at least once in consecutive Q years Table 4. Probability Pr of tsunami occurrence such as March 11, 2011 at least once in conse- cutive Q years Q [Year] 10 30 100 500 800 Pr 0.012 0.037 0.118 0.465 0.632 So far, the discussion was made on the probability of occurrence of tsunami, and this has only a little bearing on the safety of tsunamibreakwater.Now, let us next consider the reliability and safety of the tsunami breakwaters. Let d represent the height of a breakwater, then the safety margin Z may be expressed by Z = d−H ­ 0 (4.3) Again by applying the AFOSM, the reliability index β can be expressed by β= µZ σZ = µd−µH σZ = µd−µH √ σ2 d +σ2H = µd µH −1 √ σ2 d µ2 H + σ2 H µ2 H = νd−1 σH µH √ σ2 d σ2 H +1 = νd−1 VH √ σ2 d σ2 H +1 = νd−1 √ V 2 d ν2 d +V 2H (4.4) where νd =µd/µH represents the factor of safety of a breakwater; µZ, µd, µH and σZ, σd, σH represent the expected values and the standard deviations of Z, d, H, respectively. Further- more, Vd and VH designate the coefficients of variation of d and H, respectively. Now, upon differentiation of Eq. (4.1), Eq. (4.5) can be obtained dH H0 = 1 logT0 dT T (4.5) Safety of bridges and breakwaters estimated... 881 FromEq. (4.5), the relationship between σH and σT of H and T , respectively can be found as σ2H = ( H0 T logT0 )2 σ2T (4.6) Thus, the reliability index can be expressed by Eq. (4.7) evaluated at the critical design point H0 and T0 as β= νd−1 √ V 2 d ν2 d +V 2H = νd−1 √ V 2 d ν2 d + ( VT logT0 )2 (4.7) where VT represents the coefficient of variation of T . Then the reliability index β and the probability Pf of tsunami height exceeding the height of the breakwater is shown by Fig. 15 andFig. 16, respectively.Furthermore,under conditionof VT =0.2and Vd =0.1, theprobability of failure (overflow) Pf is obtained as shown in Fig. 17 when the return period of tsunami T0 and the predictedmaximum tsunami height H0 are varied. Just as for the structural reliability problems, in the case of a low frequency but high consequence occurrence of a huge tsunami, the reliability index becomes by neglecting the coefficients of variation VH and VT β= νd−1 νdVd (4.8) Fig. 15. Reliability index of breakwater Fig. 16. Probability of tsunami exceeding height of the breakwater It is surprising to know that according to the current designmethod of a breakwater against tsunami, since the factor of safety is very close to one, the probability of failure, namely tsunami exceeding the height of the breakwater is 50%. Even if the factor of safety is improved to be 1.2, the probability of failure may be something like 10%. 882 E.Watanabe Fig. 17. Change of probability of failure Pf when the return period T0 is varied (VT =0.2 and Vd =0.1) 5. Conclusions The current Japanese Specifications for highway bridges are under revision in the format of performance-based design in the genuine sense although there may be more detailed laborious effortsyet tobemadeby thewritersof theSpecifications suchas time-takingdetailed calibrations for practical purposes. Thus, attempts are made in this paper to roughly evaluate the general reliability with special emphasis on the reliability index and probability of failure of the bridge design. Special attempts are made also to evaluate the reliability of tsunami breakwaters. It is found that the current practice of the design of breakwater is such that the factor of safety of tsunami height exceeding the height of the breakwater is just one. In view of the fact that the overflow of tsunami over the top of the breakwater eventually results in the fatal catastrophe of the scouring, sliding or turning over of breakwaters, the review or reexamination of the present factor of safety of the breakwater may become worthwhile in recognition of larger factors of safety, 1.7, of bridge design employed at present in preparation of the future low frequency but high consequence tsunami in the near future. As a matter of fact, Japan is threatened with a linked series of huge earthquakes of Tokai (East Sea), Tonankai (South-East Sea) and Nankai (South Sea).These linked earthquakes in a domino reaction are predicted to occur at any moment immediately near future in the similar scale of the Great Eastern Japan Earthquake. Finally, how to be prepared for natural disasters of a low frequency but high consequences is a great challenge not only for the Japanese people but for all the people in the world, and for this purpose people of the world should work together and make effort to find the best solution for their protection. References 1. AnnakaT., SatakeK., SakakiyamaT., YanagisawaK., ShutoN., 2007, Logic-tree appro- ach for probabilistic tsunami hazard analysis and its applications to the Japanese coasts,Pure and Applied Geophysics, 164, 2/3, 577-592 2. Frangopol D.M., Das P.C., 1999,Management of bridge stocks based on future reliability and maintenance costs, [In:] Current and Future Trends in Bridge Design, Construction and Mainte- nance, Thomas Telford Publishing, London, 45-58 3. JapanRoadAssociation (JRA), 2002,Specifications for Highway Bridges, Part I, CommonDesign Principles, Maruzen, Tokyo [in Japanese] 4. Japan Society of Civil Engineers (JSCE), 1987, Design Guide for Steel Structures, Part A, Steel Structures Series 3A, Giho-do, Tokyo [in Japanese] Safety of bridges and breakwaters estimated... 883 5. Melchers R.E., 1987, Structural Reliability Analysis and Prediction, Prentice Hall, Englewood Cliffs, NJ 6. Stahlbau Handbuch, 1982,Für Studium und Praxis, Stahlbau-Verlags-GmbHKöln 7. Subcommittee on commondesign codeon steel and concrete structures, 1992,General principles for the limit state design of steel and concrete structures,Report of Committee, Proc. JSCE,450/I-20, 13-20 [in Japanese] 8. Tanaka T., Watanabe E., et al., 2010, Illustrative Science of Bridges, Kansai Chapter of JSCE, Blue Backs B1676, Kohdansha, Tokyo [in Japanese] 9. Thoft-ChristensenP., 1999,Estimation of bridge reliability distributions, [In:]Current andFu- tureTrends inBridgeDesign, Construction andMaintenance, ThomasTelfordPublishing,London, 15-25 10. Thoft-ChristensenP.,BakerM.J., 1982,Structural Reliability and its Applications, Springer- Verlag, Berlin-Heidelberg-NewYork. 11. Thoft-Christensen P., Murotsu Y., 1986,Application of Structural Systems Reliability The- ory, Springer-Verlag, Berlin 12. Yoshida I., Takeuchi T., OhmoriM., Fujii N., YanagisawaK., 2007,Amethod for estima- tion of collision hazard of tsunami floating objects, Proc. of Earthquake Engineering, JSCE, 29, 912-917 [in Japanese] Szacowanie bezpieczeństwa mostów i falochronów na podstawie ogólnych zasad przyjętych przez Japońskie Stowarzyszenie Inżynierów Lądowych dla konstrukcji narażonych na graniczny stan obciążenia Streszczenie Mimo, że zalecenia projektowedlamostówdrogowychbudowanychwJaponii sąwłaśniemodyfikowa- ne pod kątemwłaściwości eksploatacyjnych, szacowanie stopnia bezpieczeństwa takich konstrukcji wciąż pozostaje problemem nie do końca zrozumiałym jednoznacznie. W prezentowanej pracy przedstawiono próbę przybliżenia zagadnienia bezpieczeństwa i wskaźników niezawodności dla mostów projektowanych zgodnie zwytycznymiujętymiwJapońskichSpecyfikacjachDlaMostówDrogowych.Regulacje te definiu- ją ogólne zasady stosowanew konstrukcjach narażonychna eksploatacjęw granicznym stanie obciążenia. Wskaźniki niezawodności i prawdopodobieństwo uszkodzenia oszacowano numerycznie, przyjmując nor- malny rozkład wytrzymałości na obciążenie. Do obliczeń wykorzystano metodę momentów pierwszego i drugiego rzędu dla zależności pomiędzy prawdopodobieństwem zniszczenia i wskaźnikiem bezpieczeń- stwa. W dalszej części odniesiono się do problemu falochronów zabezpieczających przed tsunami jako konstrukcji narażonych na rzadkowystępujące, lecz groźnew konsekwencjach obciążenia. Dyskusję prze- prowadzono w kontekście Wielkiego Wschodnio-Japońskiego Trzęsienia Ziemi z 11 marca 2011 roku. Z analizy porównawczej jasnowynikło, że istnieją ogromne rozbieżnościw filozofii projektowaniamostów drogowych i falochronów i tym samym różna ocena wskaźników bezpieczeństwa tego typu konstrukcji. Manuscript received January 25, 2012; accepted for print February 9, 2012