1. A SAMPLING OF BRIDGE PERFORMANCE CRITERIA BY MARK YASHINSKY, CALTRANS OFFICE OF EARTHQUAKE ENGINEERING Most bridge owners have adopted design criteria that limits damage to specified levels based on the size of the earthquake and the importance of the bridge. A good source of information for bridge performance criteria is ATC-18, “Seismic Design Criteria for Bridges and Other Highway Structures: Current and Future.” However, a lot has changed since this report was written in 1997. There are three parts to performance criteria: 1. The design earthquakes and their associated hazards. 2. Different categories of bridges. 3. Different damage states.

2. Japan’s Design Earthquake All bridges are designed for three levels of earthquakes. 1. A high probability/low intensity earthquake and two very large, low probability earthquakes. 2. a 1923 Tokyo subduction zone event 3. a 1995 Kobe crustal event. For smaller, more frequent events both ordinary and important bridges are to survive without damage. For larger, low probability events, ordinary bridges must survive without collapse and important bridges must have limited damage. I) The Design Earthquake. New Zealand Design Earthquake All bridges are designed for three levels of earthquakes. 1. A design earthquake with a return period of 450 years. 2. A smaller earthquake adjusted using a Poisson Distribution. 3. A very rare earthquake adjusted using a Poisson Distribution.

3. Chapter 9 of the AREMA manual provide maps from the USGS that give the peak rock accelerations for 100 year, 475 year, and 2,400 year return period earthquakes for Canada, the United States, and for Mexico. Base acceleration coefficients (as a percentage of g) are taken from the 100 year, 475 year, and 2,400 year maps and linearly interpolated with the return periods obtained using the importance classification factor to obtain the acceleration for the three limit states. AREMA (Railway) Bridge Design Earthquake Railway bridges are designed for three earthquakes Level 1 ground motion that has a reasonable chance of being exceeded during the life of the bridge. Level 2 ground motion that has a low probably of being exceeded during the life of the bridge. Level 3 ground motion for a rare, intense earthquake. The return period for each limit state is determined by multiplying the difference in the average return period in the table above by an importance classification factor ‘I’; dividing the product by 4; and adding the result to the minimum return period. I) The Design Earthquake continued.

4. Caltrans Design Earthquake Safety Evaluation Ground Motion (Up to two methods of defining ground motions)  Deterministically assessed ground motions from the maximum earthquake as defined by the Division of Mines and Geology Open-File Report 92-1 (1992).  Probabilistically assessed ground motions with a long return period (approx. 1000-2000 years). For important bridges, both methods shall be given consideration, however the probabilistic evaluation shall be reviewed by a CALTRANS approved consensus group. For ordinary bridges, the motions shall be based only on the deterministic evaluation. In the future, the role of the two methods for these bridges shall be reviewed by a CALTRANS approved consensus group. Functional Evaluation Ground Motion: Probabilistically assessed ground motions that have a 60% probability of not being exceeded during the useful life of the bridge. A CALTRANS approved consensus group shall review the determination of this event. Although the performance criteria provide functional evaluation requirements for ordinary bridges, these structures do not require an explicit functional evaluation if they meet the safety evaluation performance criteria. In other words, Caltrans designs ordinary bridges for only one event, the deterministically assessed ground motion from the maximum earthquake.

5. Designing bridges for different return periods makes sense in regions where large earthquakes are rare. However, in California, there is only a moderate difference between the 100 year event and the 10000 year event as shown in the figure below. I) The Design Earthquake concluded.

6. II) Bridge Categories. Japan’s Bridge Categories Bridges shall be classified into two groups of importance considered in relation with road classes and bridge functions and structures: Class A bridges are of ordinary importance. Class B bridges are of high importance. New Zealand’s Bridge Categories Bridges are divided into three categories based on traffic and importance (jurisdiction). Caltrans Bridge Categories After the Northridge Earthquake (Housner, 1994), Caltrans adopted performance criteria for important and ordinary bridges, where an important bridge meets one of three criteria:  A bridge required for secondary life safety, such as providing the only access to a hospital.  A bridge formally designated as critical by a local emergency plan.  A bridge whose loss would cause a major economic impact. All other bridges are considered ordinary. AASHTO LRFD Bridge Categories Bridges are divided into three categories of importance, but these categories are not specifically addressed in design.

7. II) Bridge Categories concluded.

8. III) Damage. Japan’s Bridge Damage For smaller, more frequent events both ordinary and important bridges are to survive without damage. For larger, low probability events, ordinary bridges must survive without collapse and important bridges must have limited damage. New Zealand Bridge Damage 1. After a design event, the bridge should remain usable for emergency traffic, although some repairs may be needed. Moreover, the bridge should be repairable to its initial condition. 2. After an event with a return period significantly smaller than the design value, damage should be minor, and without disrupting traffic. 3. For an event with a very large return period, the bridge should not collapse. Moreover, it should be usable to emergency traffic after temporary repairs and it should be capable of being brought back into service, perhaps at a lower level of service. AREMA (Railway) Bridge Damage The serviceability limit state provides for train safety after a moderate event. The ultimate limit state provides structural integrity after a large event. The survivability limit state prevents bridge collapse for intense events.

9. III) Damage continued.

10. The FHWA Retrofit Manual uses damage criteria developed by Mander and Basoz in their “Seismic Fragility Curve Theory for Highway Bridges” to prioritize bridges for retrofit. It uses NBI fields to determine the bridge fragility. III) Damage continued.

12. III) Damage concluded.

13. IV) CONCLUSIONS. 1. Higher performance criteria (above ‘no collapse’) is exceedingly difficult to achieve. When ductility is limited, displacements are reduced and accelerations are greatly increased as shown in the figure below.

14. IV) CONCLUSIONS. For instance, the Southern Freeway Viaduct was retrofit to have a ductility demand of 4.0, but due to weak soil and large ground motion, the retrofit ended up costing 1.3 times the replacement cost of the structure.

15. IV) CONCLUSIONS. Originally, BART engineers were hoping to keep their trains running after a major earthquake. However, they soon realized that anything beyond Life Safety had an unacceptable benefit to cost ratio.

16. IV) CONCLUSIONS. The Performance Criteria for all the new, important bridges, such as the East Bay Crossing, the new Benecia Martinez Bridge, and the I-880 replacement was to simply reduce the ductility demand to 3.0.