Despite a 6 hour delay in departing San Francisco (and thereby arriving at the hotel in Tiachung at about 2 AM), we had a good first day investigating earthquake damages to the north east of the city. In the morning, we met with the ASCE lifelines reconnaissance group and some southern California structural engineers who were also staying at the Tiachung Howard Prince Hotel. On our way through the city to our first damage sites, we met with Prof. Kawashima (Tokyo Institute of Technology) to get updated information on bridge damage locations. He provided some useful maps and new descriptions of damage prepared by the Japanese investigation team.		  
   
 
  Tiachung and its suburbs constitute a large metropolitan area in a broad valley between coastal and inland mountain ranges. During our excursion today through Tiachung, we see very little damage. There are many modern high rise reinforced concrete structures.
 
   
 
  We see a 6,18 and 50 story welded steel frame buildings under construction. All of these utilize box columns. None appear to be damaged during our superficial drive by viewing.
 
   
 
  The following discussion is subdivided by observations made at various Stations recorded in our GPS log. During the first day, our objective was to visit cities to the northeast of Tiachung, near the northern edge of the fault rupture.  
   
 
   
 
  On our trip to the northeast out of Fengyuan on Highway 3, the terrain began to rise into a mountain valley along the Dajia River. Surprisingly, we did not notice any damage until we came upon a large ground rupture. In this case, a nearly vertical offset of about 3 meters was observed. Buildings located on the rupture were either destroyed or severely tilted. Buildings on either side of the rupture, had virtually no structural damage. In some cases, because of the slope of the terrain (and the cobbly nature of the soil in the valley near the river) buildings slightly away from the fault tilted due to moderate to minor spreading and land slides.  
 
 
  The form of construction in this area consisted of narrow two story buildings of various types: reinforced concrete stores with apartments over them, three or four story reinforced concrete apartment buildings and various buildings constructed from light metal framing, unreinforced masonry, mixed wood-masonry-etc. Occasional light industrial buildings were seen in this area.
 
 
 
  Nearly all buildings in this region that were not affected by the fault/ground movement escaped any significant structural damage. Other immediately visible forms of damage were toppling of unreinforced concrete and masonry fences, and tilting of retaining walls.  
 
 
  We continued to follow the rupture until we came to the river. At this point, the faulting continued under a multi-span reinforced concrete bridge. The bridge consisted of a series of simply supported spans; the deck was supported on prestressed concrete I-girders resting on single column hammerhead bents. The bridge girders appear to be supported on elastomeric pads. No additional restraint is provided for girder movement in the longitudinal direction of the deck. Two small keys are provided to provide restraint in the transverse direction.
 
 
 
  The rupture passed under the bridge at a skew. In viewing the photos, it is important to realize that this bridge was originally straight and level. The three spans closest to the viewer in the photos fell off their supports. The fault crosses the bridge at a skew angle such that the bridge tends to shorten. The bridge appears to have moved toward the viewer in the photos (i.e., towards the south), resulting in the northern bridge spans shoving the southern ones off their supports. In addition, the second pier from the southern abutment appears to have been founded within the fault rupture zone and collapsed.
 
 
 
  The bridge deck pushed through the southern abutment by about 2 meters. The near abutment and first pier have raised and shifted to the left.
 
 
 
  The waterfall in the river is an artifact of the fault rupture. The height of the waterfall was estimated by local observers to be about 9-10 meters, consistent with the relative movement of the abutment. The southern end of the bridge appears to have shifted about 3 meters to the west relative the northern end. The roadway that can be seen in the photos at the far end of the bridge was reportedly straight prior to the earthquake, suggesting another fault rupture in that area as well. As it turns out, this whole area is crisscrossed by a number of ruptures, each raising anywhere from a few centimeters to many meters.  
 
 
 
 
  Continuing up the valley a short distance we arrived at a small rail station in Shrgang.  
 
 
  Another rupture scarp passes through this area, as can be seen. The rail tracks and the retaining wall in foreground of the photo can be seen to have raised about 4 to 5 meters in the foreground (to the east) and shifted 2.5 meters to the right (north).
 
 
 
  This rupture trace progressed through the adjacent town resulting in the destruction and tilting of many buildings
 
 
 
  While the building in the center of the last photo is severely tilted, the buildings originally to its left and right (as well as numerous other buildings in the area) were destroyed and construction crews have already removed many of them.
 
 
 
  In this area, several local buildings away from the fault (on the hanging wall side) were seen to be damaged. In some cases, these were older buildings. Infill panels (masonry) had begun to work within the surrounding frames or, in some cases, had fallen out.  
 
 
  A couple of small steel frame buildings were noticed in this town. One was a frame with end plated connections. The connection shown was exposed due to the collapse of the adjacent building. The connection is seen to have worked with localized yielding. The member sizes are quite small, and this two story building had numerous nonstructural features that also helped provide lateral resistance and damping. This building was immediately across the street from the rupture trace.
 
 
 
  Another nearby steel building that was on top of the trace collapsed. The main reason for the collapse was likely the large differential displacements developed under the building, rather than any particular structural feature of the building.
 
 
 
  However, the quality of the construction can be assessed in these two additional photos. In general, the welds appear to be fillets, and of very poor quality. A number of these types of frames are subsequently noted throughout the area.
 
 
 
 
 
  We traveled a short way upstream to another long bridge crossing the river
 
 
 
  In this case, the bridge again appeared to shift as a whole towards the south, with the various spans eventually pushing the last two bays on the south end off their supports
 
 
 
  At the far south end, there was evidence of the bridge shoving past the abutment (an indication of the amount of permanent displacement at the abutment region is shown). No evidence of fault rupture through the bridge was immediately apparent. There is evidence of compression pounding between adjacent spans of the bridge deck on the north end of the bridge.
 
 
 
  The military has constructed a temporary bypass so traffic can get around the collapsed portion of the bridge and back on the remaining portion of the bridge. The bridges along this river appear to be constructed similarly, with the precast I-girders supported on elastomeric pads or sliders. Substantial displacement is noted in the longitudinal direction of the roadway in many of these bridges. The keys used to restrain transverse displacement suffered varying amounts of damage depending on the bridge.  
 
 
 
 
  A dam used to supply drinking water was located between Stations 2 and 4. From a distance, there is damage on the right (north) side of the dam. On closer inspection, it appears that a rupture extends through this portion of the dam, with the south end ending up approximately 9-10 meters higher than the north end.
 
 
 
  The rupture traces in this area appear to be multiple, and have complex orientations. Several buildings in the vicinity are severely damaged or collapse.  
 
 
 
 
  Following the river further upstream, the valley narrows. The bridge at Station 6 is similar in construction to previous bridges inspected down stream, but it is considerably shorter, and constructed at a significant skew. There is evidence of several small ruptures in the rock down stream of the bridge that may extend through the bridge.
 
 
 
  One of the bridge piers shows evidence of inclined shear cracking (consistent with loading in the upstream (westerly) direction). Again the bridge spans appear to have shifted to the south, pushing the southernmost spans off their supports and damaging the southern abutments. The southern abutment appears to be higher than suggested by the alignment of the remaining bridge, consistent with the south end of the bridge rising and moving towards the north.
 
 
 
 
 
  Moving upstream further we came to Dungshr. This is a city of somewhere between 45,000 and 75,000 people. Reportedly, 710 people died in the city.  
 
 
  There is greater damage to buildings as a result of shaking than apparent in any of the previously visited areas. There appear to be numerous examples of weak first story behavior, including many collapses
 
 
 
  This appears to be attributable in part to the practice in reinforced concrete buildings of making architectural partitions and cladding out of lightly reinforced concrete panels, integrally cast with the underlying moment-resisting frame construction
 
 
 
  The ground floors in most buildings in cities are predominantly used for commercial purposes. These floors do not have as many walls as the residential or office structures above. As a result, the first stories are often (substantially) weaker than the higher stories. In many reinforced concrete buildings, lightly reinforced concrete or unreinforced masonry (brick) walls are provided perpendicular to the street providing partitions between stores. The frames parallel to the street are often open to provide access at the front (and sometimes) back of the store. This difference in the orientation of these stiff nonstructural components may have a significant influence on the distribution of damage in many areas.  
 
 
  The older, central part of Dungshr was seen to be heavily damaged. Many older forms of construction were observed including structures of mixed construction, even from adobe constructions. Many buildings in this area were observed to be demolished and removed. Many buildings had large piles of brick and concrete rubble in front of them, suggesting considerable nonstructural damage to interior partitions.
 
 
 
  An interesting pair of modern 12 story buildings was seen in Dungshr. These buildings were part of the same mixed commercial residential complex. One building suffered significant damage to the nonstructural concrete and masonry walls over the bottom four or five stories and the other suffered a failed ground story level. The building that did not collapse had commercial stores in the lower three levels.
 
 
 
  These stores were small, resulting in numerous masonry and lightly reinforced concrete walls throughout the lower levels of the building. In the other building, the lower level was kept open as an open air lobby and playground. The columns in this level suffered severe damage (associated with the soft story mechanism)and the few nonstructural concrete walls provided at this level failed in a brittle manner.
 
 
 
  Detailing of these buildings appears typical of buildings seen in the area. Column transverse reinforcement consists of ties with 90 degree hooks, crossties are non-existent or few. In these buildings, spacing of the ties is relatively close, but in many buildings in the area spacing appears to be about the minimum column width dimension. Longitudinal reinforcement in the columns is spliced at the base of the column, immediately above the floor level. This results in distress in this area in this and other buildings. Beam to column joints have limited transverse reinforcement. These details appear to have a significant influence on the details of the local damage observed.  
 
 
  A few other interesting structures were noted. One of these was a simple, single bay, welded steel moment frame building. This structure yielded during the earthquake, with a significant permanent residual displacement resulting.
 
 
 
 
 
  A long and wide bridge across the Dajia River at Dungshr suffered some interesting damage. Three parallel bridge structures have been constructed, each generally similar to those described down stream. The bridge has 17 bays, none of these collapsed.
 
 
 
  All three bridges suffered significant damage to the keys intended to limit transverse movement. Some of the supporting piers suffered damage. Two newer bridges were added, one immediately on either side of the original bridge. All of them shifted significantly upstream (because of a twist in the river, this motion is to the southeast).
 
 
 
  As a result of this shift, the outer I-girder on the upstream edge of the bridge is about to fall off the end of the bent cap. Note the spalling at the end of the cap beam shown and the fact that the original bearing for the girder has fallen out of place. To repair this damage, heavy shims have been placed under most girders to improve support on the bent caps (some of these shims can be seen in the previous photo). In addition, heavy shoring is provided for several bent caps, as well as at the ends of several girders. A variety of lattice truss, and chevron-braced frames are used for this shoring. These are supported on new concrete foundations. In some of these cases, it appears that the bridge piers may have been damaged by the earthquake. Several columns have been encased in large concrete blocks. These shorten the length of the columns so their behavior during another earthquake may be questioned.
 
 
 
 
 
  Nightfall ends the first day of investigation.  
   
  Steve Mahin