Today, we are traveling south from Tiachung, basically following route 3. This route follows along the base of the mountain range that borders the eastern edge of the valley in which Tiachung is located. One of the major north-south fault ruptures east of Tiachung surfaces along the base of these mountains. Again, observations are reported chronologically, according to the GPS station identification tags assigned to the site visited.  

 
 
We have passed by a complex of four highrise reinforced concrete buildings on Route 63 near the southern edge of Tiachung. The southeastern building in this complex has sustained a partial collapse, associated with the collapse of columns at the third or fourth level (above a change in plan). About half of the building has dropped one story vertically.  
 
 
We progress directly down Route 63 to Nantou. A large steel viaduct has been closed. This viaduct crosses a river, curves and extends for about a kilometer through part of town. The steel deck is very wide. The steel plate girders are continuous over the supports and are supported generally by steel bent caps. The steel bent caps are supported in turn by circular reinforced concrete columns. Several of these columns have suffered obvious inclined cracking, and pounding damage is apparent at the southern abutment. Since this bridge was inspected by an earlier PEER team, we press on southward.  
 
We note a few older looking buildings that appear damaged as we pass through Nantou, Mingjian and nearby cities on Route 3. But even unreinforced masonry structures appear to have faired well. The reported ground acceleration levels for this area seem inconsistent with the low level of damage in this area.  
 
 
As we cross over the wide Juoshuei River to begin our climb into he mountains, we notice that the southern end of four-lane bridge has collapsed and has been removed.
 
 
Two single column bents support the bridge. A temporary earth embankment has been run up to the side of the remaining bridge so that traffic can continue to move across the bridge. Of the bridge's original 28 bays, 7 have been removed.
 
 
It appears that the southern abutment of the bridge has raised about 3-4 meters relative to the northern end, and shifted sideways about 5 meters.
 
 
Evidence can be seen of the rupture passing across the levee upstream of the bridge and passing across the river and traversing the bridge near its southern end. Because of all of the demolition work, it is hard to see exactly where the rupture was located.
 
 
Some pounding damage is apparent at the remaining expansion joints suggesting that considerable longitudinal movement occurred during the earthquake,
 
 
and the southern abutment has considerable damage (in addition to its permanent offset).
 
 
An adjacent concrete batch plant was apparently destroyed by the earthquake. There was a lot of highway construction throughout the area and a significant number of these large, but movable concrete batch plants collapsed during the earthquake. These plants incorporated large steel bins elevated on steel braced frames.  
 
We turn off Route 3 onto Route 16A and approach Jiji, which is near the epicenter of the earthquake. There is little damage to buildings in this largely rural area, even to unreinforced brick buildings. A local reinforced concrete school was damaged. We also pass by a large buttress dam which appears to be undamaged.  
 
 
Jiji is a small-sized town. Many three story buildings of reinforced concrete and mixed construction were built in its commercial district. A few taller buildings have been built in town, including a new reinforced concrete frame under construction (apparently undamaged).  
 
There are numerous vacant lots full of rubble in the older part of town, suggesting severely damaged and collapsed buildings that have been demolished and removed.
 
 
A major cable stayed bridge was in the final stages of completion at the time of the earthquake. This connected Jiji to a smaller town (Lu) across the Juoshuei River. Viaducts ran from both banks of the river to the ends of the cable stayed bridge.
 
 
Cable stayed bridge was supported on a single pylon and two rows of cables (greased post-tensioning strands in a protective cover) connected the pylon to the center of the roadway.
 
 
 
 
Significant liquefaction and settlement was noticed around the base of the viaduct piers.
 
 
The viaducts were supported by two single column bents. Prestressed, precast beams were supported on the bent caps. The roadway was cast in place, continuous over supports, with an expansion joint provided about every fourth span. The closure pours in the deck between the continuous spans were heavily damaged near the cable stayed bridge suggesting large force transfers between spans at these locations.
 
 
Some damage is associated with the expansion joints, especially as the cable stayed bridge is approached. Keys were placed on top of the bent caps and a small number of short restraining cables connected the adjacent spans in order to limit the displacement of the prestressed girders relative to the bent caps and one another.  
 
The keys were damaged, especially as the cable stayed bridge was approached. The keys were highly damaged and in several cases punched through the end diaphragm connecting the prestressed girders.
 
 
In the end span, where the viaduct connects to the cable stayed bridge, one of the viaduct decks deflected sufficiently relative to the bent cap that it fell transversely off the supporting bearings.
 
 
The lateral movement knocked off a small enclosure wall on the outside face of the bridge, and dislodged the bearing on outermost girder.
 
 
The main cable stayed portion of the bridge sustained a variety of damage.
 
 
The bridge had not yet been completed at the time of the earthquake. A closure strip remained to be placed on one side of the bridge at the center pylon.
 
 
Prestressed panels were to be placed to complete the bottom side of the deck, and a cast in place top surface would complete the roadway.
 
 
Severe damage occurred in the deck on the southern side of the bridge as can be seen in the photos below.  
 
 
 
Additional damage occurred in the pylon. Below the roadway, the pylon showed evidence of only small cracking. However, above the roadway there was severe spalling of the cover and extended upward nearly to the level of the first cables.
 
 
The severity of the spalling may have been in part due to the very close spacing of the transverse reinforcement in this area. In many regions, there was no apparent gap between the transverse steel. The perimeter ties appear to be made from U-shaped bars, lapped along the long edge of the pylon. In addition, cross ties with long extensions on 90 degree hooks were used. This resulted in a nearly solid wall of steel on the inside face of the cover.  
 
One of the cables pulled out of its top and bottom attachments.
 
 
The bottom cable appears to have pulled out due to a fracture of a tube holding the nut that locked the cable in place.
 
 
The fracture initiated at the section where threads were started in the tub to allow for attachment of the nut. There was apparently severe movement of the cables as suggested by the waterproofing shields and stoppers walking out of their keepers and the local failure of other attachment details.
 
 
The ends of the decks were supported by a two-column bent. The end of the viaduct was also supported on this bent.
 
 
A large keyway was intended to restrain lateral displacement of the bridge deck at this location. No restraint was apparently provided to limit or resist vertical movement. It is clear from the damage that the both ends of the cable supported deck lifted considerably and moved in both directions laterally by as much as a quarter of the width of the roadway.
 
 
Based on the impact damage to the keyway, it appears that the abutment was subjected to repeated impacts from the deck. The key detail used resulted in an eccentric contact with the support when it impacted off center, thereby introducing torsional reloads into the deck. These may be responsible in part for the cracking patterns at the connection of the bridge deck to the pylons which are consistent with torsion loading on the roadway.  
 
 
From Jiji we traveled further east on highway 16 to Shueili. This is a mid-sized town. One building was being demolished, but several multistory reinforced concrete "administrative" style buildings appeared to not have any damage. A reinforced concrete school similar to those seen on Day 2 was only lightly cracked. Several unreinforced masonry buildings collapsed, but several others remained standing without apparent damage.  
 
 
From Shueili, we turned north on Highway 131 towards the Sun Moon Lake resort area. An earth dam is used to dam the lake.  
 
The photo shows a separate intake structure. Although there is no apparent damage to the dam, a drill crew is coring down through the center of the dam.
 
 
The lake is surrounded by a large number of resort hotels. Some of these appear to be quite old. At least six reinforced concrete hotel structures collapsed. Representative of these is the main building of the Teachers Hostel. This is a large complex of buildings of varying ages. The main building was irregularly shaped, with a large lobby in the center. This area collapsed during the earthquake.
 
 
Details of the portion of the hotel that collapse are similar to those seen previously in other reinforced concrete buildings. From the photo it can be seen that the failures occurred at the top and the bottom of the columns. Very little transverse reinforcement was provided for the splices at the bottom of the column and in the joints. As a result, these areas disintegrated before other failure modes could develop. Several other newer looking structures on this site, including a large auditorium, were largely undamaged.
 
 
One nearby hotel (Evergreen) was only five years old. The building appeared to be in excellent shape structurally and architecturally. However, a three or four story tall retaining wall collapsed against the building causing local damage and requiring replacement of the retaining structure.  
 
 
From Sun Moon Lake we travel further northward through the mountains to go to Puli. On the way, we pass the National Chi Nan University. This is a brand new university campus, built on a plateau at the top of a hill. Our understanding is that the campus is all new, built within the last three years. However, the campus has been abandoned following the earthquake, with students being distributed among other national universities. Walking through several of these buildings indicates no structural damage. However, the cladding consists of thin reinforced concrete wall panels as previously discussed.  
 
Many interior walls are unreinforced brick. Cracking occurred in many of these partitions and exterior panels. Generally, cracks were small, but on occasion, they were larger
 
 
Damage was also seen in the suspended ceiling system used. Reportedly, the university administration preferred to remove students and plan the repair activities, rather than have them in place during the planning and repairs.  
 
 
As sunset approached, we arrived in Puli. Interestingly, this fairly large city has suffered as much damage as any city we have seen thus far. Each block seems to have several severely damaged or collapsed buildings. The city hall (police station) has suffered a weak story collapse. A major brewery collapsed. Several other modern looking reinforced concrete buildings collapsed, but nightfall limited our ability to carryout a more comprehensive investigation at this site.  
 
In the older portions of town, many of the typical three story commercial buildings, that escaped damage in other towns were damaged here. In the direction perpendicular to the street, these had unreinforced brick walls infilling concrete frames. However, in the transverse direction, the resistance is provided by non-ductile frames.
 
 
In Puli, and in certain other areas, the columns were severely damaged.
 
 
In Puli, numerous buildings had their columns shored.
 
 
In some cases, repairs had begun, however the effectiveness of some of these repairs is questionable. For instance, in the photo, the existing concrete column at the ground level is being enlarged, but not being attached to the upper columns or beams. Some owners suggested that they could not afford repair and might want to simply leave the temporary shoring in place permanently.
 
 
A new looking parking structure in the center of town had some structural damage.
 
 
The short columns at the one end of the building suffered inclined cracks in a classical x-pattern
 
 
With the coming of darkness, we return to Tiachung along Highway 14. We pass through several short dual bore tunnels along the way. Some of these have clearly suffered cracking. In several cases, one tunnel bore has been closed down, and both directions of traffic are diverted to a single tunnel.  
 
Steve Mahin