PRELIMINARY FIELD OBSERVATIONS OF THE CHILEAN EARTHQUAKE OF 3 MARCH 1985

This paper is an initial appraisal of the effects of the M = 7.8 earthquake centred near San Antonio, Chile, which occurred on 3 March 1 985. The author arrived in Chile six days after the main shock and spent a week investigating as many aspects as pos sible . The earthquake caused heavy damage to a large area including the major cities of San Valparaiso and Santiago, and resulted in 700,000 people being made homeless. Damage to engineered structures appeared to be less than might be expected for an earthquake 'of this magnitude , but non structural damage was heavy. Eight or nine accelerographs recorded the shaking in different parts of the more heavily affected area and should help enrich the findings of the numerous valuable post-earthquake studies which will be possible.

Antonio, Valparaiso and Santiago as well as many other smaller communities in an area exceeding 200 by 80 kilometres. The main shock has been assigned a Richter magnitude M = 7.8 and two weeks after the event the official damage statistics included 177 people killed, 2 ,500 injured, 1 70,000 homes destroyed and 700,000 people homeless• Early estimates of the cost of the earthquake were set at $US600 million, but some locals expected that this figure would be considerably exceeded. Considering the large amount of damage to houses the number of deaths and injuries was mercifully low. This has been attributed to the fortunate time of the earthquake which occurred : during the evening promenading period when the local custom is for most people to walk about in the streets.
Despite the large amount of adobe construction, there were a significant number of casualties in engineered buildings .

SOME SEISMOLOGICAL ASPECTS
Based on the uncorrected data, the peak horizontal ground accelerations varied between about 0.16 g and 0.75 g with vertical accelerations being in much the same range.
There is a possibility that in one direction acceleration increased rather than decreased with distance from the source for a distance of some tens of kilometres. If true, this will pose some interesting questions about the nature and distribution of the rock layers in the transmission path. Firstly the MM intensities in what was thought to be the epicentral area were relatively low for an earthquake of M = 7.8, the maximum intensity probably being MM9, which was observed in one small area only, namely on the poor soil of downtown San Antonio. The relationship between the definitive focal mechanism and the isoseismals will hopefully help to explain this phenomenon.

FIELD RECONNAISSANCE ARRANGEMENTS
The author was in London at the time of the earthquake, and decided on 7 March to fly the next day to Santiago with Edmund Booth from Ove Arup & Partners 1 London office.
After 24 hours travelling we arrived in Santiago on the afternoon of Saturday 9 March.
As damage was very widespread and travel arrangements initially uncertain we decided to be based in Santiago, travelling daily to different areas of interest. in Chile, Sunday 10 to Saturday 16 inclusive, we were in the field, mostly travelling two hours or so each way to get to the daily zone of interest.
To make maximum use of the daylight, we did not start for base till dark most nights, and consequently usually started dinner about 10.00 pm.
Everyone found this routine exhausting, and many of us suffered from the local "tummy bug" at some stage.
However, such opportunities to learn from the real thing are not common, and the effort was well worth making.

Adobe and Unreinforced Brick Buildings
Adobe (mud brick) is used in Chile for many buildings including houses, farm buildings, light industrial buildings and modest churches.
Naturally the pure adobe buildings were the worst affected and as the majority of the populace live in adobe houses, the high figures for homelessness mentioned above can be easily explained.
As has occurred in previous earthquakes , the older style of buildings framed with timber studs with adobe infill performed much better than pure adobe; while such buildings sometimes sustained severe damage, total collapses appeared to be rare.
Even quite crudely detailedframing of almost any sort gave considerable protection to many other adobe buildings.

Reinforced Masonry Buildings
Relatively little reinforced brick or block construction was observed, but one housing estate of flats made of reinforced brick was visited.

The Villa Santa Carolina in Santiago comprised
three-storey blocks of 1 984 construction all of which were heavily damaged ( Figure 1 ) .
While the materials and workmanship appeared to be generally of good quality, the bricks had perhaps excessive vertical holes and the amount of vertical reinforcement was considerably less than would be required by the New Zealand code.

Timber Buildings
For a country with a large indigenous timber resource, it was surprising to find few buildings of timber construc-  (Figure 2) (Figure 3) were heavily damaged ( Figure  4)  Only one further factor noted in Table 2 will be commented on here, namely that of previous earthquake damage. Two of the 16 buildings visited had been redamaged in members that had been damaged in the 1 971 earthquake and subsequently repaired.
In the case of the Acapulco Building at Vina del Mar (Figure 5), there was evidence of heavy rusting of reinforcement in the previously damaged zone, which raises the serious question of the likely effectiveness of repairs to damaged buildings.

4.7
Non-Structure and Secondary Structure -Cost and Safety Questions  Figure  4), none of which was separated from the structure. This, of course, would be expected in the momentresisting frame building, but the other two buildings had shear wall structures. Some investigation of the deformations in these latter two buildings might therefore provide further data on the degree of protection to non-structure provided by shear wall stiffness above.    The first principle of earthquake resistant design is to save lives. This has traditionally been approached by attempting to ensure that the structural frame does not collapse, while heavy damage to other items has been tolerated, albeit with increasing regard to danger from falling non-structural elements ( Figure 5) .
An analysis of the causes of casualties in engineered structures by Durkin of the EERI team' will therefore be important in ascertaining (a) the nature of the dangerous items , and (b) whether the methods used currently to protect the integrity of these items are appropriate to the level of integrity provided for the structure.
The damage distribution noted above, such as that in the Canal Beagle apartments, stresses the importance of this topic.

ROADS AND BRIDGES
As usual in large earthquakes there was some disruption to road traffic as a result of the earthquake, due to debris from buildings, slips, subsidence and cracking of roadways and bridge failures. The level of initial disruption was not particularly great, most roads and streets being made passable within a few days of the earthquake.
At least along the main arteries there were only a few slips on faces of cuttings as the hillsides were composed of stable rock and soil types.
However, on the hillsides facing the ocean, localised slips had been sufficient to close a number of urban streets till higher priority works had been done. As might be expected, embankments gave more trouble, with a rnoderate amount of subsidence and tension cracking of pavement surfaces, especially at bridge approaches.

Initially
it was reported that about ten bridges had failed, but as the affected region was very, large this proved hard to substantiate , and only three bridges will be commented on here. Just south of San Antonio the Lo Gallardo Bridge ( Figure 6)  This concrete bridge crosses a wide shingle river bed and consists of many short spans, and would be similar in principle to many bridges in similar terrain in other countries including New Zealand.
The tops of piers just below deck level had suffered substantial damage due to lateral and longitudinal bending moments, but apparently had been deemed still to have sufficient vertical load carrying capacity for the bridge to remain in service. No load restriction had been placed on the bridge which hopefully would not prove to have the same result as with the bridge at Palmilla.

SILOS
A group of wheat silos at the flour mill, Casablanca, were heavily damaged. These comprised reinforced concrete and bolted corrugated steel silos, the latter being in a state of collapse.
The concrete silos were mostly built in the 1 960s, and some had been damaged in the 1 965 or 1971 earthquakes.
Further damage occurred to the strengthened silos in this earthquake. A considerable length of the concrete wall forming the quay had disappeared under the water and the associated subsidence of the adj acent paved area was enormous, causing massive damage to the quayside buildings and overturning ship-loading cranes. A few sand-boils were found in crevices in part of th^ subsided zone, but an initial appraisal did not find large scale liquefaction to be likely. However, the evident increase in pore water pressure in the embankment fill and the underlying sands may well have contributed to the quay wall failure and to the subsidence that seemed to have occurred in parts of the embankment not affected by the quay wall failure.
Other quays within the harbour had not failed but there was no time to inspect them for signs of distress. However, their difference in behaviour may help in back-analysis of the breakwater collapse through comparative analyses•

LIFELINES
Although the author did not