 |
 |
|
|
|
|
|
Experience with Concrete Overlayed
Bridges in Tasmania Introduction There are different ways of using concrete in a timber bridge (see Figure 1). These including:
 This paper will look at how bridges of these types have been built in Tasmania and how they have tended to perform. In particular it will look at:
Successful
Timber and Concrete Bridges Bridge builders have been building timber and concrete bridges successfully for a long time. One of the first detailed studies of these bridges was carried out by the Oregon State Highway Department in the USA. They built composite timber and concrete highway bridges as long ago as 1932. In Australia, timber and concrete bridges have been built since at least 1945. The technology was probably encouraged if not imported by the US military Bridge near Railton The oldest concrete and timber bridge examined by this research is a single span bridge in service near Railton in Tasmania. This bridge was built in 1949. It is 6 m long, 7 m wide and spans 3.6 m clear over Whitewater Creek on the Elizabeth Town Main Road. It has 200 - 250 mm thick concrete deck laid on six timber beams at 1 500 mm nominal centres. These beams are 450 deep round timbers that have been squared on one side only. This was to allow for the removable formwork use in construction. The deck of the bridge was built with up stand curbs at each side and these act as edge beams to the bridge. The concrete was returned down behind the back of the beams and then the formwork stripped away so that there is now an air gap between the concrete and the end of the beams. The substructure is driven timber piles with timber headstocks. This bridge was built with timber beams because of a shortage of reinforcing steel in 1949. It was built with a concrete deck because the original builders hoped to get improved service life of 25 to 35 years from the timber. The bridge is now over 45 years old and has performed reasonable well. Maintenance on the bridge has been low. Where the timber beams are completely under cover they are still in very good condition. There is no evidence of decay. One beam was tested and returned a moisture content of only 17. 5%, 125 mm inside the beam, meaning that the beam has effectively seasoned. One can infer that the timber beams should last as long as the concrete deck continues to keep the water out. There has been some problems with the bridge however. The timber work in the bridge has shrunk from its original size and the structure has settled. Though it appears that it was an allowance for the deck to settle also, this was probably insufficient. As a consequence, the beams and the deck separated to such an extent that when the bridge was 6 years old, the beams had to be jacked up to the deck and packed. This appears to have been a continual problem with the bridge and the two outside beams still sag noticeably below the surface of the concrete. Decay is now evident in those parts of the bridge which have not be fully protected by the concrete deck. Much of the deterioration of the exposed timber has been ignored and the base of the piles have been encased in a concrete footing. This footing has also been used to further prop the capwales and indirectly the beams. We tested the bridge with full gravel truck loading and measured the deflection of both the beam and the concrete beside the beam. The maximum deflection recorded was 2 mm but the results did show that the concrete and timber had separate responses. On average, where the concrete deflected about 1. 5 mm, the timber would only deflect about 0.75 mm. Bridge on the Little Rapid River at Smithton A newer and much larger timber and concrete bridge is Forestry Tasmania's bridge across the Little Rapid River outside Smithton. This bridge is 19. 5 m long with two roughly equal spans of 9. 7 m, and 4 m wide. It was built in 1975. It has 160 mm thick concrete deck laid on four timber beams at 1 000 mm nominal centres. These are between 520 and 600 deep round timbers. Timber palings were used as sacrificial formwork. The concrete deck was returned down the back of the beams against traditional gravel boards. The substructure is timber capwales supported on timber piles resting on a sill log. This bridge has now served 21 years old on a low use road. Maintenance appears to have been very low. All the timber elements appear in very good condition, except where they are open to the weather or to ground moisture. The piers appear to be in good condition. We load tested the bridge and the results are set out in Figure 2 below. The span to deflection ratio for the bridge centrally loaded with a 20 t. gravel truck was about 1 in 1200, while it was about 1 in 950 for loading on the side of the bridge. Generally deflections were symmetrical.  This bridge also has had problems with timber shrinkage. Here, there was not an allowance for the deck to settle as the beams shrank radially so a 25 mm gap has opened up between the concrete and the timber beam at each end of the bridge. This gap tapers for about 1 meter until the deck drapes down onto the beams. Also the bridge was built with a central reinforced expansion joint. This has allowed the beams to sag under the dead load of the concrete. Neither of these problems appear to effect the performance of the bridge. However, they are important points to consider for bridges that may be built in more trafficked areas. Comparison of two bridges In the last newsletter, we included the site test results for two bridges that are still in service. We now report on two bridges we have tested fully. The first bridge was the Devil's Creek Bridge on the Tasman Highway near Falmouth on the East Coast. This was a three span single lane bridge built in about 1979 and concrete overlayed in 1981. The second bridge was a two span single lane bridge over the Guide River near Burnie. It was built in about 1979. The spans on both bridges were about 7. 2m per span and as both bridges were to be bypassed, we had the opportunity to test the complete bridges on site and then test the beams to destruction at our facility in Launceston. The load test deflections are shown in Figure 3 below. These charts show the deflection at the centre of the beams for the bridge loaded about the centre of the bridge. The field deflection of the Guide River bridge are dramatically higher than those at Devil's Creek. This is for two reasons. The beams at Guide River were generally weaker than those at Devil's Creek and the end decay, and subsequent crushing at the supports, played a significant part in the total deflection of the beams at Guide River. The end decay of the beams at Devil's Creek was negligible.  After the bridges were removed from service we tested each beam to destruction. The results for these test been analysed and Figure 4. shows the bending stress for three population of beams: eight beams recovered from Guide River; twelve beams recovered from Devils Creek; and two pristine bridge beams. The results have been reduced to bending stress per deflection so that the differing span and dimensions of the logs can be taken into account. Comparing the results illustrates some interesting points: The Guide River Bridge still had at least three very strong beams left in it. It had three more beams that were suspect and it had two beams that had very low capacity. The failure strength of these beams was right at the working load one would expect from normal traffic loading. The moisture content of one of these failed beams was 77% in the outside 75 mm of the log. The beams from the bridge at Devil's Creek were considerable stronger and stiffer than those at Guide River. Though this population also had its weaker members, even the weakest still had a safety factor of 2.5 to 3 over the working load one would expect from normal traffic loading. A significant proportion of this population was also stiffer than the two pristine logs tested. They were not stronger in the long run however. The moisture content of the beams at Devil's Creek were only about 17% in the outside 75 mm of the log and rose to about 25 % at the centre of the log. These beams had effectively seasoned and as they did, they suffered considerable radial checking. Unlike one would expect from unroofed bridges however, there was little or no biodegradation in these checks and the fibres that spanned across the checks remained intact.  Failures in
Timber and Concrete Bridges Before looking at specific failure of timber and concrete bridges, it is important to realise exactly what the concrete in a concrete overlayed bridge is supposed to do. Its two main functions are to provide a well wearing running surface and to keep the water out of the timber. It also performs a secondary role of providing some continuity between spans in multi span bridge, if it does not have expansion joints over the piers. The overlay will serve its to main functions is it sufficiently strong to handle traffic impact load, if the deck underneath it is sufficiently strong to support it and if the surface is impervious. This last point includes protecting the timber beams from surface water and ensuring that water runs off the bridge and away from the timber quickly. If the any of these points are compromised by poor detailing or bad workmanship, then quite expensive problems will arise. If the deck is pierced, or breaks up or is not impervious, water will seep under the overlay, be retained against the timber. This leads of decay of the deck and beams. The deck than becoming weaker very quickly with less support for the overlay. So more water gets in and the cycle continues. A bridge in this condition can not be simply redecked and effective remedial work is expensive. Bridge on the Golconda Rd near Scottsdale The problems with this bridge include just about everything that could go wrong with a concrete overlayed bridge. This bridge is a two span bridge, 17.8 m long and 7.1 m wide on a heavily travelled road. It was built new in about 1980 and had a concrete overlay installed at construction. In an effort to have the bridge operational as quickly as possible, the concrete for the overlay was specified as rapid setting but the day it was laid was very hot. As a result, the concrete went off very fast, so fast that two loads of concrete had to be dumped in the. The concrete that did get onto the bridge was spread quickly, with poor compaction. As a result the bridge is effectively covered with eight separate strips of concrete. These move independently and allow water to get between them. This water has settled in the deck and it has decayed. Consequently, part of the deck has rotted out completely and the deck has been punctured by the traffic. The water has also gotten into the beams and one span has had to be propped with an additional support. Soft flashings were used over the outside of the bridge and these were turned up under the beams. These has filled with water and keep the outside beam almost continually in a pool of water. These beams have also decayed. This bridge is still in service but not for very long. The Long Bridge near Branxholm In the early 1980's, the Works Department or what is now Works Tasmania, overlayed a series of multi span timber bridges throughout the state. Executed to what were probably similar standard drawings, these bridges nevertheless had a minor range of different details used in them. In most cases that we have examined, the decks were laid without expansion joints, except at the Long Bridge near Branxholm. The Long Bridge is 5 span 37. 3 long single lane bridge that was built in about 1978 and overlayed a short while latter. There appears to be two associated problems with the overlay to this bridge. First the bridge was overlayed without a camber so that with the build up of material on the edge, water is retained on the deck of the bridge. Secondly, the bridge was built with expansion joints. As time has gone by, the movement of the beams and the rocking of the corbels has lead the corking in at least one of these joints to work out and the joint has leaked. This is not a noticeable problem when it first happens but as the deck tends to pool water, this water seeps into the deck, where it is retained between the timber and the concrete. Decay incubates in these situations and again a section of overlay has been punched out where the deck members below them have rotted through. This is only a small problems for this whole bridge but it is a very complicated one to fix effectively. A new deck member cannot just be installed as one would do with a straight timber bridge. New
practice in overlay construction Working with this research project and subsequent to experimental bridges built and tested at our Launceston workshops, Forestry Tasmania recently built a 6.1 m span composite timber and concrete bridge near Beulah in north west Tasmania. This bridge used eight timber beams to form the log bed and 16 mm steel dowels as the shear connectors. The dowels were driven into tight pre drilled holes at 300 mm centres along each log. The bed was then overlayed with 140 mm of concrete. The logs were only trimmed where they sat on the bed logs and the outside beams were protected with a flashing strip and a drip line edge to the concrete. This experimental bridge was built in 3 days at a cost of about $10,000: $320 a square metre of bridge deck. With a crew trained in this type of bridge and with a refined design, both construction time and cost could be reduced further. After the deck had cured, we load tested the bridge. With a 20 tonne gravel truck right over to one side of the bridge at the centre, we recorded our maximum deflection, 2 mm over a 6 m span.  |
|
|
|
|
|
|
Timber Research
Unit | |
|
| ||||
