Embedment Strength of Tropical Timber Species

The objective of this study was to determine the embedment strength of tropical timber species. The study was limited to five timber species representing densities from 300 kg/m to 900 kg/m. The embedment strengths were determined in directions parallel to and perpendicular to the grain of timber and were determined on timber thicknesses of 25 mm, 38 mm and 50 mm using bolts of diameter 9.5 mm, 12.5 mm and 15.9 mm. The experimental results obtained were compared with the recommendations ofthe Eurocode. It was found that these recommendations need to be modified for tropical timber species and this paper provides the modified equations.


Background
The embedment strength of timber is essential to calculate the design resistance capacity of timber bolted joints when using the European Yield Model (EYM). The EYM is now the basis for the determination of strength of timber joints in most codes of practice [1]. The application of the EYM to bolted joints of timber species indigenous to Sri Lanka has been discussed previously [2].
The embedment strength is defined in EN 383: (1993) [3], as the average compressive stress at maximum load in a piece of timber or timber-based sheet material product under the action of a stiff linear fastener while the fastener's axis is perpendicular to the surface of specimen and the fastener is loaded perpendicular to its axis.
The embedment strength can be determined, experimentally, in accordance with EN 383: (1993) [3]. This test can be regarded as a symmetrical three-piece joint test, using steel side members and a timber or timber-based sheet material central member in which a bearing failure is enforced under lateral load (Fig: 1

Test Programme
This paper discusses a test series that was conducted to check the validity of the EC5 equations to predict embedment strength for some indigenous hardwoods of Sri Lanka. The species selected for the test series are Lunumidella, Ginisapu, Hora, Kumbuk, and Palu. The oven dry densities of these timbers range from 300 kg/m 3 to 900 kg/m 3 . Bolted joints made from these timber species were tested in both parallel and perpendicular to grain directions. The joints were made with bolts of diameters, 9.5 mm, 12.7 mm and 15.9 mm. The timber specimens tested were of 25 mm, 38 mm and 50 mm thicknesses. However, for the 9.5 mm diameter bolts, the specimens were of 25 mm and 38 mm thickness in order to comply with the requirement of maintaining the specimen thickness within the range between 1.5 and 4 times the bolt diameter. Three replicates from each specimen were tested and thus a total of 240 tests were carried out (Table 1). Specimen size in relation to bolt size is shown in Fig: 2. The apparatus used in the test series is illustrated in Fig: 3. It consisted of adjustable steel side plates, which have a circular hole at the bottom portion of the plate. The hole diameters for fasteners were 2 mm larger than the bolt size and were so selected such that there is no rotation nor vertical movement of bolt within the hole. No tightening was applied to the bolts. Side plates were attached to the horizontal beam for convenience so that the distance between plates could be adjusted according to the thickness of the test specimen. No friction was applied on the timber specimen as the side plates did not touch it.

Details of the test apparatus are given in [10]
.   Figures 4.1 to 4.3), it can be seen that the agreement is best for smaller (9.5 mm) bolt diameters while with the increasing of bolt diameter up to 15.9 mm, deviation between embedment strengths predicted by EC 5 [4] and experimental results increases. It can be seen that the opposite is true for perpendicular to grain loading. The best agreement is seen for the specimens tested with 15.9

It is also evident that the deviation between the predicted and the experimental results increases with increase in density.
There appeared to be no significant dependence of the embedment strength on moisture content in the moisture content range tested (from 10% to 20%).