EPS CONCRETE COMPOSITE STRUCTURES
LAB TESTING AND FEA MODELLING
M.Haimana, M.Rakb, J.Krolob, Lj.Hercegb, V.Čalogovićb
a Department for Structures, Faculty of Civil Engineering University of Zagreb, Kaciceva 26, 10000 Zagreb, Croatia,
bTechnical Mechanics Department, Faculty of Civil Engineering University of Zagreb, Kaciceva 26, 10000 Zagreb, Croatia, , , ,
ABSTRACT
This paper presents the results of the numeric and laboratory tests of the «T» section of EPS concrete-timber composite ceiling structures. Sheared connections of EPS concrete board and timber girder elements have been tested in several different ways by means of mechanical connectors. Various types of dowels were used for sheared connections. Shearing forces in the connection plane were assumed by means of dowels, then by means of nails, and by gluing the connection of EPS concrete and timber with epoxy glue.
All the tests were numerically modelled by means of FEA software COSMOS/M 2.8 and the results were compared with those of the laboratory tests.
Application of these structures in civil engineering practice is multiple, particularly in reconstruction of old timber girder floor structures or manufacture of new ones at the reconstruction of old buildings in areas exposed to frequent earthquakes.
1. INTRODUCTIN
At the reconstruction of ceiling timber structures, they can be strengthened with reinforced concrete board composed with the existing timber girder. The new lightweight EPS concretes weighting 1200 to 1400 kg/m3 provide new possibilities of application in composite structures. Besides very high solidity achievable with such structures, their advantage is also the increased fireproof quality and very good acoustic properties. Furthermore, such structures are of significantly lower weight applicable in areas exposed to frequent earthquakes in reconstruction of old timber girders or manufacture of new ceiling structures.
In efforts to test and determine the rules of behaviour of such structures at static loads, several types of dowels were tested in the laboratory. The tests were carried out within the framework of the scientific research program supported by the Ministry of Science and Technology of the Republic of Croatia at the Laboratory of the Technical Mechanics Institute Faculty of Civil Engineering, University of Zagreb.
2. LAB TESTS
Sheared connections of a timber beam and an EPS concrete board were done for four different cases of connections in the connection plane.
In Type 1, the connection was done with Ø25 mm steel dowels placed, two of them in a single section. The distance between the dowels in longitudinal direction was 200 mm. In Type 2, the same Ø25 mm steel dowels were used, but they were placed alternatively, 1 dowel in section, at distances of 200 mm. In these two types, the holes for dowels were of a slightly larger diameter than dowel diameter. Epoxy glue was placed in the holes and then the dowels were installed.
In Type 3, Ø6 mm nails were used for connection. In a single section, three nails were driven in up to the depth of 100 mm. Mutual distance between the nail groups was 100 mm.
Type 4 was tested with the connection by means of epoxy glue placed on timber beam elements directly before placement of EPS concrete.
Presentations of all the four tested laboratory samples are at Fig. 1 and Fig. 2.
Features of lightweight EPS concrete are the following:
Eclw=12400 MPa, σcc=14,5 MPa, σct=4,5 MPa, tcs=2,5 MPa
and features of glued laminated timber beams are the following:
EIIv=12800 MPa, σII vc=55 MPa, σII vt=72 MPa and σIIvs=15,5 MPa.
Fig. 1. Types of Timber-EPS concrete composite beams
Fig. 2. Types of in-built connectors
The testing of laboratory samples by bending was carried out on beam systems with 3.2 m span.
The cross section of a glulam timber beam element was 160/180 mm, and the thickness of the EPS concrete board was 80 mm. The width of the tested «T» section was 600 mm.
Load was added by linear force at the midspan, and displacements were measured at the place of maximum expected shear and installed dowels, and vertical displacement in the midspan. Four LVDT sensors were symmetrically placed at the connection plane between concrete and timber. The sensors measure the relative deformation between concrete and timber. Seven LVDT sensors were placed at the midspan to measure deformations on concrete and timber. Three LVDT sensors for measuring the displacement were placed at the midspan and on the bearings. Two LVDT sensors were placed at the end of the beam to measure common sliding between concrete and timber. There were total 15 measuring spots connected to A/D converter and a laptop interface. To monitor and record data during the test, the “National-Instruments – Virtual Bench Logger” software was used.
The bending tests were done for two characteristic load levels.
The first load level was with concentrated linear force of 20 kN at the midspan. The force size is the accounting force of the permitted load. The other level of load was by force until failure, corresponding to breaking load of the testing models.
Measuring places are presented on Fig. 3., and test of one of the models is presented at Fig 4.
Fig. 3: Review of the measuring places
Fig. 4: Testing of EPS concrete composite beam Type 1
Behaviour of the tested composite girders within the limits of permitted load up to 20kN force is almost linear, with small differences in midspan deformation, as seen on Fig. 9. Very high differences were obtained by laboratory test of each of the aforesaid types until failure as presented on diagram (Fig. 5). Types 1 and 4 have shown the best bearing capacity characteristics until breaking, while Type 3 with the nails yielded the worst test results.
Failure of all the tested types occurred by exceeding the tensile strength of the glulam timber beams.
Fig. 5: Force – deformation ratio until failure, for tested types 1 to 4
3. NUMERICAL MODELLING OF THE TESTED MODELS
Numerical tests of the tested types 1 to 4 have been carried out by complex analysis by means of COSMOS/M 2.8 software. The modelling was done by use of two types of elements. By means of SOLID elements, EPS concrete board was modelled as well as laminated timber beam. Mechanical features of timber and concrete in space were taken into account, as manifested with different elasticity and shearing modules.
The following material features were used, as obtained by individual tests of the laboratory samples of elements used in this test:
For glulam timber:
EX = 13000 MPa NUXY = 0.027 GXY = 600 MPa DENS = 4.0E-6 N/mm3
EY = 2060 MPa NUYZ = 0.6 GYZ = 60 MPa
EZ = 600 MPa NUXZ = 0.033 GXZ = 650 MPa
For EPS concrete:
EX = EY = EZ = 12720 MPa NUXY = NUXZ = NUYZ = 0.22
GXY = GXZ = GYZ = 5100 MPa DENS = 1.4E-5 N/mm3
For steel dowels:
EX = 210000 MPa NUXY = 0.3 GXY = 81000 MPa DENS = 78.5E-6 N/mm3
The boundary elements network of the concrete part of the section was not numerically connected to the boundary elements network of the laminated timber beam, instead their distance was 1 mm. The connection of the networks was made only at places of installation of dowels modelled by means of round section, 25 mm diameter BEAM3D elements. The modelling of uniform behaviour of the dowel connection point was done by CPDOF command, where identical displacements were modelled in all the three directions for two neighbouring points (nodes).
On the bearing part, a steel washer was modelled to obtain as realistic status as possible concerning the transversal pressure stresses of timber in the zone.
FEA model and border conditions are presented on Fig. 6.
Fig. 6. FEA numerical model and boundary conditions of Timber - EPS concrete composite structure,
a)
b)
c)
Fig. 7 a) Deformation status for model Type 1, for 20 kN force in midspan ; b) stress status SX for analysed load; c) status of shearing stresses in timber and EPS concrete. Values are in MPa.
The Fig. 7 shows several numerical results for the modelled TYPE 1. Under a) there is presentation of the status of deformations in the elastic area for the laboratory analysed status of load with 20 kN force. Calculated deformation is about 3 mm.
Under b) is the status of main stresses in longitudinal X direction in elements of timber and EPS concrete, and c) is the status of shearing TXY stresses in timber and concrete.
All the analysed statuses of deformations and stresses are within the limits of permitted values for the analysed load. The status of stresses in steel dowels is lower than maximum permitted.
Distribution of bending moments per dowels is presented on Fig. 8.
Fig. 8. Distribution of bending moments per dowels in EPS concrete composite structure connection
Values obtained in laboratory tests have been compared with the values obtained in this numerical test, for each analysed case.
4. RESULTS
The results of the lab and numerical tests have been compared and are presented on the diagram (Fig. 9). It should be noticed that for the analysed cases Type 1 to 4, very good compliance of the obtained results was obtained in the elastic area.
It has been shown that the cases with two installed dowels per section (Type 1) as well as gluing the connection with epoxy glue (Type 4), are best solutions for sheared connections between timber beams and EPS concrete board.
Nail connections, otherwise a very good solution in timber structures, have shown the worst results for sheared connection of timber and EPS concrete, as contact surfaces in timber or concrete are too small and there is maximum pressing and yielding of the sheared connection.
Fig. 9. Diagram of comparison of the values obtained by experimental and numerical analysis
5. CONCLUSION
EPS concrete composite structures account for relatively simple and inexpensive structures usable in repair or reconstruction of old timber beam ceilings. Such a combination is very advantageous for use in areas exposed to frequent earthquakes, where as lightweight structure as possible needs to make sufficiently solid and strong solution of floor structure strengthening. These solutions contribute to the general stability and solidity of buildings where such structures are applied.
By composite timber – EPS concrete structures, the floor structure bearing capacity is increased, and deformations for maximum exploitation loads are reduced.
The laboratory tests have shown that the use of epoxy glue as connection between EPS concrete board and timber beam is the best connecting solution.
Use of nails in connection plane for transfer of shearing forces of connection has brought the worst results and is not recommended for use in civil engineering practice.
All the numerical tests have shown very good compliance of the results with the results obtained in the lab. This provides additional certainty to the results of this scientific research.
REFERENCES
[1] H. J. Blass, Mario. L.R. Van Der Linden, M. Schlager: «Timber Floors Strengthened with Concrete», Saving Buildings in Central and Eastern Europe – IABSE Colloquium Berlin 4-5 June, 1998.
[2] A. Frangi: Elasto-Plastic Model for Timber-Concrete Composite Beams with Ductile Connections, SEI Volume 13, 1/2003
[3] T. Godycki-Cwirko, J. Pawlica:, «Construction and Design of Composite Concrete – Timber Floors», Saving Buildings in Central and Eastern Europe – IABSE Colloquium Berlin 4-5 June, 1998.
[4] H.Werner: «Holz – Beton – Verbunddecke mit einer neuartigen Fugenausbildung»; Bauen mit Holz 4/92
[5] H. J. Blass, M. Schlager: «Trag – und verformungsverhalten von Holz-Beton-Verbundkonstruk-tionen, Bauen mit Holz 5/96
[6] V.Rajčić, M. Rak: «Continuous Shear Connecting-The Best Way to Compose Timber and Lightweight (EPS) Concrete», The 7th World Conference on Timber Engineering, WCTE 2002, August 12-15, 2002, Shah Alam, Malaysia.
[7] M.Haiman: «Modelling of composite Wood/Concrete Floor Structures»; Građevinar 10, Vol. 48, Zagreb 1996.