Australian Journal of Basic and Applied Sciences, 8(1) January 2014, Pages: 397-406

AENSI Journals

Australian Journal of Basic and Applied Sciences

Journal home page: www.ajbasweb.com

An E-Glass/Polyester FRP, TAUFIQ-Girder: Load-Deflections Relationship and Its Hyperelastic Behaviour Under Flexural Loads
1

Taufiq Rochman, 2Agoes Soehardjono, 3Achfas Zacoeb

Doctoral student of Civil Engineering Doctoral Program of Brawijaya University, Malang Indonesia. Professor of Civil Engineering, Doctoral Program of Brawijaya University, Malang Indonesia. 3 Assistant Professor of Civil Engineering, Doctoral Program of Brawijaya University, Malang Indonesia.
1 2

ARTICLE INFO Article history: Received 20 November 2013 Received in revised form 26 January 2014 Accepted 29 January 2014 Available online 25 February 2014 Key words: loads-deflections, TAUFIQ girder, Eglass/polyester, non-destructive test, hyperelastic behaviour

ABSTRACT An all-FRP system can be a good alternative due to its high strength-weight ratio instead of full reinforced concrete girder. FRP in the civil engineering applications have been researched mainly focused on the retrofitting the old bridges using strip system, but very rarely focused on all-FRP material application, if any, deck-slab system applications. An all-FRP girder is still interesting to develop and this proposed full FRP one is named TAUFIQ (Totally Audacity U-Fibered Intelligent Quality)-girder. A set of uniaxial mat 0 and biaxial mat 45 fiber of GFRP E-type and polyester resin was used to form the composite laminates into the U-shaped. Several non-destructive load of 200kN through 3-points and 4 points static bending tests are done to describe their performance. A quite good asymptotic result is obtained at load over than 105 kN to 200 kN, both deflection (47mm average) and strain (0.00047 strain average) describe a very ductile behaviour. This phenomenon is rushed up due to fiber angle coersion to be straightened and collaborated together with uniaxial fiber resist the tension stress, that fiber more tightened in higher load. These asymptotic strain hardening will contribute a better design safety with CIP (cast in place) and without CIP system.

2014 AENSI Publisher All rights reserved. To Cite This Article: Taufiq Rochman, Agoes Soehardjono, Achfas Zacoeb., An E-Glass/Polyester Frp, Taufiq-Girder:Load-Deflections Relationship And Its Hyperelastic Behaviour Under Flexural Loads. Aust. J. Basic & Appl. Sci., 8(1): 397-406, 2014

INTRODUCTION Although full reinforced concrete girder is known well as bridge material, but in a longer span actually its selfweight is predominant than its residual nominal bending capacity. The usual scaffolding and formwork concrete bridge construction, huge weight of traveler launcher, etc all are also have many constraints and potency to enlarge initial deflection. A usual corrosive lack characteristic of steel section and large deflection on larger span due to its dominant selfweight over live load are also stimulate human to find an alternative lighweight structures instead of them. Lightweight structures nowadays become an important component in almost overall industries. (Zemcik, 2006). Not like most traditional material such as wood, steel or concrete, FRP (fiber reinforced polymer) as laminated composite lightweight density material have a high specification in stiffness and strength so it can be used instead of conventional structural material. These characteristics performance are high strength-weight ratio, corrosive and chemistry resistance, durability, long fatigue life, crack propagation resistance, crashworthiness and energy absorbtion (Ghassemieh, 2011) impact and thermal resistance (Thomas et al, 2011). Also can be cheap easily found and formed into arbitrary geometries, boltless, weldless, and no-rivet connection needed so perfect surface finishing can be achieved and low thermal conductivity. FRP is also having low cost production and resist any biological attack where result in low maintenance cost. The using of FRP have been discussed and researched in civil engineering application such as Tarek (2010), Freymond-Maceri (2005), Bank (2006), USACE (2007) and Qasrawi (2007) that almost all of them focused on the retrofitting the old bridges using strip or prestressed strip system of this intelligent material, but very rarely they focused on the full FRP material application, if any, some of them discuss a slab system applications. Actually, pultruded section of full FRP is also included in their discussion but in the term of construction easiness many shape should be examined well, because in the pultrusion process lateral direction have not been pressed well indeed the longitudinal axis the laminated composite is well done the crinkle in the curved section raise the interlaminar delamination. More over the kind of FRP section is very limited in the I-shaped and hollow shaped pultruded FRP, and yet its dimension.
Corresponding Author: Taufiq Rochman, Doctoral student of Civil Engineering Doctoral Program of Brawijaya University, Malang Indonesia E-mail: taufiqrochman@gmail.com

398

Taufiq Rochman et al, 2014 Australian Journal of Basic and Applied Sciences, 8(1) January 2014, Pages: 397-406

Objectives and Scopes: The objection of this research are to produce a fundamental understanding of the behaviour and response of TAUFIQ-girder structure through experimental testing and collapse patterns identification and contributes FRP design and analysis to the development of any girder made of FRP materials for civil engineering structural system. Based on the research variety needs of full FRP section, good full FRP girders are still needed to develop to be proposed named TAUFIQ (Totally Audacity U-Fibered Intelligent Quality)-girder. The girder is named so that the tested shape was U-shaped, laminated, and can be developed into CIP system that distribute truck axle through concrete slab into the girder. The real condition of the moderately span bridge, first, need to be computed to determine their loads per section area or their applied stresses. A simple analytical bending calculation should be implemented also to know the global response of this girder. Full section of box shaped should be investigated firstly, in order to ignore the concrete bondslip phenomenon in the U-shaped or CIP system. The observed dimension then be determined based on that results. This U-form and its dimension must have satisfactory bending performance and predictable capacity but it still have to be adjusted to ensure that the girder will be performed in the condition such that not so sensitive to the collapse of the second order such as huge buckling in both of the top flange and web and also delamination. Several assumption should be made that are TAUFIQ-girder have transversely isotropic in the longitudinal direction, laminated and homogenizationly analysed, the ignorance of transverse shear effect and normal strains in the thickness direction, axial strains are not to be measured interlaminarly, E-Glass fiber type and ordinary general polyester thermosetting resin as matrix, it means thermal neglection. Laminated Composite Material: Laminated composite material is made of two or more very fine (5-15m diameter) Kevlar, carbon or glass fiber layers that were glued in polymer matrix, a strong adhesive substance such as epoxy resin. Matrix unite all together and protects the fiber from damage to the outside as well as transfer and distribute the applied load to the fiber and its also contribute the required properties such as ductility and strength. Each of layers or lamina is transversely isotropic in the longitudinal direction of the reinforcing fiber which holds the main stress because it have high characteristics of quality, but light and brittle. Although composite materials are heterogeneous materials that must be materially nonlinear analysed, there are many theories that treat homogeneously in a particular direction each layer, and even those that assume a fully homogeneous like ESL theory (equivalent single layer) by Tanov (2000). The fiber angle orientation of each lamina greatly affect the performance of laminated composite structure, Young transverse shear modulus ratio (E/G) as well as ply stacking sequence of the lamina. By inputing the corresponding fiber angles orientation of each lamina, the strength and its stiffness parameters can be obtained (Kant-Khare, 2004). FRP Application for civil engineering: A fully FRP material for structure have been constructed in Aberfeldy 113m span cable stayed footbridge, in Scotland (Skinner, 2009 and Busel, 2009). This bridge have 63m main span and 2.23m width and 2 planes 40 stay cables in the 18m height pylon. Its overall weight is 14.5 tons fully FRP section including pylon, girders, deck and all cables. The decks, parapets and pylons made of GFRP such as E-type and isophthalic polyester resin, while cable from aramid Kevlar fiber protected by coating polyethylene. It only need 8 weeks to build whole structures. The Aberfeldy bridge and the girder shape are shown in Fig 1a) and Fig 1b) respectively. Pre-Backward Analysis Method: In order to get an accurate and representative experimental result, a globally backward analysis is needed to check and predict the applied service loads in the real bridge, to determined loading parameter, material mechanical properties, stress concentration in the observed local section, and failure behaviour in predicting dimension, the specifications of specimens and testing equipment capacity related both given monotonic and cyclic loading rate thus girder specimens keep loading responsive while keep thoroughly recorded full-length data before its failure. Real Bridge Tipical Dimension: Based on Indonesian Bridge Standard, the bridge dimension follows T-type A-class with loading class BM 100 (100% T-loading and D-loading) at 25m span, and the vehicle traffic slab width of 7.0m, kerb width 2 x1.0m thus have approximately total width of 9.6 m, there are 5 prestressed concrete girders or 6 nonprestressed concrete girders shown in Fig 3a), Fig 3b) and Fig 4 respectively.

399

Taufiq Rochman et al, 2014 Australian Journal of Basic and Applied Sciences, 8(1) January 2014, Pages: 397-406

Fig. 1a: Fully FRP Aberfeldy bridge Source: Skinner, 2009

b) The girder of Aberfeldy bridge Source: Busel, 2009

Below are the pultruded FRP sections and a large magnification zoom of microscopic electron scan of the fiber and matrix.

Fig. 2.a: FRP wide flange section Source: http://www.diytrade.com

b) Large magnification zoom of fiber and matrices Source: Nanotech/AAC

side view and longitudinal section

a)

cross section

Centerline

Mid section Top view

Support section

b)

Girder cross section

Fig. 3:Tipical dimension of T-type prestressed girder for 25m span (in cm), a) side b) top view Source: Department of Public Works, 1993

400

Taufiq Rochman et al, 2014 Australian Journal of Basic and Applied Sciences, 8(1) January 2014, Pages: 397-406

25m span

Side view

Cross section

Fig. 4: Tipical dimension of T-type non-prestressed girder for 25m span (in mm), Source: Department of Public Works, 1993 Stress Analysis of TAUFIQ-girder: Bridge is analyzed through general purpose finite element software STAADPro v8i 2007, this simplified is such done because the standard and conventional displacement based structural analysis software is sufficient enough to preliminarily analyze the global behaviour of layered composite structures for design purpose, because the loads and the stresses are given external disturbances occurs from the outside environment, hence they depend on their dimension and independent from the material characteristics, where that are not so with the strains and the deflections that come under internal structural responses that closely related with material parameters that fully affect the nominal capacity against that external loads or stresses. Thus, a simple linear 3D analysis is deemed sufficient for the pre-analysis to determine and check the systems dimensions of the static load response. Bridge girders are arranged at a distance of 2000mm, height of 1120 mm and width of 30mm. Typical thickness of bridge deckslab was 200 mm while the thickness of the laminated composite part is approximately set to 30mm thick. Hence the bridge dimension used TAUFIQ-girder system is described in Fig. 5 below:
9600 200 mm 1120 mm

2000

2000

2000

2000

Fig. 5: The dimension of FRP TAUFIQ-Girder system for 25m span Concrete deckslab is modeled into 3D solid brick elements while the FRP girder laminates are modeled into 2D flat shell elements as shown in Fig 6.

401

Taufiq Rochman et al, 2014 Australian Journal of Basic and Applied Sciences, 8(1) January 2014, Pages: 397-406

Flat Shell element

3D solid element

Fig. 6: a) Cross sectional of TAUFIQ-girder system and its element models. b). Integrated full span finite element modeling: 3D solid and flat shell element. Actually, to investigate FRP girder behaviour under pure normal both tension and compression stresses, for instance, the full scale bridge can be concrete in-filled without any shear connector needed, but it was not be performed in this preliminary study, its initial behaviour should be examined firstly to avoid the all nonlinear influence of buckling, slip, torsion and also delamination. There are two loading methods can be choosen. First, using D load, and simplified into uniformly dead load UDL as q=2.2 t/m and knife edge load (KEL) as P=12 ton as shown in Fig. 7.

Fig. 7a: Loading diagram uniformly dead load D (UDL) and knife edge load (KEL) b) Truck axle load T as 2.5 t (25 kN) and 10 t (100 kN) Source: Department of Public Works Secondly, T truck axle load in every axle are 2.5 t and 10 t respectively for front and rear axle. Both of these method agree in almost identical results. Results:

a)

402

Taufiq Rochman et al, 2014 Australian Journal of Basic and Applied Sciences, 8(1) January 2014, Pages: 397-406

b)

(c) Fig. 8: The maximum axial stress Sxx in the mid girder of TAUFIQ bridge system is 78.5 MPa a) top view, b) bottom view c) Stress distribution of Sxx in the interior girder Based on this analysis as shown in Fig. 8, the 25m span FRP girder of 30 mm thick will be adequate to hold tension normal stress as amount of 78.5 Mpa, because according many references that failure stress of unidirectional fiber at angle 0 varies in amount around 500 Mpa.

a)

b)

Fig. 9: System local stresses a) SYY value is predicted as 155 Mpa (in the top view) b) SXY value is predicted as 37.9 Mpa (in the bottom view)

403

Taufiq Rochman et al, 2014 Australian Journal of Basic and Applied Sciences, 8(1) January 2014, Pages: 397-406

Refers to Fig. 9a) and stress diagram in Fig. 10a), compressive normal stress in the compressive region have a large enough value around 100-155 MPa, get closer to the compressive failure value of 120 200 MPa for angle fiber of 90 means there is needed thicker laminate or filled in with the concrete especially in the support region to avoid support failure or even buckling in the web of FRP girder.

Fig. 10: Local stresses a) Syy distribution stresses value in the support region of mid girder b) Sxy distribution stresses value in the support region of mid girder

Fig. 11: Stress value of Sxx of the system real span in the support region 168.7 MPa and 198.25 MPa for compression and tension respectively. Based on shown Fig. 9b) and stress diagram in Fig. 10b), the value of shear strain is large enough as 37.9 MPa close to the 40 MPa limit. It can be seen in Fig. 11 that more detailed local stresses using MIDAS FEA even shows larger value. A local failure will be occurs around the support of 168.7 MPa for compressive stress and 198.2 Mpa for tension stress hence is needed several local stiffener of concrete in-filled in the 30cm length support region. Discussion: Using above preliminary analysis, the model specimen dimension with certain capacity of both loads equipment and test portal then can be determined. In this experiment, TAUFIQ-girder specimens were 2.2m Ushaped (only 2m clear span), consist of two kinds E-glass fiber mat type, unidirectional mat and biaxial mat. All of the web and flange except the bottom flange that hold big amount of tension stress are consist of 9 layers biaxial [45]9 mats, while the bottom flange is consist of 17 layer of [453/04/453/04/453] lay-up configuration means 3 layers of 45 double bias mats and 4 layers of 0 unidirectional mats. General isophthalic polyester unsaturated resin was used as matrix. This system is moulded on plastic-protected wood mold and glued together using vacuum bag injection methods that was better in both resin absorption and its fiber fraction than conventional hand lay-up methods. Bending test was performed in 4 point bending loads, using 0-200 kN range statics load, there are 3 kinds of girder specimens were examinated with various girder depth as 15cm, 12.5cm and 10 cm called A, B and C

404

Taufiq Rochman et al, 2014 Australian Journal of Basic and Applied Sciences, 8(1) January 2014, Pages: 397-406

girders respectively. Below Fig. 12 shows both loading set position and strain gauges, while Fig. 13 a) shows the specimen 4 point loading test, and Fig 13 b) depict the maximum strains.
P

/2

/2
SG1 and SG2

1m

SG1 and SG2

1m

constant moment area

Fig. 12: a) Setting of 4 point loading position

Main tensile b) strain gauges position in stress the midle of the girder 400
Failure stress (MPa)

Main tensile stress 25 9 11 15 19 12 6 17 18 26

300

average= 148.64 MPa

200 1 100 3

29 27

13 14

32 40

0 0 20

Specimen No.

Fig. 13: a) Loading test setup of girder specimen under static bending b) Failure stress average at the mid-bottom coupon specimen
SEMUA TARIK 350
1 250

300

250

R2=0.897
200

Stresses (MPa)

0.32435 1.5

200
Stresses (MPa)
150

150

100

0.4 1.5

100

50

50

0 0 50
o

100

150

20

40

60
o

80

100

Strains ( / oo)

Strains ( /oo)

Fig. 14: a) Proposed equation for main stresses (bottom region) in the coupon specimen; b) Refined equation to fulfill the goodness of fit

405

Taufiq Rochman et al, 2014 Australian Journal of Basic and Applied Sciences, 8(1) January 2014, Pages: 397-406

Strains were automatically recorded using data acquisition system, while loads and deflections were manually recorded. Maximum deflection and its strains is shown in Fig. 15 b).

(a) Fig. 15: Both a) deflections and b) strains tends to be asymptotic at the higher loads

(b)

The great behaviour shown in Fig 15 a) and 15b) for these 4 point bending experiments, the girder tends to shows asymtotic hardening both in deflections and strains in certain higher loads about over 105kN to 200kN. The phenomenon occurs in the deflections that varies from 30 mm to 90mm, and have about 47mm deflection average, while for the strains this phenomenon occurs in the strains varies from 0.0002 to 0.0007, and have about 0.00047 strain average. Almost all of the speciment give almost the same results. Conclusion: A great asymptotic hardening behaviour of TAUFIQ-girder system is obtained at load over than 105 kN to 200 kN, both deflection at the value of 47mm average and strain at the value of 0.00047 describe a very ductile behaviour. This phenomenon is rushed up due to fiber angle coersion to be straightened and collaborated together with uniaxial fiber resist the tension stress region, that fiber more tightened in higher load. This asymptotic strain hardening will contribute a better design safety with CIP (concrete filled in) and without CIP system. As a summary, TAUFIQ girder is better than steel reinforcing against the bending moment due to its lightweight, erection easiness, and its ductility with and without CIP (full FRP section). ACKNOWLEDGEMENTS I would like to express my deep gratitude to Professor Agoes Soehardjono and Assistant Professor Achfas Zacoeb, my research supervisors, for their patient guidance, enthusiastic encouragement and useful critiques of this research work. My grateful thanks are also extended to Mr. Victor Yap, Mr. Muliadi Ho and Mr. Bayu Nugraha from TRIAXIS Composites company industry for their help, tutorial and explanations in doing the preparation for the FRP experimental specimen, and I would also like to extend my thanks to the technicians of the laboratory of the Civil Engineering department for their help in offering me the resources in running the program. Finally, I wish to thank my parents for their support and encouragement throughout my study. REFERENCES Bakis, C.E., L.C. Bank, F.V.L. Brown, M.E. Cosenza, J.F. Davalos, A.M.J.J. Lesko, A. Machida, S.H.F. Rizkalla, And T.C. Triantafillou, 2002. Fiber-Reinforced Polymer Composites For ConstructionState-Of-

406

Taufiq Rochman et al, 2014 Australian Journal of Basic and Applied Sciences, 8(1) January 2014, Pages: 397-406

The-Art Review. Journal Of Composites For Construction. 150th Anniversary Of The American Society Of Civil Engineers (Asce). Bank, Lawrence. 2006. Composites for Construction: Structural Design with FRP Materials. John Wiley & Sons, Inc. Bridge Management System (BMS). 1993. Upper structure standard of prestressed girder. Department of Public Works - Directorate General for Highway Construction (Bina Marga). Burgoyne, C.J., P.R. dan Head, 1993. Aberfeldy Bridge: An Advanced Textile Reinforced Footbridge. TechTextil Symposium, Frankfurt, Germany. Busel, J.P., Composites Industrys Perspective on Transportation Infrastructures Opportunities. Virginia Fiber Reinforced Composites Showcase, Bristol. Darby, J.J., 1999.Role of bonded fibre-reinforced composites in strengthening of structures in Strengthening of Reinforced Concrete Structures, eds. L. C. Hollaway & M. B. Leeming, Woodhead Publishing Ltd, CRC Press LLC, Cornwall, England, pp. 1-10. Oehlers, D. J. 2000, 'Development of Design Rules for Retrofitting by Adhesive Bonding or Bolting either FRP or Steel Plates to RC Beams in Bridges and Buildings. ' In Composites in the Transportation Industry, eds. S. Bandyopadhyay et al., University of New South Wales, Sydney, Australia, pp: 110-119. Elaheh Ghassemieh, 2011. Materials in Automotive Application, State of the Art and Prospects, New Trends and Developments in Automotive Industry, Prof. Marcello Chiaberge (Ed.), ISBN: 978-953-307-999-8, InTech. Fremond, M. dan F. Maceri, 2005. Mechanical Modeling and Computational Issues in Civil Engineering. Lecture Notes in Applied and Computational Mechanics. Springer Khare, K., T. Kant, K. dan Garg, 2004. Free Vibration of Composite and Sandwich Laminates with a Higher Order Facet Shell Element. Composite Structures 65 : 405-418. Maunsell Structural Plastic. Nanni, A., P.C. Huang and G. Tumialan, "Strengthening of Impact-Damaged Bridge Girder Using FRP Laminates," 9th Int. Conf., Structural Faults and Repair, London, UK, July 4-6, 2001, M.C. Forde, Ed., Engineering Technics Press, CDROM version, pp: 7. Nanni, A., T. Alkhrdaji, G. Chen, M. Barker, Y. Xinbao and R. Mayo, 1999. "Overview of Testing to Failure Program of a Highway Bridge Strengthened with FRP Composites," Selected Presentation Proc., 4th International Symposium on FRP for Reinforcement of Concrete Structures (FRPRCS4), Baltimore, MD, pp: 69-80. Qasrawi, Yazan, 2007. Flexural Behaviour Of Spun-Cast Concrete Filled Fibre Reinforced Polymer Tubes For Pole Applications. Master Thesis. Queens University of Canada. Rizzo, A., N. Galati, and A. Nanni, 2005. Strengthening of Off-System Bridges with Mechanically Fastened Pre-Cured FRP Laminates. Skinner, J.M., 2009. A Critical Analysis of The Aberfeldy Footbridge, Scotland. Proceeding of Bridge Engineering 2. Conference in University of Bath, UK. Tarek, 2010. Flexural Behaviour of Sandwich Panels Composed Of Polyurethane Core and GFRP Skins and Ribs. PhD. Thesis. Queens University of Canada. Thomas, S., K. Joseph, S.K. Malhotra, K. Goda and M.S. Sreekala, 2011. Advances in Polymer Composites: Macro- and Microcomposites State of the Art, New Challenges, and Opportunities. Polymer Composites: Volume 1, First Edition. Wiley-VCH Verlag GmbH & Co. KGaA. Zemcik, R. dan R. Rolfes, 2006. High Performance 4-Node Shell Element with Piezoelectric Coupling. Mechanics of Advanced Material and Structures. 13: 393-401.