Test Unit W2 willbe discussed. The initial stiffness in the simulation of Test UnitW2 which is displayed in Fig. 11a with a dotted line accuratelymatches the one observed in the test. The shape of the simulatedhysteresis loops and the computed strength of the test unit are alsoin fair agreementwith the experimental results. Only in the secondcycle at a displacement ductility of 2, the numerical simulationclearly underestimates the maximum strength. This is a directconsequence of the inability of the nonlinear isotropic/kinematichardening model of the steel to match the hysteretic behavior  of the real reinforcing bars in the strain range occurring duringreloading to ductility 2 [4].A further comparison between numerical and experimentalresults is shown in Fig. 12where the top displacement of Test UnitsW2 and W3 is broken down into its three components. In orderto compute the displacement components from the numericalresults virtual instruments were used. In the numerical modelnodes were defined in correspondence with the fixture of theLVDTs that were mounted on the test units (see Fig. 6b). From therelative displacement of these nodes, virtual readings of the LVDTscould be computed. Afterwards, the displacement componentswere computed using the same procedure as described in Section 3for the experimental data.Due to the proposed structural system for the HFC structuralwalls, the largest part of the top displacement is made up bythe fixed-end component. Its hysteretic behaviour is stronglygoverned by the longitudinal reinforcement and its predictionis quite accurate over the entire deformation range. On theother hand the hysteretic behaviour of the flexural and theshear displacement components is strongly affected by the cyclictensile behaviour of HFC and as described in Section 4.1 thisbehaviour is only known and modelled with large uncertainties.The simulation of Test Unit W2 was able to accurately predict theflexural deformation up to first cracking of the wall. Afterwards,the behaviour of the numerical model was too stiff leading tomaximum flexural deformations (Fig. 12e) of about only 50% ofthe flexural deformations computed from the experimental data(Fig. 12b). A similar observation holds true for the flexural (Fig. 12hand k) and shear (Fig. 12i and l) deformations of Test UnitW3, thisdespite Fig. 10 showing a good agreement between the 3-pointbending tests and their simulation which is also governed by thetensile behaviour of HFC. Owing to the large uncertainties involvedin such a procedure, it was not deemed reasonable and beyondthe scope of this investigation to further improve the numericalsimulation by iteratively adjusting the parameters defining thecyclic tensile behaviour of HFC such as to match the overallbehaviour of the test units.5. Conclusions and outlookHybrid Fibre Concrete (HFC) structural walls are able to un-dergo large inelastic deformations while ensuring an easier con-structability and superior post-earthquake functionality compared to conventional reinforced concrete walls. The tests and the nu-merical simulations on threeHFC structuralwalls allowthe follow-ing observations:1.
It was possible to build structural walls without any transversereinforcement for shear, confinement and stabilisation of thelongitudinal reinforcement. The HFC used for this purpose hada fibre volume content between 3.5% and 6.0%. Due to the tightcontrol of the rheology, the HFC had self-compacting propertiesdespite the high fibre content.2. The HFC structural walls were able to reach ultimate displace-ment ductilities in excess of 8 corresponding to drifts rangingbetween 3.2 and 4.2% which is comparable to the deformationcapacity of well-detailed capacity designed reinforced concrete(RC) walls and is larger than the capacity demand required tosurvive most severe earthquakes. Furthermore, the deforma-tion capacity of HFC walls can easily be adjusted by changingthe length of the sleeves placed onto the longitudinal reinforc-ing bars.3. Provided that the thickness of the cover concrete was largeenough to accommodate the biggest fibres, the concrete coverdid not spall and thereby buckling of the flexural reinforcementwas prevented.4. Test Units W2 and W3 did not suffer any significant structuraldamage up to failure. However, residual displacements uponunloading were of the same order of magnitude as thoseexperienced by RC walls, and they are of course affecting thepost-earthquake functionality and reparability of HFC walls. Inorder to fully exploit the excellent properties of HFC, structuralsystems characterized by small residual displacements shouldbe investigated.5. For all test units, the cracks remained small (<0.3 mm)throughout the whole test and the performance in shear wasvery satisfactory. Test Units W1 and W2 were able to sustainnominal shear stresses up to 2.4 MPa and Test unit W3 up to7.1 MPa without failure. This latter value should, however, beinterpreted with caution because of the barbelled section andthe relatively low aspect ratio of Test UnitW3. Additional testsare needed in order to define criteria for the shear failure of HFCwalls and other HFC structural elements.6. The global hysteretic behaviour of HFCwalls could be simulatednumerically. However, an accurate simulation could only beobtained after introducing changes in the models that werebased on engineering judgment rather than evidence. In orderto allow for better modelling of the cyclic global and localbehaviour of HFC structural elements further research aimingat the characterization of the tensile behaviour of HFC underaxial cyclic loading is needed.AcknowledgementsThe Hybrid Fibre Concrete used for the tests presented in thispaper was developed by Mr. Patrick Stähli from the Institute forBuilding Materials (IfB) at the ETH Zurich working under theguidance of Prof. Dr. Jan G.M. van Mier.The test units were built and tested in the laboratories ofthe ETH Zurich. Mr. Thomas Jaggi, Mr. Heinz Richner and Mr.Patrick Stähli from the Institute for Building Materials (IfB) wereinstrumental in the construction of the test units and also playedan essential role in preparing and testing all material samples.Mr. Markus Baumann designed and implemented the test controlscheme. His contributionwas also essential for themanagement ofthe test unit instrumentation and for the actual testing of all testunits. Mr. Christoph Gisler manufactured several special parts forthe test setup and helped assembling them.The ductile D5.2 reinforcing barswere donated by Ferriere NordSpA, Gruppo Pittini, 33010 Osoppo, Italy.References[1] Paulay T, Priestley MJN. Seismic design of reinforced concrete and masonrybuildings. New York: JohnWiley & Sons; 1992.[2] Priestley MJN, Seible F, Calvi GM. Seismic design and retrofit of bridges. 非线性循环行为混杂纤维混凝土结构墙英文文献和中文翻译(6):http://www.751com.cn/fanyi/lunwen_57270.html