Finite element analysis of the dynamic behavior of concrete-filled double skin steel tubes (CFDST) under lateral impact with fixed-simply support

Size: px

Start display at page:

Download "Finite element analysis of the dynamic behavior of concrete-filled double skin steel tubes (CFDST) under lateral impact with fixed-simply support"

Geraldine Malone
5 years ago
Views:

1 Structures Uner Shock an Impact XIII 383 Finite element analysis of the ynamic behavior of concrete-fille ouble skin steel tubes (CFDST) uner lateral impact with fixe-simply support M. Liu & R. Wang Architecture an Civil Engineering, Taiyuan University of Technology, China Abstract Concrete-fille ouble skin steel tubes (CFDST) are a type of steel-concrete composite structure. This structure has been wiely use in high-rise builings, briges an power facilities. The current research about it focuses on the performance about axial compression, bias, bening, an low cyclic loas. However, uring the whole life cycle, these structures may inevitably suffer from various impact loas. In orer to unerstan the ynamic response of these structures uner lateral impact, this paper sets up a finite element moel to analyze the estruction moal, impact force (F), eflection ( ) as well as other ynamic responses of CFDST uner lateral impact loa. One en of the CFDST is fixe while the other en is simply supporte. The whole process an failure mechanics of the ynamic response of CFDST uner lateral impact are analyze basically, lai the founation for further parametric analysis an practical calculation metho. Keywors: concrete-fille ouble skin steel tubes (CFDST), lateral impact loa, ynamic response, finite element moel analysis. 1 Introuction Compare with the traitional concrete-fille steel tubes (CFST), concrete-fille ouble skin steel tubes (CFDST) [1] a a smaller concentric steel tube, an fill the space with concrete between the outer steel tube an the inner steel tube. Since the better bening rigiity, expaning cross-section, seismic an fire oi: /susi140331

2 384 Structures Uner Shock an Impact XIII resistance, light weight, CFDST primarily applies as loa-bearing components to subway, gymnasium, support columns of offshore platform, piers an other builings requiring high bearing capability. During the life cycle, these components may inevitably suffer acciental impact. For instance, explosion of meteorite, collision of crashe plane, impact of cars or ships [2]. For example, 50% of briges in Beijing suffere from vehicle collision. The survey ata form Department of Transportation in New York show that there were 520 of 1000 collapse briges ue to impact, while only 19 of 1000 ue to earthquake. Once an accient occurs, the structure may be severely amage or even completely lose bearing capacity, leaing to huge losses of life an property. Therefore, it s necessary to stuy the ynamic behavior of CFDST uner lateral impact loa. Serious literature has been one about CFST s impact performance, e.g. Chen et al. [5], Xiao et al. [6], Huo et al. [7], Bambach [8], Wang et al. [9] an Hou [10]. Currently, research about CFDST mainly focus on the static performance of axial compression, pure bening an the hysteretic behavior of bening components, as well as torsional properties of torsion members [11 14]. The experimental analysis an theoretical research of CFDST uner lateral impact loaing is still relatively less. In this paper, the finite element analysis moel of CFDST uner lateral impact with fixe-simply support is establishe by the ABAQUS finite element software. The work mechanic an full range of the ynamic response are analyze. To verify the correctness of the moel, the finite element analysis moel of CFST uner lateral impact is establishe firstly, the analysis results are compare with existing experimental results. Then the finite element analysis moel of CFDST uner lateral impact is built, an the whole process an failure mechanics are analyze. The results showe that the there is a stable platform of impact force of the CFDST uner lateral impact with fixesimply supporte an the specimen has a goo anti-impact performance. 2 Finite element analysis (FEA) moel Reference to the literature [9], this paper set up the moel of CFDST uner lateral impact with fixe-simply supporte in ABAQUS/Explicit moule (Hibbitt, Karlsson & Sorensen Inc.) [17]. 2.1 Materials A five-phase elastic-plastic constitutive moel introuce by Han [15] is use herein to simulate the steel, take account of the impact of strength ue to the strain rate uner ynamic loaing, Cowper-Symons moel is employe: p / 1 ( / D 1/ s ) where, is strain rate of steel, σ is stress value when strain rate of steel is, σ s is stress of steel uner the static stress, D an p are material parameters. In this paper D=6844s -1 an p =3.91 [10]. For steel, elastic moulus, yiel strength an Poisson s ratio is 2.06e 5 MPa, 298 MPa, 0.3 [9]. For supports, elastic moulus (1)

3 Structures Uner Shock an Impact XIII 385 an Poisson s ratio is 5e 5 MPa, 0.3. For bolt, elastic moulus an Poisson s ratio is 2e 5 MPa, 0.3. The concrete amage plasticity moel is use to simulate the behavior of concrete [15], which consiers the restraints of the outer steel pipe. Meanwhile, the strain rate uner ynamic loaing must be consiere. The moel in CEB- FIP, 1993 [12] is employe: / s ( / s ) 30s (2) f t / f ts ( / s ) ( / s ) 1/ 3 1s 1s where, σ is concrete s compressive strength uner impact loaing, σ s is concrete s static compressive strength, is the ynamic strain rate, s is the strain rate uner static loaing. Uner pressure, s = s -1, while in tension, s= s -1. f t is tensile strength of concrete uner impact loaing, f ts is tensile strength of concrete uner static loaing. An =1/(5+9f c /f co ), f c =f ck +8(N/mm 2 ), f co =10(N/mm 2 ), where f ck is the characteristic value of concrete compressive strength. log, (1+8f c /f co ); In this paper, for concrete, aopt C40, E c =4370f 1/2 c MPa, Poisson s ratio is 0.2. Also, since the inertial force, the ensity of steel is taken as 7850 kg/m 3,an the ensity of concrete is 2500 kg/m Interface contact The contacts of steel hammer, steel concrete, steel supporters an supporters bolts are mae of general contact, which in normal take har contact, in tangential take the Columbia Friction moel to simulate the relative slip. The friction factors of Coulomb friction were recommene by Han [15]. 2.3 Bounary an initial conitions Figure 1 show the fixe-simply bounary conition. The shapes of simply supporter an fixe supporter are shown in figure 1. In the fixe supporter, the upper part is connecte with the lower part by eight bolts with 20mm iameters. The bottoms of imply supporter an fixe supporter are fully constraine In this moel, hammer is 1mm up to the upsie of the specimen, its mass is 203.9kg an the initial velocity of hammer is given accoring to ifferent impact energy. 2.4 Element type an mesh In the FEA moel, steel tubes are simulate by using 4-noe shell elements with reuce integration, hammer uses the 4-noe iscrete rigi shell elements with reuce integration, while the other parts are moele by the 8-noe soli elements (3)

386 Structures Uner Shock an Impact XIII After comparing the mesh quality an calculation time of ifferent mesh sizes, this moel aopt local refinement, in the non- encrypte area the mesh size is 20mm

3 Results of finite element analysis (FEA) Four circular CFDST members are simulate, the length of all specimens is 1500mm, the iameters of outer steel tube an inner steel tube are 170mm an 100mm

The etaile information of each specimen is presente in Table 1, where F for fixe supporters an S for simply supporters, while the number enotes the height of the hammer.

4 386 Structures Uner Shock an Impact XIII After comparing the mesh quality an calculation time of ifferent mesh sizes, this moel aopt local refinement, in the non- encrypte area the mesh size is 20mm an in the encrypte area the mesh size is 5mm. Figure 2 shows the mesh of this FEA moel. Bolt Specimen Hammer Fixe-supporter Inner Concrete Simply-supporter Test be Outer Figure 1: Abaqus moel an mesh. 3 Results of finite element analysis (FEA) Four circular CFDST members are simulate, the length of all specimens is 1500mm, the iameters of outer steel tube an inner steel tube are 170mm an 100mm respectively, an the thickness of outer steel tube an inner steel tube are 3mm an 2mm respectively. The etaile information of each specimen is presente in Table 1, where F for fixe supporters an S for simply supporters, while the number enotes the height of the hammer. Take s-fs3 for example, enotes that the hammer rop at a height of 3m uner fixe an simply supporte. Then, H, V 0 an E 0 represent the height, the velocity, an the energy of the impact, F stab, t represent plateau value an time of the impact force, an is the ultimate lateral eflection of specimen at mi-span. Table 1: Specimen information. No. H(m) V 0 (m/s) E 0 (kj) F stab (kn) t(ms) (mm) FS FS FS FS The failure moes In Figure 2 the preicte typical failure moel of CFDST is shown. It can be observe that the failure moel inclue the integer bening eformation an local buckling of steel tube on the upsie of CFDST at the mi-span. A Figure 2: A Failure moes of CFDST member. A-A

5 Structures Uner Shock an Impact XIII Impact force an eflection Figure 3 compares the impact force between ifferent impact energy of ynamic loas an static loas. It can be seen that with the same static flexural capacity(m u ), the resistant capacity of the specimen uner impact is 4% 14% higher than that uner static impact. M u reference to the practical formula propose by Hang Hong [16], an it is relate to the section an material only. M u =γ m1 W scm f scy +γ m2 W si f yi (4) As can be seen from the curves of figure 3, plateau value (F stab ) an time (t) of the impact force increases along with the rise of the velocity (V 0 ). An at the impact height of 9m there is a istinct peak impact force. It certifies the amage of CFDST uner impact is uctile failure, these specimens have well anti-impact performance. Impact force/kn Static s-fs3 s-fs5 s-fs7 s-fs9 Figure 3: Time/s Contrast of F stab uner static an impact loa. Figure 4 compares the lateral eflection of specimen at mi-span uner ifferent impact velocity in FEA moel. This figure shows that the eflection ( ) increases with the rise of the velocity (V 0 ). Reference to the curves, the members appeare total eformation in 8ms, the specimens reboune after the max eflection until final equilibrium. 3.3 Dynamic response analysis of the whole process Take the simulate result of specimen s-fs9 as the example to analyze the work mechanic an full range of the ynamic response.

6 388 Structures Uner Shock an Impact XIII Deflection/mm FS3 FS5 FS7 FS9 Time/s Figure 4: Contrast of eflection uner ifferent height. Figure 5 shows the imensionless results of impact force (F), eflection ( ), the velocity of hammer (V 0 ) an the velocity of member (V). Obviously, these time-curves can be ivie into the following four stages: Impact force/kn A B C D E CF U V(hammer) v(member) Time/s Figure 5: Impact force (F), eflection ( ), velocity (V) versus time. First, peak stage (from point A to B): During this stage, the impact force increases rapily, the velocity of the hammer ecreases while the velocity of the CFDST member increases. At the en of this perio, the hammer an the member achieve the same spee, the impact force rises to peak, meanwhile the member s initial eformation appears.

7 Structures Uner Shock an Impact XIII 389 Secon, escent stage (from point B to C): At this stage, the spee of hammer continues to reuce while the spee of member continues to increase until to the maximum spee. Due to elastic recovery of the local eformation, the hammer an specimen on t separate, but the impact force eclines for a short time. Thir, Recovery Platform stage (from point C to D): In this phase, the spee of hammer ecreases rapily an reboun relatively, which cause the impact force come to peak again. The spee of hammer an CFDST member get convergence at the platform of the impact force. Last, unloaing stage (from point D to E): At the point D, the spee of hammer an CFDST member reuce to zero, the isplacement of the specimen rise to the limits. During this stage, the CFDST member rebouns while the impact force graually unloae to zero. 4 Conclusions 1) With the same static flexural capacity (M u ), the resistant capacity of the specimen uner impact is 4%-14% higher. Plateau value (F stab ) an time (t) of the impact force increases with the rise of the velocity (V 0 ).The eflection ( ) increases with the rise of the velocity (V 0 ). 2) The failure moel inclue the integer bening eformation an local buckling of the steel tube on the upsie of CFDST at the mi-span 3) The full range of the ynamic response base on FEA moel proves that time-curves can be ivie into four stages: Peak stage, Descent stage, Recovery Platform stage, Unloaing stage. References [1] Tao Z, Yu Q. New composite columns: Test, Theory an Methos. Sciencep. BJ [2] Wang R. Dynamic response an amage failure of CFST uner lateral impact. PhD thesis of Taiyuan University of Technology [3] Lu XZ, He ST, Huang SN. Failure Mechanism, Design methos an protective measures of superstructure uner impact. China Architecture & Builing Press. BJ [4] Maria Garlock, Ignacio Paya-Zaforteza, Venkatesh Kour, Li G. Fire hazar in briges: Review, assessment an repair strategies. Engineering structures, 35(07), pp , [5] Chen ZY, Luo JQ, Pan XW. Performance of CFST as protective structures. Earthquake an Antiknock Engineering Research Center of Tsinghua University. BJ [6] Xiao Y, Shan JH, Zheng Q, Chen BS, Shen YL. Experimental stuies on concrete fille steel tubes uner high strain rate loaing. Journal of Materials in Civil Engineering, 21(10), pp , [7] Huo JS, Ren XH, Xiao Y. Impact behavior of concrete-fille steel tubular stub columns uner ISO-834 stanar fire. China Civil Engineering Journal. 45(04), pp. 9-20, 2012.

8 390 Structures Uner Shock an Impact XIII [8] Bambach M R. Design of hollow an concrete fille steel an stainless steel tubular columns for transverse impact loas. Thin-Walle Structures, 49(10), pp , [9] Wang Rui, Han LH, Hou CC. Behaviour of concrete fille steel tubular (CFST) members uner lateral impact: experiment an FEA moel. Journal of construction steel research, 80 (01), pp , [10] Hou CC. Stuy on Performance of Circular Concrete-fille Steel Tubular (CFST) Members uner Low Velocity Transverse Impact. Journal of Tsinghua University [11] Tao Z, Han LH. Development in the research of concrete fille oubleskin steel tubes [J]. Journal of Harbin Institute of Technology. 35, pp , [12] Huang H, Tao Z, Han LH. Mechanism of concrete-fille ouble-skin steel tubular columns (CHS inner an CHS outer) subjecte to axial compression [J]. Inustrial Construction Vol. 36, No. 11, [13] Chen XJ, Chen MC, Huang C. Hysteretic moel of concrete-fille oubleskin steel tubular bening members. Railway Engineering. pp , 2011(4). [14] Huang H, Chen MC, Huang BJ. Twist stuy of concrete-fille ouble-skin steel tubular. Journal of experimental mechanics 27(3), pp , [15] Han LH. Concrete-fille steel tubular structures theory an application. Sciencep. BJ [16] Huang H. Behavior of concrete fille ouble-skin steel tubular beamcolumns. PhD thesis of Fuzhou University [17] Hibbitt, Karlsson & Sorensen Inc ABAQUS/Explicit User s Manual. Vision Hibbitt, Karlsson & Sorensen Inc; 2005.

OF CHS. associated. indicate. the need. Rio de Janeiro, Brazil. a) Footbridge Rio. d) Maria Lenk. CHS K joints

OF CHS. associated. indicate. the need. Rio de Janeiro, Brazil. a) Footbridge Rio. d) Maria Lenk. CHS K joints EUROSTEEL 2, August 3 September 2, 2, Buapest, Hungary A NUMERICAL EVALUATION OF CHS T JOINTS UNDER AXIAL LOADS Raphael S. a Silva a, Luciano R. O. e Lima b, Pero C. G. a S. Vellasco b, José G. S. a Silva