Influence of Microstructure Difference of Stainless Steels on the Strength and Fracture of Bolted Connections

Transcription

1 , pp Influence of Microstructure Difference of Stainless Steels on the Strength and Fracture of Bolted Connections TaeSoo KIM 1) and TaeJun CHO 2) * 1) Deptartment of Architectural Engineering, Hanbat National University, Daejeon, Korea. 2) Department of civil engineering, Daejin University, Sundan-dong, Pocheon, Gyeunggi-do, Korea. (Received on May 25, 2012; accepted on October 12, 2012) Recently, since stainless steel has a variety of significant characteristics such as corrosion resistance, durability, aesthetic appeal and fire resistance, the use of stainless steel in construction has been steadily increasing. Also, stainless steels can be classified into five main groups; austenitic, ferritic, martensitic, duplex and precipitation hardening according to metallurgical microstructure. Austenitic stainless steels are the most widely utilized in buildings. This paper focused on comparing material properties and ultimate behaviors such as strength, fracture mode and curling effect of single shear bolted connections fabricated with two different stainless steels; austenitic and ferritic types. Bolt arrangement and end distance in the parallel direction of applied load are considered as main variables. Specimens have same edge distance perpendicular to the direction of load, plate thickness, bolt diameter, pitch and gauge. A monotonic tensile test has been carried out for specimens under shear and some bolted connections with long end distance were accompanied by curling; out of plane deformation, which caused strength reduction. Furthermore, it is found that the fracture shape and curling effect patterns of two kinds of bolted connections differ from two stainless steel materials. KEY WORDS: austenitic stainless steel; ferritic stainless steel; bolted connection; curling; strength reduction. 1. Introduction Since stainless steel has a variety of significant characteristics such as corrosion resistance, durability, aesthetic appeal and fire resistance, the need of stainless steel in construction has been steadily increasing. Stainless steels can be classified into five main groups; austenitic, ferritic, martensitic, duplex and precipitation hardening according to metallurgical microstructure. 1,2) Austenitic stainless steels are the most widely utilized in buildings. Most of studies have also focused on the behaviors of these austenitic types. In the early 1960s, Johnson, and Winter 3) sponsored by American Iron and Steel Institute (AISI) initiated a set of researches regarding the structural behaviors of stainless steel columns and beams in order to provide the design specification of austenitic type stainless steel structural members. The first edition of Specification for the design of light gage coldformed stainless steel structural members (1968) 4) and Stainless steel cold-formed structural design manual (1974) 5) were published by AISI. In the early 1980s, South African steel company developed the ferritic stainless steels such as Types 409, 430, 3Cr12. Since then design criteria development for ferritic stainless steels had been made by Merwe 6) and Berg 7) and as a result, Specification for the design of cold-formed stainless steel structural members * Corresponding author: cho_taejun@hanmail.net DOI: (ASCE 8-90) including austenitic and ferrictic types was also published by the American Society of Civil Engineers (ASCE) in ) The latest editions of design standards for cold-formed stainless steel members have been specified by Structural Engineering Institute (SEI)/ASCE; SEI/ASCE-8-02 (2002)[2], Australian/New Zealand Standards; AS/NZS 4673 (2001), 9) Stainless Steel Building Association of Japan (SSBA) (2006) 10) and Eurocode 3 Part 1.4 (2006). 11) Kuwamura et al. had conducted the experimental research about ultimate behaviors of cold-formed austentic stainless steel (SUS 304 in Japan Industrial Standards; JIS) shear bolted connections with variables such as bolt arrangement end distance and plate thickness. 12,13) It is known that curling; out of plane deformation in the vertical direction of load was observed for specimens with a relatively thin walled steel plate and long end distance and the curling resulted in sudden strength reduction. Subsequently, Kim et al. have verified the applicability of finite element analysis for predicting the fracture mode and ultimate strength of austenitic stainless steel bolted connections and investigated the influence of curling on structural behaviors ) In addition, mechanical behaviors for bolted connections in carbon steel and in stainless steel using finite element analysis have compared and it is found that connections with austenitic stainless showed a higher strength reduction rate than connections with carbon steel and stress distribution was different each other. 18) It has been reported that stainless steels exhibit different ISIJ

2 properties in tension, compression, anisotropy and work hardening capability according to alloy and heat treatment. Berg conducted research on the local buckling of partially stiffened Z and hat sections made from 304, 430 and 3Cr12 stainless steels in order to investigate the influence of these material characteristics on compressive members 7) and Lecce et al. have focused on the experimental and numerical investigation of cold-formed steel (austenitic 304, ferritic 430 stainless steel and ferritic-like 3Cr13 chrominum weldable steel) sections subject to distortional buckling under compression. 19,20) Strength design guidelines of compressive members with local bucking and distortional buckling modes were proposed for austenitic and ferritic stainless steels. However, the influence of the difference of stainless steel material properties on the strength and fracture mode of single shear bolted connections has not been discussed till now in calculating the ultimate strength. Austenitic stainless steels contain 18% chromium and 8% nickel as a principle alloy content. Especially, since nickel is a high-priced element, austenitic stainless steel is considerably expensive compared to ferritic stainless steel and martensitic stainless steel. For those reasons, ferritic stainless steels which do not contain nickel and are low-priced have been recently applied to many buildings as structural members. Ferritic stainless steels have chromium in the range of 11.5 to 18 percent as major alloy and no nickel except American Society for Testing and Materials (ASTM) type ,22) Even if corrosion resistance, ductility, formability and weldability are not as good as in the austenitic stainless steels, ferritic stainless steels have also good corrosion resistance in mildly corrosive environments and superior resistance to stress corrosion cracking. In this paper, the specimens of single shear bolted connections with two types of bolt arrangement were planned and tested for austenitic 304 (SA series) and ferritic 430 (SF series) stainless steels. The purpose of this paper to compare the ultimate behaviors such as curling occurrence, ultimate strength reduction by curling, fracture mode and strain distribution of single shear bolted connections with two different stainless steels. have a fixed long edge distance (b=60 mm) perpendicular to the direction of load in order to fail by shear-out fracture or block shear fracture instead of the net section tensile fracture. Specimens have one test part and two rigid parts described in Fig. 2. Test part of specimen is fastened with rigid plates for coupling, which have two kinds of high strength bolts (F10T, equivalent to AISI A490) of 12 mm and 20 mm diameter as depicted in Fig. 2. Test part is fabricated with bolt of 12 mm diameter (d ), bolt hole of 13 mm diameter (φ), pitch (p) and gage (g) of 36 mm (=3.0d), nominal plate thickness (t n=3.0 mm) as constant dimension and end distances (e=2.0d, 2.5d, 3.0d, 4.0d, 5.0d) as main variables. Specimens are listed in Table 1. Figure 3 shows the set-up of specimens and location of measuring gages. The (a) Plane figure of test part and rigid part 2. Experimental Investigation 2.1. Specimen Plan and Set-up Specimens; single shear bolted connections are fabricated by two types of bolt arrangement; 1 2 (SA2 and SF2 series) and 2 1 (SA3 and SF3 series) array as shown in Fig. 1 and (a) SA2 and SF2 series (1 2 bolt array) (b) SA3 and SF3 series (2 1 bolt array) Fig. 1. Geometry and bolt array of specimen. (b) Assembly and side elevation of test part and rigid part Fig. 2. Assembly of specimens (SA2 and SF2 series) ISIJ 366

3 Table 1. List of specimen. Table 2. Chemical composition of stainless steel. Specimen SA2T30E24 Bolt array row column Nominal plate thickness t n [mm] End distance e[mm] SA2T30E30 30 SA2T30E SA2T30E48 48 SA2T30E60 60 SA3T30E24 24 SA3T30E30 30 SA3T30E SA3T30E48 48 SA3T30E60 60 SF2T30E24 24 SF2T30E30 30 SF2T30E SF2T30E48 48 SF2T30E60 60 SF3T30E24 24 SF3T30E30 30 SF3T30E SF3T30E48 48 SF3T30E60 60 Fig. 3. Common facotrs location of loading center (the center of plate thickness) in test part with 3.0 mm thick plate is planned to coincide with that of clamping part (combination of 20 mm thick plate and 37 mm thick plate as given in Fig. 2). In other words, when specimen is set up in the testing machine (UTM), if the 24 (a) Transducers (SA2T30E60) Bolt diameter (d)=12 mm Bolt hole (ø) =13 mm pitch (p), gauge (g)=36 mm edge distance (b)=60 mm SA2 and SF3 series (b) Strain gages Set-up of specimen, transducers (LVDTs) and strain gages. Steel Material Specification Chemical composition C Si Mn P S Ni Cr KS D STS304 ASTM A KS D STS430 ASTM A thickness center of test part (3.0 mm thick plate) is identical to that of clamping rigid part (37 mm thick plate), the eccentricity in geometry and load will not occur during test. For specimen notation, for example, SA2T30E24, SF3T30E24 in Table 1, SA and SF mean austenitic and ferritic, respectively, second, 2 and 3 are 1 2 and 2 1 bolt array, T30 is plate thickness ( T30 =3.0 mm) and E24 is end distance ( E24 =24 mm) Material Properties Steel material types for specimens are austenitic; STS304 (equivalent to ASTM type 304) and ferritic stainless steel; STS430 (equivalent to ASTM type 430). 1,21) The chemical composition of the stainless steels is summarized in Table 2 based on Korean standards (KS) 1) and ASTM. 2) ASTM and KS requirements for yield strength in both STS340 and STS 430 are 206 MPa and 205 MPa, respectively. Tensile coupon test results and nominal stress-strain curves are summarized in Table 3 and Fig. 4, respectively. It can be seen that apparent yield point was not exhibited on the stress-strain curves of material test. ASTM A370 specifies that the yield strength of material is defined as 0.2% proof stress. 22) In the specification of Japan Industrial Standards (JIS, JIS Z 2241: Method of tensile test for metallic materials and JIS G 4321: Stainless steel for building structure), 23,24) the yield strength of stainless steel is defined by 0.1% offset method. Since JIS definition may be reasonable to provide a sound deflection control at the serviceability limit state and more conservative value in estimating the width to thickness ratio for plate local bucking strength, 0.1% offset (proof) method in this paper is adopted for deciding the yield strength of stainless steels. 22) Austenitic stainless steel has yield strength (0.1% offset method) of MPa, tensile stress of MPa, yield ratio (F y/f u; ratio of tensile stress F u to yield strength F y) of 36.46% and elongation of 61.6%. For ferritic stainless steel, yield strength is MPa, tensile stress is MPa, yield ratio is 69.15% and elongation is 38%. Austenitic type 304 has lower yield strength by 17% and higher tensile stress by 60% than ferritic type 430. In other words, ferrtic type has 1.9 times yield ratio (YR=F y/f u) compared to that of austenitic type. The elongation of austenitic type is lager by 60% than that of ferritic type. From this comparison, it can be found that austenitic type, 304 stainless steel has excellent strength enhancement (F u/f y=2.74, extensive strain hardening) and ductility until material reaches tensile strength after yielding compared with ferritic type 430 (F u/f y=1.45) because the high work-hardening exponent of austenitic stainless steel (type 304) is attributable to transformation induced plasticity (TRIP) ISIJ

4 Table 3. Tensile coupon test results. Coupon Actual plate thickness te [mm] E [GPa] 0.1% offset yield stress Fy [MPa] Tensile stress Fu [MPa] Yield ratio YR Fy/Fu [%] Fu/Fy [%] Elongation EL [%] SAT SAT SAT Mean C.O.V SFT SFT SFT Mean C.O.V Fy_304/ Fy_430 (0.1%) Fu_304/ Fu_ Fig. 4. (a) Complete range of strain (b) Initial range (strain: 1%) (c) Initial range (strain: 10%) Stress-strain curves of STS304 (SAT30-1) and STS430 (SFT30-1). 3. Experimental Procedures and Results 3.1. Experimental Procedures Clamping part and test part of specimens are clipped by upper and lower jigs of UTM as given in Figs. 2 and 3. A tensile force was applied to the specimen with a displacement control mode in order to obtain load-displacement (stroke) curve. The enforced displacement (stroke) parallel to the direction of loading was obtained using the average of two transducers (LVDTs) attached in test machine as shown in Fig. 3(a). One transducer was also used to measure curling displacement (out of plane deformation), which was placed at the location 30 mm apart from the center of bolt bole toward plate end of specimen as displayed in Fig. 3(a). Uniaxial strain gage was attached to plate surface of each specimen. The location of the strain gage (G1) is shown in Fig. 3(b). Gage, G1 in all of specimens were located at point, 30 mm apart from the center of bolt hole (No. 1 bolt) in order to observe the strain change of load direction (1- axis direction, E11) during test Fracture Mode and Ultimate Strength of Test Results Figures 5 and 6 exhibit fracture shapes of specimens and load-displacement curves, respectively obtained from test results. Also, ultimate strength, curling occurrence, fracture mode and strength ratio based on ultimate strength of specimens with the shortest end distance were summarized in Table 4. Fracture modes are classified into two types according to ultimate state and test end point. It should be noted that fracture modes at the end point of test did not necessarily coincide with those at ultimate state (ultimate strength) of bolted connection. For SA2 and SF2 series with 1 2 bolt array, all specimens except SA2T30E24 showed block shear fracture mode. Although specimens, SA2T30E24 and SF2T30E24 had same end distance (e=24 mm), the fracture modes of two specimens were different each other; SA2T30E24 failed in shear-out fracture (see Fig. 5(a)) and SF2T30E24 showed a typical block shear fracture (see Fig. 5(b)). This is because material properties (elongation, yield ratio) and microstructure (atom arrangement) between austenitic stainless steel and ferritic stainless steel are different. Austenitic stainless 2013 ISIJ 368

5 SA2T30E24 (Shear out) (a) SA2 series SA2T30E48, SA2T30E60 (Curling+tensile fracture) (a) SA2 Series SF2T30E24 (Block shear) (b) SF2 series SF2T30E48 (Block shear) SF2T30E60 (Curling+tensile fracture) (b) SF2 Series SA3T30E36 (Shear fracture+curling) SA3T30E60 (Shear fracture+curling) (c) SA3 Series (c) SA3 series SF3T30E36 (Shear out) SF3T30E60 (Shear out+curling) (d) SF3 Series Fig. 5. Fracture shapes of specimens in each series. steels (high ductility and elongation) can be easily deformed under loading when compared with ferritic stainless steel (low ductility and elongation). Therefore, when external force is applied to bolted connections, for SA2T30E24 fabricated with austenitic stainless steel, stress and deformation (strain) in the area adjacent to bolt can be concentrated in the parallel direction of loading and consequently, shear out fracture parallel to the direction of loading occurred at first. (d) SF3 series Fig. 6. Load-displacement curves of test results. On the contrary, for SA2T30E24 fabricated with ferritic stainless steel, since stress and strain are concentrated between two bolt holes with smaller critical section area ISIJ

6 Table 4. Test results for specimens. Specimen Ultimate strength P ue [kn] Curling occurrence Fracture mode at ultimate strength Fracture mode at the end of test Strength ratio Pue/Pue (emin) SA2T30E Shear fracture of plate end Shear-out fracture (E) 1.00 SA2T30E Tensile fracture between 1.17 SA2T30E two bolts perpendicular to the direction of load Block shear 1.24 fracture (BS) SA2T30E Tensile fracture SA2T30E SA3T30E Shear fracture of plate end 1.00 SA3T30E Shear-out SA3T30E Shear fracture between two bolts 1.06 fracture (E) SA3T30E parallel to the direction of load 0.96 SA3T30E SF2T30E SF2T30E Tensile fracture between 1.15 two bolts perpendicular Block shear SF2T30E to the direction of load 1.23 fracture (BS) SF2T30E SF2T30E Tensile fracture+curling 1.69 SF3T30E SF3T30E Shear fracture of plate end 1.15 SF3T30E Shear-out fracture (E) 1.28 SF3T30E Shear fracture between two bolts 1.25 SF3T30E parallel to the direction of load 1.36 against applied force, tensile fracture occurred first, followed by shear fracture. Especially, specimens such as SA2T30E48, SA2T30E60, and SF2T30E60 with a relatively long end distance were accompanied by curling (refer to Figs. 5(a) and 5(b)). The curling occurred prior to shear fracture toward the end of plate (refer to Figs. 6(a) and 6(b)). Curling is assumed to be a kind of local buckling of thin-walled plane plate (out of plane deformation) which occurs in the perpendicular direction of applied load due to the bearing action of bolt. Where, block shear fracture is defined by the combination of tensile fracture between two bolts and shear fracture of two gross sections as can be seen in Fig. 5(b). Specimens of SA3 and SF3 series with 2 1 bolt array showed shear-out fracture mode at the end of test. All specimens of SA3 series were accompanied by curling and failed in shear fracture as shown in Figs. 5(c) and 6(c). For specimens, SF3T30E24, SF3T30E30 and SF3T30E36 with a short end distance (e=24, 30, 36 mm), typical two near-longitudinal shear out fractures (refer to Figs. 5(c) and 5(d)) extending from the plate part between two bolts to the end of plate were observed. However, for SF3T30E48 and SF3T30E60 with a long end distance (e=48, 60 mm), curling occurred before ultimate state and shear fracture governed the ultimate state as can be seen in Figs. 5(d) and 6(d), where the curling caused the transient strength reduction on the curves. It is also known that even if the geometry and bolt array of bolted connection are identical, behaviors such as fracture mode, curling occurrence, curling effect and loaddisplacement curve pattern can be different from mechanical properties of stainless steel types; austenitic and ferritic as can be seen in Figs. 5, 6 and Table 4. For instance, in specimens with 24 mm end distance, SA2T30E24 fabricated with austenitic stainless steel showed shear fracture at ultimate state (see the left side of top in Fig. 5(a)), on the contrary, for SF2T30E24 fabricated with ferritic stainless steel, the ultimate state was determined by tensile fracture between two bolts and block shear fracture mode was observed at test end (see the left side of top in Fig. 5(b)). In addition, it is known that austenitic type specimens (SA series) was vulnerable to curling occurrence compared with ferritic type specimens (SF series) as given in Table 4. Since austenitic stainless steel has an excellent ductility (elongation) and low yield strength compared to ferritic stainless steel as known in Table 3 and Fig. 4(a), in-plane and outof-plane deformations are more likely to be concentrated around the bolt group and the stress condition of connected plate reaches first the yield stress of material in case of bolted connections (SA series) with austenitic stainless steel. Table 4 contains test results such as ultimate strength (P ue) and strength ratios (P ue/p ue(emin)) according to end distance. The values of P ue/p ue(emin) mean the ratio of ultimate strength of specimen (P ue) to ultimate strength of specimen with the minimum end distance (P ue(emin), e=24 mm) based on the same bolt arrangement. For specimens with no curling occurrence, ultimate strengths (Pue) and strength ratios (P ue/p ue(emin)= ) in Table 4 get larger as the end distance (e) increases. However, since specimens with remarkable curling showed a sudden strength drop caused by the curling as can be seen in Fig. 6, there seemed to be no big increase in ultimate strength. Especially, all specimens of SA3 series had curling occurrence, ultimate strengths (P ue= kn kn) of SA3T30E48 and SA3T30E60 were lower than those (P ue= kn kn) of specimens with shorter end distances ranged from 24 mm to 36 mm (see Table 4). For 2013 ISIJ 370

7 curled specimens of SA2 series and SF3 series with 48 mm and 60 mm end distance, the increase of ultimate strength was not high. It can be found like fracture mode that influence of curling occurrence on ultimate strength of single shear bolted connection and structural behaviors between curled specimens and uncurled specimens differed from mechanical properties and microstructure of stainless steel types. Table 6 exhibits the comparison (P uesa/p uesf) of ultimate strengths for specimens with same bolt arrangement between austenintic (SA series) and ferrictic (SF series) type. In Table 4, mean tensile strength ratio (F usa/f usf) of austenitic type (STS304) to ferrittic type (STS430) was Influence of Curling Occurrence on Structural Behaviors Curling Effect on Fracture Mode and Ultimate Strength As already mentioned in section 3.1, curling (out of plane deformation) occurrence perpendicular to the direction of applied force was affected by stainless steel types and also led to the ultimate strength reduction of single shear stainless steel bolted connections. Deformed shape and fracture development taken at specified displacement points during test are displayed in Fig. 7 in order to grasp the influence pattern of curling on the ultimate strength of bolted connection according to the microstructure of stainless steel. In addition, Table 5 presents the additional investigation results regarding enforced displacement and strength obtained from two critical points; one is for ultimate state and the other is for strength drop by curling for curled specimens given in Table 4. For curled specimens with 1 2 bolt array, the curling occurrence in austenitic stainless steel bolted connection (SA2T30E48 and SA2T30E60) had an influence on strength reduction but did not decide ultimate strength as can be seen in Table 5, Figs. 6(a) and 7(a). Ultimate strength of ferritic stainless steel bolted connection, SF2T30E60 was determined by curling and after specimen reached the ultimate state; tensile net section fracture between two bolts occurred (refer to Table 5). For curled specimens with 2 1 bolt array, the curling in bolted connections of two material types was observed at the beginning of enforced displacement (a) SA2T30E60 (e=60 mm, b=60 mm) (b) SF2T30E60 (e=60 mm, b=60 mm) (c) SA3T30E60 (e=60 mm, b=60 mm) (d) SF3T30E60 (e=60 mm, b=60 mm) Fig. 7. Fracture and curling occurrence with the increase of load ISIJ

8 Table 5. Additional investigation of strength and displacement at critical point for curled specimens. Specimen Fracture mode At ultimate state Displacement d ue [mm] Ultimate Strength P ue [kn] Strength drop point by curling Displacement d uec [mm] Strength P uec [kn] P ue/p uec SA2T30E Tensile fracture SA2T30E SA3T30E24 Shear fracture of plate end SA3T30E SA3T30E36 Shear fracture between two bolts parallel SA3T30E48 to the direction of load SA3T30E SF2T30E60 Curling/Tensile fracture between two bolts 13.71/ / SF3T30E48 Shear fracture between SF3T30E60 two bolts parallel to the direction of load Table 6. Specimen Comparison of ultimate strength of specimens in austenitic and ferritic type. Ultimate strength P ue [kn] With curling occurrence With strength drop effect by curling Fracture mode at the end of test Strength ratio PueSA/ PueSF SA2T30E E 1.51 SA2T30E SA2T30E BS SA2T30E SA2T30E SF2T30E SF2T30E SF2T30E BS SF2T30E SF2T30E SA3T30E SA3T30E SA3T30E E 1.40 SA3T30E SA3T30E SF3T30E SF3T30E SF3T30E E SF3T30E SF3T30E (d uec = mm in Table 5) and shear fracture between two bolts parallel to the direction of loading led to ultimate strength as given in Figs. 5(c) and 5(d). However, specimens of SA3 series (P ue/p uec=46 81%) in strength rise after curling occurrence became higher than specimens of SF3 series (P ue/p uec =26%) because of the high strength enhancement (strain hardening effect, F u/f y) and ductility of austenitic stainless steel material given in Table 3. Load-displacement curve shape and strength reduction pattern in curling occurrence point (d uec, P uec in Table 5) and ultimate state (d ue, P ue in Table 5) are presented in Fig. 7. Table 6 includes the ratio (P ue_304/p ue_430) of ultimate strength of specimens in austenitic to ferritic type and classification about which specimens had negligible strength drop by curling. In particular, for SA3T30E24, SA3T30E30 and SA3T30E36 among SA3 series specimens, curling occurred in the vicinity of enforced displacement ranged from 6.36 mm to 8.03 mm (see Table 5). However, since the ultimate strength of bolted connection tended to be on rise (P ue/p ue(emin)= , in Table 4) with the increase of end distance, above mentioned three specimens (SA3T30E24, SA3T30E30 and SA3T30E36) were excluded from the groups of specimens with strength drop effect by curling as given in Table 6. For specimen SA2T30E60, curling began to occur at the initial enforced displacement. In point 1 (displacement: 8.31 mm, strength: kn) of load-displacement curve, sudden strength drop by curling was observed. Since then, with the increase of enforced displacement and curling deformation, strength (load) showed a tendency to be on the rise again as shown in the point 2 of Fig. 7(a) (displacement: mm, strength: kn). In point 3 (displacement: mm, strength: kn), tensile fracture between two bolts led to the ultimate state of bolted connection. Specimen SF2T30E60 in point 1 (displacement: 3.69 mm, strength: kn, where these pictures were taken during test) of Fig. 7(b) remains elastic and did not show any deformation in shape. It can be seen from Fig. 7(b) and Table 5 that curling occurrence and tensile fracture made the specimen SF2T30E60 reach the ultimate strength almost simultaneously. With the increase of strength (load), tensile stress concentrated on net section between two bolts and curling deformation began to occur. Therefore, in point 2 (displacement: mm, strength: kn) of loaddisplacement, temporary strength drop by curling (in displacement mm written in Table 5) was observed and right after the curling, tensile fracture initiated between two bolts. There was no big difference in strengths between two ultimate states; curling and tensile fracture. Finally, in point 3 (displacement: mm, strength: kn), shear fracture. Parallel to the direction of load and severe curling was observed. Accordingly, this indicates that the curling had little impact on the strength reduction of SF2T30E60. Figure 7(b) represents the deformed shapes and fracture modes taken at each point of 1, 2 and 3. It has already stated in section 2.2 that material properties 2013 ISIJ 372

9 such yield ratio (Fy/Fu) and elongation (EL) of austenitic and ferritic stainless steels were different, that is, austenitic stainless steel has considerable ductility and excellent strength enhancement until material reaches tensile strength after yielding when compared with ferritic stainless steel. For these reasons, specimens SA2T30E24, SA2T30E30 and SA2T30E36 with no curling effect showed higher maximum displacement and strength increase rates (P ue_304/p ue_430= ) than those of SF2 series at ultimate state. In addition, curling in austenitic type (for SA2T30E60, d uec,=8.81 mm in Table 5) also occurred at an earlier stage of displacement compared with ferritic type (for SF2T30E60, d uec,= mm in Table 5) and the pattern of load-displacement curve after curling occurrence was different each other (see Figs. 7(a) and 7(b)). Figures 7(c) and 7(d) represents the load-displacement curve deformed shapes of SA3T30E60 and SF3T30E60 taken at each point 1 and 2. For specimens SA3T30E60 with 2 1 bolt array, transient strength reduction by curling was observed at point 1 (displacement: 3.72 mm, corresponding strength: kn). The curling occurred at the early step of enforced displacement (point 1; d uec,=3.72 mm in Table 5) and reached the ultimate strength (point 2; d ue,=25.67 mm in Table 5) after enforced displacement proceeded sufficiently until shear fracture between two bolts parallel to the direction of loading. It can be known from Figs. 7(a) and 7(b) that the strength increase rate of SA3T30E60 (P ue/p uec = kn/68.11 kn=1.81) after strength drop by curling got higher than that of SA2T30E60 (P ue/p uec = kn/ kn=1.19). For specimens SF3T30E60 with 2 1 bolt array, sudden strength reduction by curling was observed directly after getting through point 1 (displacement: 5.11 mm, corresponding strength: kn). Unlike bolted connections, SF2T30E60 with 1 2 bolt array, curling occurred at the early step of enforced displacement (point 1; d uec,=5.11 mm in Table 5) and reached the ultimate strength (point 2; d ue,=23.29 mm in Table 5) after enforced displacement proceeded sufficiently until shear fracture between two bolts parallel to the direction of loading Curling Deformation and Strain Distribution Figure 8 exhibits the curling-enforced displacement relationship of four specimens (SA3T30E48, SA3T30E60, SF3T30E48, SF3T30E60) with curling influence in SA3 series and SF3 series. At the beginning stage of enforced displacement, curling deformation was increased slightly. After a transient strength reduction was caused by the curling, the sharp increase of curling deformation at the displacement of around 4.0 mm was observed. A strain gage (G1) was attached on the plate surface of location given in Fig. 3(b) in order to investigate the strain change pattern in the plate toward loading direction. Strain displacement relationship was displayed in Fig. 9. The horizontal axis of Fig. 9 indicates enforced displacement, the vertical axis represents strain value taken in G1. Since plate part on which strain gage was placed is under compressive pressure due to bearing action of bolt shank, in general, the strain value in G1 is expected to be compressive ( ). Therefore, for specimen SF2T30E48 without curling effect; the compressive strain value of G1 tended to get larger with the increase of enforced displacement as shown in Fig. 9(c). (a) SA3 series (b) SF3 series Fig. 8. Curling-displacement curves for curled specimens. (a) SA2 series (Curling) (b) SA3 series (Curling) (c) SF2 series (d) SF3 series (curling) Fig. 9. Strain distribution of specimens in gage 1 (G1). However, specimens with strength reduction effect by curling such as SA2T30E48, SA2T30E60, SA3T30E48, SA3T30E60, SF3T30E48 and SF3T30E60 showed different patterns in strain change. In case of SA2T30E48 and SA2T30E60 with remarkable strength drop by curling, compressive strain in G1 continued to be on rise with the increase of displacement, the compressive strain has been falling owing to the progress of severe curling (see Fig. 7(a)). On the contrary, SF2T30E48 and SF2T30E60 had been relatively less affected by curling compared to the others and compressive strain values also continued to rise like those of SA series at an early displacement. After strength drop point by curling, the increase of strain value was slowed down. Also for specimens (SA3 series and SF3 series in Fig. 9) with 2 1 bolt array, curling occurrence had much of impact on the ultimate strength reduction and also led to the transfer of strain distribution from compressive values to tensile ones as shown in Figs. 9(b) and 9(d). Also, as can be known in Table 4 and Fig. 9, the curling occurrence and strain distribution were different between 1 2 bolt array and 2 1 bolt array. As the number of bolt in the parallel direction of loading gets larger, the restraint effect for curling occurrence tends to be larger (the possibility of curling occurrence will be lower). At the beginning of enforced displacement, since the width of SA3T30E48 and SA3T30E60 (2 1 bolt array) was smaller than that of SA2T30E48 and SA2T30E60 (1 2 bolt array), it can be found that curling occurrence in SA3 series went first compared with SA2 series. Nevertheless, after earlier curling occurrence, the more strength enhancement in SA3 series (i.e., the smaller strength reduction by curling) resulted from the higher restraint effect of curling by two bolts of loading ISIJ

10 direction. This is why the area and depth of SA3 series (2 1 bolt array) with two rows in the parallel direction of loading which can resist out of plane deformation toward the end of connected plate is larger than those of SA2 series (1 2 bolt array). 4. Conclusions Experiments for single shear bolted connections with thin-walled stainless steels have been performed and the structural behaviors such as ultimate strength, fracture mode and curling effect have been investigated. Common geometries for specimens are plate thickness, t e: 3.0 mm, bolt diameter, d: 12 mm, gauge & pitch: 36 mm and edge distance perpendicular to the direction of loading, b: 60 mm. Main parameters are two stainless steels (austenitic type (STS304) and ferritic type (STS430)), two bolt arrangements (1 2 and 2 1) and end distance (e) parallel to the direction of applied load (e: 24 mm 60 mm). In this paper, material properties between austenitic (304 type) and ferritric (430 type) stainless steels were compared. As a result, because of the difference of microstructure in two stainless steels, Young s modules, tensile stress and elongation of austenitic stainless steel (304 type) were higher by 3%, 60%, 62%, respectively than those of ferritric stainless steel (430 type). For yield stress and yield ratio, ferritic stailess steel was higher by 19%, 90%, respectively than austenitic stainless steel. That is, it can be found that austenitic type, 304 stainless steel has excellent strength enhancement (F u/f y=2.74, extensive strain hardening) and enough ductility until material reaches tensile strength after yielding compared with ferritic type 430 (F u/f y=1.45). Also, the ultimate behaviors of single shear bolted connections were also investigated compared according to the difference of mechanical properties in two stainless steels and main variables. Fracture modes at the test end were identical for most of bolted connections with two stainless steels and bolt arrays. Specimens with 1 2 bolt array failed in block shear fracture (a combination of tensile net section fracture between two bolts and shear out fracture) except SA2T30E24 (end distance=24 mm, shear fracture mode) and specimens with 2 1 bolt array failed in shear fracture in the parallel direction of loading. For uncurled specimens, shear strength and block shear strength tended to get larger with the increase of end distance. The mean ultimate strength ratio (=0.60) of austenitic type specimens to ferritic type specimens was almost identical with the mean tensile stress ratio (=0.53) obtained from the material test results of two stainless steels. Curling (out of plane deformation) in the plate thickness direction (in the perpendicular direction of loading) occurred for bolted connections with a relatively long end distance. Curling resulted in temporary strength reduction and so there was no big difference in ultimate strength regardless of the increase of end distance for bolted connections. The strength of specimens showed a tendency to rise again after curling occurrence. The rates (ultimate strength to curling strength) of strength increase for bolted connections with austenitic type were higher than those for ferritic type due to excellent strength enhancement (strain hardening effect) of austenitic stainless steel material. Furthermore, curling deformation and strain distribution pattern for curled specimens were similar between two stainless steels. Strain taken from the plate surface of bearing part adjacent to bolts was slowed down for 1 2 bolt array, while for 2 1 bolt array, the strain value was changed into tensile strain values from compressive ones after the strength reduction by curling occurred. Afterwards, especially, for bolted connections with the strength reduction effect of curling, the ultimate strength of test results is need to compare the strength predicted by current design specification. Acknowledgements This research was supported by Basic Science Research Program through the National Research Foundation of Korea (NRF) funded by the Ministry of Education, Science and Technology (No ). REFERENCES 1) ASTM: Standard Specification for Stainless and Heat-resisting Steel Bars and Shapes, ASTM Designation A276-85a, American Society for Testing and Materials, Pennsylvania, USA, (2010). 2) American Society of Civil Engineers: Specification for the Design of Cold-formed Stainless Steel Structural Members, SEI/ASCE-8-02, ASCM, Virginia, (2002). 3) A. L. Johnson and G. Winter: J. Struct. Div., ASCE Proc., 92 (1966), 97. 4) American Iron and Institute: Specification for the Design of Light Gage Cold-formed Stainles Steel Structural Members, AISI, New York, (1968). 5) American Iron and Steel Institute: Stainless Steel Cold-formed Structural Design Manual, AISI, New York, (1974). 6) V. Merwe: Development of Design Criteria for Ferritic Stainless Steel Cold-formed Structural Members and Connections. PhD Thesis, University of Missouri-Rolla, (1987). 7) G. J. van den Berg: J. Constr. Steel Res., 54 (2000), ) American Society of Civil Engineers(ASCE): Specification for the Design of Cold-formed Stainless Steel Structural Members, ASCE 8-90, ASCE, New York, NY, (1991). 9) Australian Standard/New Zealand Standard: Cold-formed Stainless Steel Structures, AS/NZS 4673, Australian Standard/New Zealand Standard, Sydney, (2001). 10) SSBA: Design Manual of Light-Weight Stainless Steel Structures, Stainless Steel Building Association of Japan (SSBA), Tokyo, (2006). 11) EN , Eurocode 3, Design of Steel Structures Part 1.4: General Rules Supplementary Rules for Stainless Steel, CEN, London, (2006). 12) H. Kuwamura: Experimental Report on Strength of Bolted Connections in Thin-walled Plates. Steel Structure Laboratory, The University of Tokyo, Tokyo, (2001). 13) H. Kuwamura and A. Isozaki: J. Struct. Constr. Eng. (Trans. Archit. Inst. Jpn.), 556 (2002), 159 (in Japanese). 14) T. S. Kim and H. Kuwamura: Thin-Walled Struct., 45 (2007), ) T. S. Kim, H. Kuwamura and T. J. Cho: Thin-Walled Struct., 46 (2008), ) T. S. Kim, H. Kuwamura, S. H. Kim, Y. T. Lee and T. J. Cho: Thin- Walled Struct., 47 (2009), ) T. S. Kim and H. Kuwamura: Mater. Design, 32 (2011), ) T. S. Kim, H. Kuwamura, T. J. Cho, S. H. Shin, S. H. Kim and Y. T. Lee: ISIJ Int., 48 (2008), ) M. Lecce and K. Rasmussen: J. Struc. Eng. ASCE, 132 (2006), ) M. Lecce and K. Rasmussen: J. Struc. Eng. ASCE, 132 (2006), ) KS: Cold Rolled Stainless Steel Plates, Sheets and Strips, (KS D 3698), Korean Standards, Seoul, (2008). 22) ASTM: Standard Test Methods and Definitions for Mechanical Testing of Steel Products, (ASTM A370-10), American Society for Testing and Materials, Pennsylvania, USA, (2010). 23) JIS: Method of Tensile Test for Metallic Materials, JIS Z 2241, Japan Industrial Standards, Tokyo, (2010). 24) JIS: Stainless Steel for Building Structure, JIS GZ 4321, Japan Industrial Standards, Tokyo, (2010) ISIJ 374