Fatigue properties and failure characterization of spot welded high strength steel sheet

Fatigue properties and failure characterization of spot welded high strength steel sheet

International Journalof Fatigue International Journal of Fatigue 29 (2007) 879–886 www.elsevier.com/locate/ijfatigue Fatigue properties and failure...

2MB Sizes 0 Downloads 58 Views

International Journalof Fatigue

International Journal of Fatigue 29 (2007) 879–886

www.elsevier.com/locate/ijfatigue

Fatigue properties and failure characterization of spot welded high strength steel sheet Xin Long, Sanjeev K. Khanna

*

Mechanical and Aerospace Engineering Department, University of Missouri, Columbia, MO 65211, United States Received 31 July 2005; received in revised form 1 May 2006; accepted 14 August 2006 Available online 30 October 2006

Abstract Fatigue properties and failure characterization of high strength spot welded steels, such as DP600 GI, TRIP600-bare and HSLA340Y GI, have been conducted. Tensile shear and coach peel samples have been used in this investigation. HSLA340Y GI samples were used as the baseline material for comparison. Microhardness was measured to study the hardness change across the weld nugget. Under low load and high cycles situation all the materials show very similar fatigue strength. Crack initiation and propagation during the fatigue loading history has been experimentally determined and discussed. Microhardness tests show that DP600 GI samples have the highest hardness, about 420 HV in weld nugget and 250 HV in base metal, followed by TRIP600-bare, about 400 HV in weld nugget and 220 HV in base metal, and HSLA340Y GI, about 320 HV in weld nugget and 160 HV in base metal. It was also found that both DP600 GI and HSLA340Y GI display softening during high load and low cycles fatigue tests with HSLA showing more prominent softening behavior.  2006 Elsevier Ltd. All rights reserved. Keywords: Fatigue properties; Dual phase steel; Transformation induced plasticity steel; High strength low alloy steel

1. Introduction With increasing demands for higher fuel efficiency in automobiles, new structural materials are being considered among other strategies. New materials are being evaluated for their light-weight or their superior strength, with the ultimate goal of reducing the weight of the vehicle, which will result in lower fuel consumption. In this regard, aluminum alloys and advanced high strength steels (AHSS) are being investigated for substituting currently used low-carbon steels and high strength low alloy (HSLA) steels. Although aluminum alloys, such as 6111 and 5754, have only one-third the density of steel, their widespread use in automobiles has been limited because of its higher costs and difficulties in manufacturing, such as forming and welding [1].

*

Corresponding author. Tel.: +1 573 884 9109. E-mail address: [email protected] (S.K. Khanna).

0142-1123/$ - see front matter  2006 Elsevier Ltd. All rights reserved. doi:10.1016/j.ijfatigue.2006.08.003

AHSS is obtained by controlled cooling from an intercritical annealing process followed by lower-temperature austenite transformation [2,3]. Depending on different alloy elements and cooling process, AHSS has different finial microstructures. Dual phase (DP) steel consists of ferrite and martensite, and the mechanical properties are controlled by the volume fraction of martensite and ferrite grain size. Transformation induced plasticity steel (TRIP) consists of ferrite, bainite and austenite. Retained austenite plays a very important role in the TRIP steel’s mechanical performance. When TRIP steel is subjected to plastic deformation in manufacturing or in service, the retained austenite can transform to martensite accompanied by a large elongation. This is the so called transformation-induced plasticity (TRIP) effect. AHSS, such as dual-phase steel DP600 GI and transformation induced plasticity steel TRIP600-bare have tensile strengths over 600 MPa, compared to conventional high strength steels within the range of 400–440 MPa, though they have similar yield strengths and are good candidates for making lighter-weight vehicles. These advanced steels typically exhibit higher yield

880

X. Long, S.K. Khanna / International Journal of Fatigue 29 (2007) 879–886

strength, low yield to tensile strength ratio, good formability, high work hardening rate, and high strain energy absorption capabilities. As a result some of the sections of an automobile such as doors, and front and rear longitudinal rails, can be made of a thinner section that reduces the weight of the vehicle while absorbing significant deformation energy due to the high work hardening rate and ductility. Thus, thinner AHSS sheets, which decrease the weight of an automobile, can be used without loosing any strength, and having a similar or higher level of crash energy absorption. The materials mentioned above can be joined by a variety of methods, but spot welding still remains the primary joining method in automobile body manufacturing till today. A typical vehicle could contain more than 3000 spot welds [2]. AHSS steels, as any other material used in automobiles, can undergo significant fatigue damage, which is related to the reduction in sheet thickness [4–7]. Thus, it is imperative to investigate the fatigue behavior of spot welded joints in thin section AHSS for achieving a safe and reliable design. The purpose of this study is to experimentally investigate the weld fatigue properties of AHSS steels (DP600 GI and TRIP600) and compare with the baseline material HSLA340Y GI for both coach peel and tensile shear type specimens under tension–tension sinusoidal loads. The fatigue crack initiation and propagation has been studied and the failure mechanisms have been discussed.

2. Experimental procedure 2.1. Materials and specimen The chemical composition and mechanical properties of AHSS steels and HSLA steel are shown in Table 1 and Table 2, respectively [2]. Specimens used for fatigue tests were tensile shear type and coach peel type, as shown in Fig. 1. Samples were supplied by Auto Steel Partnership. The thickness is 1.53 mm for DP600 GI, 1.64 mm for TRIP600, and 1.78 mm for HSLA340Y GI. Tensile shear type specimens have a length of 217 mm and width of 38 mm. Coach peel specimens have a length of 155 mm and width of 38 mm. In tensile shear specimens suitable tabs were glued to the specimen ends to reduce bending deformation at the joint during the fatigue tests. The average diameter of spot weld nugget was measured to be approximately 7.0 mm for all samples. 2.2. Fatigue test procedure An Instron universal servohydraulic testing machine with a 8800 Plus controller was used to conduct load-controlled tension–tension fatigue test. The samples were subject to various sinusoidal load with a load-ratio R = 0.1 or R = 0.3, at a frequency of 5 Hz for coach peel samples and 10 Hz for tensile shear samples. For the case of R = 0.3, a constant load amplitude was set when compared with that

Table 1 Nominal chemical composition of AHSS steels and HSLA steel (wt%) Steel

C

Mn

P

S

Si

Cu

Ni

Cr

Mo

Al

V

Cb

DP600 GI TRIP600 HSLA340Y GI

0.081 0.101 0.053

1.760 1.470 0.620

0.017 0.002 0.008

0.006 0.001 0.005

0.013 1.536 0.214

0.040 0.016 0.052

0.02 0.02 0.02

0.19 0.05 0.01

0.180 0.010 0.000

0.048 0.027 0.039

0.002 0.005 0.001

0.004 0.005 0.016

Table 2 Nominal mechanical properties of AHSS steels and HSLA steel Steel

0.2% offset yield strength (MPa)

Ultimate tensile stress (MPa)

Uniform elongation (%)

Total elongation (%)

DP600 GI TRIP600 HSLA340Y GI

432.6 420.9 369.2

671.4 672.8 448.5

13.6 20.6 15.9

22.1 29.3 31.7

Fig. 1. (a) Tensile shear type and (b) coach peel type fatigue specimens.

X. Long, S.K. Khanna / International Journal of Fatigue 29 (2007) 879–886

3. Results and discussion 3.1. Quasi-static tension properties

Fig. 2 shows typical load vs. displacement curves for various spot-welded samples and the average maximum load sustained by these samples are listed in Table 3. It can be seen from Table 3 and Fig. 2 that TRIP600 tensile shear samples have the highest strength among three types of Table 3 Maximum load carrying capacity of tensile shear and coach peel samples Material

Tensile shear samples (kN)

Coach peel samples (kN)

DP600 GI TRIP600 HSLA340Y GI

20.95 23.58 17.82

3.64 3.29 4.84

a

5000 4500

Load Amplitude (N)

of R = 0.1. Thus, the maximum, minimum and mean load at R = 0.3 were higher from that of R = 0.1 case, but the amplitude was the same. The maximum dynamic load used was typically about 60% of the maximum load obtained during a quasi-static tensile test. During the fatigue test, the maximum load, minimum load, maximum and minimum cross head displacement, and the frequency were monitored by using the Instron Wave Maker Data Acquisition software and 8800 Plus digital controller. For microstructure and fracture morphology observation of the failed sample, they were cut along the center of the spot nugget in the direction of the length of sample. The cross section of the cut samples was polished and then etched. The etchant used was 3% Nital (3 mL HNO3 and 97 mL alcohol). Microhardness tests were conducted on the surface of the sectioned and polished samples used for microstructure observation. A Micromet II microhardness tester was used to conduct Vickers hardness tests with a load of 300 g.

881

Quasi-static tension tests were conducted with tensile shear and coach peel samples for each of the materials.

4000 3500

DP600-TS-R01

3000

DP600-TS-R03

2500

DP600-CP-R01

2000

DP600-CP-R03

1500 1000 500 0 1.0E+02

1.0E+03

1.0E+04

1.0E+05

1.0E+06

1.0E+05

1.0E+06

1.0E+05

1.0E+06

Cycles to failure

Load amplitude (N )

b

5000 4500 4000 3500

TRIP600-TS-R01

3000

TRIP600-TS-R03

2500

TRIP600-CP-R01

2000

TRIP600-CP-R03

1500 1000 500 0 1.0E+02

1.0E+03

1.0E+04

Cycles to failure

c

5000

Load amplitude (N )

4500 4000 3500

HSLA340-TS-R01

3000

HSLA340-TS-R03

2500

HSLA340-CP-R01

2000

HSLA340-CP-R03

1500 1000 500 0 1.0E+02

1.0E+03

1.0E+04

Cycles to failure Fig. 2. Typical load vs. displacement plots under quasi-static loading of: (a) tensile shear and (b) coach peel samples.

Fig. 3. Load amplitude vs. cycles to failure curves for: (a) DP600 GI, (b) TRIP600 and (c) HSLA340Y GI (TS, tensile shear; CP, coach peel, R01, R-ratio of 0.1; and R03, R-ratio of 0.3).

882

X. Long, S.K. Khanna / International Journal of Fatigue 29 (2007) 879–886

materials (32% higher than that of HSLA340Y GI sample and 13% higher than that of DP600 GI). For coach peel samples, however, the situation is just the opposite. HSLA340Y GI samples display the highest load carrying capacity and TRIP600 has the lowest maximum load. 3.2. Fatigue properties The fatigue life curves of the three materials under consideration are shown in Fig. 3. It can be seen from Fig. 3 that the fatigue strength for low cycles (higher load) of DP600 GI samples is very similar to that of TRIP600 samples, and both of them have lower fatigue strength than that of HSLA340Y GI samples for both tensile shear and coach peel specimens. However, this difference is mainly due to the higher sheet thickness of HSLA specimens. For high cycles (lower load), the fatigue strength is approximately the same for all three types of materials and for both tensile shear and coach peel samples. The slightly higher fatigue strength of HSLA specimens, as seen in Figs. 3 and 4, is again due to the somewhat higher sheet thickness of HSLA specimens. For tensile shear samples, the fatigue strength for infinite life of DP600 GI and TRIP600 spot-welded joint is about 6–8% of the static joint strength. For coach peel samples, the above value is about 4%. Thus a high strength of base materials does not necessarily mean high fatigue strength of spot welded joint. These results agree with previous studies that the spot weld fatigue performance is independent of base materials strength at low load and high cycles [9,10]. The fatigue strength depends on the nature of the loading on the spot weld and the stress concentration factor of the circumferential notch around the weld.

3.3. Microstructure and fatigue crack characterization 3.3.1. Tensile shear samples Microstructures and fatigue crack surface morphologies are shown in Figs. 5–7. Fig. 5 and Fig. 6 show the microstructure and fatigue crack path in DP600 GI and TRIP600 spot welded joints, respectively. The microstructures of DP600 GI and TRIP600 spot welds that were subjected to fatigue loading do not show any special characteristics under optical microscopy at low magnification, compared to an ordinary spot weld joint not subjected to fatigue loads. Coarse column dendritic crystals in the spot nugget are surrounded by about 0.7 mm thick heat affected zone (HAZ) areas for all three materials. It has been observed that at high load and low cycles, fatigue crack in a tensile shear sample initiated outside the weld nugget for HSLA350 while in DP600 and TRIP 600 initiated close to the weld nugget and then propagates along the workpiece interface for a short distance (from HAZ of spot nugget to edge of nugget, about 0.8 mm), and then propagates perpendicular to the workpiece interface (that is out of the plane of the sheet) to complete fracture, as shown in Fig. 5b. For the case of low load and high cycles, the crack directly propagates perpendicular to the workpiece interface from the HAZ area of spot nugget, as shown in Fig. 5c. Unlike DP600 GI spot weld, it was found that a ‘‘tongue’’ formed between the two workpieces of TRIP600 tensile shear samples around the spot nugget, which is shown in Fig. 6a. This tongue is usually the location for initiation of a fatigue crack under both high cycle and low cycle conditions, as shown in Figs. 6b and c. This could be the reason for the low fatigue strength of TRIP600 spot welded joints even though they have a high quasi-static tensile strength.

7000 HSLA340_0.1 DP600_0.1 HSLA340_0.3 DP600_0.3 TRIP600_0.1 TRIP600_0.3

Load Amplitude, N

6000 5000 4000 3000 2000 1000 0 1000

10000

100000

1000000

10000000

Number of Cycles to Failure Fig. 4. Load amplitude vs. cycles to failure curves for: (a) DP600 GI, (b) TRIP600 and (c) HSLA340Y GI tensile shear type specimens (extension_0.1 represents a R-ratio of 0.1, _0.3 a R-ratio of 0.3).

X. Long, S.K. Khanna / International Journal of Fatigue 29 (2007) 879–886

883

Fig. 6. (a) Microstructure of TRIP600 spot weld, and crack growth in TRIP600 tensile shear specimens: (b) for fatigue life of 6920 cycles, (c) for fatigue life of 677,748 cycles. (All pictures at 5· magnification.) Fig. 5. (a) Microstructure of DP600 GI spot weld, and crack growth in DP600 GI tensile shear specimens: (b) for fatigue life of 51,360 cycles, (c) for fatigue life of 1,647,440 cycles. (All pictures at 5· magnification.) Note: BM, base metal; HAZ, heat affected zone; and SN spot nugget.

It has been reported in a previous study [2] that TRIP steels need high current intensity and high squeeze force to be spot welded. This may be the cause for ‘‘tongue’’ formation during welding and could be minimized by a better

choice of welding parameters. The tongue may form a very weak interfacial bond with the base material due to its surface oxidation [2], or directly forms a sharp crack-like notch as shown in Fig. 6a. Under the application of a fatigue load, the sharp crack-like notch easily forms a crack. It is interesting to note that the crack path is not along the HAZ border as in the case of DP600 GI spot weld, but only through the base metal.

884

X. Long, S.K. Khanna / International Journal of Fatigue 29 (2007) 879–886

gue load conditions. However, it is another crack that is perpendicular to the workpiece interface which causes the final fatigue failure. For high load condition, this perpendicular crack starts in the HAZ area, while for low load condition; it starts in the base metal adjacent to the HAZ area. The differences in the nature of fatigue crack initiation and propagation for different materials and fatigue loads is possibly due to the interplay between the generation of plastic zone at the circumferential notch, microcracking, and the relative mechanical properties of the base metal, HAZ, and the weld nugget. A detailed fracture analysis of fatigue crack propagation is not within the scope of this investigation.

Fig. 8. Fatigue crack growth in HSLA340Y GI coach peel sample with a life of 5103 cycles (magnification: 5·).

Fig. 7. (a) Microstructure of HSLA340 spot weld, and crack growth in HSLA340 tensile shear specimens: (b) for fatigue life of 9571 cycles, (c) for fatigue life of 1,728,126 cycles. (All pictures at 5· magnification.)

The crack path in HSLA340Y GI tensile shear samples is different from that in DP600 GI or TRIP600 samples. From Figs. 7b and c, it can be seen that crack propagates along the workpiece interface into the weld nugget for a very short distance in both high fatigue load and low fati-

Fig. 9. DP600 GI coach peel fatigue sample failure modes: (a) for a fatigue life of 995 cycles, and (b) for a life of 1,524,765 cycles.

X. Long, S.K. Khanna / International Journal of Fatigue 29 (2007) 879–886

Microhardness HV300g

3.3.2. Coach peel samples The fatigue fracture morphology of coach peel samples was similar for all three types of materials studied. At high load and low cycles, crack propagates parallel to the weld interface and into the spot nugget for a short distance, and then propagates perpendicular to the weld interface till complete fracture, as shown in Figs. 8 and 9a. For the case of low load and high cycles, the crack propagates parallel to the weld interface and into the spot nugget for a relatively long distance, as shown in Fig. 9b.

550 500 450 400 350 300 250 200 150 100

Non-fatigue 51K cycles 1624K cycles 0

4 2 Distance From Nugget Center (mm)

885

3.4. Microhardness measurements Microhardness tests were conducted along the centerline on a plane at half thickness of spot-weld and on a plane at quarter thickness of spot nugget for each of the three materials using a Vickers pyramidal indenter at a load of 300 g. Since the results of microhardness on the two planes were very similar, only results at the half-thickness plane are shown in Fig. 10. In Fig. 10, the distance from 0 to 3.5 mm represents the weld nugget, from 3.5 mm to about 4.2 mm the heat affected zone (HAZ) and beyond 4.5 mm the base metal. It can be seen from Fig. 10 that the DP600 GI spot weld has the highest hardness of about 420 HV in weld nugget and 250 HV in base metal, followed by TRIP600 with about 400 HV in weld nugget and 220 HV in base metal, and HSLA340Y GI, has a hardness of 320 HV in weld nugget and 160 HV in base metal. It has been observed that DP600 GI shows some softening during high load and low cycle fatigue tests, while HSLA340Y GI samples showed more significant softening in the spot nugget and the HAZ region. This behavior is similar to that reported in a previous study [7,8]. The low load and high cycle tests for all three materials did not show any significant softening or hardening effect.

DP600 GI 4. Conclusions

550 Microhardness HV300g

500

An experimental investigation of the fatigue properties of DP600 GI, TRIP600-bare and HSLA340Y GI spot welded samples led to the following conclusions:

450 400 350 300

Non-fatigue

250

7K cycles

200

678K cycles

150 100

0

4 2 Distance From Nugget Center (mm)

TRIP600

Microhardness HV300g

500 450

Non-fatigue

400

10K cycles 151K cycles

350 300 250 200 150 100

0

4 2 Distance From Nugget Center (mm)

HSLA 340 Fig. 10. Microhardness variation in the half-thickness plane of: (a) DP600 GI, (b) TRIP600 and (c) HSLA340Y GI spot welded samples as a function of number of cycles to fatigue failure.

1. Fatigue strength, under constant amplitude sinusoidal fatigue loading, of spot welded DP600 GI, HSLA340Y GI and TRIP600 samples was approximately the same in the case of tensile shear and coach peel type specimens for both low and high cycle life. 2. For tensile shear samples, different spot welded materials exhibited different fatigue crack paths. At high loads, fatigue cracks in DP600 GI samples propagated along the workpieces interface and into the spot nugget for a short distance and then propagated perpendicular to the workpieces interface until complete fracture, while cracks in HSLA340Y GI samples propagate perpendicularly in the HAZ area. For low fatigue loads, cracks in DP600 GI samples directly propagate perpendicular to the workpieces interface from the HAZ area of spot nugget, while cracks in HSLA340Y GI samples propagate perpendicularly in the base material adjacent to HAZ area. In TRIPsteels, for both high and low load conditions, the cracks propagate from the tongue area perpendicular to the workpieces interface. 3. For coach peel samples made of all three materials at high loads, cracks propagate along the workpieces interface into the spot nugget for a short distance, and then propagates perpendicular to the workpieces interface

886

X. Long, S.K. Khanna / International Journal of Fatigue 29 (2007) 879–886

until complete fracture. For low loads, the cracks propagate along to the workpieces interface into the spot nugget for a relatively long distance. 4. Microhardness tests showed that DP600 GI spot weld has the highest hardness, about 420 HV in weld nugget and 250 HV in base metal, followed by TRIP600, about 400 HV in weld nugget and 220 HV in base metal, and HSLA340Y GI, about 320 HV in weld nugget and 160 HV in base metal. It was also observed that DP600 GI displayed slight softening in the HAZ region for high load and low cycle fatigue, while HSLA340Y GI showed significant softening in the nugget and HAZ during high load fatigue testing.

Acknowledgments The partial financial support of Auto Steel Partnership and the National Science Foundation under grant NSFCareer 0196390 is gratefully acknowledged. The Auto Steel partnership also supplied all the spot-welded samples. The authors are thankful to Dr. Benda Yan, Mittal Steel, USA, for many helpful comments.

References [1] Kelkar A, Roth R, Clarl J. Can aluminum be an economical alternative to steel. JOM 2001;53(8):28–32. [2] Personal communication with Inland Ispat co. [3] Lindgren C, Sperle JO, Jonsson M. Fatigue strength of spot welded beams in high strength steels. Weld World 1996;37(1):90–104. [4] El-Sayed ME, Stawiarski T, Frutiger R. Fatigue analysis of spotwelded joints under variable amplitude load history. Eng Fract Mech 1996;55(3):363–9. [5] Stephens RI, Fatemi A, Stephens RR, Fuchs HO. Metal fatigue in engineering. 2nd ed. A Wiley-Interscience Publication, John Wiley & Sons, Inc.; 2001. [6] Sakuma Y, Kimura N, Itami A, et al. Next-generation high-strength sheet steel utilizing transformation-induced plasticity (TRIP) effect. Nippon Steel Technical Report No. 64, March 1995, p. 20–25. [7] Fredriksson K, Melander A, Hedman M. Influence of prestraining and ageing on fatigue properties of high-strength sheet steel. Int J Fatig 1988;10(3):139–51. [8] Cycle counting for fatigue analysis, ASTM Standard E-1049, Annual book of ASTM standards, vol. 03.01, ASTM, West Conshohocken (PA); 1998. p. 693. [9] Rathbun RW, Matlock DK, Speer JG. Fatigue behavior of spot welded high-strength sheet steels. Weld J 2003;82:207s–18s. [10] Davidson JA, Imhof Jr EJ, The effect of tensile strength on the fatigue life of spot-welded sheet steels, SAE Technical Paper 840110, Warrendale (Pa): Society of Automotive Engineers; 1984.