1 Introduction

Hot-dip galvanizing (HDG) is a viable option for surface coating to have sufficient lifetime against the environmental corrosion. On the other hand, a production of lightweight and high-performing steel components requires use of high-strength and ultra-high-strength steels (HSSs/UHSSs). To reliably utilize the potential of HDG coatings in steel structures, extensive studies on its effects on ultimate and fatigue capacities are needed. However, only few studies have been published on the mechanical performance of HDG HSS and UHSS joints. Najafabadi et al. [1] studied S960 tubes’ properties after HDG and Lipiäinen et al. [2, 3] conducted fatigue performance-related studies on S960 t = 6-mm plates with laser-cut edges.

The EN 14713–2 standard [4], instructing the applications of zinc coatings, recommends grinding thermal cut edges before HDG since changes can occur in the chemical composition at the vicinity of a cut edge due to the presence of oxygen in the kerf. In [5], the chemical composition of a structural steel was found to change with reactive fusion cutting whereas fusion cutting affected chemical composition only at the cut edge defects. In addition to the local changes in chemical composition, heat-affected zone (HAZ) is formed in laser cutting which may have an influence on tensile strength. The cracking of CF beams during HDG has been found in previous studies [6]. Consequently, following recommended bending radii, utilizing UHSS-specific low friction tools, and applying pre-heating before HDG is necessary to prevent cracking.

The manufacturing method of steel grade has been found to affect behavior at subzero temperatures in studies of the uncoated specimen. In [7], the impact toughness of CF S500 DQ steel grade at the subzero temperatures was found significantly higher than conventional thermomechanical controlled processing (TMCP) manufactured S500 grade. S960 steel grade base material and welded specimens’ tensile strength and elongation increased at − 80 °C in comparison to RT in [8]. Nevertheless, in the context of the mechanical performance of HDG welded and laser-cut UHSSs, no contributions have been made. The aim of the current investigation is to shed light on the effects of HDG on the ultimate load and deformation capacities of UHSSs, and thus investigate the applicability of HDG coatings in high-strength components. Ultimate tensile strength and deformation capacities of a direct-quenched (DQ) UHSS grade S960 with a plate thickness of 6 mm are obtained by quasistatic tensile testing welded and plate specimens in the as-rolled condition and after HDG coating. Instead of evaluating the properties of as-treated material, this study focuses on engineering applications, and standard manufacturing procedures for material and components in which a variety of different fabrication methods and post-treatments are applied. In this study, the applied fabrication methods include laser-cut and machined plate edges, welded joints, and cold-formed (CF) and welded beam structures. Quasistatic tensile tests are conducted to verify the usability of HDG UHSSs for producing lighter and more efficient structures, such as those in the transportation and construction sector to mitigate life-cycle CO2 emissions from manufacturing to end-of-life.

Gas metal arc welding (GMAW) was used to manufacture the test specimens. The investigated joint types include the most typical details, i.e., butt welds (BW), fillet-welded non-load-carrying T-joints (NLCT), and longitudinal gusset joints (NLCG), as well as load-carrying cruciform joints (LCX). A nominal throat thickness of a = 4.5 mm was applied in the NLCT and NLCG weldments whereas LCX joints were welded with a = 4 mm fillet welds to study the criticality of weld and influence of heat input (HI) on the joint capacity that was found important factor by Amraei et al. [9].

2 Experimental methods

This paper investigates the tensile behavior of a S960 UHSS in the uncoated condition and after HDG process. The tested components included base material specimens in the as-cut and machined conditions, welded small-scale specimens (BW, NLCT, NLCG, and LCX joints), and welded CF beam flanges. For the same UHSS grade and HDG procedure, fatigue tests on the cut edges and welded longitudinal T-joints were performed in [3, 10].

2.1 Materials

Although this study is focused on the S960 MC steel grade, for the sake of comparison, standard designated S420 MC and S700 MC specimens with machined edges were considered as reference materials. The nominal chemical composition of the materials is presented in Table 1. For all steel grades in this study, the sheets with silicon content of 0.18–0.20 wt% were selected from standard production patches to ensure comparable results. The S690QL material, the feature of higher silicon content as it was not intended for HDG. However, it was used in beam structure compression flange which is non-critical part of the test specimen. Based on the experimental observations, strength-matching (nominally undermatching) Union X96 filler wire with 1.0-mm diameter was used to reach a matching strength, as shown in Table 1, despite its high silicon content.

Table 1 Nominal material properties and chemical composition of the studied materials. fy and fu are the yield strength and ultimate tensile strength of material
Full size table

2.2 Preparation of the specimens

The work dealt with six different types of welded specimens typically found in steel structures (Fig. 1). In addition to those, flat coupon-shaped base material specimens, similar shape and dimensions to those of the butt-welded specimens (Fig. 1d) were tested. In these specimens, both as-cut (laser cutting) and machined conditions were considered. The welded specimens were manufactured using robotic GMAW process to achieve high and consistent quality in the uncoated and HDG specimens. In each welding case, a spray arc transfer mode was applied. A mixture of Ar + 10%CO2 was used as a shielding gas in the GMAW process. The T-profile specimens with gussets (Fig. 1f) were tack-welded first from the specimens’ ends to avoid irregular root geometry at the investigated area. The T-profile specimens were clamped at the base plate and top of the gusset to avoid excessive deformation during welding.

Fig. 1
figure 1

The shape and dimensions of the studied joint types: a NLCT, b CF beam, c LCX, d BM, e NLCG, and f T-profile specimens

Full size image

The welding preparation was carried out according to the limitations given by the manufacturer [11]. The maximum recommended HI and t8/5 cooling time for the studied S960 grade are 0.75 kJ/mm and 15 s, respectively [11]. Since this UHSS is a low alloy steel manufactured via direct quenching technique, its sensitivity to high HIs and, consequently, to desirable cooling rates is higher than in traditional quenched and tempered UHSSs. Due to these reasons, the applied HIs and resulting cooling times followed the given recommendations, as shown in Table 2 (calculated according to Eqs. 1 and 2). Thermal efficiency k = 0.8 was used in the calculations. Cooling time t8/5 was calculated with 2D-method (Eq. 2) as recommended in EN1011-2 [12].The longitudinal welds in the compression flange of the CF specimens (Fig. 1b) were welded with a higher HI (Q = 0.82–0.84 kJ/mm) to achieve high welding quality but these joints were out of the investigated area. Welding parameters are given in Table 2.

Table 2 Welding parameters of test specimens (realized values)
Full size table

It can be noted that weld between S690 QL t = 20 mm and studied S960 MC t = 6 mm was not optimal due to high HI and t8/5. A throat thickness of 3.5 mm was achieved. A bevel with a 2-mm root face without air gap was applied as per a recommendation presented by Skriko et al. [13]. Commercially available Bystronic 10-kW fiber laser setup was used to cut the specimens. The laser cutting parameters are given in Table 3 for specimens with laser cut or base material specimens with machined edges. The laser cutting process and cutting speed influence on the HAZ width that is an important factor on HDG. HAZ features increased hardness which may cause initial cracks and liquid metal embrittlement.

Table 3 Laser cutting parameters. Data from [10]
Full size table
$$Q= k\cdot U\cdot \frac{I}{v\cdot 1000}$$
(1)
$${t}_{8/5}=(4300-4.3\cdot {T}_{0})\cdot {10}^{5}\cdot \frac{{Q}^{2}}{{d}^{2}}\cdot \left[{\left(\frac{1}{500-{T}_{0}}\right)}^{2}-{\left(\frac{1}{800-{T}_{0}}\right)}^{2}\right]\cdot {F}_{2}$$
(2)

The temperatures in a T-profile specimen were monitored during welding with thermocouples from cut edges in the centerline of the specimen at 10-mm (base plate cut edge) and 25-mm (T-profile top edge) distances from weld toes. Temperatures slightly over 300 °C were measured from laser the cut edges. T-profile specimens were clamped until 50 °C was reached and then cooled to room temperature between welds. For CF-beam manufacturing, thermal imaging was used to illustrate the welding of the brackets. Figure 2b presents the specimen in true position beam being in vertical position welding after welding has been finished. Moreover, improved weld toe geometry is beneficial for the fatigue performance.

Fig. 2
figure 2

The temperature during a T-profile and b CF-beam bracket welding

Full size image

A typical industrial HDG treatment was applied to the specimens, as shown in Fig. 3. The CF beam specimens were galvanized without the end plates applied in the four-point bending device. The plates were welded before testing using a manual GMAW process after removing the zinc layer by grinding. The specimens were prepared with 16% hydrochloric acid with an adequate amount of effective pickling inhibitor to avoid the risk of hydrogen embrittlement. The temperature of the drying oven was 70 °C and drying acted as pre-heating before galvanizing to reduce thermal stresses during galvanizing. The specimens were then kept at 450 °C in a 99% zinc path for 2–4 min.

Fig. 3
figure 3

HDG of CF-beam specimens

Full size image

Macrosections of the HDG specimens are shown in Fig. 4 together with bevelling and measured ideal throat thicknesses. A throat thickness of a = 4.5 mm was achieved in NLCG and NLCT joints, and CF beam specimens. In the LCX joints, a = 4 mm was targeted to increase the potential for weld failures. The polished section of CF beam specimen was prepared from a fatigue-loaded specimen and failure location at weld toe is marked with an arrow. The zinc layer on weld metal surface was thicker (250–300 μm) than on the hot-rolled surface of the S960 base material (60–100 μm). In the S690 compression flanges, zinc thicknesses up to 400 μm were found due to the difference in the silicon content (Table 1).

Fig. 4
figure 4

Macrosections and throat thicknesses of specimens

Full size image

2.3 Quasistatic tensile testing

Quasistatic tensile tests were conducted with a 1200-kN servo-hydraulic test rig as displacement-controlled tests with 0.02–0.05 mm/s displacement. Aramis digital image correlation (DIC) system with 1-Hz image frequency was used to evaluate local strain values. Figure 5c shows base material failure with HDG LCX specimen in 30° to loading corresponding to the critical plane failure criterion of sheet metals [15].

Fig. 5
figure 5

Welded specimens: a BW, b NLCT, c LCX, d NLCG, e welded T-profile, and f cut edge specimens at the test rig for RT tensile tests

Full size image

The cold temperature tests for the coupon specimens were performed inside a cold chamber with air cooling. Due to the more complex geometry, the CF beam specimens were insulated from outside ambient temperature and solid CO2 cooled ethanol flow through the beam was used for cooling. The specimen temperature was monitored with thermocouples and kept between − 40 and − 44 °C by controlling the air and ethanol flow.

3 Results and discussion

3.1 Hardness and microstructural evaluation

To identify the local material hardness at the weld metal, base material, and HAZ, hardness measurements were carried out. The hardness contours for the HDG welded joints were obtained using the HV3 hardness with a 0.375-mm step size between the measurement points. Material softening, typical for DQ UHSSs, can be observed at the HAZ, and it was the most significant in the BW specimens (Fig. 6e). The difference in the measured hardness of 4.5-mm (Fig. 6a–c) and 4-mm (Fig. 6d) throat thicknesses is significant. The lower HI in the a = 4-mm welds results in a narrower softened zone and the decrease in base material hardness is also significantly smaller. In terms of local hardness, the difference between the coupon and CF beam specimens was found minor and similar softened zones were found in both cases.

Fig. 6
figure 6

Hardness mapping of a NLCT, b NLCT CF, c NLCG, d LCX a = 4 mm, and e BW welded joints in HDG condition

Full size image

The microstructure of a HDG BW specimen’s weld metal, base material, and HAZ was characterized and is shown in Fig. 7. In addition to martensitic weld metal and BM, prior austenite (PAG), ferrite (F), and granular bainite (GB) grains in the HAZ close to the weld area. Ferritic microstructure in the HAZ of the BW specimen explains decreased hardness (Fig. 6e) after two heat cycles in welding.

Fig. 7
figure 7

Microstructure of weld metal, HAZ, and base material from S960 BW specimen. Note to the different magnifications of the images

Full size image

HDG was found to increase grain size by 20 μm close to the zinc layer as shown in Fig. 8. An etch-resistant, very fine, and equiaxed structure which is probably the most zinc contaminated region. Zinc content in grain boundary is shown in Fig. 8c with a white arrow. Liquid metal embrittlement caused 50–150-μm initial cracking is illustrated in Section 3.4.

Fig. 8
figure 8

EBSD microstructure of S960 HR surface from two different samples (location 1 in a and c and location 2 in b and d), LCX weld toe (e–f), and weld metal (g) in BW-reinforced section

Full size image

3.2 Tensile tests at ambient temperature

3.2.1 Base material and laser cut edges

Tensile tests were performed to evaluate the effect of HDG on the ultimate load and deformation capacity of welded joints. Digital image correction (DIC) system was used in tensile tests. Yield strength was determined from a sharp yield point (upper yield strength) or using the proof strength Rp0.2 (0.2% plastic strain) value. HDG modified the material behavior especially with S700 and S420 and sharp yield point with increased yield stress can be found. With S960 material, a minor decrease in yield strength after HDG was found. According to the quasi-static tensile test results, the ultimate tensile strength of S960 decreased after HDG whereas a small increase was found with S700 and S420 materials. The ratio between yield and tensile strengths (fy/fu) was determined to describe the material behavior. It was found that HDG increased the ratio and with S700 material upper yield strength was higher than tensile strength. The results are shown in Table 4. The behavior of UHSS, HSS, and mild steels coupon base material tests were found than in [1] .

Table 4 Cut edge tensile test results
Full size table

The T-profile specimen with the longitudinal welds had similar deformation capacity to the coupon specimen with FL cut edges which can be seen when comparing Fig. 9a and 9c. With the DIC measurement, it was observed that the longitudinal weld ruptured and revealed decreased deformation capacity (Fig. 9). Such decrease in the deformation capacity can be explained by zinc inclusion and LME reported in [2].

Fig. 9
figure 9

Tensile test results of a S960 BM and FL cut, b S700 and S420 BM, and c S960 T-profiles

Full size image

3.2.2 Welded joints

The tensile tests of welded joints did not reveal major differences between the uncoated and HDG specimens. The ultimate tensile strength and the deformation capacity are the most important values in joint and component testing. However, proof strength Rp0.2 was used to evaluate the influence of HDG, and results should be considered indicative. HI significantly decreased fy/fu ratio and values under from 0.75 to 0.80 were estimated from BW and NLCT specimens with a slight increase after HDG. HDG was found to increase local strain at CF-beams HAZ while the width of HAZ remained the same (Fig. 10d).

Fig. 10
figure 10

Tensile test results of the welded joints for a load-carrying cruciform and non-load carrying T joints, b non-load carrying gusset and butt-welded specimens, c stress-displacement curve over the specimen length for load-carrying cruciform joints and non-load carrying gusset specimens, and d force-strain plot for CF-beam in bending tests shown together with HAZ strains before final rupture

Full size image

The ultimate tensile strength values for the CF-beam specimens with welded NLCT gussets were estimated by using a plastic section modulus and converting test force to nominal bending moment. Similar ultimate capacity was found in both the CF beam specimens and NLCT joints with similar transversal attachments (Table 5).

Table 5 Obtained strength properties of the welded specimens
Full size table

DIC measurement data was used to inspect differences between the uncoated and HDG specimens. Figure 10 illustrates the failure modes of the uncoated and HDG specimens. DIC measured strain values significantly increased in the HDG CF-beam specimen compared to the uncoated specimen. Figure 11 presents overall view on the stress–strain for coupon specimens and force-strain for CF-beams where strain was measured with DIC close to the weld as illustrated in Fig. 10d. Curves and detailed view from the softened zone before the final rupture of specimens. The DIC strain values are extracted from corresponding locations at the end of the CF section toward the bottom of the beams.

Fig. 11
figure 11

Tensile test results of the coupon specimens (a–c) and CF beam specimens (d) at the subzero (− 40 °C) temperature

Full size image

Based on the obtained tensile strengths of the load carrying welds (butt welded and LCX specimens), the strength properties of the filler metal remained at reasonable level in respect to joint capacity although the silicon content exceeded the optimal range for HDG (Table 1). The hardness contours of the specimens explain the distinguishing behavior of the welded joints below the yield stress. The LCX specimens show the lowest strain at Rp0.2. Welding around the gusset end increases heat input locally, which can be seen also from the hardness contour (Fig. 6c). The softened zone has lower yield strength while the tensile strength of the specimen is similar to that of the NLCT specimen due to local strain hardening. However, DIC obtained strain values are not comparable between NLCG and NLCT specimens. The uniform elongation decreases along with increasing material strength. Moreover, HDG had only a minor influence on the uniform elongation of base material.

3.3 Tensile tests at subzero (− 40 °C) temperature

The subzero tensile tests were performed similarly to the tensile tests at the ambient room temperature (RT). The stress-displacement curves of the RT and − 40 °C tests are presented in Fig. 11a–c for coupon specimens. Weld fracture was observed in the uncoated LCX specimens, and a direct comparison of the uncoated and HDG specimens cannot be thus performed. It was found that the applied subzero temperature increased both yield and tensile strength in all specimens. The sharp yield point of HDG specimens is also visible in the subzero tests of the coupon specimens.

The LCX specimens were tested at the subzero temperatures to verify weld metal properties and ductile behavior. Higher elongation was measured in HDG LCX specimens when compared to RT tests. The softened zones of LCX joints were significantly narrower compared to other welded specimens as shown in Fig. 6. However, LCX yield strength is achieved with higher strain than cut edges specimen due to local softening at HAZ.

The subzero temperature test result of the CF beam specimen (Fig. 11d) showed similar trends than coupon specimens, i.e., an increase in both ultimate load and displacement capacities were found. The deformation at fracture of the CF beam specimens was increased after HDG in both RT and − 40 °C tests.

The S960 base material specimens with machined and laser-cut edges were found to have higher strength properties at the subzero temperature. The yield strength of the uncoated specimens and HDG specimens increased by 9–10% and 4%, respectively (Table 6). An increase in the ultimate tensile strength was found also from 2% (uncoated S960 with FL O2 cut edges) to 5% (HDG S960 with machined edges). Yan et al. [16] found similar behavior when testing uncoated S690 and S960 steel grades from + 20 °C to − 80 °C temperatures.

Table 6 Tensile strengths of the base material and welded specimens at the subzero (− 40 °C) temperature
Full size table

3.4 Failure modes

Figure 12 illustrates fracture surfaces from an uncoated NLCT joint (Fig. 12a) and corresponding HDG joint (Fig. 12b–c). HDG longitudinal weld specimen fracture surface is shown in Fig. 12d. Polished sections on the fracture planes of BW and NLCT specimens are illustrated in Fig. 12e–g. Based on the observations, HDG did not influence on the failure mode — final rupture occurred at the HAZ in all specimens. CF-beams revealed crack in weld 1 in all quasistatic tests. In welding the first weld, the heat is transferred via cooling from the weld to the beam while in welding the second weld, heat is partially transferred via the first weld which also decreases cooling time when specimen was cooled to room temperature between the welds.

Fig. 12
figure 12

The fracture surface of a uncoated NLCT HAZ; b, c HDG NLCT HAZ; and d HDG longitudinal fillet weld and transition on the weld surface. Polished section of specimens after test e–f butt welded and h–g NLCT-specimens

Full size image

The zinc layer induced a 50–150-μm initial crack like layer on surface at HAZ of NLCT joint while performance remained the same than tested uncoated NLCT specimen. Consequently, the yield and tensile strength properties of the studied S960 material were not found to be influenced by HDG. However, the HDG longitudinal fillet weld (T-profile) was examined with SEM (Fig. 12d) due to limited deformation capacity and significant initial crack caused by zinc was found. A similar zinc inclusion caused a decrease in the fatigue performance in [2].

3.5 Further research and design recommendations

Post weld treatments (PWT) are typically used to improve fatigue strength of weldments, particularly in HSSs and UHSSs. In this study, only as-welded joints were tested. However, HI from TIG dressing and possible microcracks from high-frequency mechanical impact (HFMI) treatment could affect tensile strength. Consequently, a future study on tested specimens’ fatigue strength is needed to ensure the structural performance and safety of HDG UHSS structures. However, based on initial results on the cut edges and welded T-profiles in [2], reasonable fatigue performance of welded joints can be expected.

As a design recommendation, a use of undermatching filler materials is good practice. According to this study, filler materials are more susceptible on hydrogen embrittlement than the studied S960 steel grade. Undermatching filler material should be used especially when root side of weld is exposed to zinc like in single sided load carrying welds studied in uncoated condition by Skriko et al. [17]. Heat input should be minimized to achieve the best possible performance. HI in welding DQ UHSSs should be controlled and minimized and, in case of fillet welds, smaller throat thicknesses and multipass welding may yield better results.

Risk for detrimental effects of HDG increases along material strength. Experimental studies should be conducted to verify the material suitability and performance on HDG structures before utilization. Moreover, Lipiäinen et al. [3] performed fatigue test for S1100 base material and found similar performance than S960 specimens tested in in this study.

4 Conclusion

Hot-dip galvanized DQ UHSSs can be used in welded steel structures based on tensile tests when recommended practice guidelines in the galvanizing process are followed. Tensile tests at the ambient room (+ 20 °C) and subzero (− 40 °C) temperatures HDG decreased yield slightly with S960 material whereas upper yield strength of S420 and S700 material increased from 468 to 519 MPa and from 740 to 828 MPa, respectively. Sharp yield point was found after HDG S420 and S700 materials.

Tensile strength of S960 base material decreased from 1114–1130 MPa to 1015–1039 MPa after HDG. However, HDG was not found to influence on S960 butt-welded or non-load carrying transverse attachments -joints performance. An increase in deformation capacity after HDG was found in cold-formed beam with non-load carrying attachment bending tests in both room and in subzero temperatures. Subzero temperatures tests revealed 2–10% (uncoated specimens) and 2–5% increase with HDG specimens in both yield and tensile strength. Deformations measured in subzero tests were significantly higher than room temperature tests.