1 Introduction

The service environment of metallic structure is becoming more and more complicated as the in-depth advancement of industrialization. Traditional homogenous materials have more problems dealing with the service environment required for corrosion resistance effects. So that, laminated composite materials were designed out to simultaneously meet both mechanical and corrosion resistance requirements for industrial equipment [1, 2]. Stainless steel clad plate (SSCP) is the most common type of binary laminated composite material which always contains 10–20% stainless steel plate except for base carbon plate [3]. Generally speaking, parent base plate is made of low-cost carbon steel such as Q235, Q345, or X65 so as to meet the mechanical property requirements, while parent flyer plate is always single-phase austenitic or duplex stainless steel such as 304, 316L, and 2205 for the sake of corrosion resistance performance. Excellent performance and great cost advantage make the SSCP has already been rapidly promoted in shipbuilding, petrochemical engineering, nuclear power, and automobile manufacturing [4, 5]. It is regarded as one of the most promising types of structural metallic materials in the future.

Due to the great industrial demand, high-efficiency and high-quality preparation technology for SSCP has been developed maturely [6, 7]. In contrast, there are still many problems with its welding and joining procedure. Two component metals with distinct chemical composition, physical and mechanical properties, will lead to microstructural deterioration [8, 9] and excessive internal stress [10] during welding, and hence make the joint properties undesirable. As thus, welding technology has become a major factor restricting the large-scale application of SSCP.

Over the past years, a lot of studies have been done on technological development of SSCP welding. In this paper, the welding status and issues of conventional multi-layer and multi-pass welding, highly efficient welding methods, and the numerical simulation research for SSCP were first reviewed. And meanwhile, the relevant next works were planned.

2 Conventional multi-layer and multi-pass welding

The dilution between the base plate and flyer plate during welding is the leading cause of microstructural deterioration. The interdiffusion of main elements in base (C, Fe) and flyer (Cr, Ni) plates may result in hardening region generation on the one hand and corrosion resistance decrease on the other hand [11]. In view of this problem, engineers posted that multi-layer and multi-passes welding should be applied for stainless clad steel plate welding, namely, welding the joint into base, transition, and flyer seam layers.

2.1 Influence of welding sequence

For such type of welding procedure, welding priority of three layers has a great influence on the joint properties. Generally speaking, the welding sequence with base layer first is commonly adopted in most industrial parts manufacturing since it can usually get better joint performance. For instance, Wang et al. [12] welded 2205/16MnR plate with the base layer first and a high-quality joint was obtained with no precipitation of sigma phase (σ) or M23C6. Dhib et al. [13] had prepared the 316L/A283 clad plate joints under three different welding sequences and subsequently compared their mechanical properties. Results were shown in Fig. 1. As we can see, the toughness values of weld metal and heat affected zone (HAZ) in joint under welding sequence with base seam layer welded first were higher than that of the other two joints.

Fig. 1
figure 1

a Passes arrangement schematics of welded joints prepared with different sequences. b Measured Charpy V-notch energies [13]

Full size image

Although other welding sequences always result in deterioration of joint performance, they are inevitable to be adopted in some special cases in engineering, such as the final closure welding of small diameter pipes and pressure tanks in petrochemical industries. In such conditions, single-side welding started from flyer layer is usually adopted. To improve joint performance, engineers always apply stainless steel fillers into all beads welding [14]. For example, Torbati et al. welded 2205 duplex stainless steel clad pipe with flyer plate first by all duplex stainless steel wire metal and a desirable welding joint was obtained under optimized welding methods and parameters [15].

2.2 Influence of the usage of mixed filling metals

The usage of mixed filling metals, namely, applying the stainless steel welding wire for the flyer and transition layer welding and using carbon steel wire for the base layer welding, is very common under the welding sequence started from base layer. And generally, the obtained joints can meet the post-welding test evaluation [12]. But due to the application of mixed filling metals, the multi-layer welding process of stainless clad steel plate essentially belongs to the dissimilar metal welding (DMW) between low carbon steel and stainless steel. Dilution will occur in the local areas near seam layers’ interfaces and inevitably result in the generation of martensite phase according to previous study [16]. Li et al. [17, 18] applied the post flyer plate welding process to weld the L415/316L clad plate. Hardness and microstructure characterization showed that about 0.6–0.8-mm-thick hardened martensite layer generated at the interface between stainless and carbon steel seam layers and hence resulted in the crack initiation on the bending test samples although the tensile and bending performance of joints did meet the standards. Similar microstructure was found in another study too [19]. Although effects of the interfacial martensite layer on the long-term service performance such as fatigue and creep resistance of SSCP joints has not been deeply studied so far, such a thin martensite layer is the common microstructure at interface between stainless and carbon steel DMW joints. From the point of structural integrity, the occurrence of such a hard and brittle layer at interface can obviously cause a lot of problems in subsequent service life of DMW joint [20, 21]. Some targeted controlling methods have already been proposed in DMW, such as inserting buttering layer or using some special groove types [22, 23]. Obviously, these methods are not suitable for interfacial martensite layer controlling in SSCP welding. Extra solutions need to be explored furthermore.

Notably, the local hardening problem due to the usage of mixed filling metals will become more serious under the other welding sequences. For instance, Gou et al. prepared the 2205/X65 bimetallic butt joint according to the welding sequence with flyer layer first. The impact test results showed that impact energy of weld metal (WM) and HAZ of layer 1 (transition and flyer layers) exhibited lower than other two layers, as shown in Fig. 2 [24]. Li et al. [25] welded the 321SS/Q345 bimetallic clad plate with mixed filling metals. Results showed that the martensite formed in the first seam bead filled with carbon steel wires and led to a large area of local hardening zone (LHZ) generated in the WM, as shown in Fig. 3. Although the LHZ had no effect on the tensile properties of the joint, it was still possible to induce crack initiation in the bending test. This kind of joint performance is obviously a hidden risk for safe service of SSCP industrial equipment.

Fig. 2
figure 2

a Morphology of the cross section of the welded joint. b Scheme of Charpy impact test. c Charpy impact test results for weld metal. d Charpy impact test results for HAZ [24]

Full size image
Fig. 3
figure 3

The welding sequence, microhardness measured lines, and corresponding measured data of joints: a, b Single metal filled samples. c, d Mixed metals filled samples. e Bulging appear in the face bending test sample of mixed metal filled joints. f Cracks appear in the side bending test of mixed metal filled joints [25]

Full size image

Reported studies demonstrated that the harden martensite phase is bound to generate at the junction of transition layer and the base layer as long as the mixed filling metals were used. And the area of hardened zone would greatly increase under the welding sequence with flyer layer welded first. But the use of mixed filling metals is the inevitable tendency for technological development of SSCP welding since it can greatly reduce the welding cost. So that, improvement solutions need to be found to reduce the amount of hardened microstructure in the future. Some special welding methods may become the viable options. Górnikowska et al. [26] found that the width of the martensite band could be highly limited by using Cold Metal Transfer (CMT) technique for weld overlaying since it introduced low amounts of heat input compared to traditional overlaying methods. Similarly, there are some other feasible welding methods such as TIP Tungsten Inert Gas Arc Welding (TIP TIG) [27], Hot-Wire TIG welding (HW TIG) [28], and Pulsed TIG Welding (P-TIG) [29]. Because all these methods can reduce the arc heat input to the specimen and hence reduce the dilution rate compred to the conventional TIG welding. However, limited comparative studies have been done on the performance between joints prepared by these improved and traditional welding methods.

In addition to the generation of hardening zone in the use of mixed welding materials, the multi-layer and multi-pass welding of SSCP still has a fatal drawback, namely, the low welding efficiency. So far, some efforts [30,31,32,33,34,35,36] have been done on the high efficiency welding.

3 High efficiency welding

3.1 Laser beam welding

The most direct way to improve welding efficiency of SSCP is to reduce the number of welding passes. Mossori et al. [30] made try to prepare the 304 stainless steel clad plate (2.5 + 6.5mm) by single-pass high-penetration laser beam welding (LBM). The joints obtained showed satisfactory mechanical properties, while Ni-based alloy filler metal was put into the joint in the shape of strips. Similarly, Gou et al. [31] used laser welding to do the single-pass joining of the 2205/X65 SSCP (2 + 2 mm) with the flyer plate set on the bottom side. Results showed that the upper and lower parts of the dumbbell welding seam did show extremely high inhomogeneity on account of the Marangoni convection, as shown in Fig. 4. The high cooling rate of LBM induced the formation of martensite in upper fusion zone (FZ) and hence resulted in the increase of local hardness. Meanwhile, high cooling rate led to the excessive amount of ferrite (unbalanced ferrite/austenite ratio = 65/35) and the second-phase particles within austenite grains in the lower FZ. The undesirable microstructure finally contributed to this area preferentially to fail during the tensile test.

Fig. 4
figure 4

a Schematic sketch of 2205/X65 bimetallic sheet laser butt welding process. b Relationship between the flow behavior of the molten pool and the construction of bimetallic sheets. c Morphology and the test position of the Vickers hardness test line on the cross section. d Results of the microhardness profile along line 3 and 4 on the cross section. e Stress-strain curves of the tensile test and the corresponding localized strain distributions on the surface for the same tensile specimen at points A, B, C, and D [31].

Full size image

Subsequently, Gou et al. introduced a lagging Metal Inert Gas Welding (MIG) arc (laser-arc distance ≥ 15 mm) so as to counterbalance the excess α-ferrite formed in the lower FZ in the laser beam welding joint, as shown in Fig. 5 [32]. MIG arc heated up the front laser welding seam and allowed a reduction in the cooling rate of laser weld, and hence resulted in more nucleation and growth of austenite in the FZ of flyer layer. But the study pointed out that a desirable joint asked for the extremely strict optimized welding parameters, especially the peak temperature at lower FZ reheated by MIG arc.

Fig. 5
figure 5

a Schematic sketch of laser-MIG tandem welding process for 2205/X65 bimetallic sheet. b Finite element results of thermal histories of lower FZ of laser welding (Group A) and laser-arc hybrid welding (Groups B and D, with 15-mm and 20-mm laser-arc distances, respectively). Microstructures in the FZ: c Laser welding samples. d Laser-MIG tandem welding samples. Microstructures in HAZ: e Laser welding samples. f Laser-MIG tandem welding samples [32]

Full size image

3.2 Laser-arc hybrid welding

Laser-arc hybrid welding is one of the high efficiency welding techniques developed rapidly in recent years. It owns many merits such as fast welding speed, deep penetration, and high welding efficiency [33]. Kang et al. [34] applied laser-arc hybrid welding to weld the 304/Q235 SSCP (0.5 + 4.5mm) with different laser-arc distance (DLA). Results showed that tensile properties of joints can be easily ensured under different DLA from 3 to 9 mm, but the corrosion resistance performance strongly depended on the values of DLA. The accepted joint could be obtained with nearly equal corrosion resistance to parent metal by increasing the DLA to 9 mm. The increasing DLA led to the separation of laser and arc heat sources or molten pools and hence reduced the dilution effect of the base layer to increase Cr content and promote the growth of homogeneous cellular grains, as shown in Fig. 6.

Fig. 6
figure 6

a The experimental set-up. b Fracture location of tensile test samples (The DLA of three samples from top to bottom were 3, 6, and 9 mm, respectively). c Potentiodynamic polarization curves of samples with different DLA. d Effects of DLA on the corrosion potential (Ecorr), the corrosion current (Icorr), the passive potential (Epit), and the passive current density (Ipit), obtained by polarization curves. Microstructure of 304SS cladding layer surface of the welds: e Nearby fusion line of the weld at DLA 3 mm. f FZ centre of the weld at DLA 3 mm. g Details of area W3. h Nearby fusion line of the weld at DLA 6 mm. i FZ centre of the weld at DLA 6 mm. j Details of area W6. k Nearby fusion line of the weld at DLA 9 mm. l FZ centre of the weld at DLA 9 mm. m Details of area W9 [34]

Full size image

Subsequently, Meng et al. [35] studied the influence of type of filling wires imposing on the corrosion resistance of laser-arc welding stainless clad steel joint. Results showed the corrosion resistance property improved as the Cr content increases in the filling wire. Furthermore, Meng et al. [36] found that laser power has greater impact on the weld corrosion resistance compared to wire filling rate.

Studies demonstrate the feasibility of single-pass welding of SSCP by high efficiency welding methods such as laser beam welding and laser-arc hybrid welding, but apparently, the joint performance obtained by reported welding process is very unstable. More work needs to be done on improving the stability of single-pass welding process. Predictably, the maturity of single-pass welding technology will promote a great revolution of the manufacturing technology of SSCP structures.

3.3 Welding efficiency improvement in multi-pass welding process

Single-pass welding is only applicable for the joining of plates with small thickness (≤9mm). For the medium and thick SSCP, the feasible welding efficiency improvement way is to improve the cladding efficiency of each pass during multi-pass welding. Limited research has been reported in this regard. But Twin-Electrode TIG Welding (T-TIG) welding method may be the workable option [37]. Wu et al. [38, 39] have done lots of investigations on the arc physics and process features of T-TIG. Results found that coupling arc pressure and penetration for T-TIG are lower than single arc (TIG) under the same welding parameters, and meanwhile, the welding speed is much higher. This welding technology is a promising method for high efficiency welding of medium and thick SSCP.

4 Numerical simulation of SSCP welding

Besides the improvement of microstructure and mechanical property deterioration, residual stress controlling is also an important aspect of SSCP welding. The magnitude and distribution of residual stresses have a significant impact on structural integrity and stress corrosion cracking resistance of welding joints [10]. Due to the significant differences in physical, mechanical, and metallurgical properties of two component metals, the generation of large amount of residual stress inside SSCP joints is predictable. Because of the difficulty in accurate residual stress measuring by experiment, finite element (FE) method has already become an important approach to welding residual stress prediction. However, there were only few numerical researches reported about welding residual stress of stainless steel clad plates or pipes.

Jiang et al. [40,41,42] established the sequential-coupled FE models of repair welding process for 304/Q345 clad plates based on Abaqus and studied the influence of several welding factors imposing on the residual stress evolution. They [40] pointed out that large and nonuniform residual stress was generated in the repair welds and the increase of repair weld width could reduce the peak residual stress value, as shown in Fig. 7. In addition, they also found that the heat input and repair welds layer number had great effects on residual stress values. As the heat input and welding layer number increased, the residual stresses decreased [41]. Furthermore, Jiang et al. [42] integrated the FEM with neutron diffraction method so as to get the residual stress distribution feature on the clad plate surface and along the thickness direction of the clad plate joint. Results showed that surficial residual stress was concentrated in the heat affected zone and it was much bigger than the yield strength of 304 stainless steel resulted from the work hardening. And meanwhile, the bending stress would generate in the thickness direction.

Fig. 7
figure 7

a FE meshing of the repair model (repair weld width = 8 mm). b Residual stress contours (transverse stress S11, longitudinal stress S33). c Effect of repair width on transverse stress along P4. d Effect of repair width on longitudinal stress along P4 [40]

Full size image

In addition, Dong et al. [43] numerically simulated the butt welding of 2205/X65 bimetallic pipe. Its results showed that residual stress around circumferential direction had homogenous distribution except for the welding start/end region. And welding speed has a significant influence on the distribution of inside and outside surficial residual stress. As welding speed decreased, the position of the maximum stress was farther away from the weld centre, as shown in Fig. 8.

Fig. 8
figure 8

a Dimensions of welded pipe and groove dimension of FE model. b Residual stress distribution of hoop and axial stress on the inside surface. c Residual stress distribution of hoop and axial stress on the outside surface [43]

Full size image

It is not hard to find that large amount of welding residual stress will be generated inside the SSCP joint and its peak value and distribution law are strongly influenced by welding parameters. However, there are few studies focusing on the influence of the welding factors during multi-layer and multi-pass welding procedure on the residual stress evolution feature, such as welding sequences, pass arrangements, groove types, and filling metal types. The influence law of these factors on residual stress is very important to the stress prediction and controlling during welding procedure of SSCP. More intensive work should be done in this regard.

5 Conclusion and future work

This paper reviewed the welding status and issues of SSCP from aspects of conventional multi-layer and multi-pass welding, high efficiency welding, and welding finite element simulation. Based on the reported results, several conclusions can be drawn as follows:

  1. (1)

    Harden martensite phase was bound to generate at the junction of transition layer and the base layer as long as the mixed filling metals were used in the multi-layer and multi-pass welding.

  2. (2)

    It is feasible to weld thin SSCP (≤9mm) with single-pass welding by laser beam welding or laser-arc hybrid welding method. But more work needs to be done on improving the stability of single-pass welding process.

  3. (3)

    Due to the multi-component design, large amount of welding residual stress will be generated inside the SSCP, but few studies focusing on studying the residual stress evolution and controlling of SSCP during welding procedure.

Based on this, we regard that future work can be conducted from the following three aspects:

  1. (1)

    Improvement solutions suitable for SSCP need to be found which is capable of reducing the amount of hardened microstructure. Some low heat input welding methods may be the feasible options.

  2. (2)

    Specific methods capable of improving the welding efficiency of medium and thick SSCP still require further studies. Twin-Electrode TIG Welding (T-TIG) method may be the workable way.

  3. (3)

    The influence of factors during multi-layer and multi-pass welding procedure on the residual stress evolution law, such as welding sequences, passes arrangements, groove types, and filling metal types, can be further explored.