Thread Content
01 Causes of fatigue failure in welded structures. The main reasons for fatigue failure in welded structures are as follows: (1) Objectively speaking, the static load-bearing capacity of welded joints is generally not lower than that of the base material； However, under cyclic loading, its load-bearing capacity is much lower than that of the base metal, and it is closely related to the type of weld joint and the form of the welded structure. This is a major factor that causes some structures to fail prematurely due to fatigue of the weld joints ； (2) Early welding structure design focused primarily on static load strength, with no consideration given to fatigue resistance; or the fatigue design standards for welding structures were not well-developed, which led to the creation of many welding joints that appear unreasonable by today’s standards ； (3) Engineering design technicians have insufficient understanding of the characteristics of the fatigue resistance of welded structures; consequently, the welded structures they design often simply adopt the fatigue design criteria and structural forms used for other metal structures ； (4) Welded structures are being used increasingly widely, and the blind pursuit of low cost and light weight in their design and manufacturing leads to increasingly higher design loads for such structures ； (5) Welded structures are showing a trend toward higher speeds and greater loads; there are increasing demands for their ability to withstand dynamic loads, yet research on the fatigue strength of welded structures lags behind relatively. 02 Main factors affecting the fatigue strength of welded structures 2.1 Influence of static load strength on the fatigue strength of welded structures In the study of steel materials, there is always a desire for these materials to have high specific strength, that is, to be able to bear large loads with a relatively low weight; this is because structures of the same weight can possess great load-bearing capacity ； Or the same load-bearing capacity can reduce its own weight. Thus, high-strength steel was developed; it also possesses high fatigue strength. The fatigue strength of base metals always increases as their static strength increases. However, the situation is different for welded structures, as the fatigue strength of a welded joint has little to do with the static strength of the base material, the static strength of the weld metal, the microstructural properties of the heat-affected zone, or the strength of the weld metal. In other words, as long as the details of the welded joint remain the same, the fatigue strength of high-strength steel and low-carbon steel is identical, and they exhibit the same S-N curve. This principle applies to various types of joints, including butt joints, fillet joints, and welded beams. Maddox studied the fatigue crack propagation in carbon-manganese steels with a yield point ranging from 386 to 636 MPa, as well as in the weld metals and heat-affected zones welded using six types of electrodes. The results showed that the mechanical properties of the materials have a certain effect on the crack propagation rate, but this effect is not significant. When designing welded structures subjected to alternating loads, it is pointless to attempt to meet engineering requirements by using steel grades with higher strength. Only when the stress ratio is greater than +0.5 and the static strength condition plays a dominant role should high-strength steel be used for the base metal of welded joints. The reason for the above result is the presence of slag wedge defects similar to undercut at the weld toe of the joint along the fusion line; these defects have a thickness of 0.075–0.5 mm, with a tip radius of less than 0.015 mm. This sharp defect is where fatigue cracks originate; it corresponds to the fatigue crack initiation stage. Therefore, the fatigue life of the joint under a certain stress amplitude is primarily determined by the fatigue crack propagation stage. The presence of these defects results in all weld joints of the same type of steel having the same fatigue strength, with little relation to the static strength of the base metal and the welding materials. 2.2 Effect of stress concentration on fatigue strength Figure 2.2.1 Influence of joint type. The main types of welded joints are butt joints, cross joints, T-joints, and lap joints; at these joints, stress concentration occurs due to the disruption of the force transmission paths. The force lines in butt joints experience minimal interference; therefore, the stress concentration factor is small, and their fatigue strength is also higher than that of other joint types. However, experiments show that the fatigue strength of butt joints varies over a wide range, as a number of factors affect the fatigue performance of such joints. Factors such as the size of the specimen, groove shape, welding method, electrode type, welding position, weld shape, post-welding weld processing, and post-welding heat treatment all have an impact on it. Butt joints with permanent gussets have reduced fatigue strength due to severe stress concentration at the gusset location. The fatigue cracks in such joints originate from the junction of the weld and the gasket, rather than at the weld toe, and their fatigue strength is generally equal to that of butt joints with the worst geometry without a gasket. Cross joints or T-joints are widely used in welded structures. In such load-bearing joints, due to the significant change in cross-section at the transition from the weld to the base metal, their stress concentration factor is higher than that of butt joints; as a result, the fatigue strength of cross or T-joints is lower than that of butt joints. For joints connected by fillet welds without grooves and grooved joints with partially penetrated welds, when the welds carry working stresses, fatigue fracture may occur at two weak points: the junction between the base metal and the weld toe, or within the weld itself. For cross joints with full penetration obtained through grooving, fracture generally occurs only at the weld toe, rather than in the weld seam. The fatigue strength of T-joints and cross joints, in which the welds are not subjected to working stresses, depends mainly on the stress concentration at the junction between the welds and the plates that bear the main stresses. T-joints have higher fatigue strength, whereas cross joints have lower fatigue strength. The fundamental measure to improve the fatigue strength of T-joints or cross joints is to weld with bevels and to shape the weld transitions to ensure a smooth transition; through such improvements, the fatigue strength can be increased significantly. The fatigue strength of lap joints is very low, due to the severe distortion of the force line. It is highly unreasonable to use butt joints with so-called \"reinforced\" covers, as this increases the effect of stress concentration; as a result, the butt joints, which originally had high fatigue strength, are significantly weakened by the use of these covers. For load-bearing cover plate joints, fatigue cracks can occur in the base metal as well as in the welds. Additionally, changing the width of the cover plate or the length of the welds alters the stress distribution within the base metal, thereby affecting the fatigue strength of the joint. That is, as the ratio of weld length to cover plate width increases, the fatigue strength of the joint increases, due to the more uniform stress distribution in the base metal. 2.2.2 Influence of weld shape: Regardless of the type of joint, they are all composed of two types of welds: butt welds and fillet welds. Different weld shapes result in different stress concentration coefficients, leading to considerable variability in fatigue strength. The shape of the butt weld has the greatest influence on the fatigue strength of the joint. (1) Effect of the transition angle: Yamaguchi et al. established a relationship between fatigue strength and the transition angle (external obtuse angle) between the base metal and the weld metal. In the experiments, W (weld width) and h (height) varied, but the h/W ratio remained constant. This means that the angle remains unchanged, and test results show that the fatigue strength also remains unchanged. However, if W remains constant while the variable parameter h changes, it is found that as h increases, the fatigue strength of the joint decreases, which is clearly a result of the decrease in the external angle. (2) Influence of the weld transition radius. The research results of Sander et al. indicate that the weld transition radius also has a significant effect on the fatigue strength of joints; that is, as the transition radius increases (while the transition angle remains constant), the fatigue strength also increases. The shape of the fillet weld also has a significant impact on the fatigue strength of the joint. When the ratio of the calculated thickness a of a single weld to the plate thickness B, namely a/B0.7, is within a certain range, fracture generally occurs in the base metal. However, increasing the weld size is only effective in improving fatigue strength within a certain range. Since an increase in weld size does not change the strength of the base metal at the other weak section, namely the weld toe, it can at most not exceed the fatigue strength of that area. Soete and Van Crombrugge used 15mm-thick plates welded with different fillet welds, and tests under axial fatigue loading showed that when the root thickness of the weld was 13mm, fracture occurred at the weld toe in either the base metal or the weld itself. When the root depth of the weld is less than this value, fatigue fracture occurs at the weld ； When the weld leg size is 18 mm, fracture occurs in the base metal. Based on this, they proposed the limit weld leg size: S=0.85B, where S is the weld leg size and B is the plate thickness. It can be seen that even when the root gap size reaches the plate thickness (15 mm), fracture at the weld seam still occurs, a result that is in good agreement with the theoretical predictions. Image 2.2.3 Effects of welding defects. There are numerous types of defects at the weld toe; these various defects lead to premature initiation of fatigue cracks and a significant reduction in the fatigue strength of the base material (by 80%). Welding defects can be broadly divided into two categories: surface defects (such as cracks and lack of fusion) and volumetric defects (such as pores and inclusions). The severity of their impact varies, and the effect of welding defects on the fatigue strength of joints depends on the type, direction, and location of the defects. (1) Cracks: Cracks in welding, such as cold and hot cracks, are severe sources of stress concentration; in addition to having a brittle microstructure, they can significantly reduce the fatigue strength of the structure or joint. Earlier studies have shown that in butt joint specimens made of low-carbon steel with a width of 60 mm and a thickness of 12.7 mm, when there are cracks measuring 25 mm in length and 5.2 mm in depth within the weld (which account for approximately 10% of the specimen’s cross-sectional area), the fatigue strength at a life of 2×10^6 cycles under alternating load conditions decreases by about 55%–65%. (2) Lack of penetration: It should be noted that lack of penetration is not necessarily considered a defect, as sometimes it is deliberately required that certain joints be fully penetrated; a typical example is the design of nozzles on certain pressure vessels. The lack of full penetration defect can be a surface defect (in single-sided welds) or an internal defect (in double-sided welds); it may be localized or affect the entire area. Its main effects are to reduce the cross-sectional area and cause stress concentration. Compared with the test results for fatigue life at a reduction area of 10%, the fatigue strength decreased by 25%, which indicates that the impact is less severe than that of cracks. (3) Non-welding: Due to the difficulty in preparing specimens, there are extremely few studies on this topic to date. However, there is no doubt that lack of fusion is a planar defect; thus, it cannot be ignored. Generally, it is treated as equivalent to incomplete penetration. (4) Undercut. The main parameters characterizing undercut are the undercut length L, undercut depth h, and undercut width W. The main parameter affecting fatigue strength is the kerf depth h; currently, either the depth h or the ratio of depth to plate thickness (h/B) can be used as a parameter to evaluate the fatigue strength of the joint. (5) Porosity is a volumetric defect. Harrison analyzed and summarized previous experimental results, finding that the decrease in fatigue strength is mainly due to porosity reducing the cross-sectional area, and there is a certain linear relationship between them. However, some studies have shown that when machining methods are used to process the specimen surface such that pores are located on the surface or just beneath it, their adverse effects become more pronounced; they act as stress concentrators and serve as initiation points for fatigue cracks. This indicates that the location of pores has a greater influence on the fatigue strength of joints than their size; surface or subsurface pores have the most detrimental effect. (6) Relevant research reports on slag inclusions IIW indicate that, as a volumetric defect, slag inclusions have a greater impact on the fatigue strength of joints than pores. As can be seen from the above discussion, the effect of welding defects on the fatigue strength of joints depends not only on the size of the defects but also on many other factors. For example, surface defects have a greater impact than internal defects, and planar defects perpendicular to the direction of the applied force have a greater influence than those in other directions ； The effect of defects located in the residual tensile stress region is greater than that in the residual compressive stress region ； Defects located in stress concentration areas (such as crack at the weld toe) have a greater impact than identical defects in a uniform stress field. 2.3 Effect of welding residual stress on fatigue strength Welding residual stress is a characteristic unique to welded structures; therefore, its impact on the fatigue strength of such structures is a matter of great concern, and numerous experimental studies have been conducted on this topic. Tests often involve conducting fatigue tests on specimens with welding residual stresses and those whose residual stresses have been eliminated through heat treatment, for comparison purposes. Since the generation of welding residual stresses is often accompanied by changes in material properties caused by the welding thermal cycle, and since heat treatment eliminates residual stresses while restoring or partially restoring the material’s properties, coupled with the variability of test results, different interpretations of these results arise; consequently, there are also differing evaluations regarding the effects of welding residual stresses. To illustrate this issue clearly, let us take some research works conducted by various researchers in the early and recent periods as examples. Regarding the results of 2×10^6 cycle tests on butt joints with overfill, different researchers reached different conclusions. It has been found that the fatigue strength of stress-relieved specimens after heat treatment is 12.5% higher than that of identical specimens in their as-welded state ； Others have found that the fatigue strength of as-welded and heat-treated specimens is similar; in other words, there is little difference between them ； However, it has also been found that although fatigue strength increases after heat treatment is used to eliminate residual stresses, the increase is far below 12.5%, among other things. The same is true for the test results of butt joint specimens with surface grinding; some tests suggest that heat treatment can increase fatigue strength by 17%, while other test results indicate no improvement in fatigue strength after heat treatment. This issue has puzzled people for a long time, until a series of experiments conducted under alternating loads by some scholars from the former Soviet Union gradually clarified it. Among them, worthy of mention is Trufyakov’s research on the effect of welding residual stresses on the fatigue strength of joints under different stress cycling characteristics. The test used ordinary low-alloy structural steel 14Mn2; the specimen had a transverse butt weld, with one longitudinal weld bead deposited on each of the front and back surfaces. One group of specimens underwent heat treatment to relieve residual stresses after welding, while the other group was not subjected to any heat treatment. The fatigue strength comparison test used three stress cycle characteristic coefficients of r = -1, 0, +0.3. Under alternating loading (r=-1), the fatigue strength of the specimens with residual stresses removed was close to 130 MPa, whereas that of the specimens without residual stress removal was only 75 MPa. Under pulsating loading (r=0), the fatigue strength of both groups of specimens was the same, at 185 MPa. When r=0.3, the fatigue strength of the specimens with residual stresses removed through heat treatment was 260 MPa, which was slightly lower than that of the untreated specimens (270 MPa). The main reason for this phenomenon is that at higher values of r, such as under pulsating loads (r=0), the fatigue strength is higher; under higher tensile stresses, the residual stresses are released more rapidly, and thus the influence of residual stresses on fatigue strength is reduced ； When r increases to 0.3, the residual stress further decreases under load, and it no longer has any effect on fatigue strength. Heat treatment softens the material while eliminating residual stresses, which as a result causes the fatigue strength to decrease after heat treatment. This experiment provides a good illustration of the effect of residual stress and material changes caused by the welding heat cycle on fatigue strength. It can also be seen from this that the effect of welding residual stress on the fatigue strength of joints is related to the stress cycling characteristics of the fatigue load. That is, when the cyclic characteristic value is low, the impact is relatively large. As previously mentioned, due to the residual stresses in the structural welds that reach the material’s yield point, in joints subjected to stress cycles of constant amplitude, the actual stress cycles experienced near the welds will fluctuate below the material’s yield point, regardless of the characteristics of the original stress cycles applied. For example, if the nominal stress cycle is from +S1 to -S2, then its stress range should be S1+S2. However, the actual stress cycle range in the joint will be from Sy (the stress amplitude at yield) to Sy- (S1+S2). This is very important when studying the fatigue strength of welded joints, as it has led to some design codes replacing the cyclic characteristic r with stress ranges. Furthermore, during the testing process, the dimensions of the specimen, the loading method, the stress cycle ratio, and the load spectrum also have a significant impact on fatigue strength. 03 Process methods for improving the fatigue strength of welded structures. The typical initiation sites of fatigue cracks in welded joints are at the weld root and weld toe. If the risk of fatigue crack initiation at the weld root is minimized, the critical points for the welded joint then become concentrated at the weld toe. Many methods can be used to improve the fatigue strength of welded joints. (1) Reduce or eliminate welding defects, especially open defects ； (2) Improve the geometry of the weld toe to reduce the stress concentration factor ； (3) Adjust the welding residual stress field to generate a residual compressive stress field. These improvement methods can be divided into two major categories, as shown in Table 1. Welding process optimization methods are considered not only to improve the fatigue strength of welded structures, but they also bring significant benefits to various aspects such as the static load strength of these structures and the metallurgical properties of the weld joints. There is ample literature on this topic, so no further details will be provided here.