·Table of Contents
·Industrial Plants and Structures
Problems Linked to Laser Welding of components having Large ThicknessesG. Caironi
Dipartimento di Meccanica - Politecnico di Milano
consultant, Riva Techint
|Parent metal||Thickness [mm]||Power [kW]||Welding speed [m/min]||Heat input [kJ/m]||Wire feed rate [m/min]|
|3||Fe430B||15||I - 14.0||0.84||1110||1.60|
|II - 12.1||0.96||680||630|
|4||Fe430B||25||I - 16.0||0.6||1440||1.75|
|II - 13.1||0.6||1180||7.90|
|III - 12.2||0.72||916||7.90|
|5||Fe510C||34||I - 15.35||0.60||1381||1.25|
|II - 15.0||0.60||1350||7.0|
|III - 15.0||0.60||1350||7.9|
|IV - 13.0||0.48||1462||7.9|
|Table 1: Welding parameters used in production.|
1. WELDED JOINT WITHOUT FILLER WIRE
The sample was obtained from butt welding two sheets having a thickness of 12 mm with straight edges (fig. 1).
|Fig 1: Macrograph of the joined sheets of 12 mm thickness without filler wire.|
The junction was made in a single pass without using filler wire. The sample's specifications and the final parameters used for the welding are given in table 1 under point 1. The base metal, Fe 430 B, is characterised by a ferritic-pearlitic structure with 30-35% pearlite. Areas of bainite and martensite were detected in the melted area and at the same time there was a presence of porosity. The pearlite's dissolution with small areas of bainite and martensite is instead evident in the ZTA.
The joint's microhardness profiles are given in figure 2.
|Fig 2: Microhardness in weld and in parent metal.|
The increases of hardness in correspondence with the ZTA and Z.F. can be associated with the presence of martensitic and bainitic structures detected in the microstructural analysis.
The bead has the characteristic nail-head shape, which is typical of laser welding.
The Z.F. has dendrites with a columnar type of growth in a direction that is at a right angle to the axis itself. The size of the dendrites decreases while moving in the direction of the apex due to the increase of the cooling speed.
The welding has an underthickness defect caused by the lack of material and by excessive evaporation during its interaction with the laser beam. There is a longitudinal crack starting from the bead's head that is caused by the solid external mass's poor contribution of heat and excessive thermal removal, which brings about rapid solidification of the melted metal and a consequent formation of the crack.
There is a gas inclusion in the intermediate part. Its formation is due to the high viscosity of the bead that results in incorporating the gas bubbles present during the solidification phase. Their origin is attributable to the presence of substances that can be easily vaporised on the welding edges and to the air contained in the gaps between the edges. These defects can be reduced by thoroughly cleaning the edges, with proper presentation of the edges to be welded and by modifying the working parameters to reduce or improve the metallic bead-agitation.
The inclusion of S and Ca was observed in the Z.F. at the centre of the bead (figure 3).
Fig 3: Inclusion of S and Ca observed at the center of the bead.
2. WELDED JOINT WITH FILLER WIRE
The sample was obtained from the butt welding of two sheets having a thickness of 12 mm (fig. 4) with straight edges. In comparison with the previous case, the power capacity was increased and the gap size was decreased from 0 to 0.5 mm.
The junction was made in a single pass by using filler wire having a 1.2-mm diameter and in WA2 material to try to eliminate, during a preliminary research stage, the longitudinal crack that showed up in the previous sample.
The sample displays the presence of a gas cavity near the apex of the welding.
The Fe 430 B base metal presents the same microstructure of the previous sample. The sample's features and the final parameters used for the welding are given in table 1 under point 2.
Metallographic observation detected a structure that is similar to the previous case. In the ZTA, instead, dissolution of the pearlite with small areas of bainite and martensite is evident.
The sample presents a macroscopic defect that is attributable to gas inclusion when the melted bead is re-mixed. Analysis of the same defect permits us to show that it involves only one cavity coming to the surface in various places. Analysis of the SEM showed uniformly distributed porosity in the Z.F. It was probably caused by the evaporation of some elements in the alloy, of uncontrolled agitation and the bead's viscosity that may be associated with an incorrect position of the welding head, by the presence of polluting substances that can easily be vaporised on the welding edges and by air and/or moisture trapped inside the gaps on the edges. The presence of a non-metallic inclusion made up of calcium sulphide was observed.
3. WELDED JOINT IN THREE PASSES
The sample was obtained from the butt welding of sheets having a thickness of 25 mm and straight edges (fig. 5).
The junction was made with three overlapping passes and by using WA2 filler metal in 1.2-mm diameter wire. The base metal is Fe 430 B. The sample's features and the final parameters used for the welding are given in table 1 under point 4.
The Z.F.'s structure is made up of pearlite with the presence of bainite and martensite. Furthermore, a ferritic-pearlitic structure with some tempered martensite was observed in the intersecting area between the two passes. The heat treatment's effect caused by the following passes on top of the previous ones is obvious. The dendrites show up as being particularly coarse in the second and third pass. The passes show a satisfactory alignment, which is a sign of a fair centring of the beam. For welding components having a long length, centring of the beam is done by using a joint follower made with a low-power laser to keep the beam centred on the other passes. Even when using beam-centring systems, you can verify slight maladjustments of the passes due to the large thickness. The sample shows a defect of incomplete penetration of the first pass. Also a longitudinal macro crack, found on the third pass's Z.F. axis and caused by poor heat contribution, is evident. The Z.F. of each pass shows numerous non-metallic inclusions of sulphides.
|Fig 4: Macrograph of the joined sheets of 12 mm thickness using filler wire.||Fig 5: Macrograph of the joined sheets of 25 mm thickness using filler wire with three overlapping passes.||Fig 6: Macrograph of the joined sheets of 34 mm thickness using filler wire with four overlapping passes.|
4. WELDED JOINT IN FOUR PASSES
The sample was obtained from the head junction of sheets having a thickness of 34 mm (fig. 6).
The welding was performed in four passes with WA2 filler wire with a 1.2-mm diameter. The base metal is Fe 510 C. The sample's features and the final parameters used for the welding are given in table 1 under point 5.
Compared to the previous trial, the type of base metal, thickness, and power and speed values were changed. Furthermore, the number of passes was increased from 3 to 4 in order to give the welding a greater thickness.
The four passes' Z.F. show a ferritic-pearlitic structure with more or less marked traces of martensite in quantities that slightly vary according to the various passes. In the third pass's Z.F. we also detect the presence of tempered martensite. In the ZTA, dissolution of the pearlite is shown, with some martensite and upper bainite. Finally, we observed the presence of tempered martensite, caused by the previous pass's heat treatment, in the passes' overlapping area. The sample shows a cavity, attributable to a gas inclusion, in correspondence with the apex of the fourth pass. The dendrites are particularly coarse in the second, third and fourth passes due to the lower cooling speed. The passes display satisfactory adjustment, which is the sign of a fair centring of the beam. The sample presents small cavities in correspondence with the apex of the fourth pass, and this is attributable to gas inclusions. The presence of small manganese sulphide inclusions in the melted area were detected, and their origin is attributable to the base metal. Uniformly distributed porosity was detected in the melted areas of all four passes, and with a more marked distribution at the apex of the first pass. Observation of the ZTA did not reveal the presence of cracks. We verified the onset of longitudinal cracks that involve the entire bead quite often in the first two passes. The causes are associated with the high heat contribution and rapid heat disposal caused by the different melted bead volume compared to the high thickness of the surrounding metal. A problem linked to this type of defect is the inspection bodies' requirement that there be no defects after each pass. The execution of later passes causes remelting of the area affected by cracks and the consequent elimination of them.
|© AIPnD , created by NDT.net|||Home| |Top||