Joint geometries and preparation

The plate thickness of all the studied HSSs was 8.0 mm (excluding one piece of 690 MPa yield strength QT HSS, steel G, which was only started being delivered to Finland in 2008 at a minimum thickness of 12 mm). A one side V-groove as seen in fig. 17 was used. The V-groove angle was 60 degrees with an air gap of 1.5 mm and root edge of 1 mm. Test pieces with dimensions of 150 mm x 400 mm were welded together. Fig. 20 shows the preparation of the groove, where run-on and run-off plates were used as tacking. Tacks of welds were first welded to fasten test pieces together, with an advance angle of three degrees estimate angular distortion. Fig. 20 illustrates a good example of a complete penetration.

Fiberglass tape was used as a backing ring, as shown in figs. 18 and 19. The groove was welded on one side with two welding beads, and the 12 mm HSS was welded with three welding beads.

56 Figure 17. One side V-groove preparation.

Figure 18. Fiberglass tape was used as a backing ring.

Picture 19. Glued backing ring.

57

Figure 20. Run-off plates pictured root of groove.

Weldable pieces were fixed between holders in the mechanized machine. The welding torch was installed downright upon the welded groove. Fig. 21 shows the plate and completed torch installation.

Figure 21. Fixed weldment in mechanized machine.

58 6.3. Test set up

While the backing ring was primarily used to make sure that the root edges were completely melted, it is also important to consider that the backing ring shapes the surface of the backing weld. The surfaces of joint preparation were polished between weld passes and the interpass area between welds was not subjected to any heat input, and only experienced room temperature.

Additionally, the amount of free wire, which depends on current levels and the pass that is being welded, was adjusted to appropriate lengths before welding.

The gas run was also adjusted, and all of the welding parameters used, including pWPS’s appear appendices 1, 2 and 3. All the welding parameters used were collected and stored using Kemppi ProWeld Data computer software for research use.

The equation used for t8/5 time was (two dimension heat conduction) according to standard SFS-EN 1011-2

k = thermal efficiency (0.8 in MAG welding) E = welding energy (kJ/mm)

d = thickness of welded piece (mm)

F2 = Coefficient depending the type of joint in two dimensional heat conduction (it is 1.0 when the cooling curve (t8/5) in two dimensional heat conduction is in the oblique area)

Used equation for heat input (Q) was

𝑄=ɳE (2)

59

ɳ is 0.8 through 8 mm plate in MAG welding according to the standard SFS-EN ISO 4063.

𝐸=_1000·𝑣^{60·𝑈·𝐼} (3)

where

E = welding energy (kJ/mm) U = arc voltage (V)

I = welding current (A) v = welding speed (mm/min)

For thicker plates, a three dimensional equation will be used, as follows:

t8/5=(6700-5·T0)·k·E·[( ¹

500−𝑇₀)−( ¹

800−𝑇₀)]·F3 (4) where;

t8/5 = cooling time between 800-500 °C (s) T0 = work temperature (°C)

k = thermal efficiency (0.8 in MAG welding) E = welding energy (kJ/mm)

F3 = Coefficient depending on the type of joint in three dimensional heat con-duction (it is 1.0 when the cooling curve (t8/5) in three dimensional heat conduc-tion is in the oblique area)

The welding parameters used in this study are shown below in table 4.

60

Table 4. MAG welding values in three different test procedures.

HEAT

The calculated cooling times for plates are in table 5.

Table 5. Cooling times for heat inputs of 1.0 kJ/mm, 1.3 kJ/mm and 1.7 kJ/mm when plate thickness is 8 mm, 12mm or 15 mm.

Heat input

The root pass has a lower cooling time. The heat input of the root pass was 0.97 kJ/mm and it is the leader in cooling times with 8 mm plate in 17 s and with 12 mm plate in as low as 7 s.

Using equation 4 the maximum heat input will be 5.0 kJ/mm and the cooling time will again be 21s – the same as was used in the 8 mm thick plate with a heat input of 1.0 kJ/mm. The three dimensional equation for heat input can be used if the plate thickness is more than 46 mm. This plate thickness can be

61

calculated so that equations 1 and 4 will be set even and then the thickness of plate will be calculated using a heat input of 1.0 kJ/mm. For example, if the plate thickness is 20 mm then the heat input can be 2.7 kJ/mm using t8/5 21 s.

6.4. Material properties

HSSs, made by either the TMCP or QT manufacturing methods were the core materials of this study. The chemical properties of these steels are presented in table 6. The chemical properties are specified in the inspection certificate 3.1 (EN 10 204-3.1 2004) provided by the manufacturer. All manufacturers have stated that their steels are made according to the conditions specified in these certificates that were supplied for this study. The conditions under which the steels were created were carefully controlled.

The mechanical properties of the steels in the research are not similar, as shown in table 7. The tensile strength of these steels, which have the required yield strength of 690 MPa, varies between 798 and 879 MPa. One of steels has a tensile strength of 769 MPa, but the standard yield strength value of it is 650 MPa. The change of highest tensile strength is 10 % compared to the lowest value, which is 798 MPa. Additionally, the elongation of HSS is lower than structural steel, at yield strength 235 and 355 MPa, respectively. The change of elongation in the steels used in the experiment is between 15 and 22 %. The lowest elongation percentage was seen in steels B and D, at 15 %, while the highest elongation percentage was found in steel F, at 22 %.

The impact ductility of the steels in this investigation changed between 40 and 194 J, at a temperature of -40 °C, however steels A and C were tested at a temperature of -20 °C. An impact value of at least 27 J is needed for impact ductility. That means that all the reported values in the material certificates are quite exceptional compared material standards, however HSS’s have larger strength tolerance than structure steels.

62

Furthermore, table 6 presents the chemical properties of the steels that are tested with in this research and illustrates that there are varied amounts of alloying elements used in these different steels. For example, steels E and F have the most alloying elements, as Sn is found in steel F and Zr is found is steel E. Comparatively, steel H has much fewer alloying elements. The base elements in HSS are C, Si, Mn, P and S. In addition to these five core elements, steel H only includes two more elements, Cr and Mo. Mo is in all steels in this investigation, while Cr has been used in all QT steels, and Ni has been used in all irrespective of H steel. Carbon is used in the formation of all steel. Steels A and C had the lowest amount of carbon, each with 0.05 % C. The carbon content in the other steels used for the experiments was closer to 0.15 %.

However, this can be explained by the fact that steels A and C are made by the TMCP method and all the others are made by the QT method. Steel B was produced through the quenched and tempered method but additionally has a low notch toughness temperature. The grade of this steel B was S690QL.

These manufacturing specifications emphasize the tough features of Steel B in cold environments up to -40 ⁰C, according to standards SFS-EN 10025-6 + A1.

Aluminium is also found in HSSs and is used in the deoxidation process. Of all the steels used in the scope of this research, only steels G and H do not have any Al. Furthermore, another element found in HSSs is nitrogen, which plays a role in making nitrides such as TiN. Of the steels used for this research, nitrogen is found in five of the eight steels; namely A, C, D, E and F.

Boron is an important alloying element that aids to the hardness of the steel.

Only small amounts of B are needed to do an adequate job, mostly under 0.005

%. B is found in steel B, D, E and F, and the hardness of steels in this investigations was between 270 HV5 and 290 HV5. Other micro alloying elements used in these steels were Nb, V, Cu and Ti.

63

Table 6. Chemical properties of various steels used in the research (wt %). STEELDelivery temper Thick- ness mmC %Si %Mn %

P %S %Al %Cr %Ni %Mo %B %Nb %V %Cu %Ti %N %Sn %Zr %CEVCETPCM A M8 0,0520,191,640,0100,0030,029 - - 0,009 - 0,0460,011 - 0,0910,006 - - 0,340,220,14 B QL8 0,1590,330,820,0080,0010,0490,3 0,050,2230,00170,0040,0100,0250,019 - - - 0,410,280,25 C M8 0,0490,171,860,0080,0040,025 - - 0,008 - 0,0810,009 - 0,0920,005 - - 0,380,240,15 D QT8 0,1300,301,200,0090,0020,0440,260,040,1480,0020,0210,0070,010,0150,004 - - 0,420,280,23 E QT8 0,1370,2761,3900,0130,00130,0610,0520,0660,0290,00210,0220,0010,0200,0020,0050 - 0,00020,390,280,23 F QT8 0,1400,401,410,0110,0040,0370,020,020,0020,0020,0320,060,010,0260,00460,002 - 0,3930,280,24 G QT120,1400,371,210,0130,004 - 0,070,0010,11 - - 0,0010,002 - - - - 0,380,280,22 H QT8 0,1600,240,870,0110,001 - 0,35 - 0,22 - - - - - - - - 0,4190,290,24 M = TMCP QL= Quenched and Tempered + Low notch toughness temperature QT= Quenched + Tempered 𝐶𝐸𝑉=𝐶+𝑀𝑛 6+𝐶𝑟+𝑀𝑜+𝑉 6+𝑁𝑖+𝐶𝑢 15 (1) 𝐶𝐸𝑇=𝐶+𝑀𝑛+𝑀𝑜 10+𝐶𝑟+𝐶𝑢 20+𝑁𝑖 40 (2) 𝑃𝐶𝑀=𝐶+𝑆𝑖 30+𝑀𝑛+𝐶𝑢+𝐶𝑟 20+𝑁𝑖 60+𝑀𝑜 15+𝑉 10+ 5𝐵 (3)

64

Table 7. Mechanical properties of steels used in the research.

STEEL

QL= Quenched and Tempered + Low notch toughness temperature QT= Quenched + Tempered

The filler metal for all these steels was ESAB 12.51. The chemical analysis of which can be seen in table 8. It is an undermatched filler metal, because it has

65

a yield strength 470 MPa. The mechanical properties for this filler metal are in table 9 and a 1.2 mm fillet solid wire was used in the welding. Additionally, the shielding gas was an AGA mixing gas composed at 15 % CO2 and 85 % Ar.

Table 8. Chemical Analysis of filler material OK AUTROD 12.51 (ESAB 2008).

CHEMICAL

Table 9. Mechanical properties of filler material OK AUTROD 12.51 (ESAB 2008).

To know the real content of the weld metal, the area of the first and second pass must be measure from figure first and then calculated (figs. 22 and 23).

Every different alloying element will be calculated one to one. Dilution will happen between the base material and the filler material.

The first pass has a weld metal area of 52 mm x 54 mm= 2808 mm² (the measurements 52 mm and 54 mm are measured from fig. 22). Smelted base material areas are 7 mm x 71 mm= 497 mm² and 6 mm x 53 mm= 336 mm². The sum of the smelted base material areas are 497 mm² + 336 mm² = 833 mm². This is 30 % from all the weld area. A concentration of the alloy elements can be calculated:

The concentration of QT HSS C of the first pass:

Cweld = Cbase material * 0.3 + Cfiller material * 0.7 = 0.137 *0.3 + 0.07 * 0.7 = 0.0901%.

66

The same equation was applied for all the alloy elements in the first pass. The concentrations of the first pass to the welded TMCP HSS E are:

Si = 0.7058%, Mn = 1.432%, P = 0.0123%, S = 0.0144%, Cr = 0.0506%, Ni = 0.0478%, Cu = 0.216%, N = 0.005% and Ti = 0.0076%. Other alloy elements are only in the base material. Then the content of the alloy elements in the weld is 30 % of the base material content. It is likely that the content of Mo was 0.3 x 0.029 % = 0.0087 % in weld and the content of Al was 0.0183%, Nb = 0.0066

%, V = 0.00006 % and B = 0.00063%.

The second pass will be calculated between the base material, the first pass and the filler material.

The second pass has the weld metal which will be calculated in four parts. The first area is 39 mm x 118 mm= 4608 mm². Two triangles, (22 mm * 22 mm)/ 2 = 242 mm² and (22 mm * 27 mm)/ 2 = 297 mm² and second rectangle 13 mm * 22 mm = 286 mm². The sum of weld metal is 5433 mm². Smelted base material areas are 19 mm x 26 mm= 494 mm² and 18 mm x 31 mm= 558 mm². The sum of the smelted base material areas are 494 mm² + 558 mm² = 1052 mm². Smelted first pass was 12 mm x 30 mm= 360 mm². Filler metal was 5433 mm² – 1052 mm² – 360 mm² = 4021 mm². This is 74 % from all the weld area.

Smelted base material was 19 % and smelted first pass was 7 % from all weld metal.

A concentration of the alloy elements can then be calculated:

The concentration of C of the second pass to the QT HSS E,

Cweld = Cbase material * 0.19 + Cfiller material * 0.74 + Cfirst pass * 0.07= 0.137 *0.19 + 0.07 * 0.74 + 0.09 *0.901 = 0.086%.

The same equation will be used for all alloy elements. All concentrations to second pass of welded TMCP HSS E are:

67

Si = 0.76%, Mn = 1.437%, P = 0.0122%, S = 0.016%, Al = 0.0129%, Cr = 0.0504%, Ni = 0.0455%, Mo = 0.00612%, B = 0.00044%, Nb = 0.0046%, V = 0.0002%, Cu < 0.241%, N = 0.005%, Ti = 0.0083% and Zr = 0.00004 %.

Smelted base material Weld area

Figure 22. Principle the figure to calculate the weld metal dilution of the first pass. Aspect ratio of 1:500.

Second pass

Fusion line First pass

Surface of first pass after polishing

Weld metal Smelted base

metal Smelted first

pass

a) b) c)

Figure 23. Principle figure to calculate weld metal dilution of the second pass. a) fusion line and surface of the polished first pass before welding, b) weld metal area, c) smelted base metal and first pass.

Content of alloy elements in weld after welding are in table 10.

68

Table 10. Content of alloy element of QT HSS E in the first and second pass.

Dilution between base material and filler material has happened in all QT HSSs in about the same proportion. This means that the content of alloy elements were at the same levels. In TMCP HSS content of C was less of than in QT HSS. It leads to smaller content of C in the weld of TMCP HSS. As in TMCP HSS C, the C content was in the first pass was 0.3 * 0.05% + 0.7 * 0.07% = 0.064 %, while in the second pass of TMCP HSS C, the C content was 0.05

*0.19 + 0.07 * 0.74 + 0.09 *0.064 = 0.067%.

For all the other alloy elements, the content differences between TMCP and QT HSSs were not large. The content of other alloying elements in the weld was at same level in TMCP HSS as in QT HSS.

69 6.5. Standard tests

A welding procedure test is an inclusive test for welded structures. Using this test, the usability of the welded structure can be examined. In the standard SFS-EN ISO 15164-1 welding procedure test, all of the applicable areas are tested. Testing includes both non-destructive testing (NDT) and destructive testing which shall be in accordance with the requirements of table 11. A description of these tests is provided in the enclosed standards, and all of the welding procedure tests done on all welded pieces were carried out by the chief researcher.

The first test to be conducted was a visual examination of all of the pieces.

Radiographic tests were made using an industrial X-ray machine, RUP-300.

Additionally, penetrant testing was made to all pieces using red penetrating liquid and white development of dye.

Table 11. Examination and testing of the test pieces (standard SFS-EN ISO 15164-1).

70

Metallographic specimens were polished and etched with 4 % Nital (HNO3 + ethanol) before being placed under a conventional light microscope. The polishing automat machine was a Struers TegraPol-31.

Macro- and microscopic examinations were made to all the welded test pieces.

The test machine for the macro photography was Wild M400 macroscope and an Olympus 4040 camera. In addition, microscopic examinations were made on all of the weldments including the HAZ area. Microfilming was made using a light microscope, Zeiss MC63, and the computer software was Isolution Lite.

Additional microscopic test were done at St. Petersburg State Polytechnic University laboratory (StPSPU) using light metallographic microscope LEICA DMI5000M with magnification up to x1000 to clarify the exact microstructure of the HAZ.

Additionally, the impact toughness test was measured using the standard Charpy V-notch impact test (standard SFS-EN ISO 148-1). The test temperature was -40 °C and test machine was model VEB Werkstoffpromachine Leipzig VBN with a load of 150 N. The 5 x 10 mm Charpy test pieces were shaped with a “V” notch of 2 mm depth with the notch tip in conformity with the standards of the HAZ and the weld.

Vickers hardness tests were also performed on the welded specimens, to the SFS-EN ISO 6057-1 standard, using a 5 kg load. Test machine was a Zwick 3202.

Four transverse bending tests were made using standard SFS-EN ISO 5173 to all welded structures, two from the weld surface and two from the root, and the machine used was a bend machine, WPN 20. The same WPN 20 machine was used to make tensile tests with an extensometer. The standard used with the tensile test was SFS-EN ISO 6892-1. The computer software used for this information was PicoLog for windows PLW recorder, and two tests were made to all welded structures.

71

When conducting a transverse bend test on HSSs, the diameter of pusher and opening of drums must be considerably larger than when testing lower yield strength steels. All of these values are found in standards SFS-EN ISO 5173 and SFS-EN ISO 15614-1. For example, an 8 mm thick plate must have a pusher diameter 45 mm and a drum opening of 65 mm, while 12 mm thick plate must have a pusher diameter 75 mm and a drum opening of 105 mm.

According to the standard SFS-EN ISO 5173, the bending angle at which to conduct the bending test should be 180, however the bending machine that was used was limited to a maximum 150 angle.

Equation is

d=(100 x ts)/A-ts and (5) d+3 x ts≥ l >(d+2 x ts) (6) where

d= diameter of pusher

A= minimum ultimate elongation of base metal ts= plate thickness

l= opening of drums

6.6. Additional material test

To confirm that standard SFS-EN ISO 15164-1 test has shown realistic results, an additional material test had to be conducted. CTOD tests and microstructure analysis, like analysing different faces and micro hardness, were done. HAZs were also calculated and CTOD tests were done to all HSS steels. Additional microstructure tests were performed on QT HSS steel (steel E) and TMCP HSS steel (steel C).

72 6.6.1. CTOD test

In order to check if the impact toughness values were correct, a CTOD test was made on all welded structures. Fig. 24 shows the construction of the welded pieces with dimensions of 8 x 15 x 50 mm. The 8 mm in thick, 50 mm length and 15 mm lateral pieces were cut from the whole plate. Using tack welds these pieces were welded together and the fusion faces were machined. The gap was about 1.5 mm and root edge was about 1.0 mm, while a single-bevel (½-V) groove was used with a flank angle of 45 degrees.

Fig. 25 illustrates the welded pieces before they were separated by saw. The beginning and end of the groove were made with assisting pieces. The test pieces with dimension 5 mm width, 10 mm high and 50 mm long were made by machining after cutting.

Figure 24. Used one side single be-vel (½-V) groove in CTOD tests.

Figure 25. CTOD test pieces after welding.

73

In fig. 26 there is an etched CTOD test piece where the red line indicates the fusion line, the blue line indicates the start notch, the yellow line indicates the fatigue notch, and the green line indicates the test area. The groove was welded with three or four beads according to the heat input. The same three heat inputs (1.0; 1.3 and 1.7 kJ/mm) were used as in previous tests.

Figure 26. Etched CTOD test piece.

The CTOD test equipment was made by the Welding Technology Laboratory at Lappeenranta University of Technology. Fig. 27 illustrates the pusher and its counterpart, while fig. 28 is a picture of the actual machine used for the testing.

74

Figure 27. CTOD test components and test piece.

COMPUTER AND SOFTWARE PUSHER AND ITS

COUNTERPART

COOLING UNIT

FATIGUE TEST MACHINE

Figure 28. CTOD test machine.

75

The testing temperature was -40°C. Ethanol was used to guarantee the con-stancy of the temperature, while and temperature adjustments were made with the application or removal of dry ice. Fig. 29 illustrates the equipment at the -40

°C test temperature.

Figure 29. Isolated equipment at -40 °C and liquid intermediate test agent.

As the size of the CGHAZ is quite small, a study of this region is particularly difficult in real welds. Therefore, a thermal simulation was used to generate a relatively large region of CGHAZ, which allowed the notch to be reliably located in the correct microstructure. The steels were subjected to a welding thermal simulation. Thermal simulation test blanks were cut from the surface position of each plate, with the test piece axis transverse to the rolling direction, in T-L direction. Fig. 30 shows the test blanks, 8 x 17 mm in size. After the thermal simulation, these blanks were machined down to a 5 x 10 mm size appropriate for CTOD test pieces. The weld HAZ thermal simulations were performed on a Gleeble 3800 simulator, as the one shared in fig. 31, which is owned by the StPSPU.

76

Figure 30. Test pieces proportion to rolling direction.

Figure 30. Test pieces proportion to rolling direction.

In document The effects of welding heat input on the usability of high strength steels in welded structures (sivua 57-0)