Influence of thermal field in the GMAW process: modelling and comparison with experimental results.

Marta, Constantin ; Doroftei, Ioan ; Suciu, Lenuta 等

1. INTRODUCTION

As in the literature one has already studied the equations of the mathematical model of the welded seam formation, it is only a question of monitoring the evolution of the thermal field in components (basic material), with the purpose of observing the possible modifications of the welding bath, due to the change of the welding regime parameters. For the study one used two methods: an analytical method, consisting in the explicit determination of the thermal field with the help of the formulae from the literature, and a numerical method with finite elements, which will be compared to the results obtained experimentally, for the validation of the utilised software. The safety requirements and the high prices of components are the reason for the use of simulation, and thus manufacture may be optimised if the residual strains occurred as a result of welding influence the deformations appeared at welding through ulterior heat treatments.

2. ANALYTICAL DETERMINATION OF THE THERMAL FIELD

The literature presents equations of the thermal fields, the most frequently encountered in the welding processes, and with their help one may calculate the temperature in a certain point from a welded part, at any moment, both in the welding period, and in the cooling one.

One calculated the thermal fields at different moments in time, with the help of the equation (1), used for permanent mobile source of high power and speed, for the case of plate-shaped bodies:

[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] (1)

[FIGURE 1 OMITTED]

One performed calculation for three distinct positions along the welded seam, with the MATHCAD calculus software and they are presented under a table form, in Table 1, for a medium position of the welded seam, and in Figure 1 these values are compared to the results obtained experimentally.

3. NUMERICAL PROCEDURE

The thermal analysis was performed in non-stationary (transitory) regime. In the case of such an analysis we must establish the duration of the analysis (TIME). Keeping in mind that the length of components is of 120 mm, and the welding speed is of 6 mm/s, from a simple calculus it results that the duration of welding is of 20 seconds. The Ansys software determines the value of the dimensions in knots within certain time intervals, beginning with the initial moment [t.sub.0]=0. Between each two moments consecutively determined there is a time step [DELTA]t. In order to obtain a very high precision of the simulation results, it is necessary that this step be as reduced as possible, but this leads to a very long calculus time and consequently to the increase of the file dimensions. Moreover, one imposes a minimum value of the time step [DELTA]t, so that during it the thermal waves should cross each finite element. For an optimum precision of results, [DELTA]t=0.2 seconds. Knowing that the welding duration is of 20 seconds, one will have 100 time steps. The geometric model of the components (1) crossed by the electric ark (2), produced by electrode (3), during welding, is protected by the protective gas curtain (5), passing through the gas nozzle (4), is presented in Figure 2.

The theoretic study of the welding bath shape was approached by numerous researchers, among whom the Russian researchers Rikalin and Prohorov, who elaborated equations of the isothermal surface corresponding to the melting temperature, surface.

[FIGURE 2 OMITTED]

[FIGURE 3 OMITTED]

The issue is very complex, as each welding procedure requires a separate study, the profile in longitudinal and transversal section of the welding bath being determined by a series of factors connected to the characteristics of the procedure, the basic metal and added material, the nature of the electrode, the manner of source motion, etc. For arc welds, accurate results are obtained with a power density distribution in which surfaces of constant power density are ellipsoids and on radial lines the power density obeys a Gaussian distribution (Goldak et al., 2007). Nevertheless many authors use as tridimensional model a double ellipsoid suggested by the English researcher Goldak, whose model is presented in Figure 3. The software allows the visualisation of the temperature distribution, both in 2D coordinates and in 3D ones, through an isometric representation, according to the same established time periods.

4. RESULTS AND DISCUSSION

The software allows the introduction of the properties of material according to temperature, which leads to the correct estimation of the evolution in space of the thermal field. (Mughal et al., 2005). After the modelling and simulation of the welding processes in an environment of protective gases, it is absolutely necessary to proceed to the confirmation of the theoretical premises by means of the experimental programme in real, effective welding conditions. Thus one measured the temperature in three areas situated along the welded seam, with the help of thermocouples of the R 87%Pt-13%Rh/Pt type, with the diameter of 0.5 mm which can measure temperature up to 1800[degrees]C, in a short period. The situations of the temperature variation in different points on the welded seam, simulated with the help of the Ansys software and calculated analytically, were compared to the results obtained experimentally. One may notice the accuracy with which the curve of the simulated values follows the curve of the measured values, compared to those obtained through analytical calculus.

The application of the finite elements method, as a preliminary method of the analysis for the thermal transfer, the estimation of temperatures in the welded joints and extension of the thermal influence areas in the heterogeneous welded joints constitute a very actual and useful solution for the checking and optimization of the welded technologies, correlating the power produced by the electric arc and welding speed (Wang et al., 2003). If we note that the maximal values of the temperature surpass the ones given by the specialized literature, the diminution of the arc power is imposed, maintaining the same welding speed or the increase of the welding speed is imposed maintaining the same power of the electric arc.

Thus, the finite elements analysis allows a correlation of the electric arc power with the welding speed and the establishment of the optimal welding technology so that the final joint should correspond to the desired characteristics, (Dilthey & Pavlyk 2005).

One presented the values of the temperatures measured experimentally TM, and those obtained by simulation, noted [T.sub.MEF], for the same periods of time, according to Table 2.

[FIGURE 4 OMITTED]

The experimental checking of the temperatures distribution in the welded joints confirms the results obtained through the finite elements analysis which means that the numerical modeling with finite elements of the thermal transfer and the simulation of the continuous process represent useful instruments in establishing the temperatures values, their variations in different areas of the welded joint and the estimation of the welding bath dimensions and of the thermal influence area.

5. CONCLUSION

The necessity of knowing the thermal field in the welded joints is extremely important in the study of the residual tensions and deformations occurring in the welded structure. The application of method of the finite elements, as preliminary method of the thermal transfer analysis, estimation of temperatures in the welded joints and extension of thermal influence areas in the welded joints constitutes a very modern and useful solution in the verification and optimization of welding technologies, correlating the power developed by the electric ark and the welding speed.

If one finds that the maximum values of the temperature exceed those furnished by the literature, it is necessary to reduce the power of the ark, maintaining the same welding speed, or to increase the welding speed maintaining the same power of the electric ark.

The study refers to the possibility of quantifying the modifications which take place due to the temperature variation in the material to be welded and in the welded joint during and at the end of the welding process.

The situation of the temperature variation in different points on the welded joint was simulated with the help of the Ansys program and the experimental results juxtaposed with those obtained by calculation/simulation can validate the accuracy of the calculation program.

6. REFERENCES

Dilthey, U., Pavlyk, V., Integrative MIG/MAG Welding Process Simulation, (2005), In: Proceedings of the 4th German-Japanese Seminar "Materials, Processes and Components", Juli, 2005, Karlsruhe.

Goldak, J., Chakravarti, A., Bibby, M. (1985), A double ellipsoid finite element model for welding heat sources, IIW Doc. 212-603-85.

Goldak, J.; M. Bibby, J. Moore, R. House, & B. Patel, (2007), Computer modeling of heat flow in welds, Metallurgical and Materials Transactions B, ISSN: 1073-5615, Springer Boston, pg. 587-600.

Mughal, M. P., R. A. Mufti, H. Fawad, (2005), Deformation modelling in layered manufacturing of metallic parts using gas metal arc welding: effect of process parameters, Proceedings of the Institution of Mechanical Engineers, Part B: Journal of Engineering Manufacture, ISSN: 09544054, pg. 1499-1509.

Wang, G., Huang, P. G., and Zhang, Y. M., (2003), Numerical Analysis of Metal Transfer in GMAW, University of Kentucky, source: Internet.

Tab. 1. The analytical calculus o the thermal field at the middle
of the welded seam

r[m] 0,06 0,06 0,06 0,06 0,06

t[sec] 0 2 4 6 8

T[[degrees]C] 20 1434, 30 1013, 019 827, 307 716, 554

r[m] 0,06 0,06 0,06 0,06 0,06

t[sec] 10 12 14 16 18

T[[degrees]C] 641, 004 583, 13 541, 74 506, 766 477, 792

r[m] 0,06

t[sec] 20

T[[degrees]C] 453, 814

Tab. 2. Comparative temperature values at the middle of the
welded seam

Timp(s) 0 2 4 6 8 10
[T.sub.M][[degrees]C] 20 1515 1466 1367 1250 1037
[T.sub.MEF][[degrees]C] 20 1535 1421 1327 1200 998

Timp(s) 12 14 16 18 20
[T.sub.M][[degrees]C] 890 721 555 434 311
[T.sub.MEF][[degrees]C] 851 673 505 400 278