• No results found

Application of Expert Methods for Predicting Liquidus Temperature of TIG Weldments

N/A
N/A
Protected

Academic year: 2022

Share "Application of Expert Methods for Predicting Liquidus Temperature of TIG Weldments"

Copied!
5
0
0

Loading.... (view fulltext now)

Full text

(1)

Abstract—Liquidus temperature is regarded as the minimum temperature required for an alloy to completely transform into the liquid state. Uncontrolled temperature leads to excessive heat generated in the material which create wider heat affected zone, alters the microstructure of the material and also induce residual stresses in the material. Optimizing this process is one sure way of producing a quality weld. In this study, the application of expert systems such as response surface method to optimize the liquidus temperature was pursued. The central composite design matrix was employed to collect data from the sets of experiments. The specimen was made from mild steel plates and welded with the tungsten inert gas process. The result of the response surface method shows that current has a very strong influence on the liquidus temperature. The model for optimizing liquidus temperature has a P-value < 0.0001. The model developed had a very high noise to signal ratio (S/N). Finally, the numerical solution obtained shows that a current of 130Amp, a voltage of 20.94volts, and a speed of 0.48m/min produced a result with liquidus temperature of 1365.05oC.

Index Terms—Liduidus, RSM, TIG, Weld.

I. INTRODUCTION

Welding is the process of joining two pieces of metal by creating a strong metallurgical bond between them by either applying heat or pressure, or both to obtain a permanent bond [1],[2]. The tendency of atoms to bond is the fundamental basis of welding. Radhakrishnan [3] stated that as the demand for welding new materials and larger thickness components increased, mere gas flame welding, which was utilized by earlier welding engineers, is no longer up to par with current demands, and over the years welding has discussion improved by incorporating better suited welding methods such as Metal Inert Gas Welding, Tungsten Inert Gas Welding (TIG), and Electron and Laser Beam Welding [4],[5].

According to [6], Tungsten inert gas (TIG) welding is a thermal process that depends upon heat conducted through the weld joint materials. The melting temperature necessary to weld materials in the TIG welding is obtained by inducing an arc between a tungsten electrode and the work piece. The weld pool temperatures can advance up to 2500o. In TIG welding, a non-consumable tungsten electrode of diameter between 0.5 to 6.5 mm is employed with an inert shielding gas [7]. The shielding gas shields both the tungsten electrode and the weld pool from the detrimental effects of surrounding atmospheric gases. Argon is the most

Published on May 9, 2019.

Authors are with the Department of Production Engineering, University of Benin, Benin City, Edo State, Nigeria. (frank.uwoghiren@yahoo.com,

common shielding gas employed in welding low carbon steels and stainless steels, both power supply sources (AC or DC) can be used in this process. In general, this process uses a direct current (DC) arc, where the tungsten electrode has a negative polarity, consequently the tungsten electrode turns into the cathode and the work piece turns into the anode and the polarity is known as straight polarity or direct current electrode negative (DCEN) [7]. All welding position capabilities are present in this process while others are limited to one or a few welding positions. There are different methods and standards adopted and there is still a continuous search for new and improved method of welding.

Radhakrishnan [3] stated that as the demand for welding new materials and larger thickness components increased, mere gas flame welding, which was first known to the welding engineer, is no longer satisfactory and has been improved to welding methods such as metal inert gas welding, tungsten inert gas welding, electron and laser beam welding. Welding is the process of joining two pieces of metal by creating a strong metallurgical bond between them by heating or pressure or both. A welded joint is obtained when two clean surfaces are brought into contact with each other and either pressure or heat, or both are applied to obtain a bond [8]. The tendency of atoms to bond is the fundamental basis of welding. The basic equipment for TIG welding comprises a power source, a welding torch, a supply of an inert shield gas, a supply of filler wire and water cooling system. TIG welding finds its application in pressure vessels, aero, rocket, missile, nuclear and marine industries. TIG welding has experienced great improvement by researchers over the years since its invention (late 1930s) [9].

II. MATERIALS AND METHODS

The key parameters considered in this work are welding current, welding speed and welding voltage. The range of the process parameters obtained from literature is shown in Table I. 100 pieces of mild steel coupons measuring 60mm x 40mm x 10mm were used for this experiments. The experiment was performed 20 times, using 5 specimens for each run. Fig. 1 shows the TIG welding setup. The welding process uses a shielding gas to protect the weld specimen from atmospheric interaction, 100% pure Argon gas was used in this research study. Fig. 3 shows the. A cable is used to connect the thermocouple to the weld specimen as shown in Fig. 4. The central composite design matrix was developed, using the design expert software, producing 20 experimental runs. The input parameters and output parameters make up the experimental matrix and the responses recorded from the weld samples were used as the

Application of Expert Methods for Predicting Liquidus Temperature of TIG Weldments

F. O. Uwoghiren, A. Ozigagun and T. B. Adeleke

(2)

data. Table II shows the central composite design matrix

TABLEI:PROCESS PARAMETERS AND THEIR LEVELS Parameters Unit Symbol Coded value Coded value

Low(-1) High(+1)

Current Amp A 100 180

welding speed, M/min F 0.10 0.6

Voltage Volt V 16 22

Fig. 1. TIG equipment

Fig. 2. Shielding gas cylinder and regulator

Fig. 3. Weld samples

Fig. 4. Thermocouple connection cable

TABLEII:CENTRAL COMPOSITE DESIGN MATRIX (CCD) std Run A: Current (Amp) B: Voltage (V) C: Welding speed (m/min)

15 1 145,00 19,50 0,35

16 2 145,00 19,50 0,35

17 3 145,00 19,50 0,35

18 4 145,00 19,50 0,35

19 5 145,00 19,50 0,35

20 6 145,00 19,50 0,35

9 7 119,77 19,50 0,35

10 8 170,23 19,50 0,35

11 9 145,00 16,98 0,35

12 10 145,00 22,02 0,35

13 11 145,00 19,50 0,10

14 12 145,00 19,50 0,60

1 13 130,00 18,00 0,20

2 14 160,00 18,00 0,20

3 15 130,00 21,00 0,20

4 16 160,00 21,00 0,20

5 17 130,00 18,00 0,50

6 18 160,00 18,00 0,50

7 19 130,00 21,00 0,50

8 20 160,00 21,00 0,50

III. RESULTS

The randomized design matrix comprising of three input variables namely; current (Amp), voltage (V), welding speed (m/min) and four response variables namely (arc length, liquidus temperature, heat input and heat affected zone) in real values is presented in Table III.

TABLEIII:DESIGN MATRIX SHOWING THE REAL VALUES AND THE EXPERIMENTAL VALUES

std Run A:

Current (Amp)

B:

Voltage (V)

C: Welding speed (m/min)

Liquidus Temp(degree C)

15 1 145,00 19,50 0,35 1187

16 2 145,00 19,50 0,35 1208

17 3 145,00 19,50 0,35 1208

18 4 145,00 19,50 0,35 1214

19 5 145,00 19,50 0,35 1236

20 6 145,00 19,50 0,35 1220

9 7 119,77 19,50 0,35 1257

10 8 170,23 19,50 0,35 1523

11 9 145,00 16,98 0,35 1326

12 10 145,00 22,02 0,35 1265

13 11 145,00 19,50 0,10 1265

14 12 145,00 19,50 0,60 1280

1 13 130,00 18,00 0,20 1367

2 14 160,00 18,00 0,20 1548

3 15 130,00 21,00 0,20 1116

4 16 160,00 21,00 0,20 1302

5 17 130,00 18,00 0,50 1236

6 18 160,00 18,00 0,50 1361

7 19 130,00 21,00 0,50 1420

8 20 160,00 21,00 0,50 1548

The model statistics computed for liquidus temperature based on the different model sources is presented in Fig. 5.

(3)

Fig. 5. Model summary statistics for liquidus temperature

The model summary statistics of models fit shows the standard deviation, the r-squared and adjusted r-squared, predicted r-squared and the PRESS statistic for each complete model. In assessing the strength of the quadratic model towards maximizing the liquidus temperature, one- way analysis of variance (ANOVA) table was generated for maximizing the liquidus temperature and result obtained is presented in Fig. 6.

Fig. 6. ANOVA table for validating the model significance towards maximizing the liquidus temperature

Analysis of variance(ANOVA) was needed to check whether or not the model is significant and also to evaluate the significant contributions of each individual variable, the combined and quadratic effects towards each response.

Values of "Prob > F" less than 0.0500 indicate model terms are significant. In this case A, BC, A2, B2, C2 are significant model terms. Values greater than 0.1000 indicate the model terms are not significant. The "Lack of Fit F-value" of 5.05 implies the Lack of Fit is not significant relative to the pure error. There is only a 65.00% chance that a "Lack of Fit F- value" this large could occur due to noise. Non-significant lack of fit is good as it indicates a model that is significant.

Fig. 7. GOF statistics for validating model significance towards maximizing LT

The optimal equation which shows the individual effects and combine interactions of the selected input variables (current (Amp), voltage (V) and welding speed (m/min)) against the mesured response (liquidus temperature) is presented based on the actual factors in Fig. 8.

Fig. 8. Optimal equation in terms of actual factors for maximizing the liquidus temp.

To assess the accuracy of prediction and established the suitability of response surface methodology using the quadratic model, a reliability plot of the observed and predicted values of liquidus temperature obtained is presented in Fig. 9.

Fig. 9. Reliability plot of observed versus predicted liquidus temperature

The high coefficient of determination (r2 = 0.9723) as observed in Fig. 10 was used to established the suitability of response surface methodology in maximizing the liquidus temperature. To study the effects of combine input variables on the response variable (liquidus temperature), the 3D surface plot presented in Fig. 10 was developed.

(4)

Fig. 10. Effect of current and voltage on the liquidus temperature

The 3D surface plot as observed in Fig. 10 shows the relationship between the input variables (current, voltage and welding speed) and the response variable (liquidus temperature). It is a 3 dimensional surface plot which was employed to give a clearer concept of the response surface.

Although not as useful as the contour plot for establishing responses values and coordinates, this view may provide a clearer picture of the surface. As the colour of the curved surface gets darker, the arc length, heat input and heat affected zone decreases proportionately while the liquidus temperature increases. The presence of a coloured hole at the middle of the upper surface gave a clue that more points lightly shaded for easier identification fell below the surface.

The numerical optimization produced twenty (20) optimal solutions which are presented in Fig. 11.

Fig. 11. Optimal solutions of numerical optimization model

From the results of Fig. 11, it was observed that a current of 130.00amp, voltage of 20.94v and a welding speed of 0.48m/min will result in a welding process with Liquidus Temperature of 1365.090C

Finally, based on the optimal solution, the contour plots showing each response variable (liquidus temperature) against the optimized value of the input variable is presented in Fig. 12

Fig. 12. Predicting liquidus temperature using contour plot

As presented in Fig. 12, the contour plot can be employed to predict the optimum values of the input variables based on the flagged response variables.

IV. DISCUSSION

In this study, the response surface methodology was used to optimize and predict weld parameter. The solution of the optimization process was to determine the optimum value of each input variable namely: voltage(volt), current(Amp) and speed(m/min) that will maximize the liquidus temperature.

A statistical design of experiment (DoE) using the central composite design method (CCD) was done. The design and optimization was executed with the aid of statistical tool called Design Expert 7.01. An experimental design matrix having six (6) center points (k), six(6) axial points (2n) and eight(8) factorial points (2n)resulting to 20 experimental runs was generated.

From the model design summary, for liquidus temperature, the minimum value was observed to be 11180C, with a maximum value of 1548.000C, mean value of 1305.450 and standard deviation of 119.151. To test how well the quadratic model can explain the underlying variation associated with the experimental data, the lack of fit test was estimated for liquidus temperature. Model with significant lack of fit cannot be employed for prediction.

Results of the computed lack of fit for liquidus temperature is presented in Fig. 11.

The computed standard errors for the selected responses are presented in Fig. 9. It was observed that the model possesses a low standard error ranging from 0.27 for the individual terms, 0.35 for the combine effects and 0.26 for the quadratic terms. Standard errors should be similar within type of coefficient; smaller is better. Variance inflation factor (VIF) of approximately 1.0 was observed in fig. 9 which was good since ideal VIF is 1.0. VIF's above 10 are cause for alarm.

In assessing the strength of the quadratic model towards optimizing a target response, one-way analysis of variance (ANOVA) table was generated for each response variable and result obtained is presented in Fig. 6. To validate the adequacy of the quadratic model based on its ability in maximizing the liquidus temperature, the goodness of fit

(5)

statistics presented in Fig. 7 was employed.

V. CONCLUSION

The quality and integrity of welded joints is highly influenced by the optimal combination of the welding input parameters. This study developed a model using expert systems, such as Response Surface Methodology to optimize and predict weld liquidus temperature from input parameter such as current, voltage and welding speed. The result from one Response Surface Methodology analysis shows that a current of 130.00amp, voltage of 20.94V, speed of 0.48m/min will produce a liquidus temperature of 1365.0oC with a desirability of 0.962. Welding operations carried out, using the established results from this study, has improved the integrity of welded joints. It is therefore recommended that the optimal liquidus temperature be employed so as to achieve the accurate viscosity of the molten metal weld to use minimum heat input and minimize heat affected zone and arc length, so as to achieve a favorable liquidus temperature and to use the optimized input parameters obtained in this study.

REFERENCES

[1] J.I. Achebo and C.E. Etin-osa. (2017). Optimization of Weld Quality Properties Using Analytical Hierarchy Process (AHP). Journal of Nigerian Association of Mathematical Physics. Vol 39, pp 389 – 396.

[2] Tham, G., Yaakub, M.Y., Abas, S.K., Manurung, Y.H. and Jalil, B.A.,

"Predicting the gmaw 3f t-fillet geometry and its welding parameter", Procedia Engineering, Vol. 41, (2012), 1794-1799.

[3] RadhakrishnanV.M.,“Welding Technology& Design”, McGraw Hill, New York, 2006.

[4] Reisgen, U., Schleser, M., Mokrov, O. and Ahmed, E. (2012):

Optimization of laser welding of DP/TRIP steel sheets using statistical approach, Opt. Laser Technol., 44, pp. 255–262, DOI:

10.1016/j.optlastec.2011.06.028.

[5] Pan,L. K., Wang,C. C., Hsiso,Y. C., and Ho,K. C., (2004):

Optimization of Nd-YAG laser welding ontomagnesium alloy via Taguchi analysis. J. of Optics& Laser Technology, Vol. 37, pp. 33-42 (2004).

[6] Bansal, Jitender, Umed Kumar Khod, and Deepak Kumar. "Optimize the Various Effects of Welding Parameters of AL Plate by TIG Welding." gas 4.2,2015.

[7] Rishi Kumar, Ramesh N Mevada, Santosh Rathore, Nitin Agarwal, Vinod Rajput,AjayPalSinh Barad. (2017). “Experimental Investigation and Optimization of TIG WeldingParameters on Aluminum 6061 Alloy Using Firefly Algorithm”.IOP Conference Series: Materials Science and Engineering.

https://iopscience.iop.org/article/10.1088/1757- 899X/225/1/012153/pdf.

[8] Collins Eruogun Etin-osa and Joseph IfeanyiAchebo. (2017). Analysis of Optimum Butt Welded Joint for Mild Steel Components Using FEM (ANSYS). American Journal of Naval Architecture and Marine Engineering DOI: 10.11648/j.aas.20170206.12.

[9] Vasudevan, M. (2009) ‘Soft computing techniques in stainless steel welding’, Materials and Manufacturing Processes, Vol. 24, No. 2, pp.209–218.

[10] www.halcyon.com/pub/journals/21ps03-vidmar.

References

Related documents

22:00 - 22:50 : Ataturks Principles And Turkish Revolution History-II     ANS Eğitim 5 Blok - 125      Lecturer MERAL ŞAHİN Turkish Foreign Policy during Atatürk Era.

Abstract: Sixteen reinforced concrete laboratory specimens were used to evaluate a number of corrosion prevention methods under an accelerated (6 months) testing program.. The use

The respondents were asked to indicate if they held discussions with their subordinates after they had filled the appraisal forms on the training needs

Therefore, it requires good parenting models, especially for those who have digitally native teens so that the use of gadgets does not make the teens out of control.

Dynamic changes in gene expres- sion following chemotherapy have been associated with survival in ovarian cancer pre-clinical models and in breast cancer patients treated in

specifically into the implementation or successes of the Intel English language Program in Saudi Arabia. 3) This study provides a better understanding of the effectiveness of using

Discharged Runoff Volume 77 Figure 32: Concentration-Based “First Flush” for Dissolved Aluminum ...81.. Figure 33: Concentration-Based “First Flush” for Dissolved

High chromium white iron surface layers (HWSLs) have been produced on a ductile cast iron EN-GJS-700-2 substrate by diode laser surface alloying with a direct injection of a