Optimization of Hardness in Barreled Cylindrical Billets – A Taguchi Approach

Optimization of Hardness in Barreled Cylindrical Billets – A Taguchi Approach

Optimization Of Hardness In Barreled Cylindrical Billets – A Taguchi Approach

N.Prudhvi1, Rama Krishana.CH 2, D.Satyanarayana3 1C.R.Reddy Engineering College, Eluru.

2Assistant Professor in ME Department, C.R. Reddy Engineering College , Eluru, A.P

3 Senior Assistant Professor, in ME Department, C.R. Reddy Engineering College , Eluru, A.P.

The present work focuses on the application of Taguchi Approach. In the cold forging process at the die/billet interface plays a significant role. were as deciding the billet material, lubrication ,and number of stages to make products .in this work, the influence of varied friction conditions on the hardness is studied. The friction factors of these lubricants molybdenum disulphide(m=0.29), Teflon grease(m=0.37)and silicon grease(m=0.44) were obtained from ring compression test(RCT). The whole configuration assumed to be axi- symmetric and the AA2014 solid cylindrical billets compressed between the dies. At each and every incremental level of deformation process, the following results were noted down. From the simulation study mainly hardness depends upon three major factors as friction effect, aspect ratios, and temperature effect. By using these parameters were optimized to get much uniform distribution of hardness in the billet, to reduce the deformation load and barrelling effect.

Keywords Ring compression test(RCT);Ageing; Machining; forging.

1. Introduction

   1. Factors Effecting hardness:

      The parameters which influence the hardness are ductility of the material, friction, rate of deformation, heat treatment process such as annealing, ageing etc and height to the diameter ratios of the billets. Out of these parameters, the aging time will improve the ductility of the material. Less friction at the billet interface allows the free flow of material in the radial direction in simple upsetting process and high friction

      makes the die to stick to the billet and constraining the material to flow in the radial direction. This inability of flow of material reduces the plastic strain and thereby decreasing the hardness. More height to the diameter ratios makes the material to fail at lower strain and bulge more at the equator. This makes loosening of the material near the equator of the billet and hardness inside the billet is not uniform.

   2. Investigation plan : Taguchi approach will be followed to optimize the hardness which requires a minimum of 3 parameters.

      1. Friction at the die/billet interface: Friction coefficient varies with the lubricant applied which was evaluated from the Ring Compression test.

      2. Height to the diameter ratios of the billet : AA2014 cylindrical billets of diameter 24mm were machined to different height to the diameter ratios of h/d=1, h/d=0.75 and h/d=0.5.

      3. Heat Treatment process: The billets were heat treated in a furnace at a temperature of 502 0c and then water quenched. The solutionized samples were aged at 1600c for different times of 2hrs, 4hrs and 6hrs to improve the ductility of the material.

      After performing the above steps the samples need to be deformed by applying the necessary conditions for the optimization approach.

      The objective of the present work is to optimize the process parameters involved in billet expansion process by using Taguchi and ANOVA methods. The settings of process parameters were determined by using Taguchis experimental design method. Orthogonal arrays of Taguchi, the signal-to-noise (S/N) ratio, the analysis of variance (ANOVA), and regression analyses are employed to find the optimal levels and to analyze

      the effect of the process parameters on experimental values. The parameters that affect the process were determined using Taguchi method, and the most significant process parameters and their percentage contribution were determined by using ANOVA technique. Confirmation test with the optimal levels of

      process parameters was carried out in order to illustrate the effectiveness of Taguchis optimization method.

2. Dissimilar Materials

   AA2014 is the material used for the current investigation. The major constituents of this alloying material are Al-0.8Si-4.4Cu-0.8Mn-0.4Mg. This is one of the most widely used wrought aluminum-copper alloys. This is a heat-treatable alloy and has the behavior of corrosion resistance. AA2014 has the ultimate tensile strength of 483 MPa, yield strength of

   414 MPa, Rockwell hardness (RHN-B) of 82 and percentage elongation of 13%. This alloy is used for heavy-duty forgings, air craft fittings and truck frames..

3. Ring Compression test

   A series of ring compression tests were conducted by UTM (universal testing machine) on Aluminium 6 ring specimen For each condition of lubrication, 6 ring samples were compressed to different heights within different loads. In the present study, the friction calibration curves were constructed and the deformation behavior of the geometry of the ring specimen was examined. This method has the advantage of determining friction coefficient based on the dimensional changes of the geometry of ring specimen. They obtained the calibration curves for estimating friction by plotting the decrease or increase of hole diameter versus percentage change in height reduction of the ring specimen. Problems such as barreling of ring specimen, mushrooming of the hole were analyzed by employing the lubricants such teflon grease, silicon grease , molibdunm disulphate at the die billet interface.

   After the experimental process is done friction was defined at each different lubricants, and the material changes will taken place at inner dia, height.

   Fig 1: RCT Equipment

   Table. 1 Processing Conditions for the Ring Compression test

   Ring specimen size (OD: ID:H=6:3:2)	24:12:8
   Total No of ring samples required for each and all lubrication conditions	(6 + 6 + 6 = 18- total)
   Compression testing machine	Capacity of 1000 ton
   Two Dies	Rigid (H-13 steel)
   Measurement of geometry parameters	Digital Vernier calipers
   Temperature	Cold working (under room temperature)

   This indicates good lubrication condition. After noting down all the results at each and every incremental % reduction in height, the above parameters will be noted down and the friction calibration curves will be plotted. The mean value of the change in the hole diameter to the height reduction of all the 6 specimens gives the value of the friction coefficient m.

   Fig of MOS2

   Fig Teflon grease

   Fig of Silicon grease

   Fig 2: Deformed ring samples for different lubricants

   After upsetting six ring specimens to different height reductions as illustrated in Fig. 2, the change in the hole diameter was noted down. The mean value of the change in the hole diameter to the height reduction of all the 6 specimens gives the value of the friction coefficient m. For the condition of no friction between the die/billet interface the friction coefficient m=0 and m=1 for sticking Low friction causes expansion of the hole and high friction causes bulging of the internal hole of the ring (Fig. 2). For molybdenum

   disulphide, teflon grease and white grease,the values of m are 0.24, 0.33 and 0.44 respectively

4. Eperimental Procedure

   1. Upsetting Procedure

      Solid cylindrical test specimens of different aspect ratios were compressed between two rigid flat dies to different strain levels. As the material undergoes plastic deformation a bulge near the equatorial region can be observed. When this plastic deformation becomes severe, the material damages and a crack will appear on the surface of the cylindrical specimen. The non- uniform deformation of the cylinder is because of the high friction between the contact surfaces of the billet and the dies. By applying proper lubrication, the cylindrical billets can be deforming without failure and utilizing the material and processing conditions effectively

      Table:2 Processing conditions for conducting experiments

      —————————————————————————————–

      Material used AA2014

      Aspect ratio h/d=1, h/d=0.75, h/d=0.5

      Condition of Lubrication Grease/Molybdenum disulphide,

      Grease/

      on the top/bottom faces

      Die Rigid

      Billet Plastic

      Mode Axi-symmetric configuration

      A series of Aluminium solid cylinders will be compressed between the rigid cylindrical dies for the lubricant conditions using a 3000 ton capacity UTM. At each and every stage of the deformation process the following parameters will be noted using digital vernier calipers. For each and every lubricant condition, to find the ductility property for 3 different categories,in 9 different pieces, by applied on the faces of the billet and were reduced to 30% in height. Thereby the geometry parameters will be compared for all the conditions of the lubrication as mentioned.

      Table3. Factors and levels for the Upsetting process

      Factors	Factor description	Level 1	Level 2	Level 3
      Aspect ratio (h/d)	A	0.5	0.75	1
      Aging time in hrs (t)	B	2	4	6
      Friction coefficient (m)	C	0.24	0.33	0.4

      Factors	Factor description	Level 1	Level 2	Level 3
      Aspect ratio (h/d)	A	0.5	0.75	1
      Aging time in hrs (t)	B	2	4	6
      Friction coefficient (m)	C	0.24	0.33	0.4

      After conducting all the 9 experiments, the results were noted down in Table. 2. The response R1 indicates the difference in the Rockwell hardness- B at the center of the billet and at the equator is indicated by

      RHN-B.

      C. Barrel radius

      h height of the cylinder before upsetting d diameter of the cylinder before upsetting Hf Height after deformation

      The barrel radius after deformation was also calculated for all the experiments using the equation proposed by Narayanasamy et al. [17]. Where hf is the height after deformation of the billet, Db is the bulged diameter, Dc is the contact diameter and Ra is the barrel radius

      Db Central bulging after deformation

      R Bulged radius at the center of the billet

      Ra

      2

      h

      h

      f

      4(Db Dc)

      b

      Dc Contact diameter at the die/billet interface Ra barrel radius after deformation

      Fig. 3. Geometry profile of the billet after upsetting.

   2. Hardness

   Table 4: Response charecterstic measured table

   Experiment No	Factor A	Factor B	Factor C	(RHN- B)c	(RHN- B)e	Response RHN- B (R1)	Barrel radius (R2)	Load (R3)
   1	0.5	2	0.24	91.5	86.8	4.7	29.86	464
   2	0.5	4	0.33	92.5	86	6.5	22.25	441
   3	0.5	6	0.4	93	86	7	18.17	422
   4	0.75	2	0.33	93.5	86.5	7	31.509	458
   5	0.75	4	0.4	94	86.5	7.5	26.92	427
   6	0.75	6	0.24	92.5	87	5.5	43.89	390
   7	1	2	0.4	96	86.7	9.3	33.83	445
   8	1	4	0.24	94	87	7	52.07	404
   9	1	6	0.33	94.5	87	7.5	43.08	382

5. Results and Discussions:

   1. Taguchi Analysis

      (i) Minimum deviation in hardness distribution, S/N ratio smaller the better:

      Table 5: Response table for signal to noise ratio (S/N) and Means

      Response(R1)for signal to noise ratio				Response (R1) for means
      Level	A	B	C	Level	A	B	C
      1	-15.53	-16.4	-15.05	1	6.06	7.00	5.73
      2	-16.4	-16.57	-16.89	2	6.66	7.00	7.00
      3	-17.92	-16.89	-17.92	3	7.93	6.66	7.93
      	2.39	0.48	2.87		1.86	0.33	2.2
      Rank	2	3	1	Rank	2	3	1

      To obtain the results of minimum deviation in hardness distribution, the response (R1) smaller the better has been chosen for which the S/N ratio is as follows.

      Table 6:Response table for signal to noise ratio (S/N) and Means

      Response for signal to noise ratio				Response for Means
      Level	A	B	C	Level	A	B	C
      1	27.21	30.02	32.34	1	23.43	31.73	42.61
      2	30.47	30.07	29.87	2	34.11	34.41	32.28
      3	32.64	30.24	28.12	3	43.66	35.05	26.31
      	5.43	0.22	4.21		20.23	3.31	16.3
      Rank	1	3	2	Rank	1	3	2

      (iii) Minimum load absorption, S/N ratio smaller the better:

      Response for singal to noise ratio (S/N)				Response for Means
      Level	A	B	C	Level	A	B	C
      1	-52.91	-53.17	-52.43	1	442.3	455.7	419.3
      2	-52.55	-52.54	-52.58	2	425	424	427
      3	-52.25	-51.99	-52.69	3	410.3	398	431.3
      	0.69	1.18	0.27		32	57.7	12
      Rank	2	1	3	Rank	2	1	3

      Response for singal to noise ratio (S/N)				Response for Means
      Level	A	B	C	Level	A	B	C
      1	-52.91	-53.17	-52.43	1	442.3	455.7	419.3
      2	-52.55	-52.54	-52.58	2	425	424	427
      3	-52.25	-51.99	-52.69	3	410.3	398	431.3
      	0.69	1.18	0.27		32	57.7	12
      Rank	2	1	3	Rank	2	1	3

      Table 7: Response (R3) Table for Signal to Noise ratio and Means

      1

      S/N ratio =-10log n

      n

      2

      2

      Yi

      i 1

      (2)

      Where Yi is the measured value of the response characteristic and n is the no. repetitions for the experimental condition. Signal to noise ratio (S/N) values for different levels and different factors are calculated from the above formula and given in Table 5.

      Reduces the load absorption capacity of the billet. Shorter billets absorb more load than the longer billets, the response (R3) smaller the better has been chosen for which the S/N ratio is as follows.

      (ii) Maximum barrel radius, S/N ratiolarger the better:

      To obtain the results of maximum barrel radius, the response larger the better has been chosen for which

      1

      S/N ratio =-10log n

      n

      2

      2

      Yi

      i 1

      (2)

      the S/N ratio is as follows.

      1 n 1

      S/N ratio =-10log n Y 2

   2. Analysis of variance (ANOVA)

      ANOVA is a decision making tool which aids in evaluating the most significant factor among all the

      i1 i

      (3)

      process parameters. The output quality characteristic is judged based on the variance, F-ratio and percentage contribution ratio and the results are tabulated in Table

      4. Using 95% confidence level, the results obtained from ANOVA are tabulated in Table 4. The F-ratio and

      percentage contributions reveal that friction has contributed 54.31% while the aging time contributed 1.64% and the size of the billet has contributed to 40.45%.

      Table 8 : Analysis of variance for Means(R1)

      Source	Degree of Freedom	Sum of squares	variance	F- ratio	Contribution ratio (%) (R1)
      A	2	5.4489	2.7244	11.30	40.45
      B	2	0.222	0.111	0.46	1.64
      C	2	7.3156	3.6578	15.17	54.31
      % Error	2	0.4822	0.2411		3.58
      Total	8	13.468			100

      Table 9: Analysis of variance for Means (R2)

      Table 10: Analysis of variance for Means (R3)

      So urc e	Degree of Freedom	Sum of squares	varia nce	F- ratio	Contribution ratio (%) (R3)
      A	2	1539.6	769.7 8	71.4 2	22.68
      B	2	5004.22	2502. 11	232. 15	73.73
      C	2	221.56	110.7 8	10.2 8	3.26
      % Err or	2	21.56	10.78		0.31
      Tot al	8	6786.89			100

      A similar procedure has been adopted for investigating the effect of process parameters on deformation load as in the case of studying hardness variation and barrel radius. Using 95% confidence level, the results obtained from ANOVA are tabulated in Table 8. The F- ratio and percentage contributions explain that aging time has contributed 73.73% while the size of the billet has contributed 22.68% and the friction at the die/billet interface contributes to 3.26%.

   3. Prediction of Optimum value for R1, R2 and R3

      From the results of S/N ratio and the mean values, the optimum control factors and levels have been chosen as A1, B3 and C1 for R1; A3, B3 and C1 for R2 and R3. The mean optimum values (MR1, MR2 and MR3) for the responses can be determined as below.

      Source	Degree of Freedom	Sum of squares	variance	F-ratio	Contribution ratio (%) (R2)
      A	2	614.72	307.358	188.07	58.84
      B	2	18.57	9.283	5.68	1.77
      C	2	408.01	204.007	124.83	39.06
      %	2	3.27	1.634		0.31
      Total	8	1044.56			100

      Source	Degree of Freedom	Sum of squares	variance	F-ratio	Contribution ratio (%) (R2)
      A	2	614.72	307.358	18.07	58.84
      B	2	18.57	9.283	5.68	1.77
      C	2	408.01	204.007	124.83	39.06
      %	2	3.27	1.634		0.31
      Total	8	1044.56			100

      MR1 = Y (A1 Y ) (B3 Y ) (C1 Y )

      MR2 = Y (A3 Y ) (B3 Y ) (C1 Y )

      MR3 = Y (A3 Y ) (B3 Y ) (C1 Y )

      (4)

      (5)

      (6)

      Using 95% confidence level, the results obtained from ANOVA are tabulated in Table 6. The F-ratio and percentage contributions confirm that size of the billet has contributed 58.84% while the aging time contributed 1.77% and the friction at the die/billet interface has contributed to 39.06%.

      Where; Y is the average expansion of all the 9 experiments in the Taguchi orthogonal array, A1, B3 and C1; A3, B3 and C1 are the response values of the mean with process parameters at optimum levels.

   4. Confirmation run

   The confirmation tests were conducted to evaluate the results predicted from Taguchi technique. The predicted results and the experimental results are tabulated in Table 9. It can be examined that the results obtained are within the span of predicted 95%

   confidence level and the predicted values are in close agreement with the experimental results.

   Table 11: Test data confirmation run

   Responses	Optimum	Predicted	Experimental	%Error
   R1	A1, B3,	4.67	4.46	4.49
   R2	A3, B3,	54.07	56.12	3.79
   R3	A3, B3,	375.84	385	2.44

6. Graphs

   Graph 2: Main effects plot for Means to investigate hardness distribution.

   Graph 1: Main effects plot for S/N ratio to investigate hardness distribution.

   Graph 3: Main effects plot for S/N ratio to investigate barreling.

   Graph 4: Main effects plot for Means to investigate barreling.

   Graph 5: Main effects plot for S/N ratio to investigate deformation load

   Graph 6: Main effects plot for Means to investigate deformation load.

7. Conclusions

   The following conclusions can be drawn from the present work:-

   1. The significant factor responsible for the indifference hardness distributions throughout the billet is friction which contributes 54.31% among the total value.

   2. The major parameter affecting the barreling behavior is size of the billets which contributes 58.84% and the ageing condition contributes 73.73% on the deformation load.

   3. The optimum factors to control the deviation in hardness distribution are A1, B3 and C1; A3, B3 and C1 for barreling behavior and minimum deformation load.

   4. The optimum value obtained from the Taguchi method is 4.67 for hardness distribution, 54.07 for barreling behavior and 375.84 for deformation load.

   5. The values obtained from the optimum condition of Taguchi method are in good agreement with the confirmation experiments of hardness distribution, barreling behavior and deformation load.

8. References

1. Fazil, O. Sonmez.; Ahmet D.Analytical relations between hardness and strain for cold formed parts, Journal of Materials Processing Technology 2007, 186, 163173.

2. Haluk T.; Fazl O. Sonmez. Optimum shape design of die and preform for improved hardness distribution in cold forged parts, Journal of materials processing technology 2009, 209, 1538 1549.

3. Muller, M.; Barrans, S.M.; Blunt, L. Predicting plastic deformation and work hardening during V- band formation, Journal of Materials Processing Technology 2011, 211, 627636.

4. Muller, M.; Barrans, S.M.. Ultimate axial load capacity of V-band clamp joints.In: University of Huddersfield (Ed.), Proceedings of Computing and Engineering Annual Researchers Conference. Huddersfield, UK, 2009 pp. 1418.

5. Muller, M.; Barrans, S.M.;. Impact of flange geometry on the ultimate axial load capacity of V- band clamps. In: 9th International Conference on Turbochargers and Turbocharging, 2010 London, UK.

6. Paralikas, J.; Salonitis, K.; Chryssolouris, G. Investigation of the effects of main roll-forming process parameters on quality for a V-section profile from AHSS, International Journal of Advanced Manufacturing Technology 2009, 44,

   223237.

7. Paralikas, J.; Salonitis, K.; Chryssolouris, G. Optimization of roll forming process parameters a semi-empirical approach, International Journal of Advanced Manufacturing Technology 2010, 47,

   10411052.

8. Lindgren, M. Cold roll forming of a U-channel made of high strength steel, Journal of Materials Processing Technology 2007a, 186, 7781.

9. Lindgren, M. An improved model for the longitudinal peak strain in the flange of a roll formed U-channel developed by FE-analysis, Steel Research International 2007b, 78 (1), 8287.

10. Babakhani, A.; Kiani-Rashid, A.R.; Ziaei, S.

    M. R. The Microstructure and Mechanical Properties of Hot Forged Vanadium Microalloyed Steel, Materials and Manufacturing process 2012,

    27 (2), 135-139.Jiang, Li.; Fuguo, Li.; Fengmei, Xue.; Jun, Cai.; Bo, Chen.Micromechanical behavior study of forged 7050 aluminum alloy by

    microindentation, Materials & Design 2012, 37,

    491-499.

11. Mohan-Raj, A.P.; Selvakumar, N. Deformation behavior of Sintered Fe-C-Mn Composite During Cold Upset forming, Materials and Manufacturing Processes 2011, 26 (6) 1388-

    1392.

12. Sumathi, M.; Selvakumar, N. Investigation of Cu-SiC Composite Preforms during Cold Upsetting, Materials and Manufacturing Processes 2011 26 (6) 826-831.

13. Jiang, Li.; Fuguo, Li.; Fengmei, Xue.; Jun, Cai.; Bo, Chen. Micromechanical behavior study of forged 7050 aluminum alloy by microindentation, Materials & Design 2012 37

    491-499.

14. Zhehe, Yao.; Gap-Yong, Kim. et al. Acoustic softening and Hardening of Aluminum in high frequency vibaration Assites Micro/Mesoforming, Materials and Manufacturing Process 2013, 28 (5),

    584-588.

15. Narayanasamy, R.; Pandey, K.S.; Salient features in the cold upset forming of sintered aluminium-3.5 percent alumina powder composite performs. Journal of materials processing technology 1997, 72, 201-207.

Biography

N.Prudhvi was born in Machilipatnam in India, on feb 14, 1989. He was graduated from

D.M.S S.V.H College of Engineering, Guntur in 2010 and student of M.Tech MACHINE DESIGN at SIR C.R Reddy Engg College, eluru India. His areas of interest are Design, upset forging,

related topics.

Mr.Ch.Ramakrishna is graduated from VRSCE. He completed his M.Tech from JNTU KAKINADA. He is presently working as Assistant professor at SIR C.R Reddy Engg College, Eluru, His areas of interest are Design, Manufacturing and related topics.

.