2212-8271 © 2013 The Authors. Published by Elsevier B.V.Selection and peer-review under responsibility of Professor Pedro Filipe do Carmo Cunhadoi: 10.1016/j.procir.2013.06.047

Procedia CIRP 7 ( 2013 ) 646 – 651

Forty Sixth CIRP Conference on Manufacturing Systems 2013

A study on the heating process for forging of an automotive crankshaft in terms of energy efficiency

Hong-Seok Park*, Xuan-Phuong Dang Lab for Production Engineering, School of Mechanical and Automotive engineering, University of Ulsan, San 29, Mugeo 2-Dong, Namgu, Ulsan,

P.O. Box 18, Ulsan 680-749, South Korea * Corresponding author. Tel.: +82-052-259-2294; fax: +82-052-259-1680. E-mail address: phosk@ulsan.ac.kr.

Abstract

This work studies the in-line induction heating process before hot forging of an automotive crankshaft to find the potential solutions for improving the energy efficiency. We optimized the process parameters and proposed an insulating system to reduce the radiation and convection losses at the open spaces between adjacent heaters. The results obtained from the analytical model of the heat transfer show that using insulating covers can roughly reduce 9% of heat losses compared to the energy stored in the workpiece. Approximately additional 6% of energy can be saved by process parameters optimization. © 2013 The Authors. Published by Elsevier B.V. Selection and/or peer-review under responsibility of Professor Pedro Filipe do Carmo Cunha Keywords: Optimization; Manufacturing; Forging; Heat treatment

1. Introduction

The manufacturing community is trying to reduce the energy consumption due to the energy cost problem, ecological and environmental policies. Increasing the energy efficiency of the manufacturing process is best way to resolve these issues and strengthen the competition of manufactures. The methods for energy efficiency at the production or factory level in many manufacturing aspects are being intensive studied. The state of the art of energy efficiency in laser based manufacturing processes, machining, machine tools and manufacturing system can be found in the literature [1].

For hot forging industry, induction heating process has been considered as a high productivity, repeatable quality, and green heating technology compared to fuel-fired furnaces. This is the reason that induction heating, a best available heating technology, is preferred in forging manufacturing [2-3]. Induction heating before hot forging requires a huge amount of electrical energy for heating a steel workpiece with a large volume from the ambient temperature to approximate 1150 1250 C. Therefore, the increase in the thermal efficiency of the heating system significantly saves the consumed energy.

Similar to other manufacturing technologies, energy-saving solutions for induction heating are important issues that the manufacturers and researchers always pay their attention. Solutions for saving energy for industrial induction heating may include the energy management, innovative components of induction devices, energy recovery, and adaptive control [4]. One effective way to resolve the energy-savings problem in induction heating is carrying out the optimization process through parameters study. Diverse published works devoted to optimization of induction heating [5-11], but most of them focus on how to minimize the temperature deviation at the end of the heating process.

This work studies the in-line induction heating process before hot forging of an automotive crankshaft in order to find the potential solutions for improving the energy efficiency. The inappropriate characteristic of the of the current in-line induction heating systems is the heat losses caused by radiation and convection between heaters. Therefore, we proposed an insulating system to reduce theses thermal losses. In addition, when the thermal efficiency is increased as well as the throughput is changed, the reconfiguration of the heating process parameters is necessary. A new heating strategy was also proposed for higher heating flexibility and better thermal

Available online at www.sciencedirect.com

© 2013 The Authors. Published by Elsevier B.V.Selection and peer-review under responsibility of Professor Pedro Filipe do Carmo Cunha

647 Hong-Seok Park and Xuan-Phuong Dang / Procedia CIRP 7 ( 2013 ) 646 – 651

efficiency. This works also addresses the issues that are

and power distribution along the heating line, a controversial problem which is rarely discussed in the literature [12].

2. Analyze the heat losses of the in-line induction heating system for crankshaft forging

The manufacturing line of an 4-cylinder automotive crankshaft is depicted in Fig. 1. The induction heating step is the one that consumes a great amount of energy compared to other stages in the forging process. It is also the first important step in the manufacturing process.

Fig. 1. (a) An automotive crankshaft; (b) the manufacturing process

The in-line induction system which consists of seven heaters for heating a long steel bar with a diameter of 97-mm is shown in Fig. 2(a). The steel bar moves continuously through the in-line heaters with a designated velocity that is equivalent to the cycle time before going to the forging step.

Fig. 2. (a) The in-line induction heating with seven heaters; (b) the heat losses at the open spaces between adjacent heaters

The total rate of heat loss from the surface of the hot workpiece is

1 1 1conv radq q q (1)

where q1 con and q1 rad are the heat losses caused by convection and radiation, respectively. q1 is the heat flux per unit length tha transfers from the outmost of the insulating cover to the ambient air (natural convection in the infinite space).

1 1 1 1( ) ( )conv wq d t t Nu t t (2)

where , , dw, t1, t are the heat-transfer coefficient, thermal conduction, diameter of the workpiece, surface temperature of the workpiece, and ambient temperature, respectively.

1Nud

(3)

and Nu is calculated by the formulas [13]

1/4

9/16 4/9

0.518( Pr)0.36[1 (0.559 / Pr) ]

GrNu (4)

The Grashof number is calculated as: 3

12

( )g t t dGr (5)

where 1

ftand 1

2ft tt is the mean film

temperature

The thermal properties of air are assumed constant and are taken to be the values at the mean film temperature

The heat loss caused by radiation is calculated as

4 41 1 1( )radq d T T (6)

In the in-line induction heating, because the induction heaters are connected one by one and there are open spaces between adjacent heaters due to the existence of the rollers (see Fig. 2(a)), the losses caused by radiation and convection are remarkable. We carried out the investigation into the real induction system and found that the heat losses caused by convection and radiation at the open spaces where the heated billet exposes to the ambient air account for a significant amount of energy. Figure 2(b) shows the thermal losses of a induction heating line with the rating power of 4250 kW and 850 kW of effective power (the energy needed to heat the workpiece per cycle time). The heat losses account for approximate 6.9% of the energy stored in the workpiece. It is clear that the reduction of this kind of losses can

1 2 31

4 51

6 7812 0 C 8890 C 9500 C 10300 C 11600 C 12000 C 12200 C

Heater 2 Heater 3 Heater 4 Heater 5 Heater 6 Heater 7

Total thermal losses at open spaces: 60.8 kW(Accounts for 6.9% of engergy stored in the workpiece)

0.5

4.35.8

8.0

11.0

14.1

17.1

0.02.04.06.08.0

10.012.014.016.018.020.0

Position1

Position2

Position3

Position4

Position5

Position6

Position7

Ther

mal

loss

es (k

W)

(a)

(b)

(b)

Raw material Induction heating Cutting

Forging Handling Cooling

(a)

648 Hong-Seok Park and Xuan-Phuong Dang / Procedia CIRP 7 ( 2013 ) 646 – 651

save the energy effectively. Therefore, the insulatingsolution was proposed to reduce the heat losses. In this work, the energy efficiency is defined as the ratiocounted by % of the heat energy stored in the workpiece and the input electric energy to the heaters whensimulating. The electrical efficiency of the power supplyis temporarily ignored due to the scope of the study.

3. Development of the insulating covers to reduceheat losses for the in-line induction heating system

To reduce the heat losses at the open spaces, wherethe extreme hot billet (from approximate 800 ~ 1220 C)directly exposes to the ambient air, we proposed theinsulating covers as shown in Fig. 3. Due to the high temperature of the billet, the special materials (ceramicbrick and super wool) that can withstand thistemperature were selected. These materials have a good manufacturability that can be machined or formed intothe required shape.

The insulating device is consider as the multi-layercylindrical wall including an annulus cylindrical air gap,a ceramic tube, and a thin stainless steel cover. In fact,there is an open space at the bottom of the insulating cover (see Fig. 3); however, this area accounts for a small portion. Consequently, this open area is ignored and the insulating device is treated as a perfect cylindrical walls. The schematic axisymmetric layout of the insulating cover is illustrated in Fig. 4.

The governing equation for determining the heattransfer rate and the temperature is the law of energybalance. The heat transfer through the cylindrical layersmust be conservative.

12 24

24 4

q q12

q q24

(7)

where q12, q24, 44 and q4 are the heat flux per unit length (W/m) transfers through the air-gap 11, through theceramic and stainless cover, and from the outmost of theinsulating cover to the ambient air.

4 44 4 4 1 4 )4q Nu t t d t t44 4 4 1 4( ) () ( 4

4 1 44 14 14 (8)

4 41 1 2

12 1 22 1

1 1 2 2

2 ( )2 1 11ln 1

ed t t4

1 1 21q12 1 2( 1 21e

d2 d11d1 d2

(9)

2 424

4

2 2 3 3

1 ln ln3

2 2

t t2 4q d d41333

d d2 3 32 3222 3

(10)

All of the components in the system of Eq. (7) can be calculated by the analytical heat transfer equations.Assuming that the thicknesses of the insulating layersare predetermined, the system of two equations (7) can be used to determine the temperature t2t and t4t . Whenever t2t or t4 is known, the rate of heat transfer (rate of heat losses) is obtained. Then, we can compare the heat losses for the case of with and without insulating cover.

For each value of the thickness 11 and 22, the temperature t2t and t4t has a certain value. As the result,the value of heat transfer rate also changes if 11 and 22vary. Therefore, we can find the optimum value of 11and 22 that minimize the heat losses through the cover.

Fig. 3. The proposed thermal insulating covers in the in-line induction heating system

Fig. 4. Schematic axisymmetric layout of the insulating cover

SSSSSSSStS eeeeeeeeeeeelllllllbbbbbbbbbiib lllllleeeeeeeettttttttttttetttt

e

22

3333333333333333333333333333

aira

nt airAmbien

air

cera

mic

Stainless steel cover

t1

t2t

t3 t4t

t

d1

dd2dd3d

dd4d

rr

q

11112222

SSSSySyySySySySySySySyyyyyyySym

mm

mm

mm

mm

mm

mm

mm

mm

mm

mm

mm

mm

mm

ttetetetetetetetetetetetiiiririririririririririrc

acacacacacacacacacacaalllllllllllllll ll

ilililililililililililililinenenenenenenenenenen

812 0 C 8890 C 9500 C 10300 C 11600C 12000 C 12200 C

Heater 2 Heater 3 Heater 4 Heater 5 H Heater 7Heater 6

Heated billet

Insulating ceramic tube

Induction heater

Metallic cover

Support plates

Supporting roller

H

649 Hong-Seok Park and Xuan-Phuong Dang / Procedia CIRP 7 ( 2013 ) 646 – 651

The optimization problem is minimize the heat

transferred through the insulating cover to reduce the heat losses within some constraints of the design space of the input parameters. The problem is stated as:

- Minimize the rate of heat transfer:

4 44 4 4 1 4( ) ( )q Nu t t d t t (11)

- Subject to: Side constraints)

1 0.07 and 2 0.08 Inequality constraint

t2 Equality constraints

q12 = q24 and q24 = q4 In this optimization problem, it is noted that the

objective function also be treated as a constraint due to the system of Eq. (7). The input parameters (design variables) are 1 , 2, and t2. These parameters change within their constraints during the optimization process. The results of the design and optimization of the insulating covers are presented in the results and discussion section.

4. Optimization of the process parameters of the in-line induction heating

For induction heating, the temperature profile of the heated workpiece and energy consumption are complicated functions of current density, frequency, material properties, coil design, the coupling between coils and workpiece, and the characteristic of the power supply. In this work, we did not focus on improving the hardware or equipment such as power supplies, inverters, or induction coils that were made by the induction heater manufacturer. Instead, the optimum voltages and frequencies, which are the changeable process parameters of the heating system, must be figured out by a scientific approach and elaborated studied rather than practice or experience.

Induction heating is a complex electromagnetic and heat transfer process because of the temperature dependency of electromagnetic, electrical, and thermal properties of material as well as skin effect. Therefore, the exact analysis method is very difficult to implement. Thus, general-purpose FEM was employed to analyze and simulate the induction heating process. The theoretical background of the mathematical and numerical modeling of electromagnetic field governed

presented in the literature [5, 14-16] so that it is unnecessary to duplicate in this paper.

The FEM-based model of the induction heating process was developed using ANSYS Parametric Design Language. To reduce the computational time, only one turn of the coil is considered. This simplification is

sufficient when modeling an induction heating system with a classical solenoidal inductor. Because the input of

meanwhile the induction heaters are fed by voltage supplies, the circuit-coupled FEM model was employed as shown in Fig. 5 to calculate the current density.

Flexible manufacturing requires the induction heating systems to have an ability of adapting to the change of the throughput while keeping the reasonable energy efficiency. In addition, the energy efficiency depends on the heating pattern along the heating line. Therefore, the heaters are divided into three groups that are fed by independent power supplies as shown in Fig. 6. The

temperature; the group 2 is responsible for heating above

remaining portion of energy to heat the billet to the target temperature. All the groups of heaters are fed the same voltage but different frequencies. The frequencies f1 and f2 of group 1 and 2 are parametrical studied, and the frequency f3 is estimated by our developed algorithm because the target temperature must be reached at the end of the heating line.

Fig. 6. Heating strategy that divides the heaters into three groups fed by independent power supplies

Fig. 5. Circuit-coupled FEM model of induction heating simulation

billet

coilair water

Refractory (insulation)

air

Coi

l pi

tch

Coil (finite element domain)

AC

R

L

C

Finite element domain

0

200

400

600

800

1000

1200

1400

0 100 200 300 400 500Time (s)

Tem

pera

ture

(°C

)

Center (3) Surface (1) Average (2)

1

Heater 1

Heater 2

Heater 3

Heater 4

Heater 5

Heater 6

Heater 7

2

3

Group 1 (U, f1) Group 2 (U, f2) Group 3 (U, f3)

Target temperature

Initial temperature

Length of the induction heating line

650 Hong-Seok Park and Xuan-Phuong Dang / Procedia CIRP 7 ( 2013 ) 646 – 651

Full factorial design matrix was applied for parameter

studies. Three process parameters were chosen including voltage U, frequency f1, frequency f2. Each parameter (factor) is divided into three levels, so there are 33 = 27 experiments. Initial and final temperatures of the workpiece are 25 C and 1220 C, respectively. The emissivity or coefficient of radiation is selected as 0.75, and the ambient air temperature is 25 C. The billet moves with a speed of 460 mm per 25 seconds (cycle time). The length of each heater is 1000 mm, and the distance between adjacent heaters is 300 mm. The data obtained from simulation was used to construct the relations between U, f1, f2 and all necessary outputs by using response surface methodology (see Fig. 7). The considered outputs include current, power, heat losses, temperature distribution in the radial and axial direction, temperature deviation in the cross-section of the workpiece at the end of heating, and the energy efficiency.

For hot forging, both of the energy efficiency and quality related to the uniformity of temperature distribution in the workpiece are important criteria. Based on the approximate relations between process parameters and energy efficiency as well as the temperature deviation, we found that minimizing the temperature deviation will decrease the energy efficiency. In practice, it is unnecessary to reduce the temperature deviation down to a too small value. Therefore, temperature deviation is treated as a technical constraint. In case of hot forging automotive crankshaft, the temperature deviation across the cross-section of the billet is practically chosen in the range of 2 3 C, a strict value for better heating quality. The optimization problem is stated as follows:

Maximize the energy efficiency Subject to 2 temperature deviation 3; 700 f1

1100, 900 f2 1300, and 450 U 550.

Temp

eratu

re de

viatio

n (C

)

(a) (b)

Temp

eratu

re de

viatio

n (

C)Ef

ficien

cy

Effic

iency

Fig. 7. The regression relation between frequencies and the total efficiency (a), frequencies and the temperature deviation (b)

5. Results and discussions

The results of the design and optimization of the insulating covers as well as the optimization of the process parameters are presented in this section.

Table 1. The heat losses for the case without insulating cover

Position t1 ( C) q conv (W/m) q rad (W/m) q sum (W/m)

7 1220 4571 55785 60356

6 1200 4488 52851 57339

5 1160 4322 47330 51653

4 1030 3786 32326 36113

3 950 3458 25069 28527

2 889 3208 20413 23621

1 812 2894 15496 18390

Table 1 shows the heat losses at seven open spaces of the induction heating line for the case without insulating cover. Total heat losses is 275,998 W/m, equivalent to 82.8 kW. If the insulating cover is applied, the total thermal losses at these corresponding position drop to 8,540 W/m (see Table 2), equivalent to 2.6 kW. It is clear that approximate 80 kW can be saved. The optimum thickness of the insulating layers are shown in the two last columns of Table 2.

Table 2. The heat losses and the optimum thickness of the insulating covers for the case of using insulation

Position t1 ( C) t2 ( C) t4 ( C) q (W/m) 1 (m) 2 (m)

7 1220 1210.0 151.2 1748.1 0.045 0.078

6 1200 1191.9 140.2 1365.8 0.020 0.080

5 1160 1151.5 137.2 1317.3 0.020 0.080

4 1030 1020.1 127.1 1159.9 0.020 0.080

3 950 939.3 120.7 1063.1 0.020 0.080

2 889 877.5 115.6 989.4 0.020 0.080

1 812 799.4 109.1 896.5 0.020 0.080 Figure 8 shows the result of multi-objective

optimization (maximize efficiency and minimize the temperature deviation) solved by genetic algorithm. It was found that the optimum values of voltage, frequency f1, and f2 are 472 V, 920 Hz, and 1160 Hz, respectively. The frequency f3 was estimated as 1115 Hz. The maximum energy efficiency is 62.9% (see Fig. 9), greater than the worse case in 27 simulation experiments 6.2% (62.9% compared to 56.7%). Minimum temperature deviation is 2.5 C. It can be seen that

Fig. 8. History of the value of the objective functions using GA

Temp

eratu

re de

viatio

n (C)

Number of run Number of run

651 Hong-Seok Park and Xuan-Phuong Dang / Procedia CIRP 7 ( 2013 ) 646 – 651

optimizing process parameters significantly increases the energy efficiency and the heating accuracy.

Figure 9 shows the graph of the electromagnetic efficiency and the total energy efficiency obtained by simulation and calculation. It can be found that when the temperature increases, the energy efficiency decreases. It is clear that putting more power for the heaters at the front-end of the heating line is more energy benefits.

6. Conclusions

Process parameters optimization, proper heating strategy, and practical thermal insulation are the effective ways for increasing the energy efficiency of the induction heating. The inappropriate characteristic of the of the current in-line induction heating systems is the heat losses caused by radiation and convection between heaters. Therefore, the insulating system is an effective solution to reduce theses thermal losses. The results obtained from the analytical model of the heat transfer show that using insulating covers at the open spaces between adjacent heaters of the in-line induction heating system can approximately reduce 9% of heat losses compared to the energy stored in the workpiece. When the thermal efficiency is increased, the reconfiguration of the heating process parameters is necessary. For better electromagnetic and thermal efficiency, the simulation-based optimization of the operating parameters of the induction heating system, including voltages and frequencies, was done using design of experiment and meta-model. A new heating strategy was also proposed for higher heating flexibility and better thermal efficiency. Through the research results, it was found that approximately additional 6% of energy can be saved by process parameters optimization. The optimization in conjunction with the combination of analytical and practical approaches applied to the heating process of a hot forging of the automotive crankshaft makes a contribution for the sustainability and the development of the automotive industry. However, this work starts with the simulation techniques due to the complication of the manufacturing system, the scheduling condition of the demo site, and the experimental cost. The real

physical experiment and implementation of this work at the demo site will be the further works in order to compare and verify the simulated and practical results.

Acknowledgements

This work was supported by the Ministry of Knowledge Economy, Korea, under the International Collaborative R&D Program hosted by the Korea Institute of Industrial Technology.

References

[1] Apostolos, F., Panagiotis, S., Konstantinos, S., George, C., 2012. Energy Efficiency Assessment of Laser Drilling Process, Physics Procedia, vol. 39, p. 776.

[2] Levacher, L., Hita, I., Bethenod, C., Hartmann, S., 2009. "Energy efficiency in industry: from existing technologies to innovative solutions," in ECEE 2009 Summer Study, La Colle sur Loup, France, p. 1091.

[3] Wrona, E., Nacke, B., 2001. "Rational use of energy in induction heaters for forging industry," in International Scientific Colloquium, Modelling for saving resources, Riga, p. 153.

[4] Paya, B., 2008. "Investigation on energy recovering in an induction coil for continuous through heating billet," in XVI International congress on Electricity applications in modern

[5] Favennec, Y., Labbé, V., Bay, F., 2003. Induction heating

processes optimization a general optimal control approach, Journal of Computational Physics, vol. 187, p. 68.

[6] Bodart, O., Boureau, A.-V., Touzani, R., 2001. Numerical investigation of optimal control of induction heating processes, Applied Mathematical Modelling, vol. 25, p. 697.

[7] Galunin, S., Zlobina, M., Nikanorov, A., Blinov, Y., 2005. "Numerical optimization of induction through heating for forging," in Science and Technology, KORUS 2005. Proceedings. The 9th Russian-Korean International Symposium on, p. 313.

[8] Pleshivtseva, Y., Rapoport, E., Efimov, A., Nacke, B., Nikanorov, A., 2008. "Special method of parametric optimization of induction heating systems," in International Scienticfic Colloqium, Modelling for Electromagnetic Processing, Hannover, Germany, p. 229.

[9] Rapoport, E., Pleshivtseva, Y., 2006. Optimal Control of Induction Heating Processes: CRC Press, p. 273.

[10] Rudnev, V.I., Loveless, D., Schweigert, K., Dickson, P., Rugg, M., 2000, Efficiency and Temperature Considerations in Induction Re-Heating of Bar, Rod and Slab Industrial Heating, vol. June, p. 39.

[11] Zgraja, J., 2005, The optimisation of induction heating system based on multiquadric function approximation, COMPEL: The International Journal for Computation and Mathematics in Electrical and Electronic Engineering, vol. 24, p. 305.

[12] Rudnev, V., Brown, D., Tyne, C.J.V., Clarke, K.D., 2008. "Intricacies for the Successful Induction Heating of Steels in Modern Forge Shops," in Proceedings of 19th International Forging Congress, Chicago, IL.

[13] Holman, J.P., 2002. Heat transfer, 9th edition ed.: McGraw-Book Company, p. 77.

[14] Bay, F., Labbe, V., Favennec, Y., Chenot, J.L., 2003, A numerical model for induction heating processes coupling electromagnetism and thermomechanics, Int. J. Numer. Meth. Engng vol. 58, p. 839.

[15] Sadeghipour, K., Dopkin, J.A., Li, K., 1996, A computer aided finite element/experimental analysis of induction heating process of steel, Computers in Industry, vol. 28, p. 195.

[16] Rudnev, V.I., Loveless, D., Cook, R., Black, M., 2002. Handbook of Induction Heating CRC Press, p. 150.

Fig. 9. Electromagnetic efficiency and the total energy efficiency

89.6

79.8

63.359.9

55.349.9

44.4

62.9

89.984.0

71.6 71.8 72.4 72.6 72.878.3

0.0

10.0

20.0

30.0

40.0

50.0

60.0

70.0

80.0

90.0

100.0

Heater 1

Heater 2

Heater 3

Heater 4

Heater 5

Heater 6

Heater 7

Total

Effi

cien

cy (%

)

Thermal & electrical efficiencyElectromagnetic efficiency

0

200

400

600

800

1000

1200

1400

Ave

rage

bill

et's

tem

pera

ture

(°C

)