Study on the grinding of Inconel 718

Study on the grinding of Inconel 718

Jmlra~ d Materials Processing Technology ELSEVIER Journal of Materials Processing Technology 55 (1995) 421-426 Study on the grinding of Inconel 718...

636KB Sizes 0 Downloads 10 Views

Jmlra~ d

Materials Processing Technology ELSEVIER

Journal of Materials Processing Technology 55 (1995) 421-426

Study on the grinding of Inconel 718 Pei-Lum Tso Department o/'Power Mechanical Engineering, National Tsing Hua University, Hsinchu, Taiwan, ROC Received 12 May 94; in revised form 22 December 94

Industrial summary

The grinding process has been investigated in the machining of Inconel 718 with GC, WA and CBN grinding wheels. The machinability discussed here includes surface roughness, dimensional accuracy, grinding force, grinding-wheel wear and grindingwheel life. The CBN grinding wheel is found to possess particularly good grinding properties and is most suitable for grinding Inconel 718 when compared with WA and GC grinding wheels, but it has high cost. 1. Introduction

Inconel 718 not only has high strength, corrosion resistance, heat resistance, and fatigue resistance, but also possesses lower thermal conductivity [1]. It has a wide variety of applications, e.g. aircraft gas turbines, stack gas reheaters, reciprocating engines, etc. For those special material properties, high cutting force, tool wear, and cutting temperature are the characteristic features in the machining process. The machinability of Inconel 718 has received rather limited attention, so that understanding its behavior in the grinding process is therefore of vital importance. Grinding burn in steel workpiece materials was observed by Peklienik [2] and Yoshkawa [3-1 to occur when the wear fiat area reaches a critical value. Malkin and Cook [4, 5-1 classified grinding wheel wears as attritious wear, grain fracture and bond fracture. Most of the wear consists of grain and bond fracture particles with relatively more bond fracture occurring with softer wheels. Both the vertical and horizontal grinding force components generally increase linearly with the wear fiat area. The grinding force and grinding energy have been investigated in several previous research efforts. Kannappan and Malkin [6] described the effects of grain size and operating parameters on the mechanics of grinding. The specific cutting energy in grinding, i.e. the total grinding energy minus the specific energy due to sliding between the wear flat area and the workpiece, was indicated from their results to be independent of grain size and to decrease with increasing table speed and down-feed. In 1974, Malkin and Anderson [7, 8] calculated the distribution of the total grinding energy between chip formation, plowing, and the sliding-energy component, and the portions of each of these energies that are conducted as 0924-0136/96,'$15.00 ',(" 1996 Elsevier Science S.A. All rights reserved SSDI 0 9 2 4 - 0 1 3 6 ( 9 5 ) 0 2 0 2 6 - I

heat into the workpiece. The selection of operating parameters was investigated by Malkin [9] for surface grinding of steels, when the grinding-wheel tool life is limited by workpiece burn, results indicating that the optimum down-feed for minimum production time is greater with lower workpiece velocities.

2. Experimental

A K E N T KCF-52 NC-type surface grinder was used, equipped with variable wheel speed, and down-feed controis. The grinding wheels' size was 140 x 13 x 50.8 (mm), the wheels being designated WA46K8V, GC60J8V and CBN100P75V. For all experiments, the wheel speed, table speed, and down-feed were in the range Vs = 650-1320 m/min, Vw = 2-5 m/min and d = 5 15 ~tm, respectively. The workpiece material was Inconel 718 of hardness 40 Rockwell C after age hardening [10]. The workpiece specimens were 108 mm along the grinding direction, 105 mm wide and initially 10 mm high. The horizontal and vertical grinding force components, F , and Fv, were measured with a Quartz 3-component dynamometer with its output displayed on a 386 personal computer. Fn is the power force component acting tangentially to the wheel, whilst Fv is the contact force normal to the wheel. The attritious wear and the observation of chip types were obtained via a microscope. The calculation of wear fiat area is described as: (i) using the microscope with Polaroid camera to obtain the wear pictures of the grinding wheel; (ii) utilizing a scanner to process the wear pictures; and (iii) writing a program for calculation of the wear fiat area. The wear image is either black or white, so

422

P.-L. Tso / Journal of Materials Processing Technology 55 (1995) 421-426

that the wear flate rate can be calculated from white A = (white + black) × X '

(1)

where A is wear flat rate; white is the white area of the wear picture; black is the black area of the wear picture; and X is the magnification. The hardness of the CBN grains is very close to that of diamond. Therefore, the wheel velocity was reduced to 600 m/min and the radial infeed was 5 lam, so as to prevent diamond wear. The cross-feed velocity was 320mm/min. Moreover, the dressing process was required for the CBN wheel whilst the wheel rotated at the truing velocity after the turning process. The surface roughness was measured with a Hommel Tester T1000, the mean roughness, (Ra), being used for presenting the machining quality of the workpiece surface, both parallel and perpendicular to the grinding direction.

3. Results

3.1. Surface roughness The CBN grinding wheel is observed to produce a better surface roughness than the GC and WA wheels; in addition, the surface roughness in down-grinding is better than in up-grinding with the same grinding passes. Furthermore, the surface roughness increases with decreasing grinding-wheel speed and increasing table speed and down-feed. On the other hand, the workpiece surface raised another problem of concern here. A photograph of the workpiece surface with the CBN grinding wheel is provided in Fig. 1. The workpiece surface is observed in this figure to have chips which adhere to it because of both a high grinding temperature and also because of a rubbing trace parallel to the grinding direction. In this situation, grinding fluid is important regarding its ability to lubricate, conduct heat, and carry away swarf from the grinding surface.

3.2. Dimensional accuracy The actual grinding depth is less than the down-feed due to the deflections of workpiece and the grinding wheel and the wear of the grinding wheel. The deviation between the actual grinding depth and the down-feed would seriously affect the dimensional accuracy. The method for measuring the actual grinding depth is as follows: A plane is first ground as the basis plane on the workpiece. The down-feed is then set to be equal to

Fig. 1. The workpiece surface for the CBN grinding wheel.

10 ~tm and grinding continued on the basis plane. The dial indicator is finally used to measure the discrepancies between the grinding plane and the basis plane. The results for one-pass spark out and no spark out with the CBN grinding wheel are 9.5 and 8 !am, which are much better than the values for the WA grinding wheel, i.e. 8 and 5 lam. The reason for this is apparently the high stiffness of the CBN grinding wheel; consequently, the deflection of the grinding wheel is reduced.

3.3. Grinding force The relationship between grinding force and the grinding passes for the GC wheel is shown in Fig. 2. The grinding forces are found in this figure to increase with the number of grinding passes due to the dulling of the grinding wheel. Additionally, the grinding force increases rapidly when the grinding fluid is used. The GC grinding wheel are therefore considered as not being suitable when the grinding fluid is used. For the WA wheel, the grinding fluid is not observed to influence the grinding force; however, it seriously affects the grinding wheel wear. Fig. 3 shows the grinding forces versus grinding passes for, the CBN grinding wheel when grinding fluid is not in use: the forces increase rapidly and result in workpiece burn. The relationship between grinding forces and grinding passes when grinding fluid is used are shown in Fig. 4. A transition period occurs during the grinding process after the truing process has been done, the grinding forces having a long stable period after this transition period. In Fig. 5 it is seen that the forces have a less intense variation in the transition period when the CBN wheel has been dressed by A1203 stick. This dressing process does not eliminate the transition period, but it can reduce its influence dramatically.

P.-L. Tso / Journal o f Materials Processing Technology 55 ~1995) 421 426

423

Ft(dr~9

--o-

1000 : :: : . . . . . . . . . . . . . . .

- - ~ - Fn(dry) Ft Fn

~_.__~?~;

100

.....?...2:5....~ ............?...........~

IO

1

6

I1

16

_L

:

21

26

3I

......................... '

36

41

46

51

Grin ding Passex Fig. 2. Grinding force versus grinding passes for the GC grinding wheel (V,,. = 3 m'min; 1/~ = 1050 m/rain: a = 10 tam}.

200 Ii 180

. . . . . . . . . . . . . .

Ft

..................

"~i

- - Err

160

/;~\"tJ

~ 140

-- *

~.

/

Ft(harden)

~ 120 100 "~ ..~

................~ i

80

y

60

........... ~ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

.... ,y

4O l / , ~ , ~

~:.::~

............................

/

;...

.

.... / . . . . . . . .

....................................................................... J , .....i ............

/

//-~g

20 1

3

5

7

9

I1

13

15

17

19

21

Grimling l'a.s'.ses Fig. 3. Grinding force versus grinding passes for the CBN grinding wheel (V,,. = 3 m/min; V, = 1050 m/min; a = 10 pm).

120 ~

Ft

~ . ~

F"

I00 ...............................................

~

~o

"~

8O

I.

60

.

.

.

.

.

.

.

.

.

. . . . . . . . . . . . . .

.

.

I ~

.

.

el'"

.

.

.

~

.

.

.

.

~

- - 1

61

71

--

v T

~

- - r

--

"

"

--

I""

=

"

40 20

1

11

21

31

41

51

81

91

I01 111 121 131 141 151

Grinding Pa~ses Fig. 4. Grinding force versus grinding passes for the CBN grinding wheel (V,, = 3 m/min; Vs = 1320 m/min; a = 5 ~tm; grinding fluid 9.4 1/min).

424

P.-L. Tso / Journal o f Materials Processing Technology 55 (1995) 421 426

120

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

100

---*--

Dressing

--tJ--

N° Dressing

4o 20

................................................................................................................................................................................

0 i

_,

1

11

2I

3!

41

5I

61

71

8I

,

91 101 111 121 1.71 14I 151

Grinding Passes Fig. 5. The influence of dressing on the grinding force (V~ = 3 m/min; I{~ = 1320 m/rain; a = 5 ~tm; grinding fluid 9.41/min).

300 ................................................................................................

250

....

200 150

. . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

'

....

I00 50

w

0 1

3

5

7

9

11

13

15

17

19

21

23

Grinding Passes Fig. 6. Normal grinding versus grinding passes for up-and down-grinding (Vw = 4 m/min; Vs = 1050 m/rain; a = 10 ~tm).

In Fig. 6 it is seen that the grinding forces in upgrinding are initially larger than those in down-grinding when the grinding wheel is sharp. Contrary is, the grinding forces of down-grinding are greater than those of up-grinding when the grinding wheel is dull. Malkin [4] divided the grinding force into cutting force, FH¢, and sliding force FH~, the sliding forces increasing proportionally with wear flat area whilst the cutting forces remain constant. Up-grinding produces a short chip length and a large undeformed chip thickness; therefore, it has a larger cutting force than sliding force when the wheel is dull and dominates the total grinding force, down-grinding then having a greater grinding force due to its greater sliding force. 3.4. Grinding-wheel wear Fig. 7. Wear photograph of the C B N grinding wheel.

Grinding-wheel wear affects the size and the surface roughness of the ground part. Fig. 7 shows a photograph of the CBN grinding wheel wear, the large white areas being loading wear and the small white spots being

attritious wear which is caused by the dulling of the abrasive grain rubbing against the workpiece surface. Each loading wear area has a small chip adhering to the

P.-L. Tso / Journal of Materials Processing Technology. 55 (1995) 421 426 12

-,

425

14

12

10 ~10

~.8

,~

8

6

2~ % ~. 4

.

4

2

.................

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

..............................

.

....

i

.

.

.

.

.

.

.....................

i ............................................................................................. ~o cq

0 0

20

40

60

80

1

100

Grinding Passes Fig. 8. Wear flat area versus grinding passes (CBN) (V. = 3 m/rain; V~ = 1050 m/rain; a = 5 lain; grinding fluid 9.4 l/min).

IO0 Ft

3

4

Workpiece

Speed

(re~rain)

5

6

Fig. 10. Specific cutting energy versus workpiece speed.

-,16 •

2

i

/ ~

80

...........

K

. . . . . . . . . . . . . . . . . . . . . . . . . . . .

S

~60 u,

8 6

k~ 2O

.................................. i . . . . . . . . . . . . . . .

___~_ 2

6

10

12

14

Wear Flat Area (0.01%)

i

................ !...........

0

0

I

20

30

40

50

Material Removal Rate (10 "6 m 2 / m i n )

Fig. 9. Grinding force versus wear flat area (CBN) (V~ = 3 m/rain; V~ = 1050 m/rain: a = 5 lain; grinding fluid 9.4 l/min).

wheel which will drop off after some passes. The wear flat area has been calculated and plotted versus grinding passes in Fig. 8. This figure shows that the wear flat area increases intensely during the transition period because of wheel loading; then decreases and maintains a constant value thereafter, and it also indicates that the unstable grinding forces in Fig. 3 are caused by wheel loading. Fig. 9 shows that the grinding force increases linearly with wear flat area for the CBN grinding wheels. Malkin [7, 8] divided the total grinding energy into plowing, chip formation, and sliding energy, the specific cutting energy, uc, which is the sum of plowing and chip formation energy, being calculated from FHc × V w

u~ -- b x ~ x d'

.......................................

i

L__ 4

4

(2)

Fig. 11. Specific cutting energy versus metal removal rate.

where FH¢ is the cutting component force of the horizontal grinding force; V,~ is the workpiece speed; b is the grinding width; Vs is the grinding wheel speed; and d is the down-feed. Values for the specific cutting energies from Eq. (2) are plotted in Figs. 10 and 11 as a function of workpiece speed and material removal rate. Decreasing either the workpiece speed or the material removal rate is observed in these figures to lead towards a steep rise in the specific cutting energy because of the size effect. On the other hand, as either the workpiece speed or material removal rate is made very large, it would appear that uc approaches a value of about 30 GJ/m 3. Additionally, the specific cutting energy is about 13.8 GJ for steel [7-] because of the lower material strength of this material than that of Inconel 718.

P.-L. Tso / Journal o f Materials Processing Technology 55 (1995) 421-426

426

Table 1 The grinding-wheel life for the GC grinding wheel Vw (m/min)

V~ (m/min)

a (pro)

Fluid (l/min)

Hardness (HRC)

Life (mm 3)

2 3 4 4 4 5 3 3 3 4 2

1050 1050 1050 1050 1050 1050 680 1050 1320 1050 1050

10 10 5 10 15 10 10 10 10 10 10

0 0 0 0 0 0 0 0 0 0 9.4

8 8 8 8 8 8 40 40 40 40 40

155.52 181.44 77.76 155.52 155.52 129.6 129.6 129.6 181.44 207.36 51.84

Table 2 The grinding-wheel life for the WA grinding wheel Vw (m/min)

Vs (m/min)

a (lira)

Fluid (l/rain)

Hardness (HRC)

Life (mm 3)

3 3 5 3 3 3 3 3 3 5

1050 1320 1050 680 1050 1320 1050 1320 1320 1050

10 10 10 I0 10 10 10 10 20 10

0 0 0 0 0 0 9.4 9.4 9.4 9.4

8 8 8 40 40 40 40 40 40 40

155.52 233.28 311.04 129.6 103.68 116.64 259.2 259.2 518.4 259.2

Table 3 The grinding-wheel life for the CBN grinding wheel

estimating the grinding-wheel life. From these tables, it can be found that dry grinding is better for the GC wheel, whilst the CBN grinding wheel is observed to have the longest grinding-wheel life if proper grinding conditions are provided. 4. Conclusions The following conclusions could be made on the basis of the above discussion: 1. The surface roughness increases with decreasing wheel speed and increasing work speed and down-feed for Inconel 718 material. Additionally, the surface roughness in down-grinding is better than that in up-grinding. 2. Using grinding fluid for the GC grinding wheel results in a high grinding force, quicker dulling of the grinding wheel and high surface roughness, so that this wheel is better suited to dry grinding. 3. The CBN grinding wheel has the optimum dimensional accuracy, due to the high stiffness of this grinding wheel. 4. A transition period occurs of the CBN grinding wheel in the grinding process, no matter whether the dressing process is used or not. However, an adequate dressing process can reduce the fluctuation of grinding force, grinding wheel wear, and surface roughness in the transition period. 5. The CBN grinding wheel is the most suitable for grinding Inconel 718 no matter what machinability index is considered, i.e. grinding force, surface roughness, dimensional accuracy, grinding wheel life, except cost.

Vw (m/rain)

Vs (m/min)

a (~m)

Fluid Hardness ( 1 / m i n ) (HRC)

Li~ (mm 3)

References

3 3 3 3 2 2 3 3 3 3 3

680 1050 1320 1050 1050 1050 1050 1320 680 1050 1320

I0 10 10 10 5 10 5 5 10 10 10

0 0 0 0 9.4 9.4 9.4 9.4 9.4 9.4 9.4

51.84 25.92 25.92 155.52

[1] C.T. Lynch, Practical Handbook of Material Science, CRC Press, Boca Raton, FL, 1989, second printing, 1990. [2] J. Peklenik, Untersuchen uber das Versehleisskriterium biem Schleifeus, Industrie-Anzieger, 80 (1958) 280-289. [3] H. Yoshikawa, Criterion of grinding wheel tool life, Bull. Japan Soc. Grinding Eng., 3 (1963) 29-32. [4] S. Malkin and N.H. Cook, The wear of grinding wheels, Part 1 attritious wear, J. Eng. Ind. Trans ASME Ser. B, 93(4) (1971) 1120--1128. [5] S. Malkin and N.H. Cook, The wear of grinding wheels, Part 2 93(4) (1971) 1129-1133. [6] S. Malkin and S. Kannappan, Effects of grain size and operating parameters on the Mechanics of grinding, J. Eng. Ind. Trans ASME Ser. B 94 (1972) 833-842. [7] S. Malkin and R.B. Anderson, Thermal aspects of grinding, Part l-energy partition, J. Eng. Ind. Trans ASME Ser. B 96 (1974) 1177-1183. [8] S. Malkin and R,B. Anderson, Thermal aspects of grinding, Part 2 - surface temperatures and workpiece burn, J. Eng. Ind. Trans ASME, Set. B 96 (1974) 1184-I 191. [9-1 S. Malkin, Selection of operating parameters in surface grinding of steels, J. Eng. Ind. Trans ASME Ser. B 98 (1976) 56 62. [10] Metals Handbook, 9th ed., Vol. 4, Heat Treating, American Society for Metals, 1981.

8 8 8 40 40 40 40 40 40 40 40

Qc

oc oc

3.5. Grinding-wheel life Many methods are available to determine the grinding-wheel life. The criterion for the grinding wheel is the dramatic change of grinding force, as described above. The grinding-wheel life of GC, WA and CBN wheels are presented in Tables 1, 2 and 3, respectively. Different grinding parameters have been used for these testing processes, the volume of material removed being used for