1
A. INTRODUCTION OF DISSERTATION
1. Dissertation title
Optimization on determination of dressing parameters, lubricant
conditions and exchanged grinding wheel diameter in internal cylindrical
grinding process
2. Rationale of the study
Nowadays, according to the great development of technologies,
machining processes have to satisfy more and more requirements of
mechanical products for quality as well as productivity. In reality, among
machining processes, grinding is commonly used to obtain the high quality
of surface finish. Especially, it predominates in machining annealed
products with high hardness, high strength. It accounts for about 20-25% of
the total expenditures for mechanical parts in industries. Because of these
reasons, improvement of grinding performance and reduction of machining
expenditure while remaining accuracy requirement have been interested in
researchers.
In comparison with other type of grinding, internal cylindrical grinding
process is implemented in difficult conditions and tight spaces. For that
reason, it is more difficult to study the process of internal grinding.
Therefore, the research of the internal grinding process is less interested by
scientists than studying external grinding or surface grinding.
In order to improve the internal grinding performance, many solutions
have been proposed such as using high standard grinding wheels (diamond
or CBN wheel), high speed grinding and optimizing grinding process
parameters (cutting, dressing and lubricant parameters). Among these
solutions, optimization of grinding process parameters has been considered
in many studies.
90CrSi is steel alloy with high mechanic strength and abrasion
resistance. It is commonly applied to make molds, low speed cutting tools
and machine parts required high durable and abrasion resistance. In medical

grinding productivity; Study on the calculation model of internal grinding
cost and the influences of grinding process on grinding cost; Determination
of optimal exchanged grinding wheel diameter.
7. New contributions
This study has analyzed the internal grinding cost and the influence of
grinding process parameters on the grinding cost.
Determining model to calculate the optimal exchanged grinding wheel
diameter (or optimum wheel lifetime) in internal grinding process and the
influence of grinding process parameters on the optimal exchanged wheel
diameter.


3
The lubricating-cooling parameters and dressing parameters have been
analyzed and optimized based on grinding experiments of 90CrSi steel
alloy.
8. Dissertation structure
The dissertation includes the following parts: Introduction, 5 chapters,
conclusions and appendix.
Chapter 1: Overview of internal grinding process.
Chapter 2: The model of efficiency improvement of internal grinding
process and experimental system.
Chapter 3: Experimental study on influence of lubricating-cooling
parameter in internal grinding process
Chapter 4: Experimental study on influence of dressing parameter in
internal grinding process
Chapter5: Determination of optimal exchanged grinding wheel diameter.
9. Significances
Science significances
This dissertation has studied the influence of lubricating-cooling

grinding cost such as Tarasow – Shaw, Field and Ebbrells – Rowe are
reviewed and analyzed.
1.4. Proposal solution to improve grinding efficiency

-

Determination of appropriate lubricating-cooling conditions;
Determination of optimal dressing parameters;
Determination of optimal grinding wheel life (optimal exchanged
grinding wheel diameter).

CHAPTER 2. MODEL TO IMPROVE THE EFFICIENCY OF
INTERNAL GRINDING PROCESS AND DEVELOPMENT OF
EXPERIMENTAL SYSTEM
2.1. Model to improve the efficiency of internal grinding process
Normally, researches focus on the technical efficiency of the grinding
process to improve the accuracy and the ground surface quality; reduce
force, heat, vibration or increase productivity. In order to solve both
directions, the dissertation develops a model to improve the efficiency of
the internal cylindrical grinding process. This model has been proposed


5
including three parts to increase accuracy and machining surface quality and
reduce grinding cost in internal cylindrical grinding process.
Part 1: Input parameters.
Internal grinding process is complex under the influences of many input
parameters. These input parameters can be classified into five groups
including: grinding machines and cutting parameters; workpieces; grinding
wheels; dressing technologies and lubricating-cooling technologies. Among

internal grinding process for one product.

Cmt,p
Ct,pmin; Lop

Cgw,p
Ct,p

Grinding wheel life- L (hour)

Figure 2. The relationship between the grinding wheel life and grinding
cost
The longer the grinding wheel life, the lower the cost of grinding wheel
is. In contrast, the cost for machines, labors and management linearly
increases with the machining time. The total machining cost for a part
includes the expenditures of grinding wheel, machines, labors and
management ... In Figure 2, a certain optimal grinding wheel life always
exists.
2.2. Experimental system
Experimental system includes technical system and measurement
devices
2.3. Conclusion of chapter 2.
1. The input and output parameters have been analyzed and determined
as the following:
- Input parameters: Vđ, Vct, fa, fr, ae,tot, Cm,h, Cwa,h, dw, Rld tg, Srg, D0, Bgw,
wpd, Cgw, tw, tsđ, Ssđ, nsđ, NĐ, LL.
- Output: Ra, Ct,p and De,op


7

2
3
4
5
6
7
8
9
10
11
12
13

P6
P8
P3
P2
P9
P1
P9
P7
P9
P4
P5
P9
P9

-1
1
0

1
4
2,5
4
2,5
4,6
2,5
2,5
2,5
1
0,3
2,5
2,5

Concentration
(%)
2
2
5,6
5
3,5
3,5
3,5
1,3
3,5
5
3,5
3,5
3,5


reduced. However, the space of internal grinding is limited by the gring
wheel dimension, increasing the flow rate does not increase amount of the
coolant in cutting area. In addition, increasing the flow rate increases the
concentraion of the coolant in the cutting area and also increases the chips
on the workpiece surafce. That is an interaction effect between two
parameters on the roughness Ra.

Figure 3. Regression surface of Ra for Caltex Aquatex 3180 oil
b. Emulsion


9
Table 2. Experimental results for Emulsion oil

No.

Points

1
2
3
4
5
6
7
8
9
10
11
12

-1
-1,4
0
0

-1
-1
1,4
1
0
0
0
-1,4
0
1
0
0
0

Uncode
Concentr
Flow rate
ation
(l/m)
(%)
1
3
4
3
2,5

0,366
0,311
0,371
0,487
0,452
0,354
0,356

Results and Discussions
Using Minitab software, analyzing the experiment results, we obtained
the regression equation:
Y= 0,218 – 0,006x1+0,038x2 - 0,016x1x2 + 0,016x12 + 0,004x22
(2)

Figure 4. Regression surface of Ra for Emulsion oil


10
In figure 4, when the flow rate is low, the concentration strongly affects
to the roughness value. Increasing the concentration, surface roughness
increase. When flow rate equal 4 l/min, the roughness value is almost
constant for all concentration value. In general, the more Emulsion oil will
increase the surface roughness. This is because Emulsion solution is high
density, makes it difficult to escape chips and clean the machining surface.
3.1.2. Optimization of the concentration and the flow rate
a. Caltex Aquatex 3180 oil
Using response surface method, the relation between the concentration
and the flow rate with the roughness Ra is shown. From the optimization
plot, there exists an optimal set of these parameters to obtain a minimum
roughness Ra. The solution of the optimal parameters are shown, the

1
2
3
4
5
6
Non-feeding
1
CK
0
1
2
3
4
5
dressing times
Coarse
2
dressing
ttho
0,02 0,025 0,03
depth (mm)
Coarse
3
ntho
1
2
3
dressing times
Fine dressing



12

Mean of Ra

4.2. Effect of dressing parameter to surface roughness (Ra) and
material remove rate (MRR) in internal grinding.
4.2.1. Experiment results and single-objective optimization.
a, Influence of dressing parameter to Ra
From the analysis of variance – ANOVA, it is clearly seen that the nonfeeding dressing times has the largest effect on surface roughness Ra, The
other parameters, which have the effect on Ra, sequence: coarse dressing
depth, coarse dressing times, fine dressing depth, fine dressing times and
dressing feed rate.
Table 4. The effect of dressing parameters on Ra at their levels
Level
CK
ttho
ntho
ttinh
ntinh
Ssd
1
0,4043
0,4929
0,4797 0,5193 0,5146 0,5023
2
0,4407
0,4808
0,5034 0,4836 0,5059 0,5144

If dressing has not the non-feeding dressing times, topography of
grinding will become rougher. The space for escaping, containing chip is
larger, so cutting heat, force and roughness decrease. The more non-feeding
dressing times reduce, the more ridiculous peaks will be reduced and thus
increasing Ra
Dressing depth increases, surface is rougher, grinding wheel time life
and MRR increase (suitable for rough grinding). Coarse dressing times
increase, thus Ra increase. The reason is that coarse dressing times increase,
number of undulating peaks in grinding increases and Ra increase.
Fine dressing depth is too small, leading to the undulating height on
surface grinding small, so that difficult to contain and escape chips, leading
to Ra increase. In other way, fine dressing depth increase, the undulating
height on surface grinding is higher but quickly flattened, so that grinding
wheel is worn out rapidly and Ra increase.
b. Optimum Surface roughness

Figure 6. Effect of dressing parameters on S/N
The optimal value of Ra is determined by the parameter level (circle) in
figure 6: CK = 0 time (A1); ttho = 0,025 mm (B2); ntho = 1 time (C1); ttinh =
0,01mm (D2); ntinh = 3 time (E3); Ssd = 1,4 m/min (F3).
Optimum value of Ra
Ratoiuu  A1  B2  C1  D2  E3  F3  5.Tgg


14
Substituting all of the parameters into equation
Ratoiuu  0.404  0.481  0.48  0.484  0.493  5x0.504  0.318m

A confidence interval (CI) can be computed as:
 1

Level
CK
ttho
ntho
ttinh
ntinh
Ssd
1
2,109 2,446 2,450 2,577 2,253
2,355
2
2,033 2,463 2,384 2,314 2,382
2,426
3
2,475 2,318 2,393 2,336 2,591
2,445
4
2,462
5
2,438
6
2,937
Delta
0,905 0,146 0,066 0,264 0,338
0,090
Rank
1
4
6
3

number of fine dressing is the second most powerful factor on MRR after
the number of superfine dressing. MRR is proportional to the number of
fine dressing.


16
The dressing feed does not affect much to the MRR (similar to the
affecting to Ra).
d. Optimization of MRR
The value MRR max is determined by the following equation at levels:
CK (A6); ttho (B2); ntho (C1); ttinh (D1); ntinh (E3); Ssđ (F3).
MRRtoiuu  A5  B2  C1  D1  E3  F3  5.Tgg

And we have: MRRtoiuu  3, 42(mm3 / s)
The CI confidence interval is calculated as follows:
 1 1
CI   F 1, f e  ,Ve , 
   0, 415
 Ne R 

Mean of S/N ratio

Where, 𝐹∝ (1, 𝑓𝑒 ) = 8,5262 is a coefficient with significance level
%=90%, fe =2 is the degree of freedom of error, Ve = 0,032125 is the
average error, neff is the number of effective iterations, R = 3 is the number
of iterations of an experiment.

Figure 8. Effect of factors on S/N of MRR
𝑁𝑒 =


times, CK = 5 times, ntinh = 3 times, ttinh = 0,005 mm, Ssđ = 1,4 m/ph.
Whereby:
(𝑅𝑎) 𝑇𝑜𝑖𝑢𝑢 = 0,4929 + 0,4797 + 0,563 + 0,5193 + 0,4929 + 0,4966 − 5
∗ 0,5045 = 0,522 µ𝑚
(𝑀𝑅𝑅) 𝑇𝑜𝑖𝑢𝑢 = 2,446 + 2,45 + 2,937 + 2,577 + 2,591 + 2,445 − 5
∗ 2,4089 = 3,402 𝑚𝑚3 /𝑠

Figure 9. Main effect plot for means
4.4. Conclusions of chapter 4
1. The process of dressing should follow rough, fine and super fine
dressing steps to help stabilize the topography of the wheel. The number of
times the super fine dressing has the greatest effect on the surface roughness
and the grinding performance. The super fine dressing can reduce the
surface roughness but it can help to increase the grinding productivity


18
significantly. The greater the depth of rough dressing and fine dressing can
increase the surface roughness and reduce the MRR. Therefore, it is
advisable to choose a suitable depth of dressing. The larger of the number of
rough dressing also increase the surface roughness and reduce the MRR.
Also, the more fine dressing times will help reduce the surface roughness
and increase MRR.
2. The results of the study help to choose the optimum dressing mode
when internal grinding 90CrSi tool:
+) For minimum surface roughness (fine grinding) the optimum dressing
parameters are: (CK = 0; ttho = 0,025mm; ntho = 1; ttinh = 0,01mm; ntinh = 3;
Ssđ = 1,4m/p) Ramin = 0,318µm
+) For maximum grinding productivities (rough grinding) the optimum
dressing parameters are (CK = 5; ttho = 0,025mm; ntho = 1; ttinh = 0,005mm;

. t L  t s  t c  1  d 
Cgw 
  
60
t
(
D

D
).
t
(
D

D
).
t
w
0
e
w
0
e
w

 



5.2. Effect of parameters on the cost of the internal grinding process


20
cut should be selected appropriately, in accordance with machining
requirements in order to reduce the cost.
The most powerful impact on the grinding cost is the ratio of the hole
length to hole diameter Rld and the part diameter. When this ratio is larger
the grinding cost will increase. Meanwhile, the processing conditions will
be harsher and the horizontal feed speed cannot be large. Also, the amount
of removal material is also large, thus increasing the grinding cost. In
addition, the greater the surface roughness grade Srg will increase the
grinding cost. So to reduce grinding costs should not choose the large Rld (if
possible).  is a factor closely related to the original part diameter d w and
the initial grinding diameter D0. Increasing the ratio  will increase the
grinding cost.
As analyzed above, the increase in grinding wheel cost will increase the
grinding cost. However, the higher the wheel lifetime will reduce the cost.
Also, the impact of wheel lifetime is greater than the impact of wheel cost.
In addition, the amount of wheel wear wpd and the width of wheel B gw do
not affect the grinding cost much. Therefore, if we use high quality grinding
wheel (expensive, durable) we can reduce the grinding cost. In addition,
optimizing grinding parameters to increase the wheel lifetime also helps to
reduce the grinding cost.
The initial wheel diameter D0 and the hole diameter are two parameters
depending on the coefficient . Therefore, increasing D0 can increase the
average cutting speed and reduce the machining time. However, in this case,
the processing conditions are also changed. Therefore, the amount of
removal material increases and increases the cost of grinding. Besides, the
exchanged wheel diameter De also affects the grinding cost. When delta (De
/ D0) decreases (or De decreases), it will reduce the grinding cost.
5.3. Optimal exchanged wheel diameter

Cmin = 5.927 VNĐ
De,op = 17,5

6100
5600

13

14

15

16

17

18

19

20

Exchanged grinding wheel diameter - De (mm)

Figue10. Exchanged wheel diameter versus grinding cost
As mentioned above, because the exchanged wheel diameter greatly
affects the cost of grinding, finding the value of the optimal exchanged
wheel diameter will help to reduce grinding cost significantly. When
comparing the cost of grinding when changing the wheel at the optimum
exchanged diameter De, op = 17.5mm with the cost of replacing the wheel

such as reducing the cost of machines, the cost of grinding wheel, the labor
costs (workers, management ...); Using abrasive wheel with high durability
and studying methods to improve the wheel lifetime and determine the
appropriate amount of dressing depth of cut; Finding methods to reduce the
dressing time and the time for changing dressing tools...
2. The exchanged wheel diameter greatly affects the cost of grinding.
Also, there exists an optimal value of the exchanged wheel diameter at
which the grinding cost is minimal. In addition, a formula to determine the
optimal exchanged grinding wheel diameter De, op has been proposed.
3. The influence of these factors on the optimum exchanged wheel
diameter is as follows: The initial diameter of the grinding wheel D0 has the
strongest influence on the exchanged grinding wheel diameter De,op, next is
the grinding wheel cost Cgw, the wheel lifetime tw, the machine cost Cm,h,
the labor cost Cwa,h. Also, the dressing depth of cut aeđ. The ratio Rld, the
wheel wear wpd, the accuracy grade tg do not affect De,op. The quadratic


23
factors influence De,op are D0*tw, D0*Cgw, D0*Cmh, tw*Cgw, D0*Cwah, D0*aed,
Cmh*Cwah, Cmh*Cgw and the final is Bgw*aed.
- The economic efficiency of applying the optimum exchanged wheel
diameter helps to reduce the cost of grinding per part by 9.02%, the total
grinding time decreases by 14.7%.
CONCLUSIONS AND RECOMMENDATION
Conclusions
The objective of this thesis is to improve the efficiency of internal
grinding process. In order to do that, it is necessary to solve the following
problems: Determining a reasonable cooling lubrication mode, determining
a reasonable dressing parameters and determining the optimal exchanged
wheel diameter. The main results and new contributions of the thesis can be

can be reduced by 9.02%, the total grinding time is reduced by 14.7%. This
method is applicable in cases where the grinder is unable to change the
spindle rotation speed.
Recommendation
Although this research has found a number of solutions to improve the
efficiency of internal grinding process, there are still issues that need further
investment in research. Specifically include the following research
directions:
1) Research on the method to supply of the coolant into deep areas of
grinding.
2)

Cutting conditions when grinding small and deep holes with the
diameter less than 10 mm are very fierce. Therefore, it is needed
further researches.

3)

Investigation of the effects of coolant parameters and dressing
parameters on the mechanical and physical properties of the workpiece
surface.


25
LIST OF PUBLISHED WORKS RELATED TO THE THESIS
* Internal journal papers
1. Banh Tien Long, Vu Ngoc Pi, Le Xuan Hung, Ta Viet Cuong, A study on the
effects of coolant regimes to surfaceroughness in in ternalgrinding of steel 9XC,
VietNam Mechanical Engineering Journal, Vol 5, 2016, pp 71 – 76 (In Vietnamese)
2. Banh Tien Long, Vu Ngoc Pi, Le Xuan Hung, Luu Anh Tung, Buiding

9. Le Xuan Hung, Vu Thi Lien, Vu Ngoc Pi, Banh Tien Long, “A Study on
Coolant Parameters in Internal Grinding of 90CrSi Steel”, Materials Science
Forum, Vol. 950, pp 24-31, Apirl, 2019 Trans Tech Publications, Switzerland.
Scopus