Katholieke Universiteit Leuven

Tribology: Final assignment [H04T9a]

Solution techniques for Reynolds equation

Sandeep Banik
r0605642

June 16, 2017

Contents
1 Introduction 1

2 Reynolds equation 2
2.1 Non dimensional reynolds equation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
2.2 Solution strategy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
2.2.1 Integration method . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
2.2.2 Hybrid method . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
2.2.3 Finite difference method . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
2.3 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8

3 Viscosity variation in Reynolds equation 10

3.1 Pressure viscosity relation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
3.1.1 Low viscous oil . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
3.1.2 High viscous oil . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
3.2 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11

4 Elasto-hydrodynamic lubrication 13
4.1 EHL in FDM . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14

5 Application in journal bearing 16

6 Conclusion and future work 18

i
List of Figures
1 Pressure distribution long bearing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
2 Non dimensional pressure distribution long bearing . . . . . . . . . . . . . . . . . . . . . . 5
3 Pressure distribution short bearing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
4 Non dimensional pressure distribution short bearing . . . . . . . . . . . . . . . . . . . . . 6
5 Non dimensional pressure distribution approach 1 . . . . . . . . . . . . . . . . . . . . . . . 6
6 Non dimensional pressure distribution hybrid approach . . . . . . . . . . . . . . . . . . . . 6
7 Non dimensional pressure distribution approach 2 . . . . . . . . . . . . . . . . . . . . . . . 7
8 Non dimensional pressure distribution hybrid approach . . . . . . . . . . . . . . . . . . . . 7
9 Non dimensional pressure finite difference method . . . . . . . . . . . . . . . . . . . . . . 8
10 Non dimensional pressure distribution approach 1 . . . . . . . . . . . . . . . . . . . . . . . 8
11 Non dimensional pressure distribution hybrid approach . . . . . . . . . . . . . . . . . . . . 8
12 Evolution of viscosity with pressure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
13 Pressure distribution with constant viscosity for low viscous oil . . . . . . . . . . . . . . . 11
14 Pressure distribution with viscosity variation for low viscous oil . . . . . . . . . . . . . . . 11
15 Pressure distribution with constant viscosity for high viscous oil . . . . . . . . . . . . . . . 11
16 Pressure distribution with viscosity variation for high viscous oil . . . . . . . . . . . . . . 11
17 Pressure distribution on spherical contacts . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
18 Pressure distribution on spherical contacts . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
19 Elastic deformation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
20 Pressure distribution with elastic deformation . . . . . . . . . . . . . . . . . . . . . . . . . 16
21 Journal bearing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
22 Pressure distribution in journal bearing . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17

ii
1 Introduction
Tribology can be explained as ’the science and technology of interacting surfaces in relative motion and
of related subjects and practices’. This tribological behavior can be influenced by the physical, chemical
and mechanical properties of the surface and near surface materials.

Discovery of Reynolds equation throws light on the effect of pressure distribution in a liquid film lubri-
cating bearing. Considering newtonian fluid, negligible inertia and body forces, laminar flow regime, etc.,
this equation was initially developed. However,later few relaxation on the assumptions are granted to
produce a more generalized form through which pressure distribution on the bearing surface for arbitrary
film curvature can be determined. Further, with this calculated pressure distribution other parameters
such as friction force, flow rates etc, can be obtained easily.

Since Reynolds equation is a partial differential equation, therefore it is important to develop solution
strategies to solve the Reynolds equation.

Three main solution methods are studied and implemented. The solution strategies are compared
and each advantage and disadvantage is discussed. At the end an appropriate solution strategy for a
journal bearing of some given specifications is discussed to determine the pressure distribution.

1
2 Reynolds equation
After Beauchamp Tower performed his experiments on pressure in a lubricated bearing, Professor Os-
borne Reynols in 1886 developed the governing equation that decsribed the pressure in a fluid-flim
lubricated bearing, known as Reynolds equation. In lubrication theory, the Reynolds equation is given
as. The general Reynolds equation is:
 3   3     
∂ h ∂p ∂ h ∂p ∂ h (U2 − U1 ) ∂ h (V2 − V1 ) ∂h
+ = + + (1)
∂x 12µ ∂x ∂y 12µ ∂y ∂x 2 ∂x 2 ∂t
Where
• p is fluid film pressure
• h is film thickness
• x and y is the bearing length and width
• µ is the viscosity of the fluid
• U2 and U1 is the velocity in the x direction
• V2 and V1 is the velocity in the y direction
The right hand side of the equation 1 is the pressure term and left hand side is the source term.

Under the assumption of constant viscosity and no relative velocity in the y direction the equation 1
becomes

∂ h3 ∂p ∂ h3 ∂p
     
∂ h (U2 − U1 ) ∂h
+ =µ +µ (2)
∂x 12 ∂x ∂y 12 ∂y ∂x 2 ∂t

Further assuming one surface with zero tangential velocity,

     
∂ ∂p ∂ ∂p ∂ h (U2 ) ∂h
h3 + h3 = 12µ + 12µ (3)
∂x ∂x ∂y ∂y ∂x 2 ∂t

The equation 3 is valid with the following assumptions

1. Negligible inertia terms
2. Negligible pressure gradient in the direction of film thickness
3. Newtonian fluid

2
 4. Constant value of viscosity
5. No slip at the liquid solid interface
6. Incompressible flow
7. No relative tangential velocity in y direction
8. Rigid surfaces
9. Only inclined surface slides

2.1 Non dimensional reynolds equation

The equation 3 can be non dimensionalized, so as to carry out order analysis. Representing the dimensions
as follows [2],
x y h
.x̄ = , ȳ = , h̄ = (4)
X Y C

where, X,Y and C is the length, width and film thickness respectively. Substituting back in the equation 3,
     
∂ 3 ∂p ∂ 3 ∂p ∂ h̄C (U2 ) ∂ h̄C
(h̄C) + (h̄C) = 12µ + 12µ (5)
∂(x̄X) ∂(x̄X) ∂(ȳY ) ∂(ȳY ) ∂(x̄X) 2 ∂t

C3 C3 X2 ∂
   
∂ 3 ∂p 3 ∂p C ∂ h̄ C ∂ h̄
h̄ + h̄ = +2 (6)
6µU X 2 ∂ x̄ ∂ x̄ 6µU X 2 Y 2 ∂ ȳ ∂ ȳ X ∂ x̄ U ∂t

Further on non dimensionalization of pressure and time,

C 3P tU
.p̄ = , t̄ = (7)
6µU X 2 C

X2 ∂
   
∂ 3 ∂ p̄ 3 ∂ p̄ C ∂ h̄ ∂ h̄
h̄ + 2 h̄ = +2 (8)
∂ x̄ ∂ x̄ Y ∂ ȳ ∂ ȳ X ∂ x̄ ∂t

The equation 8 provides a basis to carry out order analysis. For example
   
X ∂ ∂ p̄ ∂ ∂ p̄ C ∂ h̄ ∂ h̄
if = 10 h̄3 + 100 h̄3 = +2 (9)
Y ∂ x̄ ∂ x̄ ∂ ȳ ∂ ȳ X ∂ x̄ ∂t

   
X ∂ ∂ p̄ ∂ ∂ p̄ C ∂ h̄ ∂ h̄
if = 0.1 h̄3 + 0.01 h̄3 = +2 (10)
Y ∂ x̄ ∂ x̄ ∂ ȳ ∂ ȳ X ∂ x̄ ∂t

In equation 9 the pressure gradient in the x-direction can be ignored, whereas in equation 10, the pressure
gradient in the y-direction can be ignored. Therefore, it is observed that reynolds equation is dependent
on the geometry.

2.2 Solution strategy

The solution strategy for the steady state reynolds equation is given by 3 approaches
1. Integration method
2. Hybrid method
3. Finite difference method
The integration method and hybrid method shown here applies to slider bearing. However, with modi-
fication it can be applied to a range of bearings.

3
2.2.1 Integration method
The integration method is based on the dimensions of the bearing in the x and y direction. If X >> Y
it is considered short bearing assumption and if Y >> X is considered to be long bearing assumption.

Long bearing assumption

In the long bearing assumption as mentioned above Y >> X. Therefore, the equation 8 can be modified as,
 
∂ 3 ∂ p̄ C ∂ h̄
h̄ = (11)
∂ x̄ ∂ x̄ X ∂ x̄

Since, the bearing under consideration is a slider bearing, the film thickness is wedge shaped. Maximum
wedge height is defined as hmax and minimum wedge height as hmin . C is defined as the average film
thickness given by,
hmax + hmin
C=
2
The film thickness in terms of maximum and minimum film thickness is defined as,

hmax − (hmax − hmin )x̄

h̄ =
C
The derivative of the non dimensional film thickness with respect to non dimensional length is given by,

∂ h̄ −(hmax − hmin )
=
∂ x̄ C
Let α = hmax − hmin . Therefore the equation 11 is written as,
 
∂ 3 ∂ p̄ C α
h̄ =− (12)
∂ x̄ ∂ x̄ XC

Integrating the equation once with respect to x,

∂ p̄ α
h̄3 = − x̄ + C1 (13)
∂ x̄ X
∂ p̄
Applying the boundary condition on 13 i.e., ∂ x̄ = 0 at x̄pmax
α
C1 = x̄pmax
X
Integrating equation 13 once again with respect to x̄,
∂ p̄ α x̄pmax − x̄
=
∂ x̄ X " h̄3 #
x̄pmax − hmax
Z Z
∂ p̄ αC α 1
=
3 dx +
2 dx
∂ x̄ X hmax
− hmax (14)
α x̄ α − x̄
" #
α2 C x̄pmax − hmax α 1
p̄ =
2 + hmax
+ C2
X hmax
− x̄ 2 α − x̄
α

Applying the boundary conditions on 14, p̄ = 0 at x̄ = 0

α2
 
hmax
C2 = − 2 x̄max + (15)
2hmax α
" #
α2 C x̄pmax − hmax α2

α 1 hmax
p̄ =
2 + hmax
− 2 x̄max + (16)
X hmax
− x̄ 2 α − x̄
2hmax α
α

4
Example case: Assume a thrust pad of 10*100 mm dimensions. Leading and trailing film thicknesses
are 0.04 and 0.02 mm respectively. Sliding speed is 20 m/s. Viscosity of oil is 10 mPa.s and determin
pressure distribution.
Using the above approach the results obtained are shown in the figure 1 and 2.

Figure 2: Non dimensional pressure distribution long

Figure 1: Pressure distribution long bearing
bearing

Short bearing assumption

In the short bearing assumption X>>Y. The equation 8 is modified as,

X2 ∂
 
3 ∂ p̄ C ∂ h̄
h̄ = (17)
Y 2 ∂ ȳ ∂ ȳ X ∂ x̄

Integrating the equation 17 with respect to y

∂ p̄ CZ 2 ∂ h̄
= 3 3 ȳ + C1 (18)
∂ ȳ X h̄ ∂ x̄

Integrating the equation 18 with respect to y again

CZ 2 ∂ h̄ ȳ 2
p̄ = + C1ȳ + C2 (19)
X 3 h̄3 ∂ x̄ 2

Applying the boundary conditions on equation 19, p̄ = 0 at z̄ = 0 and z̄ = 1.

C2 = 0
CZ 2 ∂ h̄ 1 (20)
C1 = −
2h̄3 X 3 ∂ x̄ 2


CZ 2 ∂ h̄ z̄ 2 − z̄
p̄ = − 3 3 (21)
2h̄ X ∂ x̄ 2

Example case: Assume a thrust pad of 100*10 mm dimensions. Leading and trailing film thicknesses
are 0.04 and 0.02 mm respectively. Sliding speed is 20 m/s. Viscosity of oil is 10 mPa.s. Determine the
pressure distribution. Under short bearing assumption the results are shown in the figure 3 and 4. The
results comply well with the y direction, however we know that at x=0 and x=100, we must observe a
zero pressure. This is not reflected well in the results.

5
 Figure 4: Non dimensional pressure distribution
Figure 3: Pressure distribution short bearing
short bearing

2.2.2 Hybrid method

In the previous method two approaches were introduced with long bearing and short bearing assumptions.
Both the approaches are simple to implement and provide quick results, however they overestimate the
pressure when compared to actual pressure profiles. To provide a more accurate result, the long and
short bearing approach is combined to give a hybrid method [2]. Approach 1 and approach 2 corresponds
to long and short bearing respectively.
1 1 1
= + (22)
p̄ p̄approach1 p̄approach2

Example case: For example case of hybrid approach, the examples of prior approach 1 and approach
2 are taken and the results are compared. Observing figures 5, 6, 7 and 8, it is seen that the hybrid

Figure 5: Non dimensional pressure distribu- Figure 6: Non dimensional pressure distribu-
tion approach 1 tion hybrid approach

approach provides lower magnitude of pressure distribution.

2.2.3 Finite difference method

The equation 8 can be solved with numerical methods such as finite difference method,

X2 ∂
   
∂ ∂ p̄ ∂ p̄ C ∂ h̄
h̄3 + 2 h̄3 = (23)
∂ x̄ ∂ x̄ Y ∂ ȳ ∂ ȳ X ∂ x̄

6
Figure 7: Non dimensional pressure distribu- Figure 8: Non dimensional pressure distribu-
tion approach 2 tion hybrid approach

The domain of solution is discretized into m nodes in x-direction and n nodes in the y-direction. The
discretized surface is shown in the figure
The first term of the equation 23 is expanded as

h̄3i+0.5,j .p̄i+1,j + h̄3i−0.5,j .p̄i−1,j − (h̄3i+0.5,j + h̄3i−0.5,j ).p̄i,j

 
∂ ∂ p̄
h̄3 = (24)
∂ x̄ ∂ x̄ ∆x̄2
The second term of the equation 23 is expanded as

h̄3i,j+0.5 .p̄i,j+1 + h̄3i,j−0.5 .p̄i,j−1 − (h̄3i,j+0.5 + h̄3i,j−0.5 ).p̄i,j

 
∂ ∂ p̄
h̄3 = (25)
∂ ȳ ∂ ȳ ∆ȳ 2
The right hand side of the equation 23 is expanded as

∂ h̄ h̄i+1,j − h̄i−1,j
= (26)
∂ x̄ 2∆x̄
Resubstitute in equation 23 [2],

X 2 ∆x̄2 3 C∆x̄
h̄3i+0.5,j .p̄i+1,j + h̄3i−0.5,j .p̄i−1,j + 2 h̄ (p̄i,j+1 + p̄i,j−1 ) − (h̄i+1,j − h̄i−1,j )
Y ∆ȳ 2 i,j X2 (27)
X 2 ∆x̄2 3
 
= h̄3i+0.5,j + h̄3i−0.5,j + 2 2 h̄ .p̄i,j
Y ∆ȳ 2 i,j


h̄3i+0.5,j h̄3i−0.5,j
p̄i,j =   2
 p̄i+1,j +   p̄i−1,j
∆x̄2 3 2 ∆x̄2 3
h̄i+0.5,j + h̄i−0.5,j + 2 X
Y2 ∆ȳ 2 h̄i,j h̄i+0.5,j + h̄i−0.5,j + 2 X
Y2 ∆ȳ 2 h̄i,j

X 2 ∆x̄2 3
Y 2 ∆ȳ 2 h̄i,j ∆x̄C (h̄i+1,j + h̄i−1,j )
+  (p̄i,j+1 + p̄i,j−1 ) −  
h̄i+0.5,j + h̄i−0.5,j + 2X
2 ∆x̄2 3 2X h̄ X 2 ∆x̄2 3
Y 2 ∆ȳ 2 h̄i,j i+0.5,j + h̄i−0.5,j + 2 Y 2 ∆ȳ 2 h̄i,j

(28)

Initially all the nodal pressures are set to an initial condition of p̄i,j = 0. As most of the nodal pressures
are unknown an iterative loop is employed. The iterations is performed until a convergence criteria is
satisfied given by,
P  P 
n Pm n Pm
| i=1 j=1 p̄i,j − i=1 j=1 p̄i,j |
iterationk iterationk−1
P  ≤
n Pm
| i=1 j=1 p̄ i,j |
iterationk

7
 is defined as the error. If the error is kept at a low enough value, satisfactory convergence is achieved.

Example case: For example case we consider a thrust pad of 10*100 mmˆ2. Leading and trailing
film thickness are 0.04 and 0.02 mm respectively. Sliding speed is 20m/s. Viscosity of oil is 10mPa.s.
Considering 50 nodes in each direction with an error requirement of 0.0001. The results obtained are
shown in. This example is the same as the long bearing assumption.

Figure 9: Non dimensional pressure finite difference method

Figure 10: Non dimensional pressure distribu- Figure 11: Non dimensional pressure distribu-
tion approach 1 tion hybrid approach

From the results shown in the figure 9, 10 and 11, it is observed that the hybrid approach is close to
the finite difference solution. Therefore, for quick approximate solution hybrid approach can be used.

2.3 Conclusion
The following conclusion can be drawn from the solution strategies
1. Hybrid approach provides better results compared to integration method.

8
2. Finite difference method provides better results compared to integration and hybrid method. How-
ever, the computation time is high.
3. The accuracy of the finite difference method depends on the number of nodes in each direction and
number of iterations.

4. To reduce the computation time of the finite difference method initial values of the hybrid approach
can be used as initial values.

9
3 Viscosity variation in Reynolds equation
In the above analysis of reynolds equation, the fluid was assumed to be of constant viscosity. However,
in practice viscosity is a function of various factors.

Barus equation: The barus equation shows the dependency of viscosity with pressure.

η = η0 eαP

As an example, consider a initial viscosity of 158.6 mPa.s and a pressure increase from 105 to 108 . The
increase in viscosity is shown in the figure 12.

Figure 12: Evolution of viscosity with pressure

On substituting the viscosity relation in the equation 8, the following equation is obtained [2].
 3   3 
∂ h ∂p ∂ h ∂p ∂h
+ = 6U (29)
∂x η0 eαP ∂x ∂y η0 eαP ∂y ∂x

   
∂ ∂p ∂ ∂p ∂h
h3 e−αP + h3 e−αP = 6U η0 (30)
∂x ∂x ∂y ∂y ∂x
1−e−αP ∂q
Considering q = α . Therefore, ∂x = e−αP ∂P
∂x
   
∂ 3 ∂q 3 ∂ ∂q ∂h
h +h = 6U η0 (31)
∂x ∂x ∂y ∂y ∂x

3.1 Pressure viscosity relation

The equation 31 is solved using finite difference method. Since the solution is not in terms of pressure,
the solution is scaled to account for actual pressure. The results are shown for low viscous and high
viscous oils.

10
3.1.1 Low viscous oil
The geometry under consideration is thrust pad of 10*25 mmˆ2 with a sliding velocity of 20 m/s. The
leading and trailing film thickness is 0.04 mm and 0.02 mm. The viscosity at atmospheric pressure is 18.6
mPa.s with a pressure coefficient of 20∗10− 9m2 /N . The results with and without viscosity consideration
is displayed in figure 13 and figure 14.

Figure 13: Pressure distribution with constant vis- Figure 14: Pressure distribution with viscosity vari-
cosity for low viscous oil ation for low viscous oil

It is observed that the pressure distribution is not much affected with viscosity variation.

3.1.2 High viscous oil

The geometrical conditions are identical to the above low viscous oil. The viscosity of the oil at atmo-
spheric pressure is 158.6 mPa.s with a pressure coefficient of 20.7 ∗ 10− 9 m2 /N . The results with and
without viscosity consideration is displayed in figure

Figure 15: Pressure distribution with constant vis- Figure 16: Pressure distribution with viscosity vari-
cosity for high viscous oil ation for high viscous oil

3.2 Conclusion
1. The pressure distribution has a significant effect for high viscous oil.

2. The pressure distribution is not much effected for low viscous oil.

11
3. Temperature effects are not taken into consideration. There is a strong correlation of viscosity and
temperature.

12
4 Elasto-hydrodynamic lubrication
In previous analysis effect of pressure on viscosity was considered. Now considering the effect of pressure
on elasticity of the contacting (or non-contacting) materials. It is generally known that elastic defor-
mation of the material depends on “Young’s modulus” and “Poisson’s ratio”. Under compressive load,
elastic deformation of surfaces over a region surrounding the initial point of contact occurs. Associated
stresses are highly dependent on geometry of the surfaces in contact as well as loading and material
properties [2]. Elastic deformation defined by Timoshenko and Goodier is given as,

1 − ν2 F
δ1 =
2πE r
The pressure distribution in EHL between spherical object shown in the figure 18 can be modeled as,
r  r 2
p = pmax 1 − (32)
b

Figure 17: Pressure distribution on spherical con- Figure 18: Pressure distribution on spherical con-
tacts tacts

q
r 2


1−ν 2 Z b Z 2π pmax 1− b
δ1 (r, θ) = rdθdr
2πE 0 0 r

2 q
1 − ν2
Z bp
max 1 − rb (33)
δ1 (r, θ) = 2π rdr
2πE 0 r
r
1 − ν2 b
Z  r 2
δ1 = pmax 1 − dr
E 0 b

13
On assumption r = bsinθ
π/2
1 − ν2
Z
δ1 = pmax (cos2θ + 1)dθ
E 0
π/2
1 − ν2

sin2θ
δ1 = pmax + θ)
E 2 0
π/2 (34)
1 − ν2

sin2θ
δ1 = pmax +θ
E 2 0
1 − ν2
δ1 = b pmax π/2
2E
From the figure the deflection δ1 can be approximated as,

δ1 = OB − OC
p
δ1 = R − OA2 − AC 2
p
δ1 = R − R2 − b2
 s 
 2
b
δ1 = R 1 − 1 − 
(35)
R
 2 !!
1 b
δ1 = R 1 − 1− + negligible terms
2 R
b2
δ1 = R
2R2
The load for a spherical contact is given by
2 2
F = πb pmax
3
Using equation 34 and 35, the following relation is obtained,

1 − ν2
b3 = 0.75R F (36)
E
3 1 − ν2
δ1 = F (37)
8b E

4.1 EHL in FDM

The deflection can be incorporated in the finite difference method as [2],

1 − ν2
Z Z
p(x, y)dxdy
δ1 = p (38)
πE A x2 + y 2
 
n m
1 − ν 2 X X p(i, j)dxdy
δl,k = p  (39)
πE i=1 j=1
(i − l)2 dx2 + (j − k)2 dy 2

From the equation 39 it is observed, pressure at one node affects the entire contour. During the summa-
tion for the deflection at the specified node, the pressure at the node under consideration is avoided as
this will lead to an infinite value. Therefore the total deflection at any node is,

hi,j = hhydrodynamic + δelastic

14
However, this cannot be incorporated directly into the reynolds equation. In the finite difference method
the film thickness must be evaluated at each node instead of half node and updated after every iteration.

0.75(h̄i+1,j − h̄i−1,j ) + h̄i,j h̄i,j − 0.75(h̄i+1,j − h̄i−1,j )

p̄i,j = X2
p̄i+1,j + 2
p̄i−1,j
2h̄i,j 1 + X


2h̄i,j 1 + Y 2 Y 2
 2 "
X 0.75(h̄i,j+1 − h̄i,j ) + h̄i,j
+ 2
p̄i,j+1 (40)
Y 2h̄i,j 1 + X Y2
#
h̄i,j − 0.75(h̄i,j+1 − h̄i,j−1 ) C∆x (h̄i−1,j − h̄i+1,j )
+ X2
p̄i,j−1 + 2

2X h̄3i,j 1 + X


2h̄i,j 1 + Y 2 Y2

Using the equation 40 the pressure and deflection is evaluated at every node and iteration. The pressure
distribution for the specification as stated in the viscosity section is used, with E = 200 GPa. The results
are shown in the figure 19 and figure 20

Figure 19: Elastic deformation

As observed from the deformation profile the surfaces under consideration undergo some amount of
deformation, in turn effecting the pressure distribution. From the pressure distribution curve 20 it is
observed to converge to a more uniform pressure distribution as compared to figure 15.

15
 Figure 20: Pressure distribution with elastic deformation

5 Application in journal bearing

The finite difference method with viscosity variation is applied to a journal bearing bearing. In a journal
bearing the journal rotates to bring the lubricant into a convergent gap which produces the hydrodynamic
pressure. The pressure produced help to balance the load which acts on it. A steady state position of
the journal inside a bearing is shown in the figure. The journal center is determined by the deviation
angle ψ and eccentricity e. The eccentricity ratio is defined as  = e/c, where c is the radial clearance
and is given by c = R1 - R2. R1 is the radius of the bearing and R2 is the radius of the journal. From
the figure 21 it can be seen that [1],
e R1
=
sinα sinθ
2
e2
 
2 e
cosα = 1 − sin2 α = 1 − 2 sin2 θ = 1 − sin2 θ + ...


R1 2R12
If e/R1 ≤≤ 1, it can be omitted. Therefore,

h = ecosθ + c

The Reynolds equation for the journal is modified as,

 3 
∂ h ∂p h3 ∂ 2 p ∂h
2
+ 2
=ω (41)
R ∂θ 6η∂θ 6η ∂y ∂θ

Example case: The journal bearing Reynolds equation is solved using finite difference method. A
bearing of diameter 200 mm and length 160 mm is rotating at 1200 rpm. It has an eccentricity ratio of
0.7 with a radial clearance of 0.18 mm.

16
 Figure 21: Journal bearing
[1]

Figure 22: Pressure distribution in journal bearing

17
6 Conclusion and future work
The above numerical numerical strategies can be adopted for various kinds of bearing geometry. The
finite difference method is a generic solution method. The initial conditions to evaluate the pressure
distribution needs to be modified for various geometries for the solution. The pressure distribution with
viscosity variations only considers the effect of pressure and no effect of temperature. Therefore the
accuracy of the method is limited. Some of the future work could include,
1. Incorporation of temperature effect in calculation of pressure distribution
2. Incorporation of mass convering algorithm to determine the true distribution of pressure.

18
References
[1] Ping Huang. Numerical calculation of lubrication methods and programs.
[2] Prof. Dr. Harish Hirani. Tribology lecture series.

19