ARTICLE IN PRESS

Tribology International 42 (2009) 14331438

Contents lists available at ScienceDirect

Tribology International
journal homepage: www.elsevier.com/locate/triboint

Squeeze process under impact, in highly compressible porous layers,

imbibed with liquids
Mircea D. Pascovici , Traian Cicone, Victor Marian
University POLITEHNICA of Bucharest, Spl. Independentei 313, Bucharest 060042, Romania

a r t i c l e in fo

abstract

Article history:
Received 8 July 2008
Received in revised form
23 December 2008
Accepted 7 May 2009
Available online 15 May 2009

The paper proposes a model for the squeeze process under impact for highly compressible porous layers
imbibed with uids (HCPL). It is assumed that the normal forces generated by elastic compression of the
bers comprising the solid phase are negligible compared to the pressure forces generated in the
imbibed uid, within the porous layer. The process, named ex-poro-hydrodynamic (XPHD) lubrication,
is strongly dependent on porosity variation and consequently on permeability variation. Closed form
analytical solutions are given for circular and rectangular contacts with various boundary conditions.
Numerical applications show the existence of optimum initial compacticity, s0, that minimizes the
maximum impact force (i.e. the maximum damping effect). These interesting results can be useful for
the design of squeeze dampers and shock absorbers based on XPHD lubrication.
& 2009 Elsevier Ltd. All rights reserved.

Keywords:
Hydrodynamic lubrication
Porous medium
Compliant layer
Impact loading

1. Introduction

2. Problem formulation and main assumptions

During the last 1015 years, a new lubrication mechanism

applicable to highly compressible porous layers imbibed with
uids (HCPL) acting as self-sustained lms has been developed
[14]. This new type of lubrication requires that the normal forces
generated by elastic compression of the bers comprising the
solid phase are negligible compared to the pressure forces
generated in the imbibed uid, within the porous layer. The
porous layer can be represented by

Fluid ow through a porous media can be approximated by

Brinkmans (momentum) equation [10], which has the following
general form:


1
rp Z r2  um
(1)

 the unwoven textile materials, as felt, or similar materials used

as wash-cloth [1,5,6];

 the articular cartilage [1,7];

 the endothelial surface glycocolyx that coats the microvessels



[6,9];
the fresh powder snow [1,2];
the goose down [4].

This type of lubrication is strongly dependent on porosity variation

and consequently on permeability variation hence, the name of exporo-hydrodynamic (XPHD) lubrication was proposed [3].
The present paper analyzes the squeeze process under impact
in XPHD conditions, for two different congurations: circular,
respectively rectangular. The results suggest the possibility to use
such processes for squeeze dampers and shock absorbers.
 Corresponding author.

E-mail addresses: mircea@omtr.pub.ro, mdpascovici@yahoo.com (M.D. Pascovici).

0301-679X/$ - see front matter & 2009 Elsevier Ltd. All rights reserved.
doi:10.1016/j.triboint.2009.05.006

where um is the mean uid velocity through the channels in

porous media.
For low permeability, f, this equation becomes an alternative
of Darcy law [10]:

rp 

Zqp
fh

(2)

where qp umh is the rate of ow in a porous media of thickness h.

The geometry and kinematics of a friction joint containing
HCPL determine a variation in space and/or in time of the
thickness and porosity of the layer. This variation implies a
variation of permeability, also in space and time. These variations
are related essentially with the occurrence of a self-sustained
pressure distribution, i.e. XPHD lubrication mode.
The analytical modeling of this self-sustained lm requires the
following assumptions:
1. The uid imbibed in the porous layer is Newtonian and the
ow is laminar and isothermal/isoviscous.
2. The porous layer is homogenous, isotropic and relatively thin;
consequently the pressure across its thickness is constant.
3. The local deformation is produced only in the normal direction
of the porous layer. For the sake of clarity the term of local

ARTICLE IN PRESS
1434

M.D. Pascovici et al. / Tribology International 42 (2009) 14331438

Nomenclature
a, b
b
d
D
f
F
Fs
F s
g
h
H
Hf
k
M

M
p
r
R
Re

rectangular plate Ralf dimensions

rectangular plate aspect ratio, b/a
ber diameter of HCPL
complex parameter in the KozenyCarman equation
form factor
induced load for constant velocity
load at impact (shock) squeeze
dimensionless load, h0 D=ZV 0 S2 F s
gravity acceleration
layer thickness
dimensionless layer thickness h/h0
height of fall
dimensionless constant in KozenyCarman equation
mass of impact
dimensionless impact load, MDV 0 =ZS2
pressure
radial coordinate
radius of the disc
outer diameter of the porous domain

compacticity, s, is introduced s 1  , i.e. the instantaneous

solid fraction. The solid mass is conserved throughout the layer
during deformation process. This assumption was used in all
the previous papers dealing with this subject [3,5,8,11]:

sh s0 h0

(3)

where s0 and h0 are the uncompressed compacticity (solid

fraction) and layer thickness.
4. In the previous papers published by Pascovici [3,5,9] as well in
the present analysis, the local permeability is related to
porosity according to KozenyCarman law [10]:
D

1  2

dimensionless outer diameter, Re/R

upper plate area
time
squeeze/normal velocity
axial coordinates
dimensionless axial coordinates, x x=a; z z=b

Greek letters

e
Z
s

porosity
viscosity
compacticity
permeability

Subscripts
c
r
0

corresponding to circular conguration

corresponding to rectangular conguration
initial (corresponding to undeformed layer)

Consequently, two geometrical cases, with particular boundary

conditions will be considered: extended model that assumes Re4R
and restrained model that assumes Re R.
The ow conservation equation for a given normal velocity,
V, can be expressed function of HCPL properties as


2pr fh dp
pr 2 V
Z dr

(5)

for internal area and



2pr f0 h0 dp
pR2 V
Z
dr

(6)

R e
S
t
V
x, z
x ; z

D1  s

s2

(4)

where D d2/16k and k 5C10.

Recently, Hakanson et al. [6] have demonstrated, for the case
of anisotropic ber web, the complexity of permeability
variation in function of compacticity, s, and consequently,
the limits of KozenyCarman equation. However, based on the
validity of this law, proved for 2D porous structures and its
possible extension for 3D structures [12], we decided to keep
this convenient form in order to obtain closed form analytical
solutions.
5. The behavior of HCPL in impact squeeze conditions is
described using the component properties measured in static
conditions. This assumption was also used by Bowden and
Tabor [13].
6. During the squeeze process the rigid plates remain parallel so
that the thickness of the porous layer is constant in space and
varies only in time.

for external area.

Introducing Eqs. (3) and (4), in Eqs. (5) and (6) we get
dp
Zrs2 V

dr
2D1  s2 h

(7)

for internal area and

ZR2 s20 1  sV
dp

dr
2D1  s0 3 h0 r

(8)

for external area.

Eqs. (7) and (8) represent the equivalent Reynolds equations of
the squeeze problem in XPHD lubrication conditions.

M
Hf

V0
V

3. The model for circular plates

Fig. 1 shows the schematic of a circular conguration. The
impact is produced by dropping a mass M from an initial height Hf
on a circular weightless rigid plate (disc), 4, in contact with
HCPL, 2, xed on a rigid, circular plate, 1, with a diameter, Re,
grater than that of the upper plate, 4. The external surface of HCPL
is protected with a thin impermeable membrane, 3.
The circular area dened by r  R will be named internal area,
and the annular area (for Ror  Re ) will be named external area.

h0
2

h
pb

r
R

Re

p(r)
Fig. 1. Geometry of the circular conguration under impact loading.

ARTICLE IN PRESS
M.D. Pascovici et al. / Tribology International 42 (2009) 14331438

Assuming zero pressure on the outer edge of the porous layer

(p 0 for r Re), and considering pressure continuity at the edges
of the upper disc (p pb for r R), after integration and some
algebra, we obtain the pressure distribution on each area:
"
#
Zs20 V R2 H  s0 Re
R2  r 2
(9)

ln
p
2DHh0 1  s0 3
R 2H  s0 2
for internal area r  R and
p

ns20 V R2 H  s0
Re
ln
2DHh0 1  s0 3
r

(10)

for external area Ror  Re .

The force acting on the moving part, 4, is obtained by
integrating the pressure distributionEq. (9):
"
#
Z R
pZs20 R4 4H  s0 Re
1
V
ln

pr dr
(11)
F 2p
2 h
8D
s

R
H1

HH  s0
0
0
0
For the analysis of the squeeze process under impact/shock
conditions the method proposed by Bowden and Tabor [13] will
be used. It consists of the impulse conservation application:
M dv F dt

(12)

Noting that V=h0 dH=dt and V V0, for H 1, Eq. (12) can be
easily integrated to get the velocity variation of the upper disc
with the dropping mass, M, on it:
"
4s20
pZR4
Re
 1  H s0 ln H
ln
V V0 
8DM
R
H1  s0 3

H  s0
s0 1  H

(13)
ln
H1  s0 H  s0 1  s0
p
where V 0 2gHf is the dropping mass speed at the contact
with the upper plate. Note that we assume a perfectly plastic
contact between the dropping mass and the upper plate.
Combining Eqs. (11) and (13), the variation of the dimensionless impact force, F sc , versus dimensionless HCPL thickness H, for a
 c MDV 0 =ZS2 , is obtained
given dimensionless mass M
c
" 
#

2
s0
Dh0
H  s0 3

F sc
F

4
ln
R

1
e
1  s0
ZV 0 s2c sc 8pHH  s0 2
"

4s20
1
1  H s0 ln H
ln R e
 1

8pM c
1  s0 3

H  s0
s0 1  H

(14)
ln
H1  s0 H  s0 1  s0
By making R e ! 1 in Eq. (14) the particular form of the impact
load equation for restrained model yields


s20
1
H  s0
1

ln
F sc
c
H1  s0
8pM
8pHH  s0 2

s0 1  H
(15)

H  s0 1  s0

1435

Rearranging the terms, we obtain theequivalent Reynolds equations of squeeze problem in XPHD lubrication conditions for
internal area (zone). Using Eqs. (3) and (4), Eq. (16) becomes

s20
@2 p @2 p
ZV


Dh0 HH  s0 2
@x2 @z2

(17)

In dimensionless form, Eq. (17) takes the classical form of Poisson

partial differential equation:
b

@2 p
@x

1 @2 p
b @z2

(18)

1

where
p

4h0 D HH  s0 2
p
ZVSr
s20

According to the particular assumption for rectangular conguration, zero pressure boundary conditions on the edges of the plate
are written as
p0

at x 1 and z 1

Using the separation of variables, the solution of the equation can

be written as an innite series [14]:
i
8 h
1 <sin 2k1p x 1
2
1  x2
16 X

p
2k  13
2b
p3 b k1 :
h
i9
p =
cosh b 2k1
z
2
h
i

(19)
cosh b 2k1p ;
2

The dimensionless induced load for given velocity, V, results by

simple analytical integration:
Z 1Z 1
16h0 DHH  s0 2

 x ; z dx dz 4
Fr 4
F r
p
2
2
s0
nVSr
0
0
3b


 2k  1p
1 tanh b
256 X
2

2
2k  15
p5 b k1


A simple parametric analysis shows that for any value b41
this
series converges rapidly. If only the rst term is retained, the

dx
x
O
z

dz
4. The model for rectangular plates
The analysis of the rectangular conguration is similar to that
of the circular one. However, in order to obtain an analytical
solution, only the case of restrained conditions will be analyzed.
The schematic of the rectangular plate is presented in Fig. 2.
The ow conservation inside the internal area, for a given
normal velocity, V, is


fh @p
@p
(16)
dz
dx V dx dz

Z @x
@z

(20)

a
Fig. 2. The geometry of the rectangular plate.

ARTICLE IN PRESS

Fr

M.D. Pascovici et al. / Tribology International 42 (2009) 14331438

Fr (b = 1)

1436

1.0

general form:

0.9

F s

0.8

f s20
2

HH  s0

1

 


f
H  s0
s0 1  H

ln

H1  s0
1  s0 H  s0
M
(25)

0.7
0.6
0.5
0

b = b/a

 depending
where the particular forms of the variables S; F s ; f ; M
on the plate geometry are synthetically presented in Table 1.
It is worth noting that in the particular case of a narrow

rectangular plate (for b410)
the following approximate expression results:
fr

Fig. 3. Impact load variation with aspect ratio of rectangular plates.

errors are less than 1%, with a maximum for

p

4
256

b
tanh

F r
2
2
3b p5 b

(21)


Note that the symmetrical congurations, dened by b and 1=b,
have the same pressure distribution and load capacity, so that

only the case b41
will be considered from now on.

The variation of dimensionless load capacity, F r with the

rectangle aspect ratio, b, is presented in Fig. 3. One can remark
that the maximum reacting force is obtained for a square shape
(b a), as it should be expected. It is also worth to be noted that
for an aspect ratio, b 4, which we consider as a limit for some
practical cases, load capacity is reduced at half of its maximum
value.
Following the same analytical procedure like in the case of
circular plate, we have successively:
the equation of impulse conservation:
M dV F dt

f r ZS2r s20
HDH  s0 2

dH

(26)

Consequently, it is possible to compare the behavior of the circular

and rectangular congurations in dimensionless form, remarking
that the results correspond to the case of equal area: Sc Sr.
This comparison is shown in Fig. 4, for a typical value
 0:2. Three cases are
of dimensionless dropping mass, M
plotted in parallel:

 circular conguration with restrained boundary conditions

(HCPL only under the circular plate);

 circular plate with extended boundary conditions (assuming

the outer diameter of the HCPL ve times greater than the
upper plate diameter, R e 5);
square plate with restrained boundary conditions.

Note that the force evolution in time must be followed from right
to left (initially, HCPL layer is undeformed and correspondingly,
lm thickness is maximum, i.e. H 1). Obviously, the maximum
deformation is limited by the condition H s0, that is the solid
Table 1
Particular equations for circular and rectangular plate.

(22)


where f r F r =16 is a modied form of the dimensionless induced
load.
The normal speed during impact:
 


f ZS2
H  s0
s0 1  H

V V 0  r r ln
MD
H1  s0
1  s0 H  s0

1
12b

(23)

the dimensionless impact force, F sr for the rectangular plate

(combining Eqs. (20) and (23))

 


f r s20
fr
H  s0
s0 1  H

1

ln
F sr
r
H1  s0
1  s0 H  s0
M
HH  s0 2
(24)

Circular plate (analytical, exact

solution)
S Sc pR2
F s Dh0
F sc
ZV 0 S2c
f 1
8p
 MDV 0
M

Dh0

ZV 0 S2r

F sr

16
tanh b

2

2

12b p5 b
MDV 0
1

ZS2r

Circular plate - restrained Re = 1

Circular plate - extended Re = 5
Square plate - restrained b = 1

0 = 0.2

M =

Dh0

0 = 0.4

MDV0

 S2

= 0.2

0 = 0.8

0 = 0.6

Fs =

VS2

Fs

1.5

In the previous two chapters analytical forms for force

variation after impact during the damping process (during
deceleration) have been developed. This variation is governed by
three parameters: lm thickness ratio, H, initial compacticity of

the HCPL, s0, and dimensionless dropping mass, M.
If we analyze in parallel the two forms of dimensionless impact
force (for circular, respectively, rectangular conguration), one
can remark that they are identically non-dimensionlized, as
the expression of the area of the upper part is not explicitly
given. Hence, Eqs. (15) and (24) can be written in the following

Sr 4ab

ZS2c

where, obviously, the dimensionless mass is calculated with the

area of the rectangle, Sr.

5. Numerical results and discussion

Rectangular plate (analytical, series

approximation)

0.5

0
0.2

0.3

0.4

0.5

0.6
0.7
H0 = h/h0

0.8

0.9
t

1
0

Fig. 4. Impact load versus lm thickness for circular and square congurations.

ARTICLE IN PRESS
M.D. Pascovici et al. / Tribology International 42 (2009) 14331438

104

Dh0
F
VSc2 sc
Fsc =

Rectangular configuration - restrained case

0 = 0.2

2.5
Dh0
F
VSr2 sr

M=

MDV0
 S2r

= 0.2

Rectangular plate b = 4

1.5

Rectangular plate b = 2

0 = 0.6

Square plate b = 1

0.5
0
0.2

0.3

0.4

0.5

0.6
H = h/h0

0.7

0.8

0.9
t

1
0

 on dimensionless impact load, F s .

Fig. 7. The effect of rectangle aspect ratio, b,

Finally, Fig. 7 shows the variation of the impact load with

dimensionless lm thickness for three rectangular congurations.
As expected, the maximum impact load for square plate has
a lower value than the corresponding values for the other two
rectangular plates. The effect is more pronounced for low
compacticity.

Circular plate - restrained case

M = 20

103

6. Conclusions
M=2

102
0 = 0.2

10

0 = 0.6

M = 0.2

M = 0.02

10-1
10-2
0

0.2

0.4

0.6
H = h/h0

0.8

1
0

Fig. 5. Impact force versus lm thickness for circular congurations with

restrained boundary conditions.

Circular plate - restrained case

102

M=2

10

M = 0.2

10-1

Analytical models for impact load variation with lm thickness

under squeeze conditions have been developed for circular,
respectively rectangular, plates. For the circular plates a generalized form of the analytical model for the case extended porous
layer has also been obtained.
The parametric analysis reveals an optimum initial compacticity for minimizing the maximum squeeze induced force (i.e. the
maximum damping effect). This optimum depends strongly on
the dimensionless impact mass.
It is demonstrated that, for the rectangular plates, the square
minimizes the maximum squeeze induced load.
Preliminary experiments for impact loaded squeeze of HCPL
have been started in our laboratory but rst results obtained
cannot be used yet for comparison with the predicted theoretical
data. Unfortunately, no experimental data have been found in the
literature to validate these models. However, the experiments
reported by Wu et al. [4] despite some differences in geometry of
the porous layer and in measured parameters support qualitatively our models (see Fig. 4 from [4]).
For the future a development of these models for the case of
nonparallel plates is envisaged.

Acknowledgments

Fsc =

Dh0
F
VSc2 sc

Fsr =

fraction equals the volume of HCPL under the moving plate (nil
porosity).
From this gure it results that the circular and square
congurations, under the restrained conditions, lead to comparable maximum impact forces. The square conguration produces
greater impact forces than circular congurations for low initial
compacticity, s0. Targeting the damping effect we are looking for
minimizing the maximum of the impact force. The case of HCPL
with extended boundary conditions produces most favorable
damping effects, as shown in Fig. 4.
Fig. 5 shows the variation of the impact load function of lm
thickness, for two typical values of the initial compacticity for
circular plates, under restrained boundary conditions. A series
 covering the
of four values for dimensionless dropping mass, M,
most probable practical cases are considered. The difference
between maximum impact forces for the two compacticities is
very sensible to the dropping mass and changes the sign.
At the same time, one can remark the existence of optimum
initial compacticity, s0, that gives maximum damping characteristics. The optimum locations, in the case of circular plates, under
restrained boundary conditions, are presented in Fig. 6, for three
different dropping masses. One can remark again the sensitivity of
the minimum impact force and its location, with the dropping

mass, M.

1437

Financial support for the work described in this paper was

provided by National Council for Academic Scientic Research
(CNCSIS) under Grant ID_912/2007.

M = 0.02

References
10-2
0

0.2

0.4

0

0.6

0.8

Fig. 6. The geometric loci of optimum initial compacticity for circular congurations with restrained boundary conditions.

[1] Pascovici MD. Procedure and device for pumping by uid dislocation.
Romanian Patent 109469; 1994 [in Romanian].
[2] Feng J, Weinbaum S. Lubrication theory in highly compressible porous media:
the mechanics of skiing from red cells to humans. J Fluid Mech 2000;422:
281317.

ARTICLE IN PRESS
1438

M.D. Pascovici et al. / Tribology International 42 (2009) 14331438

[3] Pascovici MD. Lubrication by dislocation: a new mechanism for load carrying
capacity. In: Proceedings of second world tribology congress, Vienna, 2001.
p. 41 [paper on CD].
[4] Wu Q, Andreapoulos Y, Weinbaum S. From red cells to snowboarding: a new
concept for a train track. Phys Rev Lett 2004;93(19):194501-14.
[5] Pascovici MD. Lubrication processes in highly compressible, porous layers. In:
16emes Journees Francophones de Tribologie, JFT 2007, 2223 May, Poitiers,
FranceInvited paper; 2009, [to be published].
[6] Hakanson JM, Toll S, Lundstrom TS. Liquid permeability of an anisotropic ber
web. Text Res J 2005;75(4):30411.
[7] McCutchen CW. The frictional properties of animal joints. Wear 1962;5:117.
[8] Han Y, Weinbaum S, Spaan JAE, Vink H. Large-deformation analysis of the
elastic recoil of ber layers in a Brinkman medium with application to the
endothelial glycocalyx. J Fluid Mech 2006;554:21735.

[9] Pascovici MD. Lubrication of red blood cells in narrow capillaries. A heuristic
approach. In: Proceedings of second Vienna international conference on
micro- and nano-technology,Viennano 07, March 1416, Austria, 2007.
p. 95100 [on CD].
[10] Scheidegger AE. The physics of ow through porous media. 3rd ed. New York:
University of Toronto Press; 1974.
[11] Pascovici MD, Cicone T. Squeeze-lm of unconformal compliant and layered
contacts. Tribology Int 2003;36:7919.
[12] Ghaddar CK. On the permeability of unidirectional brous media. A parallel
computational approach. Phys Fluids 1995;7(11):256386.
[13] Bowden FP, Tabor D. The friction and lubrication of solids. Oxford: Clarendon
Press; 1950. p. 25984.
[14] Haberman R. Elementary applied partial differential equations. 2nd ed.
London: Prentice-Hall; 1987.