You are on page 1of 9

Food Hydrocolloids 20 (2006) 483–491

www.elsevier.com/locate/foodhyd

Soft lubrication of model hydrocolloids


J. de Vicentea,b,1, J.R. Stokesc, H.A. Spikesb,*
a
Food Research Centre, Unilever R&D Colworth, Colworth House, Sharnbrook MK44 1LQ, UK
b
Tribology Section, Department of Mechanical Engineering, Imperial College London, Exhibition Road, London SW7 2BX, UK
c
Unilever Corporate Research, Unilever R&D Colworth, Colworth House, Sharnbrook MK44 1LQ, UK
Received 21 October 2004; revised 3 March 2005; accepted 19 April 2005

Abstract
Many food products are highly structured, complex hydrocolloids that are characterized by either a gel-like or shear thinning behaviour
and are thus strongly non-Newtonian systems. Their sensory perception during use is dependent on their behaviour in very thin film
conditions at high shear rates. In order to better understand their behaviour in thin films, we have formulated two kinds of solution which are
typically present in real systems. We examine the lubricating properties (i.e. tribology) of thin films of structured and non-structured fluids
between a hard and soft surface (‘soft-EHL’ lubrication), when the film thickness is of a similar size to the microstructural elements. The
present work involves relating the tribology of both shear thinning polymer solutions and swollen microgel suspensions to their rheological
and microstructural properties in thin films and at high shear rates in the boundary, mixed and fluid film lubrication regimes.
We show that the polymer solutions are typically entrained into rubbing contacts to form mixed fluid/boundary lubricating films while
model yield stress fluids generate only boundary lubricating films. Soft-EHL theory satisfactorily predicts friction measurements at high
entrainment speeds in full film lubrication. The microgels are found to form a confined film which minimizes contact between the surfaces,
causing a lowering of the friction coefficient in the boundary-regime, while the friction in the mixed-regime is also dependent on the high-
shear viscosity.
q 2005 Elsevier Ltd. All rights reserved.

Keywords: Tribology; Lubrication; Soft contacts; Isoviscous-elastic; Carbopol

1. Introduction astringency. While initial thickness/viscosity perception


may be governed by relatively low shear rates in the mouth
The consumption and use of food, beverages and oral (e.g. 10–50 sK1), taste perception is expected to be
care products in the oral cavity are comprised of a complex influenced by the high shear rates experienced by the
set of operations from ingestion and mastication to material as it is confined to very thin films and mass transfer
swallowing. During this process, fluids and soft solids are near the oral surfaces. We investigate the process of nano-
confined to very thin films between rubbing, deformable and micro-scale confinement and lubrication in order to
surfaces and their response to this influences the sensory understand the behaviour of complex materials during oral
perceptions of ‘mouthfeel’, texture and taste. Malone, processing.
Appelqvist, and Norton (2003) reviewed the influence of The lubrication properties of food-related products,
lubrication properties of hydrocolloids and emulsions on ranging from chocolate to emulsions and biopolymer
oral percepts and found that these strongly influenced solutions, have been previously investigated using modified
sensory properties such as fattiness, smoothness, and tribometers and the surface forces apparatus. The behaviour
of such materials in lubrication processes is found to be
* Corresponding author. Tel.: C44 20 7594 7063/7064; fax: C44 20
governed by the bulk rheology, the surface properties of the
7823 8845. constituents, and the confinement or exclusion of micro-
E-mail address: h.spikes@imperial.ac.uk (H.A. Spikes). structural components from the contact zone. Kokini (1987)
1
Present address: Department of Applied Physics, Faculty of Sciences, highlighted the role played by friction forces and lubrication
University of Granada, C/Fuentenueva s/n, 18071 Granada, Spain. processes in foods such as whole milk, cream and
0268-005X/$ - see front matter q 2005 Elsevier Ltd. All rights reserved. butter. Lee, Heuberger, Rousset, and Spencer (2004) used
doi:10.1016/j.foodhyd.2005.04.005 a pin-on-disc tribometer to study molten chocolate samples
484 J. de Vicente et al. / Food Hydrocolloids 20 (2006) 483–491

and found that the latters’ lubrication properties are All load All load
determined by particle behaviour in the inlet region of the supported supported by
sliding contact. Luengo, Tsuchiya, Heuberger, and Israe- at asperity fluid film
lachvili (1997) used the surface forces apparatus to show contact pressure
that fat constitution and average particle size are important
in determining the type of friction force generated between
rubbing surfaces separated by thin films of chocolate.

log(friction coefficient)
Giasson, Israelachvili, and Yoshizawa (1997) correlated
tribological properties with composition and texture of
mayonnaise samples. The thin-film rheology and friction of
thin mayonnaise films between two shearing surfaces
appeared to depend mainly on the particle properties, such
as size, stiffness and hydrophobicity. Cassin, Heinrich, and
BL mixed hydrodynamic
Spikes (2001) studied the lubrication properties of pig
gastric mucin and guar gum solutions in a simulated
biotribological contact consisting of a rolling, hard ball on a log(film thickness)
soft flat surface, and showed that these fluids significantly
reduced friction. Mucin, a major component of saliva, in Fig. 1. Typical Stribeck curve as a function of entrainment speed and film
thickness.
particular was shown to be a strong boundary lubricant.
Using the same apparatus, the friction in the mixed-
(EHL) regime, a film of lubricant, whose thickness depends
lubrication regime was correlated by Malone et al. (2003)
with sensory perception of slipperiness for guar solutions on the viscosity and entrainment speed, is entrained to fully
and perceived fattiness for a complex emulsion system. separate the solid surfaces. The friction now depends on the
However, several difficulties have been found when trying rheological properties of the lubricant film in the contact, at
to accurately simulate the biotribological behaviour of the high shear rate condition which prevail there. In the
complex materials. These include; not including a signifi- mixed lubrication regime, which lies between boundary
cantly deformable surface, not accounting for the influence lubrication and EHL, both the boundary film and bulk
of either the bulk or thin-film rheological properties in the lubricant play a role in determining friction. In terms of the
contact zone, and difficulties in interpreting results due to in-use process of food products, the initial behaviour of the
the complex multiphase materials utilised. material may be governed by its bulk rheological properties
In this work we study the tribological performance of when the product film between surfaces is relatively thick.
contacts between a silicone elastic surface and steel (a ‘soft’ However, as the product is sheared and broken down to a
contact which should operate in the isoviscous-elastic much thinner film during consumption, it can be envisaged
lubrication regime) lubricated by aqueous solutions of that the mixed regime is entered. Once the product is cleared
model hydrocolloids which have non-Newtonian rheologi- from the oral cavity, surface residue may control the
cal properties. Using this approach, we expect to gain a boundary lubrication. Thus we might expect to move from
better understanding of the oral behaviour of real food right to left along the friction curve shown in Fig. 1.
hydrocolloids. In the isoviscous-elastic (soft-EHL) regime, Throughout the process, the interaction of the product with
the elastic deformation of one or both of the interacting saliva should also be considered but for simplicity this is
solids has a significant effect on the thickness of the fluid largely ignored in the current work
film separating them, but the pressure within the contact is We explore the influence of several very different types
quite low and insufficient to cause any substantial increase of hydrocolloids on the three lubrication regimes. Foods
in fluid viscosity (Esfahanian & Hamrock, 1991). We adapt usually contain a multitude of surface active ingredients,
a tribometer to measure the friction for a soft contact and consist of multiple phases and have complex microstruc-
explore the influence of rheology and microstructure over tures, all of which can contribute to the rheological and
the nano- to micron-scale on the frictional response. tribological properties at different time and length scales.
Lubrication properties are classically represented in the The rheology itself contributes to the tribological properties,
form of a Stribeck curve, where the coefficient of friction is so it is useful to investigate the influence of such non-
given either as a function of entrainment speed or film Newtonian characteristics as shear-thinning and yield stress
thickness, as shown in Fig. 1. The frictional behaviour is or gel-like rheology. We, therefore, seek to examine the
usually divided into several regimes. In the boundary behaviour of different hydrocolloids in isolation as a prelude
lubrication regime, which occurs at slow speeds when there to examining more complex food systems.
is negligible fluid entrainment into the contact, the load is The first part of the paper examines the Stribeck curve for
carried by the contacting asperities and is dependent on the Newtonian liquids in a soft-EHL contact. The results are
surface and interfacial film properties at the molecular scale. compared to those derived theoretically and measured
In the hydrodynamic or elastohydrodynamic lubrication previously on rubber surfaces. The second part examines
J. de Vicente et al. / Food Hydrocolloids 20 (2006) 483–491 485

the Stribeck curves for two sets of shear-thinning polymer Xanthan gum (XG) is used in many thickening
solutions containing polymeric food hydrocolloids, xanthan applications, for example in salad dressings. It is an
gum and guar gum, such that the effect of shear-thinning in exocellular polysaccharide and may be considered to be
the mixed-regime can be examined and compared to the an anionic polyelectrolyte. The molecule has a cellulosic
Newtonian results. The final section examines the effect of backbone which is rendered water-soluble by the presence
microgel particle suspensions on the Stribeck curve. These of short side chains attached to every second glucose residue
microgels, commercially known as Carbopol, swell at in the main chain. As a consequence of this structure,
neutral pH to create a model gel-like material which xanthan typically exists in an ordered conformation in
forms non-thixotropic, highly shear-thinning (or yield solution having a semi-rigid rod-like structure (Norton,
stress) fluids. Goodall, Frangou, Morris, & Rees, 1984). The polymer used
in this study is a Keltrol RD sample obtained from CP
Kelco, Wilmington, DE, USA. The molecular weight of
2. Test materials and rheology Keltrol xanthan gum is reported to be in the range 2–7!106
by Zirnsak, Boger, and Tirtaatmadja (1999).
Three sets of materials are examined: Guar gum (GG) is a high molecular weight (approx.
1–1.5!106) carbohydrate made from galactomannan units.
– Newtonian fluids, aqueous sugar solutions. The specific polysaccharide component of guar gum is
– Shear-thinning fluids, aqueous polymer solutions. guaran, a galactomannan where about one-half of the b-D-
– Gel-like or yield stress fluids, soft-microgel mannopyranosyl main-chain units contain an a-D-galacto-
suspensions. pyranosyl side chain (Frollini, Reed, Milas, & Rinaudo,
1995). Guar is a long-chain, linear molecule. In this work,
the guar used is a THI quality sample obtained from
2.1. Newtonian fluids Hercules Inc., Kennedy, TX, USA.
Both xanthan and guar solutions were prepared by first
Corn syrup/water mixtures were used to produce a dissolving the appropriate amount of polymer into Millipore
variety of Newtonian fluids with viscosities ranging from deionised water. The polymer was added gradually to the
7.5!10K4 to 3.12 Pa s (Table 1). These fluids were water while continually swirling the container, such that the
Newtonian at shear rates between 1 and 102 sK1, as polymer was dispersed throughout the solution and did not
indicated by measurements in a controlled strain ARES associate into clumps. 0.02% w/w NaN3 (sodium azide) was
rheometer (Rheometric Scientific, Piscataway, NJ, USA). added to the stock solutions to act as a biocide.
The Carbopol test fluids studied are aqueous suspensions
of Carbopol C934 microgels (B.F. Goodrich, USA). These
2.2. Hydrocolloids
hydrophilic microgels are a highly cross-linked anionic
polyacrylic acid that swells upon neutralisation to form
Solutions of three polymers are investigated in this work:
electrically charged particles of approximately 2–4 mm
xanthan gum, guar gum and cross-linked polyacrylic acid.
diameter (de Vicente, Spikes, & Stokes, 2004a). Suspen-
Xanthan gum is a high molecular weight, food grade
sions were prepared by dilution with water from a stock
polysaccharide with a rigid conformation. Guar is a food
solution to 0.1, 0.15, 0.2, and 0.5 wt% (unneutralised,
grade, high molecular weight carbohydrate. Cross-linked
pHw3), and to neutral pH with 0.15, 0.2, 0.3 and 1 wt%
polyacrylic acid, commercially known as Carbopol, consists
triethanolamine, respectively.
of a suspension of microgel particles that swell upon
neutralisation and is commonly used as a bioadhesive for
pharmaceutical applications. While these are not necessarily 2.3. Rheology of test fluids
all food grade products, they have been chosen as models to
investigate the effect of their physical properties and The results described in this paper were obtained at 35 8C
microstructure on soft lubrication. using a controlled-strain Rheometrics ARES-LS rheometer
and a controlled-stress Haake RS1 rheometer, primarily
Table 1 with a 50 mm, 0.04 rad cone-and-plate configuration. For
Viscosity of corn syrup water mixtures at 308C some of the Carbopol suspensions, 50 mm parallel plates
Corn syrup content (wt%) Viscosity (Pa s) were covered with rough emery paper to avoid slip effects.
The rheological properties of xanthan gum and guar gum
0 0.00075
50 0.007 are shown in Fig. 2. As can be seen, both polymer solutions
72 0.032 are strongly shear thinning.
86 0.23 Rheological properties of Carbopol are shown in Fig. 3.
92 0.90 The unneutralised (UN) solutions have essentially constant,
93 1.14
low viscosity, while the neutralised (N) solutions are highly
95 3.12
shear-thinning. Concentrations above 0.1% have an
486 J. de Vicente et al. / Food Hydrocolloids 20 (2006) 483–491

(a) (a)
Water Water
100 0.005 wt %
103 0.1 wt % UN
0.15 wt % UN
0.02 wt % 0.2 wt % UN
0.07 wt % 102 0.5 wt % UN
0.2 wt % 0.1 wt % N
Viscosity (Pa s)

Viscosity (Pa s)
0.15 wt % N
10-1 101 0.2 wt % N
0.5 wt % N
100

10-2 10-1

10-2
Xanthan gum Carbopol
10-3 10-3
10-3 10-2 10-1 100 101 102 103 104 105 10-3 10-2 10-1 100 101 102 103 104
Shear rate (s-1) Shear rate (s-1)
(b)

Storage (G’) and loss (G’’) modulus (Pa)


0
(b) 10 103
Water
0.05 wt %
0.2 wt %
0.4 wt % 102
10-1 0.6 wt %
Viscosity (Pa s)

101
G' -0.5 wt %
10-2
G'' -0.5 wt %
100 G' -0.2 wt %
G'' -0.2 wt %
Guar gum Neutralized Carbopol G' -0.15 wt %
G'' -0.15 wt %
10-3 10-1
10-1 100 101 102
10-3 10-2 10-1 100 101 102 103 104 105
Frequency (rad/s)
Shear rate (s-1)
Fig. 3. Rheology of Carbopol (a) steady shear, and (b) dynamic oscillatory
Fig. 2. Steady shear rheology of (a) XG and (b) GG solutions.
shear in the viscoelastic linear regime.

apparent yield stress (i.e.hf g_K1 ) and gel-like properties Both surfaces are independently driven to obtain a
(i.e. G 0 OG 00 ). It should be noted that the maximum shear lubricated contact at a fixed slide-roll ratio over a wide
rate obtained in the rheometers is around 1000–10,000 sK1, range of mean rolling speeds. The slide-to-roll ratio SRR, is
while it is expected that a higher shear rate is present in the defined as the sliding speed us (the difference in surface
tribometer. speed between the ball and disc) divided by the entrainment
speed U (half the sum of the ball and disc surface speeds).
The ball shaft is tilted to ensure that there is no spin in the
contact and the friction force is measured by a load cell
3. Friction measurement method
attached to the ball drive shaft. A schematic diagram of the
A modified form of the Mini Traction Machine MTM test set-up is shown in Fig. 4.
(PCS Instruments, London, UK) was used to measure In the current study the applied load, W was 3 N and the
friction in mixed sliding/rolling conditions in a non- reduced elastic modulus, E* was 5.45 MPa. E* is defined as
conforming point contact. Using this equipment it is ðð1K v21 Þ=E1 C ð1K v22 Þ=E2 ÞK1 , where E1, E2 are the elastic
possible to measure the friction coefficient over a wide modulus and n1, n2 the Poisson’s ratio of the two, contacting
speed range and hence to fully map the Stribeck friction solids. These conditions give a theoretical Hertz contact
curve. radius aHertzZ1.6 mm, and a maximum Hertz contact
A rotating stainless steel ball (AISI 440; radius, pressure pmaxZ0.57 MPa. The low elastic modulus and
RZ9.5 mm) is loaded against the flat surface of a silicone low contact pressure means that, when a lubricating film is
elastomer disc (NDA Engineering Equipment Ltd, Kemp- present, the contact should operate in the isoviscous-elastic
ston, UK) and the contact immersed in lubricant at a or ‘soft-EHL’ lubrication regime (Esfahanian & Hamrock,
controlled temperature. The root-mean-square roughness of 1991).
the steel ball used in this study (RqZ10 nm) is negligible in Silicone surfaces are normally hydrophobic (HB).
comparison to the elastomer roughness (RqZ800G100 nm). However, hydrophilic (HL) surfaces can be easily prepared
J. de Vicente et al. / Food Hydrocolloids 20 (2006) 483–491 487

4.1. Approximate method

Load The approximate approach assumes that elastic defor-


W mation produces a thin film, circular Hertzian contact region
Steel ball
of film thickness, hc, and that effectively all of the friction
Silicone elastomer
originates from Couette flow within this region (Jin et al.,
1994). In Couette flow, the shear stress is given by
hus
tZ (2)
hc
where us is the sliding speed. The traction force, F, can be
calculated by integrating the shear stress over the circular
Hertz contact of radius a:
Fig. 4. Schematic representation of the point contact. Adaptation of the ð rZa
MTM device to study soft-EHL contacts. FZ t2pr dr (3)
rZ0
by oxygen plasma oxidation. This allows us to study the Substituting Eqs. (2) and (3) we obtain:
effect of hydrophobic/hydrophilic character of the surfaces
u
on tribological performance. For more details in this F Z hpa2 s (4)
hc
technique as well as surface treatment/cleaning the reader
is referred to de Vicente, Stokes, and Spikes (2004b). The friction coefficient, m, is thus given by:
In this work, the temperature was fixed at 35 8C and F u
friction coefficient measurements were carried out at a fixed mZ Z hpa2 s (5)
W hc W
slide-to-roll ratio of 50% over a range of entrainment speeds
between 4 and 1200 mm/s. Friction measurements were the From Hertz theory
mean of values measured with the ball travelling faster and  
2 RW 0:67
slower than the disc and experiments were carried by pa Z 2:6 (6)
progressively decreasing the entrainment speed. A new E
elastomer disc was used for each test experiment. so the friction coefficient can be written as:
m Z 2:6SRR R0:67 ðhUÞW K0:33 EK0:67 hK1
c (7)
Therefore, the apparent friction coefficient is given by
4. Friction in soft-EHL
m Z KðhUÞ0:34 (8a)
The tribology of materials in soft contacts has received where
relatively little attention despite its importance in a number
of engineering systems, such as seals, wiper blades, tyres K Z 0:8ðSRRÞRK0:09 W K0:12 EK0:22 (8b)
and rubber or plastic machine components as well as in
bio-lubrication processes such as are present at joints, eyes
4.2. Full soft-EHL solution
and skin.
According to elastohydrodynamic theory (EHL) theory,
Full solution of the soft-EHL problem was also carried
the central film thickness in the iso-viscous elastohydrody-
out numerically in order to rigorously calculate friction
namic regime for a ball-on-disk contact is approximately
coefficient within a soft-EHL contact and compare this with
4/3 the minimum film thickness, and is given by (Esfahanian
both experiments and the approximate approach described
& Hamrock, 1991):
above. This involved simultaneously solving the isoviscous
hc z3:2R0:76 ðhUÞ0:66 W K0:21 EK0:45 (1) 2D Reynolds equation and the elasticity equation, Eqs. (9)
and (10), respectively, using a forward iterative approach as
where U is the entrainment speed, and h is the effective described by Hamrock and Dowson (1978)
viscosity. This equation was obtained by regression-fitting    
computed results over a wide range of conditions of v 3 vp v 3 vp dh
h C h Z 12Uh (9)
entrainment speed, viscosity, load, modulus and radius. vx vx vy vy dx
Unfortunately, similar regression equations do not seem to ðð
have been published to predict friction coefficient so that 1 pðx 0 ; y 0 Þ 0 0
uðx; yÞ Z dx dy (10)
this can be compared with measurements in this study. 4pE r
A
Two approaches have, therefore, been used by the authors,
an approximate one and a full soft-EHL solution. where p is the local pressure at point (x 0 , y 0 ), and u(x, y), the
488 J. de Vicente et al. / Food Hydrocolloids 20 (2006) 483–491

total elastic deflection at point (x, y). The undeformed ball (a)
Water
on flat geometry, S(x, y) was approximated by a parabola as 50% CS + 50% Water
shown in Eq. (11): 72% CS + 28% Water
86% CS + 14% Water
92% CS + 8% Water
x2 y2 93% CS + 7% Water

Friction coefficient,
Sðx; yÞ Z C (11) 95% CS + 5% Water
2R 2R 10-1
The solution approach was a standard finite difference
one with an evenly spaced 128!128 grid. One operating
variable which has little effect on film thickness but may
affect the friction is the extent to which the contact inlet is
filled with fluid; i.e. the extent to which the contact is fully 10-2
flooded. This depends on the depth of lubricant supplied to
cover the elastomer disc in the MTM rig. In the current
101 102 103
work, the effect of several different supply thickness values
Entrainment speed, U (mm/s)
between 0.7 and 4 mm were modelled and the lubricant was
assumed to fill the inlet half of the contact with a circular (b)
pool of fluid of radius determined by this thickness. In
practice, this corresponded to inlet distances of between 2
and 8 Hz contact radii. 10-1

5. Results and discussion Friction coefficient, Wa ter


50% CS + 50% Water
72% CS + 28% Water
5.1. Dry contact 86% CS + 14% Water
92% CS + 8% Water
10-2 93% CS + 7% Water
95% CS + 5% Water
In the absence of any lubricant (i.e. dry contact) at 3 N Approx. EHL Theory
load and 50% slide-to-roll ratio, a negligible dependence of Numerical solution

the friction coefficient on entrainment speed was observed


10-6 10-5 10-4 10-3 10-2 10-1 100
over the range investigated. Interestingly, higher friction
Entrainment speed * viscosity, Uh (N/m)
was measured in the case of hydrophilic elastomer surfaces
(mHLZ0.60G0.03) than hydrophobic ones (mHBZ0.39G Fig. 5. Stribeck curve for Newtonian fluids (a) friction vs. entrainment
0.02). We interpret this as resulting from a stronger speed curve, and (b) friction vs. speed!viscosity curve. Dashed line
adhesion between hydrophilic/hydrophilic (HL/HL) than corresponds to the approximate EHL solution (Eq. (8)). Solid line
corresponds to the numerical solution (Eqs (9)–(11)).
hydrophobic/hydrophilic (HB/HL) surfaces.

5.2. Stribeck ‘master-curve’ for aqueous Newtonian fluids A master Stribeck curve can be generated for these
aqueous Newtonian fluids by multiplying the entrainment
The friction coefficient as a function of entrainment speed by the dynamic viscosity of the lubricant, so long as the
speed with the hydrophobic elastomer is shown in Fig. 5(a) roughness, elasticity of the surfaces, the load and the slide-to-
for corn syrup/water mixtures. It can be seen that the friction roll ratio are kept constant. This is shown for the corn syrup
curve varies strongly with the concentration of corn syrup samples in Fig. 5(b), where the results are plotted as friction
and thus the viscosity, even in the mixed-regime. For the coefficient versus (Uh). There is overlap between the curves
lowest viscosity fluid, water, at slow speed, a friction and the three lubrication regimes are clearly seen. This curve
coefficient of about 0.2 is observed, which is only just below represents a baseline against which to test both variations in
that for a dry contact, indicating that the contact boundary lubricating properties and non-Newtonian effects.
operates essentially in the boundary regime, i.e. no The approximate and full predictions of friction
significant entrainment of the fluid occurs. For solutions coefficient calculated as described in Sections 4.1 and 4.2
of 50 and 72% corn syrup, near boundary lubrication is above are shown as lines in Fig. 5(b). It is apparent that the
seen at slow speeds, with a friction coefficient of around 0.1, approximate approach of assuming only friction from
but due to their higher viscosity, some fluid entrainment Couette shear in the theoretical Hertzian contact does not
occurs, which leads to a decrease in friction, with yield correct predictions. However, as shown by the solid
increasing speed. For the highest concentration of corn line, the full solution predicts both the measured slope of
syrup, mixed-lubrication is observed at slow speeds but the soft-EHL curve and also the absolute friction coefficient
full film lubrication occurs at high speeds, leading values measured for an appropriate inlet thickness of ca.
ultimately to an increase in friction coefficient with 0.8 mm. Thus, the full numerical solution confirms the
increasing speed. experimental data which indicates that the friction
J. de Vicente et al. / Food Hydrocolloids 20 (2006) 483–491 489

coefficient depends on the product of the viscosity and


entrainment speed to a power close to 0.5 instead of the
value of 0.34 predicted by the simple theory.
10-1

Friction coefficient,
5.3. Shear-thinning polymer solutions

Fig. 6 shows measured friction coefficient as a function of


XG (wt %) GG (wt %)
entrainment speed obtained using a range of different
Water Water
concentrations of xanthan and guar solutions in water. Both 0.005 0.05
HB and HL surfaces were investigated. It can be seen that, in 10-2
0.02 0.2
general, the friction coefficient falls with increasing concen- 0.07 0.4
tration of polymer over the whole speed range. The HL 0.2 0.6
elastomer gives higher friction than the HB one at almost all
10-6 10-5 10-4 10-3 10-2
conditions.
Entrainment speed * Scaling factor, Uk (N/m)
Xanthan and guar biopolymers are used throughout foods
and beverages for stabilisation and thickening. They Fig. 7. Master Stribeck curve for XG and GG polymer solutions for
typically form shear-thinning fluids, i.e. their viscosity hydrophobic elastomer surfaces. Solid line corresponds to the averaged
decreases with increasing shear rate. We test the influence of Newtonian Stribeck curve (Fig. 5(b)).
rheology and polymer conformation on the tribological
properties of several biopolymer solutions at different as the polymer concentration is increased, the friction
concentrations by comparing the measured friction against
coefficient in the boundary-to-mixed regime decreases
the Newtonian master curve. It was previously observed that
(Cassin et al., 2001; de Vicente et al., 2004b), in the
absence of any strongly adsorbing species. Fig. 7 shows the
(a)
friction coefficient for these two different types of polymers
in aqueous solution plotted as a function of speed multiplied
by a shift factor K. The value of K is adjusted such that the
curves overlap with the Newtonian master curve (solid line),
10-1 and a comparison is made between K and the dynamic
Friction coefficient,

viscosity measured at 1000 sK1. Reasonable agreement is


found, suggesting that in the gap between the rubbing
surfaces at high speed, the friction is determined by the
Water
0.005 wt %
physical properties of the fluid at the appropriate shear rate
0.02 wt % for a given surface, load, and roughness. The value of K
0.07 wt %
Xanthan gum effectively corresponds to the viscosity of the fluid in the
0.2 wt % gap, and is only marginally greater than that of water,
10-2 despite the highly viscous nature of the fluids at low shear
101 102 103 rates. This suggests that the shear rate in the film is so high
Entrainment speed, U (mm/s) at all speeds that the effective fluid viscosity is low and
constant. It should be noted that because of the thin gap, the
(b)
shear rate will vary only very slightly across the contact and
hence all the fluid entrained in the gap has effectively the
same apparent viscosity, equal to the bulk value at the
appropriate shear rate. The conformation of the polymers
Friction coefficient,

-1
10 appears to have little effect on the frictional response, expect
in governing the rheology properties, despite the confine-
ment to film thicknesses of less than a micron.
Water
0.05 wt %
0.2 wt % 5.4. Boundary and mixed lubrication of soft-microgel
0.4 wt % Guar gum suspensions
0.6 wt %

Fig. 8(a) shows the friction coefficient, m for water and


101 102 103
Entrainment speed, U (mm/s)
various concentrations of Carbopol between HL–HL
surfaces (i.e. hydrophilic elastomer). At low U, the
Fig. 6. Stribeck curve for polymer solutions (a) XG, and (b) GG. Open unneutralised (UN) Carbopol suspensions have similar
(solid) symbols for hydrophilic (hydrophobic) elastomer surfaces. friction coefficients to those of water. At high speeds,
490 J. de Vicente et al. / Food Hydrocolloids 20 (2006) 483–491

(a) N-Carbopol passes through a maximum in the transition into


Water
0.1 wt % UN the mixed-regime where m decreases with increasing
0.15 wt % UN
0.2 wt % UN entrainment speed. Friction coefficient in the mixed-regime
0.5 wt % UN is similar to that for the UN-suspensions of microgels,
0.1 wt % N
Friction coefficient,

0.15 wt % N except for the 0.5% UN-Carbopol which has a much


0.2 wt % N
0.5 wt % N smaller m.
10-1 Stribeck curves for the Carbopol suspensions between
HB–HL surfaces (i.e. the hydrophobic elastomer) are shown
in Fig. 8(b). The magnitude of m at slow speeds is lower than
for the HL–HL contact because greater adhesion occurs
between the latter surfaces. The Stribeck curve for the UN-
Carbopol, hydrophilic
Carbopol is similar to that of water, although there is a
slightly higher m at low speeds and a lower m than water at
101 102 103
high speeds, the latter presumably due to differences in
Entrainment speed, U (mm/s)
viscosity at the high shear rates in the film. The Stribeck
(b) curve for N-Carbopol passes through a maximum between
Water the boundary- and mixed-regimes. However, at low speeds,
0.1 wt % UN
0.15 wt % UN m is the same for each N-Carbopol concentrations above
0.2 wt % UN
0.5 wt % UN 0.1%, but the transition to mixed regime occurs at different
Friction coefficient,

0.1 wt % N speeds. Friction coefficient, m in the mixed regime is also


0.15 wt % N
0.2 wt % N different at different concentrations, with 0.5% UN-
0.5 wt % N
10-1
Carbopol giving the lowest friction although m does not
strictly decrease with increasing concentration.
The lubrication properties in the boundary- and mixed-
regimes for the neutralized Carbopol may be related to the
confinement of swollen microgel particles in the interfacial
Carbopol, hydrophobic film rather than bulk rheological properties. There is likely to
be some degree of association between the microgels and the
101 102 103
surfaces, resulting in differences in lubrication behaviour on
Entrainment speed, U (mm/s)
either the hydrophilic or hydrophobic elastomer. In the
Fig. 8. Friction coefficient as a function of entrainment speed for boundary-regime, there seems to be more of a decrease of m
unneutralized (UN) and neutralized (N) Carbopol solutions. (a) Hydrophilic with increasing concentration with the hydrophilic elastomer
and (b) hydrophobic elastomer surfaces. than with the hydrophobic one. However, in the mixed-
regime, there appears to be a stronger concentration
dependence using the hydrophobic elastomer. The formation
mixed-lubrication is observed and the UN suspensions have of a confined film of microgel is similar to the observations
a lower m than that of water, presumably due to the slightly made in squeeze flow experiments by Meeten (2001) for
higher viscosity caused by the presence of microgels. The structured fluids that are similar to Carbopol. In squeeze flow
Stribeck curve for the more structured N-Carbopol it has been shown that the yield stress increases when the film
suspension has a very different shape to that of water and thickness decreases below a size similar to that of the
UN-Carbopol. At low speeds, boundary lubrication is microstructural elements. This may indeed be occurring at
observed as a slight increase in friction with increasing low rotation speeds in the MTM, where the film thickness
speed. Friction coefficient in the boundary-regime decreases decreases to approximately the size of the Carbopol
with increasing concentration for the HL–HL surfaces. The microgels (i.e. microns). The frictional response in this
large increase with speed is unusual for the boundary- boundary regime increases with a similar dependency on
regime, where a plateau is generally observed for layers of speed as for hydrodynamic lubrication and may be a
absorbed species such as surfactants. The charged hydro- reflection of the microrheology of the confined
philic Carbopol microgels may associate with the surface to microstructure.
form a layer which is difficult to penetrate and which
therefore prevents the surfaces from effectively contacting.
The increase in friction with increasing speed in this domain 6. Concluding remarks
may then be related to the physical or rheological properties
of the film of microgels as they are confined between In order to obtain a better understanding of the in-mouth
the surfaces. Since the microgels themselves are of order tribological behavior of foods, we have studied three types
microns in size, it may be expected that the film thickness of model materials in a simulated soft-EHL contact. The
itself may be of similar size. The Stribeck curve for contact consisted in a stainless steel ball loaded against a flat
J. de Vicente et al. / Food Hydrocolloids 20 (2006) 483–491 491

silicone rubber. A rolling-sliding relative motion is applied de Vicente, J., Spikes, H. A., & Stokes, J. R. (2004a). In J. W. Lee, & S. J.
and the friction mapped as a function of the speed. Lee, Complex fluids in thin lubricating films between soft surfaces.
Proceedings of the XIVth international congress on rheology. Publ.
The model fluids used in the work were: Newtonian corn- University of Seoul.
syrup water mixtures, biopolymer solutions and microgel de Vicente, J., Stokes, J. R., & Spikes, H. A. (2004b). In G. Fuller,
suspensions. A master Stribeck curve was obtained using Tribology of polymer solutions in isoviscous elastohydrodynamic
the Newtonian fluids having a wide range of different contacts. Proceedings of the eighth international tribology conference.
viscosities. This master curve was later used to study the NA08. University of Pretoria.
Esfahanian, M., & Hamrock, B. J. (1991). Fluid-film lubrication regimes
lubrication properties of xanthan and guar biopolymer
revisited. Tribology Transactions, 34, 628–632.
solutions. It was found that, in spite of their conformation Frollini, E., Reed, W. F., Milas, M., & Rinaudo, M. (1995). Polyelectrolytes
being very different and their being strongly shear thinning, from polysaccharides—selective oxidation of guar gum—a revisited
these biopolymer solutions behaved effectively as New- reaction. Carbohydrate Polymers, 27(2), 129–135.
tonian fluids with an apparent constant viscosity corre- Giasson, S., Israelachvili, J., & Yoshizawa, H. (1997). Thin film
morphology and tribology study of mayonnaise. Journal of Food
sponding to the one measured at high shear rates. On the
Science, 62(4), 640–646.
other hand, microgels particles mainly operated in the Hamrock, B. J., & Dowson, D. (1978). Elastohydrodynamic lubrication
boundary regime and interestingly showed an increase in of elliptical contacts for materials of low elastic modulus.
friction with increasing the entrainment speed in this regime Transactions of the ASME Journal of Lubrication Technology,
suggesting that some kind of confinement of the swollen 100, 236–245.
microgels is taking place. Jin, Z. M., Dowson, D., Fisher, J., Rimmer, D., Wilkinson, R., &
Jobbins, B. (1994). Measurement of lubricating film thickness in low
Analytical and numerical full solutions of the isoviscous elastic modulus lined bearings, with particular reference to models of
elastohydrodynamic problem have been briefly presented, cushion form bearings for total joint replacements. Part 1: Steady
and the results compared to the master curve. It has been state motion Proceedings of the Institution of Mechanical Engineers,
shown that the approximate analytical model does not J208 pp. 207–212.
satisfactorily predict the friction in soft-EHL conditions; in Kokini, J. L. (1987). The physical basis of liquid food texture and texture
taste interactions. Journal of Food Engineering, 6, 51–81.
this case the full solution is needed. Lee, S., Heuberger, M., Rousset, P., & Spencer, N. D. (2004). A
tribological model for chocolate in the mouth: General implications
for slurry-lubricated hard/soft sliding counterfaces. Tribology Letters,
16(3), 239–249.
Acknowledgements Luengo, G., Tsuchiya, M., Heuberger, M., & Israelachvili, J. (1997). Thin
film rheology and tribology of chocolate. Journal of Food Science,
A Post-doctoral Marie Curie Fellowship, from the E.U. 62(4), 767–772.
awarded to J. de Vicente is acknowledged. We are Malone, M. E., Appelqvist, I. A. M., & Norton, I. T. (2003). Oral behaviour
of food hydrocolloids and emulsions. Part 1. Lubrication and deposition
grateful to Unilever Research for permission to publish
considerations. Food Hydrocolloids, 17, 763–773.
these results. Meeten, G. H. (2001). Squeeze flow between plane and spherical surfaces.
Rheologica Acta, 40, 279–288.
Norton, I. T., Goodall, D. M., Frangou, S. A., Morris, E. R., & Rees, D. A.
(1984). Mechanism and dynamics of conformational ordering
References in xanthan polysaccharide. Journal of Molecular Biology, 175,
371–394.
Cassin, G., Heinrich, E., & Spikes, H. A. (2001). The influence of surface Zirnsak, M. A., Boger, D. V., & Tirtaatmadja, V. (1999). Steady shear and
roughness on the lubrication properties of adsorbing and non-adsorbing dynamic rheological properties of xanthan gum solutions in viscous
biopolymers. Tribology Letters, 11(2), 95–102. solvents. Journal of Rheology, 43(3), 627–650.

You might also like