ON THE MEASUREMENT OF MASSECUITE FLOW

PROPERTIES
J. N. Ness
Bureau of Sugar Experiment Stations, Bundaberg, Australia

ABSTRACT
Various flow models have been used in the literature to explain the flow
properties of massecuites. Massecuites have been considered to be Bing-
ham plastic, pseudo-plastic and Newtonian fluids.
Two different viscometers, an efflux type pipeline viscometer and a
Brookfield RVT rotational viscometer, were used to study the flow proper-
ties of a number of samples of low grade cane massecuites. All samples
showed pseudo-plastic behaviour and the power law model gave a good fit
to the data from both instruments. Some samples tended towards constant
apparent viscosity at the higher shear rates in the pipeline viscometer.
The flow behaviour index obtained from the pipeline studies was inde-
pendent of tube dimensions. The measured consistency was dependent on
tube diameter and length. The rotational viscometer gave a lower flow
behaviour index ( a higher degree of non-Newtonian behaviour) than the
pipeline viscometer.

INTRODUCTION
T h e rheology of massecuite is significant in the processing in the raw sugar
factory a s it affects t h e operation of t h e vacuum pans, the crystallizers and the
fugals as well as pumping a n d conveying of the material in the mill. Massecuites
a r e usually considered t o be non-Newtonian materials-although the precise nature
of the material has been variously described. T h e different conclusions reached
by various investigations may arise from inherent differences in the materials
studied (high o r low grade massecuites, beet o r cane material. influence of diffe-
rent types of non-sugars, presence of any deterioration products) but may have
also been influenced by the experimental method used t o study the massecuite
flow p r o erties.
P
D o n e , ~ n a m e n s k i i ' , P o ov a n d ~ r o i n o 'all considered massecuites t o be Bing-
P
ham plastic materials. D o n also commented o n thixotropic properties. ~ d k i n s '
noted that massecuites would flow under very low shear rates and concluded it was
pseudo-plastic. Silin -',from studies with a rotational viscometer, found massecui-
t e t o conform best t o the laws of flow of Newtonian fluids. Gromkovskii and
Ivanov " considered that massecuite did not follow any of the classical forms of
flow but was nearest t o the Newtonian fluid. They also noted slip at the wall of
the inner cylinder for a modified rotational viscometer, especially for massecuites
with high crystal contents.
O t h e r investigators using rotational viscometers e . g. Behne ', Kot et al,"
'
Awang a n d White, Garcia, "' and Gebler er a1 ' I have concluded that massecui
tes were usually pseudo-plastic fluids. T h e last mentioned authors also thought
that a t crystal contents higher than 55 per cent by weight Bingham plastic beha-
viour was approached.
Ozer investigated beet molasses a n d massecuites with a pipeline viscometer
and concluded that whereas molasses is a pseudo-plastic fluid, massecuite was
Newtonian. H e attributed the non-Newtonian behaviour of molasses to air which
was not present in the massecuitk samples. Ness 13. l4 has reported measurements
on the rheology of massecuites using different sizes of tubes in an efflux type
pipeline viscometer. The results showed that massecuites were pseudo-plastic
fluids, with some samples approaching Newtonian behaviour (constant apparent
viscosity) at higher shear rates.
Maudarbocus l5 discusses the design of an orifice viscometer and showed that it
gave results consistent with the pipeline viscometer when used to measure masse-
cuite rheology properties. H e assqmed massecuites obeyed the power law fluid
model.

EXPERIMENTAL PROCEDURE
Two instruments were used in this study: a pipeline viscometer and a Brook-
field model R V T Synchro-Lectric viscometer. The pipeline viscometer used has
been described previously (Ness 13* 14). The Brookfield viscometer has been well
'
described in the literature (e. g. Behne, ~ o w l e set a l . , l 6 Rosen "). In these
tests the cylindrical number 7 spindle was used and the sample was held in a water
jacketted copper vessel of 89 mm internal diameter and 115 mm de pth. The guard
legs supplied with the instrument were not fitted.
Various massecuites were investigated with both types of instruments, and a
comparative study was also done in which the same five massecuite samples were
investigated with both viizometers.

Calculation of Results
The raw data from the pipeline viscometer consists of the applied air pressure
and the mass of efflux collected in a measured time interval. The analysis of the
data is based on being able to relate these basic measurements to the shear stress
tw and shear rate Yw at the tube wall. For fully developed flow the wall shear
stress is:

and the apparent wall shear rate is:

The true wall shear rate is given by:

In the above relationships D and 1 are the tube diameter and length, V is the
average flow velocity in the tube and A P is the full developed pressure drop. A
Y
log-log plot of tw vs 8VlD yields a curve of slope n . For a power law fluid this
plot is a straight line and n' is constant and usually written as n.
The viscosity of the fluid is then defined by:
 The measured pressure drop consists of the fully developed pressure drop
including the head of fluid plus end effects and exit losses. The measured flow
velocity V will be higher than it should be if slip occurs at the wall as a result of
particle migration away from the tube wall. A method for correcting raw data for
these extra pressure losses and wall slip effects has been given by Sarmiento et
al. l8 These procedures require that measurements be made with a range of tube
diameters (at least three) and tube lengths (at least three).
The data from the Brookfield viscometer consists of the dial reading and the
speed of rotation. Knowing the spring constant for the instrument and the spin-
dle dimensions the shear stress and shear rate at the surface of the rotating spindle
can be determined (Rosen 17). For the results in this paper the shear rate was
calculated using the formula:

1, N - Speed of rotation, rpm. r, and r2 - Radii of the spindle and container

n Slope of the line of log t vs log N.

Power Law Model

The power law rheology model relates shear stress and shear rate as:

where K is the consistency, the higher K the more viscous the fluid, and n the flow
behaviour index, is a'measure of the degree of non-Newtonian behaviour.

RESULTS AND DISCUSSION

/ Pipeline Viscometer
In general the power law rheology model was found to give a good fit to the data
from the pipeline viscometer, at least over a limited range of shear rate. At higher
shear rates constant apparent viscosities, as has been reported elsewhere (Ness '7,
were noted with some massecuite samples and, on occassions even indications of
dilatancy. These observations were noted in samples containing significant quantities
of gas.
Figure 1 shows a typical results from a pipeline viscometer measurement. The
shear stress has been plotted against the apparent shear rate. The curve is seen to
change slope at an apparent shear rate of about 1 s-I for this particular sample.
Power laws were fitted to each section of the graph resulting in a flow index of
0,86 at the lower shear rates and 1,00 at the higher rates.
Table 1 gives the results of meassurements obtained on two different massecui-
te samples using a range of tube sizes. Table 2 gives the dimensions of the
various tubes used in the study. The power law parameters (flow behaviour index
 Figure 1. Flow curve from pipeline viscometer.

DAPm
n, consistency K) were calculated by a least square fit to the raw data (
8Vm 4L '
-) where APm is the measured pressure drop across the system and Vm the
n
-
measured average flow velocity. In all cases the correllation coefficient was
better than 0,998 in log-log space.
The data in Table 1 show that as the tube length is reduced the calculated
consistency increase. This is consistent with the influence of end effects in the
results where the additional pressure loss due to tube entrance, tube exit and the
exist kinetic energy of the efflux stream becomes relatively greater for shorter
length tubes. The results also show that as the tube diameter is increased K is
also increased. This observation is consistent with wall slip, where because of
particle migration away from the tube wall a higher flow rate is obtained than
would otherwise be the case. The effect is more pronounced with smaller diame-
ter tubes.
Table 1. Power Law ~ k a m e t e r sin Different Size Tubes

Sample 1 Sample 2

Tube Shear*Rate ' Shear Rate

K K
n N.sn.m-2 Range n Range
N.sn.m-'
scl s-I

Table 2. Tube Dimensions

Tube ' Diameter (mm) Length (mm)

However, both these observations are also consistent with time-dependent flow
properties and it is difficult to quantify this property with capillary type viscome-
ters. End effects can be overcome if sufficiently long tubes are used, while slip
effects are reduced by using tubes of large enough diameters. Measurements with
longer tubes than used in this study may help to clarify some of these observa-
tions.
The flow behaviour index n does not seem to be influenced in any systematic
way by the tube dimensions.

Brookfield Viscometer
As noted in the introductiop various investigators have used rotational visco-
meters t o study massecuite flow behaviour. Some of these (Done, Behne, ' '
Awang and White ') have commented on an apparent thixotropic behavior with
such viscometers. ' Other (e.g., Gromkovskii and Ivanov ') have referred to slip at
the inner rotating cylinder, especially for high crystal content massecuites.
Figure 2 shows a set of readings obtained with one massecuite sample. The
dial reading is seen t o decrease with each revolution especially at higher speeds of
rotation. This the viscosity calculated using the highest readings at each speed
would differ significantly from the final readings.
I
Revolution number

I Figure 2. Brookfield dial reading as a function of number of revolutions.

Comparative Study
1 A comparative study of the two viscometers was undertaken on five different
massecuite samples collected from successive strikes in the one factory during the
I 1981 season. A ~ i n g l etube size (17,04 X 401 mm) was used for the pipeline
/ viscometer and so corrections for end effects or wall slip cannot be made. Figure
3 shows the typical flow curves (shear stress vs shear rate) obtained with the
1 viscometers for both instruments.

11 Table 3 gives the power law rheology parameters obtained on all five samples
for which the chemical analysis are given in Table 4. For both viscometers the
data werewell correlated by the power law flow model. In all samples it is seen
'
that the Brookfield viscometer gives a lower flow index than the pipeline viscome-
ter i.e. it indicates a much higher degree of non-Newtonian behaviour. The
values of n obtained in these trials were somewhat lower than are usually observed
with the Brookfield viscometer. Typically n is in the range of 0,7 to 0,9 whit this
instrument. However previous experience had indicated that the Brookfield vis-
cometer would give lower flow behaviour index values than the pipeline viscome-
ter on the same sample. The Brookfield viscometer also gave lower consistency
figures than the particular pipeline viscometer tube used in these trials. Referen-
ce to Table 1 shows that this tube (3L) gave higher consistency figures than the
smaller diameter tubes.
It is also seen that the Brookfield viscometer operates over lower shear rates
than the pipeline viscometer. The limits on the pipeline viscometer are between
zero applied air pressure i.e, when operating under the head of the massecuite
only and the upper safe working limit of the sample container, about 300 kPa for
this instrument. 2'. .
\ 2: ,<
Table 3. Power Law Rheology Parameters

Pipeline Brookfield
Sample
K Shear Rate K Shear Rate
n n
N.s".~-' , scl N.sn.mc2 S-

For the Brookfield viscometer the working range is determined by the spindle
used, the spring constant and the speeds of rotation available. For these samples
observations were restricted to speeds of rotation from 0,5 to 10 rpm, with only
the lowest three speeds possible in the most viscous sample.

Table 4. C-Massecuite Analysis '

Sample RS S DS TP CC

' RS, per cent reducing sugars; S, per cent sucrose; DS, per cent dry substance; TP, per cent true
purlty; CC, crystal content per cent massecuite by w e ~ g h t .

CONCLUSIONS

All samples of massecuite examined with both instrufments showed pseudo-

plastic behaviour. The power law model gives a good fit to the observed beha-
viour. Over a wider range of shear rate it might be necessary to use separate
power laws to adequately fit all the data.
The flow behaviour index obtained from pipeline viscometry was independent
of the tube dimensions. The consistency was affected by tube diameter and tube
length implying that if the true shear stress-shear rate curve is to be obtained
tubes must be selected that minimize end effects and wall slip effects. Alternati-
vely measurements can be made with a range of tubes such that allowances can be
made for these effects in the data.
The Brookfield viscometer tends to indicate a lower flow behaviour index
(higher non-Newtonian) than the pipeline viscometer. ~

The pipeline viscometer can be designed to operate over a wider range of shear
rates than the Brookfield viscometer.
The extent of thixotropic behaviour has not yet been satisfactorily quantified.

1302
 REFERENCES

1. Done, W. G . (1950): The Flow Properties of Molasses and Massecuites. B. App. Sc. Thesis.
Univ. of Qld.
2. Znamenskii, G . M. (1951): As Reference by Kot et al
3. Popov, V. D . and V. P. Troino (1958): As Reference by Kot et al.
4. Adkins, B. G . (1951): Notes on the Viscosity of Molasses and Massecultes. Proc. Old. Soc.
Sugar Cane Technol. 18th Conference, pp. 43-52.
5. Silin, N. P. (1953): As Reference by Kot et a1 R .
6. Gromkovskii, A . I. and Ivanov, S. R . Flow of Massecuites. Sakhar. Prom, 40 (40), pp. 17-23.
7. Behne, M. F. (1964): Viscometry in Massecuites. Proc. Qld. Soc. Sugar Cane Technol. 31st
Conference, pp. 289-296.
8. Kot, Y. D . ; Yassinshava, T. V. and Suschenko, A . K. (1968): VISCOUS Properties of Massecuites.
Sakhar. Prom., pp. 106-125.
9. Awang, M. and White, E . T. (1976): Effect of Crystal on the Viscosity of Massecuites. Proc.
Qld. Soc. Sugar Cane Technol., 43rd Conference, pp. 263-270.
10. Garcia, A . D . (1977): Rheological Study for Predicting the Effect of Temperature and Crystal
Content on the Viscosity of Molasses and Massecuites. Cuba Azdcar, Jul-Sept, pp. 45-51
11. Gebler, J . ; Giz, K , and Hobikova, V. (1978): Rheological Properties of Massecuites 111. Incline
Flow and Outflow from a Vessel, Listy Cukrovarnlcke, 94:80-83.
12. Ozer, E . (1977): Rheological Properties of Molasses and Massecuite. Turklsh Sugar Institute
13. Ness, J . M. (1979): The Rheology of Massecu~ieby Pipeline Vlscometry. First Nat. Conf. on
Rheology, Melbourne, May 30-June 1, pp. 47-50.
14. Ness, J . N. (1980): Massecuite Viscosity-Some Observations With a Plpeline Viscometer. Proc
Aust. Soc. Sugar Cane Technol., pp. 195-200.
15. Maudarbocus, S. M. R . (1980): The orifice viscometer: A new technique for measuring rheologl-
cal properties of massecuites and molasses. Proc., Int. Soc. Sugar Cane Technol. 17th Congress,
Manilla, pp. 2257-2263.
16. Bowles, R . L.; Dawe, R . P. and Todd, W. D . (1955): A Method for the Interpretation of
brookfield Viscosities, Modern Plastics. 33 (3), Nov., pp. 140-148.
17. Rosen, M. R. (1971): A Rheogram Template for Power law- Fluids. J Colloid and Interface
I
Science. 36 (3), July, pp. 350-358.
18. Sarmiento, G . ; Crabbe, P . G . ; Boger, D . R . and Uhlheer, P. H . T . (1979): Measurement of the
Rheology Characteristics of Slowly Settllng Flocculated Suspensions. Ind. Engng Chem. Pro-
cess Des. Dev. 18 (4), pp. 746-751.

MESURAGE DES PROPRIETES D'ECOULEMENT DES

MASSES CUITES
J. N. Ness
Bureau des stations expkrimentales du sucre, Bundaberg, Australie

RESUME
La littkrature offre une grande vari6tk de moddes servant a expliquer
les proprikt6s d'kcouleument des masses cuites, class6es en fluides plasti-
ques Bingham, pseudo-plastiques et newtoniens.
Deux viscosim6tres on kt6 utilis6s: l'un adapt6 a un tuyau et l'autre de
1303
type rotatif Brookfield RVT, afin d'Ctudier les propriCtCs d'Ccoulernent de
plusieurs Cchantillons de masses cuites bas produits de cannes.
Tous les Cchantillons ont en un comporternent pseudoplastique et la loi
de puissance s'est bien adaptCe aux donnCes obtenues par les deux instru-
ments. Certains Cchantillons avaient tendance B une viscositC apparente
constante lorsque les taux de cisaillement dans le viscosirnktre adapt6 au
tuyau Ctaient ClevCs.
L'indique d'Ccoulernent obtenu des Ctudes faites sur les tuyauteries
nYCtaitpas en rapport avec la dimension du tuyau, alors que la consistance
dCpendait du diarnetre et la longueur de celui-ci.
Le viscosirnktre rotatif a donne un indice d'Ccoulement plus faible (un
degrC plus ClevC de cornporternent non newtonien) que celui du viscosirnk-
tre adapt6 au tuyau.

SOBRE LA M E D I C I ~ NDE LAS PROPIEDADES

DE FLUIDEZ DE LA MASACOCIDA
J. N. Ness
Bur6 de Estaciones Experimentales del Azucar, Bundaberg, Australia

RESUMEN
E n la literatura se han utilizado diversos rnodelos para explicar las
I
propiedades de fluidez de las rnasacocidas. Las rnasacocidas han sido cata-
logadas corno fluidos plasticos Bingham, seudopl~sticosy newtonianos.
Dos viscosirnetros diferentes -un viscosirnetro de tuberia y uno rotato-
rio Brookfield RVT- fueron utilizados para estudiar las propiedades de
fluidez de una serie de rnuestras de rnasacocidas de cafia de baja pureza.
Todas las rnuestras presentaron un comportarniento seudoplastico y el rno-
delo de la ley de potencias se ajust6 bien a 10s datos de arnbos instrurnen-
tos. Algunas de las rnuestras evidenciaron una tendencia hacia la viscosidad
constante aparente a lo: rnayores indices de coste en el viscosirnetro de
tuberia.
El indice de comportarniento de la fluidez obteiido a partir de 10s
estudios del tub0 fue independiente a las dirnensiones de este ultimo. La
consistencia registrada si dependi6 del diarnetro y de la longitud del tubo.
El viscosirnetro rotatorio arroj6 un indice de cornportarniento de fluidez
inferior (mayor grado de cornportarniento no newtoniano) que el viscosirne-
tro de tuberia.