Forum of Technology

In this article the range of Hycast ball clays are presented, one of which was used in vitreous sanitaryware bodies with
china clays to demonstrate the benefit of NSC kaolin to improve the resistance of a body to cracking. The concept of
a Cracking Tolerance Number is introduced as a means to characterise the tendency of a body to crack and this con-
firmed other evidence that NSC china clay imparts superior plasticity and packing characteristics to a body. These cor-
relate with low cast moisture contents and are attributed to the enhanced fluidity of the NSC product.

Clay Systems
for Improved Performance
NSC China Clay: Benefits to Sanitaryware.

N.P. Glasson, N.R.Forbes, UK*

Introduction Particle size range -5 -2 -0,5 D50

Kaolins 60...70 35...45 15...25 3...4
At ECC International over the years, considerable Ball clays 80...90 70...80 45...55 0...5
investment has continued to be applied to the under- Non-plastics 20...25 8...12 0...5 10...15
standing of the fundamental properties and variables Body 50...60 35...40 20...25 4...5
that affect the body properties of sanitaryware casting
slips. For convenience, in this article, the properties Tab.1 Typical particle size parameters for sanitaryware bodies [%]
are divided into two types, i.e. objective and subjec-
tive, where the latter infers an element of operator
judgement or opinion in the measurement. Examples
(The non-plastics are assumed to be suitable and con-
include firming-up, texture, drainage and even plas-
sistent for this analysis, where typical particle size
ticity by some potters’ standards. Objective properties
parameters for non-plastics, clays and bodies are as
represent those parameters that are unequivocal and
presented in Tab 1). For the rheology, the influence of
central to the manufacturing aim, i.e. casting time (or
the clay fraction can be expressed as the effect on the
casting rate) and cast strength. It is a relatively
balance of deflocculant needed to retain constant
straightforward exercise to formulate a body to pro-
thixotropy, where the components can be represent-
duce the desired objective properties, using base data
ed in the form of a hypothetical relationship as fol-
obtained on the components. However, the associat-
lows (assuming complete mechanical dispersion):
ed performance and subjective behaviour for some
options will be far more acceptable than others.
Surface Water Added Effective
An earlier investigation [1] showed that improve-
Charge + Soluble = Chemical + Organic
ments to de-moulding properties could be obtained
related props. Salts Dispersant Colloid
by using NSC kaolin in traditional and pressure cast-
ing formulations. In that work it was demonstrated
For example, for constant casting slip thixotropy, as
that NSC, in combination with Hycast VC ball clay,
the factors on the left-hand side of the equation are
produced a lower cast moisture content body with less
increased, the deflocculant requirement for the slip
shrinkage and greater plastic tolerance. This enabled
must be increased to return the thixotropy back to the
castware to be de-moulded earlier than usual, with
constant value. Conversely, when the factors on the
less tendency to crack and therefore improved yields.
right-hand side of the equation are increased, the
Although these results were supported in factory prac-
deflocculant requirement of the slip must be decreased
tice, there was no currently universally accepted labo-
to recover the value for constant thixotropy.
ratory test method to confirm the differences. How-
Essentially, this means that increases on the left-hand
ever progress was being made at ECC International
side can be balanced by increases on the right-hand
and one of the techniques developed, defined as a
side, within acceptable and practical limits of rheo-
“Cracking Tolerance Number” (CTN) is now described
logical control for the application.
in this article. The results underpin the other evidence
Sanitaryware manufacturers will be familiar with the
obtained i.e. that NSC kaolin imparts improved per-
effects of these variables on the casting slip rheology,
formance to a sanitaryware body.
but possibly not be aware of the significance they can
have on the subsequent casting performance. In this
Dependency on Clay Properties paper, these are briefly discussed with reference main-
Objective casting properties are absolute and are ly to packing and permeability parameters. Clearly to
dependent on the packing and permeability of the achieve the objective of speed of factory throughput,
casting slip composition. This is inter-dependant with a minimum permeability is essential, but with good
the casting slip rheology and is related to the clay packing and corresponding low cast moisture con-
properties used to formulate the sanitaryware body. tent. Cracking of castware can be linked to this cast
moisture content and associated shrinkage in the
* N.P. Glasson, N.R.Forbes, ECC International Ltd.( a member of region above the critical moisture content; particular-
IMERYS - group of companies),St. Austell PL25 4DJ, Cornwall ly in the mould. Choice of clay system can limit this

E22 cfi/Ber. DKG 78 (2001) No. 3

Forum of Technology
feature and provide a balance of plasticity and deflocculated condition of about 0,7 ... 0,9 Pa·s by
strength to resist these stresses responsible for crack- Brookfield viscometer. The equivalent measure by the
ing. The NSC/Hycast VC clay system provides a com- Gallenkamp torsion viscometer is typically 300/45,
bination of optimum surface charge related proper- degrees over-swing fluidity/degrees difference for a
ties, low water soluble salts, no residual chemical and 90 s thixotropy.
a controlled amount of effective organic content to To study the casting behaviour associated with these
match this requirement. body casting slips, it is necessary to use more than just
a single measurement at the preferred condition for
casting. This is important because ceramic properties
are often associated with high levels of statistical error.
Type of Measurement To avoid this, a range of casting slips are removed
To evaluate these effects, at ECC International a “stan- from the bulk preparation at usually five under-
dard body” is used which consists of: deflocculated conditions. The casting properties are
25,0 % Ball Clay then determined for each of the five slips, plotted
28,5 % Kaolin, and against the respective V60 thixotropy value and from
46,5 % Non-Plastics, this, the result at the casting condition of 6 Pa·s
vitrifying at 1200°C to less than 0,5 mass % water thixotropy interpolated, as illustrated in Fig 1. In the
absorption. example, the results refer to the packing density
With these evaluations it is important to emphasise (1,786 g/cm3) and permeability (11,1x10-14 cm2) of a
that the slip density in the preparation is not pre-set, body consisting of Remblend kaolin and Hycast
but instead, the total solids in the slip are adjusted to Zenith as the ball clay.
achieve full deflocculation of the casting slip at a con- This procedure was used to obtain the results on the
stant viscosity of 0,5 Pa·s (5 poise), by Brookfield visco- majority of the properties measured, but for the test of
meter. This viscosity then represents a minimum cracking, a further bulk sample was adjusted to the
value. For thixotropy this is measured over the longer V60 of 6 Pa·s and used to determine the “cracking tol-
period of 60 min, which is a time more relevant to the erance number”.
casting period used in practice. This is called a V60
long term thixotropy and from experience and evalu-
ation of many factory casting slips, a typical value Cracking Tolerance Number
found in factory practice is 6 Pa·s. To obtain this This method is based on the concept that the mechan-
thixotropy, the casting slip is usually at an under- ical stresses exerted on a piece of castware in the
mould are influenced by the strength, plasticity, and
shrinkage of a body whilst drying to the critical mois-
ture content. It was thought that a combination of
these measurements could be used to calculate a char-
acteristic “cracking tolerance number”, for which
higher results would suggest increased resistance to
cracking. Since drying shrinkage is a tedious measure-
ment to undertake, the change in moisture content
from casting to de-moulding, which is known to be
equivalent to the change in volume of the body, was
used as a measure of shrinkage.
Ideally, the “initial cast moisture” should be used as
the first moisture content, but the experimental deter-
mination of this value is highly dependent on when
Fig. 1 the piece can be de-moulded. Therefore, the casting
Interpolation of cast slip moisture content was used, on the grounds that it
body properties at was a more reliable measurement to make, although
constant thixotropie it was recognised that it was not an equivalent term to
the initial cast moisture content. The maximum flex-
ural deformation (i.e. at the point of rupture) was used
as a measure of plasticity.
The expression used to calculate the cracking tolerance
number (CTN) for each moisture content condition
was therefore as follows:

CTN = MoR x Plasticity/ (Moistureslip -Moisturecast)

As stated earlier, the slip moisture content is not an

ideal measurement of “initial moisture content”, and
hence the product of modulus of rupture and plasti-
city was thought to be a useful measure of the relevant
physical properties above the critical moisture con-
tent (cmc). This is referred to as the Cracking Tolerance
Product (CTP), because it is the multiple of the MoR
and plasticity:

CTP = MoR x Plasticity

E 24 cfi/Ber. DKG 78 (2001) No. 3

 Forum of Technology
To illustrate the differences between bodies, the ered detrimental to the associated tolerance of the
results are usually represented as a cluster group on a body to cracking.
graph of CTN plotted against cast moisture content. A
grouping which is high and to the left of the graph Added Chemical Dispersant
then signifies a body that is more resistant to cracking This subject was also discussed fully in [2]. Briefly, the
than a cluster that is low and to the right. However results showed that it was extremely important to
this is not always a convenient method of presenting maintain a tight control on the ratio of deflocculants
the results, especially when there are a limited num- used in the casting slip preparation. In particular, in
ber of bodies in the comparison. To obtain a single Western Europe the sodium carbonate/sodium silicate
overall value therefore, a further variation is to use the combination is a system which is commonly used and
critical moisture content (cmc.)as the reference con- it is generally accepted that sodium carbonate is the
dition (the cmc. being determined by an independent least effective of the two as a deflocculant. However,
means) and utilise the following modified equation sodium carbonate also activates the organics fraction
(where 0,33 is the conversion to a linear expression): of the UK type ball clays and releases fines, which
increases the packing density and thereby contributes
CTNcmc=(MoRcmc x Plasticity):([Moistureslip - cmc] x 0,33) favourably to a reduction in cast moisture contents.

The CTN is determined as follows: Effective Organic Constituents

Approximately 20 bars (thickness: 6mm) are cast, With regard to the effect of organic constituents on
then de-moulded as early as possible (usually after the performance, most of the work reported by ECC
about 40...50 min) and then left to dry in air for vari- International was done in the late 1970s [3]. The
ous times before sealing in plastic bags and storing “effective” organics are contributed by the ball clay
overnight. MoR and peak deformation are both meas- fraction and for the Hycast ball clay range these are
ured on these bars using a 60 mm span on a Testomet- donated by the same South Devon carbonaceous seam
ric Flexural (3 point bending universal tester) and the components. It was shown that the minimum per-
moisture content of the broken bars determined. The missible level for a typical sanitaryware body was in
data obtained is plotted in the form of graphs of plas- the region of 0,3 to 0,4 mass % carbon, as an indica-
ticity (peak deformation) and strength for the respec- tor of the organic content. For the Hycast products,
tive moisture contents as conditioned. this equates to a carbon content of not less than
Now, before moving on to discuss individual factors 1,2 % (as 25 % in body) in the size fraction below 125
which affect the packing behaviour and thereby the μm, which is the screening specification commonly
influence on cracking, a key feature of the evaluations used for vitreous sanitaryware manufacture. Table 2
to emphasise is that the casting slips are prepared to a shows the fundamental properties of the Hycast range
constant rheology for casting, which is typical of that used in which the carbon values are listed.
in factories and is achieved by adjusting the slip density up
or down, as appropriate. The relevance of this procedure Surface Charge Related Properties
to the importance of NSC on the casting properties Table 2 lists amongst the fundamental properties, the
will become evident later in the discussion of the values of Methylene Blue Index (MBI), which for
results. Hycast ball clays are quoted for convience as mg dye/g
of dry clay. For natural i.e. non-chemically treated
products,which applies to the Hycast ball clays, a
Factors in Hypothetical Relationship measure used for this factor is the MBI. This is quoted
The effect of the variables in the hypothetical rela- for convenience as mg methylene blue dye adsorbed
tionship on the packing, permeability and associated per g dry clay and the values obtained for the indi-
rheology have been investigated by the methods vidual clays can be used to formulate for objective
described with the following result: property performance in the body (Tab 3).
This range of Hycast ball clay products was achieved
Water Soluble Salts by producing a blend of the organic South Devon
The effects of water soluble salts i.e. calcium and sodi- fluid clays with the acknowledged plasticity associat-
um sulphates in particular but including sodium chlo- ed with the Dorset types but using a special new
ride and sodium nitrate for interest, were investigated process of finer shredding, blending, paddle-mixing
in full in [2]. That study showed that the added salts and in-process quality control. The data in Tab.3 show
were responsible for a reduction in the packing densi- that, at 10 bar pressure and room temperature, in
ty and an increase in permeability over the range combination with Remblend, this body can produce a
investigated. The porosity and the cast moisture con- cast thickness of 8 mm in 6 min (4 min at 40°C).
tent were raised and therefore this would be consid- These three products then formed the basis for further

Hycast type Unit Rapide Mega VC Visa Zenith Classic Tab 2

SiO2 mass % 55 55 53 56 55 56 Fundamental
K2O mass % 1,7 1,8 2,0 2,0 1,9 2,2 properties of Hycast
clays for sanitary-
L.O.I. mass % 12,1 11,6 12,4 10,9 11,0 10,0 ware
Carbon mass % 2,0 1,9 2,2 1,8 1,4 1,4
Water soluble salts mass % 0,11 0,12 0,12 0,25 0,16 0,16
<5 μm mass % 85 85 91 87 86 86
<0,5 μm mass % 41 43 50 49 51 53
MBI mg/g 19 22 26 27 29 12

cfi/Ber. DKG 78 (2001) No. 3 E 25

Forum of Technology
Hycast type Unit Rapide Mega VC Visa Zenith Classic
cast strength mass % 2,7 2,9 2,9 3,2 3,1 3,4
packing density mass % 1,729 1,751 1,760 1,762 1,786 1,780
Tab 3 permeability mass % 19,8 16,8 14,7 13,5 11,1 8,9
Comparison of casting rate (20°C) mass % 1,56 1,25 1,14 1,10 0,88 0,69
Hycast products time for 8 mm mass % 41 49 56 58 73 93
in the standard time for 8 bar mass % 10 12 14 15 - -
Remblend kaolin time for 10 bar mass % 6 7 8 9 - -
body
MBI calculated mg/g 9,1 9,9 10,9 11,1 11,6 12,4

MBI = Methylene Blue Index

combinations, in response to customer specific Remblend. Table 4 summarizes the fundamental

requirements, e.g. Bodystar slurry products from properties of the NSC product compared with Rem-
Cliffe Vale, UK, and Hycast Rapide/Hycast Classic blend and confirms a casting concentration for NSC
blends to produce Hycast Mega (for high pressure of 69 % solids. This is an indication of the higher flu-
application) and Hycast Visa (for medium pressure idity of the NSC product solids when compared with
application). This latter product was developed in the 66,5 mass % solids specified for Remblend, for the
association with a pressure casting equipment suppli- same viscosity condition.
er to mirror the properties of Hycast VC, but with a When evaluated in the standard body specified, the
lower carbon content to accommodate faster firing. Remblend/HycastVC plastics combination provides
The most recent addition to this portfolio was the an 8 mm cast in 60 min and for conventional casting
introduction of the non-chemical refined ball clay it is probably slightly faster than preferred. Adjusting
with the name of Hycast Zenith. the balance in favour of more ball clay to kaolin
From the work described above, it was apparent that, would modify this condition, but one other alterna-
for conventional casting slips, the calculated MBI tive is to replace the Remblend with NSC. The results
should be in the order of about 11,0 ... 11,6 mg/g. of this body are presented in Tab. 5 in comparison
Tab. 3 shows that a body based on Remblend and with the Remblend/Hycast VC body prepared at the
Hycast Zenith corresponded with this criterion. How- same time (as opposed to the series of experiments
ever, as accomplished by sanitaryware manufacturers listed in Tab. 3 and studied on a separate occasion).
all around the world, this objective can be reached by The results show that there has been an extension of
several combinations. Other factors will therefore 7 min to the casting time for an 8 mm cast thickness,
have an influence on the decision of the final compo- and the properties engineered into the NSC product
sition and part of this will be the nature of the sub- by the feed clay selection and subsequent mechanical
jective properties. As indicated above this, to some processing have imparted significant benefits to the
extent, is dependent on the judgement and prefer- body. Because the method of evaluation adjusts the
ence of the operatives. slip density to a fixed minimum viscosity of 0,5 Pa·s,
this is evident as a higher slip density with improved
packing (unfired bulk) density. The associated dried
Mechanically Beneficiated Kaolin cast strength is lifted and the cast moisture content
Remblend is the kaolin recognised as the sanitaryware reduced from 19,3 ... 18,0 mass %. This body has
industry standard from ECCI. Naturally it would excellent drainage, texture and “firming-up” behav-
therefore occupy the place as the kaolin in the stan- iour and is also very plastic, as measured by the abili-
dard body. It is produced by blending selected com- ty of the body to deform (reported in mm) before
ponents for chemistry and mineral consistency and breaking.
processed through a variety of size separation stages to Fig. 2 and 3 are graphs representing the data obtained
arrive at the specification listed. Quality control for the determination of the Cracking Tolerance Num-
includes conformance to casting and fluid properties. bers for the two bodies listed in Tab. 5. They show the
NSC kaolin is a product which, in addition to the pro- results of strength and plasticity plotted against cast
cedure described for Remblend, is subjected to moisture content condition respectively. These indi-
mechanical processing during its production. As a cate that the critical moisture content, i.e. the point
kaolin, it reports a particle size analysis equivalent to of inflection, occurs at a higher value for the Rem-
Remblend. However, by the application of the blend compared with the NSC body (determined as
mechanical treatment into the production flowsheet, 15,2 versus 14,5 mass % by an independent method)
higher fluidity is achieved for the same strength as and although the strength over the undried condi-

Properties Unit Remblend NSC Properties Unit Remblend NSC

SiO2 mass % 48 48 Slip density g/l 1801 1827
Al2O3 mass % 36,5 36,5 cast strength MPa3 3,1 3,3
K2O mass % 2,0 2,0 packing density g/cm 1751 1800
Fe2O3 mass % 0,96 0,90 permeability cm2x10-14 15,1 12,8
>10 μm mass % 20 19 casting rate (20°C) mm/min 1,07 0,96
< 2 μm mass % 40 40 time for 8 mm min 60 67
MoR MPa 1,1 1,5 cast moisture content % 19,3 18,0
MBI mg/g 12 14
plasticity mm 2,8 3,0
Casting rate mm2min-14 1,5 0,9 MBI calculated mg/g 10,6 11,2
Casting concentration mass % 66,5 69,0
Tab 4 Tab 5
Fundamental properties of kaolins for sanitaryware bodies Comparison of kaolins in bodies with Hycast ball clay at constant
thixotropy

E 26 cfi/Ber. DKG 78 (2001) No. 3

 Forum of Technology

Fig.2 Cast strengths of NSC and Remblend bodies at different cast Fig 3 Plasticity of NSC and Remblend bodies at different cast moi-
moisture contents ( cast at constant V 60 = 6 Pa·s thixotropy) sture contents ( cast at constant V 60 = 6 Pa·s thixotropy)

tions only slightly favoured the NSC body, the main A previous study had shown that clay systems
difference was in the improved plasticity (Fig. 3). This based on NSC kaolin with Hycast ball clays imparted
reflects in a higher CTNcmc value of 28 for the beneficial properties to a body during de-moulding.
NSC/Hycast VC clay system compared with 23 for the For a comparable casting time the NSC system
Remblend/Hycast VC combination in bodies. enabled a higher slip density to be achieved for the
It is clear from the results that the NSC process in same casting slip rheology.
combination with Hycast VC, produces a superior These bodies reported lower cast moisture contents
packing arrangement into the clay performance with and wet to dry shrinkage and better plastic tolerance
additional plasticity which translates into greater behaviour. This inferred a greater resistance to the
cracking tolerance. This is supported by results in stresses responsible for cracking and that the time for
industrial practice. Some manufacturers have been de-moulding could be made shorter. The study lacked
enjoying the benefits of this system for the 20 years the supporting evidence of an accepted independent
that both clays have been in existence, but now with measure of cracking in castware.
the continuing pressures of the economic climate, the In this article, the concept of a “Cracking Tolerance
system has been subject to a far more searching analy- Number” was introduced. This consisted of a formula
sis. The body combination of NSC/Hycast VC has based on measurements of strength (MoR), plasticity
recently been introduced into more new factories and (maximum deformation at break) and shrinkage. A
clay losses have been significantly reduced. Customer variation of the technique enabled a single value to be
confidentiality does not permit the reporting of deduced, with reference to the critical moisture con-
specific data but these are not isolated cases and tent, for both of the bodies reported.
the financial benefits to the users have been substan- The Cracking Tolerance Number results for the NSC
tial. and Remblend bodies were 28 and 23 respectively.
This supported the other evidence previously report-
ed and in particular, that observed in practice, which
confirmed that NSC kaolin, in combination with
Conclusions Hycast VC ball clay, contributed to bodies with
In general, for a specified practical casting time, the reduced clay losses.
body which reports a lower cast moisture content will
give the better factory yields. This is
related to the particle packing in the body and References
depends upon factors which include the surface char- [1] N.P. Glasson, D.G. Stockley, Keramica, 245 1-2/2001
ge related properties of the clay system in use, the [2] N.P. Glasson, International Ceramics Journal, Feb. 1998, 34 -
associated water soluble salt content, residual process 42
deflocculant (if used) and the organic constituents [3] C.S. Hogg, ECCI Publication: Further Aspects of the Rheology
present. of Sanitaryware Casting Slips, 1984

cfi/Ber. DKG 78 (2001) No. 3 E 27