CHAPTER 15

Testing control and quality

assurance

The inkmaker has at his disposal an exten ive inventory of chemical

materials from which to manufacture printing ink. It i unfortunate that
the term, 'raw materials', is used for these components since few are
naturally occurring substances intended (or use after minimal processing.
Modern pigments, resins, additives, solvents and oils are the sophisticated
products of industrial chemical manufacturers.
A great many of these ingredients are specially made for the ink
industry to meet stringent specifications. The final performance of print-
ing ink relies on the selective choice of starting materials, on the formu-
lation, and, of no less importance, the care with which the ink is
manufactured.
Consist nt product quality is th rightful expectation of an nd u r. To
meet this objective, careful scientific testing is essential at all stages of
manufacture.
Since the selection of starting materials is an expensive part of the ink
development programme, it follows that having made a suitable choice of
component, reproducibility is essential. This can only be ensured by the
rigorous testing of each batch of the approved materials as they are made.
Such testing must be carried out and be guaranteed, within the manu-
facturer's domain. It should also be systematically checked by the user to
ensure that compliance to the agreed standards is being attained.
The inkmaker must always agree with the supplier a full specification
and the expected quality parameters for all starling materials. A similar
expectancy arises between the inkmaker and the printing ink end user,
the printer. To this end product specification and testing is a central
activity of the printing inkmaker.
Testing of inks and starting materials falls into two main areas:
• qualitative methods which rely on visual and comparative techniques
such as now, con i tency, shade;
• quantitative methods which give numerical results such as tack,
viscosity, yield value.
SAMPLING T CHNIQUE 805
With the development of newer methods of ink testing many subjective
tests can be replaced by more scientifically based procedures.
This chapter provides a guide for both the testing of raw materials, and
for the testing of finished inks or intermediate base components of ink. It
is not possible within the confined space of the chapter to offer a complete
account of all the possible tests since some are specific to definite prod-
ucts. Te ts which are considered essential are covered in some detail.
Where standard tests are included only the salient points are given and
the reader is advised to refer to the publi hed text for the complete details.

15.1 STANDARD TESTS

It is essential that standardization of test procedures should be estab-
lished, so that results obtained in one laboratory can be directly compared
with those from another laboratory. This is true both with raw materials
(between the inkmaker and raw materials supplier) and also with the
printed product (where an ink may have to comply with an end-use
speciIication).
Some tests have been widely agreed and standardized by such author-
ities as:
• The Briti h Standards Institute (BSI);
• European Confederation of Paint Printing Ink and Artists Colours
Manufacturers Associations (CEPE);
• The Technical Association of tlle Pulp and Paper Institute (TAPPI);
• The American Society o[ Testing Materials (ASTM);
• The Institute of Petroleum (IP);
• Printing Industries R search Association (PIRA);
• The ationa! Association Of Printing Ink Manufacturers, loc USA
(NAPI1\1).

15.2 SAMPLING TECHNIQUE

Por raw material testing 1'0 be meaningful. tl1e sample raken from a
delivery of raw material must be representative of the entire delivery. A
delivery may indude more than one batch, and so batch numbers should
be carefully checked when sampling. Suppliers should be asked to make
deliveries which consist of as few different batch s as pracricable. This will
reduce the amount of testing required. If suppliers can also provide
predelivery samples, these can be tested well in advance of me delivery,
and only an occasional delivery spot check is then necessary.
The inkmaker may choose to test all materials at the time of delivery-
tanker quantities are an example - or to carry out regular spot checks to
ensure that agreed standards are being achieved.
It isimportanr that materials which are not tested on delivery or are
only POt checked are provided by approved supplier. Approved sup-
pliers are those who have a proven ability to supply consistently materials
of agreed quality. They should have a recognized quality standard, or be
806 TESTING CONTROL Af'fD QUAUTY A SURANCE

prepared to give certificates of conformity with each delivery. It may be

desirable to visit the supplier and carry out an audit of their qualily
system to ensure that the required tandards are being anained.
The following paragraphs are intended as a guide to sampling different
kinds of materials.

Resins
Re ins which are delivered in lump form may well contain some smaller
pieces and fine powder. A composite sample should be taken which is
representative of the proportions of each state present. It must also be
remembered that there is usually a specific reason for resins being deliv-
ered in lump form. Usually this is to prevent oxidation. These resins
should only be crushed immediately prior to the actual testing procedure.

Pigments
The vibration in transit of a cask of pigment may well have caused some
segregation 01 particles of different size. Hence the cask should either be
tumbled before opening, or several samples taken at varying levels in the
cask with a special auger-like tool.

Liquids
Liquids should be thoroughly stirred or the barrels tumbled before open-
ing, to ensure that no separation into distinct phases has occurred. H the
material is being d livered by tank r then ach compartment within the
tanker should be sampled.

15.3 PIGMENT TESTING

Quality control of pigments is particularly diHicult as a single pigment can
be used in a wide variety of vehicle systems to give inks for many
different end uses. Hence, the testing procedure must be selective and it
can logically be divided into three sections. First. limited testing can be
carried OUI on the dry pigment itself to ensure the batch meets the
speci[ication for shade, panicle size, oil absorption, moisture and volatile
content. The second and far more important consideration is to compare
the inkmaking characteristics of the balch against the master standard
in the inkmaker's own vehicle ystems. Finally, provided the batch meets
the specification, prints [rom the sample and standard inks are checked
[or their resistance properties.
In reality most inkmakers would be unlikely to carry out very intensive
testing o[ incoming pigments, as most pigments originate from manufac-
turers who have recognized quality standards. Many of the resistance
properties are a factor of the chemical nature of the pigment and are
unlikely to vary con id rably from batch to batch, or are related to the
vehicle systems being used, and are hence tested [or as part of the finished
PIGMENT TESTl 'G 807
ink quality-control checks. Furthermore, testing for such parameters as
particle size will involve tests and equipment which will generally be
beyond the scope of the average quality controf department.

Tests on the dry pigment

Particle size
Since the pigment particle size will a[(ect such fundamental properties as
strength, hue, rheology and stability of the finished ink, it might seem
desirable to check panicle size or at least the average size distribution, as
part of the quality control procedure. However, in the size range of most
pigments (less than 1 f.Lm) the majority of the standard techniques for
sizing, e.g. gravitational sectimentation and optical microscopy, are slow
and inaccurate. Only two methods are suitable for particle size measure-
ment in this range - electron microscopy and centrifugal sectimentation -
both of which are generally beyond the scope of the average quality
control department, and would not be considered as normal control
techniques for inkmakers. However, even the inkmaker without analyti-
cal facilities can use a local university to assist in particle size determin-
ation iI necessary.
Electron microscopy is an expensive method and requires an experi-
enced operator, but is capable of producing very accurate readings.
The disc centrifuge (Pig. 15.1) is perhaps the preferred technique and
results compare favourably with those obtain d using denron micros-

DISC CENTRIFUGe:

~ ~
~ ~._­
a.a I

Fig. 15.1 A Joyce Loebl disc centrifuge.

808 TESTING CONTROL AND QUALITY A SURANCE

copy. However, as neither method is simple and quick enough to be used

as a standard quality control procedure the inkmaker places greater
emphasis on the di persion of the pigment particles in the finished ink
(discussed later). For further reading on particle size determination by
centrifugal sedimentation the reader should consult Paint Technology
(1965) 29 (5), 30 and J. Beresford (1967) J. Oil Col. Chem. Assoc. 50,
594-614.

Oil absorption
The oil absorplion of a pigment wiU depend on:
• the chemical nature of the pigment;
• particle size and size distribution;
• particle shape;
• surIace characteristics of the sample, such as those obtained by treat-
ment to improve wettability or even simply due to absorbed moisture.
[f the oil absorption of the pigmem varies from batch to batch, then
formulation changes may have to be made to the inks, and clearly this
is undesirable. Since the factors affecting oil absorption cannot easily be
measured, the test is reduced to a purely empirical comparison between
the batch and the master standard. The actual test method is given in BS
3483: part B7: 1974, the result being quoted as the mass (or volume) of
oil required to wet 100 g of pigment. Alternatively, the sample and
standard can be dispersed in a measured amount of oil and the relative
consistencies of the two samples can be compared.

Moisture and volatile content

Absorbed moisture may affect the ease of dispersion of some pigments,
and this can b d termin d according to the method in BS 3483: part 86:
1982. A temperature of I05°e is chosen for this test, as at higher temper-
atures volatile maller may be lost which is an essemial part of the
pigment. This is especially true or carbon black and pigments on an
aJuminium hydrate base.

15.4 INKMAKING CHARACTERISTICS

All pigments should be checked wherever possible in the vehicle systems
in which they will be used in production, but in certain cases where the
pigment is used in several di[(erent types of inks, this may be roo time
con uming and a selective range of formulations must then be chosen.
This should include oil and liquid ink formulations where appropriate.
The method of dispersion in the laboratory should:
• simulate as near as possible the production process for miXing and
grinding;
• be rapid, comparative. and reproducible;
• utilize small quantities of pigment so as to retain a single master
standard for tesling as many batches as possible.
INKMAKfNG CHARACTERISTI S 809
The first test must be to ensure that the batch of pigment disperses at least
as readily as the standard and to an equivalent degree, so that costly
production time is not lost. The degree of dispersion is measured using a
(jneness of grind gauge IFig. 15.2). This is a stainless steel block contain-
ing a shallow groove which is graduated in depth from zero to, usually, 25
or 50 microns. A sample of ink is placed at the deep end, drawn down
with a teel blade towards the shallower part, and the degree of di persion
is taken as the point at which the continuous ink film breaks down across
the groove. However, care must be taken in interpreting the result if
discontinuity in the ink film occurs higher up the gauge in the form of
scratches due 10 the odd oversize particle. It is then important to deter-
mine whether these scratches on the gauge are caused by undispersed
pigment agglomerates, or by, for example, undispersed paste or dried ink
or varnish skin. Also, care should be taken that these odd scratches on the
gauge are not merely caused by airborne dust panicles that have settled
on the gauge or contaminated the ink sample after it was taken. This will
be more of a problem on the wider channel 50-micron gauges than the
narrow channel 25-micron gauges which are more commonly used, as
most quality control specifications will require a dispersion of at least
5 microns and below. The wide channel 50-micron gauge is more com-
monly used for examining the degree of dispersion at the mixing stage
belore the ink is ground, but can be very useful for checking for particU-
late contamjnation from skin, etc. which may be present but may not
actually show up on a narrow 25-micron gauge. If unusual scratches are
observed on a gauge. higher up than the overall degree of dispersion,
then the examination of the ink under a microscope is recommended to
determine the cause.
Once the correct djspersion level has been allained, the trial ink can be
compared with the standard for strength. hue, gloss and flow (or consist-
ency). First, th flow is ch cked as it is impossibl to obtain m aningful
comparative prints unless the batch and standard are similar in consist-
ency. Liquid inks having near-Newtonian rheological properties can be
compared by viscosity measurements using any of the standard tech-
niques (e.g. Ford, Sheen and Zahn cups). Oil inks, however. are not
eWlOnian and require more difficult instrumental measurement uch
as a Laray falling bar or cone-and-plate viscometer. It is important that
the viscosity of the trial ink matches that of the standard or comparative
flow tests will not be meaningful. To match the flow alone by comparing
their consistency with the aid of a palette knife, or by comparison of their
relative flow properties on an inclined or vertical glass/perspex plate is
not adequate as the Dow properties of the pigment may vary [rom batch
to batch and should also be checked. Similar flow properties may be
obtained, but at different viscosities, which is obviollsty not meaningful
as production barches will be primarily adjusted to a spedfic visco ity or
tack (section 15.13), rather than being adjusted 10 a specific flow without
regard to the latter parameters.
If the flow of tJle trial ink is different from the standard at a similar
viscosity, then this may indicate that there is some variation between
batches of pigment, or that the degree of dispersion of the twO ample is
not equal.
810 TESTING CO TROL AND QUALITY A SURANCE

Fig. 15,2 A fineness of grind gauge.

-
The strength of the trial pigment is then assessed on a drawdown at full
strength and al reduced strength as a 'bleach'. The drawdown is carried
out by placing small quantities of the standard and test inks about one
inch apart on the top of a securely held piece of paper, and then firmly
scraping them towards the operator wilh, [or example, a 3-inch wide
blade. Initially the blade should be held nearly vertical to reproduce a film
thickne s comparable to a printed Eilm weight and then finally the blade
should be brought almost parallellO the pad to give a thick film weight-
the maSSlone of the ink. The drawdown should be carried out on a semi-
absorbent paper on which a broad black band around 19 mm wide bas
previously been printed, so that the opacity (or tran parency) of the batch
and standard pigments can also be compared.
The bleach technique involves dilution by mixing (not grinding) a
coloured ink with a much larger quantity of opaque white, so as to
eliminate any minor diHerence of glos and hue. In the case of yellow
inks, the eye cannot differentiate small strength differences, and so suf-
fident blue is added to the bleach to bring the hue into the blue-green
region.
Hue can be measured by instrumental technique (Chapter 3), but
normally a visual comparison of a drawdown and print is sufficiently
accurate. The undertone and toptone of the inks can be assessed using the
thin film of the drawdown and the thicker masstone respectively. Gloss
can also be checked on the drawdown, but more accurately by examin-
ation of a print of the batch versus standard on an appropriate stock.
INKMAKING CHARACTERISTI S 811

Resistance tests
Provided the pigment matches th standard on the tests conducted so far,
the appropriate resistance checks are carried out. Many of the character-
istic propertie of a pigment are a function of its chemical nature, and
hence are unlikely to vary from batch to batch, unless an impurity is
present. Pigments are, however, coated to provide easier dispersion in differ-
em vehicle systems and this can give rise to batch variation, especially
when the pigment is obtained from an alternative supplier.
All the resistance tests are carried out on prints of the batch and
standard inks, rather than dry pigments themselve .
It is important that a vehicle is chosen which is appropriate to the
intended application of the final ink.

Lightfastness
As the Iightfastness of a pigment is a function of its chemical nature, it is
only necessary to carry out random spot checks, but auentjon should be
paid here to pigments used in inks where lightfastness is an imponant
criterion, e.g. wallpaper inks.
Pigments will show a higher lightfastness at full strength and therefore
it is important that a realistic strength print be submiued for test. For
example, a PMTA pigment should rarely be tested at full strength, since
its main application will be as a toner pigment. As pigments for use on
wallpapers could often be used in tims, it is advisable to test these at both
full and lint strengths. Prints should be prepared on a good quality paper
which does not yellow on exposure to light.
The exposure of prints to daylight is too lengthy a procedure to be
suitable and the usc of an artificial source is therefore adopted. The only
light source to have an emission spectrum close to that of daylight is the
xenon arc and provid d a filter is used to reduc the amount of UV
radiation, a result comparable to daylight exposure is obtained in a
much shorter time. Other light sources such as the carbon arc and
mercury vapour lamp are not recommended as the results bear little
comparison to those obtained under daylight conditions.
Another factor which greatly affects the lightfastness figure is the
effective humidity, which is dependem on the relative humidity and
temperature of the surrounding air and the surface temperature of the
pattern. The average eflective humidity value for daylight exposure in
Britain is 45% and therefore the test apparatus should be capable of
approximating to this figure.
tn order to classify numerically the Jightfastness results obtained, the
prints are compared with a set of eight pieces of wool, each dyed with
different blue dye which has been carefully graded to fade after a set
exposure to light. This is known as the Blue Wool scale, and the scale
ranges from one (poor Iightfaslness) to eight (excellent) according to the
number of the wool dyes which have faded under the same exposure as
the test method. The procedure for mounting and assessing samples is
given in BS 4321:1969, but it is worthy of note that the term 'fading'
includes not only a reduction in strength but also changes in hue, bright-
ness, darkening, or any combination of these £actors.
812 TESTING CO TROL AND QUAUTY A SURANCE

Heat resistance
Hear resistance of pigments is a critical factor in the areas of tin-printing
inks (and other metal coatings) and wallpaper ink, which will be sub-
j cted to hot embossing. Tlle pigments are simply tesred by placing prints
in a suitably heated oven for a set period of time, the acrual conditions
depending on the end usage o[ Ule print. The prints are tllen compared
with an unheated sample for changes in strength, hue and brightness.
Again, prints should be at a realistic strength, as heatfastness, like light-
fastness, will vary from full to tint strength.

Soap and detergent resistance

Pigments to be used on soap and detergent packaging should be checked
to ensure no impurities are present which are liable to cause bleed
problems. Although test method for both soap and detergent re istance
are given (135 4321 test methods 6 and 7 respectively) the manufacturers
of these products have their own test methods which are generally more
severe than the British Standard tests.
A typical soap re i tance test, as provided by a leading manufacturer, i
as follows. A 25% soap gel is made up and allowed to set. The gel is
moistened and a print oj the pigment to be tested is [jrst soaked in water
to prevent curing (1 minute for paper and 5 minutes for board). The print
is then placed face down on the gel and left for 15 minutes, after which
time it is examined. Arter this period there should preferably be no bleed,
although a slight bleed could be considered acceptable.
Testing for detergent resistance is becoming more complex with the
advent of enzyme and 'solvent' powders. Provided a pigment passes the
above soap test, it is unlikely to fail with the standard soap powders.
However, enzyme-detergent resistance should be assessed with an enzy-
mic powder. (To dat no standard test has b en published for resiSlanc to
'solvent detergents'.)
A typical detergent resistance test is as follows. A 5% solution is made
up using the appropriate enzymic detergent in water. Two filter papers
are soaked in the solution, one is placed on a glass slab and rh print to be
tested is placed face down on the filter paper. The second filter paper is
then laid on top and a further glass slab provides a sandwich effect. The
print is then examined after three hours and although ideally there
should be no bleed, a sHght bleed could be considered acceptable.
The disadvantage in interpretation of these test methods is that the
degree of stainjng is only expressed in vague terms such as 'slighl bleed',
whereas both the British Standard tests (numbers 6 and 7) quote
a numerical means of assessment of stairung by use oj the grey scale
(BS 2663C).

Alkali resistance
This rest can be conducted either by streaking dilute sodium hydroxide
solutions onto a print 01 rhe pigme.nt (e.g. 2, 3 and 5%) and examining
the dried streaks for djscolouration, or as in the British Standard test (BS
4321 Te t method 5) by andwiching the print between filter paper
previously soaked in a dilute alkaline solution. In this case, the filter
INKMAKING CHARACTERISTI S 813
paper are dried after a set contact period, and then the degree of staining
is assessed with the aid of a grey scale.

Acid resistance
Acid resistance can be carried out in a manner similar to BS 4321 test
method 5 above, but replacing the alkaline solution with either dilute
hydrochloric acid or dilute acetic acid. The latter reagent is normally used
where the print could come into contacl with vinegar or pickles.

Wax resistance
Wax resistance is essential [or inks on waxed paper wrappers. ice cream
cartons, certain vending machine cups, and similar applications where
the subslrate is made more water resistant by immersion in molten wax
(usually para[(jn wax). The standard test (BS 4321 Test method 10)
involves immersion of a prim of the pigment in molten wax for [jve
minutes, after which the wax is allowed to run over the margin sur-
rounding the print. The margin and the wax in the basin are then
examined for discolouraLion.

Plasticizer bleed
This method is extremely important for pigments that will be used in ink
to be printed on plasticized PVC such as [or vinyl wallpaper and shrink
wrap applications. Prints of an ink made from the pigment and a standard
ink on plasticized pvc are sandwiched between plain sheets of the un-
printed material and placed between glass plates in an oven at 45°C for
3 days. A small weight is often placed on top of the sandwich 10 increase
contact pressure. On examination o[ the unprinted substrate adjaceIll to
the print there should be no evidence of colour bleed at the end of the test
period.

Deep-freeze resistance
This test is becoming increasingly important as more and more products
are being stored under deep-freeze conditions. On carton board, the
vehicle system (or overprint varnish) will normally provide adequate
deep-freeze resistance and testing of the pigment is nOI necessary. How-
ever, for polyethylene printing the deep-freeze resistance of the finished
pack is a [unction not only o[ the vehicle (usually polyamide), but also o[
the pigment as well. The deep-freeze resistance of the pigment in these
cases is not so much a function of its chemical nature, but of the pigment
coating. and therefore quality control of each batch o[ pigment is essential
for this application. The tests are carried out on prints made [rom the
pigment under test.
The severest condition met by the print is nOI when the pack is actually
under the deep-freeze conditions, but during the stage of thawing on
removal from the freezer, and hence the test normally conducted is one
of water immersion.
A typical test is as [ollows. The print i immersed in water [or a period o[
at least 1 hour and a maximum of 5 hours. The print is then flexed
between the knuckles len times and should show no more than 5%
814 TESTING CONTROL AND QUALITY A SURANCE

removal. A second portion of the print is then blotted dry and should
show no more than 5% removal when checked by the Scotch tape test.

15.5 CHIPS, PREDISPERSIONS AND FLUSHED

PASTES
In chips and predispersions the physical process of dispersing the pigment
in the resin system has already taken place, and in theory an ink can be
made simply by solubilization of the resin component in the solvent
system. However, in production the dissolution usually takes place on a
high-shear mixer and therefore the laboratory technique should simulate
this method.
A useful comparative method of testing such predispersions is to solu-
bilize the pigment 'chip' in the solvent by shaking in a sealed container
on a Red Devil paint mixer for a set period. Finally, if the disper ion is less
than adequate both batch and standard can be further processed for a
short period on a high-shear laboratory mixer such as a SiJverson blender.
A comparison between batch and stClndard inks can then be made accord-
ing to the test methods given in the section on pigment dispersion
(section 15.3). The method of first shaking in a sealed container and
then mixing under shear enables the tester to ensure that dissolution
takes place without excessive solvent loss, as well as providing a tech-
nique for assessment of relative ease of dispersion of the standard and
batch.
If the predispersion is in the form of a preground paste this can be
t Sled as if it wer a finished ink, such parameters as degree of disper ion,
strength, hue, and rheology being examined (section 15.13)
Flushed pastes are slightly different in that the pigment is not ground
into the varnish base, but 'flushed' into it as part of the manufacturing
proce s. Flushing is the process of transferring the pigment from the
aqueous phase into the oil phase without the normal conversion into
dry powder. The pigment particles within the Oushed paste are therefore
fully dispersed to give maximum strength, gloss, transparency. etc.
Flushed pastes are normally so highly concentrated and viscous that
they cannot be tested as if they were a finished ink. The most convenient
way of testing flushed pastes is to 'let-down' the paste with a precise
quantity of a standard varnish to give an 'ink' for which a specification is
set. As the varnish is standard any deviation from the specification has to
be due directly [0 variations in the flushed paste.

15.6 DYE TESTING

Solvent dyes
Solvent dye have a limited usefulness as an alternative to the more
commonly selected pigmented chips. They are tested by ensuring that
the dye is soluble 10 the required concentration in the appropriate solvents.
RESIN 815

On standing for a further period (usually up 1048 hours) no precipitation

should occur. The strength, hue and brightness of the resulting solution
may be determined on a colorimeter or by making up a suitable ink and
comparing with a mast I' standard in the normal mann r. Lightfastness
and bleed tests can then be carried Out on a print as described earlier in
thi chapter.

Basic dyes
These dyes have poor fastness properties, high tinctorial strength, and
find their greatest use as toners in flexographic inks, after laking with, for
example, tannic acid. They should be checked for solubility and resistance
properties against a master STandard in an appropriate formulation.

Acid, reactive and direct dyes

These are little used in the printing ink industry, but do find specific end
uses, such as in ecurity inks and inks for printing tis ues and paper
kitchen towels. Again these are tested by comparison with a master
standard in a typical formulation.
When these dyes are to be used for printing tissues, etc. particular
attention must be paid to the properties of the dried print (e.g. soap
bleed resistance, ease of bleachability) since the dyes play an important
parL in determining whether or not the inks will meet the customer's
specification.

15.7 RESINS
There are very few classes of ink which do not contain a resin of one form
or another, thus before discussing the importance of testing and control
procedures an appreciation of the many forms a resin may take is essen-
tial. Encompassed within the name 'resin' is a wide range of differing
polymeric chemical entities ranging from low molecular-weight oligo-
mel's to high molecular-weight complex polymers, and from low viscosity
mobile liquids to plastic solids. Beyond the e boundaries, re ins may also
be chemjcally reactive or comparatively inert, thermally stable or sensi-
tive to light and heat.
The properties of resins selected by the inkmaker are governed by their
chemical and physical nature, which in turn are controlled by manufac-
turing procedure and the purity of the raw materials. Both of these factors
may be influenced at anyone time for any number of reasons, the
economic climate and changes in health and safety legislation to name
only two. To ensure that a consistent end product is manufactured,
suitable control procedures must be applied to the materials used in
manufacture. The simplest and most useful form of resin testing would
be to incorporate it into an ink and then monitor the desired properties of
that ink. This would be very costly in terms of manpower and time if re in
used by an inkmaker had to undergo th same testing. It is th refor
816 TESTING CO TROL Af'fD QUAUTY A SURANCE

desirable to apply emphical tests which are rapid and will convey suffi-
cient information to determine whether a batch of resin is acceptable
or not.
The following methods are typical of those employed by an inkmaker
to determine whether incoming materials are acceptable against a proven
standard.

Acid value
The stability, printability. [low, gloss and pigment wetting are just some of
the properties which are affected by the acid value of a resin. Changes in
the acid value may nOI always be detrimental, a higher acid-value resin in
an oleoresinous lithographic system may improve the water take-up and
thus press performance of the ink. However, a reproducible producT is
essential and therefore this property must be controlled within set para-
meters.
The method for determining the acid value of a resin can be found in
the section on analysis.

Hydroxyl value
Residual hydroxyl gro ps in a resin may inlluence different factors
according to the end use of that re in. Hydroxyl groups may affect re in
and varnish viscosity via hydrogen bonding; solubility and solvent release
characteristics may be influenced for the same reason. The water resist-
ance of a system may be impaired if the value is too high. Similarly,
reactive ystems may not cure properly if the value is too low. Pigment
wetting and thus gloss, transparency and flow may be adversely affected
and 'can' stability may suff r if the hydroxyl value i outsid of sp cifi-
cation.
The exact method for determining hydroxyl value for re ins is
described in the section on analysis.

Solubility
Resins are often formed into a varnish prior to incorporation in an ink
and this may itself be used as a quality test on the resin. This is not always
so and in the case of plastisols and some aqueous systems the resin is not
in the solution in the finished ink at all. In plastisols the resin forms a
plastic solution in plasticizer only when the final print is heated. The
complex interactions which occur within ink systems are not as simple as
this and many different solubility factors come into play, solvent on
solvent. solvent on resin. resin on solvent. plasticizer on resin to mention
a few of the many possibilities. If the resin is required to be compatible
and stable within a defined system it is relatively simple to test this
characteristic. Solubility is a good means of testing a resin's performance
within a system and changes in the solubility of a resin are an xcellent
indication that there is some significant batch-to-batch variation. Differ-
ences in the solubility of resins manifest themselves mainly in a change in
VARNlSHES AND OIL 817

the solution viscosity. Significant variations in the viscosity of a resin

solution indicate that either the chemical or physical nature of the
resin is different. As the molecular weight of a resin increases 0 does it
solution viscosity. Similarly, subtl changes in the solvent being used may
have significant effects upon the solution characteristics being measured.
It is therefore essential to either keep a stock of standard test solution, or
compare each new batch of resin with a retained sample of standard
material under identicaL conditions.

Melting range
The melting point of a resin is also characteristic o( its molecular weight,
thlls batch-to-batch changes in melting point indicate a variation in
molecular weight distribution which will directly influence solution vis-
cosity, film hardness and gloss. The melting range of a resin is also a very
u dul test to apply when re ins are being considered for certain applica-
tions. For example a resin of too high a melting range may have no
appLication in heat-sealing inks. SimilarLy, if the melt range is too low,
marking due to resoftening in reverse printing may prove a problem.
There are two standard methods for determining melting ranges, both
of which can be found in the section on anaLysis.

Colour
Gross colour variations although immediately obvious may nOt always
have a significant effect upon the finished product. Visual colour meas-
ur ments of a resin are far more r alistic when carried out in solution,
and instrumental techniques such as a colorimeter or standard colour
guides are normally employed.
One of the most important aspects of colour testing is the indication of
ither age or overheating that may have cau ed the darkening of a resin.
U the colour of a resin is acceptable but outside normal standards, it
would be advisable to test thoroughly the other re in properties discussed
above.

15.8 VARNISHES AND OILS

Varnishes in this category may be simple solutions of a single polymer or
they may be complex mixtures such as the 'cooked' varnishes encoun-
tered in paste-ink tech ology. Oils may be pure or partly polymerized
vegetable oils or animal· and mineral-derived products.
Often tbe most successfuL test metbod for these materials is to fabricate
them into a finished ink. However, this is time consuming and therefore
would not generally be a realistic quality control procedure. Good quality
control practice should make use o( a sufficient number of reproducible,
rapid test to ensure thal the product is the correct mat rial of a suitable
grade. Invariably economic pre sures (orce the inkmaker to test only
when necessary and thus to rely upon hjs supplier to submit raw
818 TESTING CONTROL AND QUALITY A SURANCE

materials of an acceptable quality. Many of the larger varnish and oil

suppliers now have BS 5750/ISO 9000 registration, but those which have
not will normally be prepared to supply certificates of conformity either
prior to a delivery or with it. This goes some way towards giving the
inkmaker confidence only occasionally to spot-check incoming material.
This is never a wholly acceptable situation; the ideal procedure would
involve comparing an incoming raw material with a retained sample of
standard material before approval to use that particular batdl.

General tests
General tests available \0 the inkmaker for testing incoming varnishes
and oils are as follows:
• colour
• viscosity
• infra-red spectroscopy
• refractive index
• acid value
• iodine value
• ruene value (i.e. degree of unsaturation)
• saponification value
• unsaponifiablc matter
• hydroxyl value
• water content.

Colour
The colour of an oil or varnish may influence the fjnal colour of an ink or
print, but generally the colour is used as an indication of consistency of
manufacture. For comparative purposes one can use standard colour
tubes or colour comparators such as the Tintometer or Gardner lubes.
The colour of an oil or varnish, if it has darkened, may indi.cale thaI the
sample has deteriorated and, on oxidation, oils and varnishes may no
longer have the same drying characteristics.

Viscosity
According to ewton, viscosity is the ratio of shear stress applied to a fluid
versus the shear rate, and in an ideal system this ratio should be constant
at constant temperature. It is rare that the inkmaker will encounter
ewtoruan liquids except when testing some oils, as the majority of
varnish systems encountered will have some form of thixotropy depend-
ing on the resin systems being used. In some oil-ink formulations the
resins within the varnish will have been 'gelled' with a chela ling agent,
and in this case the thixotropic nature of the varnish will be marked. For
a Newtonian liquid the relationship between shear rate and shear stress is
shown in Fig. 15.3. Deviations from this enc untered in thixotropic
systems would exhibit a curve as shown in Fig. 15.4. figures 15.5 and
15.6 are typical viscosirylLime plots for a thjxotropic system. A suitable
VARNlSHES AND 01 S 819
measurement technique must be chosen which will renect the charac-
teristic of the oil or varnish.
There are many types of viscosity measurement equipment available
and the choice of which instrument to use will depend largely on the
rheological nature of the material to be tested. Newtonian oils and var-
nishes can be tested using efrIux or orifice viscometers such as the Ford,
Shell, Zahn and Sheen cups, which are the most basic types of visco-
meters. For varnishes which exhibit thixotropic properties rotational
viscometers such as the Brookfield or Stormer may be used, but other
methods such as a cone and plate, or Laray falling rod viscometer are
ideally suited, particularly where the yield value of the system is critical
(Fig. 15.7).
When testing viscosity specifications it should be noted that viscosity is
temperature dependent and all measurements should be made at a spe-
ciIic temperature. Also it is very difficult to obtain correlation between
in truments which measure the viscosity in different ways, such as
a failing bar versus a cone-and-plate rheometer. Furthermore, it is
extremely unlikely that the viscosity measured on a similar instrument
will be identical to that specified by the supplier unless the material is
ewtonian. If this is a problem the inkmaker must establish their own
specifications on the instrument being used, or work with their supplier
to establish a correlation between their own instrument and that of the
supplier.
The remainder of the tests listed, although not neces arily lengthy, are
likely to be beyond the average quality control laboratory and more
applicable ro an analytical laboratory. However, this will vary between
inkmakers and the facilities available to them, and for inkmakers who
either wish to carry out some of these tests withjn their quality control
laboratories, or have separate analytical laboratories, the test methods
can b found in th s (lion on analysis.

Shear
stress

Shear rate

Fig. 15.3 Shear stress versus shear rate for a Newtonian liquid.
820 TESTING CO TROL AND QUAUTY A SURANCE

Shear
stress

Shear rate

Fig. 15.4 Change of shear stress with varying shear rate for a
thixotropic system.

Viscosity

Time
Shear starts

Fig. 15.5 Change of viscosity with time at constant shear rate for
a thixotropic system.

15.9 SOLVENTS
Solvents are all too often considered 10 be discrete chemical compounds,
yet even highly refined single compound solvents contain low levels of
impurities. Solvents commonly lIsed in the printing ink industry are
invariably a technical grade and therefore contain residual impurities
resulting [rom their method of manufacture. Most solvents have a char-
acteristic odour which may be used as a first indication of the nature o[
the solvent, and can also be indicative of contamination. Odour [rom
SOLVENTS 821

Viscosity

Time
Shear stops

Fig. 15.6 Re-establishment of original viscosity with time of a

destructed thixotropic system.

ester or ketone contamination in mineral-oil solvents is easy to detect and

can easily occur if a road tanker is inadequately cleaned between deliv-
eries. Quality control of solvents involves a blend of modern instrumental
techniques such as IR spectroscopy and gas-liquid chromatography
(Gtc) with traditional testing procedures.
Infra-red spectroscopy may be used to give a chemical 'fingerprint' of
the solvent under test. 0 two dil'rerent compounds will give exactly the
same spectrum. The IR spectrum of the test sample should be exactly
superimposable over that of the standard. Varia Lions from this would
indicate the presence 01 impurities. It should be noted that solvents of
large extinction coefficient (absorbance in the IR region) may well con-
ceal significant levels of low extinction-coerficient impurities.
Gas-liquid chromatography, a vapour phase separation technique, may
be successfully llsed to identify the test material, its boiling range and its
purity. Spurious chromatographk peaks obtained from a single com-
pound solvent may automatically be labelled as impurities. It is not
often possible to do this for solvents derived from petroleum as the
chemical composition will vary with the raw material supply.
Petroleum solvents may be usefully 'fingerprinted' using GLe and
significant changes in composition or aromaticity may be detected.
High performance li.q id chromatography (HPlC) and column chro-
matography are also finding application in solvent comrol. A particular
application of these tedmiques is the determination of aromatic content.

Boiling range
The boiling range of a solvent may be cited as the literature value
provided that the level and nature of impurities is within the acceptable
level. The boiling range may be readily determined by distilling a known
volume of the solvent, recording the initial boiling point, dry point, and
822 TESTING CONTROL AND QUAUTY ASSURANCE

Fig. 15.7 Falling rod viscometer with temperature-controlled

waterbath. (Courtesy of Betta-Tech Equipment Ltd)

the temperatures at which specific volumes of the solvent have distilled

over. This method is covered by 85.638: 1962.
Simulated distillation by GLe may also be used especially when utiliz-
ing modern very high performance capillary columns with computerized
data handling.

Relative density
The relative density (or specific gravity (SG)) of a liquid is the ratio of the
density of the liquid and the density of air-free distilled water at the same
temperature. The density of a solvent is commonly determined using a
hydrometer or by weighing using a calibrated density borrle. Useful
information about the purity of a solvent can be obtained from density
infonnation. Aliphatic and aromatic solvents of the same boiling range
may be distinguished as the aliphatic hydrocarbons normally have a
lower density.
 OLVENTS 823

Fig. 15.8 Abel apparatus for measurement of flash point.

Flash point
The flash pOint is defined as the lowest temperature at which the vapour
above a liquid will ignite when brought into contact with a flame or
spark. The nash points of the more volatile solvents encountered in
l1exographic and gravure ink technology are usually very low. Ethyl
acetate has a Dash point (dosed cup) of -4°C. The hazards involved in
handHng, transporting and torage of Oammable solvents have led to
definitive legislation covering all aspects o[ exposure, handling, transport-
ing, packaging and labelling. There are three approved methods for flash
pOint determination using the Abel apparatus (Fig. 15.8), Pensky-
Martens apparatus and Cleveland open cup. The open cup method is
preferred for solvents of flash point greater than 80°C. The details of
these methods may be found in BS 2000: Part 4: 1983, IP 34, 35 and 36.
The nash point of a mixture of flammable solvents can usually be taken
as that of the lowest. II any doubt exists the flash pOint of a mixture
should be determined according to one o[ the approved methods.
824 TESTING CONTROL AND QUALITY AS URANCE

There has been a review of current legislation to allow for mixtures

which have a flash point when measured using one of the approved
methods, but which do not support combustion. Any solvent ,"lith a
flash point below 21 °e is deemed highly flammable whether it supports
combustion or not. Any solvent with a flash pOint between 21 and 55°e
and which supports combustion at 55°C is deemed to be flammable.
Typically an aqueous emulsion containing a low percentage of isopro-
panolmay have a flash point between 2J and 55°C but would not support
combustion. Such a mixture does not require a 'flammable' label.

Aromatic content
Commercial grades of petroleum hydrocarbons contajn varying propor-
tions of aliphatic, naphthenic and aromatic compounds. The level of
these vary according to the source of the raw materials and the methods
of refinement. The presence of aromatic compounds leads to greater
solvency, a tendency to sweJJ the compounds used in producing litho
blankets, increased odour and toxicity. The effect of small difference
in aromatic content on the solubility of resin systems should not
be underestimated, particularly when being used to produce varnish
systems using resins with limited solubility characteristics. The aromatic
content i normally measured by aniline point or kauri-butanol value,
and details of how to carry out the e tests can be found in the section on
analysis.

Water content
The presence of water in solvents invariably leads to problems such as
poor adhesion, lack of gloss and flow in inks. Details of how to determille
water content can be found in the section on analysis.

Refractive index
The refractive index of a solvent can be used as an indication of its purity,
tbe method for determining refractive index being found in the section
on analysis.

Sulphur content
Petroleum products may be contaminated with sulphur compounds from
two routes. Sulphur compounds occur naturally in crude oils and may
not be removed by distillation or purification, and may also be inlroduced
when aromatics are reduced by treaunent with concentrated sulphuric
acid. Sulphur compounds lead to tarnishing and odour problems.
The usual test for sulphur consisrs of immersing a polished copper strip
in a known volume of the solvent at elevated temperature.
Compounds of this nature may al 0 be detected by GLe using a flame
phorometric detector which is very sensitive to sulphur compounds.
RADlATlO ·CURI G PRODUCTS 825

Odour
Odour, although a subjective parameter, is becoming of increasing con-
cern to the inkmaker, particularly in the field of food packaging.
Very simple odour tests such as allowing solvent to evaporate slowly
[rom a filter paper and then examining for residual odour may be
employed.
Odour tests are normally carried out using a panel of testers and
comparisons are made against proven standard materials u ing a simple
numerical scoring system.
Particular attention should be paid to bulk tanker deliveries, and each
tanker compartment should be tested for odour, small amounts of
residual ketones or esters in a companment leading to major odour
problems.

15.10 RADIATION-CURING PRODUCTS

Products which are polymerizable by the action of UV light, or by the
bombardment of electron in the case of EB curing, are by their very
nature reactive and unstable. Any quality control procedure or testing
method must be concerned with the unique properties of these c m-
pounds. The compounds used in radiation curing inks fall into four
categories.
• Diluents - these consist of reactive monomers generally based on the
reaction product of a polyo I, or alkoxylated polyol with an ethyleni-
cally unsaturated carboxylic acid.
• Resins - this category consists of reacrive polymers of the acrylate type,
varying in molecular weight from viscous liquids to olid r sins.
• Photo-initiators - compounds whkh form free radicals when exposed
to UV radiation.
• Synergists - a class of compounds, generally amines, which promote
the action of photo-initiators.

Diluents and resins

The role of the diluent and resin is to produce a curable fUm via a printing
proce s, therefore all the usual tests for colour, viscosity, curing speed and
odour may be applied.
Some reactive monomers are a blend of isomer whose reactivity may
vary. II may therefore be of use to examine this distribution using gas
chromatography or HPLC.
Residual acrylic acid in either the monomer or p lymer can affect the
curing rate, odour, shelf-life and skin irritancy of the product and the acid
value should therefore be determined.
The unstable nature of these reactive products require that the shelf-
life should be examined. StabiUzers such as methoxyphenol are incorp-
orated into diluent monomers to prevent premature reaction leading to
viscosity increases, gelation and darkenjng, and it may be advantageous
826 TESTING CONTROL Af'fD QUAUTY A SURANCE

to monitor the levels of these. Inhibitors can be readily determined using

HPLC. If instrumental techniques are not available the conventional
method for examining shelf-life by accelerated testing may be applied.
Accelerated testing for stability may be done by placing samples of a
standard material and the test material in an oven at temperatures ran-
ging from 30 0 to 60°C and then comparing viscosities and colour after
suitable time periods.

Photo-initiators and synergists

These classes are, with the exception of amine acrylates, invariably pure
compounds and as such may be te ted using analytical methods. infra-red
spectroscopy can be used to generate a spectrum of the sample which
may then be compared with that of the standard. The majority of these
compounds may also be examined using GLC or HPLC. All of these
techniques can be used to identify the compound and its purity. The
only other applicable test is to fabricate the initiator into a standard
vehicle and assess its performance under fixed curing conditions.

15.11 MISCELLANEOUS MATERIALS

Driers
Driers are oil-soluble metallic salts or soaps in which the metal is rendered
soluble by reaction with a long-chain acid (e.g. OClOates) or complex
mixture of acids (e.g. naphthenate ). Variation from batch to batch can
therefore be associated with the organic portion or with the type and
quantity of metal ion.
Comparison o[ the batch against a known standard using IR spectros-
copy will characterize the organic portion of the drier. The type and
quantity of metal can be accurately determined by either atomic absorp-
tion (AA) or HPLC. All these tecJmiques will be discussed in the section
covering analysis.

Plasticizers and pure chemicals

These materials can be simply tested by IR spectroscopy. The spectrum
of the test material should be exactly superimposable on that of the
standard.

Waxes
Waxes fall into two maUl categories for raw material testing purposes. The
simple chemical types such as fatty acid amines can be treated as pure
chemicals and analysis by IR spectroscopy as indicated in the previous
section. The more complex polymeric waxes (e.g. paraffin and polyethylene
waxes) cannot be identified by this analytical method and the following
more traditionaJ methods have to be employed:
INK QUAUTY CONTROL 827

• melting point determination similar to that used for resins;

• solubility characteristics at room and elevated temperatures;
• empirical end-product testing.

Others
Many miscellaneous items are often included in small amounts in print-
ing inks for specific uses, e.g. pigment-wetting aids, flow additives, anti-
foams, and these can only usually be tested by assessment again t a
standard in a typical formulation.

15.12 INK QUALITY CONTROL

Once there is absolute confidence Ihat all raw materials being used for
manufacture are within specification, then the primary objective of qual-
ity control is to ensure that alJ finished products manufactured consist-
ently conform to that specification. Confidence in raw material
specification is gained, as djscussed previou Iy, by either using quality-
assured suppliers, insistiJlg on certificates of conformity, by inspection, r
by using a combination of alJ three.
Quality control of the finished product alone is inadequate and quality
control should be applied throughout the processes used Jor manufac-
ture. This is to ensure that there is n deviation from the formula during
manufacture and that the final product wiD match the master standard in
all respects.
Testing of the product can commence as early as the mixing stage, it
often being easier and more cost effective to correct major variations in
strength or rheology while the batch is still on a mixer. If a batch is not
t sted until extensive processing has tak n plac additions of dry pig-
ments or pastes can necessitate expensive reprocessing to reach the same
stage of manufacture, whkh is obviously undesirable.
In both oil and liquid ink manufacture there are fundamentally two
types of grinding apparatus, i.e. continuous and batch mills. Continuous
mills present a problem to the quality control laboratory, since while the
sample is being tested the mill, if left running, could be producing ink
inferior to the standard, or alternatively if SLOpped will result in lost
production time. Hence a series of quick tests must initially be carried
out, followed by a more rigorous test procedure when the milling
operation is complete. Batch mills present the same problem but to a
lesser extent, it often being possible to stop batch mills whilsl testing is
being carried our. Triple-roll mills present special problems for quality
control as they have a sieving effect as well as grinding the ink. This
means that although the product may be homogeneous when entering
the mill, it could well dilfer during a run, larger particles being held back
until laler in the grinding run. The other problem with triple-roll mills
is that they are very sensitive to minor alterations in mill settings,
which could be inadvertently due LO a change in mill temperature as it
warms up, or due to an operator changing ettings when a shift changes.
As mosl of the parameters of Ihe finaJ product. ranging [rom strength to
828 TESTING CONTROL AND QUALITY A SURANCE

viscosity, depend very much on the degree of dispersion this can cause
problems with lack of homogeneity throughout the batch, and regular
tests for fineness should be carried our during a run to ensure thar the
degree of dispersion has not cbanged. Whichever fonn of olill is bing
used it is recommended that rhe ink is mixed before a final sample is
taken for testing to ensure thai it is repre entative of the whole batch
rather than being just the last portion off the mill.
Care should be taken when Cinally approving a batch of ink to ensure
that there will be no change in the ink between it being pa sed by the
quality control laboratory and the ink getting into a package. Inks are
often passed from post mixers when a further process such as loose
milling into a package will take place, possibly affecting rhe physical
characteristics of the finished ink. If a batch is to be approved but is still
to undergo another process, even if that process is only running into a
package through an 'open' mm, the sample being tested should have had
the same processing, i.e. run through an 'open' laboratOry-size mill prior
to resting. It should not be taken for granted that further processes such as
potting off wi]) have no effect on the final product every step being taken
to ensure that what is passed by rhe quality control laboratory is in fact
what reaches the customer.
In any ink quality control area, it is important to have good lighting
which should be as ncar 'white' as possible. The more complex forms of
lighting used for colour matching purposes are not necessary as all quality
coorrol procedures should be on a comparative basis of batch versus
master standard, and any distortions of hue due to an off-white light
source will be equal for both inks.
For convenience, the quality control of inks wi]) be djvided into four
sections:
• short-term ink te ting
• long-term ink testing
• press performance tests
• dry print performance tests.
Under each headjng, teslS common to both oil and liquid inks will first be
con idered and then specific test on each of the two basic types of inks
will be given.

15.13 SHORT-TERM INK TESTING

There are tests which can be carried out rapidly while the ink is still on
the mm and incJudes checks for dispersion, flow, strength, hue, opacity
(or transparency) and gloss.

Dispersion
A satisfactory level of djspersion (i.e. equal to the standard) is the
fundamental requirement for any printing ink and until this has been
achieved, any further testing is unnecessary. Disp rsion can be
 HORT-TERM INK TESTING 829
adequately checked on a fineness of grind gauge as described earlier in
this chapter.

Rheology
Before comparative prints can be obtained, to compare the strength and
hue, the test ink must be of similar rheology to the standard. Provided the
ink is not poned directly from the mill, it is not necessary at this stage to
adjust the ink accurately for any small now variations as this can be
carried out after (he ink has been taken of[ the mill.
Tests on the rheology o[ oil and liquid inks are quite di£ferent and
hence will be considered separately.

Oil inks
Rheological parameters are extremely important in controlling the qual-
ity of oil inks. While many of these parameters can be assigned limits,
others involve an assessment based purely on a comparison with an
agreed standard. This subject has been discussed fully in Chapter 13 and
only points pertinem to quality control will be further considered.

Feel on the slab

Observation of the body of an ink, the length of now and 'stringiness'
resulting [rom the rapid lifting of a flexible palene knife from the body of
an ink on a slab can provide valuable information in a shon space o[ time
concerning certain vital properties of the finished product. Comparison
against the master standard can be effected in a similar manner and
substandard products can be readily identified even by the novice ink
technician.

Ink viscosity
Ink viscosity can be measured in a variety of ways, giving values for both
the plastic viscosity of the sample and its yield value. These two values
will in many instances give information concerning the anticipated
performance of a product. A lower ink viscosity will be required for
web-offset ink, for example, than for a similar product for sheet-fed
lithographic applications.
Certain applications will also demand a high yield value particularly in
metal-decorating applications where varnishing on wet inks is desired. A
low yield value, conversely, may be paramount in overprint varnishes
where good flow and subsequently high gloss may be required.

Ink tack
There are several different instruments which will measure the tack of an
ink. None of these gives an easily defined tack or stickiness value but are a
related property which is dependent upon the design and materials o[ the
instrument itself. The instruments measure the forces needed to over-
come the resistance of an ink film to splitting between two m ving
rollers. This simulate the ink-film splitting which occurs during the
distribution of ink on a printing press roller train.
830 TESTING CONTROL AND QUALITY A SURANCE

The most usual insrrument is the Tack-a-Scope tackmerer (Fig. 15.9),

which can be found in the laboratories of the larger primer as well as the
inkmaker. This instrument has the optional advantage of u ing a chart
recorder. This gives a permanent record of the results obtained. A further
use is checking tack stability over a set period of time.
Other instruments include the Thwing Albert Inkometer and rhe
Churchill tack meter. Precise tack values are very important in obtaining
the best results in multicolour wet-an-wet applications where the precise
control of ink trapping is essential.

Ink flow
Ink flow has become a more significant test in recent years with the
introduction of much higher pre s speeds, where transfer is important
but the ink must not fly or drip from the duct. Plow characterisrics can be
measured comparatively by way of reference to a standard. but care must
be taken if using this method as freshly manufactured ink may change
considerably on standing, even overnight. For higher-viscosity products.
such as sheet-fed inks, flow tests can be comparative by knHing both
samples vigorously on a bench and then visually assessing the flow from
a palette knife.
More precise tests can be carried out by applying 0.5 cc of ink, whkh
has been worked on a bench for 30 seconds, onto an inclined or vertical
plate using a narrow blade knife. By using this method numerical speci-
fications can be set for the length of flow down the glas in a 15-minute
period, and this is often used. in combination with a vertical glass or
perspex plate, as a standard test within the newsink industry.

Liquid inks
Many liquid inks wh n first sampl d from a mill will b above room
temperature and may also be quite heavily aerated. All efforts should,
therefore, be made to equalize as far as possible the batch sample to the

Fig. 15.9 The Tacl<-O-Scope tackmeter. (Courtesy of Testprint BV)

 HaRT-TERM INK TESTING 831
standard_ The viscosity can then be checked with a Ford or Zahn cup,
while the relative flow properties can be assessed by stirring the inks with
a palette knife and then comparing the flow off the knife as it is with-
drawn [rom each of the inks.

Strength, hue and opacity

The 'drawdown' technique as described earlier in this chapter is suHicient
as a quick test for comparison of the relative strength, hue and opacity of
the batch against the standard.

Gloss
Relative gloss is more dif(jcult to assess on a drawdown and there[ore
prints of the test ink against the standard should be prepared. for liquid
inks, a drawdown with a wire-wound applicator is suitable (fig. 15.10).
This technique gives an excellent comparison between two inks provided
they have similar flow. The applicators are available for manual draw-
downs. Automatic applicator models with a built-in applicator bar can be
obtained so that a consistent pressure can be applied on consecutive
prints.
Gloss may be assessed subjectively by visual means. How ver gloss
meter with the capacity for variation of the angle of incident illumin-
ation and reflected light are in common use. These provide a numerical
measurement of gloss under standardized and reproduceable conditions.
The instrumental method uses a light source which can be directed so as
to fall upon a urface to be tested at a known angle to the normal of that
surface. The reflected light is then measured by a photocell set at the
corr sponding angle to receiv th refl ct d light. The procedure i
repeated with standard gloss and matt surfaces; these are usually liIes.
The comparison of the readings i then expressed as a percentage reflect-
ance. The results obtained do not necessarily agree with those obtained
visually, since the eye will judge the relative gloss levels at a variety of
angles and will also take into account other [actors such as print imper-
fections and colour variation. Visual assessment is, therefore, also import-
ant since the printer is concerned with the total visual appeal of his
finished print. Discrepancies in the gloss level of the balch sample could
be caused by a misweigh of one or more of the ink components or, on the
continuous type of mills, it could be due to an incorrect mill setting or
condition. Hence, by checking the gloss level at this early stage, any
mechanical adjustments necessary can easily be carried out.

Additional tests on inks cured by radiation

Inks 'curing' under radiation (UV or electron beam) vary in their rate of
polymerization. The end-point, when the ink is 'dry' on the substrate, is
somewhat more speculative.
For quality control purposes, it is necessary to ascertain whether an ink
batch is curing at the same rate as the standard. This is normally carried
832 TESTING CO TROL AND QUAUTY ASSURANCE

b
Fig. 15.10 A wire-wound applicator: (a) K bars; (b) applicator.
(Courtesy of RK Print-Coat Instruments Ltd)
LONG-TERM lNl< TESTING 833
out qualitatively using a finger rub test at a given exposure time under
the source of radiation (e.g. 2 lamps at 500 fpm at a film weight of 2 m(.L).
There are certain situations, however, when it is difficult to distinguish
between cure and adhesion. An ink may be dry to the touch, but easily
removed under the finger rub test. This can best be judged by experience.
Very slow rates of polymerization can sUIl resuJt in 'chaJking' as in the
case of rapid setting conventional inks on certain absorbent stock. Cure
rate can also be measured by employing a rub test on the substrate after a
given radiation expo ure. Indeed, the toughness of a cured film, which
may reflect the degree of cure, can be checked, for example, using the
PIRA rotary rub tester. A cured ink can thus be described in terms of its
resistance to rub under given conditions.

15.14 LONG-TERM INK TESTING

The tests already outlined under section 15.13 should first be carried out
on a representative sample of the final ink. Alterations should be made as
necessary, ranging from additional pigment to increase slrength (0 dilu-
ents to lower viscosity.
GeneraUy if an ink is laO high in lack/viscosity then a diluent addition
of a solvent or oil should be added to lower the solids concentration of the
ink. If the tack/viscosity is low then it shouJd be raised by the addition of
more solids, be it pigment in the form of an extender, or resin in the form
of powder or varnish. Generally, batches will be formulated so that the
former scenario is experienced rather than the latter, as it is far easier to
lower the tack/viscosity of a product than to raise it.
If the ink has a flow whjch is too short, this again suggests that the
solids content is too high and a further addition of varnish/oil is normally
used. If the flow is too long then the ink can be bodied by increasing the
solids content or, if it is undesirable to raise the viscosity further, a
rheological shortening agent can be u ed.
A more precise strength te t can be canied out by asses ing the batch
against the standard using tJ1e bleach technique as described in section
15.3. Additionally, both oil and liquid inks can be tested for stability on
ageing. This test is usuaIJy accelerated by placing lhe inks in a well-sealed
container in an oven preset at between 35 and 50° C. After 24 hours the
sample is allowed to cool to room temperature and checked for any
separation and/or loss in flow characteristics compared with the bulk.
A comparative drawdown of the heated and unheated sample is then
prepared and examined for changes in strength and hue. Caution should
be exercised with this test, however, as some pigment-vehicle inter-
actions can cause a shade change to occur at these temperatures over a
period of time.

Specific tests on oil inks

Drying time
The drying time of an oleoresinous ink i important to know for reasons
oJ both press stability and print drying. Optimum drying rat S are a
834 TESTING CONTROL AND QUALITY A SURANCE

compromise between these two extremes and will obviously depend

upon the nature of the particular application. Before discussing methods
available to measure ink drying times, it is important to note that this
term refers to the lime taken to dry a print of a particular ink fiJm
thickness on a certain substrate (e.g. paper/board/glass).
Factor affecting drying time of an oleoresinous system include:
• temperature
• humidity
• availability of atmospheric oxygen - a single laboratory print Jeft on a
workbench will have a greater access to the atmosphere than a pro-
duction situation of printed stacks of varying pile heights.
In carrying Out a manual assessment care should be taken to ensure all
fingers are dean, as a build-up of deposit on the fingers can everelyaffect
the test. In cases where powdering of the ink may be mistaken for drying
failure, the print should be warmed in an oven for a few minutes. This
will aid drying, but have little or no effect on powdering. For quality
control purposes, one rarely needs to know the exact time of drying, but
merely whether or not the print is dry within a cerrain lime. However,
since atmospheric conditions can vary quite widely, it is always impera-
tive to carry out the test against a standard ink.

Automatic drying time testers

Exact measurements of drying times on prints is obviously both time
consuming and laborious, and one to which automation has been
applied. Unfortunately, no equipm nt xactJy reproduces a real situation.
AutOmatic test equipment can only be used to give relative drying times.

PIRA ink drying time tester

This equipment was described in a paper by Coupe entitled 'Ink and paper
testing' in the Proceedings of the PArRA Letterpress Conference, 1955.
Modirications to this equipment are currently available in other similar
forms.
The equipment utilizes a glass upon which is printed the ink in ques-
tion (or the standard) and which is subsequently rotated. A track is scored
in the wet film using a metal stylus and as the ink dries, the smooth
track pattern changes to a tearing pattern. At this pOint where the track
changes its pattern, the ink is designated dry, and the time required to
reach this point is classed as the drying time.
The main drawbacks to methods based on glass plate prints are the
reproducibility (which can be poor) and the remoteness of the system
from behaviour on absorbent substrates. It is, however, a rapid method
suitable for quality control purposes and gives information concerning
the ink drying time on non-absorbent surfaces such as rollers.

IGT drying time recorder

The IGT drying time rec rder measures the drying lime of an ink or the
suitability of a particular paper for a given ink in two way. The first is the
set-off al preselected time intervals over a period of up to 24 hours and
LO G-TERM INK TESTING 835
the second is related inasmuch as one examines the ink removed during
the dwell time of a pressing-bar, over the same period. Various ink film
weights can be examined on any paper. Whenever the moving bar is in
contact with th wet ink (jIm either disturbance of the test strip is
observed or set-off on a backing strip is noticed.

BK drying time recorder

The BK drying time recorder is very similar in lhe method of measure-
ment to the PIRA drying time tester. utilizing glass strips upon which the
ink is printed. The glass strips are placed on a flat bed down which six
hemispherical-ended needles travel, in a time between 6 and 24 hours.
The recorder defines the following stages in the drying process.
• The first stage is a pear-shaped impression corresponding 10 the time
taken for evaporation of the solvent.
• The second stage is the cutting of a continuous track, corresponding to
a sol-gel transition.
• The third stage is an interrupted track corresponding to the surface-dry
time.
• A fourth stage is where the needle no longer penetrates the film,
corresponding to the final drying time.
The advantages and disadvantages of this system are similar to those
given for the PIRA drying time tester.

Setting time
Ink setting time is governed by the choice of paper as well as the ink fiJm
weight used. A control should thus be run under similar conditions. An
approximate qualitative test for setTing time can be carried OUI' by careful
finger asse sment checking surfac tackin ss. Il is also noteworthy that
setling is olten accompanied by a change in surlace gloss and the wet
glossy appearance after initial printing is reduced as the ink sets.
A more quantitative test utilizes a 'back-up' technique on a laboralOry
pro fing press or printing unit. Back-up in this context refer t the
process whereby an extra sheet of paper is applied to the surface of the
printed sub trate and run through the press again. Wet ink wiu be
transferred to the second sheet of paper. The time interval between
printing and backing lip is kept constant for a series of prints (e.g. 1, 3,
and 5 minutes lor quick-setting inks) and the amount of ink transferred
to the backing paper is inversely related to the setting time. The time
required lor no back-transfer o( ink is thu a function of the setting rate
and gives an indication of the 'back-up' time when printing the reverse of
the she l.

Ink fly
The problem of ink 'misting' or 'flying' in the vicinity of a press is a
serious nuisance which can largely be overcome by (ormulation changes.
This phenomenon is often a function of the nature of the vehicle, but will
836 TESTING CONTROL Af'fD QUALITY A SURANCE

obviously also depend upon a multitude of other parameters, e.g. pig-

ment, extender type and level.
It is comparatively easy to monitor ink ny in the laboratory on a given
batch of ink, given a two-roller system. It is preferable 0 have one roBer
oscillating to reduce the tendency of the ink to form ridges. A useful piece
of equipment in the laboratory for this purpose is the tackmeter or
inkometer.
Comparisons between different inks can be undertaken quantitatively
by either:

• weighing the weight increase per unit area on the collection device;
• visually assessing the droplet distribution per unit area.
Any test for ink ny ideally should try to reproduce the most severe
conrutions likely 10 occur in the field on a given press. The parameters
for consideration should thus be the follOWing.

• Ink film thickness on the roller surface - generally ink ny will increase
as Ihe ink film thickness increases.
• The temperature of the roUer surfaces - certain inks wiJI dramatically
change their rheological properties as the temperature changes and will
generally decrease in viscosity as the temperature rises. As modern
roller temperatures may be appreciably above the surrounding air
temperature, this factor should be remembered when assessing ink fly.
• RoBer speed - iis important to remember that ink misting primarily
occurs in tbe ink-roller chain. It is worth noting that speed of separ-
ation is higher as the roller diameter decreases, hence a greater ten-
dency for ink fly will arise on small diameter high speed rollers such as
those in an inkjng-roller chain.
• Ink consist ncy or rh ology - a low bodi d ink will show a predictably
greater tendency to mist than a high bodied ink.
• Siting of the target - in measuring misting levels of a variety of inks, it is
important to site the target at a point which reflects this level. On a
tackmeter for instance, a convenient point is at the same height a the
nip between the two rollers, on the rear wall of the measuring cham-
ber. It is perhaps surprising that the position of maximum fly should be
on the opposIte side to the parting nip of the rollers, wh re the fila-
ments are formed and broken: the explanation appears to be that
particles are carried around in a cloud which follows the roller surface
until it collides with the cloud following the other roller which will be
at this very position. It is also important to note that a percentage of
heavy droplets, which are often termed 'sling', wUl be formed on the
base oJ tbe measuring chamber. Although this mayor may not be a
nuisance in the same way as ink mist is (depenrung upon the design of
the press), it may result in the necessity for constant press cleaning, and
does not lead to a clean operation. This should also be monitored
carefully.
• The complete endo ure of the test roller system i important to min-
imize operator exposure to ink mists, by restricting the spread of
particles, and to exclude draughts which could affect reproducibility.
PRESS P RFORMANCE TEST 837
Specific tests on liquid inks
The drying rate of I"iquid inks is a critical factor since an ink which is too
fa t drying will cause printing problems, whereas a slow drying ink could
give ri e to excessive olvent retention. Also, an incorrect solvent balance,
especially in a system where two or more resins are present, could lead to
premature precipitation 01 one 01 the resins leading to poor printability
and a drop in the overall gloss level. Tests on the drying rate can be
carried out by drawing down side by side the test and standard inks on a
grinding gauge and comparing the relative evaporation rates at different
film thickne ses.

Pigment settling
This cannot easily be tested except by examination of a batch sample over
a long period of time which is, of course, completely impracticable.
However, the batch sample should be allowed to remain undisturbed
lor 24 or 48 hours belore being examined for any signs of either a
supernatant layer or pig ent sediment on the bottom of the can.

Water-based inks
These should additionally be checked for foaming and pH variation. The
batch and standard inks are either stirred or shaken for a set period of
time and compared for the degree of foaming. Both inks are then allowed
to stand [or a few hours to check on the ea e of dispersability o[ any I am
formed on the above test. Excessive use o[ anti-foams should be avoided
as they can lead to printability problems and stereo sweHing.
The pH of a conventional water-based ink (i.e. one in which an acidic
r sin is olubilized with ammonia or an amine) is an important [actor
since if it i too high the ink would be too slow drying due to the excess of
amine present. Alternative[y, if the pH is too low, then large viscosity
fluctuations can occur and the resin can even be precipitated out of
solution. A check on the pH o[ the finished ink can simply be carried
out by using a standard pH meter.

15.15 PRESS PERFORMANCE TESTS

Oil inks
The u[timate test [or examining the performance o[ an ink is to run it on a
pre s, and there is no laboratory te t which can precisely predict the
performance on press. However, there are some tests which can be
carried out in the laborawry to predict, to a certain extent, the behaviour
of an ink on press, although many would not normally be within the
scope o[ a normal quality control specification. Quality comrol of most
printing inks will rely on the principle that i[ the standard ink performs
satisfactorily on press, and the batch being tested is made with the same
materials, by the same method, and to the same pecification, then it can
be reasonably assumed that this test batch will also perform satisfactorily.
Some inkmakers may have the 'facility to test products on presses
838 TESTING CO TROL AND QUALITY ASSURANCE

identical or similar to those ro be used by its customers, and in this case

may choose either to spot check, or test every batch for press perform-
ance. However, this will involve considerable test time and expense and
will tend to be limited to products which are extremely critical and where
larger contributions canjustHy the resources needed.
Those inkmakers which do not have press facilities available wiJI use
one or more laboratory instruments and tests in an attempt to predict the
final press performance of a batch of ink. These tests which range from
those whid1 will measure such parameters as transfer of the ink, to the
water-uptake properties. the complexity of the instruments being used
will be dependent on the resources available 10 the inkmaker.
The lGT printability tester (Fig. 15.11) exhibits certain features of the
printing operation and is well used for both ink and paper testing. It
consists of an ink distribution system and a print unit (Fig. 15.12). This
0
printing unit is comprised of a 150 secror of impression cylinder suitably
covered, on which a I-inch (25-mm) wide strip of paper is placed. Ink is
applied from the distributor to the rim of the black-centred wheels, which
act as the rotary printing plates. Power is furnished by a pendulum or
spring and in later models by electrical means. One has thus the choice of
either having a constant speed or varying speed impression. In the case
of the latter, the paper strip can be graduated along its length in terms of
speed at given pressures, which themselves can be varied.

Pig. 15.11 Distribution unit of the IGT printability tester.

(Courtesy of Testprint BV)
PRE S P RFORMANCE TEST 839

Fig. 15.12 The print unit of the IGT printability tester. (Courtesy
of Testprint BV)

Under controlled conditions of humidity and temperature, this instTll-

ment can be u ed to xamine a whole vari ty of printing properties
together with comparisons of ink and paper. The manufacturer's infor-
mation is recommended for a fuJI list of these applications.
One property missing from the TGT list involves the lithographic prop-
erties associated with water-ink interfacial phenomena. Although the
IGT is not designed specifically for examining water-ink relationships,
there are pieces of equipment available to do this. The Pope & Gray Litho-
Break Tester is perhaps the most simple and commonly u d equipment,
whereas the PriHbau Lithomat can also be extended to measure changes
in tack, film thickness and water take-up.
The Lithomat consists of a steel roller that revolves in the fountain
solution. This roller rotales a rubber roller whjch presses against another
steel roller that lie above it. Both teel rollers oscillate to ensure an even
ink distribution. The effect of the fount solution on tack and thickness of
the ink on the rolJers can be assessed by combination with a graphome-
tronic instrument which measures the tack as the force exerted 00 a roller
driven by the inked roller. The change in tack and film thickness is
recorded with time. The final water content take-up by the ink is meas-
ured by removing a sample from the centre Tllbber roller and analysing by
the Karl-Fisher method.
The Tack-a-Scope tackmeter can also be used to assess the lithographic
characteristic of an ink, in many ways being similar to the Lithomal.
840 TESTING CONTROL Arm QUALITY A SURANCE

The Tack-O-Scope has the facility to apply fountajn solution to the main-
metal roller using a shallow tray and fount roller, the effect of adrung or
removing the fount being monitored using a chart or graphometronic
recorder.
For the small inkmaker who does nOl have such sophisticated equip-
ment available there is always the manual Surland type of water take-up
test, which is carried out by mixing fount solution into a quantity of ink
in a beaker, and measuring the rate at which the fount is taken up. This
may be a relatively primitive method of investigating the lithographic
properties of an ink bu t can give usefu I clues where somethjng is radkaJly
wrong with the lithographic characteristics.
As one moves away from conventional oleoresinous systems to. for
example. ester chemjstry, as in UV curing inks. it is important to ascertain
the effect the ink may have on the rollers, blankets and plates of the press.
While this is of paramount importance in introducing a new product, it is
nOt necessary for routine quality control testing. The assumptions made
earlier should hold.

Liquid inks
The majority of solvent-based inks are primed by either the nexo or
gravure processes, although screen printing is also important in certain
specific areas. Whatever the process. the performance of the ink on the
pre s is obviously a critical factor and is one which should be carefully
examined in the quality control procedure.
Gravure printing can be carried out in the laboratory by using a scaled
down version of the production model. the rollers being either mechan-
ically driven or even simply turned by hand. AJternatively, small gravure
proofing presses are avrolable (Fig. 15.13) and these, although ideal for
quality control work. sufI r from th rusadvantage that only a singl print
can be obtained before eaning up the rollers. AJI inks should be thinned
where possible with the recommended reducers. although the use of
slower evaporating solvents may be necessary to obtain clean prints at
the relatively low printing speeds obtained in the laboratory.
Before any printing is undertaken, the proposed end use of the ink
should be carefully considered and the following questions raised.

• Is the ink expected to be used for priming solids, tone areas, or a

combination of both?
• Is the ink part of a process set?
• Is the ink to be used for surface or reverse printing?
• For coloured inks: is the ink to be used in conjunction with a white?
Conversely, for whites: is there a typical coloured ink also used when
printing this white?
• What substrate(s) should the ink be printed on?
The answers to these queries will provide the quality control tester with a
guide in assessing the press performance of the test ink. II is important in
thjs context that any inks to be used in conjunction with the te t ink (e.g.
for a process ink, the remainjng colours of the process set) should also be
PRESS P RFORMANCE TEST 841

Fig. 15.13 A gravure proofer. (Courtesy of RK Print-Coat

Instruments Ltd)

printed so that the effecl of superimposition of the inks can be carefully

examined.
Other problems a socia ted with the gravure proce s, such as examin-
ation Jor cylinder wear and scumming, are more difficult to reproduce as
a standard procedure and a great reliance must be placed here on having
a good ink formulation. Several tests have been designed for measuring
cylinder wear, but perhaps the most common uses a small trough con-
taining the test ink in which is immersed a glass slide coated with a thin
layer of chromium. A weighted arm with a brass weight on the end is
then passed back and forth in the ink over the chrome-plated slide for a
set time period. The test is then repeated using a control ink on a second
chrome slide. A comparison between the degree of chrome abrasion on the
trial and control ink slides is an indication of the wear potential of the ink.
Scumming. i.e. the formation of an ink film in non-image areas, can
result either from an incorrect press setting or from an ink faull. Assum-
ing correct mechanical adjustments on the press scumming can still be
evident, in certain extreme cases from the beginning of a print run, and
this is usually due to oversize particles in the ink causing the doctor blade
to lift fractionally away from the cylinder surface. This fault should have
been eliminated, of course, in the initial quality control, but in other cases
where scumming is obtained only after a longer press run the cause is
normally more difficult to ascertain and hence to rectify as a standard
procedure.
In fJexography, the film weight applied is a much less readily defined
factOr than in the gravure proce s. This is because the film weight carried
can be varied simply by altering. for example, the pressure between the
842 TESTING CONTROL Af'fD QUALITY A URANCE

rollers, the speed of running, the ink viscosity, as well as varying rhe
screen size of the anilox and the Shore hardness of the rubber rollers.
Hence, Oexographic proofing presses have a limHed value, and a greater
reliance is usually placed on the more versatile anilox hand roller (Fig.
15.14).
This consists of an engraved roller (termed the anilox) and a rubber
roller. A print is obtained by placing the ink in the nip of the rollers and
rolling the whole unit across rhe substrate. The film weighr applied by this
meth d hould roughly parallel a typical comm rcial []exographic print
and can be varied by:
• Altering the ink viscosity;
• Applying more or less pressure on rhe roll out;
• Using a different screen size anilox roller;
• Using a different Shore hardness rubber roller.
Some raw materials, especially some solvents and anti-foams, can cause
swelling of the stereos leading to poor printability and in certain cases it
lTlay be necessary to specify the type of stereo material (i.e. natural or a
synthetic rubber or a photopolymer plate) to be used with a particular
ink. Tests can easily be carried out by immersing a sample of che Slere
material in the ink for a short time and measuring the thickness of the
sample before and after the test using a micromerer gauge and also the
hardness before and after immersion.

Screen inks
Before the press behaviour of a screen ink can be examined, the formu-
lation of the ink should be checked since inks for this proce S vary
considerably in composition depending, for example, on the substrate
to b print d and th drying facilities availabl . Generally, screen inks can
be dried by evaporation. absorption, oxidation, polymerization or more
often by a combination f these methods. Inks drying mainly by oxidation

Fig. 15.14 An anilox roller. (Courtesy of Weller Patents

Development)
DRY PRJ T PERFORMAN E TESTS 843
(e.g. by the incorporation of a drier) resemble a letterpress ink in many
ways and would normally be manufactured in a similar manJler (e.g.
triple-roll mill). Alternatively, an in.k drying mainly by evaporation (e.g.
using a cellulose ester in solvent) is more akin to a high viscosity liquid
ink and would normally be manufactured on conventional liquid ink
milling equipment. Hence, in assessing the press performance of screen
inks, many of the difficulties encountered may be a direct result of the
formulation type and the reader should then refer to the appropriate
ection of this chapter.
In the actual printing of screen inks, the best results are normally
obtained when the inks are soft and bUtlery to alJow easy passage
through the screen and short and thixotropic to give smooth lay and
sharp definition. Testing can be carried out in the laboratory simply by
drawing a squeegee over a hand screen, the film weight applied being
varied by the use of different mesh screens (Fig. 15.15).

15.16 DRY PRINT PERFORMANCE TESTS

Paper and board
Although inks for paper and board printing may have different proper-
ties, many tests carried ut on the final print have similar performance
ratings.
Properties such as Jightfastness, scuff resistance, soap and alkali resist-
ance, etc. on the printed substrate are generally similarly te ted in both
paper and board printing and for convenience are treated together.

Fig. 15.15 A screen proofer.

844 TESTING CO TROL Arm QUAUTY A SURANCE

Lightfastness
It is not normally necessary to check th ]jghtfastness of each individual
ink except for ca e where contamination may cause a speofic problem.
Details of the test have already been described earlier in this chapter and
are also covered in BS 4321 :1969.

Chemica1resistance tests
Several chemical resistance tests on finished prints have already been
covered in the testing o[ pigments (section 15.3). These are:
• soap and detergent resistance
• alkali resistance
• acid resistance
• wax resistance.

Specific test methods

The following specHic resistance tests are all fully described in BS 4321:
1969 and are all self-explanatory.
Determination of resistance of printing ink and print to cheese: Test
Method 8, p. 27. Determination of resi tance of printing ink and prints to
edible oils and fats: Test Method 9, p. 29. Determination of resistance of
printing inks and prints to spices: Test Method II, p. 34. Determination of
resistance of printing inks and prints to water: Test Method 3, p. 17.
Determination of resistance of printing inks and prints to detergent: Test
Method 7, p. 25. Determination of resistance of printing inks and prints to
solvents: Test Method 4, p. 19.

Abrasion resistance
The resistance to abrasion is an important aspeCT of many printed articles,
particularly on paper and board. Demands for improved abrasion resist-
ance continue while the speed of packaging rises, giving greater possibil-
ity of severe scuUing from other printing surfaces or from contact with
the production machinery itself. In certain cases (e.g. oleoresinous inks
where excessive spray powder may give an abrasive surface) the package
in itself can have a very rough surface and should be examined prior to
any test. Failure on laboratOry testing may result from incorrect or over
use of anti-set-off powders. It is also worth noting that problems may
arise in certain packaging areas where spillage o[ the conteJ1lS can lead to
an abrasive material being trapped betwe n two areas of print. Any
laboratory test should be designed to cover the severest conditions likely
to be encountered.
Several instruments are available (or checking rub resistance o[ prints
and three such instruments are described in the British Standard Methods
for Measuring the Rub Resistalice of Print, BS 311 0: 1959.
While the simplest method described in BS 3110 is the Procter &
Gamble tester, it is not used much outside the soap and detergent pack
area for which it is specified. This rub tester is composed of a frame acting
as a clamp to a print sample which is held against the base of the test
instrument. The frame also acts as a guide to a block on which is attached
a second piece of print in surface contact with the first sample. Hand
DRY PRI T PERFORMANCE TESTS 845

movement of the block for a number of times at a steady speed (around

1 ftls or 30 cm/s) gives direct comparisons between prints.
The Sutherland instrument and the PIRA instrument (Fig. 15.16) are
discussed in BS 3110 which is recommended for further details.
It is important that the age of prints being tested is known. With
conventional oleoresinous prints a week is the normal time allowed to
elapse before rub testing. Even with radiation-cured prints 24--48 hours
normally elapse before testing as inks dry to the touch can change with
respect to their rub re istance properties within the e Limes.

Crease resistance
In board printing the finished print is often required to be formed into a
carton (e.g. a cigarette packet) and the ink must therefore have sufficient
flexibility to prevent cracking when the board is folded. This test is
particularly important where higher film weights of ink (and/or over-
lacquer) are carried as in gravure printing. Ideally, the test shou Id be
conducted using an apparatus such as the PIRA carton board crease tester
(Fig. 15.17) which reproduces quite accurately production conditions. A
cruder test can be carried out by either folding the board back and forth
through an angle of 180" until the first igns of ink split is apparent or by

Pig. 15.16 A PIRA rub tester. (Courtesy of PIRA)

846 TESTING CONTROL AND QUALITY A SURANCE

folding the board once rhrough an angle of 180 0 and lightly scuffing the
surface of the fold with the Hnger nail.

Film printing
Immediate adhesion of the ink to the substrare should normally be
obrained in film prinring to prevent set-off occurring in the reel. All
film prinrs should be tested for adhesion, flexibility, scuff, slip and set-
of( as outlined below.

Scotch tape adhesion

A lengrh of standard Scotch tape is placed over the printed areas,
smoothed down with the ball of the finger and then pulled back lowly
for half of irs length and snatched back rapidly for the remainder. The
print is examined for ink removal and the tape [or any ink transfer. Care
should be exercised with coated films so as not to remove the base coating
and so obtain a completely false result.

Flexibility
The printed fjJm is held firmly between the thumb and forefingers of each
hand with about half an inch (13 mm) in between the hands and flexed
10 times. The print is rhen examined for any ink removal.

Fig. 15.17 A PIRA carton board crease tester. (Courtesy of PIRA)

DRY PRINT PERFORMANCE TESTS 847

Scuff
The print is repeatedly scuffed with Ihe back of the fingernail, then
examined for ink abrasion.

Slip
The friction of fUm to film and film to metal are important factors in
processing packaging materials. J( the coefficient of friclion is too high,
this can result in:
• movemenl of the packaging material over fixed elements becoming
di[[jcult;
• problems in controlling web tension;
• print which is stacked becomes 'adhered' and this can give rise to
blocking;
• generalion of a high static e£ectrical change;
• excessive creasing during packaging.
Alternatively, if the coefficient of friction is too low, then:
• guiding rollers on the packaging line will not rolate and this results in
irregular flow;
• web tension is diWcuh to adjust;
• Finished packages tend to slip during transportation.
Two types of friction can be measured, i.e. static and dynamic friction. Tn
static friction there is no relative velocity between the boundaries of
the materials and the static coeHicient of friction can be described as the
energy required to break down the localized 'adhesion' between the
substrat s in order to allow tbem to mov freely. The static coefficient
of friction can be measured using the tilting plane where a print is
fastened face upwards to the plane and a metal sled is then placed on
the film. The plane is slowly raised until the sled slowly and evenly moves
down the plane and the tangent of tbe angle of inclination is equal 10 the
static coeHicient of friction. Alternatively, the static coefficient of friction
between two printed surfaces can be measured by allacbjng a print 10 the
metal sled and repeating the test as above.
Dynamic frinion measures the friction between two materials which
have a relative velocity ro each other and is caused by molecules on the
surface of the two materials colliding. These collisions are imperfectly
elastic and hence energy is consumed. The dynamic friction can be
measured on a flat bed by drawing a sled across a print at an even velocity
(Fig. 15.18) or alternatively by the pendulum method where the print is
held in a fixed position and the pendulum is rotated.

Blocking
The print immediately after drying is placed in contact with the back of
the same film and subjected to a pressure of between 750 and 3000 psi for
16 bours in a hydraulic ram to simulate conditions produced in a reel of
printed film (Fig. 15.19). The pieces of film are then slowly pulled apan
and the te t ink examined (or degree of cling relative ro the standard and
any signs of ink transfer to the back of the plain film. Additionally, the
848 TESTING CONTROL Arm QUALITY A SURANCE

Fig. 15.18 Daventest dynamic friction tester. (Courtesy of

Daventest Ltd)

blocking tests can be repeated at elevated t mperarure if th facility

exist for heating the platens of the hydraulic ram. This test can be used
to simulate the effect of entering the re-red stage while still warm or
aliernatively as a guide when film i used for packaging hot goods. [n
carrying out these tests, care must be taken in selecting a realistic rem-
perature, time and pressures.

Additional tests for film

The following tests are dependel1l on the type of ink and its end applica-
tion and need only be checked where stated in the ink specification.

Heat resistance
Heat-resistance inks are necessary where the film is printed in the heat
seal area so as to prevent ink pick-orr onro the heat seal jaws. An area of
print in contact with virgin foil is placed between the heat seal jaws
(preset at the heat seal temperature of the film under test) for a short
time dwell (usually I second). After cooling, the print and foil are pulled
apart and examined for ink pick-orr and ink Iran fer on to the foil. Ideally,
the prinl and foil should show no cling and there should be no ink
tran fer from the print to the foil. A further and more stringent test
involves folding the print inside the foil so that the two ink surfaces are
in contact. The heat resistance test i then repeated and when the ink
surface are separated there should be no cling (Fig. 15.20).

Adhesive lamination
[f nece sary the adhesive is mixed according to the manufacturer'
instructions and reduced to a suitable printing viscosity with solvent.
DRY PRINT PERFORMAN E TESTS 849

Fig. 15.19 Hydraulic ram for blocking tests. (Courtesy of

Specac Ltd)

Th print under test is s t asid for 12 hours to allow for any wax S to
migrate to the surface and then coated with the adhesive. After evapor-
ation of the solvent, the tacky print is smoothed onto the laminating film
(or foil) taking care that all air bubbles are removed. In order to allow for
easy access into the laminate after the adhesive has cured, a strip oIpaper
is wedged between the substrates down one edge of the adhesive layer.
The bond strength oI the laminate is tested after 24 hours and ideally also
after 1 week, on a spring balance. Laminates for boil-in-the-bag work also
require an additional bond strength test after being immersed in boiling
water for 30 minutes.

PVdC coating
In certain types of packaging (particularly biscuits) the print is coated
with a PYdC emulsion to allow for better seal integrity on A-B and A-A
seals. To simulate this process, the PYdC emulsion is simply applied by a
wire wound applicator over the print and the adhesion checked with
Scotch tape. An additional problem can often occur after heat sealing,
when interJayer adhesion between the PYdC and the ink film or the ink
film and the base Iilm breaks down. To ensure good interlayer adhesion is
maintained the coated print is heat sealed (PYdC to PYdC) at I 35°C/0.5 sl
0.5 psi and the strength of seal tested on a spring balance as for a laminate
850 TESTING CONTROL Arm QUALITY A SURANCE

Fig. 15.20 A BCL heat sealer. (Courtesy of British Cellophane Ltd)

described in the preceding section. If poor interlayer adhesion occur a

poor seal strength will result.

Cold-seal packaging
Cold-seal packaging is being increasingly adopted to prevent heat con-
tamination of ensitive products and for faster packaging line speeds. In
this process a cold-seal adhesive is applied in line to the reverse side of the
print d film so that on re-reeling th adh siv is in direct contact with the
inks or release lacquer. It is therefore imperative that:
• no set-off of inks on to the adhesive occurs or blocking will result;
• pigments, dyes, waxes do not migrate into the cold seal and cause
discolouration and/or loss of seal;
• no chemical interaction takes place between components in the inks or
lacquer and the adhesive which might poison the adhesive.
Testing of inks and lacquers must be such as to show if any of the above
situations can occur in practice. The normal test involves both short- and
longer-term blocking tests with inkllacquer in contact with the cold-seal
adhesive and with a maximum pressure of 750 psi in the hydraulic ram.
In addition to no cling being obtained on separation of the inksllacquer
from the cold-seal adhesive, the cold-seal adhesive must be sealed to itself
to ensure no drop in seal strength is given compared with a print of
adhesive which has not been in contact with the ink/lacquer.

Deep-freeze
This test was fully described earlier under pigment quality control (see
page 813).
DRY PRJ T PERFORMANCE TESTS 851

Odour and taim

Since many liquid inks arc used on films for food packaging, it is neces-
sary that they must have low odour and do nOt taint the packaged
product. Samples of the test print, standard prim and plain film (as a
reference) are placed in separate tightly sealed containers, and heated in a
50°C oven for 1 hour. On removal the samples are tested for residual
odour while still hot, then rapidly resealed and tested again after cooling
to room temperature. A more elaborate odour test and taint test proce-
dure is given in BS 3755: 1974.

Non-curl
A few films are especially prone to curl when printed with certain rypes of
ink, causing great problems on the packaging lines. In cases where curl
could be a problem. a sample of film, printed all over with the test ink,
should be allowed to lie undisturbed on a bench. After a period of
lime (e.g. 1-2 hours), there should be no igns of the film curling at the
edges.

Grease resistance
Printed film used for packaging certain foods, e.g. potato crisps, must be
resistant to the oils used in the manufacture of these goods. A test can be
carried out by rubbing tbe printed film a standard number of times wilh a
piece of cotton wool saturaled in the appropriale oil. followed by a
flexibility test on the greased print. The print and cotton wool are then
examined for print breakdown and stain respectively.
Alternatively, the grease rub test can be carried out using the SATRA
rub tester by impregnating the felt pad with a measured amount of the oil
and then rubbing the test print for a set number of revolutions with an
agr ed pr ssure. Thi t Sl, although appar ntly more scientific. is subject
to wide variations in results due lo variability in the felt pads and an
average re Ull of three readings should be taken and compared with those
of tests on a master standard print carried out at the same time.

Foil
The testing of foil prints closely resembles the tests already outlined for
plastic films. Care should be exercised to ensure the correct type of foil is
used whelher it be, for example. the simple virgin foil (such as used on
many chocolate wraps) or the paper/board-backed foil which has a sur-
face wash (normally either nitrocellulose, shellac or vinyl) to prevent
tarnishing and to improve print receptivity.
The procedure for film printing should be adopted, where appropriate.
with the exception of the flexibility lest. For virgin foil, f1exjbility is tested
by folding the printed foil through a 180° bend and then with the thumb
and forefinger at either side of the crease the foil is rolled back and forth
through the fold ten times. The ink should then show no signs of break-
down at the fold. Paper- and bard-backed foil is examined for crease
resistance u ing te t methods already outlined for paper and board (see
page 845).
852 TESTING CONTROL AND QUALITY A SURANCE

Metal
While some of the tests already described Ior paper and board prints will
apply in the case of metal decorating inks, others are extension of these
(e.g. abrasion resistance), to take into consideration the tougher film
involved, or the subsequent processing requirements of the metal con-
tainer (e.g. high temperature process resistance), that is not possible with
other substrates.

Abrasion resistance
The tougher film formed on metal surfaces, necessitated by the abrasion
resistance requirements of the ultimate package. demands a different
approach to testing the scratch susceptibility or hardness of that film.
Several instruments are available for these tests.

Scratch testers
The apparatus consists of a weight-loaded needle arm which is placed in
contact with the panel under test. Operation of the lever gently lowers
the needle on to the tesl surface and slowly withdraws the panel so that
the needle makes a scratch on the surface. As the load on the needle
increases. a point is reached where penetration of the ink film is achieved
and the needle strikes the metal substrate. This contaCT completes an
electrical circuit as signified by the illumination of a lamp. Results are
thus expressed as the maximum weight at which the film pa ses the test.

Hardness tester
Two types of hardness tester are used in the main: these are the Sward
type and the pendulum type (DN 53-157). The Sward type utilizes a
rocker action and the hardness is described as the percentage of oscitla-
lions mad by th wheel typ rock rs on th t st plate as opposed to the
reference glass plate.
The pendulum type hardness tester is primarily used on the continent.
The apparatus consists of a triangular shaped pendulum, free to swing on
two ball point which rest on the sample. An electronic counting device
records the number of swings made by the pendulum between the 6- and
3-degree marks on the scale. Comparison is then made between the
number of swings on the glass reference plate and the number on the
test surface.

Heat resistance
This te t depends primarily upon the properties and concentration of the
pigment, but may be affected by the medium.
While knowledge of tJle pigment will provide the basis for a selection.
to b decisive, a test should always be carried out on the final print.
Routine checking on ink batches need not be necessary for reasons
outlined earlier. In the ca e of new products, however. this te t must
always be carried out where heat resistance is required.
The test itself. using a print of known film thickness on the metal to be
used, employs a stoving operation in an oven at a suitable temperature
for a realistic time. An average schedule is 150 C, for 5 minutes, but this
Q
DRY PRINT PERFORMANCE TESTS 853
can vary depending upon the type of container under test, and higher
temperatures may be encountered for example on DWl can lines. There
are spednc applications (such as solder seaming) where much higher
temperatures (e.g. 230°C) are encountered, albeit for much shoner times.
The stoved panel is tJlen compared for hue, strength and brightness
with the control panel, making due aJlowance for the wet state of the
latter.

Flexibility or elongation properties

Simple f1exibiJity tests can be carried out by bending the printed metal
substrate and examining for failure, but the method can show poor
reproducibility and often has little relationship to the stresses that the
printed substrate is subjected to in commerdal situations.
Reproduction of the actual stress situation on a tin-printing assembly
line is not easy in the laboratory, but some equipment is avaiJable that
can give inCormation on the flexibility of a printed or coated film. One
such piece of equipment is the conical bent test apparatus designed to
measure the elongation potential of coated metal panels. The panels are
damped in position and formed round the conicaJ mandrel by rotation of
the roUer Cram . The 200 mm long mandrel tapers from 37 to 3 mm
diameter along its length and after the test the panel is examined to
determine the minimum diameter at which failure by stretdling
occurred.

Adhesion
Adhesion to the metal and good interfilm adhesion on overprinted areas
are both important requirements for printing inks in metal-decorating
applications. Failure of abrasion resistance tests may indicate that insuf-
ficient adhesion to the base metal or the underlying film is the cause. The
principal test for po l' adh sion is probably a simple scuffing t st with a
hard, blunt surface. Soft films or brittle films can be disturbed relatively
easily. Poor interfilm adhesion can also be noticed in the same test as
indicated by removal of surface layers only. The use of Scotch tape can
also prove helpful, particularly for radiation-cured inks where adhesion
has not been assisted by thermal treatment. lt is particularly useful to
predict the behaviour of cured ink films on interdeck drying as poor
adhesion or interfilm adhesion can give rise to stripping of the ink from
the sheet on to subsequent blankets.

Pasteurization and processing

Film cracking, softening or adhesion failure can aJso occur during the
processing or pasteurizing of the printed metal. Where the ink is to be
used in this application, the necessary laboratory tests should be per-
formed.
Processing tests are normally carried out under steam pressure (lS psi)
for various times, depending upon the particular application. The print is
then checked for failure.
Pasteurization varies somewhat from laboratory to laboratory but an
average te t use partial immersion of the print in a water batch at 150 0 P
(66°C) for 30 minutes. Print failure tests are performed as for processing.
854 TESTING CONTROL AND QUALITY AS URANCE

It is importanr when carrying out these tests to ensure that the print bas
been treated as would the finished article. Where for instance bending or
stretching occurs this should be carried out under laboratory conditions.
Also where a base coating or lacquer is present, it is important that lh
correct material is used in conjunction wilh the ink under test.

Plastics
The test required for prints on polymeric surfaces (plastic conrainers,
etc.) lend to lie between the paper/board and metal areas.
As many of these containers hold food, the main tests are contact tests
with the material in question. These are similar to those described for
paper and board, and are carried out by immersing the printed substrate
in the material for a certain period and testing for film weakness or
pigment bleed.

Ink adhesion and scuff resistance

Ink adhesion is tested using Scotch tape in a similar manner to that
described for film inks (page 846). While complete resistance to Scotch
tape may not be required, the inks must possess good scuff resistance.
A fingernail te t as performed for film inks generally suffices.

15.17 STATISTICAL PROCESS CONTROL

As mentioned in the previous section, statistical process control (SPC) is
an important tool in the improvement of manufacturing industrial pro-
cedures. SPC is a technique for monitoring the quality of processes rather
lhan products, which makes it diff rent from lh majority of oth r quality
improvement strategies.
Failure to accept SPC as a quality tool is often due to lhe perception lhat
a detailed knowledge of mathematics is required. This is not so, for
although the derivation of lhe theory is complex, it is only necessary to
have a basic understanding of mathematics to apply lhe technique in
practice.
Unlike traditional quality control measurements, SPC monitors the
performance and the stability of the manufacturing process, and is not
dependent upon final product inspection. This means that it can be used
as a tool for improving lhe quality of the manufacturing process and
thereby reduce lhe time spent on final product inspection. 11 SPC is
being used to monitor and adjust processes within manufacture, the
products are more likely to meet specification with less final adjustment
required. Furthermore. lhere will be a smaller incidence of product
rejection. This gives a saving in time and materials which fully justifies
the time required to set up and use such a system.
Before considering the details of applying statistical process control
techniques, it is useful to have an understanding of the elements which
go to make up a process. A proce s usually contains five distinct variable
elements: