Note: Descriptions are shown in the official language in which they were submitted.
CA 02337335 2001-O1-12
WO 00/07010 PCTlSE99/01237
Method to predict and/or control the strength properties of a foil-like
material
The present invention relates to a method to predict and/or control the
strength properties of a foil-like material in relation to machinery and
processes for
manufacturing this material.
It especially relates to properties with regard to the tensile strength and
tensile strength index (tensile index) of the foil-like material in question.
The invention is
based on the use of and access to statistical information regarding local
mechanical
properties, especially local tensile stiffness and local tensile stiffness
index. When the
foil-like material is paper, a specially interesting characteristic of the
invention is that it
makes it possible to achieve a suitable trade off between the tensile strength
of the paper
and the variation in other important measured properties of paper.
Unfortunately, it is not always possible to simultaneously achieve the sought
after minimum variations or the desired levels of key parameters, such as
local tensile
stiffness and local basis weight. For example, the tensile strength is not
always greatest
when variations in basis weight are minimised. In practice, this means that a
compromise
must be made. In many processes for manufacturing paper, it would thus be of
greatest
importance if the mechanical strength of the paper could be frequently
predicted to allow
rapid control measures to be taken and also to make it possible to achieve the
desired
quality in the final product.
Today, the tensile strength of paper is predicted by means of models that
are based on information regarding the mean value of modules of elasticity
that are
measured in different directions in the material. These modules of elasticity
are
increasingly often estimated on the basis of measurements that give the phase
velocity of
ultrasound in the material. Predicting tensile strength based on information
about the
spatial variations in basis weight of paper, such as the formation number,
have also been
tested in practice but have had only limited success.
The lack of success regarding the desired prediction is not surprising in the
light of the discoveries on which the present invention is based. The spatial
resolution,
which is normally 10 cm for the ultrasound measurements that are carried out
on moving
paper webs, is not sufficient in relation to the resolution that, according to
the invention,
would have been required. It has been shown that a spatial resolution in the
order of
millimetres gives a very good result. In addition, theoretical calculations
and
CA 02337335 2001-O1-12
WO 00/07010 PCT/SE99/01237
2
measurements show that the relation between the measured wave velocity for
ultrasound
and the module of elasticity of the material in question can vary when the
process
conditions change, such as variations in the beating of the pulp; see Fig. I.
It should be
seen that the variation shown in Fig. 1 is significant in connection with
control
applications of the type referred to. When one attempts to use data that give
variations in
local basis weight to achieve the desired prediction, one encounters the
problem that the
local basis weight does not always reflect the important local mechanical
properties.
In accordance with the principles of the invention it now becomes possible
to realize a method of the type mentioned in the introductian by the
measurements being
performed locally on the foil-like material and with a high spatial resolution
- from 20
mm and less - to obtain a mean value and variation regarding at least one
local
mechanical property of the foil-like material, whereby the strength value or
control signal
worked out from the measurement results obtained is used to achieve the
strength
prediction and/or the process and quality control desired.
Different advantageous embodiments of the new method are evident from
the non-independent claims 2-8. In this way, the said measurements can be of
the direct
and/or indirect measurement type. In particular, an adjustment of the strength
value
and/or control signal is made as a result of structural differences in the
foil-like material
based on local measurements of at least one further mechanical property,
preferably
bending stiffness. The measurements are preferably carried out on a mm-scale
with a
high frequency of repetition in relation to the dynamics of the process. The
foil-like
material comprises paper especially. At least one of the Local mechanical
properties is
measured in what is per se a known manner, especially by means of the
arrangement
described in US 5 361 638.
The invention will now be described in more detail below with reference to
the enclosed drawings wherein:
Fig. 1 shows in diagram form the relation between a standard test and an
ultrasound test on paper at different indices of tensional stiffness and
beating levels for
pulp;
Fig. 2 shows in diagram form the relation between mean values of index of
tensional stiffness and index of tensional strength for different levels of
beating;
CA 02337335 2001-O1-12
WO 00/07010 PCT/SE99/01237
3
Fig. 3 shows in diagram form the relation between measured index of
tensional strength and, according to the invention, the predicted index of
tensional
strength for different levels of beating;
Fig. 4 shows in diagram form the relation between strength quotient and
stiffness quotient at varying conditions of orientation (81 samples, 3 forming
units and 3
different pulps).
The new observation on which the present invention is based is that
successful prediction of the strength properties of paper, for example tensile
strength,
requires the use of data for both the mean value and the statistical
variations in the local
IO tensile stiffness (lts) and/or the local tensile stiffness index (ltsi)
measured with a spatial
resolution at the millimetre level.
The statistical variations can be expressed by the measured maximum and
minimum values, the standard deviation a, the coefficient of variation CV or a
number
of other parameters that can be obtained from the statistical frequency
function for the
measured local mechanical properties. (Information regarding the local
variations in
tensile stiffness index can be obtained from the local tensile stiffness data
if the local
basis weights have also been measured at the same locations).
Until very recently no measurement method has been available that has
been able to provide the type of local data that are required for the said
prediction.
Fortunately, it has now become possible to obtain the required data by
utilising a new
arrangement that is described in US 5 361 638.
As already mentioned, the type of predictive model that we have found to
be very useful is based on information about both mean values and statistical
variations in
local tensile stiffness and/or local tensile stiffness index measured on a
millimetre scale.
In accordance with the principles of the invention, the mean value m of the
tensile stiffness and the tensile stiffness index firstly give a value of the
upper limit of the
tensile strength that can be achieved. This upper limit can only be achieved
if the paper
has a completely even structure, i.e. if it lacks disturbing local variations.
In accordance
with the principles of the invention, variations in local tensile stiffness,
secondly, give
rise to a spatially variable strain in the paper that will reduce the tensile
strength of the
paper to below the said upper limit.
CA 02337335 2001-O1-12
WO 00/07010 PCT/SE99/01237
4
There are significant variations in local tensile stiffness measured on a
millimetre scale in commercially manufactured paper. To obtain a good
predictive result,
it is therefore very important to take account of these local variations,
whose amplitude
varies with varying conditions of production. The relation between measured
mean
values of tensile stiffness and tensile strength will vary significantly with,
for example,
beating level (beating of the pulp) if the local variations are not taken into
account, see
Fig. 2.
Results of prediction experiments that are based on data obtained by means
of the arrangement described in US 5 361 638 are shown in Fig. 3. The contents
of the
said US patent, which in this way contributes to the realisation of the
present invention in
an ingenious manner, are therefore intended to be included by reference in the
present
patent matter. By use of a prediction model according to the invention and
according to
what is evident in detail from the following equation, a very good prediction
(rz = 0.997)
is obtained. This result has been obtained under significantly variable
process conditions,
in this case varying beating levels (degree of beating).
Tensile strength = f(m, a)"~; = const.* m'~5/(1+ a /m)
It is known that structural differences in the sheet as a result of the
varying
degree of orientation (MD/CD ratio; where MD is the direction of the machinery
and
CD is the transverse direction) as well as varying tension (restraining force)
during the
drying of the paper affect the relation between measured tensile stiffness and
tensile
strength. With the aid of the arrangement described in US 5 361 638, data can
be
obtained that allow compensation for changes as a result of structural
differences.
Examples of data that can be obtained with the measurement method are local
stiffness in
different directions (MD, CD and ZD) in the paper, i.e. tensile stiffness and
bending
stiffness in MD and CD as well as compressional stiffness in the thickness
direction ZD.
The result that we have obtained when evaluating experimental data shows
that when orientation increases (increasing MD/CD ratio), the tensile strength
in MD in
relation to the tensile strength in CD will increase somewhat more than the
tensile
stiffness in MD in relation to the tensile stiffness in CD; see Fig. 4. With
knowledge of
this relationship and the actual tensile stiffness in the said directions, the
influence of a
CA 02337335 2001-O1-12
WO 00/07010 PCT/SE9910I237
varying MD/CD relationship can be compensated by the introduction of a
function g of
the stiffness in MD and CD, which gives:
Tensile strength index = f* g(stiffness-MD, stiffness-CD)
5
It is widely known that the tensile stiffness increases to a larger extent
than
the tensile strength with increased drying tension. At the same time, the
relation between
stiffnesses measured in different directions in the paper changes. Thus, with
the present
invention, even this effect can be compensated for, in this case by the
introduction of a
function h of the stiffnesses in MD, CD and ZD, i.e. in the machine,
transverse and
thickness directions. When determining the tensile strength of paper with
different levels
of local variations in stiffness, varying MD/CD ratios and different tensions
during
drying the following predictive function is utilised:
Tensile strength index = f* g *h(stiffness-MD, stiffness-CD, stiffness-ZD)
The present invention can also lead to an improved prediction of the tensile
strength of a material on the basis of indirect measurements. In these cases,
the method is
complemented by utilising a relation between data from a number of other local
measurements and the local tensile stiffness and/or the local tensile
stiffness index, which
are required to obtain a better prediction. This naturally applies if a
relation can be
confirmed at the actual process conditions. Examples of indirect measurements
that may
be used (separately or in combination) to generate data for this type of
prediction are
local basis weight, local optical density and local thickness of the paper.
Finally, it is probable that other strength properties of paper, for example
tear strength and compression strength, are similarly dependent on variations
in the local
mechanical properties, such as the local stiffness (in the plane and/or the
thickness
direction) and the local rigidity to bending, i.e. properties that are
measured with a high
spatial resolution.
