1981 Volume 2
During the early 1970’s the old experimental paper machine at STFI was left too far behind by technical developments to be of any great use for basic process research. In the mid 1970’s, therefore, the planning began of new equipment at STFI. The aim was to design the experimental system required for a basic process research programme during the period 1980 – 2000.
It has to include all process units involved in recirculating material flows. This means that stock preparation, wire section, press section and backwater system should be included, while drying can be performed off-line.
The wire section should include the two basic twin-wire forming principles as well as a Fourdrinier section. In the wire section it should further be possible to form multi-ply products, either from a multi-ply headbox or from several headboxes. The roll width in the wire section will be 800 mm, which means that headboxes up to 500 mm width can be applied. The press section should avoid open draws and include double felted as well as single-felted nips. To simulate industrial processes running according to modern environmental standards, a filter should be used to recover suspended material from the excess back-water, which could then be re-used for showers, dilution, etc.
The dynamic properties of the FEX System were calculated using computer simulations. As a result it was found that 2 tons of pulp will be needed to run the system into equilibrium. It was decided that the storage capacity in the stock preparation plant should allow the system to be run into equilibrium four times on one batch of pulp.
An experimental paper machine often has to produce more data than paper. The FEX System is therefore heavily instrumented, and the requirements of the different measurements often had a strong impact on the design of the system. A central minicomputer is to administer a number of microcomputers which will do the actual job of process control and data collection and processing, and also automatically start up and shut down the complete system.
Increased use of multi-ply forming and paper-making additives are two important development trends in the paper industry, and extensive research within these areas is planned.
The characteristics of the disturbances affecting the paper web are discussed with particular reference to grammage variation, wet press vibration, and calender barring. Some of the difficulties of investigation are pointed out. It is shown that the usual assumptions of stationarity and linearity may not hold, so that frequency domain analysis must be applied with caution.
The key to any investigation lies in the comparison of corresponding spectral components in signals from different points in the system. Two contrasting approaches are compared and discussed: that adopted in the conventional two channel analysis of random data, and that of considering the interrelationship as a vector function of time. The latter approach, embodied in the Vector Correlator, gives considerable insight into the underlying relationship between the variables. Some practical examples are given.
In this paper we present a historical review of the development of image analysis hardware and describe the types of equipment which are commercially available today. We also discuss probable future trends in the development of automatic image analysis systems. Against this background the problems specific to analysing images of fibres are described and illustrated by experimental results obtained working with phase contrast microscope images of asbestos fibres.
Much has been done in studies of the light and beta-ray images of paper for formation and Mass Density Distribution respectively. This paper presents the comparison of the images obtained by the two methods using random data analysis techniques, such as filter theory, auto-correlation functions, transfer functions and coherence. Only objective measurements will be used so that visual optical illusions will not confuse the results. Types of paper to be studied will include calendered and uncalendered papers.
Experimental and analytical techniques are discussed for the application of image analysis to the measurement of fibre length, width, coarseness, and curl. It is shown that automated and manual procedures agree well and the crossed fibre problem can be reduced to insignificance by appropriate sample preparation procedures and the use of a curl factor as a recognition function for crossings. Image analysis is used to characterise the introduction of curl in kraft pulp fibres by high consistency beating, and the removal of curl in the hot disintegration of refiner mechanical pulp.
The mechanical equilibrium of two fluids separated by a curved interface requires the existence of a hydrostatic pressure, Δp between them; this is given by the Laplace equation:
Δp = -yc
Cambridgepp 707-722The Application of Image Analysis to Evaluate Small Scale Basis Weight Variations in PaperAbstractPDF
An image analysis technique was developed by means of which the mass distribution of paper was characterised by measurements of the light transmitted through it. To establish the ability of the image analyser to assess the mass distribution accurately, comparisons were made with the established method of beta-radiography. On the basis of these comparisons it was concluded that the small scale mass distribution can be assessed by image analysis of the light transmitted through paper.
The remaining aspects of the study demonstrate applications of the image analysis technique which include: analysing the variation in the mean and variance of the optical densities of different commercial newsprint samples: investigating the relationship between the mass distribution of newsprint and machine type: and comparing this with other methods used to evaluate mass distribution and formation of paper.
Cambridgepp 685-705Coating Pore Structure Analysis by Fluid Penetration and PermeationAbstractPDF
A fluid penetration technique is described which enables measurements to be made of penetration rate into coatings, of total pore volume, and of the subsequent permeation rate of fluid through the film. From these measurements, the number of pores and their length and diameter can be calculated assuming a simple model of pore structure. The lengths of pores calculated in this way are shown to be unrealistically large and a more complex model structure is postulated involving large and small pores serially connected. In order to determine pore sizes for this model it is necessary to generate an additional equation and this is achieved by measuring air permeation through the film. Pore sizes determined in this way are comparable in magnitude with those measured by other techniques.
Evidence is offered which suggests that air flow through coatings is molecular rather than viscous and that fluid pressures within the coating pore structure can be negative.
Cambridgepp 655-683Multidimensional Analysis of Paper-Related Factors in the Subjective Evaluation of Print QualityAbstractPDF
Multidimensional scaling, a statistical technique that permits separation and identification of the principal factors used by people when judging differences and preferences between pairs of test stimuli, has been adapted to the subjective evaluation of print quality. Numerical values of the factors involved in subjective print quality evaluation are used to establish the relationship with corresponding physical print qualities and related paper properties. Information is also generated concerning the preferences and reliability of each judge and the degree to which each judge agrees with other judges in a professional group.
Multidimensional analysis of the subjective evaluation of wire-mark in solid letterpress prints indicates that the degree to which lines appear in the wire-mark pattern is as disturbing as the overall wiremark intensity.
Mottle, show-through, contrast, and paper colour are found to be of importance in the judgement of stereo letterpress print quality; while mottle, liming, show through, and set-off are found to be significant in polymer plate letterpress printing trials. Physical tests are compared for their ability to predict these subjective print quality factors.
The factors which are responsible for the surface reflection of coated paper at specular angles have been explored by means of models of specular reflection and of coating roughness. The model of specular reflection used is based upon the addition of the roughnesses which arise from independent sources, i.e. the roughness caused by the base paper on the one hand, and by the coating pigments on the other. Coating surface roughness is approached theoretically by the incorporation in a model of the influences of particle shape and size, size distribution and the hypothetical statistics of particle position at the surface.
The theoretical predictions have been tested with the aid of experimental data, on specular reflection, Hunter gloss, profilometric roughness, and particle size. The samples studied were different types of coated paper, blade-coated on a pilot scale.