GB1601578A - Moisture profile control - Google Patents

Description

(54) MOISTURE PROFILE CONTROL (71) We, REED INTERNATIONAL LIMITED, a British Company of 82, Piccadilly, London WIA 1EJ, do hereby declare the invention, for which we pray that a patent may be granted to us, and the method by which it is to be performed, to be particularly described in and by the following statement:- This invention relates to paper making and is particularly concerned with the improvement of the moisture profile of a web of paper as it leaves a paper-making machine. "Moisture profile" refers to the variation of the moisture content of a web across the width of the web.

It is generally desirable that the moisture profile of a web of paper should be as uniform as possible so that sheets cut from the web shall have uniform qualities.

Further, if the variability of the moisture content is reduced, it is possible to increase the average moisture content with less risk of the moisture content being excessive at some point across the width. This enables the production of paper with a higher average moisture content when this is desired.

The moisture content of paper is generally cbntrolled by a series of presses and steam-heated drying cylinders. If a machine is dryer limited", i.e. the dryers operate at their maximum capacity, the machine speed can be increased if an increase in average moisture content is permissible, since the dryers are required to remove less moisture. This of course results in an increased rate of production. In a non dryer limited situation, when the machine speed is limited by other factors, reduction of moisture profile variations results in a product of improved quality with a decrease in the amount of rejected material.

The force ("end loading") applied to the ends of the press rolls and the steam pressure in the drying cylinders may in general be varied to vary the moisture content of the product. The rollers of the presses are invariably crowned (i.e. have a wider diameter at the centre than at the ends) so that generally uniform pressure can be applied across the web when the ends of the rolls are loaded to a predetermined extent.

It has been proposed to provide variable crown rolls, the profile of which can be varied. Two forms of variable crown rolls are the swimming roll (or Kiister (RTM) roll) comprising a rotating shell filled with oil in which the side loading and oil pressure can be varied and the Nipco (RTM) roll which is divided across its width into a number of cylindrical chambers, the pressures in which can be varied independently. By measuring the moisture content at various positions across the width of the paper as it leaves the dryers and controlling a variable crown roll (either manually or automatically) in dependence on the measured moisture, it is possible to effect improvements in the moisture profile. A serious disadvantage of variable crown rolls is that they are necessarily large and strong and consequently very expensive.

The great majority of paper-making machines in operation at present use in the press section fixed camber presses (i.e. presses with a fixed profile) generally comprising a steel shaft covered with granite. It is possible by altering the end loadings on a fixed camber press to alter the pressure distribution across the press.

When both end loadings are equal to a particular design value, the fixed camber press roll is resiliently compressed across its width such that the pressure exerted by the roll is uniform across its width. If both end loadings are equally reduced, the centre pressure becomes greater than that at the ends (but both are reduced) owing to the profile of the roll. If one end loading is altered more than the other, the pressure distribution becomes asymmetrical. However, it is not possible simply to vary the end loadings of a fixed camber press for controlling the moisture profile because any variation in the end loadings from the design value will vary the average line pressure and hence cause an undesirable variation in the average moisture content.

According to the invention there is provided a method of controlling the moisture profile of paper made in a paper-making machine having a fixed camber press and a dryer, comprising measuring the moisture content of the paper at a plurality of positions across its width as it leaves the dryer, automatically producing three control signals dependent on the measured moisture content values, using two of the control signals to control respective end loadings of the press whereby to reduce relative variations in the moisture profile, and using the third control signal to control a machine variable which causes uniform moisture changes across the width of the paper whereby to compensate any change in average moisture content caused by a change in said end loadings.

The machine variable may be, for example, the rate of drying of the dryer. the speed of operation of the machine, the pressure on a further press which is substantially stiffer than said fixed camber press or a combination of these variables. The dryer may comprise conventional drying cylinders in which case the rate of drying may be controlled by controlling the pressure of steam introduced into the cylinders. If the machine is dryer limited, it is appropriate to control the machine speed, but if not, the steam pressure could be varied. The average line pressure of a further stiff press could be controlled alternatively or additionally in both cases. If the machine speed is controlled it is, of course, necessary to control other related variables, such as the rate of fibre feed, in the appropriate time relationship and magnitude.

The machine may comprise more than one fixed camber press, in which case it is possible to control the end loadings on two fixed camber presses, when a total of five control signals would be required. More than two fixed camber presses could be controlled, although higher numbers bring diminishing economic returns.

Preferably the control signals are produced by linear combinations of the measured moisture values, the coefficients of the linear combinations preferably being preselected so as to minimise the standard deviation of the final moisture profile.

The invention also provides a paper-making machine comprising a fixed camber press, a dryer, means for measuring the moisture content of paper at a plurality of positions across its width as it leaves the dryer, control apparatus coupled to the measuring means and arranged to produce three control signals dependent on the measured moisture content values, means for controlling in dependence on two of the control signals the respective end loadings on said fixed camber press whereby to reduce relative variations in the moisture content of the paper across its width, and means for controlling in dependence on the third control signal a machine variable which causes uniform moisture changes across the width of the paper whereby to compensate any change in average moisture content caused by a change in said end loadings.

The control apparatus is preferably an electronic controller arranged to produce three electrical control signals in dependence on the measured moisture content values. The control signals may be supplied to indicators to present their values to an operator and the operator may then institute the necessary control variations of the end loadings and the machine variable. Preferably, however, the controller is connected to control automatically the end loadings and machine variable.

As mentioned above, the controlled machine variable may be, for example, steam pressure, machine speed, or the pressure of a relatively stiff further press.

The automatic control of the end loadings of a press may be achieved by an electropneumatic device.

Viewed from another aspect, the invention provides an electronic controller for controlling the moisture profile of paper made in a paper-making machine having a fixed camber press, a dryer and a moisture gauge arranged to measure the moisture content of paper as it leaves the dryer, the controller comprising means for producing three control signals in dependence on measured values of moisture content at a plurality of positions across the width of the paper, and being arranged such that two of the control signals, when applied to control respective end loadings of the fixed camber press, cause a reduction of the relative variations in the moisture profile, and such that the third control signal, when applied to control a machine variable which causes uniform moisture changes across the width of the paper, causes any change in average moisture content caused by a change in said end loadings to be compensated.

Certain embodiments of the invention will now be described by way of example and with reference to the accompanying drawings, in which: Figure 1 is a simplified schematic diagram of a paper-making machine according to the invention; Figure 2 is a simplified block diagram of the controller of Fig. 1; Figure 3 is a more detailed block diagram of the controller of Fig. 2; Figure 4 is an explanatory diagram of a fixed camber press; Figure 5 is a graph of the line pressure distribution for two evenly loaded presses with different pressure distribution numbers; Figure 6 is a graph of the line pressure gradient as a function of pressure distribution number; and Figure 7 shows a number of simulated graphs of the moisture profile across the paper for the presses of Fig. 5, showing the effect of the present invention.

Referring to the drawings, the paper-making machine includes a conventional source of fibre material 1 which supplies material in a continuous web along an endless wire screen to a series of presses 2 and a dryer 3. The presses 2 are of the fixed camber type and comprise a steel shaft covered in a profiled granite coating which is wider at the centre than at its ends. The dryer 3 comprises 5060 drying cylinders around which the web passes; the cylinders are heated by steam supplied via one or more steam valves 4. A moisture gauge 5 is positioned to measure the moisture content of the paper as it leaves the dryer and is operative to provide moisture content values at a plurality of positions across the width of the paper. As an example, the gauge 5 could sample the moisture content at 33 positions across the paper.

An electronic controller 6 is coupled to the moisture gauge 5 and is arranged to produce three control signals, each being derived by a linear combination of the measured moisture content values. The coefficients in the linear combinations are preselected so as to minimise the standard deviation of the moisture profile of the paper, in a manner described below. Two of the control signals are applied to respective electro-pneumatic devices at the ends of a fixed camber roll 9 to control the end loadings of the roll so as to reduce relative variations in the moisture profile of the paper. The third control signal is used to control a machine variable which causes uniform moisture changes across the width of the paper whereby to compensate any change in average moisture content caused by a change in the end loadings.

The third control signal may, for example, be supplied to the steam valve 4 so as to control the supply of steam to the drying cylinders and hence the rate of drying. If necessary such third output signal could be delayed in the controller 6 by a time corresponding to the time taken by the paper to move from the controlled press 9 to the dryer 3.

The third control signal may alternatively be applied to control the speed of the paper-making machine as indicated at 7 and this is particularly appropriate in the case of a dryer limited machine, i.e. one in which the steam valve 4 is fully open.

In controlling the machine speed it is necessary first to vary the rate of fibre feed and then correspondingly to alter the speed of subsequent parts of the machine with the appropriate time delays.

Alternatively or in addition to the above two control methods, the third control signal could be used to control the end loadings on a further press 8 which is substantially stiffer than the controlled fixed camber press 9. As will be shown below, if a press is very stiff, its line pressure distribution remains substantially uniform at its end loadings are equally varied. Thus the further press 8 could be used to produce substantially uniform moisture content changes.

The controller 6 may if desired be arranged to produce two further control signals for controlling the end loadings of another fixed camber press 10. By controlling simultaneously two presses 9 and 10 the moisture profile may be made more uniform. Further slight improvements may be made by controlling more than two fixed camber presses.

The controller 6 comprises a computing and memory unit 11 in which the linear combination of the moisture content values is carried out. The coefficients by which the moisture content values are to be multiplied may be entered via an engineer's panel 12 and stored in the unit 11 or may be preprogrammed. A signal conditioning unit 13 is connected to receive signals from the moisture gauge 5 and convert them into suitable digital form for supplying to the unit 11. The results of the linear combinations are supplied to an output unit 14 including amplifiers for producing the three control signals. The gains of the output amplifiers may be controlled from the panel 12 and an operator's panel 15 is also provided to enable the outputs to be manually overriden and to set the average moisture content target.As shown in Fig. 3, the computing unit includes an arithmetic unit 16, a fixed program unit 17, a store 18 for the coefficients, and a profile store 19 in which the moisture content values are stored. The profile store 19 also stores a complete set of true or machine direction filtered moisture profile values which is continuously updated by the arithmetic unit 16.

A further explanation of the operation of the invention and a method of calculating the coefficients used in the controller will now be given with reference to Figs. 4, 5 and 6.

Any elastic press on a paper-making machine is capable of a certain amount of selective water removal across the sheet width, in that the line pressure distribution is changed as the roll deflects under the effect of varying end loadings. This behaviour offers the opportunity to modify the moisture profile of the sheet as it passes through the press and ultimately the moisture profile at the dry end. The behaviour of a press system can be predicted from a knowledge of the roll geometry and covering materials.

Most paper machine press rolls are given a camber or crown to provide a predetermined line pressure distribution at the working load. The reason is that press rolls behave as beams which deflect under applied load. The following analysis of the press system is based upon the theory of short beams on elastic foundations. The line pressure distribution in a "sandwich" of elastic material between two elastic beams, as shown in Figure 4, can be determined in this way.

For ease of presentation it will be assumed that the bottom roll is rigid. This is substantially true in practice. The top roll is therefore assumed to be a beam supported along its entire length by an elastic medium and subjected to two vertical forces given by the two independent end loadings. Because of these forces the beam will deflect, producing continuously distributed reaction forces in the supporting medium (line pressure distribution). Where the deflection is directed downward (positive) there will be a compression.

By considering the equilibrium forces on an element of the beam coupled with the equation of a beam in bending it is possible to derive the following differential equation for the deflection curve or line pressure distribution of a beam supported on an elastic foundation.

d4y El =-ky (1) dx4 y is the deflection at point x along the beam given by the beam loadings, El is the flexural rigidity of the beam and k is the effective spring rate or resilience of the foundation.

The solution of (1) is given by y=eAX(C, cos Ax+C2 sin li)+eAx(C3 cos Ax+C4 sin Ax) (2) Here A includes the flexural rigidity of the beam as well as the elasticity of the supporting medium. A is an important parameter influencing the shape of the elastic line or line pressure distribution; for this reason A is called the characteristic of the system, and, since its dimension is length -1, the term '/A is referred to as the characteristic length. If L is the length of the roll, AL will be an absolute number (the "line pressure distribution number") which determines uniquely the line pressure distribution across the width of the press that is obtained for any given side loadings.

In order to give an idea of the significance of different pressure distribution numbers, Figures 5 and 6 have been included. The line pressure distributions for two presses with AL=1.39 and 2.45 have been calculated and are shown in Fig. 5. Here, the side loadings have been increased by equal amounts to cause a change in average line pressure of 100 pli (17.5 kN/m). It can be seen that the line pressure gradient to the centre (i.e. the difference between the local line pressures at the roll edge and roll centre) is greater for a higher value of AL. Figure 6 shows this gradient calculated for a 100 pli change in average line pressure as a function of the pressure distribution number AL. For AL greater than X (3.142) the pressure distribution curve flattens out at the roll centre.For lower values of at the distribution becomes progressively flatter. The ideal range for AL is 2.0 to 3.14 for a fixed camber press; in a calender stack it may be as high as 5.0 and for a very stiff press it may be as low as 1.0.

Where the press loading system is fitted with plain parallel bearings, as is usually the case, there is a consequent turning moment applied at the edge of the press roll face which influences the pressure distribution. The moments caused by an overhang 'an between the roll edge and the bearing centre-line have the effect of increasing the line pressure gradient obtained in the press by a factor (1+4.9 a/L).

This effect is allowed for rigorously in our mathematical treatment. In Figure 4 these moments are indicated by MAS MB.

The effectiveness of any control system depends on obtaining a value of AL which is sufficiently high to obtain a sensible shaping of the line pressure distribution by changing the side loadings.

The value of AL for a beam on an elastic foundation is determined by the effective spring rate or resilience k of the foundation and the flexural rigidity El of the beam:

k varies between different materials: typical values would be about 4000 psi (27000 kN/m2) for a normal first press (in which a hard nip is to be avoided), about 12000 psi (80000 kN/m2) for a Microrok (RTM) covered roll on a PVC covered Venta-Nip (RTM) press and about 500000 psi (3500 MN/m2) for paper in a calender nip.

The value of EI varies with the material of construction of the press and with its geometry, the most important geometrical factor being the ratio of diameter D to face length L. As a rough approximation, for the k value of 12000 psi mentioned above, AL~0.30 L/D for a granite top roll AL~0.24 LID for a steel top roll.

A survey of a number of presses showed LID ratios in the range 4:1 to 9:1.

From the analysis above it can be seen that a high ratio, e.g. in the range 6:1 to 10:1, is advantageous to profile control, otherwise the roll is too stiff to allow the desired shaping of the line pressure distribution.

It is therefore possible to select a press combination from different roll constructions and coverings to obtain the desired line pressure distribution number AL. Such a selection would of course be subject to the normal engineering limitations on roll design.

The general solution of (1) given by (2) must also satisfy the boundary conditions existing at both ends of the beam. The four constants of integration can be determined by making use of the boundary conditions in terms of side loadings and moments.

In the particular case when the two side loadings are equal and the moments are zero, the line pressure distribution is given by cos hk'. cos Ax'+cos hAx' cos ,Ix key=2 AP (3) sin hAL+sin AL where x'=L-x.

Similar analytical solutions exist for the elastic line for a beam with arbitrary forces PA and PB and moments MA and MB at both ends. The general functional for the line pressure distribution kyf(PA, Ps, MAS MB, x) (4) may be transformed into a linear matrix equation where the line pressure, at the selected values of x equally spaced across the width of the roll, is collected in a line pressure vector 1 and related to the side loadings by l=FAp (5) where F is an mx2 constant matrix dependent upon the roll characteristics and p is the 2x I forcing vector (PA, PB)T.

We have determined that in the operation of a typical press a change of 55 pli (9.8 kN/m) in average line pressure causes a change of approximately 10; in final moisture content. The above line pressure distributions can therefore be transformed into corresponding moisture changes.

If it is assumed that the changes in line pressure distribution due to changes in the external loadings also change, linearly, the moisture distribution in the sheet across the width of the machine, then the following matrix equation describes the changes in the moisture profile x due to changes in the external loadings Ap, thus x=F A Ap (6) Given (6) we can seek to effect moisture changes x by loading changes Ap.

However, the use of a single press to shape the moisture profile necessitates changing the average line pressure which in turn affects the average moisture content. This unwanted side effect must therefore be compensated in any profile control system which is based on controlled changes in press side loadings. The compensation mentioned above can be included in the model equations by augmenting matrix F by an extra centre column of elements.

The new augmented system can therefore be written x=A Au (7) where A is an mx3 constant dimensionless matrix (with all the centre column elements equal and the two other columns are the same as those of F described previously, u is a 3xl forcing vector (APA, AP, AP5)T where AP is the average hypothetical line pressure compensation which must be converted optimally and applied to one of the machine variables in the system that causes change in the average moisture such as i) steam pressure setpoint ii) machine speed iii) loading of a second rigid press ()LL1 .0).

From a control point of view we are therefore seeking to find an optimal vector uO for u which corrects the measured moisture profile according to a chosen criterion.

A suitable criterion in this case would be to minimize the square of the different (or standard deviation) between the measured profile y and the profile changes that are possible with the roll, A u", i.e.

Min S=Min 211 y-A uO 112 2 (8) subject to the following necessary conditions: 1) The sum of the control outputs is zero.

2) All the control outputs are zero if all the elements of the measurement vector are equal.

Condition (1) ensures that there is no change in average moisture content and Condition (2) ensures that a level measurément profile does not produce control outputs.

The criterion function (8) may be modified to account for various other constrains on u" as well as process dynamics, to make the solution fit the requirements on any particular plant.

The minimisation of (8) leads to a feedback control law-i.e. the controls are given as a linear combination of all the measurements across the moisture profile, thus u"=-D y (9) where D is a 3xm control matrix.

The possibilities offered by the fixed camber press control system have been calculated for a Venta-Nip (RTM) press for which At=1.39. For comparison.

simulations were also carried out assuming a hypothetical press for which AL=.45.

(It should be noted that it would not be difficult to construct such a press in practice, within the normal limitations of press roll strength).

Some of the simulated profile results are plotted in Figure 7. The uncontrolled profiles are shown as histograms; the predicted profiles resulting from control actions are shown as lines joining single points. The predictions are shown for both presses having different values of the line pressure number AL.

It can be seen that the profiles are generally bowed in shape, with dry edges. In order to counteract this shape using the press controller it is obviously necessary to reduce load on the press. The calculated load changes are shown in Table 1. The inevitable reduction in line pressure resulting from these changes will allow the sheet moisture content to rise and corrective action must be taken for this on either machine speed or steam pressure (or a second press).

According to the criterion function (8) it may be seen that the standard deviation of the profile variations is reduced to an absolute minimum for a particular roll. Other criteria can be used such as reduction of relative variations- i.e. the fluctuations between the maximum and minimum moisture values are smoothed out to some extent; these criteria are however mathematically intractable.

TABLE 1 Simulation results for single presses

Control Actions Standard Possible Front Load Back Load Average Line Dryer Limited Deviation Moisture Change Change Pressure Speed Change Profile before after Shift tO lb lb Change pli 1.22 0.64 -3977 -5968 -89 -0.8 a 1.54 0.88 1.32 -2200 -4085 -56 +0.2 0.53 0.64 -1507 +229 -11 b 0.85 0.51 0.68 -1239 +442 -7 +0.4 0.96 -------------------- 0.38 -4369 -3155 -67 -0.7 c 1.15 0.70 0.90 -2984 -1779 -90 -0.6 1.21 0.56 -6013 -4345 -93 -0.9 d 1.49 0.84 1.30 -4056 -2501 -116 -0.6 Note: The upper results refer to a press for which AL=1.39 The lower results refer to a press for which #L=2.45 Clearly, even with a comparatively stiff press system considerable improvements in the moisture profile can be made. These can be considerably increased by using a more flexible press roll design.

It may be seen that the average moisture content may be increased by as much as 10. This may be compared with the increase in production speed of about 3 Ó obtainable with the use of a considerably more expensive swimming roll.

Reference has been made above to the use of an existing press which has a pressure distribution number #L=1.39 and to a hypothetical press for which AL=2.45. Computer simulations have therefore been made for a system in which both these presses are used together. One of the difficulties in calculating the control outputs for the system is that in order to achieve the absolute minimum standard deviation of a given moisture profile it may in theory be desirable to apply a high positive load to one press and to relieve all the load on the other. This is of course impossible in practice and it is therefore necessary to impose even greater restraints upon the control calculations.

It was, however, found that improvements could be made in the predicted profile standard deviation by simulating the effect of two sequentially operated presses with average moisture content compensation. The results are compared with the single press case in Table 2. It can be seen that the addition of a second press of different line pressure distribution number from the first can indeed be beneficial, providing that practical limitations are borne in mind. The controller would of course need to be designed for 5 outputs in this case.

TABLE 2 Simulation results for one and two presses

Standard Deviation before after press after press after both Profile control #L=1.39 #L=2.45 presses a 1.54 1.22 0.88 0.84 b 0.85 0.53 0.51 0.51 c 1.15 0.96 0.70 0.67 d 1.49 1.21 0.84 0.79 In principle, control could be extended to press sections having more than two presses. However, it is unlikely that existing press sections would have the necessary range of line pressure distribution number AL and the cost of replacing more than one press roll would not be justified by the small improvement likely to be achieved.Careful selection of roll material, construction and covering will allow a value of AL to be achieved which will permit good profile control using one or two fixed camber presses.

One of the related benefits of an automatic control system is that observation of the controller outputs over a period will indicate whether the roll camber is correct for the normal operating loads. Where a press roll had too little camber, the controller would persistently reduce the press loadings in order to achieve the best line pressure distribution. Conversely, too high a camber would be indicated by high press loadings.

Similarly, the gradual wear of press rolls would be compensated for by the controller until a limiting condition was reached, at which point the necessity of regrinding the roll would be obvious.

WHAT WE CLAIM IS: 1. A method of controlling the moisture profile of paper made in a papermaking machine having a fixed camber press and a dryer, comprising measuring the moisture content of the paper at a plurality of positions across its width as it leaves the dryer, automatically producing three control signals dependent on the measured moisture content values, using two of the control signals to control respective end loadings of the press whereby to reduce relative variations in the moisture profile, and using the third control signal to control a machine variable which causes uniform moisture changes across the width of the paper whereby to compensate any change in average moisture content caused by a change in said end loadings.

2. A method as claimed in claim 1 comprising using the third control signal to control the rate of drying of the dryer.

3. A method as claimed in claim 2, comprising using the third control signal to control the pressure of steam supplied to the dryer.

4. A method as claimed in claim 1 comprising using the third control signal to control the speed of operation of the paper-making machine.

5. A method as claimed in any preceding claim comprising using the third control signal to control the pressure on a further press which is substantially stiffer than said fixed camber press.

6. A method as claimed in any preceding claim wherein the machine has two fixed camber presses, comprising automatically producing five said control signals dependent on the measured moisture content values and using four of the control signals to control the respective end loadings on the two fixed camber presses.

7. A method as claimed in any preceding claim wherein the control signals are produced by linear combinations of the measured moisture values.

8. A method as claimed in claim 7 wherein the coefficients of the linear combinations are preselected so as to minimise the standard deviation of the final moisture profile.

9. A method of controlling the moisture profile of paper made in a papermaking machine, substantially as hereinbefore described with reference to the accompanying drawings.

10. A paper-making machine comprising a fixed camber press, a dryer, means for measuring the moisture content of paper at a plurality of positions across its width as it leaves the dryer, control apparatus coupled to the measuring means and

**WARNING** end of DESC field may overlap start of CLMS **.

Claims (17)

**WARNING** start of CLMS field may overlap end of DESC **. TABLE 2 Simulation results for one and two presses Standard Deviation before after press after press after both Profile control #L=1.39 #L=2.45 presses a 1.54 1.22 0.88 0.84 b 0.85 0.53 0.51 0.51 c 1.15 0.96 0.70 0.67 d 1.49 1.21 0.84 0.79 In principle, control could be extended to press sections having more than two presses. However, it is unlikely that existing press sections would have the necessary range of line pressure distribution number AL and the cost of replacing more than one press roll would not be justified by the small improvement likely to be achieved.Careful selection of roll material, construction and covering will allow a value of AL to be achieved which will permit good profile control using one or two fixed camber presses. One of the related benefits of an automatic control system is that observation of the controller outputs over a period will indicate whether the roll camber is correct for the normal operating loads. Where a press roll had too little camber, the controller would persistently reduce the press loadings in order to achieve the best line pressure distribution. Conversely, too high a camber would be indicated by high press loadings. Similarly, the gradual wear of press rolls would be compensated for by the controller until a limiting condition was reached, at which point the necessity of regrinding the roll would be obvious. WHAT WE CLAIM IS:

1. A method of controlling the moisture profile of paper made in a papermaking machine having a fixed camber press and a dryer, comprising measuring the moisture content of the paper at a plurality of positions across its width as it leaves the dryer, automatically producing three control signals dependent on the measured moisture content values, using two of the control signals to control respective end loadings of the press whereby to reduce relative variations in the moisture profile, and using the third control signal to control a machine variable which causes uniform moisture changes across the width of the paper whereby to compensate any change in average moisture content caused by a change in said end loadings.

2. A method as claimed in claim 1 comprising using the third control signal to control the rate of drying of the dryer.

3. A method as claimed in claim 2, comprising using the third control signal to control the pressure of steam supplied to the dryer.

4. A method as claimed in claim 1 comprising using the third control signal to control the speed of operation of the paper-making machine.

5. A method as claimed in any preceding claim comprising using the third control signal to control the pressure on a further press which is substantially stiffer than said fixed camber press.

6. A method as claimed in any preceding claim wherein the machine has two fixed camber presses, comprising automatically producing five said control signals dependent on the measured moisture content values and using four of the control signals to control the respective end loadings on the two fixed camber presses.

7. A method as claimed in any preceding claim wherein the control signals are produced by linear combinations of the measured moisture values.

8. A method as claimed in claim 7 wherein the coefficients of the linear combinations are preselected so as to minimise the standard deviation of the final moisture profile.

9. A method of controlling the moisture profile of paper made in a papermaking machine, substantially as hereinbefore described with reference to the accompanying drawings.

10. A paper-making machine comprising a fixed camber press, a dryer, means for measuring the moisture content of paper at a plurality of positions across its width as it leaves the dryer, control apparatus coupled to the measuring means and

arranged to produce three control signals dependent on the measured moisture content values, means for controlling in dependence on two of the control signals the respective end loadings on said fixed camber press whereby to reduce relative variations in the moisture content of the paper across its width, and means for controlling in dependence on the third control signal a machine variable which causes uniform moisture changes across the width of the paper whereby to compensate any change in average moisture content caused by a change in said end loadings.

11. A machine as claimed in claim 10 wherein the control apparatus is an electronic controller arranged to produce three electrical control signals in dependence on the measured moisture content values.

12. A machine as claimed in claim 11 wherein the controller is connected to control automatically the end loadings and machine variable.

13. A paper-making machine substantially as hereinbefore described with reference to the accompanying drawings.

14. An electronic controller for controlling the moisture profile of paper made in a paper-making machine having a fixed camber press, a dryer and a moisture gauge arranged to measure the moisture content of paper as it leaves the dryer, the controller comprising means for producing three control signals in dependence on measured values of moisture content at a plurality of positions across the width of the paper, and being arranged such that two of the control signals, when applied to control respective end loadings of the fixed camber press, cause a reduction of the relative variations in the moisture profile, and such that the third control signal, when applied to control a machine variable which causes uniform moisture changes across the width of the paper, causes any change in average moisture content caused by a change in said end loadings to be compensated.

15. An electronic controller as claimed in claim 14 arranged to calculate said control signals as linear combinations of the measured moisture values.

16. An electronic controller as claimed in claim 15 wherein the coefficients of the linear combinations are preselected so as to minimise the standard deviation of the final moisture profile.

17. An electronic controller for controlling the moisture profile of paper made in a paper-making machine, substantially as hereinbefore described with reference to the accompanying drawings.