GB2168944A

GB2168944A - Heave compensator

Info

Publication number: GB2168944A
Application number: GB08531417A
Authority: GB
Inventors: Michel Chatard
Original assignee: IFP Energies Nouvelles IFPEN
Current assignee: IFP Energies Nouvelles IFPEN
Priority date: 1984-12-28
Filing date: 1985-12-20
Publication date: 1986-07-02
Also published as: GB8531417D0; NL192535C; FR2575452B1; NL8503578A; NO177017C; NO177017B; FR2575452A1; NO855256L; JPS61179993A; GB2168944B; NL192535B

Abstract

A method and device for withdrawing an element fastened to a mobile installation (1) while compensating for the movements of this installation, comprises at least one actuating cylinder (21, 22), at least one accumulator (26) and several pulleys (12, 13, 14, 16), arranged so that the mechanical and oloepneumatic forces are substantially equal. The pulleys are mounted on levers (19, 23, 24, 25) and a cable passes from a winch (7) over the pulleys to a fixed anchorage (8). The cable extends over a block (4) which is immobile with respect to the sea bed 5 and a movable block located between the pulleys. <IMAGE>

Description

SPECIFICATION A method and device for withdrawing an element fastened to a mobile installation from the influence of the movements of this installation BACKGROUND OF THE INVENTION 1. Field ofthe invention The present invention relates to a device for withdrawing an element fastened to a mobile installation from the influence of movements of limited amplitudes of this installation.

The device of the present invention may be used, more particularly in the marine sphere, as antipounding slide or pounding compensator. In fact, at sea, the swell causes, among other effects, the pounding or heave of floating equipment. When these latter support well drilling apparatus, it is necessary to compensate for the pounding so that the drilling tool is permanently in contact with the bottom of the hole.

For this, there exist three large families of devices, namely: -those which are placed in the drilling string, -those which are inserted between the drilling string and the lifting system of the drilling apparatus, -those which are integrated in the lifting system.

The present invention, when it is applied to the marine sphere, relates to a particular device of the third family. This device forms part of those which solve the problem by making the fixed block movable, corresponding to the "crown block". This block will be designated hereafter by the expression "first block", and the mobile block corresponding to the "travelling block" will be designated by the expression "second block".

2. Description of the prior art The systems connected directly to the lifting device, in the prior art, generally comprise at least one actuating cylinder or jack itself connected to pneumatic accumulators. These accumulators occupy a large volume, which is penalizing.

The prior art may be illustrated by the American patents US-A-3 791 628 and US-A-3 749 367, by the German patent DE-A-2 221 700 and the French patent FR-A-95 453 as tvell as by the-article entitled "Heave compensating devices" published in n" 8 of the 10 September 1973 on pages 4 and 8 of the Oil Report Review.

The volume of these accumulators is an important parameter in the determination of a lifting system.

Another important magnitude is the variation of the force to be exerted on the second block as a function of the stroke of the mobile installation with respect to a constant value which this second block should withstand.

The difference between the real force and this constant value will be termed error.

The present invention reduces the volume of the accumulators and/or reduces the error.

Thus, the present invention relates to a device for withdrawing an element fastened to a mobile installation from the influence of the movements of this installation, said device comprising a first and second blocks, this latter serving for fastening said element, said first block being connected both to the shaft of a first intermediate pulley and to the shaft of a second intermediate pulley by a first and second rod respectively, a first pulley and a second pulley fixed with respect to the mobile installation, the shaft of the first fixed pulley, respectively of the second fixed pulley, being connected by a third rod, respectively by a fourth rod, to the shaft of the first intermediate pulley, respectively of the second intermediate pulley, a first and a second retaining member, a cable connecting these two retaining members together by passing successively, starting from the second retaining member, over the first fixed pulley, the first intermediate pulley, the first block and the second block while forming at least one loop. The second intermediate pulley and the second fixed pulley, at least one jack one end of which is connected to the first block and the other is connected to the mobile installation and at least one accumulator in oleopneumatic relation with said actuating cylinder, the first and the second rods having an identical length equal to C, and the third and fourth rods also have an identical length equal to B, the semidistance separating the shaft of the first and the second fixed pulleys being equal to A and the distance separating the shaft of the first block from the plane joining the shafts of the first and second pulleys is equal to G.

SUMMARY OF THE INVENTION The device of the invention is characterized in that the magnitudes A, B, C, G and the passage of the cable are determined so that the mechanical Fm and oleopneumatic Fv forces are substantially equal over a portion at least of the stroke.

The device of the invention may comprise an auxiliary correction cylinder or jack whose force is adjustable.

The device of the invention may comprise a measuring means for measuring the force exerted by the second block and means for driving theauxiliary cylinder.

It still comes within the scope of the,present invention if the angle formed by the straight line joining the shaft of the first, respectively, the second, fixed pulley and the first, respectively the second, intermediate pulley with the straight line containing the portion of the cable joining these two pulleys together is at least equal to 30 .

This angle may be at least equal to 45 . Good results may be obtained with an angle close to 65 or more.

The first block may comprise ballasting means. In the case where the main actuating cylinder(s) or jack(s) are parallel to the stroke of the first block, the expression of the mechanical Fm and oleopneumatic Ffornes may be given respectively by the expressions:

and

Fv = P5v$1 k11 - Reduced strokeg in which Q = force coming from everything that contributes to the tension of the cables, N = number of strands of the block, ss = angle formed by the strand of the cable situated between the fixed pulley and the intermediate pulley and the straight line joining the centers of these two pulleys, U = fraction of Fm independent of the tension of the cables, (p = angle defined by the direction of the straight line joining the center of the first and the second fixed pulleys and the direction of the straight line joining the axes of the first fixed pulley and of the first intermediate pulley, y = angle formed by the cable strand joining the intermediate pulley and a pulley of the block and the straight line joining the centers of these two pulleys, 0 = angle formed by the direction of the straight line joining the center of the first and second fixed pulleys and the direction of the straight line joining the centers of the first block and of the first intermediate pulley, Pg = the preinflation pressure of the accumulators, Bv = section of the actuating cylinders, K = Va/SvCcdc with V5 = volume of the accumulators, Ccdc = total stroke of the first block, Reduced stroke = actual stroke/Ccdc = = expansion coefficient of the gases.

The magnitudes A, B, C, G and the path ofthe cable may be determined for making the linearized expressions of Fm and Fv identical.

The expression of the oleopneumatic forces may be linearized by only considering the forces given by the mathematical expression of these forces in two limit positions of the travel of the first block.

The present invention also relates to a method for determining the geometry of a device for withdrawing an element fastened to a mobile installation from the influence of the movements of this installation, this device comprising a first and a second block, this latter serving for fastening said element, said first block being connected both to the shaft of a first intermediate pulley and to the shaft of a second intermediate pulley by a first and second rod respectively, a first pulley and a second pulley fixed with respect to the mobile installation, the shaft of the first fixed pulley, respectively of the second fixed pulley, being connected by a third rod, respectively by a fourth rod, to the shaft of the first intermediate pulley respectively of the second intermediate pulley, a first and a second retaining member, a cable connecting these two retaining members together while passing, successively, from the second retaining member, over the first fixed pulley, the first intermediate pulley, the first block and the second block while forming at least one loop, the second intermediate pulley and the second fixed pulley, at least one actuating cylinder one end of which is connected to the first block and the other is connected to the mobile installation and at Jeast one accumulator in oleopneumatic relation with said actuating cylinder, the first and the second rods having an identical length equal to C, and similarly the third and fourth rods have an identical length equal to B, the semidistance separating the axis of the first and of the second fixed pulleys being equal to A and the distance separating the axis of the first block from the plane joining the axes of the first and second pulleys is equal to G.

In this method, the magnitudes A, B, C and G and the passage of the cable are determined so that the mechanical Fm and oleopneumatic Forces are substantially equal over a portion at least of the stroke.

In a variant, the magnitudes A, B, C, and G and passage of the cable may be determined so that the linearized expressions of the mechanical Fm and oleopneumatic Forces correspond to curves parallel with each other.

In another variant, the magnitudes A, B, C, G and passage of the cable may be determined so that the linearized expressions of the mechanical Fm and oleopneumatic F, forces correspond to.curves having at least one common point.

BRIEF DESCRIPTION OF THE DRAWINGS The present invention will be better understood and its advantages will be more clearly understood from the following description of a particular embodiment, illustrated by the accompanying Figures in which: Figure 'shows a simple compensator device of the prior art, the right hand and left hand parts of this Figure corresponding to different positions of a mobile installation, Figures 2 and 3 show different arrangements of pulleys for guiding the cables (it should be noted that for each of these Figures, the right hand and left hand parts represent different arrangements, in actual fact it is preferable to provide symmetrical configurations), Figures 4a to 4c show schematically different positions of the device of the invention in operation, Figure 5 is an explanatory diagram for defining certain variables which characterize the device of the invention, and Figure 6 shows the response of the mechanical system and of the oleopneumatic system.

DESCRIPTION OF THE PREFERRED EMBODIMENTS The example described below relates to a loop compensating system.

It is recalled that, when the pounding of a floating body 1 is compensated for by movement of a first block 2 with respect to the derrick 3 of a drilling apparatus, for example, it is necessary and sufficient to move the first block 2 by a distance less than the amplitude of the pounding so that the second block 4 is immobile with respect to the sea bed.

If, in fact, the distance from the sea bed 5 to the floating body 1 increases, the distance fromthe first block 2 to the floating body 1 must decrease. But, if the argument is based on a constant length of cable 6, since the first block draws closer to winch 7 and the fixed point 8, the second block 4 moves away from the first one.

The stroke for compensating the pounding is therefore less than the pounding.

However, during this movement cable 6 is wound on and off the pulleys of one and other of blocks 2 and 4 and this is unacceptable from the point of view of working of the cable. On the other hand, the tension in the different portions of the cable has remained constant, movement of the first block 2 only causing negligible variations of the angle of the dead and fast strands of the cable with the feet of the derrick.

Thus, then movement of the first block 2, i.e. the variation of the distance from the first block to winch 7, must not cause a variation in length of the path of cable 6.

It is sufficient for that to provide, for example, a cable path such that it passes through two sides of fixed length of a deformable triangle, the third side of variable length, which ensures the variation of distance from the first block to the floating body, not being travelled over by the cable (Figure 2).

In fact, the apices of this triangle correspond to the centers 9, 10 and 11 of pulleys 12, 13, 14 over which the cable passes. It can be shown that the cable length remains strictly constant if the pulleys have the same diameter.

If the intermediate pulley 13 is situated below the fixed pulley 14, the path of the cable intersects one and the other of the sides of fixed length of the deformable triangle (Figure 3, left hand part).

If it is situated above and if the triangle has only acute angles, the cable path will intersect one side (Figure 2, right hand part). But if the triangle comprises an obtuse angle, the path of the cable will remain parallel to the sides of fixed lengths of the triangle (Figure 3, right hand part) A cable path may also be formed with only two pulleys (Figure 2, left hand part) and which provides a constant length path to the extent that the distance from winch 7 and from the fixing point 8 of the neutral strand 15 to the center 17 of the intermediate pulley 16 may be considered as constant.This is achieved when, on the one hand, the mean position of the intermediate pulley 6 is situated on or in the immediate vicinity of the straight line joining the fixing point 8 of the neutral strand 15 or winch 7 to the pivoting point 18 of link 19 and of derrick 3 and when, on the other hand, the range of movement of the intermediate pulley 16 is not too great.

When the variation of length of the cable path 6 is not compensated for, all the suspended load 20 as well as the tension of the fast strand 23 and of the dead strand 15 are supported by the cylinders 21 and 22 before being supported by the feet of mast 3 (Figure 1).

When this variation is compensated for, a part of the load and of the tension of the dead 15 and fast 23 strands is transferred directly to the mast 3 through arms, or rods 24 and 25 of the device without passing through the cylinders (Figure 2).

This load fraction not passing through the cylinders 21 and 22 varies depending on the position of the first block or first pulley 12 and the law according to which this variation takes place depends on the geometric characteristics of the compensation device and of its position with respect to mast 3.

If then the variation of length is not compensated for, in order to balance a constant load, whatever the position of the first block 2, the pressure in cylinders 21,23 must be held constant. In the case of Figure 2, reference 27 designates a liquid gas separator.

If the pressure is held constant by means of gas accumulators 26, the volume thereof must be as large as possible so that the pressure variation due to the polytropic expansion of the gas is as low as possible.

If, on the other hand, the variation of length is compensated, for a constant suspended load, the apparent load on the cylinders is variable.

By dimensioning and positioning the system, it can be arranged that, for a constant suspended load, the apparent load variation is greater or lesser.

If this variation is small, we come back in practice to the preceding problem.

On the other hand, if it is large, it may still be satisfactory provided that it is substantially identical to the pressure variation caused by the polytropic expansion of the gas of the accumulators.

This tangible identity of the apparent variation of the load on the cylinders and of the pressure variation due to the polytropic expansion forms the principle ofthe proposed compensation method.

The compensation system is positioned with respect to the low position of the first block 28 once the magnitudes A and G have been chosen (Figure 4c).

A is the distance from the center of the guide pulley or fixed pulley 29 to the axis 30 of the derrick 3 and G the difference of the dimensions from the center 31 of the guide pulley 29 and the center 32 of the pulleys of the first block 28 when these latter are in the low position (case of Figure 4c).

The system is dimensioned by the lengths B and C of the articulated arms or rods 34 and 35. it is in fact possible, once these dimensions are known, to position the intermediate pulley 33 whatever the travel distance.

The device will be perfectly defined once the path of the cable has been defined, i.e. when the passage direction of the cable over the pulleys and the dimensions thereof as well as the positions of the cylinders, in this case their inclination with respect to the stroke, have been defined.

The law of variation of the apparent load on the cylinders will therefore depend, in so far as the geometry of the system is concerned and for each stroke distance fixed by the specifications, on six independent parameters.

To simplify the description, it will be considered hereafter that the tilt of the cylinders with respect to the stroke is zero.

The force Fm which the system exerts on the top 3 ofthe derrick generally designated by the term"water-table" may be considered as the result of two forces U and Q, U being defined as the fraction of Fm independent of the tension of the cables.

U will itself be considered later as the sum of two forces Uf and Uc, Uc being the fraction of U dependent on the stroke.

Uf, independent of the stroke, corresponds to the weights of the mobile elements fixed to the support carriage 47 of the first block and driven at the same time as it with a linear movement corresponding to the pounding (first block, cylinder rods, etc).

Uc corresponds to the fraction of the weight of the intermediate pulley 33 and ofthe links, arms or rods, which is transferred to the carriage of the first block. This is a function of the stroke but also depends on the positioning and dimensioning of the system.

The force Q comes from all that contributes to the tension of the cables, i.e. from the suspended load W, from the weight of the second block and from the weight of the cables themselves.

It also depends, all other things being equal, on the dimensioning and positioning of the device, on the travel, and of course on the number of strands N of the reeving.

The most general expression of Fm, as a function ofthe angles 3, y, H and cup which are themselves explained by the stroke and magnitudes which determine the positioning and dimensioning, is written:

Figure 5 defines the angles , y. 8 and sp.

In this Figure, the reference 30 designates the axis of the derrick, the reference 28 a pulley of the first block, the reference 34 an intermediate pulley and reference 31 a guide pulley orfixed pulley.

Angle is the angle formed by the cable strand between the fixed pulley 31 and the intermediate pulley 34 and the straight line 38 joining the center 39 and 40 of these two pulleys. In Figure 5, two path variants of the cable have been considered, one of these paths drawn with a thick line, is designated by the reference 37 and the other in broken lines is designated by the reference 41, the first defines the angle e and the section the angle ,8 i.

The angle y is defined by the cable strands joining the intermediate pulley 34 and the pulley of block 28 and the straight line 42 joining the centers 40 and 43 of these two pulleys.

In Figure 5, the path of the strand of cable 44 has been shown passing outside the two pulleys and defining the angle y e.

The angle cp is defined by the horizontal direction 45 and the direction of the straight line 38. Similarly the angle 0 is defined by the direction of the straight line 42 and the horizontal direction.

It is recalled that ss and "are independent of the stroke and only dependent on the dimensioning of the system whereas H and wy depend on the dimensioning and on the stroke.

If the intermediate pulley 34 is situated below the water-table 46 (Figure 3, left hand part) hand Py cannot be zero and it is the most general formula which applies.

If the intermediate pulley is situated above the plate situated at the top of the mast or derrick 3 designated by the term "water-table", but beyond the guide pulley with respect to the pulleys of the first block, the angle # is zero if the intermediate pulleys and the pulleys of the first block have the same diameter (Figure 2, right hand part and Figure 5).

The expression of Fm/O is simplified and becomes

Jf the intermediate pulley still situated above the water table is also "between" the guide pulley and the crown block pulleys, and if these latter h'dve the same diameter, then P is also zero and the expression of Lm becomes: 0 Fm = 1+ U (Figure 3, right hand part and Figure 5) It is therefore independent not only of the dimensioning and the positioning but also of the stroke.

This latter case which amounts to obtaining, over the whole stroke distance of an oleopneumatic system, the most constant force possible is well known from the prior art.

P and V designate the gas pressure and volume of the accumulators when the stroke distance of the cylinders is equivalent to x, between 0 and Ccdc. Pg and V5 designate the preinflation pressure and the volume of the accumulators, i.e. their gas pressure and volume when the cylinders have effected the whole of their stroke Ccdc, and PM the maximum pressure for the service considered, i.e. that which appears when the cylinders are at the beginning of their stroke.

The force Fv delivered by the cylinders as a function of the stroke is written:

5FVv= Pg K1 (1- Reduced Stroke "' = Pm+ Reduc%d s1tmke) -"' (4) Reduced stroke" is the non dimensioned expression of the stroke, i.e. its value divided by the total stroke Ccdc fixed by the specifications (Reduced stroke x Ccde -K is also a variable without dimension. It is equal to the volume V5 of the accumulators divided by the total stroke Ccdc and by the section Sv of the cylinders (K = Vs/Sv.Ccdc).

In the zone where the polytropic expansion of the gas will be used, K will be between Sand SO (and even generally close to 10) and Reduced stroke between 0 and 1.

In this zone (Figure 6), the representation of Fv Pg.S is practically a straight line.

This phenomena may be linearized, with an accuracy of a few thousandths, using for example the minimum quadratic deviation method and thus we may write:

Fv = Pg [Ad (K).Reduced stroke + Bd(K) 1 (S) Sv = Pn Ald(K)Reduced stroke + B'd(K) (6) K,Ad(K), Bd(K),A'd(K) and B'd(K) are values which are determined once a single one of them is determined.

In fact, for all the operating pressures P, the polytropic expansion may always be linearized and a straight line found such that:

pFV.5v = ad(K) Reduced stroke+ bd(K) (7) It has been established that the most general expression of the force Fm which the system applies to the cylinders, related to the load Q (that which contributes to the tensioning of the cables) is written:

The angles , , 0 and ç assume a well defined value, on the one hand, for each position of the first block and, on the other, for each of the values which have been fixed for the five independent geometric parameters which position and dimension the device.

It is also possible to linearize this expression as was done for the response of the cylinders and still using for example the minimum quadratic deviation method.

In fact, if it is done by supposing that U is zero, we may write:

Fm = AmReduced stroke + Bm(U=O) U (9) The most satisfactory geometries of the system are those which make the mechanical Fm and olepheumatic Fv responses the closest to each other over the whole of the stroke. The forces Fm and Fv may be expressed by the formulae (1) and (4).

Geometries of the system may be determined by making simpler linearized equations identical, for example for the oleopneumatic system:

pF15v Reduced stroke + bd (10) and forthe mechanical system:

ffm = Am.Reduced stroke + Bm U (11) Q For that we must have (1)P.Sv=Q (2) ad = Am (3) bd = Brn U (12) 0 for all the values of 0 less than the load 0MAX fixed by the specifications.

U is here assumed independent of the stroke. This is not strictly exact, but that considerably simplifies the present description and will simplify it even more in the second part without for all that introducing inaccuracies. It is in fact sufficient, with the above defined notations, to write the most general expression by replacing U by Uf + Uc. The linearization is then written:

Fm = Reduced stroke + B'm + QUf (13) where A'm, B'm and Uf are slightly different from Am, Bm and U.

Only the computing programs will work with A'm B'm, and Uf. The description will be made with Am, Bm, and U which changes nothing since the expressions of the formulae are identical.

In fact, it is possible to keep the analytic expression of the response of the oleopneumatic system rather than the linearized expression since, as we have been, the analytic expression is almost perfectly linear.

Values of the pressure at each end of the stroke will be used which follow from the formulae already given It will be assumed that the response of this system is linear between these two points.

The initial state and the final state of the transformation are linked by the relationship: PgVa y' = PM(Vs - Sv.Ccdc)' (14) which may be written:

On the other hand, the linearized response of the mechanical system is represented by: Fm = O(AmReduced stroke+ Bm) + U (16) At each end of the stroke, this expression is written: Fmm = Q(Am + Bm) + U (17) FmM = Q Bm + U (18) but Fmm = Pg.Sv and FmM = PM SV (19) thus

Moreover, the specifications always fix the extreme operating conditions, namely the maximum pressure PMAX which must not be exceeded in the circuit and the maximum load which must be able to be handled, and which determines QMAX.

Knowing that, on the one hand, the operation with maximum load results in the appearance of the highest pressure and that, on the other hand, for a given load the pressure is maximum when the stroke is zero, it is possible to write: FmMAX = QMAX.BM + U and FmMAX = Sv.PMAX (22) therefore

With the specifications fixing PMAx and allowing QMAX to be known, and with the design department determining U, the section of the cylinders is known through the preceding formula once the mechanical system is dimensioned and positioned.

It is then also possible to determine the air volume of the accumulators which results in the oleopneumatic and mechanical systems having responses represented by straight lines with the same slope for a load Q.

For that, it is necessary that:

It then remains to cause the now parallel straight lines which represent the responses of the mechanical and oleopneumatic systems to merge.

That amounts to fixing the operating pressure at a point of the stroke of abscissa Reduced stroke to a value such that at a point Fm = Fv, i.e: Q(Am.Reducsd stroke + Bm) + U = P.SV (25)

but P.S,, = PM5v 1 + Reducedstroke "'=PgS,, Y =p - E1 Reduced stroke YT (26) therefore

P,=O Am.Reduced stroke + Bm + U/Q 127) S,, + ReducEdstroke Y' and

p a Am Reduced stroke + Bm + U/Q Po = Q (28) SV 21 Reduced ytyl If the point common to the mechanical and hydraulic responses is chosen so that Reduced stroke = 0 the expression of PM is simple and is written:

If it is chosen for the value of Reduced stroke = 1, it is the expression of Pg which becomes simple and is written:

Thus for the given specifications and for a given mechanical system, it is possible to define an oleopneumatic system (cylinder section, air reserve of the accumulators and inflation pressure) whose response is identical to the linearized response of the mechanical system for any value of the useful load.

Since the normal use of the drilling apparatus involves a whole range of effective loads, it is necessary either: - to adapt the oleopneumatic system to each effective load (by varying for example the volume of the air reserve of the accumulators by ballasting), or - adapting the mechanical system to a given oleopneumatic system (by artifically varying for example the weight U).

Otherwise, an error must be admitted.

It is recalled that: - U is, if the whole of the moving pieces is considered, the sum of the weights of those which do not contribute to the tension of the cables, - Q is, conversely, the sum of the weights of those which contribute to the tension of the cables: - after varying the five independent parameters which position and dimension each mechanical device, only those are retained whose response may be considered as linear and expressed by::

F, stroke + U (31) - = Reduced stroke + Bm The use of the apparatus throughout the range of all the useful loads will cause 0 to vary from Qmin to Qmsx and the pressure in the hydraulic circuit will never have to exceed a value fixed by the specifications and which is called PMAX Let us assume that one of the mechanical devices has been selected. Am and Bm are then determined by calculation and U by the design ofthe constructor.

Let Uconat be this value.

The section of the cylinders is therefore known, since it is determined by the formula:

The volume of the air reserve of the accumulators is then determined when any load 0 has been fixed, which will be called 0exact' and for which the responses of the mechanical and oleopneumatic systems are parallel.

This arbitary value Exact of the load 0 in fact allows the air reserve V5 of the accumulators to be determined (volume occupied by the air in the oleopneumatic circuit when the Reduced stroke iS equal to 1 ) since:

Finally, the operating pressure is determined by writing, for any point of the stroke, that the mechanical and oleopneumatic responses are equal at this point. The two response curves, already parallel, then become merged.

If it is decided to determine the maximum pressure for this load, i.e. the pressure at a zero strake we have:

It has just been shown that for an arbitrary value Exact of the load 0, the intrinsic response of the compensation system, i.e. the difference between the mechanical and oleopneumatic responses, could be practically reduced to the sole error of linearization of the response of the mechanical system.

This latter, moreover generally low, may be made extremely reduced at the price of a constraint which will be explained further on.

It is therefore possible to calculate the volume of the air reserve of the accumulators by successively choosing Exact equal to Qmin then to 0MAY' In a first solution, it is then sufficient to construct the installation with an air reserve volume equal to the largest of the two volumes found then, during use, to reduce the volume of this reserve depending on the value of the load.

This adaptation of the volume of the air reserve of the accumulators to the load may be achieved either by steps by placing certain air cylinders out of service, or continuously by ballasting or even by combining these two processes.

It should be noted that the volume of the air reserve ofthe accumulators depends on the load 0 through the ratio U/Q.

Therefore, if, despite the variation of 0, the ratio U/O remains constant, the air reserve itself remains constant. This forms a second solution.

Depending on the value selected for 0exact' the volume of the air reserve will be such that:

Since the weight Uconat of the moving parts does not cdntribute to the tension of the cables, it will have to be artificially modified, for example by the action of an auxiliary correction cylinder or corrector cylinder 49 which provides the force Uv, so that we always have:: Uconat - U,, = Uconst (36) Q Qconst (36)

If the correcting cylinder is to be a single acting cylinder, it is obvious that it will be advantageous to choose QMAX forthe value of Qexect. Otherwise, it will be advantageous to fix Exact somewhere between QMAX and Qmin depending on the hydraulic considerations.

However, if it is desirable for the value Uv, already independent of the stroke, to be independent of the load as well, Qexact could be chosen infinitely great, which is perfectly possible, and we would then have U, = Uconst.

Physically, and only considering the static phenomena, that means that the mobile assembly supporting the crown block is balanced, for example, by counterweights.

Finally, the operating pressure when the stroke is zero remains determined by the formula:

It is possible to combine these two solutions. Let us suppose that the interesting solution has been chosen where Qexact = QMAX.

The correction cylinder must be able to develop the maximum force:

In fact, with a particular value of Q between QMAX and Qmin which will be called Qinter,the operating interval defined by QMAX and Qmin can be split into two complementary intervals. The first will be defined by QMAX and Qinter whereas the second will be defined by Qinter and Qmin.

- The volume of the air reserves of the accumulators is, for the first interval, such that:

And for the second such that:

The ratio U/Q is kept constant, by the action of the correcting cylinder which provides the force Uv, so that in the first interval we have: Uconst - Uconst-Uv ~ Uconat - May (41) QMAX O Winter and in the second: Uconat Uconst - U,, Uconat - MAX Qinter 0 0mien 42 The ratio of these expressions determines Qinter,foritgives:: 02inter = QMAxtQmin

We had established that, over the whole of the interval ON1AX-Omin and for this favorable case where Exact = 0MAx we had:

All other things being equal, with large sized apparatus 20% can be gained on the maximum performances of this correcting cylinder.

This is only one example, and it is an economic survey which will show whether there is an advantage in creating several operating ranges which will preferably overlap.

Up to now, operating modes only have been considered where the responses of the mechanical and hydraulic systems are parallel.

In the case where the volume of the air reserve is matched to the load handled, an adjustment parameter remained which was used for superimposing the responses which were parallel.

In that in which the volume of the air reserve is fixed once and for all, the parallel response curves have been superimposed through the action of a second system, called correction system, which must supply a force adapted to the load handled, but constant over the whole compensation stroke. It is moreover from this point of view that this mode of operation is advantageous, for this correction, independent of the stroke and so constant for a given load, may be possibly ensured by a passive system, i.e. which does not permanently require energy from the outside.

This mode of operation may also be considered as a mode of operation with error in the case where the correction system is not installed. The error is then equal to the force which was required of the correction system.

But, it is also possible in this mode of operation where the air reserve is constant, to abandon the principle of parallelism of the responses, in favor of equality thereof at any point of the compensation stroke.

If it is contemplated not to correct the value of the stroke, for which the equality of the mechanical and oleopneumatic responses is provided, it will be chosen so that the greatest difference between the responses, that is to say the error, is minimum.

If we consider the linearized responses, it is obvious that it is in mid stroke that this condition is achieved.

This is acquired when the error is, except for the sign, the same at each of the terminals and, if we consider the linearized responses, this occurs when it is in mid stroke that the equality of mechanical and oleopneumatic responses is achieved.

Claims

1. In a device for withdrawing an element fastened to a mobile installation from the influence of the movements of this installation, comprising a first and a second block, this latter serving for fastening said element, said first block being connected both to the shaft of a first intermediate pulley and to the shaft of a second intermediate pulley by a first and second rods respectively, first pulley and a second pulleyfixed with respect to the mobile installation, the shaft of the first fixed pulley, respectively of the second fixed pulley, being connected by a third rod, respectively by a fourth rod, to the shaft of the first intermediate pulley, respectively of the second intermediate pulley, a first and a second retaining member, a cable connecting these two retaining members together while passing successively, from the second retaining member, over the first fixed pulley, the first intermediate pulley, the first block and the second block while forming at least one loop, the second intermediate pulley and the second fixed pulley, at least one actuating cylinder one end of which is connected to the first block and the other is connected to the mobile installation and at least one accumulator in oleopneumatic relation with said cylinder, the first and second rods having an identical length equal to C and similarly the third and fourth rods have an identical length equal to B, the semidistance separating the axis of the first and of the second fixed pulleys being equal to A and the distance separating the axis of the first block from the plane joining the axes of the first and second pulleys is equaLto G, the magnitudes A, B, C, G and the path of the cable are determined so thatthe mechanical Fm and oleopneumatic F,, forces are substantially equal over a portion at least of the stroke.

2. The device as claimed in claim 1, comprising an auxiliary correction cylinder whose force is adjustable.

3. The device as claimed in claim 2, further comprising a measuring means for measuring the force exerted by the second block and means for driving the auxiliary cylinder.

4. The device as claimed in claim 1, wherein the angle formed by the straight line joining the axes of the first, respectively the second fixed pulley and the first, respectively the second intermediate pulley with a straight line containing a portion of the cable joining these two pulleys is at least equal to 30 .

5. The device as claimed in claim 4, wherein said angle is at least equal to 45 .

6. The device as claimed in one of claims 1 and 2, wherein the first block comprises a ballasting means.

7. The device as claimed in claim 1, wherein said cylinder is parallel to the stroke of the first block, wherein the expression of the mechanical Fm and oleopneumatic F,, forces are given respectively by the expressions:

and

F,= = PoS{1 - (1 - Reduced stmke in which Q = force coming from all that contributes to the tension of the cables, N = number of strands of the block U = fraction of Fm independent of the tension of the cables, p = angle formed by the strand of the cable between the fixed pulley and the intermediate pulley and the straight line joining the centers of these two pulleys, ç = angle defined by the direction ofthe straight line joining the center of the first and second fixed pulleys and the direction of the straight line joining the axes of the first fixed pulley and ofthe first intermediate pulley, y = angle formed by the cable strand joining the intermediate pulley and the pulley of the block and a straight line joining the centers ofthese two pulleys, e = angle formed by the direction of the straight line joining the center of the first and second fixed pulleys and the direction of,the straight line joining the centers of the first block and of the first intermediate pulley;; Pg = preinflation pressure of the accumulators, S, = the section of the actuating cylinders, K = VslS,,Ccdc with V5 = volume of the accumulators, Ccdc= total stroke of the first block, Reduced stroke = actual stroke/Ccdc y' = expansion coefficient of the gases.

8. The device as claimed in claim 7, wherein the magnitudes A, B, C, G and the path of the cable are determined so as to make linearized expressions of Fm and Fv identical.

9. The device as claimed in claim 8, wherein the expression giving the oleopneumaticforces is linearized byonly considering the oleopneumatic forces given by said expression at the two endmost points of the stroke of the first block.

10. In a method for determining the geometry of a device for withdrawing an element fastened to a mobile installation from the influence of the movements of this installation, said device comprising a first and a second block, this latter serving for fastening said element, said first block being connected both to the shaft of a first intermediate pulley and to the shaft of a second intermediate pulley by a first and a second rod respectively, a first pulley and a second pulley fixed with respect to the mobile installation, the shaft of the first fixed pulley, respectively ofthe second fixed pulley, being connected by a third rod, respectively by a fourth rod, to the shaft of the first intermediate pulley, respectively of the second intermediate pulley, a first and a second retaining member, a cable connecting these two retaining members together while passing successively, from the second retaining member, over the first fixed pulley, the first intermediate pulley, the first block and the second block while forming at least one loop, the second intermediate pulley and the second fixed pulley, at least one actuating cylinder one end of which is connected to the first block and the other is connected to the mobile installation and at least one accumulator in oleopneumatic relation with said cylinder, the first and the second rods having an identical length equal to C and similarly the third and fourth rods having an identical length equal to B, the semidistance separating the axis of the first and of the second fixed pulleys being equal to A and the distance separating the axis of the first block from the plane.

joining the axes of the first and second pulleys is equal to G, said magnitudes A, B, C, G and the path of the cable are determined so that the mechanical Fm and oleopneumatic F,, forces are substantially equal over a portion at least of the stroke.

11. The method as claimed in claim 10, wherein the magnitudes A, B, C, G and the path of the cable are determined so that linearized expressions ofthe mechanical Fm and oleopneumatic F, forces are parallel.

12. The method as claimed in claim 10, wherein the magnitudes A, B, C, G and the path of the cable are determined so that linearized expressions of the mechanical Fm and oleopneumatic F, forces have at least one common point.