SENSING ELEMENT OF CORIOLIS FORCE GYROSCOPE

Abstract

A gyroscope includes a ring-shaped resonator mounted in a housing, and a bottom plate attached to the resonator. A plurality of openings arranged substantially circumferentially on the bottom plate, and a plurality of grooves between the openings. A plurality of piezoelectric elements are located in the grooves. The resonator is substantially cylindrical. The plurality of openings are arranged substantially symmetrically. The piezoelectric elements can be outside the resonator, or inside the resonator. A cylindrical flexible suspension connecting the bottom to the resonator to the ring shaped resonator, wherein an average radius of the cylindrical flexible suspension and the ring shaped resonator, accounting for variation thickness of wall, is the same throughout.

Description

BACKGROUND OF THE INVENTION

1. Field of the Invention

The present invention is related to gyroscopes, and more particularly, to gyroscopes having high sensitivity and high signal-to-noise ratio.

2. Background Art

Vibrational gyroscopes have many advantages over conventional gyroscopes of the spinning wheel type. Thus, a vibrational gyroscope is considerably more rugged than a conventional spinning wheel gyroscope, can be started up much more quickly, consumes much less power and has no bearings which could be susceptible to wear.

A wide variety of vibrating members have been employed in previously proposed vibrational gyroscopes, ranging in shape from a tuning fork to a pair of torsionally oscillating coaxial spoked wheels. However the present invention is particularly concerned with vibrational gyroscopes in which the vibrating member comprises a radially vibrating annular shell, such as a hemispherical bell or a cylinder for example. In such gyroscopes the axis of the annular shell (e.g., the z axis) is the sensitive axis and the shell, when vibrating, periodically distorts in an elliptical fashion with four nodes spaced regularly around the circumference and located on the X and Y axes. Any rotation about the z axis generates tangential periodic Coriolis forces which tend to shift the vibrational nodes around the circumference of the shell and thereby generate some radial vibration at the original nodal positions on the X′ and Y′ axes. The Coriolis force can be calculated based on the relationship F_c=2V×Ω, where F_c is the Coriolis force, V is the linear velocity vector of the mass elements of the resonator (shell) due to fundamental mode vibration, “x” is the vector product, and Ω is the angular velocity vector. Consequently the output of one or more transducers located at one or more of these nodal positions gives a measure of the rotation rate (relative to an inertial frame) about the Z-axis.

This highly symmetrical system has a number of important advantages over arrangements in which the vibrating member is not rotationally symmetrical about the z-axis. Thus, the component of vibration rotationally induced by the Coriolis forces is precisely similar to the driving vibration. Consequently, if the frequency of the driving vibration changes (e.g. due to temperature variations) the frequency of the rotationally induced component of vibration will change by an identical amount. Thus, if the amplitude of the driving vibration is maintained constant, the amplitude of the rotationally induced component will not vary with temperature. Also the elliptical nature of the vibrational distortion ensures that the instantaneous polar moment of inertia about the z-axis is substantially constant throughout each cycle of the vibration. Consequently, any oscillating torque about the z-axis (due to externally applied rotational vibration) will not couple with the vibration of the walls of the shell. Accordingly vibration gyroscopes incorporating an annular shell as the vibrating member offer superior immunity to temperature changes and external vibration.

However in practice, vibrational gyroscopes generally employ piezoelectric transducers both for driving and sensing the vibration of the vibrating member. In cases where a vibrating annular shell is employed, the transducers are mounted on the curved surface of the shell, generally near its rim. Since it is difficult to form a low compliance bond between two curved surfaces, the transducers must be sufficiently small to form an essentially flat interface with the curved surface of the annular shell. The output of the vibration-sensing transducers is limited by their strain capability, so that the sensitivity of the system is limited by signal-to-noise ratio. All these problems become more acute as the dimensions of the annular shell are reduced.

FIG. 1 illustrates how Coriolis forces are used in gyroscopes to measure the speed of rotation. As shown in FIG. 1, a resonator, typically in the shape of a cylinder, designated by 104 in its un-deformed state, is rotated. The vibration modes of the cylinder 104 involve “squeezing” the cylinder along with one of its two axes, thereby forming an ellipse. One of the axes, designated by X, becomes the major axis of the ellipse, and the other one, designated by Y, becomes the minor axis of the ellipse. This is the primary vibration mode of the cylindrical resonator, with the vibration mode designated by 101 in FIG. 1.

In essence, the cylindrical resonator alternates between orthogonal states, shown by 101 in FIG. 1. When the resonator rotates at an angular velocity Ω, a second vibration mode starts to appear, which is designated by 102 in FIG. 1. This is due to Coriolis force vectors 103, which result in a Coriolis force in a combined Coriolis vector 105. Therefore, the added standing vibration wave 102 is oriented at 45° relative to the primary vibration modes 101. The amplitude of the standing wave 102 is related to the angular velocity of the resonator, and is processed electronically to generate a value representative of that angular velocity. It will be appreciated from FIG. 1 that if the rotation of the resonator 104 were counterclockwise (instead of clockwise, as shown in the figure), the orientation of the resulting Coriolis force vector would be at 90° to what is indicated by 105 in FIG. 1, and would be detected accordingly.

As discussed above, conventional Coriolis force gyroscopes typically use a machined resonator cavity, or cylindrical resonator, with a number of piezoelectric elements that are attached to the body of the cylinder. Some of the piezoelectric elements are used to drive the vibration of the cylinders, and others are used to detect the standing wave due to the rotation, indicated by 102 in FIG. 1. A typical arrangement involves eight such piezoelectric elements arranged, equiangularly around the circumference of the resonator 104, such that the major and minor axes of the ellipse (X and Y in FIG. 1) have four piezoelectric elements used to generate the primary vibration mode 101, and four piezoelectric elements arranged along the axes X′ and Y′, used to detect the standing wave 102 due to the Coriolis force.

It is relatively straightforward, using current technology, to machine a very precise resonator 104, to extremely high tolerance. However, the piezoelectric elements are typically glued to the outside of the resonator. The overall structure, therefore, deviates from a perfectly symmetrical structure, since it is extremely difficult to glue the piezoelectric elements with perfect repeatability. Typical dimensions of such structures are on the order of a few millimeters to perhaps a centimeter for the smaller resonators, and larger dimensions for some of the bigger ones. The fact that the perfectly vibrating cylinder of FIG. 1 becomes an asymmetrical structure has a direct effect on the gyroscope sensitivity, and the signal-to-noise ratio, since some of the Coriolis force-driven standing wave 102 and the primary vibration mode 101 begin to overlap, rather than be at a perfect 45° angle to each other.

Accordingly, there is a need in the art for a gyroscope with high precision, high sensitivity and a high signal-to-noise ratio.

BRIEF SUMMARY OF THE INVENTION

The present invention relates to Coriolis force gyroscopes with high sensitivity that substantially obviates one or more of the disadvantages of the related art.

More particularly, in an exemplary embodiment of the present invention, a gyroscope includes a substantially cylindrical resonator mounted in a housing. A bottom plate is attached to the ring shaped resonator using a cylindrical flexible suspension, where the cylindrical flexible suspension has substantially the same radius as the ring shaped resonator but is substantially thinner than the ring shaped resonator. A plurality of openings are arranged substantially equiangularly on the bottom plate. A plurality of grooves in the bottom plate are arranged substantially equiangularly. A plurality of piezoelectric elements are arranged in the grooves.

The number of openings can be anywhere between 2 and 16, with eight openings preferred, with a corresponding number of piezoelectric elements. The piezoelectric elements and the grooves can be inside or outside the resonator.

Additional features and advantages of the invention will be set forth in the description that follows, and in part will be apparent from the description, or may be learned by practice of the invention. The advantages of the invention will be realized and attained by the structure particularly pointed out in the written description and claims hereof as well as the appended drawings.

It is to be understood that both the foregoing general description and the following detailed description are exemplary and explanatory and are intended to provide further explanation of the invention as claimed.

BRIEF DESCRIPTION OF THE DRAWINGS/FIGURES

The accompanying drawings, which are included to provide a further understanding of the invention and are incorporated in and constitute a part of this specification, illustrate embodiments of the invention and together with the description serve to explain the principles of the invention. In the drawings:

FIG. 1 illustrates how Coriolis forces are used in gyroscopes to measure the speed of rotation.

FIG. 2 illustrates a gyroscope according to one embodiment of the invention.

FIG. 3 illustrates a cross-sectional view of the resonator according to one embodiment of the invention.

FIG. 4 illustrates one embodiment of a piezoelectrode that can be used in the present invention.

FIGS. 5A-5B illustrates photographs of one embodiment of a gyroscope of the invention.

FIG. 6 illustrates an electrical schematic of how the gyroscope is controlled and angular velocity is sensed.

FIG. 7 illustrates an electrical schematic of one of the elements of the control electronics.

FIG. 8 illustrates a plan view of the sensing element of the gyroscope in another embodiment of the invention.

FIG. 9 illustrates a cross-section along the line A-A of FIG. 1.

FIG. 10 illustrates an electrical schematic for the input signals and the output signals of the resonator.

FIG. 11 illustrates maximum nodal displacement of the sensing element during its rotation for a given angular velocity Ω.

FIG. 12 illustrates an alternative embodiment of a resonator.

DETAILED DESCRIPTION OF THE INVENTION

Reference will now be made in detail to embodiments of the present invention, examples of which are illustrated in the accompanying drawings.

FIG. 2 illustrates one embodiment of the invention. As shown in FIG. 2, a resonator of a Coriolis force gyroscope, shown in a top plan view, includes a resonator body 104 and a bottom plate 206, which has been modified in a particular way. The bottom plate 206 includes a plurality of openings 210, which are preferably equi-angularly distributed around the periphery of the bottom plate 206. The bottom plate 206, therefore, in effect, has a number of “spokes” (as in wheel spokes). In between the openings 210, a number of piezoelectric elements 208 are placed on the bottom plate. 212 in FIG. 2 designates a mounting hole, which is used to secure the resonator 104.

The piezoelectric elements 208 act to both vibrate the resonator 104 in its primary mode, and to detect the secondary vibration mode of the resonator 104. It should be noted that without the openings 210, the piezoelectric elements 208 will detect mostly the vibration modes of the bottom plate 206 itself, which are generally similar to the vibration modes of a membrane, such as a surface of a drum. However, the addition of the openings 210 enables the piezoelectric elements 208 to detect the secondary vibration mode of the resonator.

Furthermore, it should be noted that the number of openings in piezoelectric elements 208 need not be eight, as shown in FIG. 2. In the degenerate case, only two openings and two piezoelectric elements 208 (positioned at 90° to the openings) can be used. Other combinations are possible, such as the use of three openings arranged at 120°, and correspondingly three piezoelectric elements 208 also located at 120°, and offset from the openings by 60°. It will be understood that the processing of the signals involved is somewhat more complex where the secondary vibration mode is not perfectly aligned with some of the piezoelectric elements 208 (as would be the case when eight openings and eight piezoelectric elements 208 are used), however, with modern computational technology, this is not a difficult computational problem to solve.

Other variations are possible, e.g., the use of 4, 5, 6 or 7 openings and corresponding piezoelectric elements 208. As yet another possibility more than 8 such openings can be used, e.g., 16. However, manufacturability is an issue, since as the number of such openings and piezoelectric elements 208 increases, the signal-to-noise ratio and sensitivity increases, but the manufacturing costs also increase as well.

The piezoelectric elements 208 can be located both inside the cylindrical resonator (in other words on the side of the bottom plate 206 that faces into the resonator 104), or on the side of the bottom plate 206 that faces outside.

Generally, it is preferred to utilize as much of the area of the bottom plate 206 as possible. In other words, whatever space is available after the openings 210 are made, should preferably be used for locating the piezoelectric elements 208. Thus, rectangular piezoelectric elements 208, such as shown in FIG. 2, do not necessarily use up all the available area. A piezoelectric element 208, such as shown in FIG. 4, takes advantage of more of the available area. Furthermore, the openings 210 need not be circular, but can have other shape, e.g., oval, etc. It should be noted that while such a shape is more efficient in terms of “real estate” utilization, it is also more difficult to manufacture. Therefore, the advantages provided by such a shape (in terms of device sensitivity, signal-to-noise ratio, dynamic range, etc.) should be balanced against manufacturability issues.

It should also be noted that the present invention is not limited to any particular method of mounting the piezoelectric elements 208 on the bottom plate 206. For example, gluing, epoxying, or any other method known in the art can be used. Furthermore, the air from the cylindrical resonator 104 can optionally be evacuated to achieve a vacuum. For relatively small resonators, on the order of the approximately a centimeter in height, this results in only a minor improvement in performance, on the order of a few percent. For larger resonators, vacuum inside the resonator 104 may be significantly advantageous.

FIG. 3 illustrates a cross-sectional side view of one exemplary embodiment of the resonator of the present invention. The cylindrical body of resonator 104, as seen in cross-section, has a diameter 2R₀, and two portions—a relatively stiff portion, designated by 303, and having a length L and a thickness H, and a relatively flexible portion, designated by 304, having a length l, and a thickness h. A fitting 321 with an opening 317 for mounting is used to mount the resonator 104 on a shaft (not shown). Note that it is important that the resonator be tightly mounted, without any “play.” The values of the parameters R₀, H, L, h, l, r and d can vary greatly, depending on the diameter of the resonator and the field of use of the gyroscope. One exemplary embodiment can have the following values: R₀=12.5 mm, H=1 mm, L=8 mm, h=0.3 mm, l=10 mm, r=4 mm, d=5.5 mm.

The gyroscope described herein works as follows. A signal generator supplies an

AC signal to opposite piezoelectric electrodes 208, which are glued on the spokes. The frequency of the supplied AC signal is close to the natural vibration frequency of the resonator 104. Due to the bending deformation of the spokes, the resonator 104 vibrates at the fundamental frequency in the 2-nd mode, oriented along the driven piezoelectric electrodes (see 101 in FIG. 1). The piezoelectric elements 208, located at 45 degrees, are therefore used to detect the signal. The signal is proportional to the angular velocity Ω, and can be demodulated, and then used to generate a signal that compensates for the displacement of the orientation of the 2× standing wave 102, as described below with reference to FIG. 6. As one option, the compensation signal can be used as the output signal that represents the angular velocity.

FIG. 5A is a photograph of a gyroscope according to one embodiment of the invention, with its cover removed. The resonator itself is covered by a lid 534 (and is therefore not visible in this figure), and is mounted on a base plate 538 which has a flange 530. Circuit boards 532 are mounted as shown and attached to a ring 536 with screws. The ring 536 is used to maintain stiffness of the overall structure. FIG. 5B shows a gyroscope with a cover 540. The dimensions of the device of FIG. 5B are about 50 mm diameter and 45-50 mm in height, although the device can easily be miniaturized further, for example, through the use of ASICs, etc. Roughly half the volume in the device of FIGS. 5A-5B is taken up by the electronics, the circuit boards 532, etc., which easily lends itself to miniaturization.

The resonator 104 can be manufactured from any number of materials, however, to ensure high stability of the measurements, it is typically manufactured from a material with low internal losses and a high Q factor. Generally, the smaller the resonator 104, the greater the error in the measurements. To reduce the error, the resonator 104 can be made out of materials with high Q factors. Also, temperature stability is also important for some applications, and various precision non-magnetic alloys with known elasticity properties can be used, or titanium alloys with damping coefficients of, e.g., δ=0.03% δ=0.022%, and a temperature coefficient of Young's modulus of e=5×10⁻⁵ l/° C.• to e=9×10⁻⁵ l/° C. Other materials can also be used, such as various alloys, fused silica, quartz, etc.

Since the thickness of the flexible suspension portion 304 is <<H, its own natural vibration frequency is shifted to lower frequencies. This is seen from the equation for the frequency of vibration of the resonator, which is given by:

$\begin{array}{cc}{\omega }_i=K\left(i\right)\frac{h^2}{R_0^2}\sqrt{\frac{E}{\left(1+v\right)\rho }}& \left(1\right)\end{array}$

where

$K\left(i\right)=\frac{i\left(i^2-1\right)}{\sqrt{\left(i^2+1\right)}}$

is the coefficient that depends on the mode of the vibration i, E is Young's modulus, v is Poisson coefficient, ρ is the density of the material of the resonator.

This means that the resonator 104 and the base on which it is mounted are widely separated in frequency space. Therefore, the flexible suspension portion 304 of the resonator 104 functions as a damper when inertial forces act on the resonator 104 (e.g., vibrational forces, shock, impacts, etc.). Furthermore, the natural frequency of the suspension is chosen such that it is significantly different from the maximum frequency of noise, which is typically around 2-3 KHz.

Reducing the thickness h of the suspension portion 304 reduces its rotational moment of inertia, which in turn reduces the demands on the precision of its manufacturing, and reduces the need for perfect symmetry of the manufactured item. This can be seen from the relationship of the moments of inertia M_K of the resonator and moment of inertia of the suspension M_S as they relate to the amplitude of the vibration of the resonator:

$\begin{array}{cc}\frac{M_S}{M_K}={\left(\frac{h}{H}\right)}^2& \left(2\right)\end{array}$

Therefore, when

$\begin{array}{cc}\frac{h}{H}\le \frac{1}{4}& \left(2\right)\end{array}$

the tolerance requirements for manufacturing of the suspension portion 304 are reduced by an order of magnitude. Only the resonator portion 303 itself needs to be precisely manufactured, not the rest of the structure, which reduces manufacturing cost substantially.

The bottom plate 206, as well as the flexible suspension portion 304, acts as elastic suspension. Since the electrodes 208 are placed on the bottom plate 206, which increase stiffness along the axes of their orientation, it is necessary to increase the stiffness of the structure between the axes X and Y to enable the resonator 104 to vibrate along the axes X′ and Y′ in FIG. 1. The eight openings therefore serve this function. The stiffness of the “spokes” (along axes X′ and Y′) is given by

$C_x=\frac{{\mathrm{Eh}}^3}{12{\left(R_0-r_0\right)}^2}$

(see FIG. 9), whereas the stiffness of the bottom plate 206 along the axes at 22.5° relative to the Y axis, is given by

$C_y=\frac{{\mathrm{Eh}}^2}{12{\left(R_0-d\right)}^2}$

where d is the diameter of the openings (for circular openings). For the resonator to vibrate along the axis X (see FIG. 8), the following condition must be satisfied:

Cx/Cy=r ₀ ² /R ₀ ²≦1   (3)

Since the electrodes 208 are placed along the axes X, and their stiffness is given by

$C_n=\frac{E_n{\mathrm{bh}}_{n3}}{12a^3}$

where b is the width of the electrodes, a is the length of the electrodes, h_n—thickness of the electrodes, E_n is Young's modulus of the electrode (e.g., piezo-ceramic Young's modulus). The spokes have the stiffness given by h_n=h if a is approximately equal to

$a\approx \frac{R_0}{2}$

and

$C_{\sum }=\frac{{\mathrm{Eh}}^3}{12R_0^2}\left(1+\frac{8E_nb}{{\mathrm{ER}}_0}\right).$

To satisfy this condition, C_Σ/C_y<1 has to hold true, or

$\begin{array}{cc}{\left(1-\frac{d}{R_0}\right)}^2\left(1+\frac{8E_nb}{{\mathrm{ER}}_0}\right)<1& \left(4\right)\end{array}$

It is clear that this condition is satisfied even when d≧R₀/2. This, in turn, demonstrates that a gyroscope with such an arrangement of electrodes will have higher sensitivity than a conventional Coriolis force gyroscope.

FIG. 6 illustrates the electronic circuit that can be used to control the gyroscope and measure the angular velocity. Piezoelectric electrodes 208A and 208E receive a driving signal in the form of a sinusoidal voltage A sin(ω_ot), where ω_o—is the frequency equal to (or close to) the second order vibrational mode frequency of the resonator 104, typically with an amplitude between 1 and 10 Volts, depending on the dynamic range of the gyroscope. A standing wave is generated, with four nodal points oriented along the piezoelectrodes 208A, 208E and 208G, 208C, and the four nodal points, located along the piezoelectrodes 208B, 208F 208H, 208D. In order to automatically maintain a stable amplitude of oscillation when the gyroscope is functioning, signals proportional to the amplitude of the oscillation are received from the piezoelectrodes 208G and 208C, are summed, and sent to the signal generator block 624, which provides positive feedback control, as well as automatic gain control (AGC). The output of the signal generator block 624 is fed to the piezoelectrodes 208A and 208E. Thus, the signal generator 624 provides for generating the vibration of the resonator 104 with autostabilization of the amplitude of the vibration using the AGC.

In the absence of rotation, when Ω=0, the signal at the nodes of the standing wave (the signal measured by the piezoelectrodes 208B, 208F and 208H, 208D) are minimal (essentially representing the drift of the zero of the gyroscope). When the resonator 104 rotates about its axis of symmetry, the piezoelectrodes 208B, 208F and 208H, 208D measure a signal, which is shifted in phase by 90 degrees relative to the driving signal A sin(ω_ot,), in other words, a cosine wave A₁ cos(ω_ot) is measured, whose amplitude A₁ is proportional to the angular velocity Ω. This signal is received from the sense piezoelectrodes 208B, 208F is summed, demodulated (see 760 in FIG. 7) using proportional and integral (PI) regulator (see 762 in FIG. 7), then is modulated (see 764 in FIG. 7) by a signal with the same frequency as the driving frequency to form a compensation signal (the signal received by the piezoelectrodes 208A and 208E). These operations are performed in block 622. The inverted signal is then supplied to the control piezoelectrodes 208H and 208D to compensate for a signal that is generated at the nodes. Thus, a negative feedback loop is implemented, which compensates for the signal at the nodes. In this case, the feedback signal from the output of the PI regulator is proportional to the angular velocity vector Ω.

To reduce the zero bias drift of the gyroscope, block 626 can be used, which provides a minimum possible signal in the nodes of the standing wave when the gyroscope is calibrated. This signal is supplied to the control piezoelectrodes 208H and 208D with an opposite phase to the signal present in those electrodes, and which is present primarily due to imperfections of the manufacturing of the resonator 104. This approach permits compensating for mass imbalances caused by differences in resonator cylinder wall thickness.

Block 620 is a programmable gain amplifier that filters the output signal, and normalizes the amplitude of the output signal of the gyroscope.

Such resonators as used in Coriolis vibrational gyroscopes have several disadvantages:

1. locating the piezoelectric element near a free edge of the resonator acts to dampen the standing wave. This, in turn, leads to a reduction in the Q-factor of the gyroscope, which in turn leads to restrictions on the scaling (of the size) of the gyroscope, and generally, on the sensitivity of the gyroscope.

2. the use in such resonators of piezoelectric elements that are glued to the surface of the resonator leads to an uneven distribution of stiffness in the resonator, when moving in a circumferential direction. This means that the access of the standing wave does not necessarily coincide with the axes of the piezoelectric elements. This in turn leads to a nonlinearity in the scaling coefficient of the gyroscope, as well as to a reduction in the sensitivity of the gyroscope. This effect is particularly visible at relatively low rates of rotation.

If the bottom of the resonator has a number of openings that are typically evenly distributed in the circumferential direction (and also symmetrical relative to the axis of the resonator), together with a number piezoelectric elements in between the openings, such a construction has the following disadvantages:

1. locating the piezoelectric elements in between the openings does not permit an exact coincidence between the axis of the standing wave and the axes of the piezoelectric elements (in other words, the coordinate system of the two effects do not coincide).

2. mechanical balancing and tuning of the resonator in this case is a fairly complex procedure, although this is generally true of many such CVGs.

A conventional CVG gyroscope is described in U.S. Pat. No. 4,644,793, which include a cylindrical cup as a resonator. The resonator is affixed to a flat flexible plate (a membrane), while the membrane is connected to a ceramic disk on which a set of electrodes is located. Deformation of the membrane, when an AC excitation signal is supplied to the resonator causes radial vibration. When rotating about their axis, the nodes of the radial vibration move on the circumference of the resonator cup due to the action of the Coriolis force. The movement of the nodes is transferred to the membrane on which sensors are located, for example, capacitive sensors, which pickup the movement of the nodes.

Such a device has a number of disadvantages:

1. the use of a membrane in a sensing element, which needs to be clamped down on its outer edge, and loaded with the cylindrical cup on its inner edge, leads to the undesirable effect when non-inertial influences are experienced, such as vibration or shock. In this case, the membrane can resonate at its own frequencies, which can be very similar, or identical, to the primary frequency of the resonator cup. This, in turn, limits the sensitivity of the gyroscope, and can also lead to significant errors in the measurement.

2. placing sensors on the membrane leads to a lack of coincidence of the axis of the standing wave would be axes of the sensors, and also leads to a reduced sensitivity of the gyroscope and a lower accuracy of measurement of the angular velocity Ω.

3. mechanical balancing and tuning of the sensing element is still fairly complex, similar to the other gyroscopes described above.

Accordingly, the resonator structure described herein is intended to improve the sensitivity and accuracy of a CVG by using a new sensing element construction, and by locating the piezoelectric elements differently, another objective is to make the job of balancing and tuning of the sensing element simpler.

One embodiment of the invention includes a cylindrical resonator element in the shape of a cylinder with a bottom, a mounting element attaches the thin walled cylinder to the base of the gyroscope, and the piezoelectric elements and the electrical pickup elements are also included in the sensing element. The cylinder has a thicker upper portion—a ring shaped resonator. The bottom of the cylinder is divided into sectors with gaps, that are generally located in a symmetrical fashion and generally oriented radially from the center to the peripheral part of the cylinder bottom. Each sector includes a groove, where several piezoelectric excitation and signal pickup elements can be located. The grooves can be on both the outer and the inner portions of the bottom of the cylinder, and furthermore, the gaps that form the sectors can be at least partly located on the body of the side portion of the cylinder.

The proposed construction of the sensing element permits a more precise coincidence of the axes of the standing wave and the piezoelectric elements, compared to conventional gyroscopes. This leads to a simplification of the correction circuit of the standing wave and to a reduction in the energy consumption, as well as to an increase in the sensitivity and accuracy of the gyroscope. Also, the mechanical balancing and tuning of the sensing element is simplified, due to a reduction in stiffness along the axes of the flexible suspension of the resonator. This is because the mass that needs to removed in order to balance the sensing element can be removed in the areas of the grooves, by changing the widths of the grooves (here, changing the stiffness of the sectors can compensate for a mass imbalance of the ring resonator).

The sensing element, illustrated in FIGS. 8 and 9, includes a thin-walled cylinder 1, that has a ring shaped resonator portion 2, formed as a cylindrical rim having a length L and a wall thickness H. A cylindrical flexible suspension 303 includes a thin portion having a length 1 and a wall thickness h. A bottom 804 is divided into sectors 805 by the gaps 806, which have a width b₂. The wall thickness h of the flexible suspension 303 is smaller than the thickness of the walls of the ring-shaped resonator 2. The ring shaped resonator 2 is attached to the suspension 303 The average radius R₀ of the cylinder 1 with variation thickness of wall is the same throughout. Inside each sector 805 there is a groove 807 having a width b₁.

Piezoelectric elements 1008 are necessary to excite the resonator and to pick up the signal from the resonator, see FIG. 10. A joint 9 is located on the bottom 804, inside the cylinder 1, and is generally oriented coaxially with the cylinder 1. Openings 10 and 11, which are coaxial with the resonator 1 and the joint 9, are used for flexible attachment of the sensing element to the base (not shown in figures).

The opening 11 in the joint 9 has a radius r, and an axially symmetrical base surface B, which is used for the manufacture of the sensing element, and which defines the direction of its axis. The joint 9 is added in order to reduce the communication between the vibrating portions of the bottom 804 and the base. This assists in the reduction of energy selection from the ring resonator 2, where the energy is transferred to the base. Connecting the sensing element to the base is done by using the tubular portion 9, using a conical coupling surface. This permits the centering of the sensing element relative to the base of the gyroscope.

The number of sectors 805, and therefore the number of gaps 807 and grooves 807, where the piezoelectric elements 1008A-1008H are located (see FIG. 10) is at least two, although more sectors and grooves 807 can be used. For example, sixteen such sectors can be used, however, the more sectors (and therefore groove 807), the more piezoelectric elements 1008 need to used, but this tends to lead to a worsening of the dynamic characteristics of the gyroscope overall. Also, with a relatively large number of such sectors and grooves, the costs of manufacturing of the resonator increases. For most practical applications, eight sectors and eight grooves is approximately the optimal configuration, where the arrangement is axially symmetrical, at 45° angles, as illustrated in FIG. 8.

The grooves 807 can be located in both sides of the bottom 804, in other words, on the outer side, as well as on the inner side. Correspondingly, the piezoelectric elements can be located either on the outer side, or on the inner side. The gaps 806, which form the sectors 805, are partially located on the body (side surface) of the cylinder 1, near the flexible suspension 303. This permits reducing the stiffness of the flexible suspension 303. In the proposed sensing element, any method of mounting the piezoelectric elements 1008 on the bottom 804 can be used, such as gluing, epoxying, or any of the other known methods.

The sensing element can be manufactured from a number of materials, however, in order to ensure high stability of the excitation, it is preferable to manufacture it from material with low internal energy losses and a high Q factor, which would provide for a high quality resonator. It is also preferable that the material of the resonator have relatively stable elastic properties in the relevant working temperature range, such as Ni-SPAN-C-alloy 902 materials, as well as other high quality non-magnetic or weakly magnetic materials. The sensing element can be manufactured from quartz, such as Suprasil, because this material has high stability elastic characterization and has high Q factor, which exceeds the Q factor of metallic materials by a factor of several X.

The gyroscope as described herein works as follows:

A generator (not shown in FIG. 10) provides a control signal to opposing piezoelectric elements (1008A and 1008E) in the form of a sinusoidal voltage A sin ω₀t, where ω₀ is the angular frequency of the signal that is equal to (or is close to) to its own frequency of fundamental form oscillation the sensing element. Due to deformation of the bottom 804, there is a bending moment, which causes elliptical deformation of the suspension 303 at the second mode of oscillation. As a result, a standing wave is generated in the sensing element, with four nodes, oriented along the piezoelectric elements 1008A and 1008E, and 1008G and 1008C, and four nodes oriented along the piezoelectric elements 1008B, 1008F, 1008H and 1008D. Signal pickup is received from piezoelectric elements 1008G, 1008C, 1008H and 1008D. The piezoelectric elements 1008G and 1008C are used to pick up the signal from the anti-nodes of the standing wave, and the piezoelectric elements 1008H and 1008D are used to pick up the signals from the nodes. Since the piezoelectric elements are arranged in a symmetrical manner, the axis of the nodes, and the axis of the anti-nodes can switch places.

When the gyroscope rotates with the oscillating sensing element about its central axis with a constant angular velocity, a Coriolis force F_c is generated, which displaces the nodes of the standing wave along the circumference of the resonator. The piezoelectric elements 1008B and 1008F located at the nodes, therefore, receive a signal which is proportional to the angular velocity Ω. The signal is then processed electronically (not shown in FIG. 10), to generate a signal for compensate for the inertial displacement of the standing wave. The compensating signal is the output from the electronic circuit, and is proportional to the angular velocity, which is the parameter that needs to be measured.

It is important to achieve a maximum possible coincidence between the axis of the standing wave and the axes of the piezoelectric elements 1008. This can be done by using finite element analysis, and should preferably take into account the geometry of the sensing element, the loading forces acting on it, and the material properties. The process is generally as follows:

each node of element has a generalized coordinate λ_i, and all the coordinates are represented as a transpose matrix T:

{λ}={λ₁, λ₂, . . . , λ_N}^T,   (5)

where N is a total number of nodal displacements.

Within each element of the finite element analysis model, for the components of displacement vectors of any point M, the approximation through nodal displacement u_i is given, which are the unknown quantities:

u _i(M)=Φ_ik(M)λ_k, i=1, 2, 3, k=1, 2, . . . , N,   (6)

Where Φ_ik(M) are the functions of the elements, which represent the coupling between the nodal displacement and the displacement vectors of the point M of the body:

{u}={Φ}{λ}  (7)

in matrix form.

The equilibrium equation is written based on the possible displacements, based on which the work performed by the internal forces is equal to the work performed by external forces due to the possible displacement:

$\begin{array}{cc}{\int }_V\sigma ·\delta \varepsilon V={\int }_V\overrightarrow{q}·\delta \overrightarrow{u}V+{\int }_S^{}\overrightarrow{p}·\delta \overrightarrow{u}S,& \left(8\right)\end{array}$

where σ is the stress tensor, δε is the deformation tensor, {right arrow over (q)} is external load distributed over the volume V of the body, δ{right arrow over (u)} is the small-scale displacement of each point of the body, permitted by the constraints, and {right arrow over (p)} is load distributed over the surface S of the body. The tensor equation for the components of the deformations through the nodal displacement, for small deformations, is given by:

$\begin{array}{cc}{\varepsilon }_{\mathrm{ij}}=\frac{1}{2}\left(\frac{\partial {\Phi }_{\mathrm{ik}}}{\partial x_j}+\frac{\partial {\Phi }_{\mathrm{jk}}}{\partial x_i}\right){\lambda }_k,& \left(9\right)\end{array}$

where i, j=1, 2, 3, x₁, x₂, x₃ are the coordinate axes, oriented along the unitary vectors {right arrow over (e)}₁, {right arrow over (e)}₂, {right arrow over (e)}₃, or, in matrix form:

{ε}={B}{λ},   (10)

where

$\left\{B\right\}=\left\{▽\overrightarrow{\Phi }\right\}=\left\{\frac{1}{2}\left(\frac{\partial {\Phi }_{\mathrm{ik}}}{\partial x_j}+\frac{\partial {\Phi }_{\mathrm{jk}}}{\partial x_i}\right)\right\}$

is the matrix that couples the deformations to the nodal displacements. The coupling between the tensor components and the deformations, for an elastic body, is given by Hookes' law:

σ_ij=D_ijklε_kl,   (11)

where D_ijkl are the elastic constants of the body, i, j, k, l=1, 2, 3, or, in matrix form:

{σ}={D}{ε},   (12)

By substituting Equation 10 into Equation 12, the dependents of the stress tensor on the nodal displacement can be calculated. Then, the equilibrium equation for an elastic body, that contains the displacement of its points is given by:

$\begin{array}{cc}\left\{\sigma \right\}=\left\{D\right\}\left\{B\right\}\left\{\lambda \right\},& \left(13\right)\\ {\int }_V^{}D▽\overrightarrow{u}·\delta \left(▽\overrightarrow{u}\right)V={\int }_V\overrightarrow{q}·\delta \overrightarrow{u}V+{\int }_S^{}\overrightarrow{p}·\delta \overrightarrow{u}S,& \left(14\right)\end{array}$

where

$▽\overrightarrow{u}=\frac{1}{2}\left(\frac{\partial u_i}{\partial x_j}+\frac{\partial u_j}{\partial x_i}\right){\overrightarrow{e}}_i{\overrightarrow{e}}_j$

is the tensor operator. Relative to the final element that has a given volume V_e, with a finite surface area S_e, the equation can be rewritten as:

$\begin{array}{cc}\delta {\lambda }_i\left\{{\int }_{V_e}^{}{\mathrm{▽\Phi }}_i·D▽{\Phi }_j·{\lambda }_jV-{\int }_{V_e}\overrightarrow{q}·{\Phi }_iV-{\int }_{S_t}^{}\overrightarrow{p}·{\Phi }_iS\right\}=0& \left(15\right)\end{array}$

where i, j=1, 2, . . . , N. Given that δλ_i are non-zero, then, to satisfy the equation, the expressions within the curly brackets must be zero. In other words, a system of linear algebraic equations can be derived, which determine the conditions of equilibrium of the finite element:

{K}{λ}={f},   (16)

where

$K_{\mathrm{ij}}={\int }_{V_e}^{}{\mathrm{▽\Phi }}_i·D▽{\Phi }_jV$

is the stiffness matrix OK of the element. Using Equations 4 and 13, this can be rewritten as {K}={B}^T{D}{B}, while

$f_i={\int }_{V_e}\overrightarrow{q}·{\Phi }_iV+{\int }_{S_e}^{}\overrightarrow{p}·{\Phi }_iS$

is the vector of the nodal forces on the element of the finite element analysis model, where i, j=1, 2, . . . , N

The set of equations in (16), for all elements of the body, and given the boundary conditions, can be represented as a set of equilibrium equations written as:

{ K}{ λ}={ f},   (17)

where { K} is a global matrix of the body's stiffness, { λ} and { f} are the vectors of the nodal displacement and the forces acting on the body, respectively.

The set of equations in (17) can be used to calculate the properties of the sensing element, and the solution defines the vector of the nodal displacement. The displacement of the points of the body can then be calculated, as well as the deformation and the tension at those nodes.

Based on d'Alembert's principle, when volumetric inertial forces

${\overrightarrow{q}}_{\mathrm{inertial}}=-\rho \frac{{\partial }^2\overrightarrow{u}}{\partial t^2}=-\rho {\overrightarrow{\Phi }}_j·{\ddot{\lambda }}_j$

are used in the integral in Equation 16, the set of equations for the displacement of the finite element can be written as:

{M}{{umlaut over (λ)}}+{K}{λ}={f},   (18)

where

$M_{\mathrm{ij}}={\int }_{V_e}^{}\overrightarrow{p}{\Phi }_i·{\overrightarrow{\Phi }}_jV$

is the matrix representing the masses of the elements, {{umlaut over (λ)}} is the second time derivative of the vector of nodal displacements, ρ is the density of the material from which the body is made. The system of equations in (18) describes the eigenmodes (or natural vibrations) of the body, in the absence of external forces, and when the nodal displacements are found in the form {λ}e^iωt, where ω and t are the frequency and vibration time of the vibrating body, respectively, the equations can be rewritten as:

[−ω²{ M}+{ K}]{ λ}=0,   (19)

given that the system of equations is equal to zero, the natural vibration frequencies of the body ω₁, ω₂ . . . etc. can be calculated. Their corresponding nodal displacement eigenvectors { λ}_i, where i=1, 2, . . . represent their own resonant frequencies, can also be found.

Based on the above approach, the natural frequencies and the deformations of the different portions of the resonator have been calculated. As shown in FIG. 11, due to the gaps 806 and grooves 807, the maximum displacement of rim of the sensing element is located in the same plane as the gaps 806 and grooves 807. This means that when the sensing element vibrates at the second mode of frequency, the standing wave will be clearly defined relative to the piezoelectric elements, and its axis will be coincident with the axes of the piezoelectric elements. The fundamental resonant frequency of the cylindrical resonator can be given as:

$\begin{array}{cc}{\omega }_i=K\left(i\right)\frac{h^2}{R_0^2}\sqrt{\frac{E}{\left(1+v\right)\rho }},& \left(20\right)\end{array}$

where

$K\left(i\right)=\frac{i\sqrt{i^2-1}}{\sqrt{i^2+1}}$

is the coefficient that depends on the mode of oscillations, E is Young's modulus of the resonator, ν is Poisson's coefficient for the material of the resonator, and ρ is the density of the material.

Analyzing Equation 20, it is clear that when h is less than H, the own frequency of the suspension 303 will shift into a lower frequency range. Thus, the sensing element and the base can be decoupled. Thus, the suspension 303 acts as a shock absorber, or a damper, when the gyroscope is subjected to non-inertial effects, such as shock, vibration, and so on. Also, the flexible suspension 303 should have its parameters selected such that its resonant frequency should not coincide with the maximum spectral component of technical noise to minimize the random component of the output.

Furthermore, reducing the thickness of the wall of the flexible suspension 303 reduces its rotational moment of inertia, which in turn leads to a generally looser requirement for its manufacturer. Also, the demands on the materials from which it is made are less stringent, as discussed earlier.

The tuning and balancing of the resonator is primarily due to geometric imperfections in the shape of the sense in elements and of the piezoelectric element 1008, during their manufacture. The sensing element can be tuned and balanced after initial manufacture by changing the dimensions of the grooves, which simplifies the procedure considerably.

FIG. 12 illustrates an alternative embodiment, where the resonator is shaped not as a cylinder but as half a hemisphere, designated by 1203 in FIG. 12. The half-hemisphere 1203 is affixed to a stem 1231. Piezoelectric elements 1210 are mounted on the resonator 1203 as shown in the figure. As various other alternatives, the resonator 1203 can be shaped as a paraboloid, an ellipsoid, a hyperboloid, or various other similar shapes. It is generally preferable to have an axially symmetric shape. As yet a further embodiment, the resonator 903 can be made in angled sections (rather than as a smooth half hemisphere).

The resonator 1203 is preferably made of metal or alloy, although other materials, such as quartz or fused silica can also be used. The piezoelectric elements 1210 can be attached to the resonator 1203 by gluing, soldering, epoxying, or similar.

Having thus described embodiments of the invention, it should be apparent to those skilled in the art that certain advantages of the described method and apparatus have been achieved. It should also be appreciated that various modifications, adaptations, and alternative embodiments thereof may be made within the scope and spirit of the present invention. The invention is further defined by the following claims.

Claims

1. A gyroscope comprising: a half-hemisphere-shaped resonator mounted in a housing;a stem attached to the resonator coaxially with its axis of symmetry for enabling the resonator to rotate about the axis;a plurality of piezoelectric elements mounted on an outside surface of the resonator;a voltage generator that drives the piezoelectric elements to generate a standing wave in the resonator; anda demodulator with a negative feedback loop to provide an output signal proportional to angular velocity of the resonator,wherein the resonator is formed of a metal or a metal alloy.

2. The gyroscope of claim 1, wherein the piezoelectric elements are mounted using epoxy.

3. The gyroscope of claim 1, wherein the half-hemisphere-shaped resonator has a smooth edge at a side opposite the stem.

4. The gyroscope of claim 1, wherein the piezoelectric elements are mounted using solder.

5. The gyroscope of claim 1, wherein the piezoelectric elements are mounted using glue.

6. A gyroscope comprising: a parabolic-shaped resonator mounted in a housing;a stem attached to the resonator coaxially with its axis of symmetry for enabling the resonator to rotate about the axis;a plurality of piezoelectric elements mounted on an outside surface of the resonator;a voltage generator that drives the piezoelectric elements to generate a standing wave in the resonator; anda demodulator with a negative feedback loop to provide an output signal proportional to angular velocity of the resonator,wherein the resonator is formed of a metal or metal alloy.