DE29723201U1 - Device for controlling the movement path of the workpiece pick-up head of an orbital vibration welding system - Google Patents

Description

DESCRIPTION

The invention relates to a device for controlling the movement path of the workpiece holder head of an orbital vibration welding system according to the preamble of claim 1.

Vibration welding devices for workpieces, especially those made of plastic, use pressure and rapid movement of the workpieces to be welded against each other to melt the contact surfaces between them, in order to achieve welding without adding additional material. The frequency of this mechanical movement is in the order of a few 100 Hz. A system consisting of a movable ferromagnetic mass, which is mounted between springs, and several electromagnets to excite the movement is usually used to generate the movement. The system is usually operated close to its mechanical resonance frequency.

Up to now, vibration welding machines have been used (DE 25 39 167 C3, US 3,920,504) whose movement only takes place in one axis (linear vibration welding systems). Here, only amplitude control is necessary, since the movement path is determined by the mechanical degree of freedom and the time course of the movement by the mechanical resonance system. However, the weld seams always show a visible extension in the direction of the movement. In addition, the movement always comes to a standstill at the two reversal points, which reduces the average speed.

For this reason, vibration welding devices were developed (EP 0 504 494 A2, US 5,160,393), whose mechanical movement is, in the first approximation, a circular path (orbital vibration welding systems). With these, a lower amount of blowout can be achieved at the weld seam, as it is evenly distributed around the circumference of the contact surface. In addition, the same average speed and thus heat output can be achieved with an amplitude that is 29% lower (λ/2 times) than with linear systems. To power the electromagnets, hand-held

Conventional frequency converters are used, which also control the movement amplitude (DE 25 39 167).

Since the trajectory of the movement is essentially determined by the mechanical system, a circular path only occurs if both the holding tools and the workpieces to be welded are almost rotationally symmetrical with respect to the axis of movement. This is generally not the case. For workpieces that are not rotationally symmetrical, an elliptical trajectory is formed, which means that the advantages of orbital welding cannot be used for such workpieces, or can only be used to a limited extent. Active control of the trajectory is therefore necessary.

The invention is based on the object of specifying a device for controlling the movement path independently of the trajectory that is set by the free movement of the mechanical system. For non-rotationally symmetrical workpieces, circular trajectories as well as freely predeterminable elliptical trajectories should be possible.

To achieve the object according to the invention, a device with the features of patent claim 1 is provided.

The advantages that can be achieved with the invention are in particular that a circular movement can be achieved even with non-rotationally symmetrical workpieces without additional equipment expenditure by means of the measures according to the invention. Commercially available frequency converters can be used to control the electromagnets, the microcontroller of which is given an additional control task.

Because a resonance system - such as the mass of the workpiece holder head of an orbital welder suspended on springs - can only be made to move far away from its resonance with extremely high expenditure of energy, elliptical paths, including their two extreme forms of circular and linear movement, are suitable trajectories that can be achieved with a reasonable expenditure of energy (construction performance of the actuator). All of these trajectories can be represented by the superposition of two oppositely rotating vectors.

ren of the same frequency. If a frequency close to the resonance frequency is chosen as the frequency, the trajectories mentioned can be imprinted without over-dimensioning the performance of the frequency converter.

In addition, there is the previously unavailable option of setting a desired, parameterizable elliptical movement for special welding tasks. Since this ellipse also includes the linear special case, both orbital and linear welding tasks can be handled with just one device.

Thus, with the control device according to the invention, a linear vibration can be achieved in a freely adjustable axis in order to weld workpieces with an orbital vibration welding system that can only be welded with a linear system. This was not possible with previous devices.

The possibility of adjusting the trajectory of the movement offers an additional benefit in that it is possible to deliberately deviate from the circular path in order to produce the extrusion (essentially) in places with an elliptical trajectory that are not visible or do not cause interference in the finished workpiece.

Advantageous embodiments of the invention are characterized in the subclaims.

The invention is explained below with reference to the embodiments shown in the drawings.
Show it:

Fig. 1 a basic arrangement of the electromechanical system

of an orbital welder in the mechanical coordinate system,

Fig. 2 a basic arrangement of the electrical system of the orbital welder in the electrical coordinate system,

Fig. 3 the entire control loop with controller, actuator, controlled system

and measurement value acquisition,

Fig. 4 a controller for the trajectory control,

Fig. 5 Locus curves of the voltage space vector for different modulation degrees,

Fig. 6 Locus curves of the voltage space vector for different angle arguments of the modulation factor,

Fig. 7A loci of setpoint and actual value and,

Fig. 7B Locus curves of the resulting control deviation in stator-fixed and rotating coordinates, and

Fig. 8 Formation of the voltage space vector by adding a

Fundamental oscillation space vector and a harmonic space vector with 3-fold frequency.

The basic mechanical structure is shown in a highly simplified manner in Fig. 1. The workpiece holder head W is held in its rest position by springs F and can swing freely around this rest position at the mechanical resonance frequency. The mechanical stops which limit the amplitude of this movement in each direction are not shown here. Electromagnets L are present to stimulate the desired movement. The movement is detected by two sensors S. The mechanical X/Y coordinate system is aligned so that the upper coil L exerts a force in the direction of the positive X axis. The direction of rotation of the coils L in the mechanical coordinate system is identical to that in the electrical coordinate system (Fig. 2).

Fig. 2 shows the arrangement and connection of the coils L of the electromagnets in the electrical coordinate system. If the coils are electrically connected in a star connection, they are referred to as La, Lb and Lc and are aligned in the electrical coordinate system as shown in Fig. 2. If the coils are electrically connected in a delta connection, they are referred to as Ll, L2 and L3 and are aligned in the electrical coordinate system as shown in Fig. 2. The coordinate axes are designated α and β to distinguish them from the mechanical system.

The coils L are therefore either delta-connected or star-connected to a frequency converter 20, whose valves 21 are connected to a mains-powered

fed rectifier 22. The valves are controlled by a PWM stage 23, which supplies pulse-width modulated signals in order to change the ratio between the on and off times of the valves 21 within a certain switching period (frequency). Control with pulse width modulation is known and is therefore not explained further. Furthermore, pulse width modulation is only to be understood as an example and can be replaced by other types of control. The PWM stage 23 is controlled by a microprocessor 24, which is connected to the output of the target voltage generator 8 shown in Fig. 4. If the computing power of the microprocessor 24 is sufficient, it can also take over the calculation of the entire control shown here.

The magnitude of the force of an electromagnet on a passive ferromagnetic body (armature) depends on the magnitude of the current through the coil, whereby the force is always an attractive force that increases with a larger magnitude of the current, regardless of the sign of the current. The magnitude of a sinusoidal current oscillates at twice the frequency of the current itself. In addition, the sense of rotation of the electrical and mechanical systems is opposite due to this very magnitude function: a current space vector rotating in a mathematically positive sense (in the electrical system) causes a force vector rotating in a mathematically negative sense (in the mechanical system) at twice the frequency. This means that the frequency of the electrical system must be half as large as the mechanical resonance frequency.

The entire control loop is shown schematically in Fig. 3. The user specifies the parameters of the desired motion ellipse as setpoint values. The current coordinates of the motion vector (vector of the motion path) are calculated from these setpoint parameters in a setpoint generator 1. These are compared with the measured actual values in 3. The control deviation found in this way is passed on to a controller 16, which calculates the required voltage space vector. This setpoint voltage space vector is then passed on to the actuator power section 17, which consists of the components shown in Fig. 2.

20 to 24 and which modulates the desired output voltage to control the coils L.

The control system itself can be conceptually broken down into two blocks connected in series. The first block 18 symbolizes the electrical system with the coils L, in which the electromagnetic force vector is generated. The second block 19 symbolizes the mechanical system with its inertial mass M, the springs F, the damping by the load and the sensors S. These supply the actual value coordinates of the motion vector to the controller, so that a closed control loop is created. The position vector of the mechanical system is dependent on the position vector.

Due to the changing air gaps of the coils L there is a direct reaction of the mechanical system on the electrical system.

Fig. 4 shows a controller for trajectory control. This includes the blocks setpoint generator 1 and controller 16 shown in Fig. 3. The setpoint generator 1 calculates the instantaneous values of the setpoint motion vector in mechanical coordinates from the ellipse parameters, namely the amplitude peak value a, the ellipticity ε and the ellipse position δ. As already explained, the setpoint and actual values of the mechanical system oscillate at twice the frequency of the electrical system. In complex representation, the relationship is as follows:

1 - ε · ■ ε

?nz _so ii= a( e ^{J(Pm (pm} + - e ^J(pm ~ ^(pm ) e ^J with &phgr;_&Igr;&Tgr;1 = 2&pgr;&Ggr;_&Ggr;&eegr;&iacgr;

The value cp _m is the angular position of the target motion vector in the mechanical coordinate system when moving on a circular path, namely the product of the angular frequency ω (oscillation frequency f _m ) and the time t. The expression Acp _m determines a deviation of the angular position cp _m at the time t = 0, which is caused by the fact that the mechanical system is delayed in time compared to the electrical system. These values cp _m and Acp _m for the angular position are fed to the target value generator 1 and the transformer 4 for synchronization.

In comparison stage 3, a comparison is made between the target and actual values of the

motion vector in mechanical coordinates. The control deviation is transformed into a rotating coordinate system in a coordinate system transformer 4. In this rotating coordinate system, the deviation between two ellipses is merely a constant, on which the amplitude difference is superimposed as alternating components with double the mechanical frequency. The trajectory controller 5 connected downstream of the transformer 4 can thus correct the trajectory error without being influenced too strongly by the amplitude error. The output variable or manipulated variable of the trajectory controller 5 is a complex modulation factor m, which is used to determine the waveform of the voltage that is in the

ε is generated by the target voltage generator 8. The modulation factor m as an input variable for the target voltage generator 8 thus determines the movement path (elliptical, circular or linear) (corresponding to the value ε) and the position of the elliptical or linear oscillation in the mechanical coordinate system (corresponding to the angle δ). This will be explained using Fig. 5 and 6.

In addition, the amplitude peak value is determined from the actual value of the motion vector. For this purpose, the absolute value x ² + y ² is determined in block 12 in a known manner from the measured values x and y of the sensors S. This value is stored in block 13 and then the amplitude lsr value is formed in the square root generator 14, which is compared with the amplitude setpoint value a in a comparison stage 6. The control deviation is fed to a amplitude controller 7. This determines the setpoint amplitude a _u of the output voltage. This ensures that, regardless of the current shape of the trajectory, the peak value of the amplitude does not become greater than the setpoint value, so that the workpiece holder head W is reliably prevented from hitting the mechanical stops (not shown) even during dynamic processes of the trajectory controller 5.

The target voltage generator 8 calculates the space vector of the target voltage from the target amplitude a _u and the modulation factor m. In the simplest case, this is composed of two voltage space vectors rotating against each other. One voltage space vector rotates at the frequency f _e of the basic electrical oscillation, while the second voltage space vector rotates in the opposite direction.

rotates at three times the frequency 3 f _e of the fundamental electrical oscillation. The nominal voltage space vector is defined by the following equation:

u = a _u ( ei<P ^e + m e"J ³ 9 ^et ) with cp _e =

As Fig. 8 shows, this modulation of the voltage space vector leads to a distortion of the locus curve. In Fig. 8, the space vector GRZ rotates in a circle with the fundamental oscillation and the space vector ORZ for the harmonic oscillation rotates at three times the frequency. This results in a total space vector SRZ

δ as the sum of the fundamental oscillation space vector GRZ and the harmonic oscillation space vector and thus the locus curve OK for the voltage space vector.

Fig. 5 shows different locus curves of the voltage space vector for different values of the modulation factor m. For m = 0, this results in a circular trajectory. As the value of the modulation factor m increases, pincushion-shaped distorted locus curves result, which then lead to increasingly pronounced elliptical trajectories in the mechanical system. With a modulation factor m = 0.5, an almost linear oscillation can be generated in the mechanical system. The space vector of the target voltage can also be composed by superimposing additional harmonic space vectors. This can be used to achieve greater dynamics and a narrower force ellipse, so that even extreme control requirements can be met. Normally, however, the composition of the space vector from the fundamental oscillation and the harmonic with three times the frequency is sufficient to meet all requirements.

Fig. 6 shows the resulting locus curves of the voltage space vector for different angle arguments arg(m) of the modulation factor m for a value of the modulation factor m of 0.33. As in Fig. 5, the target amplitude a _u = 1 in Fig. 6.

Fig. 7 A shows loci in the mechanical system. For example, the mechanical system oscillates on a circular path (actual value) with the amplitude

the peak value a = 1.0 and the ellipticity ε = 0. However, the setpoint value given to the setpoint generator 1 is an elliptical trajectory with the parameters a = 0.9, ε = 0.5 and δ = 40° in the mechanical system. From this, the control deviation is determined in comparison stage 3, which is shown in Fig. 7B. The output of comparison stage 3 thus supplies the stator-fixed control deviation, i.e. the control deviation of the stationary coils L, while the coordinate system transformer 4 generates the rotating control deviation, which is fed to the trajectory controller 5.

The following explains how the deviation A(p _m of the angular position of the target motion vector is corrected. In the phase calculator 2 of Fig. 4, a phase difference Δφ is calculated from the target value and the actual value of the motion vector. This phase difference Δφ also includes a phase shift Δφ^ of the system (not shown). The phase shift A(ps is due to the fact that the mechanical system responds to the electrical system with a time delay, as already mentioned. The consequence of this is that the control deviation that the trajectory controller 5 is supposed to regulate does not become zero due to the system's own phase shift Δφ ₃ , even if the actual locus is the same as the target locus. In order to eliminate this error, a phase controller 9 is provided, by which the phase position of the target motion vector in the target value generator 1 is set so that the mean value of the phase difference &Dgr;&phgr; becomes zero. The setpoint generator 1 therefore receives the deviation &Dgr;&phgr;_{&idiagr;&eegr;} of the angular position determined in the phase controller 9 as a correction value. An operating mode switch 11 is provided to switch through the value &Dgr;&phgr;&trade;, which in this case is set to a fixed frequency f _e = f _S0 ||.

In the other position of the operating mode selector 11, however, a fixed phase position Δφ _{η1 )Sol)} for the setpoint value of the motion vector is applied to the setpoint generator 1. To form the correction value, a frequency controller 10 is provided, which changes the frequency of the voltage space vector so that the frequency-dependent phase shift of the controlled system supplements the mean value of the phase error to zero. In this case, Aq> _m is therefore fixed and the

The angular position of the target motion vector cp _m is changed and serves as a correction value which is applied to the target value generator 1.

The adjustment described can be carried out during the control process (online tune). In this way, if the optimal phase shift of the controlled system is known, the electrical power required for a desired trajectory can be minimized so that maximum efficiency is achieved.

Claims (4)

PATENT ATTORNEYS j ^. .... .. . DR.-ING. H. NEGENBANK <-*1973> * * ! I* HAUCK, GRAALFS, WEHNEk HAMBURG · MUNICH · DÜSSELDORF VECTRON Electronics GmbH
Europark Fichtenhain A6
47807 Krefeld File: M-9609 Device for controlling the movement path of the workpiece holder head an orbital vibration welding system PROTECTION CLAIMS 1. Device for controlling the movement path of the workpiece holder head of an orbital vibration welding system, in which the workpiece holder head, which is attached to springs and driven by electromagnets, is set in an oscillation plane in elliptical, orbital oscillations for welding workpieces, in particular made of plastic, whereby the electromagnets are fed by a frequency converter, characterized in that a setpoint generator (1) is provided which calculates the current setpoint movement vector according to the desired shape of the movement path curve, which is fed on the one hand to a phase calculator (2) and on the other hand to a setpoint/actual value comparator (3) which determines the control deviation of the trajectory controller by subtracting the measured movement vector from the setpoint movement vector, which is then transformed into a rotating coordinate system (4) in which the two-dimensional trajectory controller (5) operates, which is formed from the 2 · the required complex modulation factor m is calculated from the control deviation in order to regulate the control deviation to zero, that the amplitude peak value of the desired motion vector is compared with the amplitude peak value of the measured motion vector (6) and is fed to a amplitude controller (7) which determines the desired amplitude of the voltage space vector in such a way that the control deviation at the input of the amplitude controller becomes zero, that a desired voltage generator (8) calculates the current voltage space vector from the desired amplitude of the voltage space vector and the modulation factor m and feeds it to a power section (17) which sets the output voltage of a frequency converter in such a way that it corresponds to the desired voltage space vector. 2. Device according to claim 1, characterized in that a phase controller (9) adjusts the phase position of the setpoint generator (Acpm) such that the phase error (Δφ) calculated by the phase calculator (2) becomes zero on average. 3. Device according to claim 1, characterized in that a frequency frequency controller (10) adjusts the frequency of the voltage space vector (f _e ) so that the phase error (Δφ) calculated by the phase calculator (2) becomes zero on average. 4. Device according to claims 2 and 3, characterized in that with an operating mode switch (11) it is possible to switch between the two controllers according to claim and claim 3 during operation and when not in operation.