CN1028077C - Resistance welding method and device for realizing the same - Google Patents

Abstract

In the resistance welding method of the present invention, the primary AC voltage in the pulse width modulation process is chopped with a chopping frequency in each half-wave, and the chopping frequency is a multiple of the welding current frequency to generate a specific welding current waveform. The device according to the invention has a static frequency converter with a DC intermediate circuit and a chopper as output stage which generates a primary AC voltage and sends it to a secondary circuit welding transformer connected to the welding electrodes of a resistance welding machine. The regulator of the controller is connected to the chopper to control the welding current through pulse width modulation of the primary AC voltage. The device can be used for joint roll seam welding, spot welding or resistance welding machine for metal slab welding.

Description

The invention relates to a method for resistance welding with a welding current which is generated by an alternating primary voltage and controlled by means of pulse width modulation in a periodic half-wave pulse, in particular alternating.

The invention also relates to a device for carrying out the method, which has a static frequency converter with a DC intermediate stage and a chopper as output stage, which generates the primary AC voltage and This is sent to the welding transformer whose secondary circuit is connected to the welding electrodes of the resistance welding machine.

Such a method and device are known from EP-A2-0 260 963 and will be discussed in more detail below.

In a known seam welding device (EP-A1-0 261 328) for longitudinal resistance seam welding of overlapping edges of the bodies of storage tanks or the like, the conversion of a three-phase mains AC voltage into DC voltage, DC voltage Smoothed and converted to a pulsed voltage with alternating polarity. This pulse voltage is applied to the welding electrodes of the seam welding device. The frequency of the pulse voltage is selected in such a way that the resulting welding current is continuous, for which reason the individual spot weld nuggets or weld spots produced in each case by one of the rectangular half-waves of the pulse voltage overlap each other. Since the welding current half-waves are each generated by a half-wave of the pulse voltage, the waveform of the welding current depends on the duration of the pulse voltage half-wave. When adjusting the welding current, if the pulse duration changes during the half-wave period, which leads to considerable changes in the welding current waveform, this change must be considered a disadvantage. It would be advantageous if the welding current form was not dependent on, and therefore not determined by, machine parameters, but could be preselected for optimum welding results.

Furthermore, in known seam welding devices, the welding current regulator controlling the current for each welding spot operates in each case with the measured value of the preceding welding spot. regulator response time (also Determined by the correction device controlled by the regulator) is therefore longer (response time up to 1 ms for 500 Hz welding frequency). Therefore, the regulator cannot correct for rapid changes in welding parameters (e.g. contaminated sheet metal surfaces). In order to improve the adjustment capability of such known seam welding devices, it is therefore required to shorten the response time of the controller. To this end, the switching frequency can conceivably be increased by a certain factor. As a result, however, the frequency of the welding current also increases by this factor. As a result of the strong inductive loading of the seam welding device, the welding voltage is reduced by a factor equal to the reciprocal of the factor by which the switching frequency is increased, as the impedance grows proportionally to the frequency. To compensate for this, the voltage and power of the frequency converter of the known seam welding device and of the welding transformer must be increased by the same factor as the switching frequency. Furthermore, the requirement that the welding frequency should be kept at a certain ratio to the welding speed is no longer met. In this known seam welding device, welding parameters such as, for example, the contact resistance of the welding spot (surface quality of the welding material), material properties of the welding material, etc. Adapt the welding current waveform to different welding conditions, for example to meet the requirements of processing different materials.

The aforementioned EP-A2-0 260 963 proposes the use of high frequency current sources in order to make it possible to use smaller welding transformers. Since this poses a problem for the necessary thyristor phase control, the feedforward or forward control of the welding current is performed by using a precalculated value that has been calculated in advance based on the measured value of the previous half wave within one half wave during the phase control. The device known from this publication also does not work satisfactorily in all operating states, since the welding current is likewise switched on and off only once at the welding spot. Since the controller is also operated here in each case with the determined values of the preceding solder spot, the reaction time of the controller is longer. If the pulse width is changed during pulse width modulation, the welding current waveform is also changed, for this reason the welding current is not suitable for special materials or special working conditions.

Moreover, common to both known devices is that only the quadratic mean value of the welding current is measured as its actual value, so only this mean value of the welding current can be controlled. For this reason, a constant average value of the welding current is preset as a nominal value.

A device for steplessly controlling the amplitude of a sinusoidal alternating current is known from CH-A5-668842.

Over variable portions of each half-wave of the alternating current, a controllable circuit element can be shifted from a blocking state to a transmitting state. Undoubtedly, an electronically controllable and adjustable transformer is thus provided which is practically time-delay-free, but here too the possibility of influencing the welding current is limited to one switching process per half-wave. For this reason, faster settling times are also not possible in this case.

DE-C2-30 05 083 describes a method for the manufacture of longitudinal seam welding of circular objects, in which the duration of an almost rectangular half-wave of the welding current is adapted for transmission between welding electrode rolls in order to obtain a continuous and uninterrupted weld seam The time spent by the object, for which the energy required during the welding operation can be directly controlled by superimposing a high-frequency current component on the welding current. The adjustment possibilities by superposition of high-frequency current components are of course limited not only with respect to the adjustment range but also with regard to the adjustment time.

Finally, experts are familiar with varying the welding current intensity by means of phase-shift control, for example from the Soudronic publication "Resistance Welding" MDI 00188D, pages 9 and 10. Unfortunately, the welding current waveform also changes in each case. The same applies if the welding current remains constant under varying load conditions, since the phase shift angle has to be changed in both cases. Furthermore, the phase-shifted control of the primary AC voltage of the welding transformer produces interrupted welding current, which is likewise detrimental.

It is an object of the present invention to provide a method and a device for implementing a process of the aforementioned type, whereby the welding current waveform can easily be adapted to the needs of processing different materials. Furthermore, it is intended that the method and device are suitable for rapid adjustment of welding current until adjustment is guaranteed.

Deriving from a method of the aforementioned type, the problem is solved according to the invention, characterized in that the primary alternating voltage during pulse width modulation is chopped in each half-wave at a chopping frequency which is a multiple of the frequency of the welding current , in order to generate a specific welding current waveform.

Deriving from a device of the aforementioned type, the device according to the invention contains a static frequency converter with a DC intermediate stage circuit and a chopper as output stage, which generates the primary AC voltage Up and Transfer the primary AC voltage Up to the secondary Welding transformer for electrical connection to the welding electrodes of a resistance welding machine, characterized in that it contains a control device by means of which the multiple chopping of the half-waves of the primary alternating voltage by the chopper can be controlled, the control device having a regulating The regulator, connected to the chopper of the frequency converter, is used to control the welding current I by pulse width modulation of the primary AC voltage, and has a welding current reference element, the welding current reference element of the regulator is composed of the corresponding Welding current waveform for each chopping interval, memory for current nominal value for comparison with each current actual value determined for each chopping interval.

The use of the above-mentioned device is characterized by resistance welding machines that can be used for roll seam welding of joints, spot welding or welding of metal slabs.

Thus, while in the prior art presented above the primary AC voltage is only chopped once on each half-wave during the pulse width modulation, according to the present invention the initial AC voltage can be chopped n times, where n > 1 . If the regulation is carried out in an optimal manner, short regulation times can be obtained, since several nominal-actual value comparisons of the welding current can be carried out per half-cycle and thus the duty factor can be repeatedly influenced during the pulse width modulation. Since the chopping frequency is correspondingly a multiple of the frequency of the welding current, a fast regulation is obtained in each welding spot process. This enables the regulator to correct for rapid changes in welding parameters (e.g. from contaminated sheet metal surfaces). The form of the pulses formed by chopping the primary AC voltage is approximately rectangular on each half-wave. The duty factor, ie pulse width/pulse interval, can be varied within wide ranges. As a result, the mean value of the primary AC voltage can be directly influenced and the current waveform can be preset and thus variably formed as required, which is not possible in the prior art. Where it is influenced by the system (eg during phase shift control) or determined, as explained above. Whereas in the prior art explained above the chopping frequency is either determined (eg during phase shift control) or at most equal to the welding frequency, in the method and device according to the invention it is a multiple of the welding frequency. However, the advantage of simple selection of the desired welding current waveform also results from simple control of the chopping without adjustment. Although the adjustment is still optimal.

In an optimal arrangement, the memory contains at least n Nominal value, these n nominal values are compared with each of the n actual current values determined per half-wave during the regulation of the welding current in order to obtain a setpoint capable of influencing the duty cycle during pulse width modulation .

The method and apparatus thus have the following advantages, especially for the preferred embodiment:

The welding frequency is variable, and the chopping frequency is an optional multiple of the welding frequency;

The current waveform can be preselected, so it is variable and basically not changed by the correction of the duty factor during the welding current adjustment process;

If the preselected welding current waveform is not to be maintained during work, the current waveform is corrected accordingly by means of an adjustment program, i.e. by influence on the duty cycle;

The welding current waveform stored in the memory can be selected as required, i.e. in rectangular, sinusoidal or trapezoidal form, e.g. trapezoidal with sloped pulse tops or trapezoidal with humps or valleys (depending on the required thermal energy balance within a weld, a The better the heating and cooling phases in the solder joint can be controlled, the better the welding operation can be controlled, so that materials previously considered unweldable, such as for example chrome-plated parts, can now be welded by means of the invention);

The response time of the controller is significantly shorter than in the prior art, since welding takes place n times within one half-wave and the current is regulated in each case.

These advantages are obtained with a high chopping frequency chosen according to the invention to correspond to a multiple of the welding frequency.

In one design according to the invention, the chopping frequency is chosen to be 20 times the frequency of the welding current, for example, resulting in a welding frequency of 500 Hz and a chopping frequency of 10 KHz. This chopping frequency is fixedly preselected and then left constant. Only the duty cycle is changed during adjustment of the welding current. In the example of the above-described design according to the invention, each half-wave of the primary alternating voltage is chopped 10 times.

In a further refinement of the invention in which the welding current form is preselected by means of a nominal value table, the individual welding electrical waveforms are entered in a selectable nominal value table. So, for example, for a sinusoidal current wave One nominal value table is stored for the rectangular current waveform and one nominal value table for the trapezoidal current waveform, etc.

In yet another design of the invention in which the chopper of the frequency converter includes a bridge circuit using transistors as circuit elements and a freewheeling diode connected in parallel to the transistors, the adjustment method of the invention can be most easily realized because the circuit elements used Transistors have particularly short switching times.

In some other design schemes of the present invention, the nominal value table of each required welding current waveform that can be selected through the input terminal can be obtained in the memory; Select a subtable for each welding frequency (fs); or obtain a nominal value table for each welding frequency corresponding to the required welding current waveform in the memory. A specific subtable or nominal value table applies to each welding current waveform and/or frequency. Therefore, lower welding frequencies can achieve more nominal value per half wave than higher welding frequencies.

In another design solution of the present invention, the regulator includes a PID regulation circuit, and the PID regulation circuit is equipped with a feedforward loop, which stores the change value to the subsequent current nominal value for each current nominal value in the table, and It provides the output of the PIP regulation circuit as a feed-forward value. The short reaction times of the controllers can be used to the fullest, since the current nominal value and the corresponding change between adjacent current nominal values can be easily calculated in advance and stored in a table. Preferably the first derivative of the welding current curve is stored as the variation between adjacent current nominal values. This has the advantage that regulation can occur in advance, so that overshoots during the regulation process can be substantially avoided from the start, since it is known in advance where the next nominal current value will be based on the stored variation.

In another embodiment of the invention, a multiplier is connected to the memory, and the multiplier multiplies the nominal value of the table by a selection factor that can be input through the adjustable current nominal value input terminal of the regulator; the multiplier also has Further inputs via which further factors can be entered manually or by superimposed welding machine control systems. The required amplitude of the welding current can be obtained in a simple manner, ie by multiplying the stored nominal value of the table by a corresponding factor which can be fed in as required.

The invention also relates to a method for resistance welding with a periodic half-wave pulsating current, in particular an alternating welding current. So far, this kind of welding with sinusoidal current has appeared. In tinplate welding, the welding of very thin metal sheets and/or very lightly tinned metal sheets is increasingly problematic. Especially in the welding of cans (tin-plate containers), these metal sheets can pose problems which are technically difficult to control in their production. The method also applies to the welding of untinned black steel sheets, especially coated sheet metal, and more specifically chrome-plated sheet metal. Hitherto, attempts have been made to overcome these problems with different welding current amplitudes and welding current frequencies for sinusoidal welding currents, but the results have not been satisfactory.

The problem underlying the invention is therefore to enable the welding of thin and/or lightly tinned metal sheets and other metal sheets. Within a very narrow permissible bandwidth, it is especially possible to supply energy during welding in order to avoid spattering (too high energy supply) or overlapping gaps (too low energy supply).

According to the invention, this is achieved by the method described above in which the welding current differs from sinusoidal.

Since the individual half-waves of the welding current can have any shape, the exactly necessary energy supply for an optimally welded welding point is possible. The necessary heating and cooling of the soldering point in order to create the required electrical resistance at the soldering point can be well controlled by means of the current profile, which has hitherto not been possible.

Exemplary embodiments of the present invention are described in more detail below with reference to the accompanying drawings.

1 shows a circuit diagram of a resistance seam welding machine with a device according to the invention for regulating the welding current,

Figure 2 shows a more detailed circuit diagram of the part shown above the dotted line II-II of the device of the present invention,

Figure 3 shows a detailed view of the regulator shown as a block in Figure 1,

Figure 4 shows a first example of a pulse width modulated welding transformer primary AC voltage and the resulting sinusoidal welding current,

Figure 5 shows a second example of a pulse width modulated primary AC voltage chopped in a different way than in Figure 4,

Figure 6 shows a third example of a pulse width modulated primary alternating voltage for generating a trapezoidal welding current,

Figures 7a-7c show various evolutions of a preselected trapezoidal welding current with a sloped top of the pulse,

Figures 8a-8c show various examples of a preselectable trapezoidal welding current with one or more humps at the top of the pulse,

Figures 9a-9c1 show various examples of a preselectable trapezoidal welding current with one or more humps at the top of the pulse,

Figure 10 shows an example of a delta welding current that can be preselected,

Fig. 11 shows the circuit diagram of the resistance seam welding machine that can obtain the required welding current waveform,

Figure 12 shows a more detailed illustration of the regulator shown above line II-II in Figure 11,

Figure 13 shows a detailed illustration of the regulator shown as a block in Figure 11,

14 to 40 show various optimum current waveforms.

FIG. 1 shows a simplified circuit diagram of a resistance seam welding machine for longitudinal seam welding of round can bodies, not shown, between roller welding electrodes 10 and 12 . The resistance seam welding machine has a static frequency converter 14 powered by a power supply represented by lines L1-L3, and has an input stage connected via a conventional DC intermediate stage circuit 14c to an output stage 14b designed as a chopper 14a. The output stage 14b is connected to the primary circuit of the welding current transformer 16, and delivers the primary AC voltage Up to the primary circuit. The secondary circuit of welding transformer 16 is connected to welding electrodes 10 and 12 .

According to the extended circuit diagram in FIG. 2, the input stage 14a of the static frequency converter 14 has a three-phase rectifier, which at the same time forms the input of a DC intermediate stage circuit 14c, which is generally known because it is essential for the various aspects of the present invention. The situation is not important, so a detailed description is not required. As shown in Figure 2, in the output stage 14b of the frequency converter 14 (Figure 1) the The chopper comprises a bridge circuit with transistors T1-T4 as switching elements and freewheeling diodes F1-F4 connected in parallel with these transistors. The four gate drivers are connected to the transistors and freewheeling diodes in the manner shown in FIG. 2 and are controlled by regulator 18 (FIG. 1) via line 15. Arranged in the primary circuit of the welding transformer 16 is a current transformer 20 which detects the actual value of the current flowing in the primary circuit of the welding transformer 16 .

As already indicated, the important advantages of the present invention can be obtained even with a simple control system producing the desired duty cycle of the desired current waveform. But since the invention actually makes quick adjustment possible for the first time, the following description will be made with reference to adjustment.

According to the circuit diagram in FIG. 1 , the current actual value from the current transformer 20 is transmitted via an A/D converter 22 to the input of a controller 18 designed as a process computer. On the regulator 18, the nominal value I _soll of the welding current or the welding frequency fs can be set through the potentiometers 24 and 26 . The analog voltages developed on potentiometers 24 and 26 are applied to the process computer via A/D converter 25 or 27 . Alternatively, the welding current source value I _F can be fed into the regulator 18 via an input labeled "manual" or via the welding machine control system 19 . Said value is combined with the nominal welding current _Isoll in order to allow, for example, the current on the tank to be non-constant. In this way, the welding machine control system 19, which accurately knows the position of the can body to be welded at each moment, can therefore change the set nominal value I _soll , so that welding at each point of the can body has an appropriate welding current amplitude. The regulator 18 determines the setpoint by comparing the nominal-actual value of the welding current, which is delivered to the output stage 14b of the frequency converter 14 (FIG. 1) via the A/D converter 28 and the line 15. strobe driver (Figure 2). This setting affects the duty cycle of the rectangular pulse formed by the chopper in the output stage 14b chopping the smoothed DC voltage in each half-wave from the DC intermediate stage circuit 14c, so as to thereby pass the primary AC voltage to Pulse width modulation of the affected duty cycle regulates the welding current, as described in more detail below with reference to FIG. 3 .

Various ways of generating the primary AC voltage by chopping the smoothed DC voltage into rectangular pulses are shown in Figures 4-6. In the Figure 4 example, the smoothed DC voltage is chopped into rectangular pulses that change polarity from half-wave to half-wave, producing in the mean is the sinusoidal primary AC voltage U and the welding current I is substantially sinusoidal.

The same method is applied to the example of Fig. 5, chopping the smoothed DC voltage into rectangular pulses of equal height equal in each case to twice the peak value of the primary AC voltage which is on average sinusoidal.

In the example according to FIG. 6 , the chopping of the smooth DC voltage is carried out according to the same principle as in FIG. 4 but in such a way that the welding current I is trapezoidal.

Figure 3 shows the regulator 18 in more detail. As already mentioned above, the controller 18 is designed as a process computer, only the parts necessary for the invention are shown in FIG. 3 and described below. It comprises a PID regulation circuit 50 and a welding current reference element 52 in the form of a memory containing a current scale for each chopping interval corresponding to the welding current waveform for comparison with the respective current actual value determined in each chopping interval. value. For each welding current waveform (sine wave, triangular wave, trapezoidal wave, etc.), the memory 52 contains a table of nominal values selectable via the input W _Tab . An output of the memory 52 is connected to an input of a multiplier 54 . The output of multiplier 54 is connected to summing junction 56 . A summing junction 56 combines the input signal received from the multiplier 54 with the current actual value. The output signal of the summing junction 56 formed by the nominal-actual value comparison is applied to the input of the PID control circuit 50 .

The PID control circuit 50 transmits the setpoint signal at its output to an input of a summing junction 58 . The other output of memory 52 is connected to the other input of summing junction 58 through feedforward or forward drive loop 60 . Through the feed-forward loop, the memory transfers the change value from the actual current nominal value to the next nominal value transmitted to the multiplier 54, that is, a change in the welding current curve at the actual current nominal value in the direction of the next current nominal value. The order derivative dI/dt or delta is passed to summing junction 58 . The direction data is combined with the output signal of the PID adjustment circuit 50, so the output signal of the summing point 58 constitutes a setting signal, and the welding current can be set in the correct direction and proportion with this signal, so that in the process of adjusting the current There is no overshoot.

Within the table of nominal values associated with the respective welding current waveforms, there are sub-tables specifically selectable for the respective welding frequencies _fS , which are described in more detail further below. The nominal value of the current curve selected by the input signal W _Tab and its first derivative are stored in the respective nominal value table. For each measurement and chopping interval, the corresponding nominal value from the table is multiplied by the required current amplitude in multiplier 54 and then fed to summing junction 56 as a nominal value. The desired current amplitude is fed as signal I _soll via A/D converter 25 to multiplier 54 where it is multiplied by the nominal current value from memory 52 . The required current amplitude _Isoll can also be influenced by "manual" input or alternatively or auxiliaryally by the welding machine control system 19 (Fig. Variation process, so that within one half-wave of the primary AC voltage, for example, slope the top of the pulse more and more, as shown in Figure 7a-7c, or give it more or less humps or valleys, as shown in Figure 8a- 8c or 9a-9c.

As previously mentioned, memory 52 contains one nominal value table for each type of current waveform, four nominal value tables in the exemplary embodiment shown. In each table, the required welding current waveform is stored with a number of pre-selected current nominal values. In this example, 256 nominal values are stored for each welding current cycle. With a welding frequency of 500Hz and a chopping frequency of 10KHz, 10 chopping or switching intervals of 100μs per half-wave respectively are obtained. The welding current can thus be chopped 10 times per half-wave, ie switched on and off 10 times. Of the 256 available welding current nominal values, 20 welding current nominal values can therefore be selected per cycle, that is to say, 10 nominal values per half-wave, and used for a nominal one in the regulator 18. Actual value comparison. If the welding frequency is only up to 50Hz, then the welding current can pick up 200 nominal values per cycle, so each half wave is 100 nominal values. Depending on the selected welding frequency f _S , an appropriate sub-table in the table of nominal values corresponding to the welding current waveform is selected by the A/D converter 27 .

The change amount from one welding current nominal value to the next current nominal value is also stored in the nominal value table, that is, the dI/dt values in the sequence of 256 preset welding current nominal values. If working with a welding frequency between 35 and 40 Hz, all 256 points in the nominal-actual comparison are utilized. However, a welding frequency of 500 Hz is commonly used, so that only 20 welding spots per cycle are used for the welding current in the nominal-actual comparison. Therefore, if, instead of the nominal value table containing 256 nominal values, a subtable of higher welding frequencies above _fS is selected, the computer automatically adapts the delta to it so that the delta corresponds to the welding current standard. Selected grading between scales. Another possibility is that instead of presetting the nominal value table from the start with 256 points per welding current cycle and then selecting the subtables with fewer welding current nominal values, precalculate these subtables and make them comparable to The change from nominal to nominal is collectively selected as a table of nominal values in memory 52 .

The nominal current value delivered by the memory 52 corresponds exactly to the requested welding current pattern, but still does not correspond to the requested amplitude. As illustrated, the desired magnitude is selected by means of independent factors which may be fed into multiplier 54 through the other three inputs mentioned above.

The adjustment process operates as follows: Referring to the example cited above, it is assumed to work with a welding frequency fs of 500 Hz and a chopping frequency of 10 KHz. The welding current I has a sinusoidal waveform and is obtained by pulse width modulation of the primary AC voltage U in the manner shown in FIG. 4 . The table of nominal values contains 10 nominal values per half-wave of the welding current I. The smooth DC voltage delivered by the DC intermediate circuit 14c is chopped at 10 KHz, thereby generating a welding current curve corresponding to the nominal value of the current. The measuring frequency for determining the actual value of the welding current from the current transformer 20 is equal to the chopping frequency. The actual value of the welding current is thus determined for the respective nominal value of the welding current. In each nominal-actual value comparison, it is determined whether the measured actual value is equal to the nominal value of the welding current present in the table of nominal values. If this is not the case, the summing point 56 and the PID control circuit 50 deliver an error signal from which, in the manner described above, the setpoint signal for the corresponding duty cycle is formed by means of the feedforward signal. With this setting signal the duty cycle is influenced in such a way that the ratio between the pulse duration and the pulse interval during the pulse width modulation of the primary AC voltage is used to eliminate the difference between the actual value of the welding current and the nominal value of the welding current way to be corrected.

The welding current can thus be adjusted within a half-wave of the welding current, ie within a welding spot, with extremely short adjustment times. A further special advantage of this regulation method is that the required current waveforms can additionally be stored in the form of a nominal value table and selected as required. The welding current waveform can be chosen arbitrarily within certain limits, which are actually only set by the machine (e.g. For example, if there is a maximum possible increase in the welding current curve, which cannot be exceeded due to various physical factors, etc.).

In so-called fully sinusoidal welding of tanks between upper and lower welding rollers (as shown here with welding electrodes 10 and 12), the heating distance over the total contact length between welding roller and metal sheet is divided into six phases. , these phases are caused by a welding speed of 60 m/min and a welding frequency of 500 Hz and a total contact length of 3 mm, and generate three half-waves, these phases are divided into three cold periods and three hot periods (see " Soudronic" Company Journal, First Year Publication No. 1, June 1985, p. 3). The creation of each weld spot between the welding rollers thus involves three alternations between heating and cooling. The regulation method according to the invention allows optimum control of the heating and cooling phases within a solder joint. Figures 7-9 show suitable welding current forms for this purpose. Therefore the present invention adapts to the welding situation of different materials. Metal sheets which hitherto could only be welded with sputtering can now be welded well with flat-top welding current pulses without current peaks.

Figure 14 shows the current profile in which the welding current initially increases sinusoidally in each half-wave, then decreases before reaching the peak of the sinusoidal wave, increases again and then decreases to a zero crossing. For this particular embodiment of the invention, a very effective, purposeful influence on the solder joint forming heat (no fluid phase) is obtained. In roll seam welding it is possible, for example, to work with a welding frequency of 500 Hz, a welding current of 3700 A and a welding speed of 60 m/min with very good results even with sheet metal qualities which are often difficult to control.

Figures 15 and 16 show other optimum current waveforms where there are repeated dips in the welding current in the middle of the half-wave; Figure 15 begins with a sinusoidal increase from the zero crossing; A peak grows linearly. With these current waveforms, the maximum welding speed can be obtained with a low welding frequency, which prevents overheating of the welding equipment and produces small energy losses. For example, for roll seam welding, frequency 250Hz, current 3780A, speed 60m/min can be specified.

FIG. 17 shows the optimum current waveform in each case with a slowly falling current profile in the middle of the half-waves. Depending on the quality of the sheet metal a large welding area can be obtained with this process (between bonding and sputtering limits).

Fig. 18 shows the triangular wave change process of the current. Here, advantages can be obtained in particular in the welding of (non-tinned) metal sheets with unconventional coatings.

Figure 19 shows a similar current waveform with a slower energy supply to the material being welded.

Figures 20 to 28 show the current waveforms in various cases for keeping the welding current constant for certain times within the half-wave. In the particular welding situation, this results in a particularly good energy supply to the weld zone.

Figures 29, 30 and 31 show the profile of the welding current decreasing more or less slowly during the half-wave.

Figures 32 to 39 show the current waveforms in various cases where the energy supply is significantly reduced during the half-wave by reducing the current to a value of zero, or the current is briefly inverted during the half-wave.

Fig. 40 shows a current waveform having several constant sections, where the first constant section has a larger amplitude than subsequent constant sections.

The current waveforms shown and additional current waveforms can be generated using the apparatus described below. FIG. 11 shows a simplified circuit diagram of a resistance seam welding machine for longitudinal seam welding of round tank bodies, not shown, between roller-shaped welding electrodes 10 and 12 . The resistance seam welding machine comprises a static frequency converter 14 with a mains-powered input stage 14a represented by lines L1-L3, which is connected via a conventional DC intermediate stage circuit 14c to an output stage 14b designed as a chopper. The output stage 14 b is connected to the primary circuit of a welding current transformer 16 , to which the primary alternating voltage Up is delivered, the secondary circuit of the welding transformer 16 is connected to the welding electrodes 10 and 12 .

According to the extended circuit diagram shown in Figure 12, the input stage 14a of the static frequency converter 14 has a three-phase rectifier, which simultaneously forms the input of a DC intermediate stage circuit 14c, which is generally known because it is useful for the various aspects of the present invention. The situation is not important and does not require a detailed description. As shown in FIG. 12, the chopper in the output stage 14b of the frequency converter 14 (FIG. 11) comprises a bridge circuit having transistors T1-T4 as circuit elements and freewheeling diodes F1-F4 connected in parallel with these transistors. The four gate drivers are connected to the transistors and freewheeling diodes in the manner shown in FIG. 12 and are controlled by regulator 18 (FIG. 11) via line 15. change in welding A current transformer 20 for detecting the actual value of the current in the primary circuit of the welding transformer 16 is arranged in the primary circuit of the transformer 16 .

According to the circuit diagram in FIG. 11 , the current actual value from the current transformer 20 is transmitted via an A/D converter 22 to the input of a controller 18 designed as a process computer. On the regulator 18 , the nominal value I _soll of the welding current or the welding frequency can be set via the potentiometers 24 and 26 . The analog voltages developed on potentiometers 24 and 26 are applied to the process computer via A/D converter 25 or 27 . Alternatively, the value I of the welding current source can be fed into the regulator 18 via an input marked “manual” or via the welding machine control system 19 . This value is combined with the nominal welding current _Isoll in order to allow the current to be non-constant, for example on one tank. In this way, the welding machine control system 19, which knows the position of the can body to be welded at each moment, can therefore change the set nominal value I _soll , so that welding at each point of the can body has an appropriate welding current amplitude. The controller 18 determines the setpoint by means of a nominal-actual comparison of the welding current, which is delivered to the selected output stage 146 of the frequency converter 14 (FIG. 11) via the A/D converter 28 and the line 15. pass driver (Figure 12). This setting affects the duty cycle of the chopper in the output stage 14b to the rectangular pulse formed by the smooth DC chopping in each half-wave from the DC intermediate stage circuit 14c, so as to thereby pass the primary AC voltage at the affected Pulse width modulation of the duty cycle to regulate the welding current, as described in more detail below with reference to FIG. 13 .

Various ways of generating the primary AC voltage by chopping the smoothed DC voltage into rectangular pulses are shown in Figures 14-30.

Figure 13 shows the regulator 18 in more detail. As already mentioned above, the controller 18 is designed as a process computer, only the parts necessary for the invention are shown in FIG. 3 and described below. It comprises a PID regulation circuit 50 and a welding current reference element 52 in the form of a memory containing a current scale for each chopping interval corresponding to the welding current waveform for comparison with the respective current actual value determined in each chopping interval. value. For each welding current waveform (sine wave, triangular wave, trapezoidal wave, etc.), the memory 52 contains a table of nominal values selectable via the input W _Tab . An output of the memory 52 is connected to an input of a multiplier 54 . The output of multiplier 54 is connected to summing junction 56 . A summing junction 56 combines the input signal received from the multiplier 54 with the current actual value. The output signal of the summing junction 56 formed by the nominal-actual value comparison is applied to the input of the PID control circuit 50 .

The PID control circuit 50 transmits the setpoint signal at its output to an input of a summing point 58 . The other output of memory 52 is connected to the other input of summing junction 58 through feedforward or forward drive loop 60 . Through the feed-forward loop, the memory will transmit to the multiplier 54 the variation from the actual current nominal value to the next nominal value, that is, the welding current curve at the actual current nominal value in the direction of the next current nominal value. The order derivative dI/dt or delta is passed to summing junction 58 . The direction data is combined with the output signal of the PID regulating circuit 50, so the output signal of the summing point 58 represents a setting signal, and the welding current can be set along the correct direction and ratio with this signal, so that in the process of adjusting the current There is no overshoot in .

Within the table of nominal values associated with the respective welding current waveform, another sub-table can be selected in particular for each welding frequency _fs , which is described in more detail below. The nominal value of the current curve selected by the input signal W _Tab and its first derivative are stored in the respective nominal value table. For each measurement and chopping interval, the corresponding nominal value from the table is multiplied by the required current amplitude in multiplier 54 and then fed to summing junction 56 as a nominal value. The desired current amplitude is transmitted as signal _Isoll via A/D converter 25 to multiplier 54 where it is multiplied by the nominal current value from memory 52 . The required current amplitude _Isoll can also be influenced by "manual" input or alternatively or auxiliaryally by the welding machine control system 19 (Fig. Variations such that within one half-wave of the primary AC voltage, for example, slope the top of the pulse more and more, as shown in Figures 7a-7c, or give it more or less humps or valleys.

As previously described, for each type of current waveform memory 52 contains one nominal value table, four nominal value tables in the exemplary embodiment shown. The required welding current waveforms are stored in tables by pre-selecting several current nominal values. In this example, 256 nominal values are stored for each welding current cycle. Using a welding frequency of 500Hz and a 10KHz Chopping frequency, 10 chopping or switching intervals of 100 μs each are available per half wave. The welding current can thus be chopped 10 times per half-wave, ie switched on and off 10 times. Of the 256 available welding current nominal values, 20 welding current nominal values can therefore be selected per cycle, that is to say, 10 nominal values per half-wave, and used for a nominal one in the regulator 18. Actual value comparison. If the welding frequency is only up to 50Hz, then 200 nominal values can be selected per cycle of the welding current, so each half-wave is 100 nominal values. Depending on the selected welding frequency fs, an appropriate sub-table in the table of nominal values corresponding to the welding current waveform is selected by the A/D converter 27 . The change from one welding current nominal value to the next current nominal value is also stored in the nominal value table, ie the dI/dt values in the sequence of 256 preset welding current nominal values. If working with a welding frequency between 35 and 40 Hz, all 256 points in the nominal-actual comparison will be utilized.

However, a welding frequency of 500 Hz is commonly used, so only 20 welds per welding current cycle are used in the nominal-actual comparison. Therefore, if instead of having a nominal value table with 256 nominal values, a sub-table of higher welding frequencies exceeding fs is selected, the computer automatically adapts the variation to it so that the variation corresponds to the difference between the nominal values of the welding current. between the selected ratings. Another possibility is that instead of starting with a table of nominal values with 256 points per welding current cycle preset and then selecting fewer welding current nominal value selection sub-tables, these sub-tables are pre-calculated and made comparable to those from The nominal-to-nominal deltas are collectively selected as a table of nominal values in memory 52 . The nominal current value delivered by the memory 52 corresponds exactly to the desired welding current waveform, but still does not correspond to the desired amplitude. As illustrated, the desired magnitude is selected by means of independent factors which may be fed into multiplier 54 through the other three inputs mentioned above.

The adjustment process operates as follows: Referring to the example cited above, it is assumed to work with a welding frequency f _S of 500 Hz and a chopping frequency of 10 KHz. The welding current I is sinusoidal and is obtained by pulse-width modulation of the primary AC voltage U in the manner shown in FIG. 14 . The table of nominal values contains 10 nominal values per half-wave of the welding current I. The smooth DC voltage delivered by the DC intermediate circuit 14c is chopped at 10 KHz, thereby generating a welding current curve corresponding to the nominal value of the current. The measuring frequency at which the actual value of the welding current is determined by the current transformer 20 is equal to the chopping frequency. The actual welding current value is thus determined for the respective welding current nominal value. In each nominal-actual value comparison, it is determined whether the measured actual value is equal to the nominal value of the welding current preset in the nominal value table. If this is not the case, summing junction 56 and PID control circuit 50 deliver an error signal from which a setpoint signal for this duty cycle is formed in the manner described above by means of a feedforward signal. With this setting signal the duty factor is influenced in such a way that the ratio between the pulse width and the pulse interval during the pulse width modulation of the primary AC voltage is used to eliminate the difference between the actual value of the welding current and the nominal value of the welding current way to be corrected.

The welding current can thus be readjusted within a half-wave of the welding current, ie within a welding spot, with extremely short adjustment times. A further special advantage of this regulation method is that the required current waveforms can additionally be stored in the form of a nominal value table and selected as required. The welding current waveform can be chosen arbitrarily within certain limits, which are actually only set by the machine (for example, if there is a maximum possible increase in the welding current curve, which cannot be exceeded due to various physical factors, etc.).

In so-called fully sinusoidal welding of tanks between upper and lower welding rollers (as shown here with welding electrodes 10 and 12), the heating distance over the total contact length between welding roller and metal sheet is divided into six phases. , these phases are caused by a welding speed of 60 m/min and a welding frequency of 500 Hz and a total contact length of 3 mm, and generate three half-waves, these phases are divided into three cold periods and three hot periods (see " Soudronic" Company Journal, First Year Publication No. 1, June 1985, p. 3). The creation of each spot between the welding rollers thus consists of three alternating actions between heating and cooling. The regulation method according to the invention allows optimal control of the heating and cooling phases of a solder joint. Therefore the present invention can adapt to the welding situation of different materials. Metal sheets which hitherto could only be welded with sputtering can now be welded well with flat-top welding current pulses without current peaks.

Claims (24)

1. The method of resistance welding with periodic half-wave pulsation, especially AC welding current, which is generated by the primary AC voltage and controlled by pulse width modulation, is characterized in that: in each half-wave, the pulse width is adjusted by the chopping frequency The primary AC voltage in the modulation process is chopped, and the chopping frequency is a multiple of the welding current frequency to generate a specific welding current waveform. 2. The method according to claim 1, characterized in that the chopping frequency is selected to be 20 times the frequency of the welding current. 3. A method as claimed in claim 1 or 2, characterized in that the welding current is controlled within half-waves on the basis of a nominal-actual value comparison by influencing the duty cycle according to pulse width modulation. 4. The method according to claim 3, characterized in that: each half-wave of the primary AC voltage is chopped n times to realize n times of nominal-actual value comparison of welding current and n times of influencing the duty factor. 5. A method as claimed in claim 1, characterized in that the waveform of the welding current can be preselected by means of a table of nominal values. 6. A method as claimed in claim 5, characterized in that a table of nominal values for each welding frequency corresponding to the desired welding current waveform is available in the memory (52). 7. The method of claim 1, wherein the waveform of the welding current is non-sinusoidal and substantially trapezoidal or triangular, respectively. 8. A method as claimed in claim 1, characterized in that the welding current initially increases mainly sinusoidally after the zero crossing point, drops and increases again before reaching the peak of the sinusoidal wave, and then falls mainly sinusoidally to the zero crossing point, the welding current at Initially the main sinusoidal growth after the zero crossing, then repeating the process of taking a decline and then growth, and then the main sinusoidal decline to the zero crossing. 9. A method as claimed in claim 7, characterized in that the welding current generally increases linearly from the zero crossing point, repeatedly decreases and increases again in the peak region of the half-wave, and then generally linearly decreases to the zero crossing point. 10. The method according to claim 7, characterized in that the welding current follows a triangular wave variation process. 11. The method of claim 7, wherein the welding current follows a trapezoidal wave variation process. 12. A method as claimed in claim 11, characterized in that another trapezoidal current change occurs between the shoulders of the longer trapezoidal change. 13. The method of claim 11, wherein a plurality of similar trapezoidal current changes occur between the shoulders of the longer trapezoidal changes. 14. A method as claimed in claim 7, characterized in that the welding current initially increases in a generally linear manner, then generally decreases in a linear manner and then increases, and then decreases in a linear manner again. 15. A method as claimed in claim 8, characterized in that the drop of the current within a half-wave occurs up to zero or below the zero value of the current. 16. A method as claimed in any one of claims 7 to 15, characterized in that the welding frequency is 500 Hz or 250 Hz. 17. Apparatus for implementing the method as claimed in any one of claims 1 to 16, the apparatus comprising a static frequency converter (14) having a DC intermediate stage circuit (14c) and a A chopper (14b) for the output stage, which generates the primary AC voltage (Up) and transmits the primary AC voltage (Up) to the secondary circuit connected to the welding electrodes (10 and 12) of the resistance welding machine The welding transformer (16) is characterized in that it contains a control device, by means of which the chopper can control the multiple chopping of each half-wave of the primary AC voltage, and the control device has a regulator (18), said The regulator (18) is connected to the chopper (14b) of the frequency converter (14) for controlling the welding current (I) by pulse width modulation of the primary AC voltage and has a welding current reference element, the regulator (18 The welding current reference element of ) is a memory ( 52 ) containing a current nominal value for comparison with each current actual value determined for each chopping interval, corresponding to each chopping interval of the welding current waveform. 18. The device according to claim 17, characterized in that the chopper (14b) of the frequency converter (14) includes transistors (T1-T4) as circuit elements, freewheeling diodes (F1-F4) connected in parallel to the transistors (T1-T4) bridge circuit. 19. A device as claimed in claim 17, characterized in that the regulator (18) is provided with a memory (52) and an input Wrab for selecting a different address of said memory. 20. Apparatus according to claim 19, characterized in that the regulator (18) is provided with an adjustable welding frequency input. 21. The device according to claim 19, characterized in that the regulator (18) contains a PID regulation circuit (50), and the PID regulation circuit is equipped with a feed-forward loop. For each nominal value of the current in the table, store There is a change value to the subsequent current nominal value, and the change value is provided to the output of the PID regulation circuit (50) as a feedforward value. 22. Apparatus as claimed in any one of claims 16 to 21, characterized in that a multiplier (54) is connected to the memory (52), the multiplier rated with an adjustable current capable of passing through the regulator (18). The selection factor entered at the value input (Isoll) is multiplied by the nominal value of the table. 23. Apparatus as claimed in claim 22, characterized in that the multiplier (54) has further inputs through which other factors can be entered manually or by the superimposed welding machine control system (19). 24. Use of a device according to claim 17, characterized in that the device is used in a resistance welding machine for joint roll seam welding, spot welding or sheet metal blank welding.

Images