DE4203190C1 - Regulation and quality assessing of welding esp. spot welding - has ultrasonic detecting probe attached to welding electrode to record noise emission level at weld location - Google Patents

Abstract

In resistance spot welding an ultrasonic detecting probe is attached to one of the welding electrodes to record the noise emission level at the weld location. The maximum noise emission energy for a time window in each current half-wave is stored, starting with the current-carrying section formed by phase-chopping and ending 0.5-2 ms later. An ultrasonic emission value is formed from one or a number of the maximum energies as a measure of current density, which is used to calculate nugget dia. to be produced. ADVANTAGE - Weld quality can be assessed during welding.

Description

The invention relates to a method for controlling the welding process and for the evaluation of spot welds with the process steps the contacting of a weld metal with welding electrodes, wherein an ultrasound-detecting probe on one of the welding electrodes is attached to the welding electrodes with a Welding current for a welding time, and the detection of the sound emission level the welding point, the sound emission energy in one Evaluation circuit is evaluated over the welding time.

Such a method is known from DE-Z VDI reports No. 408 (1981), known in which the sound emission energy after the welding process is measured when the respective welding point cools and from this the Quality of the weld is estimated. However, this procedure points out the disadvantage that with him only the respective welded spot in Can be assessed afterwards. A regulation of one or more Welding parameters when creating each individual weld is not possible because the assessment of each welding point is problematic is. Only the entirety of the weld and thus z. B. the one prolonged decrease in the quality of the welding tool can be recorded and assessed.

From Materials Evaluation, Vol. 42, pp. 1652-1653 (1984) is a similar one Spot welding process known in which the appearance of welding spatter followed by observing the current sound amplitudes becomes. The welding current continues according to a predetermined maximum Half-wave number switched off or before, if a predetermined number of Sound emission signals, e.g. B. on the order of a few thousand, have been detected. This is only a regulation of the welding quality possible to the extent that the loss of quality that may result from  spatter occurs, compensated by a longer welding time becomes.

Another method is from DE-Z Werkstofftechnik, Vol. 17, pp. 264-269 (1986) known and considered the sum of impulses (number of sound impulses) and the energy of sound emission during welding. The quality of spot welds is assessed by the energy sum of the high-amplitude SE signals during heating during the welding process while suppressing the noise the magnetic reversal in the area of the current zero crossings and the Sum of impulses when cooling down after completion of the weld a discriminator threshold is considered. When a previously occurred specified number of noises in the form of overshoots of the SE signals, which indicate spatter, is the supply of welding current interrupted. A welding regulation to improve the welding result is carried out in this process so that a minimal SE energy sum is specified, if this is exceeded Total the power supply for the welding machine is turned off.

EP 1 78 565 A1 describes a method with an automatic acoustic Checking a bonding process known in the event of a deviation the sound values of predetermined guide values during the connection process regulation of machine parameters shall be.

DE 37 11 771 C2 describes a further control method for welding points known in real time, in which over measurements of the electrical Electrode parameters the welding result is influenced favorably.

Based on this prior art, the object of the invention is based on specifying a method of the type mentioned at the beginning that it allowed the quality of the same during each spot weld during the welding process itself and the expected lens diameter to determine. It is a further object of this invention  Welding time and the energy consumption to be used for each weld to minimize.

This object is achieved according to the invention for a method of the type mentioned at the outset in that the maximum sound emission _energy E _{1i is formed} for a first time window "1" in each current half-wave "i", which begins after the phase gating with the current-carrying section and between 0, 5 and 2 milliseconds wide is that one or more of the first N maximum sound emission _energies E _{1i of} the first time window with 2 N N₀ an ultrasound emission _{value of} the first time window is formed, which is proportional to the welding current density, where N₀ half the number of current half-waves used for welding , is, and that in the evaluation circuit from the welding current density thus determined, the lens diameter resulting from the ongoing welding is determined.

Because the initial noise emission during the welding process can be recorded in successive current half-waves the proportional relationship of the initial sound emission with the welding current density in each case the resulting lens diameter and thus the quality of the welding point even while it is being created be determined.

If an average value E _1m and a maximum value E _{1max of} the ultrasound _emission of the first time window are formed from the first N sound levels E _{1i of} the first time window with 2 N N₀, from which the mean and maximum welding current density then results, the prediction of the probable lens diameter can be improved will.

If the maximum E _{1max is} determined from the first n sound levels E₁₁ of the first time window with 2 <n <N₀ and the minimum E _1min that follows this maximum until the end of the welding _{time is} determined from the ratio of E _1min / E _1max Degree of melting can be determined and from this a value for the diameter of the welding lens.

The fact that a second time window is provided for each feeding half-wave of the welding current, which lasts from the end of the first time window to the beginning of the next first time window, that the sound integral E _2i for a second time window "2" in each current half-wave "i "is formed in each case that a maximum value E _{2max is formed} from the sound _integrals E _{2i of} the second time window determined so far, that from this maximum value E _2max a threshold value E _{2 threshold is} formed by division with the aid of a scaling _factor , when it _{falls below} this in one of the following second Time window in the evaluation circuit, starting from the melting point t _s determined therewith, the welding current density is calculated, and that from the welding current density determined in this way the lens diameter resulting from the ongoing welding is determined, it is now a with respect to the determination of the lens diameter with the first time windows Independently possible to determine this value with the second time window.

If then in the evaluation circuit using the first and with the help of the second time window three lens diameters can be compared with a lying in a predetermined frame Deviation of the values from each other the relevant Welding point based on the average of the three lens diameter values  can be assessed and otherwise the weld is not assessable with an indication the minimum of the theoretical lens diameter determined to be expelled.

The ultrasound emission during the ατ, τβ conversion of the iron is carbon-dependent. When the carbon content in iron is low, high noise emissions occur. With high carbon contents C <0.05 percent, on the other hand, there is almost no sound emission. The sound emission of the ατ, τβ conversion is recorded in the second time window F _2i . It is advantageous that not the contribution of the sound emission but the time range of the maximum emission is assessed in the windows F _2i . The evaluation is therefore independent of the carbon content of the iron.

Is the sound emission in the windows F _2i low and shows no pronounced maximum, e.g. B. at high carbon _contents , the welding lens _{evaluation is} carried out only with the sound emission from the first time window F _1i , which is largely independent of the carbon content. This can happen, for example, if the mean value of the maximum sound emission E _sm3 in the time windows F _{2i is} not at least by a predetermined factor, e.g. B. 2, is greater than the mean value of the minimum emission E _2m2 .

The method according to the invention is both Low carbon sheets as well Sheets with a high carbon content are equally good can be used, although in the latter only one by one Factor 10 lower noise emissions can be observed, which complicates the applicability of the known methods or prevented.  

Further advantageous embodiments of the invention are characterized in the subclaims.

Exemplary embodiments are given below the invention with reference to the drawings.

Show it:

Fig. 1a, a current-time diagram over several half-waves of the desired weld feeding welding current at a good-quality weld,

FIG. 1b a sound emission time diagram over the same half-waves of the welding current shown in FIG. 1a on the same scale,

Fig. 2a shows a current-time diagram over several half-waves of the desired weld feeding welding current at a lower quality weld,

Fig. 2b is a sound emission time diagram over the same half-waves of the welding current shown in Fig. 2a on the same scale,

Fig. 3 is a graph of the integral of the sound emission versus time for a good quality weld,

Fig. 4 is a diagram of the integral of the sound emission against time for a poor quality weld, and

Fig. 5 is a diagram of the obtained lens diameter of a weld against the determined fusing point.

The Fig. 1a shows a schematic representation of a current-time diagram for several rectified half-cycles of a welding current, with which a desired weld is heated.

The two sheets to be welded are in the desired one Layered on top of each other. In the case of the drawings The sheets shown are sheets with a carbon content of 0.01 percent.

The method according to the invention is both Low carbon sheets as well Sheets with a high carbon content are equally good can be used, although only a small one for the latter Noise emission can be observed.

The welding electrodes are placed on the sheets. A sound emission probe is attached to one of the welding electrodes, abbreviated SE probe, attached, the z. B. with realized a broadband piezoelectric sensor can be. When the weld metal is heated quickly During spot welding, there are various sources of acoustic emissions effective. The main sources are thereby rapid volume changes in the weld metal as a result of phase changes and due to thermal Grain expansion of the weld metal (thermoelastic Interaction).

The output signal of the probe acts on an amplifier and a high pass with which everyone detected Signals below a cut-off frequency of e.g. B. 20 to 200 kHz be filtered out. That then z. B. by 30-50 dB amplified sound signal acts on an evaluation device, which also with the signal about the strength of the welding current is applied. To get the  Energy information is the amplitude signal of the Sound squared.

In another embodiment of a device the SE probe also on one of the parts to be welded Sheets are attached. However, this creates in the laboratory possible arrangement problems in industrial cycle operation with workpieces to be welded successively.

The signal of the square of the welding current 1 present in the evaluation device is now plotted against time 2 in FIG. 1a. The current strength is marked in arbitrary units. In the selected exemplary embodiment, the time between two vertical markings 3 is always 10 milliseconds for a 50 Hz alternating current feeding the welding device. The welding current 4 has a cut of 3 milliseconds or 54 ° in each half-wave. FIG. 1a shows half of a weld in the time axis, which in the present exemplary embodiment lasts 200 milliseconds or 20 half-waves. The last 10 half-waves, not shown, are not accompanied by sound emissions relevant to the process.

The signals arriving in the evaluation circuit in each half-wave are recorded in several, at least three time windows 5, 7 and 8 (with associated sound energies E _ai , a = index of the time window, i = index of the current half-wave, here with 10 current periods from i = 1 to 20 ) which each contain a characteristic section of each half wave. Here, further time filters can be provided within the time windows 5, 7 and 8 at predetermined time intervals in order to filter out disruptive sound components which, for. B. can occur by breaking carbides in iron sheets with high carbon content. At the same time, a possible formation of splashes, which is noticeable by extreme SE signals (E _2i ), can be recognized, here advantageously in each case in the second time window 8 .

Each current half-wave begins with a cut of 54 °, ie after 3 milliseconds after each zero crossing and is divided into two time windows (with the determined energies E _1i and E _2i ) in the _{exemplary embodiment described} . A first time window 7 begins at the time of the start of the half-wave, ie at the beginning of the current flow and the heating of the welding material. This point in time is made available to the evaluation device via a trigger signal from the welding device. It has a time width of 0.5 to 5 milliseconds, advantageously 1 to 2 milliseconds. In this initial time of each current half-wave, the sound emission resulting from the heating and thus the energy E _1i is essentially proportional to the current density. The welding lens diameter to be expected in each case is determined from the current density by means of a predetermined table stored in the evaluation circuit. For each welding machine, these have been determined beforehand by individual calibration tests with certain welding currents.

The ultrasound emission in the first time windows 7 is essentially not influenced by phase changes. However, if the respective maximum of the ultrasound _emission E _{1i lies} in a certain such time window 7 at the start or end point of the window, then this value is recognized by the evaluation circuit as a value influenced by phase conversions and is not used.

Thus, by evaluating each ultrasonic signal of the initial current-flowing section of each current half-wave, an average value E _1m and a maximum value E _1max for the sound emission and thus for the welding current _density can be determined, from which the subsequent lens size can still be determined during the welding process.

The ultrasound emission _{value of} the first time window, which is proportional to the welding current _density , is formed from one or more of the first N maximum sound emission _energies E _{1i of} the first time window with 2 N N₀. The maximum number N₀ of time windows preferably corresponds to half the number of current half-waves used for welding, ie 20 current half-waves are used in the present exemplary embodiment, so that N₀ = 10 results.

This makes it possible, in particular, to already start the welding process in the described embodiment, e.g. B. after 11 or 12 current half-waves, since then the welding lens already has about 80 percent of its previously calculated size, if this resulting diameter is larger than the minimum diameter required in quality control. A value for the melted lens volume and the lens diameter can be calculated from the ratio E _1min / E _1max .

In a development of the method described above, the sound energy E _2i is also integrated for every second time window 8 . The absolute maximum of all sound energy _values E _{2max of} the time window 8 is then simultaneously determined in the evaluation circuit and divided by a predetermined numerical factor that _defines a relative threshold value and stored as E _{2 threshold} . The numerical factor is preselected so that when the resulting calculated _acoustic energy _{falls below} E _{2 threshold} in a later second time window 8 , the phase _{conversions have} just been completed and the melting point has been reached. This is explained in more detail in connection with FIG. 3.

The third time window 5 lies in the currentless time of the cut current half-wave. Sound signals are thus recorded in this time period, which are due to phase reverse conversions of the previously heated material.

FIG. 1b is a plot of acoustic emission 6 from the time 2 on the same semi-shafts 4 of the welding current shown in FIG. 1a on the same scale. The sound emission involves detected by the SE probe supersonic total energy starting from a lower cut-off frequency at 20 to z. B. 150 kHz can be selected to filter out disturbing mechanical vibrations of the environment. The energy is plotted in arbitrary units, a scaling factor of 10⁶ being used in FIG. 1b.

As can be seen in FIG. 1b, a sound emission signal 11 results in the time of the first feeding half-wave, which in its linear energy representation is small compared to the maximum value shown. However, disturbances can also occur with these sound energy values, so that in a preferred embodiment of the evaluation circuit, the sound energy values of the second sound window 8 are only used from i = 2, ie from the second half-wave. In the second half-wave, the preheated sheet metal material is further heated, resulting in a sound emission 12 in the time windows 7 and 8 . In the third half-wave there is a strong sound emission 13 in the window 8 , which is followed by a lower sound emission 14 in the fourth half-wave. The sound emission E₂₃ of the window 8 leads to the formation of the maximum value, which is compared below with the sound emissions E _2i (i 4). With a division _factor for determining the E _{2 threshold} of approximately six, the sound emission E 2 liegt is already below this threshold value, which results in a melting point after 35 milliseconds. In this case, the phase transformations took place in the period between approximately 15 and 35 milliseconds. The remaining welding time of approximately 165 milliseconds is used to build up the welding lens in order to liquefy further material and to take it into the welding lens.

In Fig. 2a the present in the signal evaluation device of the square of the welding current 1 is plotted against time for a 2 second, lower quality weld. The current strength is again marked in arbitrary units, but the same scaling factor as in FIG. 1a was used. It can thus be seen that a current that is half as small has acted on this second welding point. Furthermore, it can be seen that the gate is larger and is 72 ° or 4 milliseconds for the welding point described here, so that a total of significantly less energy is conducted into the welding point per half-wave.

FIG. 2b shows a diagram of the sound emission 6 versus time 2 over the same half-waves 4 of the welding current according to FIG. 2a on the same time scale. The energy is plotted in arbitrary units, with a scaling factor of 10 3 used in FIG. 2b has been. The maximum sound energy that occurs in the fourth half-wave is thus, at just under 200 · 10³ units, a factor of seven smaller than the maximum sound energy that occurs in FIG. 1b.

FIG. 2b clearly shows that by applying a weaker current with a larger phase angle to the welding point, the phase conversions take a longer period of time. They only start after about 25 milliseconds and end here after about 60 milliseconds.

In the first time window, the maximum number of which is advantageously smaller than the said number N₀, the sound emission is again evaluated and the current density is determined therefrom. The expected welding lens diameter then results from this. At the end of the weld, a value for the lens diameter can also be determined from the ratio of E _1min / E _1max . This is smaller here than in the welding according to FIG. 1, because the time remaining after the end of the phase transformations until the end of the welding, i.e. until the end of 20 half-waves here, i.e. 200 milliseconds, is only 140 milliseconds compared to 165 milliseconds in the case of FIG weld illustrated. 1 ie, this time is about 15 percent smaller than the good-quality weld.

FIG. 3 shows a diagram of the integral 21 of the sound emission energy 6 against the time 2 for a good quality weld. The curve characterizing this integral 21 of the sound emission 6 is referred to below as the SE sum 22 . In the exemplary embodiment under consideration, the following curve curve results.

From an SE sum 22 zero, the integral increases slowly in the first 15 milliseconds (curve section 23 ). This is followed by a steep rise 24 , which then flattens out again (section 25 ). The slope of the curve passes through a minimum 26 and then increases slightly for the remaining part of the curve 27 . The slope of the SE sum 22 is recorded continuously and the melting point t _s at the point in time 28 is determined in the minimum 26 mentioned.

The mean value of the minimum sound emission E _2m2 determined from the time windows F _2i with 2 <i <N₀ corresponds to the curve slope in the areas of minimum emission 23, 26 and 36 in _FIGS . 3 and 4. The mean value of the maximum emission E _2m3 corresponds to the curve _slope in the areas of maximum emission 24 and 31 in FIGS. 3 and 4. By determining the largest value "i", for which E _2i <(E _2m2 + E _2m3 ) / 2 applies, the end of the phase _transition and the Determine the start of melting t _s .

To determine the point in time t _s for the end of the ατ, τβ conversion and the start of melting, the procedure can also be followed in such a way that the mean value E _{2m1 is first} determined from the sound _integrals E _2i for the time range 2 <n <N₀ . Subsequently, two other average values E _2m2 and E are determined _2m3, whereby only values of E are taken into account _2i for the averaging of E _2m2 smaller than E _2m1 are and will be considered for the averaging of E _2m3 only values E _2i are larger than E _2m1 . A threshold value is formed from the two new mean _values E _2m2 and E _2m3 , e.g. B. by forming the mean (E _2m2 + E _2m3 ) / 2. The time t _s for the completion of the ατ, τβ conversion now results from the index "i" for which E _{2i is} for the last time greater than (E _2m2 + E _2m3 ) / 2.

The melting point t _s can also be determined in such a way that the time focus of all threshold violations is determined: t _s = ΣE _{2i *} i / ΣE _2i , where E _2i <(E _2m2 + E _2m3 ) / 2.

When the melting temperature in the weld metal has been reached, the lens volume has undergone all phase changes, and the SE of the thermoelastic interaction decreases from this point in time, since the welding lens temperature has a breakpoint at the melting temperature of the iron T _s = 1536 ° C, and the thermal grain expansion d1 due to dT = 0 over

d1 = α · 1 · dT

is also 0, with α being the coefficient of thermal expansion is designated.

The sound emission in the windows F _1i is proportional to the heated total volume V and the temperature gradient dT of the weld metal before the melting point is reached. After the melting point is reached, only the volume areas that have not yet melted (VV _s ) emit ultrasound. The ratio of the minimum SE at the end of the welding process to the maximum SE at the start of the welding process E _1min / E _1max corresponds to the ratio of (VV _s ) / V, _i.e .:

E _1min / E _1max = (VV _s ) / V = 1-0.6F _s / F,

where F is the electrode area and F _{s is} the cross-sectional area of the welding lens. From this relationship, z. B. calculate the welding lens diameter at the end of the weld.

From the melting point t _s can after

the welding current density I² / F² can also be determined. F is the electrode area, I the welding current, t₀ the currentless phase dependent on the leading edge angle, d the sheet thickness, R₀ the specific electrical resistance of the sheets to be welded, T _s the melting temperature, c _v the specific heat of the welding material, σ the density of the welding material and k is the heat transfer coefficient.

The advantage of this welding current density determination over a welding current measurement exists in particular in that the current density flowing through the weld spot is determined and leakage currents through shunts and electrode area changes due to wear and tear to disturb. At a predetermined welding time t, i.e. H. in the described embodiment 200 milliseconds the diameter D of the welding lens can then be directly by

can be determined, c _{s being} the specific heat of fusion.

FIG. 4 shows a diagram, analogous to FIG. 3, of the integral 21 of the sound emission 6 against the time 2 , but for a poorer weld.

Here too there is a slow increase 23 in the SE sum 22 at the start of the weld and a flat SE sum increase in the end section 27 .

However, there is an average welding area 31 over several half-waves in which the SE sum 22 rises, but no minimum increase in the SE sum 22 below a predetermined threshold value is achieved. A minimum 32 falling below the predefined threshold value is only reached at a much later time 32 , with which the poor quality of the welding point can be determined, since a longer welding time until the melting temperature is reached results in a smaller welding lens, as is evident from the experimentally determined connection . 5 seen in FIG.

FIG. 5 shows a diagram of the lens diameter 41 obtained against the melting time 42 . A series of determined measuring points 43 are drawn in, for which the respective lens diameter 41 has been measured afterwards and has been entered together with the melting point 42 determined during the welding. This results experimentally in the essentially linear relationship represented by straight line 45 , for which straight lines 46 are also entered at intervals of two standard deviations.

Since it is assumed in the embodiment under consideration of a minimum size 47 of welding spots is of the order of 4 millimeters, thereby resulting with the associated error bars two areas 48 and 49 in FIG. 5, by the two by 47 separate line segments of the line 45 and 46 form. Welding points for which a melting point in the region 49 has been determined are good quality welding points. At a melting point that is immediately determinable for each welding point and lies in the region 48 , the existence of a bad welding point can be determined immediately.

Because the SE total is evaluated in continuous operation the quality of each and every one of them Weld spot can be determined, with a good welding point for saving electricity, for quality assurance and power supply to avoid splashes after just 12 to 16 half-waves can be adjusted. Bad weld spots will be started again in automated operation or that corresponding, a certain number of bad Workpiece with welds is made from the Production cycle removed.

If a successive series of high-quality welding points occurs, the melting times of which are each located in the straight section in region 49 , which indicates the good quality of the welding electrodes and the metal sheets used, it can be provided that the power supply is cut off after such a welding period which has flowed half of the current required for the previous welds. This enables a further large energy saving to be achieved. This is possible because with good welds, the welding lens is 80 to 90 percent of its final size after half the welding time (in energy units).

However, the lens size reaches the specified minimum size not or only barely, then the welding current increases and when the maximum current value is reached the Weld stopped and a new welding tool requested.  

A particularly advantageous and reliable method can be achieved in that the calculated welding lens size stored in each case by the current density adjustment in the first time windows 7 is coupled with the welding lens size calculated by the melting point determination with the aid of the second welding window 8 . Then the two pre-calculated welding lens diameters are compared while the welding is in progress and their calculated difference is determined.

If both values are within a specified error Intervals, as they are assessed, is the determined one The difference, however, is larger, so this weld becomes marked as not assessable and / or the smaller one of the two values of the calculated welding lens diameter issued with the remark that this value can be faulty. Is the calculated difference the welding lens size within the given Frame, so is the average of the two lens diameters rated. Thus through this connection the two different applicable in the production process Investigation proceedings independently on two different ways a theoretical lens diameter specifiable, with which still in the middle of the production step, e.g. B. after half, e.g. B. 10 out of 20, current half-waves, the quality of the weld can be assessed in advance is.

Independently of this, a third welding lens _diameter can be determined at the end of the welding from the ratio of the minimum and the maximum sound emission in the time windows F _1i from the ratio E _1min / E _1max .

Of course, are in the first and second time slots at possible points of disruptive SE influences, the time window suppressing the sound recording there arranged. This is particularly the case in the first z. B. 50 microseconds of each pulse in the due to the strong electromagnetic field strength changes when the welding piece is loaded with the high currents interference pulses occur.

This gives you the opportunity to intervene regularly. On the one hand, this can be the one already described above earlier termination of the weld to be at energy saved and the welding tool is spared. These earlier termination may also be independent of others welds before or after for each one as a well predicted weld be provided. On the other hand, the evaluation circuit with looming bad quality Welds an increase in amperage by or and cause the gate angle to decrease.

Claims (12)

1. A method for regulating the welding process and for evaluating spot weld connections with the method steps of contacting a weld metal with welding electrodes, an ultrasound-detecting probe being attached to one of the welding electrodes, and applying a welding current ( 1 ) to the welding electrodes for a welding time ( 2 ), and the detection of the sound emission level ( 6 ) of the welding point, the sound emission energy ( 6 ) being evaluated in an evaluation circuit over the welding time ( 2 ), characterized in that in the evaluation circuit the maximum sound emission _energy E _1i for a first time window ( 7 ) " 1 "is stored in each current half-wave" i ", which begins after the phase gating with the current-carrying section and is between 0.5 and 2 milliseconds wide, that from one or more of the first N maximum sound emission _energies E _{1i of} the first time window ( 7 ) 2 N N₀ a U ltrasound emission value of the first time window ( 7 ) is formed, which is proportional to the welding current density, where N₀ is half the number of current half-waves used for welding, and that in the evaluation circuit from the welding current density thus determined, the lens diameter resulting from the current welding is determined. 2. The method according to claim 1, characterized in that from the first N maximum sound emission _energies E _{1i of} the first time window ( 7 ) with 2 N N₀ an average value E _{1m of} the ultrasound emission of the first time window ( 7 ) is formed, from which the average welding current density is determined becomes. 3. The method according to claim 1 or claim 2, characterized in that in the evaluation circuit with a threshold detector at the first sound emission energy E₁₁ of the first time window ( 7 ) is determined whether a high sound emission at the beginning and / or at the end of the first time window ( 7 ) is present, and that in the presence of such a high sound emission, the first maximum sound emission value E 1 is not used to determine the current density. 4. The method according to claim 3, characterized in that from the first N maximum sound emission _energies E _{1i of} the first time window ( 7 ) with 2 N N₀ the maximum value E _{1max of} the ultrasonic _{emission is} formed, from which the maximum welding current _density is determined. 5. The method according to claim 4, characterized in that from the sound emission _energies E _{1i of} the first time window ( 7 ) the following the maximum value E _1max minimum value E _{1min of} the ultrasonic _{emission is} determined, the search for the minimum value being carried out until the end of the welding time and the ratio E _1min / E _{1max is} determined, from which the melted volume and the lens diameter are determined. 6. The method according to any one of the preceding claims, characterized in that a second time window ( 8 ) for each feeding half-wave ( 11, 12, 13, 14, 15 ) of the welding current ( 2 ) is provided, which in each case from the end of the first time window ( 7 ) until the beginning of the next time window F _1i , that the sound _integral E _{2i is formed} for each second time window ( 8 ) "2" in each current half-wave "i", that in each case a maximum value E _2max from the sound _integrals E determined so far _{2i of} the second time window ( 8 ) is formed, that from this maximum value E _2max a threshold value E _{2Schwell is} formed by division with the aid of a scaling _factor , when it _{falls below} this in one of the following second time windows ( 8 ) in the evaluation circuit based on the melting point determined therewith t _s the welding current density is calculated, and that from the welding current density determined in this way the resulting L during welding inside diameter is determined. 7. The method according to claim 6, characterized in that the melting point is determined in that from the sound integrals E _2i with 1 <i <N₀ the mean value E _2mi and in a second evaluation _step the mean _values E _2m2 and E _{2m3 are} formed, whereby for the averaging E _2m2 only the sound _integrals E _2i , which are smaller than E _2m1 , and for the averaging E _2m3 only the sound _integrals E _2i , which are larger than E _2m1 , that a threshold value E from the mean values E _2m2 and E _{2m3 2Threshold is} determined, e.g. B. by renewed averaging (E _2m2 + E _2m3 ) / 2, and that by comparison with this threshold E _2Schwell the sound _integral E _{2i is} determined, which last exceeds this threshold. 8. The method according to claim 7, characterized in that the melting point is determined by determining the center of gravity of the sound integrals e _2i , which exceed the threshold value (E _2m2 + E _2m3 ) / 2. 9. The method according to any one of claims 6 to 8, characterized in that in the evaluation circuit, the lens diameter calculated with the aid of the first ( 7 ) and with the aid of the second ( 8 ) time window are compared, that with a deviation lying within a predetermined frame Values from each other, the welding point in question is evaluated on the basis of the mean value of the lens diameter values and that otherwise the welding point is shown as not assessable with an indication of the smallest of the theoretical lens diameters determined. 10. The method according to any one of claims 6 to 9, characterized characterized in that in the evaluation circuit Presence of a good weld with a prospect sufficient lens diameter the welding time is reduced to the time in which between 50 and 80 Percent of the welding energy customary for the welding process is introduced into the weld. 11. The method according to any one of claims 6 to 10, characterized characterized in that in the evaluation circuit Presence of a bad weld with a prospected insufficient lens diameter the welding current density for the remaining welding time is increased and / or the welding time to the time is extended by the finding of bad Quality of the welding point for the welding process corresponds to the usual total welding time. 12. The method according to any one of the preceding claims, characterized in that in the first and second time windows ( 7, 8 ) at possible points of disturbing SE influences additional time window suppressing the sound recording are arranged.