WO2011027238A1

WO2011027238A1 - Device and method for controlling a welding machine, in particular for manufacturing by means of material deposition

Info

Publication number: WO2011027238A1
Application number: PCT/IB2010/053141
Authority: WO
Inventors: Giovanni Muscato; Luciano Cantelli; Giacomo Spampinato
Original assignee: Universita' Degli Studi Di Catania
Priority date: 2009-09-02
Filing date: 2010-07-09
Publication date: 2011-03-10
Also published as: EP2473308A1; IT1395636B1; ITRM20090448A1

Abstract

A device and method for controlling a welding machine, in particular for manufacturing by means of material deposition, comprising a welding machine, for example of the inert gas (TIG) type, adapted to find and maintain stable operating conditions of the welding machine, making the machining process completely automatic. In particular, the material contribution is adjusted according to the length of the welding voltaic arc, in order to ensure a constant thickness of the weld material layer, i.e. a correct distance of the head from the part being produced. The control method comprises a step of acquiring information on the distance of the head from the part being produced. Such an adjustment is carried out when depositing each single layer in order to possibly correct the imperfections of a previously deposited layer.

Description

DEVICE AND METHOD FOR CONTROLLING A WELDING MACHINE, IN PARTICULAR FOR MANUFACTURING BY MEANS OF MATERIAL DEPOSITION

Field of the invention

The present invention relates to a device and method for controlling a welding machine, in particular for manufacturing by means of material deposition.

State of the art

Automatic material deposition techniques, one layer at a time, by means of robotized welding machines are known. In particular, techniques of producing parts from CAD models are known. The main techniques are of SMD (Shaped Metal Deposition) type, as well as of DLD (Direct Laser Deposition) type.

The SMD type process relies on the progressive addition of weld material for the production of a mechanical part, in order to obtain the desired final shape. This means that there are no reject products and that the weld material and the subsequent finishing operations are reduced to the necessary minimum. Welding control systems adapted to suppress human process supervision of the material disposition process are not however known.

A welding system includes using a welding head 1 , the following parameters of which are adjustable: the heat of the source, corresponding to the intensity of the supply current of the head, the weld material mass which flows to the welding head, the thickness of the material layer deposited on the part being machined, and the travel speed HS of the welding head. The travel speed HS of the welding head and the current intensity may be maintained constant during material deposition, as may the thickness SH of the layers and therefore their width WY on a part 10 being produced. In other words, the travel speed HS of the head is imparted by the machining times, while the perfection level of the surface finish depends on the thickness SH of the deposited layers.

It is thus apparent that the amount of weld material which reaches the head is the main control parameter for SMD welding.

Figure 1 of the prior art diagrammatically shows the operation of a TIG welding machine during a deposition process.

The current distance between the welding head and the deposited part is one of the most important information on the welding process in order to ensure deposition process stability.

During a manufacturing process by deposition, the distance between the head of the welding machine and the deposited part is fundamental, because when too much material is deposited and the travel speed and the intensity of the current supplied to the head are not sufficient to melt and deposit the weld material, the height of the head from the deposited material weld pool tends to be reduced layer after layer, making the length of the arc too short and inducing vibrations in the welding head due to the excess of weld material.

The prior art thus teaches to automatically adjust some machining parameters and maintain others constant, including the flow of weld material, which is in the worst case set by hand by an operator. In particular, the operator adjusts the feed of weld material by the welding machine if weld material splattering or a welding head vibration are found. The first case corresponds to an insufficient contribution of material by the feeder to the welding head; on the other hand, the second case corresponds to an excessive contribution of material.

Regretfully, however, the results are not satisfactory because the control system does not take the machining results into account, thus often causing a deposition of non-homogenous layers despite the operator's supervision.

Summary of the invention

It is an objective of the present invention to provide a control device for a welding machine, in particular for manufacturing by means of material deposition, adapted to solve the aforesaid problem, and in particular adapted to suppress the operator's supervision, making the welding process completely automatic.

It is an object of the present invention a control device for a welding machine, in particular for manufacturing by means of material deposition, which, in accordance with claim 1 , is adapted to adjust a flow of weld material according to the distance of the head of the welding machine from a weld material deposition point and adapted to detect such a distance by means of:

- an optical measuring system either capable of measuring the distance of the head from the deposition point or capable of obtaining said distance by measuring the arc luminosity, or

- an audio detection system, or - an arc voltage detection system.

Advantageously, information on the amount of material deposited layer after layer is obtained from information on the distance of the welding machine head from the part being produced, by means of which the weld material feeder is controlled. Whereby, welding process stability is sought by avoiding the deposition of an incorrect amount of material.

Thus, by controlling the feeder, a correct distance from the welding machine head to the part being machined is ensured, rather than moving the head with respect to the part being produced.

Therefore, an arc voltage error or a distance error measured by said optical or audio system is interpreted by the control system as an error of the amount of material deposited during the deposition of one or more previous layers.

The production result advantageously leads to an optimal deposition of weld material and to qualitatively better, and entirely automated, manufacturing processes.

A further objective of the present invention is to provide a control method of a welding machine, in particular for manufacturing by means of material deposition, adapted to solve the aforesaid problems.

It is another object of the present invention a control method of a welding machine, in particular for manufacturing by means of material deposition, in accordance with claim 7.

According to another aspect of the invention, it is a further object of the present invention an automatic welding machine in accordance with claim 8.

The automatic welding machine may be of any type, including TIG or laser types. The best application of the invention is obtained by combining a robot arm with a welding machine controlled by a control device or according to the control method illustrated in this invention.

The dependent claims describe preferred embodiments of the invention forming an integral part of the present disclosure.

Brief description of the drawings

Further features and advantages of the present invention will be apparent in the light of the detailed description of preferred, but not exclusive, embodiments of a control device and method for a welding machine, in particular for manufacturing by means of material deposition, illustrated by way of non-limitative example, with reference to the accompanying drawings, in which:

Fig. 1 depicts a weld material deposition process according to the prior art,

Fig. 2 shows a controlling curve of the feeding of weld material to the welding machine, in which a stable balance point of the supply is set beforehand, according to the present invention,

Fig. 3 shows a curve similar to that shown in figure 2, according to the present invention, in which said stable balance point is iteratively identified by moving along opposite arrows parallel to the ordinate axis,

Figs. 4 and 5 show the arc voltage pattern during a deposition test and the weld material flow during the same test, respectively.

The same reference numbers and letters in the figures refer to the same elements or components.

Detailed description of a preferred embodiment of the invention

According to a first variant of the invention, the distance between the head and the part being produced, corresponding to the arc length, may be measured by an optical measuring system provided with specific software. We believe that such a solution, although valid, is not optimal due to the high brightness and temperature reached by the arc, while inducing the use of highly sophisticated, expensive systems.

According to further variants of the invention, said distance may be indirectly determined either by the intensity of the arc brightness, by the noise emitted by the arc or by the voltage associated with the arc, there being a relationship between these aspects.

We believe that a more simple, advantageous variant provides for the flow of disposition material in relation to the arc voltage. Such a solution appears even more advantageous when considering that some welding machines of the prior art already monitor the arc voltage for other purposes.

According to a preferred variant of the control system according to the present invention, the arc reacts proportionally to the distance of the welding machine head from the welding point, and in particular the flow of weld material increases when said distance exceeds a first threshold, and vice versa, the amount of weld material is reduced when said distance is reduced below a second threshold. Said thresholds may coincide with a certain predefined optimal distance. In particular, said reaction is preferably linear. Thus, indicating by WF the variable amount of weld material which reaches the welding machine in the unit of time, i.e. the feeding of the welding machine, such a variable amount is imposed to be equal to a predetermined value WF , according to which the welding process is stable, plus an amount proportional to the error distance D with respect to a predetermined reference value Dref. Preferably, said proportionality depends on a gain K. An equation representing a preferred linear control is shown below:

WF = WF + k ^■ ( D - D _ref )

The gain k may be appropriately varied according to the method of detecting the distance of the welding head from the deposition point.

Said feeding value WF , which makes the welding process stable, is closely correlated to other parameters, such as the intensity of the current generating the arc, the travel speed of the head HS and the thickness SH of the layer of material to be deposited. Thus, when the feeding WF of the welding machine is correct, it corresponds to said predetermined feeding value WF , because the distance error D - Dref is zero. Figure 4 shows a chart related to the (continuous) control curve of the welding process, in which the feeding of the welding machine is shown on the ordinate and the distance difference is shown On the abscissa.

According to said preferred variant, WF may vary within the range of two, respectively minimum and maximum, values. Whereby, the pattern of said curve saturates at said limit values.

According to a further variant of the control method and device for a welding machine, said predetermined feeding value WF is not set beforehand but is automatically identified during the welding process itself. In other words, the stability point of the welding process is identified so as to minimize the corrective action of the control system, working as far away from said limit feeding values as possible.

Figure 5 shows a chart related to the (continuous) control curve of the welding process, wherein said balance point of the feeding of the welding machine is iteratively identified. Advantageously, said balance point may be identified by integration on the average error measured when depositing a previous layer. By making the concept explicit by means of formulae and indicating by WF, a balance point identified during the deposition of the i-th layer:

WF = WF + k - {D - D_ref)

WF_n = WF

where L is a learning coefficient, i.e. representing the learning dynamics of the control system and N is the number of sampling instants of the deposition process of a layer, whereby said balance point WFj is equal to the balance point identified for the previous layer WF, . ₁, plus the average of the distance errors detected when depositing the previous layer multiplied by said learning coefficient L. The coefficient u is needed to dimensionally convert the layer thickness error SH into the feeding (flow rate) of material that the feeder sends to the welding machine. Said learning coefficient L is chosen to make the layer thickness SH error tend to zero.

According to the previous variant, when Dref is defined, if the distance from the head of the layer does not exactly correspond to said reference value of the distance, it may occur, particularly when the first layers are deposited, that said wrong initial welding is interpreted as a deposition error. This implies an incorrect modification of the flow of weld material WF,. Such a phenomenon may lead to triggering oscillations which delay the identification of the system balancing point.

In order to solve such a problem, a further variant is shown below, in which the reference distance value is expressed in relation to a layer object of deposition.

The previous formulae thus become:

WF = WF_i + k - (D - D_refi)

WF, = WF

where the distance value is expressed by means of the following formula: ^L>ref_i - ^LJref_∞

)

k_D E [0,l[

where Dref« is the steady state reference value corresponding to the optimal distance, Dmeann is the average distance measured when depositing the previous layer and is a correction coefficient with respect to Dref-, which may be preset in a continuous range 0-1.

The system advantageously converges towards the optimal WF, value much faster, thus considerably reducing oscillations.

When depositing the first layer, the Dmean value is not known, then according to the present invention, it is replaced in said formula by an average arc voltage Dmean-rigwash measured during a so-called TIG wash, i.e. a run of the welding torch without material deposition. Said average distance value Dmean-ngwash should be preferably corrected by a correction factor k owire which takes into account the influence of the material contribution on the distance measurement into account. Then, the reference voltage Dref_FirstLayer is expressed according to the following formula when depositing the first material layer:

D ^rreeff _f FFiirrssttLLaayveerr = D ^rreeff ,∞∞

- k h N_voo _wWWⁱi.rree D n mean_TjgWash )

A localized reference may be predicted if some deposition parameters need to be modified in a given layer, such as, for example, the HS speed or current because the part should have a different width in that area, for example. In this case, the learning method is further varied by defining a position p so that said feeding WF is a variable depending on the position p of the welding head with respect to the part being machined, whereby the following equations apply for each layer:

WF(p) = WF(p)_l__l + k - (D - D_ref )

WF{p = WF(p)_l__l + L - k - (D - D_ref)

WF(p)₀ = WF(p)

As the control system is discrete over time, the balance point WF^p)^ related to the previous layer with respect to a position p is not available, but lookup table values may be used to calculate it.

Finally, by considering the response dynamics of the weld material feeder, the previous equation represents a linear control that may be expressed not only with respect to the position of the welding head but also with respect to time, becoming:

WF{p{t))_t = WF(p(t + delay)) _t__x + L^■ k^■ (D - D_ref) where delay represents the time delay of the weld material feeder in executing a welding machine feeding variation command. Thereby, a feeding variation control is imparted in advance with respect to when said variation is to be implemented, so as to obtain the correct deposition of material in the correct position.

All the described variants allow to find an optimal feeding value WF which allows a homogenous growth of the structure with constant resolution SH and width WY, maintaining constant parameters, such as relative travel speed HS of the welding head and current I.

Furthermore, all equations shown are expressed in terms of distance D of the head, but may be rewritten according to a voltage V of the voltaic arc, when the method of measuring the distance is based on the measurement of the arc voltage, the relationship being direct.

By way of test, a cylinder of 60 mm diameter and of 8 mm homogenous thickness was obtained by means of a welding machine controlled according to the presented method by depositing steel 316 on a 5 mm thick base.

For such a test, the distance was estimated with the aid of the arc voltage using the following equations:

WF

The test was conducted by using the following initial settings:

- HS 3.75 mm/sec

- SH 1 mm

- Amperage 50 A, - WFo 1300 mm/min,

with voltage reference fixed to 13.6 V, which approximately corresponds to a 4mm distance of the head from the deposition point.

This initial value WF₀ was chosen arbitrarily.

An upward deformation occurred towards a side of the basement during the step of preheating the basement, the chosen thickness of which is intentionally small. This is due to said small thickness, but allowed to test the adaption abilities of the control device in extreme conditions. Indeed, as may be observed by the arc voltage pattern over time, shown in figure 4, the deformation causes voltage oscillations when depositing the first layers, which oscillations are gradually attenuated at steady state.. In particular, the number of layers which are gradually deposited over time is shown on the abscissa.

Despite these deformations, the control system was able to compensate for said deformation, and to locally correct the distortion. This is detected by observing the arc voltage after approximately 15-20 layers reaching the set reference voltage level. Firstly, during the first four layers, only a proportional linear-type control was intentionally used, while for the others the proportional integral control was activated, e.g. via software, i.e. from the fourth layer onwards the control system operates to find the optimal amount WF of weld material, which in the light of the other fixed parameters settles at 2300 mm/min, so that the structure appears regular already from before the deposition of the tenth layer.

After deposition of the 25th layer, the travel speed HS is intentionally taken to 5 mm/sec. The new working conditions require a new point of balance (also referred to as working point) with a new -flow speed value -WF of -the- weld-material of -2800 - mm/min, which ensures a stable, regular growth of the structure.

As apparent in the chart shown in figure 5, the flow of material is stabilized despite said variation of HS after approximately 15 layers, i.e. about the 40th layer.

The advantages deriving from the application of the present invention are apparent; it is possible to:

- fully automate an industrial welding process, and in particular an SMD type process, - improve the deposition process, allowing a reduction of the thickness of deposited material to obtain an increase of resolution and a better finish of machined products,

- treat areas having different behaviors,

- adapt the operation of pre-existing automatic machines to the operating modes described in this invention.

A person skilled in the art can make changes to the variants of the present methods, in accordance with his/her knowledge of iterative method implementation.

Furthermore, a person skilled in the art knows methods for implementing said method by means of programming languages running on a computer and appropriate hardware interfaces.

The elements, features and procedures illustrated in the various embodiments may be combined without therefore departing from the scope of protection of the application.

Claims

1. A control device for a welding machine for manufacturing by means of material deposition, adapted to adjust a flow of weld material according to the distance of the head of the welding machine from a weld material deposition point and adapted to detect such a distance by means of:

- an optical measuring system capable of measuring the distance of the head from the deposition point and capable of obtaining said distance by measuring the arc luminosity, or

- an audio detection system, or

- an arc voltage detection system.

2. A device according to claim 1 , adapted to increase the flow of weld material to the head of the welding machine when said distance exceeds a first predetermined threshold and, vice versa, it reduces the flow of weld material to the head of the welding machine when said distance is lower than a second predetermined threshold.

3. A device according to claim 2, wherein said first and second thresholds coincide with a predetermined optimal distance.

4. A device according to claim 2 or 3, wherein said adjustment is of the linear proportional or integrative proportional type with respect to a previously deposited layer of weld material.

5. A device according to claim 4, wherein said adjustment depends on a position of the head with respect to an area of the part being produced and/or on time.

6. A device according to claim 5, wherein said adjustment is adapted to compensate for an actuation delay of the adjustment of the weld material flow.

7. A method for controlling a welding machine for manufacturing by means of material deposition, which welding machine comprises a control device of the welding head comprising means for measuring the distance of the head from the deposition point and means for controlling a movement of the head and means for controlling a flow of weld material, the method comprising a step of adjusting a deposition amount (WF) by means of the following formula:

WF = WF + k ^■ ( D - D _ref ) where WF is a predetermined value, k is a gain factor and D ref is a reference distance value.

8. A method according to claim 7, wherein said deposition amount of weld material WF is calculated for each i-th deposited layer by means of the following system of equations:

WF = WF + k - (D - D A

WF₀ = WF

where L is a learning coefficient, N is a number of sampling instants of a deposition process, and u is a conversion parameter for dimensionally converting an error of a layer thickness (SH) into a feeding (flow rate) of weld material.

9. A method according to claim 8, wherein said reference distance Dref of the welding head is calculated for each i-th step according to the following formula:

k_D [0,l[

where Dref_M is a predetermined reference value at steady state condition corresponding to the optimal distance, Dmeanj_i is a mean distance measured when depositing a previous layer and ko is a correction coefficient with respect to Dref. 0. A method according to claim 8 or 9, wherein before carrying out a first deposition of weld material, the method comprises a further step of measuring an average distance value Dmean-ngwash during a run of the welding torch without material deposition and a subsequent step of calculating the reference distance of a first layer deposition Dref_FirstLayer by means of the followin formula:

where k_Nowire is a correction factor taking into account an influence of the material contribution on an arc voltage measurement and KD is a pre-settable parameter in a continuous range 0-1.

1 1. An automatic welding machine comprising a welding control device configured to implement the method according to claims from 7 to 10.

12. A welding machine according to claim 11 , wherein the welding machine is of the TIG or laser type.

13. A computer program comprising program encoding means adapted to implement the steps of claims from 7 to 10, when said program is run on a computer.

14. Computer-readable means comprising a stored program, said computer-readable means comprising program encoding means adapted to implement the steps of claims from 7 to 10, when said program is run on a computer.