Method and device for optimizing process parameters and establishing a connection using a reactive multilayer system

Abstract

A method for optimizing process parameters of a process for creating a connection between two components using a reactive multilayer system (108) includes creating a feature vector from adjustable process parameters (144), determining an influence of the adjustable process parameters (144) encompassed by the feature vector on an artificial intelligence-based model (142), selecting a subset (146) of the adjustable process parameters (144) depending on a magnitude of influence, and determining optimized values ​​(148) for the subset (146) of the adjustable process parameters (144).

Description

[0001] The present invention relates to a method and a device for optimizing process parameters of a process for producing a connection between two components by means of a reaction of a reactive multilayer system, and to a method and a device for producing a connection between two components by means of a reaction of a reactive multilayer system using a manufacturing device.

[0002] A reactive multilayer system, for example a reactive multilayer film (RMS), can be used to connect elements.

[0003] The state of the art is disclosed in AH Sung: Ranking importance of input parameters of neural networks, in: Expert Systems with Applications, 15, 1998, 405-411 and in J. Wang et al.: Investigating the effect of applied pressure on reactive multilayer, in: Acta Materialia, 52, 2004, 5265-5274.

[0004] Against this background, the present invention provides an improved method and an improved apparatus for optimizing process parameters of a process for creating a connection between two components by means of a reaction of a reactive multilayer system, as well as an improved method and an improved apparatus for creating a connection between two components by means of a reaction of a reactive multilayer system using a manufacturing apparatus according to the main claims. Advantageous embodiments will become apparent from the dependent claims and the following description.

[0005] The production of a compound using a reactive multilayer system can be improved by adjusting adjustable process parameters of the compounding process using an artificial intelligence-based model of the process.

[0006] A method for optimizing the process parameters of a process for creating a connection between two components using a reaction of a reactive multilayer system comprises the following steps: Reading in adjustable process parameters that influence the process for creating the connection between the two components by means of the reaction of the reactive multilayer system; Creating a feature vector from the adjustable process parameters using a creation rule; Determining the influence of the adjustable process parameters encompassed by the feature vector on an artificial intelligence-based model of the process for establishing the connection; Selecting a subset of adjustable process parameters depending on the magnitude of the influence of the adjustable process parameters; Determining optimized values ​​for the subset of adjustable process parameters using an optimization rule.

[0007] The process for creating the compound can be carried out using suitable manufacturing equipment. To achieve the highest possible compound quality, it is important that adjustable process parameters, such as pressure or temperature during the manufacturing process, are set to optimal values. For the creation of the feature vector, established methods for generating feature vectors for an artificial intelligence-based model can be used. The model can then be used to predict the quality of the resulting compound. To determine the influence of the adjustable process parameters encompassed by the feature vector on the model, the model can be trained using the feature vector and subsequently analyzed.By selecting the subset, features with a minor influence within the model can be excluded from further use. This allows less relevant process parameters to be filtered out while retaining the relevant ones. Using this subset, an optimized feature vector can be created, which may include fewer process parameters than the original feature vector. Once the process parameters with a significant impact on the quality of the manufactured compound are known, optimized values ​​can be determined for these parameters. If, in a subsequent manufacturing process, the relevant process parameters are set to these optimized values, the resulting compound is highly likely to be of high quality.

[0008] Through activation and the resulting reaction of the reactive multilayer system, a stack of components and the reactive multilayer system can be assembled into a component. The component can thus be an assembly of components. The first component can be an electronic component or a sensor. For example, the first component could be a semiconductor device, a chip, or a wafer. For example, the component could be a transistor, a diode, or a force sensor. The second component can be a substrate for the first component. For example, the second component could be a substrate for electronic components, such as a printed circuit board or a substrate, such as a copper or ceramic substrate.The second component can also be a machine element, such as a connecting element or a force or motion transmission element, such as a shaft. The reactive multilayer system can be a reactive multilayer film (RMS) capable of providing instant heat for a wide variety of applications across many industries. The reactive multilayer system can be fabricated by vapor deposition of thousands of alternating nanoscale layers, for example, of aluminum and nickel. The reactive multilayer system can be arranged as a film between the components or comprise reactive layers that have been sequentially deposited directly onto one or both components to form the reactive multilayer system. The reactive multilayer system can perform the reaction in response to activation, such as ignition.For this purpose, the reactive multilayer system can be exposed to a laser beam or a spark, an ignition current can be passed through the reactive multilayer system, or the reactive multilayer system can be ignited using ultrasonic energy. This generates sufficient thermal energy to reach the temperature required to ignite the reactive multilayer system. Ignition can occur, for example, at one or more predefined positions, at a random position, or across the entire contact area between the reactive multilayer system and one of the components. After the reaction, the reacted reactive multilayer system can form a metallurgical bond between the components. The bond can be mechanically stable and electrically conductive and is also referred to as multilayer bonding or RMS bonding.

[0009] The procedure can include a step of training the model using the feature vector and quality data characterizing the connection quality. The model can then be trained to generate a prediction of the connection quality. Advantageously, the influence of the adjustable process parameters encompassed by the feature vector can be easily determined from the trained model. Optionally, the training step can also be performed to determine the optimized values ​​for the subset of adjustable process parameters. The model can, for example, be based on an artificial neural network.

[0010] During the initial data acquisition step, the first adjustable process parameter can be the pressure acting on at least one of the components during connection creation, and the second adjustable process parameter can be the temperature prevailing at at least one of the components during connection creation. This allows for the use of important and easily adjustable process parameters. Additional process parameters can also be used.

[0011] In the feature vector creation step, features in the statistical, temporal, and / or spectral domains can be extracted from the adjustable process parameters. This allows the use of a known feature extraction method.

[0012] In the determination step, the influence can be determined by weighting the adjustable process parameters in the model. The weighting can be easily determined.

[0013] In the determination step, the influence can be determined using a heatmap. This allows the use of analysis tools known in connection with artificial intelligence.

[0014] During the selection step, highly influential adjustable process parameters can be assigned to the subset. This allows relevant process parameters to be retained. Conversely, less influential adjustable process parameters can be excluded from the subset. This allows less relevant process parameters to be filtered out. A threshold value can be used to differentiate between high and low influence.

[0015] In the determination step, the optimized values ​​can be determined using a surrogate model, an evolutionary algorithm, or an aleatoric approach as an optimization rule. This allows for the use of various well-known methods.

[0016] A method for creating a connection between two components by means of a reaction of a reactive multilayer system using a fabrication device comprises the following steps: Setting a subset of process parameters adjustable on the manufacturing device to optimized values ​​representing values ​​determined using an embodiment of said process parameter optimization method; and Manufacturing the connection using the manufacturing device, wherein the subset of adjustable process parameters is set to the optimized values.

[0017] In this way, the previously optimized process parameters can be used advantageously to establish a connection. For example, the relevant process parameters can be set once and then maintained during the successive production of multiple connections.

[0018] The approach presented here further creates a device designed to perform, control, and implement the steps of a variant of the method presented here in appropriate facilities. This embodiment of the invention in the form of a device also allows the problem underlying the invention to be solved quickly and efficiently.

[0019] A device can be an electrical appliance that processes electrical signals, such as sensor signals, and outputs control signals accordingly. The device can have one or more suitable interfaces, which can be implemented in hardware and / or software. In the case of a hardware implementation, the interfaces can, for example, be part of an integrated circuit in which the device's functions are implemented. The interfaces can also be separate integrated circuits or consist at least partially of discrete components. In the case of a software implementation, the interfaces can be software modules that are present, for example, on a microcontroller alongside other software modules.

[0020] It is also advantageous to have a computer program product with program code that can be stored on a machine-readable medium such as semiconductor memory, hard disk memory or optical memory and is used to carry out the method according to one of the embodiments described above when the program is executed on a computer or device.

[0021] The invention is explained in more detail by way of example with reference to the accompanying drawings. These show: Fig. 1. A representation of an exemplary embodiment of a manufacturing device; Fig. 2 a schematic representation of the sequence of an exemplary embodiment of a reaction of a reactive multilayer system; Fig. 3a to 3e Representations of an exemplary embodiment of a reaction process of a reactive multilayer system; Fig. 4 sectional views of an exemplary embodiment of a low-quality connection; Fig. 5 sectional views of an exemplary embodiment of a high-quality connection; Fig. 6. A representation of an exemplary embodiment of a method for optimizing process parameters; Fig. 7 a flowchart of an exemplary implementation of a method for optimizing process parameters; and Fig. 8 a flowchart of an exemplary embodiment of a method for creating a connection between two components.

[0022] In the following description of preferred embodiments of the present invention, the same or similar reference numerals are used for the elements shown in the various figures and having a similar effect, without repeating these elements.

[0023] Fig. Figure 1 shows an embodiment of a manufacturing device 100 for joining a stack 102 of a first component 104, a second component 106 and a reactive multilayer system 108 arranged between the first component 104 and the second component 106 to form a component.

[0024] The device 100 comprises an activation device 110 configured to activate the reactive multilayer system 108 in order to establish a connection between the first component 104 and the second component 106. The activation device 110 is also referred to as an ignition device.

[0025] The activation device 110 is designed, for example, to apply a spark to the reactive multilayer system 108 or to direct a laser beam at the reactive multilayer system 108. This allows sufficient energy to be supplied to reach an ignition temperature required to activate the reactive multilayer system 108.

[0026] Optionally, the device 100 is designed to subject the stack of the first component 104, the second component 106 and the reactive multilayer system 108 to pressure 120 during the reaction of the reactive multilayer system 108 in order to increase the quality of the compound to be produced by the reactive multilayer system 108.

[0027] As an example, a pressure device 111 is used, comprising a first mass 112, also referred to as m1, and a movable second mass 114, also referred to as m2. The masses 112 and 114 are coupled to each other via a spring-damper mechanism with a damper 122 and a spring 124. Pressure is exerted on the stack via the movable mass 114, specifically on the side of the first component 104 facing away from the reactive multilayer system 108. According to one embodiment, the pressure device 111 is a spring plunger, whose first mass 112 can perform a movement y1 and whose second mass 114 can perform a movement y2.

[0028] The device 100 further comprises a sensor device 130 configured to acquire and provide sensor data 132. The sensor data 132, for example, depict process variables prevailing during the manufacture of the connection between components 104 and 106. The sensor device 130 is configured, for example, to acquire a temporal profile of various process variables, such as pressure 120, the temperature of the stack 102, or the occurrence of splashes 132. To acquire the sensor data 132, the sensor device 130 comprises one or more suitable sensors or measuring devices. The sensor data 132 may include, for example, several time series of sensor values ​​and / or process data, which, for example, can be used to determine the process parameters. Fig. 6, Fig. 7 to Fig. The procedure described in section 8 can be used.

[0029] According to one embodiment, a device 140 with an artificial intelligence-based quality-predicting model 142 is used to optimize process parameters of a process for creating a connection between the two components 104, 106 by means of a reaction of the reactive multilayer system 108. Optionally, the device 140 is part of the manufacturing device 100.

[0030] According to one embodiment, the device 140 is configured to read in adjustable process parameters 144 that influence or can influence the process for making the connection and to select from the adjustable process parameters 144 a subset 146 that are particularly relevant for producing the connection with high quality. Furthermore, the device 140 is configured to determine optimal values ​​148 for the adjustable process parameters included in the subset 146, with which the connection can be produced with high quality.

[0031] An example of an adjustable process parameter 144 is, for instance, the magnitude or time profile of a pressure 120. The pressure 120 can be adjusted, for example, by appropriately dimensioning the second mass, the damper 122, or the spring 124. Another example of an adjustable process parameter 144 is, for instance, the magnitude or time profile of a temperature at the stack 102. The temperature can be adjusted, for example, by appropriately shaping the reactive multilayer system 108 or by a method of ignition of the reactive multilayer system 108 using the activation device 110.

[0032] After determining the subset 146 of adjustable process parameters and their associated optimal values ​​148, this data can be used to adjust the manufacturing device 100 and, additionally or alternatively, the stack 102 so that the connection between the components 104 and 106 achieves a high quality. For example, a spring constant of the spring 124 can be set to a value defined by the optimal values ​​148 in order to set the pressure 120 to an optimal value. Similarly, a parameter of the activation device 110 can be set to a value defined by the optimal values ​​148, for example, to achieve an optimal activation temperature or reaction temperature of the reactive multilayer system 108.

[0033] Optionally, the optimal values ​​148 determined using the device 140 are used to automatically adjust the manufacturing device 100. For this purpose, the device 140 is connected, for example, via a suitable interface to devices such as the activation device 110 and the pressure device 111 of the manufacturing device 100.

[0034] According to one embodiment, the device 140 is configured to read in quality data 150 and use it to select the subset 146 and, additionally or alternatively, to determine the optimal values ​​148. For example, the quality data 150 are used to train the quality-predicting model 142. According to one embodiment, the quality data 150 represent pre-determined data obtained from an analysis of previously produced compounds.

[0035] According to one embodiment, a quality-predicting model 142 is used which, in its trained state, generates a prediction of the quality of the compound produced by the reaction of the reactive multilayer system 108 using current process data and provides this prediction, for example, in the form of a quality signal. For example, the quality signal indicates good or poor quality of a compound produced in a current process.

[0036] The described approach thus enables AI-based RMS bond quality optimization. Artificial intelligence (AI)-based methods are used for RMS bond quality optimization, thereby facilitating the optimization of RMS bond quality. Furthermore, the key influencing parameters on bond quality can be identified, which, as previously described, can be summarized in subset 146.

[0037] According to one embodiment, the first component 104 is an electronic component, for example a semiconductor component, and the second component 106 is a support for the first component 104. For example, the second component 106 is a printed circuit board or a substrate, for example a copper substrate. For example, the first component 104 is a chip, or, by way of example, a transistor or a diode. According to one embodiment, at least one electrical connection of the first component 104 is permanently electrically and mechanically connected to an electrical contact of the second component 106 after ignition of the reactive multilayer system 108.

[0038] According to an alternative embodiment, the first component 104 is a sensor and the second component 106 is a machine element, for example, a shaft for an electric drive or a gearbox. For example, the first component 104 is a force sensor which, after ignition of the reactive multilayer system 108, is rigidly connected to the second component 106 in order to detect a deformation of the second component 106.

[0039] According to one embodiment, the reactive multilayer system 108 comprises a plurality of alternating nanoscale layers and is designed to react exothermically upon ignition, thereby establishing a metallurgical bond between the components 104 and 106. For example, the reactive multilayer system 108 is configured as a reactive multilayer film placed between the components 104 and 106. Alternatively, the layers of the reactive multilayer system 108 are grown, for example, on a surface of one of the components 104 and 106.

[0040] The ignition temperature required for the activation of the reactive multilayer system 108 is achieved, for example, using the activation device 110 by means of a laser beam.

[0041] A reactive multilayer film (RMS), which according to one embodiment can be used as a reactive multilayer system 108, is a film that provides instant heat for a wide variety of applications in many industries. This reactive multilayer film is produced by vapor deposition of thousands of alternating nanoscale layers, for example, of aluminum (Al) and nickel (Ni). Activation is triggered by a small local energy pulse, for example, from an electrical, optical, or thermal source. The reactive multilayer system 108 reacts exothermically and releases precise local heat up to 1500 °C in fractions of a second, for example, within a few thousandths of a second.

[0042] To provide a small pulse of local energy to trigger the activation of the reactive multilayer system 108, the following can be used, for example: By applying an electrical voltage directly to the reactive multilayer system 108.

[0043] Using a laser that is directed at the reactive multilayer system 108.

[0044] By applying a direct heat source to the reactive multilayer system 108.

[0045] The concepts for monitoring the reaction can of course also be applied to an RMS system in which the reactive multilayers of the reactive multilayer system 108 are grown directly on one of the components, for example chip or substrate, and thus no separate RMS film is used as the reactive multilayer system 108.

[0046] The pressure 120 required for a high-quality RMS bonding process, for example to press the chip, the RMS foil and the substrate together, is typically applied using spring plungers. These spring plungers have a moving mass, here the second mass 114, of over 100 g and can consist of several spring-damper mass systems.

[0047] However, the concept described here can also be transferred without the pressure device 111 or with a pressure device 111 in which the second mass 114 is pressed against the stack 102 in a different way.

[0048] The rapid exothermic reaction, e.g. ~7 m / s during the activation / bonding process, leads to a rapid expansion of the reactive multilayer system 108.

[0049] This rapid exothermic reaction, or the associated RMS system changes, can be monitored by sensors. However, the relationship between the measured RMS system changes and other input variables, such as film thickness, chip size, etc., is not trivial. Using the approach described here, correlations with RMS bond quality can be established. In particular, RMS bond quality can be optimized even without prior knowledge of how the individual influencing factors are related and how they should be weighted. This can resolve several disadvantages and problems: If prior knowledge is lacking regarding how the influencing factors for RMS bond quality are related and how they should be weighted, optimization can be carried out according to the approach described here without costly series of experiments, in which the result may not be predictable.

[0050] The approach described here clarifies the key levers for RMS bond quality. This significantly reduces the effort required for optional further optimization attempts.

[0051] The RMS bond quality optimization achieved through test series for one application can potentially be transferred to other applications using the approach described here, since it is now possible to adjust the parameters in a targeted manner without understanding their interaction.

[0052] Furthermore, it will be possible to optimize the RMS bond quality in parallel with production, for example to react to unavoidable variations such as humidity, manufacturing tolerances, etc.

[0053] The approach presented here, in Fig. Figure 1, exemplified for the first component 104, for example in the form of a part, e.g. a bare die, which is connected to the second component 106, for example a substrate, with the reactive multilayer system 108, for example in the form of an RMS film, consists according to one embodiment in carrying out the fast exothermic reaction, as described below with reference to the Fig. 3a to e, or related RMS system changes, are to be monitored by sensors, for example using sensor device 130, and application data, for example in the form of process parameters 144 and quality data 150, are to be determined, and the RMS bond quality of the connection between components 104 and 106 is to be optimized with the help of AI (Artificial Intelligence), exemplified here by model 142. This results in the following advantages: The relationship between the influencing factors and their weighting can be represented using an AI algorithm, for example, implemented by model 142. Optimizing RMS bond quality therefore does not need to be carried out in costly series of experiments where the result may be unpredictable, but can instead be done with the help of an AI algorithm.

[0054] The key parameters for RMS bond quality can be identified using a heatmap. This significantly reduces the effort required for optimization attempts.

[0055] The AI ​​algorithm for RMS bond quality optimization can potentially be directly transferred to new RMS bonding applications, such as different chips or substrates, thus eliminating the need for time-consuming experimental series. Alternatively, the AI ​​algorithm can be extended with new training data from the new application.

[0056] The RMS bond quality can be optimized in parallel with production in order to react to unavoidable variations such as humidity, manufacturing tolerances, machine wear, e.g. of the manufacturing device 100 or any other bonding device used, etc.

[0057] Fig. Figure 2 shows a schematic representation of the process of an embodiment of a reaction of a reactive multilayer system 108, such as that used for manufacturing a component. The reactive multilayer system 108 is, for example, composed of a plurality of first layers 270 and a plurality of second layers 272, which are arranged alternately, with optionally mixed regions 274 located between adjacent layers 270, 272.

[0058] A schematic representation of an activation position, exemplified by a spark 276, is shown. Starting from the activation position, the material of the reactive multi-layer system 108 reacts to form reacted material of a reacted multi-layer system 280. A corresponding direction of propagation 278 is indicated by an arrow.

[0059] According to one embodiment, the reactive multilayer system 108 is a reactive multilayer film. This film provides instant heat for a wide variety of applications in many industries. This reactive multilayer film is produced by vapor deposition of thousands of alternating nanoscale layers, which are in Fig. The layers are schematically represented by 270 and 272 and consist, for example, of aluminum (Al) and nickel (Ni). Activation is triggered by a small local energy pulse from an electrical source in the form of a capacitor. The reactive multi-layer system 108 reacts exothermically to generate precise local heat up to 1500 °C in fractions of a second.

[0060] In this way, a connection that is both electrically and thermally conductive can be established, e.g., between a bare chip in electronics and a leadframe, a packaged chip and a printed circuit board, etc. - i.e., generally a connection between two parts, which are referred to here by way of example as components.

[0061] Fig. Figure 3a shows an embodiment of a stack 102 comprising a first component 104, a second component 106, and a reactive multilayer system 108 arranged between the first component 104 and the second component 106. The stack 102 is exemplary according to the design shown in Figure 3a. Fig. 1 described stack executed.

[0062] To activate the reactive multilayer system 108, an activation device is used, as exemplified by the following: Fig. Figure 1 describes the following. A first electrical contact tip 341 is shown, which is brought close to the reactive multilayer system 108, and a second electrical contact tip 342, which contacts the reactive multilayer system 108. An ignition voltage is applied between the contact tips 341 and 342.

[0063] Fig. 3b shows the using Fig. 3a shown stack 102 at a time when an ignition spark caused by the ignition voltage jumps between the electrical contact tip 341 and the reactive multilayer system 108.

[0064] Fig. 3c shows the example of Fig. 3b shown stack 102 at a time when a reaction of the reactive multilayer system 108 begins due to the energy introduced into the reactive multilayer system 108 by the ignition spark.

[0065] Fig. 3D shows the based on Fig. 3c shown stack 102 at a time when the reaction has spread to the entire reactive multilayer system 108.

[0066] Fig. 3e shows a component 380, which is made from the one based on the Fig. The stack shown in Figures 3a to 3d is formed by the reaction of the reactive multilayer system to a reacted multilayer system 280. The quality of the metallurgical bond produced by the reacted multilayer system 280 between the components 104 and 106 can be determined using the aforementioned monitoring device.

[0067] Fig. Figure 4 shows sectional views of an embodiment of a connection between two components realized by a reacted multilayer system 280. The connection was produced, for example, using a manufacturing device such as that shown in Fig. 1 is described. The compound is of low quality, as many large voids 490 are present.

[0068] Depicted are in Fig. 4 four sectional views, starting at the top left and proceeding clockwise, showing sections top left, top right, bottom right and bottom left.

[0069] Fig. Figure 5 shows sectional views of an embodiment of a connection between two components realized by a reacted multilayer system 280. The connection was produced, for example, using a manufacturing device such as that shown in Figure 5. Fig. As described in section 1. The connection is of good quality, as there are few very small voids (590).

[0070] Depicted are in Fig. 5 four sectional views, starting at the top left and proceeding clockwise, showing sections top left, top right, bottom right and bottom left.

[0071] Fig. Figure 6 shows an embodiment of a method for optimizing process parameters of a process for creating a connection between two components by means of a reaction of a reactive multilayer system, wherein it is based, for example, on Fig. 1 is described.

[0072] Adjustable process parameters 144 are used, such as those specified by Fig. 1 are described. Examples of adjustable process parameters 144 are, for example, pressure and temperature. A feature vector 660 is created from the adjustable process parameters 144 and used as input data for an artificial intelligence-based model 142 of the process for making the connection. The model 142 is trained to generate a quality signal 662 that predicts the quality of the manufactured connection. In a step 664, the influence 666 of the adjustable process parameters 144 encompassed by the feature vector on the model 142 is determined. For this purpose, for example, a weighting of the process parameters or of process parameter-based features within the model 142 is analyzed, or a heatmap relating to the model 142 is evaluated. Using the influence 666, a subset of adjustable process parameters 144 is selected.In step 668 of the determination, optimized values ​​148 are determined for the adjustable process parameters included in the subset, which can be used, for example, as optimized parameter values ​​for adjusting the based on . Fig. The manufacturing device described in section 1 can be used. For example, the optimized values ​​148 are determined using a surrogate model, an evolutionary algorithm, or other approaches.

[0073] The following section discusses in more detail examples of a procedure for AI-based RMS bond quality optimization.

[0074] According to one embodiment, training data is used to train the quality-predicting model 142. This training data is derived from sensor data, for example, obtained using the data based on Fig. 1 sensor device shown, plus application data, for example the process parameters 144, which are available both during the training of the AI ​​and in a later application for quality optimization in parallel with production, and quality data that are determined specifically for the training.

[0075] Typical sensor data include: position / motion measurements, acceleration measurements, force measurements, temperature measurements, ignition voltage and current measurements, vibration measurements, camera recordings, etc.

[0076] A combination of several measurement principles and / or measured variables from multiple concepts for RMS tape quality monitoring is possible to obtain additional insights into the quality of the RMS bond.

[0077] Typical application data include: component dimensions (such as chip size, RMS film thickness, substrate thickness), RMS film type, surface coating of the components, moving mass, spring constant of the spring plunger, etc.

[0078] Typical quality data include: shear forces / pressures from destructive shear tests, X-ray images including the percentage of voids, as in Fig. 4 shown, electrical function tests, splashing behavior, lifetime tests, deviation from the optimal position, etc.

[0079] The algorithm, here encompassed by model 142, is trained based on this data.

[0080] The application of the AI ​​algorithm is implemented according to an exemplary implementation as described below.

[0081] Model 142 can be trained using the aforementioned data and employed for AI-based RMS bond quality monitoring. Model 142 can thus be considered a quality-predicting model θ P to be trained. While this model 142 can be used to predict the quality of an RMS bond non-destructively based on measurable input data, the approach described here presents a method to optimize the adjustable parameters in the process of making a connection, here the adjustable process parameters 144, without additional complex test series and test designs, so that the best possible process quality of an RMS bond is achieved.

[0082] For this purpose, only all adjustable process parameters 144 of the feature vector 660, also referred to as feature vector, are considered and subsequently analyzed using, for example, a heatmap (on the input layer) or by a detailed analysis of the weights (W) of the model θ. P the influence of the individual parameters is determined.

[0083] Following this initial feature selection and "Feature Impact Analysis," the "optimal" value or range of values ​​for the identified adjustable parameters is determined using this information. This is represented here by the optimal values ​​(148) of the process parameters encompassed by the subset. The optimization task in this multi-dimensional space can be solved in step 668 of the determination process, for example, using a surrogate model, an evolutionary algorithm, or other approaches, such as aleatoric approaches. A neural network can be trained as the surrogate model. This network is trained with the previously selected input parameters (from the training data) to represent the RMS process, and the model parameters are then optimized accordingly. A surrogate model offers advantages over classical approaches, such as...evolutionary algorithms, resources and is able to find the optimum even in non-convex, non-linear, i.e., highly complex relationships.

[0084] This approach already allows us to use the model θ P The trained data optimizes the process parameters without conducting additional and costly test series and experimental designs. If the feature vector was derived using AI and the relationship to the real process variables cannot be directly established, the optimization model of the process parameters must use the abstracted input parameters of the model θ. P Additionally, the feature vector is mapped back to the real process parameters to determine their optimal value or range of values. Therefore, a different optimization model is used for an AI-based feature vector than for a classical feature space extraction.

[0085] Fig. Figure 7 shows a flowchart of an embodiment of a method for optimizing process parameters of a process for creating a connection between two components by means of a reaction of a reactive multilayer system, wherein it is based, for example, on Fig. The procedure can, for example, be executed once to generate optimized values ​​for process parameters, as required for the application based on Fig. The 8 described methods can be used to produce a compound.

[0086] As already at least partially based on Fig. As described in section 6, the method includes a step 770 in which adjustable process parameters are read in. These parameters influence the process of establishing the connection between the two components to a greater or lesser extent via the reaction of the reactive multilayer system. For example, a first adjustable process parameter is read in, representing the time course of a pressure acting on at least one of the components during the connection process, and a second adjustable process parameter is read in, representing the time course of a temperature prevailing at at least one of the components during the connection process. Further or different process parameters can be read in accordingly and subsequently used to create a feature vector.

[0087] In step 772, the feature vector is created from the adjustable process parameters. A suitable creation rule is used for this purpose. For example, features in the statistical, temporal, and / or spectral domains can be extracted from the adjustable process parameters and used to create the feature vector. Alternatively, the adjustable process parameters can be used directly as features of the feature vector.

[0088] In step 664, the influence of the adjustable process parameters encompassed by the feature vector on the model of the process for creating the connection is determined. For example, the influence is determined by weighting the adjustable process parameters in the model or by using a heat map.

[0089] Optionally, in step 774, the model is trained using the feature vector and quality data characterizing the quality of the connection. According to one embodiment, the model is configured to generate a prediction of the connection quality, and this prediction can be compared with the quality data to train the model.

[0090] In step 776, a subset of the adjustable process parameters read in step 770 is selected based on the influence determined in step 664. For example, in step 776, process parameters with a high influence within the model are assigned to the subset. Conversely, process parameters with a low influence within the model are not assigned to the subset.

[0091] In step 669 of the determination process, optimized values ​​are determined for the adjustable process parameters encompassed by the subset. A suitable optimization formula is used for this purpose. For example, such an optimization formula is based on a surrogate model, an evolutionary algorithm, or an aleatoric approach.

[0092] Fig. Figure 8 shows a flowchart of an embodiment of a method for producing a connection between two components by means of a reaction of a reactive multilayer system using a manufacturing device, as exemplified by… Fig. 1 is described.

[0093] In step 880, a subset of process parameters adjustable on the manufacturing device are set to optimized values, which were determined, for example, by carrying out a process parameter optimization procedure as described in the Fig. 6 and Fig. 7 is described.

[0094] Optionally, in step 882, the optimized values ​​are read in, for example via an interface to which the values ​​are based on... Fig. 1 described device which, according to an exemplary embodiment, was used to determine the optimized values.

[0095] In step 884, the connection is made using the manufacturing device. This joins, for example, two components to form a single part. The subset of adjustable process parameters is set to optimized values. For manufacturing, for example, in a provisioning step, a stack of two components and a reactive multilayer system positioned between the two components is provided, and in an activation step, the reactive multilayer system is activated to join the components together. Optionally, in a pressurization step, the stack is pressurized, at least during the reaction of the reactive multilayer system.

[0096] Step 884 can be repeated to create multiple connections. Reference sign 100 manufacturing device 102 stacks 104 first component 106 second component 108 reactive multilayer system 110 Activation device 111 Printing device 112 first mass 114 second mass 120 Print 122 dampers 124 springs 130 Sensor device 132 sensor data 134 splashes 140 Device 142 quality-predicting models 144 adjustable process parameters 146 subset 148 optimal values 150 quality data 270 first layer 272 second layer 274 mixed regions 276 Ignition 278 Direction of propagation 280 reacted multilayer system 341 first electrical contact tip 342 second electrical contact tip 380 component 490 large voids 590 small voids 660 state vector 662 Quality signal Step 664 of determining 666 Influence Step 668 of determining 770th step of the reading process Step 772 of the creation process Step 774 of training Step 776 of the selection process 880th step of the setting process Step 882 of the reading process Step 884 of the manufacturing process

Claims

[1] Method for producing a connection between two components (104, 106) by means of a reaction of a reactive multilayer system (108) using a fabrication device (100), the method comprising the following steps: Reading (770) adjustable process parameters (144) that influence the process for creating the connection between the two components (104, 106) by means of the reaction of the reactive multilayer system (108); Creating (772) a feature vector (660) from the adjustable process parameters (144) using a creation rule; Determining (664) an influence (666) of the adjustable process parameters (144) encompassed by the feature vector (660) on an artificial intelligence-based model (142) of the process for establishing the connection; Selecting (776) a subset (146) of the adjustable process parameters (144) depending on a magnitude of influence (666) of the adjustable process parameters (144); Determine (668) optimized values ​​(148) for the subset (146) of adjustable process parameters (144) using an optimization rule; Setting (880) the subset (146) of process parameters (144) adjustable on the manufacturing device (100) to the optimized values ​​(148); and Manufacturing (884) the connection using the manufacturing device (100), wherein the subset (146) of the adjustable process parameters (144) is set to the optimized values ​​(148). [2] Method according to claim 1, comprising a step (774) of training the model (142) using the feature vector (660) and using quality data (150) characterizing a quality of the connection, wherein the model (142) is configured to generate a prediction of a quality of the connection. [3] Method according to one of the preceding claims, wherein in step (770) of reading in the process a pressure (120) acting on at least one of the components (104, 106) when making the connection is read in as a first adjustable process parameter and a temperature prevailing on at least one of the components (104, 106) when making the connection is read in as a second adjustable process parameter. [4] Method according to one of the preceding claims, wherein in step (772) of creating the feature vector (660) features in the statistical and / or temporal and / or spectral domain are extracted from the adjustable process parameters (144). [5] Method according to one of the preceding claims, wherein in step (664) determining the influence (666) is determined by weighting the adjustable process parameters (144) in the model (142). [6] Method according to one of the preceding claims, wherein in step (664) determining the influence (666) is determined using a heatmap. [7] Method according to one of the preceding claims, wherein in the step (776) of selection, adjustable process parameters (144) having a high influence (666) are assigned to the subset (146) and adjustable process parameters (144) having a low influence (666) are not assigned to the subset (146). [8] Method according to one of the preceding claims, wherein in step (668) of determining the optimized values ​​(148) are determined using a surrogate model, an evolutionary algorithm or an aleatory approach as an optimization rule. [9] Device which is set up to perform and / or control the steps of the method according to any of the preceding claims in corresponding units. [10] Computer program configured to execute and / or control the steps of a method according to any one of claims 1 to 8.

Images