Additive manufacturing deposition layer repair method by changing real-

Abstract

The invention relates to an additive manufacturing deposition layer repairing method based on change of real-time cladding amount, which mainly solves the problems that the surface of a deposition layer is uneven, the requirement on metal transition stability during fuse deposition is high, and the shape of the generated deposition layer is irregular when a multi-layer wall body is manufactured through fuse deposition in the prior art. According to the technical scheme, the method is characterized in that self-starting recognition of a welding machine, deposition layer position pre-calibration and real-time position obtaining in the stacked wall layer additive manufacturing process are completed through visual positioning, meanwhile, a binocular vision sensing system is adopted, data collection is conducted on the side edge contour of a deposition layer, and the data are fed back to an image processing system. The area needing to be filled of the deposition layer is calculated according to the originally set deposition layer parameters and the received position parameters, then the wire cladding amount needed for filling the deposition layer is calculated, the technological parameters are inversely solved according to the wire cladding amount, and then the technological parameters are fed back to the main control system. And the master control system changes the cladding amount by changing the wire feeding speed and the moving speed of the welding gun, so that the effect of filling a deposition layer is achieved.

Description

Technical Field

[0001] The present invention specifically relates to a repair method for an additively manufactured deposited layer based on changing a real-time cladding amount, and the method belongs to the field of additive manufacturing. Background Art

[0002] In the arc additive manufacturing process, due to the unstable metal transition and heat accumulation effect, the surface of the deposited layer will be uneven, which will reduce the precision of the product. In order to ensure the quality of the deposited layer, the defects need to be repaired. When repairing the deposited layer, it is necessary to understand the unevenness of the surface of the deposited layer in real time, so it is difficult to control the filling quality.

[0003] At present, the welding repair technologies commonly used in the repair of metal parts mainly include automatic submerged arc cladding, gas shielded cladding, plasma arc cladding and laser cladding technology. Arc welding is simple and flexible, with simple equipment, suitable for welding repairs in various occasions, but the production efficiency is not high. The submerged arc automatic welding process is stable, does not require high operating technology, and has high productivity. It is only suitable for the repair of large quantities, long deposits, and thick parts. Gas shielded welding does not require the use of welding rods or flux, and does not require slag cleaning on the welding repair surface after welding repair. It has high productivity and low repair cost, but its application occasions are greatly limited. Spraying has a wide range of use on part materials, but it is difficult to operate. Due to the high thermal conductivity of the material, the powder droplets hit the surface of the part, the contact temperature drops rapidly, and a good fusion cannot be formed. Basically, most stacking repair methods have certain inaccuracies.

[0004] Whether it is welding repair or arc additive manufacturing, a technology that can repair deposited layer defects in real time is needed. The present invention proposes a repair method for additively manufactured deposited layers based on changing the real-time cladding amount to improve the efficiency and quality of repairing deposited layers. Summary of the invention

[0005] The purpose of the present invention is to provide a method for filling defects when defects occur in the welding deposition process, and to design a method for filling the defects of the deposition layer by calculating the amount of cladding to be repaired and inverting the required repair process parameters. The binocular vision sensor system can more accurately find the location and trajectory of the defects, thereby effectively solving the problem of over-welding and under-welding caused by low precision.

[0006] The purpose of the present invention is achieved through the following technical solutions: Figure 1As shown in the figure, the device includes a binocular vision sensing system, an image processing system, an adjustable wire feeder, a welding gun, and a main control system. Its working method is: during the deposition process, the binocular vision sensing system composed of double-sided industrial high-speed cameras a and b performs real-time monitoring and taking pictures, and transmits them to the image processing system. The image processing system extracts parameters, processes data, and compares the original data, and then feeds back to the main control system. During the deposition repair process, the main control system further calls the data fed back by the image processing system and sends target instructions to the machine. The machine changes the cladding amount by changing the wire feeding speed and the moving speed of the welding gun, thereby achieving the effect of filling the deposition layer.

[0007] The overall process of the method is divided into a deposition process and a deposition repair process. During the deposition process, the binocular vision sensor system performs real-time observation when each layer is deposited, and the image processing system analyzes and calculates the data of each deposition layer corresponding to each time period t, and stores it. After each layer of deposition process is completed, the deposition repair process is entered. During the deposition repair process, the main control system will call the processed data of each deposition layer corresponding to the deposition process, and then control the automatic welding machine to repair the deposition layer. At the same time, the main control system will also control the industrial high-speed cameras a and b to continue observing, and the observation results will be judged by the image processing system whether a secondary repair is needed for this layer. In addition, when the deposition repair process is completed and the next layer of deposition process is about to begin, the image processing system will perform mean filtering on each set of data collected previously, and store the optimal data after obtaining it for engineers to refer to for improvement, and release redundant data to save memory space.

[0008] The repair method of the deposited layer adopts the control variable method, and there are two methods: first, the wire feeding speed v s Fixed, through the cladding filling function V T (x) Calculate the position x to be repaired i Required welding gun moving speed v h Second, the moving speed of the welding gun v h Fixed, through the cladding filling function V T (x) Calculate the position x to be repaired i Required wire feed speed v s .

[0009] The binocular vision sensing system is composed of industrial high-speed cameras a and b, which are carried out simultaneously with the deposition process. During the deposition process, the industrial high-speed cameras a and b respectively take real-time cross-sectional images of the deposition layer on the left and right sides of the deposition layer and feed them back to the image processing system. The image processing system first determines whether repair is needed by fitting and comparing the left and right deposition layer cross-sections fed back by the binocular sensing system. If necessary, data processing and analysis are performed on it, otherwise the image data is released from the memory.

[0010] The image processing system is to further analyze and process the real-time data fed back by the binocular vision sensor system after collating it, and then transmit the final processed data to the main control system. During the operation of the image processing system, the picture data fed back by the binocular vision sensor system in real time is first compared with the original expected ideal picture data. If the comparison error is large, the image processing system will further process the real-time data fed back by the binocular vision sensor system. If the comparison error is very small and can be ignored, the data comparison for the next time period will be performed. Secondly, the image processing system extracts the characteristic values ​​of the sedimentary layer contour image fed back by the binocular vision sensor system, and obtains the side contour function F(x) through fitting and comparison. Then, based on the side contour function F(x), the layer height h and the layer width d, the original area S, the existing area S1 and the area S to be filled are calculated. i , and further find out the position x of the sedimentary layer to be repaired i The required cladding volume function V R (x i ), and the more appropriate wire feeding speed v obtained after multiple tests s and the moving speed v of the welding gun h Function V of the required cladding volume R (x i ) is combined to obtain the cladding filling function V T (x). Finally, the image processing system feeds the processed data back to the main control system for further processing. Each time the data is processed, the image processing system will optimize and adjust the side contour function F(x) synchronously, and continuously optimize and adjust until the optimal state is reached. When the next layer is deposited at the position to be repaired, according to the cladding filling function V T (x) Adjust wire feeding speed v s and the moving speed v of the welding gun h , to control the amount of cladding at the position to be repaired, thereby filling the surface of the deposited layer and making the surface of the previous deposited layer smooth again.

[0011] The main control system mainly writes instructions and controls the welding machine by receiving the data collated and analyzed by the image processing system. When receiving the data transmitted by the image processing system, the fuzzy PID control method is used to control the automatic welding machine. The fuzzy PID control method is an intelligent control algorithm that imitates the fuzzy reasoning and decision-making process of people in terms of behavior. At the same time, the main control system will also control the working status of industrial high-speed cameras a and b at any time. During the deposition and repair process, the main control system controls it to continue to observe, that is, observe while repairing. The observation results are judged by the image processing system. If there are no non-negligible defects, the next layer of deposition is carried out. If there are non-negligible defects after the deposition layer is repaired once, a feedback process is carried out. Such a monitoring method not only greatly reduces the error of the deposition layer repair, but also takes less time and improves the repair efficiency.

[0012] The feedback process is a secondary repair after the deposition layer is repaired once to ensure the accuracy of the process. In this process, the main control system first sends a specific signal to the image processing system. The image processing system transmits the position to be repaired for the second time and the required cladding amount data to the main control system, which controls the repair.

[0013] The automatic welding machine includes a welding gun and an adjustable wire feeder. When the automatic welding machine receives a time instruction, it controls the length of time it stays in each section of the deposition layer according to the time instruction without changing other conditions; when receiving an instruction to change the wire feeding method, the welding machine changes the wire feeding method, angle, etc. to change the required cladding amount while other parameters remain unchanged, so as to perform repair.

[0014] The main features of the present invention are: firstly, a binocular vision sensor system is used to collect cross-sectional data on both sides of the sediment layer in real time, and at the same time, the data is fitted and compared to obtain a sediment layer contour image and a travel trajectory for each time period is predetermined. Then, the fitted image data is transmitted to an image processing system, which further analyzes and processes the image: extracts the characteristic value of the side contour in the image, calculates and fits the characteristic value to obtain a side contour function F(x), and then calculates a certain position x to be repaired in the sediment layer according to the side contour function F(x), the layer height h and the layer width d. i The required cladding volume V R (x i ). When the automatic welding machine deposits the next layer to the position to be repaired, according to the cladding filling function V T (x) Adjust wire feeding speed v s and the moving speed v of the welding gun h , to control the amount of cladding at the position to be repaired, so as to fill the surface of the deposited layer and make the surface of the previous deposited layer flat again. Using the idea of ​​controlling the entire process from part to whole, not only can the repair be carried out in a sequential cycle, but the entire system also uses feedback to adjust and monitor the repair process in real time to reduce errors.

[0015] The beneficial effect of the present invention is to provide a method for repairing an additively manufactured deposited layer based on changing the real-time cladding amount, to repair the unevenness and inevitable defects on the surface of the deposited layer during fuse deposition, thereby reducing the risk of collapse; at the same time, when surface unevenness problems occur, timely feedback can be given and repairs can be made, thereby improving the quality of the workpiece; the method provided by the present invention can fill and repair at the same time, reduce the defective rate, reduce material waste, and effectively improve production efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Figure 1 The overall system flow chart

[0017] Figure 2 Schematic diagram of the cross section of the sedimentary layer

[0018] Figure 3 Schematic diagram of how the deposition process works Figure 4 How the Deposition Repair Process Works

[0019] Figure 5 Schematic diagram of the metal parts repair process DETAILED DESCRIPTION

[0020] The present invention is further described in detail below with reference to the accompanying drawings and embodiments.

[0021] See also Figure 1 This is the overall system flow chart. The main devices of the repair process include industrial high-speed cameras a and b, image processing system, main control system and automatic welding machine, etc. Industrial high-speed cameras a and b are placed on the left and right sides of the sedimentary layer to observe the deposition process in real time, record the cross-sectional data of the sedimentary layer - the sedimentary layer cross-sectional image a recorded by camera a on the left and the sedimentary layer cross-sectional image b recorded by camera b on the right, and then compare and fit the image data of each time period and send it to the image processing system. After the image processing system organizes, optimizes, and fits the image data to obtain the side contour curve and calculates the side contour function F(x) and trajectory curve of the sedimentary layer. At the same time, the image processing system calculates a certain position x to be repaired by using the side contour function and the original layer height h, layer width d and other data. i The required cladding volume V R (x i ), and then according to the wire feeding speed v s , welding gun moving speed v h , side profile function F(x) and required cladding amount V R (x i ) Get the cladding filling function V T (x). The image processing system sets the wire feeding speed v based on experience. s , welding gun moving speed v h The initial value of the initial cladding filling function V T (x) Then send it to the main control system. The main control system writes the command to control the automatic welding machine to repair the deposited layer, and at the same time changes the wire feeding speed v s , welding gun moving speed v h Change the cladding filling function V T (x), and then change the cladding amount, so as to achieve the effect of repairing the deposited layer. In order to reduce the error during the repair process of the deposited layer, the binocular vision sensing system will also observe in real time and repeat the above steps.

[0022] See also Figure 2This is a side cross-sectional view of the sedimentary layer. The height h is the original sedimentary layer parameter. There are three sedimentary layers that have been repaired and one sedimentary layer trajectory curve after repair. The area below the trajectory curve is the existing area S1, and the light gray part above is the area that needs to be repaired S i . Based on the layer height h and the area S to be repaired i To calculate the volume of the required cladding, and use it as a basis for repair.

[0023] Embodiment 1: Repair of additively manufactured deposited layer.

[0024] See also Figure 3 and Figure 4 This is a working principle diagram of the deposition process and the deposition layer repair process, which consists of industrial high-speed cameras a, b, an adjustable wire feeder, and an automatic welding machine. During the deposition process, industrial high-speed cameras a and b respectively observe the left and right sides of the deposition layer in real time, and send the observed data to the image processing and analysis system after image comparison and fitting. The image processing and analysis system analyzes and calculates the data it transmits to obtain the required cladding amount V R (x i ) and cladding filling function V T (x), and then enter the deposition repair process. At this time, the main control system receives the data fed back by the image processing and analysis system, sends control instructions to the automatic welding machine, and changes the cladding amount by changing the wire feeding speed and the welding gun moving speed, so as to achieve the effect of filling the deposition layer. At the same time, industrial high-speed cameras a and b will continue to perform real-time observation to reduce the error of deposition layer repair.

[0025] Embodiment 2: Metal parts repair.

[0026] See also Figure 5 This is the process of repairing metal parts. During this process, the automatic welding machine starts to repair from the left side. After the first layer on the left side is repaired, it then repairs from the right side. After the right side is repaired, it continues to repair from the left side. This process is repeated until the repair is completed. During the repair process, the binocular vision sensor system will observe the entire process of part repair in real time. When it is observed that the surface of the repaired part does not meet the expectations, the image parameters and the image trajectory curve after fitting comparison will be transmitted to the image processing system, which will process the data and then feed it back to the main control system. The main control system will perform a second repair on the part until the surface is flat and achieves the expected effect. After the part is repaired, the excess part is cut until the expected effect is achieved. This method of repairing metal parts mainly changes the wire feeding speed v s and the moving speed v of the welding gun h, to change the cladding amount, so as to carry out repairs. At the same time, such a repair method is not only easy to implement, but also saves the cost of rebuilding a metal part. For different metal parts, when they are damaged and need to be repaired, this method is also applicable.

Claims

1. A method for repairing deposited layers of additive manufacturing based on changing the real-time cladding amount, characterized by: In order to repair the uneven surface defects of the sedimentary layer that occur during the deposition process, a binocular visual sensing system is used to observe the deposition process in real time, and the sedimentary layer contour image is obtained by comparison and fitting. The collected image is input into the image processing system, and the characteristic value of the side contour in the image is extracted. The side contour function F(x) is obtained by calculation and fitting of the characteristic value. Then, the position x to be repaired in the sedimentary layer is calculated according to the side contour function F(x), layer height h and layer width d. i The required cladding volume V R (x i ), when the next layer is deposited to the position to be repaired, according to the cladding filling function V T (x) Adjust wire feeding speed v s and the moving speed v of the welding gun h , to control the amount of cladding at the position to be repaired, thereby filling the surface of the deposited layer and making the surface of the previous deposited layer smooth again.

2. The additive manufacturing deposit layer repair method based on changing the real-time cladding amount according to claim 1 is characterized by: Cladding Filling function V T (x) is based on the wire feeding speed v s , welding gun moving speed v h , side profile function F(x) and required cladding amount. Before repairing, fill function V T (x) Inversely solve the wire feeding speed v at the position to be repaired s and the moving speed v of the welding gun h The two process parameter information are transmitted to the main control system, which controls the position to be repaired x i Wire feeding speed v s and the moving speed v of the welding gun h To adjust the cladding amount V R (x i ).

3. The additive manufacturing deposited layer repair method based on changing the real-time cladding amount according to claim 1 is characterized in that: The structure in the repair process includes a binocular vision sensing system, an image processing system, an adjustable wire feeder, a welding gun, and a main control system. The binocular vision sensing system includes industrial high-speed cameras a and b, which respectively collect the profile images of the side of the deposition layer on the left and right sides of the deposition layer in real time, that is, camera a takes a photo on the left side of the deposition layer to obtain the profile image a of the deposition layer section, and camera b takes a photo on the right side of the deposition layer to obtain the profile image b of the deposition layer section, and inputs them into the image processing system. The main control system calculates the profile image according to the cladding filling function V T (x) to solve the process parameter value, and then control the position to be repaired x i Wire feeding speed v s and the moving speed v of the welding gun h size.

4. The additive manufacturing deposited layer repair method based on changing the real-time cladding amount according to claim 1 is characterized by: The image processing system processes the position parameters x of each segment of the trajectory image fed back by the binocular vision sensor system in real time. i , extract the image feature values, and perform fitting and comparison processing to obtain a high-precision side profile function F(x). Then, the position x of the sediment layer to be repaired is calculated based on the layer height h and the side profile function F(x). i The area S that needs to be filled i , then according to S i and layer width d to calculate x i Volume function V of the required cladding amount R (x i ).

Images