F. Memola Capece Minutolo

A model for evaluation of laser welding efficiency and quality using an artificial neural network and fuzzy logic

Abstract For any welding process, efficiency and quality strongly depend on the energy input, which is the energy introduced per unit length of weld from a travelling heat source. The focused laser beam is one of the highest power density sources available to the welding industry today, which makes it possible to weld with very low energy input with respect to most of other welding processes. In this paper a number of stainless steel butt joints were produced by laser irradiation. The welding efficiencies were calculated as the melted volume-energy input ratio. Moreover, the weld crown and depth were measured in order to evaluate the joint quality. The collected data were interpolated and correlated to the process parameters using an artificial neural network. They were then clustered using a fuzzy C-means algorithm. During the training stage of the neural network algorithm, the design of experiment (DOE) technique was used for the selection of the optimized network parameters. In practice, using some artificial intelligence, a model was built to choose the most suitable laser welding process for producing high efficiency and good quality, and is now available for supporting design and research.

FEM simulation of metal sheets laser welding with wire filler material

Worldwide manufacturers have already recognized the advantages of using laser welding for a number of applications. Although laser welding is usually performed autogenously, the use of wire filler allows a broader range of welding applications. In this paper a finite element model for key-hole laser welding sources based on experimental observations is presented. The shape for the heat source can be easily changed so to simulate both autogeneous and wire filler laser welding. In the thermal and mechanical calculations the material properties were considered with full temperature dependence. Austenitic stainless and aluminum alloy welds were chosen to validate the model. Simulations of out-of-plane welds were computed using a local coordinates system. A personal computer was used for the calculations. For different plate thicknesses and weld geometries, the calculated fused zones were found in very good agreement with the experimental ones, which proved that the thermal history was computed correctly. When materials with thermal induced microstructure changes are welded, the microstructure in every point of the heat affected zones can be determined as a function of the local temperature cycle.

Laser ablation of maskant used in chemical milling process for aerospace applications

Chemical etching is a non-traditional machining process where a chemical solution is used to remove unwanted material by dissolution. To shape the etched area, before the process, a chemical inert paint (maskant) is applied on the surface. Then the maskant is trimmed away and the uncovered area is subject to the etching. The maskant cut could be obtained mechanically or by laser ablation. In this work, the effect of process parameters, cutting speed and beam power, on interaction phenomena and defect formation in laser cutting of polymeric maskant is studied, using a 30W CO2 laser source.

The use of laser technology for the welding of the automotive tree mechanical clutch subject to the action of residual intrinsic magnetic field

The material in issue is a case-hardening steel, type 20MnCr5 (UNI 8550), that has been previously undergone to a softening annealing process conferring therefore to the material one equal Hardness Brinnel 200 to 220 HB. This steel is very difficult to weld as by laser technology as by other welding technologies by fusion. It has been experienced a power at first laser of 1500 W (for requirements begins them of understanding of the process and the relative to you parameters to set up then) and of 3000 W, like previewed. They have been used a covering gas and a laser beam focusing mirror. All the welded tests have been at first subordinates to taken care of a visual examination to eye and then to an Image Acquisition System, computerized and connected to both metallographic and stereo microscope. Then cut to you with metallographic apparatus, workings to the abrasive papers and metallographic cloths, in order to end with a chemical etching of type NITAL. The best ones turn out to you have been then always visualize in shape of macro and micro-graphs, acquired with the same Image Acquisition System of type NIKON - LUCIA 4.82 vers. by LIM, to storage and automatically measure the cross-section area (melted zone) and then to calculate the efficiency level, on each joint and bead on plate, expressed in Dau unit, to verify the laser welding efficiency, correlated to the laser working parameters. To a better characterization of the produced joints many micro-hardness tests and relate family hardness trends and profiles have been carried out. At the end from the comparison of the values of ETE % and MR % (Model of Swift- Hook & Gick) and of the values of WE in Dau (Model DA.LU.), lead on some beads of the 20MnCr5, has turned out an ulterior and substantial validity and goodness (as well as easiness of employment and calculation of WE) of this last Model and Unit of Measure. In fact the values of ETE % and WE (in Dau) are appeared substantially similar, respecting both all the substantial correlations between the job parameters, to laser.

Laser cuttig model based on artificial intelligence

Since the eighties laser incorporated into production systems has made a vast impact on production methods. In the cutting process laser has been replacing both blanking and plasma torch since it is able to cut burr free, with high production rate, and economically. Nevertheless, laser processing is still limited to repetitive high-rate operations due to the trial-and-error calibrations involved in obtaining an acceptable operating condition when changes in work piece geometry, surface quality, and dimensional accuracy are imposed.Moreover, the quality inspection of the laser cuts is performed by offline inspections of the edges of the metal by practiced operators.In this paper a model based on fuzzy logic is proposed to help the planner of a CO2 assisted laser cutting process. The quality of cut is evaluated on the basis of 5 criteria that took into account the appearance of cut, frequency of striation, width of the heat affected zone, roughness of cut, width of the cutting path (kerf). The planner can evaluate the attained cut quality through the here-proposed model. The knowledge for building the fuzzy model came from experimental trials. The same tool can be used for quality inspection performed through an automated system that merges the expertise of cutting operators with mathematical model’s exactness. A practiced operator is no longer necessary. This model can be extended to other laser cutting processes.Since the eighties laser incorporated into production systems has made a vast impact on production methods. In the cutting process laser has been replacing both blanking and plasma torch since it is able to cut burr free, with high production rate, and economically. Nevertheless, laser processing is still limited to repetitive high-rate operations due to the trial-and-error calibrations involved in obtaining an acceptable operating condition when changes in work piece geometry, surface quality, and dimensional accuracy are imposed.Moreover, the quality inspection of the laser cuts is performed by offline inspections of the edges of the metal by practiced operators.In this paper a model based on fuzzy logic is proposed to help the planner of a CO2 assisted laser cutting process. The quality of cut is evaluated on the basis of 5 criteria that took into account the appearance of cut, frequency of striation, width of the heat affected zone, roughness of cut, width of the cutting path (kerf). The planner can e...

The Use of a Neural Network in Predicting the Mechanical Properties of Low Carbon Steels after Hot Rolling

A neural network using a learning procedure based on error backpropagation algorithm was employed to identify groups of hot-rolled low carbon steels with homogeneous mechanical characteristics. The mechanical properties of hot-rolled steel bars depend on two categories of variables: technological and structural. Very often, these interact with each other, and the resulting synergic effects are not always easy to evaluate. Moreover, as it is impossible to plan test sequences in which a single parameter at a time is made to vary within a sufficiently wide range, there is no way of designing a theoretical experimental model based on partial multiple regression.

Neural networks optimization of laser welding process

The employment of laser systems for completely automated welding processes needs an analysis of physical mechanisms by means of mathematical models to correlate the characteristic quality indexes of welding to working parameters. In this paper we describe the application of neural networks method in order to correlate, for different kinds of stainless steel, the melting area and the welding efficiency to working parameters, leaving out of account the knowledge of the chemical and thermo-physical properties of working materials. The results are quite satisfactory even if a direct search of the optimal process conditions is not possible.

Employment of Fuzzy Logic Modelling in Quenching of Steel Rods for Reinforced Concrete After Hot Rolling

The traditional production of steel rods for reinforced concrete, a long time considered of short technological interest, has been recently modified with the development of new procedures, which enhance mechanical characteristics without adding micro-alloys. In particular, the process outlined in this paper, said tempcore process, enables to graduate mechanical and technological properties of the steel rods, so that these characteristics can vary in assigned limits. The underlying principle of this innovative process lies in surface quenching of the rods on leaving the last stand in the hot rolling mill, followed by self-tempering. Process adjustment still relies on the experience of the production manager, taking as a parameter for assessing process efficiency the temperature reached by the rod on the cooling plate. In this paper, the influence is examined of the process variables (final rod diameter, chemical composition, water flow rate and velocity), on the ultimate rod structure, upon which all mechanical properties, in general, depend. Based on the finding of a previous study, the possibility is examined of applying fuzzy logic techniques to the heat treatment process under examination with a view to rendering it more rational and independent of operator unreliability. The study will have a dual future aim. To evaluate the possibility of predicting the properties of the reinforced concrete rods when chemical composition of steel and final diameter are known, and secondly, to graduate the intensity of the treatment to obtain rods with predetermined properties from steel with known chemical composition.