In the last decade, wire + arc additive manufacturing (WAAM), which is one of the most promising metal additive manufacturing technologies, has been attracting high interest from both academia and industry. WAAM systems are increasingly employed in the industry and academia, but there are still several challenges and barriers to process stability control. The process stability is highly dependent on how the molten feed wire is added into the melt pool, which is known as the droplet transfer mode. To ensure a stable WAAM deposition process, it is essential to maintain the transfer mode in a suitable stable status. Without an effective transfer mode control method, the operators need to determine and control the transfer mode based on their experience using manual adjustment, which is difficult to achieve in a long period of production process. In this paper, a deep learning-based technology was proposed for the control of the droplet transfer mode based on the data collected from the WAAM process. A long short term memory neural network was applied as the core transfer mode classification model. A time-series data, arc voltage, was collected and statistical and frequency features were extracted, which included 11 relevant features, as the inputs of the classification model. Then, the distance between the melted wire and the melt pool was adjusted based on the determined transfer mode to keep a suitable stability of the process. A case study was used to evaluate the proposed approach and to show its merit. The proposed approach was compared to three commonly used machine learning algorithms,k-nearest neighbours, support vector machine, and decision tree. The proposed method obtained the highest accuracy in determining the transfer mode, which was over 91%. The performance of the proposed approach was also evaluated by the single-pass and oscillated wall building. The proposed deep learning based approach improved the process stability in real-time, which resulted in better deposition qualities, in terms of geometry size and processing cleanliness compared to without control. Furthermore, this data-driven method could be applied to other WAAM processes and materials.

Aluminum alloy 7075 (with 7055 and 7150 filler wires) was welded using a digital welding machine that can switch arc mode between MIG, CMT and CMT+P modes. The transverse-motion weldability test of joints welded under different arc modes showed that the solidification cracking susceptibility was lower in CMT-technique-based welds than in MIG welds. The temperature cycle of the welding pool under different arc modes was recorded using mini-thermocouples, which showed that the cooling rate was lower in CMT welded samples than in MIG welded samples. The low cooling rate promoted the growth of α-Al dendrites through the back diffusion effect. Electron probe micro-analysis showed that micro-segregation of the α-Al dendrites was lower in the CMT welded samples than in the MIG welded samples. The T-(f_Al)^1/2 curve of each weld was calculated, which showed that CMT-based welding enhanced the bridging of adjacent α-Al dendrites, reducing the tendency for solidification cracking.

In this work, aluminum alloy 7075-T651 was welded by using customized Al-Cu-Si and Al-Cu-Mg-Zn filler wire during gas tungsten arc welding. The liquation cracking susceptibility of the joints was tested under a circular-patch welding experiment. Besides, the temperature vs. solid fraction curves (T-f_S) was calculated for different samples to reveal the formation mechanism of liquation cracking. The joint was susceptible to liquation cracking if (f_S)_weld > (f_S)_workpiece during the cooling stage. The results of the circular-patch welding experiment show that the liquation cracking susceptibility of the joint by using ER5356, Al-Cu_1.5-Si_4.5, Al-Cu_3.0-Si_2.5, Al-Cu_4.5-Si_1.5, Al-Cu_2.3-Mg_2.3-Zn_6.6 and Al-Cu_2.2-Mg_2.0-Zn_7.8 filler metal is 22.8%, 8.3%, 2.8%, 2.8%, 3.3% and 1.4%, respectively. The mechanical test shows that the data dispersion of the 7075 gas tungsten arc welding joint can be decreased by eliminating the liquation crack.