The truss structure has high specific strength and specific stiffness, and has great advantages in mechanical performance in low-density structures. Compared with traditional solid materials, the density of wire mesh materials is greatly reduced, and mesh structures with the same performance can reduce the weight by more than 70%. Compared with metal foam materials, the performance of metal mesh structures is controllable, the strength and modulus are an order of magnitude higher than those of metal foam materials, and the load-bearing efficiency is higher.
In order to solve the problem of missing pillars and pillar defects in 3D printed wire mesh structures, the team from the LLNL Lawrence Livermore National Laboratory in the United States investigated the ability to monitor construction quality during the construction process. 3D printing, which can instantly determine whether the manufacturing quality of the parts meets requirements.
3D printed lattice structure© LLNL
Address pillar gaps
It is understood that engineers and scientists at Lawrence Livermore National Laboratory (LLNL) have developed a real-time defect detection method for 3D printed metal parts. By combining monitoring, imaging technology and multiphysics simulation, it is possible to detect and detect defects. during the 3D printing process. Predicting pillar defects in 3D printed wire mesh structures.
The high strength and low density properties of wire mesh have found applications in many fields. During the laser powder bed fusion (LBPF) 3D printing process, missing pillars and defects may occur, affecting the mechanical properties of the printed array.
Lattice structure manufacturing technology
The complexity of parts required for a lightweight design filled with truss structures has exceeded the original functions of traditional CAD software. When reducing part weight using mesh technology, it is important from a DfAM (think additive) perspective to create strong connections between the mesh and the outer skin (to prevent delamination ).
Lattice Modeling Network Structure Modeling Software
To ensure the quality of wire mesh 3D printing, a post-construction inspection is often carried out, which is time-consuming and not always possible, especially for complex constructions. To address this issue, the LLNL team investigated the possibility of monitoring manufacturing quality in the field to make instant decisions about whether a part meets quality requirements.
As described in a recent article published in Additive Manufacturing Letters, the LLNL researchers used photodiodes, pyrometers (both of which measure the light emitted by the molten pool), and thermal imaging to monitor the structure of the microarray of the metal printing process. The team 3D printed normal pillars and intentionally defective “half-pillars” using the LBPF process and measured the heat output from the weld pool. The researchers then developed a method to use these thermal emissions to predict defects with high accuracy.
Currently, the LLNL National Laboratory is capable of detecting defects on multiple layers. In the future, new methods will be developed to identify defects in 3D printed layers, enabling dynamic reactions and potentially removing defects.
By validating the mechanisms behind the observed thermal radiation through high-speed thermal and optical imaging and simulations using the ALE3D multiphysics code, the researchers were able to predict the presence or absence of printed lattice pillars in 3D with over 94% accuracy. valuable information” and reflected the potential of monitoring melt pool heat emissions to identify defects in LPBF parts.
The “brains” of 3D printing process control
According to the article “Artificial Intelligence Reductions Defects – 3D Printing Process Control l Artificial Intelligence Empowers 3D Printing” in 3D Science Valley, artificial intelligence is playing an increasingly important role in each specific area, including: detection and remediation defects, in the construction process. stress and failure reduction after neutralization construction, in situ metrology and design accuracy, microstructural design, alloy design and optimization.
According to “Application of Machine Learning Methods Based on Neural Networks in 3D Printing” published by the Chinese Academy of Engineering, sensor hardware must be controlled by powerful operating software. The basic modes of control software include monitoring, recording, analysis and data storage. In a general case, such as during a laser powder bed fusion (LBPF) 3D printing process, once the hardware transmits the captured image of the weld pool to the software, it can calculate the temperature profile and extract thermal and dimensional measurements for further analysis. . Other interesting features can also be added to the detection software, such as equipping the software with algorithms (notably machine learning algorithms) to detect holes, lack of fusion, or voids, etc.
To predict and reduce 3D printing defects, artificial intelligence is becoming the “brains” of controlling the 3D printing process. In this regard, domestic research has also made some progress, for example, researchers from Shanghai Jiao Tong School of Materials Science and Engineering. University based on experimental observations The keyhole splitting phenomenon of the weld pool in LPBF establishes the thermo-mechanical-flow coupling. The model reveals a new keyhole pore (penetrating pore, abbreviated KP-pore) formation mechanism, explores the influence of powder on keyhole, molten pool and pore formation characteristics, and establishes data based on high-throughput simulation. in relation to the number of pores provides strategies to reduce or eliminate pores in the LPBF process.
Source: 3D Science Valley
Daguang focuses on providing solutions such as precision CNC machining services (3-axis, 4-axis, 5-axis machining), CNC milling, 3D printing and rapid prototyping services.
