Die casting, also known as high pressure casting, is a near net shape technology that has been widely used in automotive, aerospace, and electronics industries in recent years. In die-casting process, molten metal (usually light alloy) fills cavity at high pressure and high speed under action of punch, and cools quickly to form final casting.
Die-casting is generally divided into cold-chamber die-casting and hot-chamber die-casting. Cold-chamber die-casting is mainly used in production of large parts, such as auto parts, communication base station cooling parts, etc.; hot-chamber die-casting is widely used in production of small electronics or 3C products Process, such as USB connector, notebook case, etc.

 

Figure 1. Typical cold chamber die casting process

Compared with ordinary casting process, die casting is characterized by high speed and high pressure. Products produced are generally light alloy thin-walled parts. However, die casting technology is also applied to production of pure copper rotors. Unlike aluminum and magnesium alloys, pure copper has a high melting point. Therefore, short die life in pure copper die casting process is a big problem.
Among all casting technologies, die casting has the highest degree of automation. Modern die-casting enterprises adopt automated die-casting technology, which highly integrates die-casting machines (generally dozens or even hundreds), forming a fully automated production process. At the same time, intelligent factory technology is adopted to monitor production process of die-casting machine, grasp performance and status of each die-casting machine in real time. Through big data measurement and real-time feedback, die casting machine production process can be adjusted in time to ensure final quality of product.

 

Figure 2. Typical die-cast automobile structural parts
Real-time control of die-casting mold temperature is a simple example:
Take cold chamber die-casting as an example. In production process, mold temperature is continuously increased due to continuous filling of high temperature liquid metal. At this time, in order to ensure that mold temperature is not overheated, cooling water is generally used to pass into mold for cooling. If cooling water pipeline design is reasonable, generally speaking, we can ensure that mold temperature reaches so-called thermal balance by controlling temperature and flow rate of cooling water. From this point of view, we can design a cooling water circuit feedback system. After knowing actual value of mold temperature, temperature and flow rate of cooling water can be controlled through calculation and instant feedback system, finally temperature of mold can be controlled. This is a typical application of die casting smart factory at this stage.
In fact, above-mentioned smart control case is only a small application scenario in "smart factory". Realizing a true "smart factory" requires collection of a large amount of real-time production data, of which data related to product quality play a decisive role, such as density, porosity and oxidized inclusions, etc., because these data are indicators that customers are most concerned about, are also core indicators to measure whether a casting is qualified. At this stage, these most critical indicators are the most difficult to obtain, because for metal alloy products, we cannot directly observe internal structure of product. Method adopted by most manufacturers is to spot check castings, cut in key areas clearly specified by customer, then directly observe whether there are problems; another method is to use two-dimensional X-ray inspection technology to scan and observe local positions of sampled castings. The biggest problem with this method is that three-dimensional casting information is compressed onto two-dimensional slices, information obtained by observation cannot fully reflect actual situation.

 

Figure 3. Implementation plan of European MUSIC project in smart factory of AUDI AG Ingolstadt factory (click to enlarge) [2]

With continuous development of automobile industry, requirements for quality of parts and components are getting higher and higher. Large automobile manufacturers continue to formulate requirements for internal quality of parts and components, quantitatively specify standards that can exist for internal defects in parts. In this case, component suppliers must be able to detect and calibrate distribution of defects in all castings in real time during production process, compare standards to assess whether castings meet requirements.
So how to correctly observe and record defects inside casting? The best technology in existing technology is computer tomography, which is what we commonly call CT technology. CT technology has been widely used in medical field, application of CT technology to industrial inspection is still a technology that has just emerged in recent years.
To apply CT technology to inspection of internal quality of castings, following requirements must be met:
First, inspection speed must be high enough to match real-time production process of castings;
Among them, first two requirements are for CT detection itself, and last one is for detection software or algorithm. Taking an overview of existing CT technology, we will find that the most promising inspection instrument is Speed-Scan CT equipment (Speed-Scan) produced by General Electric, and this equipment has been used by German Volkswagen Company for actual castings inspections. However, looking at domestic die-casting industry, use of real-time CT technology to control quality of castings is severely challenging from an operational perspective. The biggest constraint is cost-cost of CT inspection equipment is extremely high, and its use in production lines generally requires a large amount of CT inspection equipment, which is unaffordable for most domestic enterprises. With continuous development of industry and continuous improvement of casting quality requirements,use of CT technology to detect internal quality of castings in real time will be a general requirement for suppliers of OEMs in the future.

 

Figure 4. High-speed CT system deployed in casting production line

Under premise that casting is inspected by CT and three-dimensional solid data is obtained, we assume that there is an algorithm that can analyze data in a very efficient manner and give all information about internal defects of casting, including type, size and distribution, etc. Then we can use this information to adjust and modify production process itself, finally obtain qualified castings without excessive defects. This process-process of obtaining casting information and revising process, we call it process of process feedback and adjustment. Of course, we can't just complete this process based on the information of one casting. The most normal situation is to obtain information about a large number of castings, solve defects of castings through statistical analysis and process-related methods.
Next question is, even if we obtain a large amount of information on distribution of defects inside casting, how can we avoid unqualified defects by adjusting process parameters? The most powerful analysis tool available is computer numerical simulation, which is well known to us Computer Aided Engineering (CAE) technology.
Using computer simulation technology, we can realize virtual production in a local sense. Especially for die casting, we can directly simulate filling and solidification process by studying velocity, pressure, flow pattern, splashing and other behaviors of fluid filling cavity to determine whether there is entrainment in the filling process; by calculating temperature changes of casting and mold under different die-casting cycles, potential hot spots, casting defects (shrinkage cavity, porosity) and die-casting heat balance behavior of mold can be judged and studied. Through this numerical simulation technology, based on certain analysis conditions, we can to a large extent judge and avoid internal defects of casting, improve performance of casting and greatly increase production efficiency, achieve purpose of process feedback and correction that we discussed before.
 
Figure 5. Simulation of filling process of clutch housing
We string together whole process: use digital technology (CT) to detect three-dimensional defect data of product in real time. If product is unqualified, data will be transmitted to CAE analysis center, analog simulation technology will be used to analyze and produce defect problem solutions. Solutions are fed back to production and process front-end for execution and re-acquired products. Product continues to undergo digital inspection and obtain three-dimensional defect data. If product is qualified, iteration ends, otherwise it continues.

It can be seen that CAE analysis plays a key role in this process, effectiveness of proposed solution will have an impact on efficiency of entire process. In fact, whether you can master core of CAE technology and whether you can apply numerical simulation technology well in actual production can measure technical capabilities of a die-casting company to a large extent, because digital technology is the only way for an enterprise, the sooner you master core of digital technology on this road, the more you can stand out in future corporate competition.
Therefore, if digital inspection technology and CAE analysis technology are well applied to existing die-casting enterprises, we can see a complete and typical digital factory scene. Among them, digital inspection technology realizes digitization of physical entities, while CAE analysis technology transforms digital information obtained from inspection into problem solutions based on virtual production. In this process, digital inspection is actually a completely automated process, and CAE analysis still requires human participation. If CAE analysis can be solidified into an algorithm and fully automated, then this is embryonic form of future intelligent digital factory.