Mold filling is the first step in forming a casting. Bad filling process may cause turbulence, mold wall erosion, cold shut, inclusions, misruns, etc. Physical measurement and numerical simulation are two common methods to designing feeding systems. Numerical simulation has been widely used in research and production. However, the simulated results of castings have been doubted due to lack of validation. Physical methods are rooted in the mind of researches, so they have tried all kinds of methods to investigate the filling process. Among them, water analog was used for the investigation of gating systems, cast specimens and simple castings, but complicated casting is costly to model. A high temperature resistant transparent window can be used in front of the mold cavity to observe the melt metal flow. Zhao used a transparent quartz pyrex window to observe the filling process of aluminum. Khodai used this method for the filling of aluminum, cast iron and steel during lost foam casting. However, this method is most suitable for experimental study and plate shaped cast specimen. In recent years, the development of high intensive X-ray provided a new method to research the filling process of test castings. Kashiwai and Zhao used in-situ X-ray to examine the fluid flow of aluminum castings. Li studied the filling process of turbine blade by cast iron replacement of nickel alloy. X-ray radiography method is suitable for experimental study, but application in production is limited by casting size and thickness, complex equipment and expense. Jong proposed the contact time method, where a set of wires connected to a circuit detected the melt flow, and applied this method to aluminum castings successfully. Li applied the method to study the filing process of an iron cast plate in lost foam casting and calculated the filling speed by the filling time and distance. However, this method hasn’t been used in steel castings yet, and too many wires or cables onsite may cause trouble in production.
In this paper, a wireless filling process based on a contact time method is proposed, along with an observation system by a high temperature resistant, high speed camcorder. These two systems were used to investigate the filling process of two real steel castings.
Development of the Measurement Method
The mold filling time measurement is performed by a wireless contact time measurement system, as shown in Fig. 1(a). The sensor is buried in the sand with the end exposed at the inner surface of the sand mold during molding, and the emitter is placed outside the flask. A radio frequency signal is used for the emitter and receiver whose transmission distance is over 200m, enough for the onsite use. The receiver is connected to a laptop computer. Software acquires the filing time of each measurement point, stores the data, and plots it on the casting CAD model in time for monitoring. The sensor for mold filling time measurement is an open circuit with a pair of electrodes pointing into the inner surface of the mold cavity to sense the filling of the liquid metal.
As the melt flows through the pair of electrodes connected by the melt, the circuit is connected and electricity passes through the circuit. A signal emitter and a bulb are connected in the circuit for sending a signal and illustration, as shown in Fig. 1(b). The developed sensor, emitter and receiver are shown in Fig. 2. One sensor is one circuit. One circuit can link an emitter, or a dozen of circuits connect a public emitter. When electricity occurs, the emitter will send out a signal to a receiver which records the time.
Meanwhile, to better observe the filling process, another observation system is used with a high temperature resistant camcorder used. The camcorder is buried in the sand mold with its lens exposed in the inner surface of sand mold. Each camcorder is connected to the computer to monitor the filing process. The number of camcorders can be installed at the necessary places. Therefore, the filling process of each area can be recorded.
The developed system was used in the production of two steel castings, a hydro turbine blade and a hub casting produced at Harbin Electric Machinery Co., Ltd.
Case study one: Filling process of a hydro turbine blade
The hydro turbine blade is made of ZG0Cr13Ni4Mo, measures 1460 mm x 1210 mm x 850 mm and weighs 0.9 tons. The gating system uses a bottom filling style. Ten measurement points were selected, and one camcorder was placed. The onsite measurement is shown in Fig. 3. The measurement results are shown in Fig.4
As seen in Fig. 4, the melt flows out from the third ingate first, and then through the first, fourth and second ingates. The filling of the bottom is fast because of its thinner section. The top corners P1 and P2 are filled far later than the middle P3, meaning the top corners are hard to fill. Good air escape measures should be taken, such as ventilation holes. The observation of the camcorder is shown in Fig. 5. The melt comes out from the third ingate earlier than the second and the fourth ingates.
Case study two: Filling process of a hub casting
The hub casting is made of B50E54D3, measures 1,460 mm x 1,210mm x 850 mm and weighs 11.3 tons. The gating system uses a bottom filling style. Eighteen measurement points were selected at the ingates, the ribs and the top flank, and two camcorders were placed on the top of the two risers. The measurement results are shown in Fig. 6.
The four ingates fill the mold cavity almost at the same time, but ingates 1 and 3 along the flow direction fill faster than the two backward ingates 2 and 4. That indicates fluid flow is always toward the forward direction first. For heavy castings, this effect can be ignored, but it should be considered for small castings. By the filling times, the flow speed of the ingates can be calculated (it is 5 m/s), which will result in a flush of the flow and erosion of the sand roof it faces. The section view passing through a pair of ingates is shown in Fig. 7. Ther is still a certain height from the ingates to the roof sand, so the opening of the ingate is correct. The filling of the bottom flank and the ribs is uniform. The increasing speed of the melt at the bottom flank is 15mm/s, the ribs 30mm/s and the top flank 16mm/s.
The observed filling processing of the hub casting by the camcorders were merged, as shown in Fig. 8. Splashing during initial pouring resulted in oxides and inclusions. The melt comes out from ingates 1 and 3 larger. At the beginning, melt overflows from the ingates, covering the whole bottom area. Then it flows back to the center. This can be seen from the photos at 1.09s, 2s, 3s, 4s, 5s and 7s. Later, the covered area of the melt on the bottom increases, as shown by the photos at 8s, 12s and 25s. Basically, the filling process is uniform. The core releases smoke as the melt wraps the core, and it becomes heavier as the filling proceeds. Slag also is found floating on the melt front the melt level gets into the riser.
A wireless measurement system for the filling process of casting based on contact time technique and an observation system for the filling process of casting based on high speed camcorder working under high temperature were developed.
By using these two systems, the filling process of a turbine blade and a hub casting were measured and observed. The filling situation of these castings was obtained, and the filling time of a number of typical positions was recorded. The result showed that liquid steel filled ingates with those along the flow direction first. The velocity of the liquid steel in the mold was obtained by the calculation based on filling times. The study also shows these two systems operate conveniently and reliably. They are effective tools for monitoring the filling process of castings and optimizing gating system and have a broad prospect in casting production.