The relationships among softening point, minimum viscosity and strength were analyzed, indicating a research direction for synthesizing phenolic resins with high softening points and high strength. The mechanism of shell peeling in resin-coated sand and the principle of evaluating resin resistance to shell peeling from viscosity change curves were also analyzed. Keywords: phenolic resin; viscosity; resin-coated sand; coating temperature; properties. Under hot-coating conditions, the physical state of phenolic resin changes from a flake-like glassy state to a thin-film viscous-flow state. After curing agent is added in this state, cooling returns it to a thin-film glassy state [1]. Changes in the resin's physical state with temperature are called thermal behavior. Resin thermal behavior is unique and therefore testable. The thermal behavior of phenolic resin has an important influence on resin-coated sand properties such as strength, resistance to shell peeling and melting point. This paper studies the effects of phenolic resin viscosity-temperature characteristics on the coating process and resin-coated sand properties. 1 Experiment. 1.1 Main experimental raw materials: The principal synthesis materials were phenol, 37% aqueous formaldehyde and oxalic acid. The resin-coated sand materials were Weichang scrubbed sand, 50/100 mesh, 33% aqueous hexamethylenetetramine, calcium stearate and KH550. 1.2 Experimental methods. 1.2.1 Resin synthesis: The synthesis apparatus comprised an automatically temperature-controlled magnetic stirring heater, a three-neck flask, a water bath with a reflux assembly, and a vacuum dehydration unit. The reaction process is shown in Figure 1. Four resins, R1, R2, R3 and R4, were prepared in the laboratory, with softening points of 71.5, 80.5, 90.5 and 97°C respectively, and their viscosity-temperature curves were measured.

Effects of Phenolic Resin Viscosity on Coated-Sand Performance

　　Three resins with softening points of 71.5, 80.5 and 90.5°C were selected. Their viscosity-versus-time curves were measured separately with 10% HMTA added at 150°C. 1.2.2 Method for testing resin viscosity-temperature characteristics: The resin was melted at a given temperature, and its viscosity at that temperature was measured using a BROOKFIELD LVDV-II viscometer. The test temperature was varied to obtain a viscosity-temperature curve. At a fixed temperature of 150°C, the viscosity change curve of resin with curing agent added was measured. 1.2.3 Resin-coated sand mixing process: With the base sand at 145°C, 2% resin was added and mixed for 60 s. When the temperature dropped to 110°C, a hexamine solution was added at 12% of the resin mass. After forced-air cooling for 60–70 s, calcium stearate was added at 5% of the resin mass. After another 10 s, the sand temperature dropped to 70–80°C and the sand was discharged. 1.2.4 Method for testing the flexural strength of resin-coated sand: Flexural specimens were prepared according to foundry resin-coated sand standard JB/T 8583-1997. Test method: A flexural specimen was placed on the two support knife edges of a strength tester, with a support spacing of 150 mm. A single loading knife edge applied a uniform load perpendicular to the center of the specimen until it fractured. The strength value was read directly from the tester. 1.2.5 Method for testing resistance to shell peeling: A self-developed resin-coated sand shell-peeling resistance tester, shown in Figure 2(a), was used. The procedure was as follows: Heat the tester mold to 250°C and pour 3 kg of prepared resin-coated sand into it. After 60 s, withdraw the mold bottom plate so that unbonded sand falls onto the collection tray. Immediately empty this sand from the tray, return the tray to its position, and weigh the discharged sand, recording its mass as W1. After another 60 s, if shell peeling occurs, weigh the peeled sand and record its mass as W2. The formulas for calculating the shell-peeling rate and curing rate are as follows:

2 Results and discussion. 2.1 Relationship between resin viscosity-temperature characteristics and resin-coated sand strength: The viscosity-temperature test results for the four resins are shown in Figure 3, and the strengths of their corresponding resin-coated sands are listed in Table 1. The results show that all four resins follow similar viscosity trends. Resin viscosity gradually decreases as temperature rises. When the temperature increases beyond a certain point, the rate of viscosity reduction begins to slow, at a viscosity of approximately 45 poise. This paper defines the temperature at that point as the resin's constant-viscosity temperature. The curves show that viscosity continues to decrease above this temperature, but at an increasingly slower rate. Within the experimental range, viscosity can be considered approximately constant. This value is called the minimum viscosity and can be read from the resin's viscosity-temperature curve.

In the process of mixing resin with base sand, resin viscosity is the key factor determining whether the coating is uniform. The resin's viscosity–temperature curve describes its transition from the glassy state to the viscous-flow state, that is, its softening behavior. Statistical analyses show that there is no direct relationship between the softening point of phenolic resin and the strength of resin-coated sand. One important reason is that the resin softening point cannot fully and accurately describe the softening process [1], whereas the viscosity–temperature curve reflects that process more faithfully. Therefore, the traditional method of determining the coating temperature by adding 40–50°C to the resin softening point [2] may need to be revised. It is more scientific to use the resin's constant-viscosity temperature to determine the coating temperature in the process parameters. Table 1 shows that the higher the resin softening point, the higher the constant-viscosity temperature and the greater the minimum viscosity. The softening points of resins R1 and R2 differ considerably, but their minimum viscosities are very close, and both also provide relatively high strength. This demonstrates that there is no causal relationship between resin softening point and strength. This result points the way toward synthesizing resins with both high softening points and high strength: a resin with a high softening point and low viscosity can provide high strength without agglomeration. The constant-viscosity temperatures of the four resins, defined as their temperatures at 45 poise, differ considerably. To obtain a resin viscosity close to the minimum, the coating temperature, or processing temperature, should be determined primarily with reference to the constant-viscosity temperature, together with the processing conditions. Of the four resins, the resin-coated sand prepared with R2 had the highest strength. A process-control principle can therefore be established: coating temperature = constant-viscosity temperature + 40°C. 2.2 Relationship between the viscosity-change curve of resin with curing agent and resistance to shell peeling. During core shooting, filling and hot pressing, the resin film on the surface of resin-coated sand softens under heat while simultaneously undergoing a curing reaction under the action of the curing agent. The combined effects of softening and curing determine the coated sand's resistance to shell peeling. To better characterize this resistance, the softening and curing reactions of the resin film on the sand grains were considered separately from the sand: resin and curing agent alone were mixed, and the viscosity-change curve was measured. The results are shown in Figure 4.

The figure shows that the minimum melt viscosities of the resins with softening points of 71.5°C, 80.5°C and 90.5°C are 105, 210 and 260 poise, respectively. The low-viscosity plateaus on the three melt-viscosity curves last 20, 15 and 10 s, respectively. The resistance to shell peeling of coated sands made with the three resins was tested. None showed visible shell peeling during the test. The specimens were cut open using the method described above, and crack widths were measured; the results are shown in Table 2. It can be seen that as the minimum melt viscosity increases, the curing ratio gradually decreases. The coated sand made with the 71.5°C resin has the greatest tendency to crack. This shows that the lower the minimum resin melt viscosity and the longer the melt-viscosity plateau lasts, the poorer the coated sand's resistance to shell peeling.　

Effects of Phenolic Resin Viscosity on Coated-Sand Performance

The shell-peeling mechanism is analyzed as follows. Thermoplastic phenolic resin is in an amorphous condensed state when solid, so its softening process is continuous and uninterrupted. At approximately 50°C, the resin film, about 5 μm thick, has already begun to soften. Molecular flexibility and diffusion increase, and the interfaces between neighboring resin films gradually disappear, allowing the coated sand grains to bond to one another. As time passes, the temperature of the specimen as a whole continues to rise, and the region at the softening-point temperature moves progressively outward. Throughout this time, the resin film on the coated sand undergoes two changes: softening and curing. Softening elongates the resin film and permits the sand grains to slip relative to one another, while curing tends to reduce this slip. When the region at the softening point is close to the plate surface, the heat source, curing proceeds faster than softening, so the interior of the specimen does not delaminate as the resin softens. When the softening-point region is farther from the heat source, the resin softens faster than it cures. Delamination occurs at the cross-section where the softening and curing rates are equal, resulting in shell peeling. In addition, both resin softening and curing require time. The overlap between these processes explains why a shell sometimes delaminates without peeling. 3 Conclusions. (1) The viscosity–temperature curve of phenolic resin and its viscosity-change curve after adding a curing agent can serve as references for establishing the sand-mixing process and as a basis for assessing coated-sand strength and resistance to shell peeling. (2) The concept of constant-viscosity temperature was defined from the viscosity–temperature curve as the temperature at which the resin melt viscosity reaches 45 poise. The process-control principle of coating temperature = constant-viscosity temperature + 40°C was proposed. (3) The minimum resin viscosity was defined from the viscosity–temperature curve. Minimum resin viscosity is directly related to the strength of resin-coated sand. Resins with different softening points may have the same minimum viscosity, pointing to a direction for research into resins with both high softening points and high strength. (4) The effect of the relationship between melt viscosity and time in resin containing a curing agent on shell-peeling resistance was studied: the lower the minimum melt viscosity and the longer it remains at that minimum, the poorer the resin's resistance to shell peeling.