Experimental Analysis of Heat Transfer at the Interface between Die Casting Molds and Additively Manufactured Cooling Inserts

In metal die casting process, particularly in die casting field, the significance of effective thermal management is widely recognized for both heating and cooling the molds (Ref 1-3). A significant portion of the casting cycle time is dedicated to the solidification phase, making the proper temperature management of the mold crucial for meeting quality standards and productivity goals. Metallic molds facilitate the heat extraction, while molten metal is continuously being injected into the die. Cooling channel systems are frequently employed to enhance the process, reducing cooling times and improving both the surface and microstructural quality of the resulting parts (Ref 1, 4).

Due to limitations in standard drilling processes, mold cooling systems have traditionally been comprised of straight channels (Ref 5). However, this configuration is not ideal for complex geometries that require optimized designs to achieve controlled cooling (Ref 1, 6). The utilization of mold inserts manufactured through additive manufacturing (AM) has been demonstrated to solve heterogeneous cooling issues in plastic injection molds (Ref 5, 7-9). Nowadays, metal AM enables the production of complex geometries, providing a wide range of cooling channel designs for the insert (Ref 10-15). On the other hand, the use of numerical simulation tools for the design of inserts with cooling channels of complex geometries enables customization of the mold cooling process and, consequently, of the produced parts (Ref 16, 17). However, the use of AM to produce inserts with cooling conformal channels restricts the range of materials to be used (chemical composition). The contact between the mold and the insert of different materials creates an interface, which constitutes a resistance to the thermal flow from the mold to the cooling channels, the thermal contact resistance (TCR) (Ref 18, 19). For each interface between two materials, the TCR is influenced by surface roughness in the contact zones between two parts, as well as by the different thermal properties of the two materials. Besides the cooling system design, the distance from the cooling channels in the insert to the mold/insert interface also affects the value of TCR at this interface. These factors as well as the thermal gradients and casting process parameters influence crack formation at the interface. An effective thermal dissipation management across these mold/insert interfaces is of utmost importance to ensure the durability of this connection without incurring any damage. Thorough analysis, simulations, and practical experience are needed to determine the specific distances and optimize them to minimize the risk of thermal cracking. As the TCR affects the thermal flow through the mold/insert interface, this is one important parameter to consider in the heat transfer numerical analysis, especially when high thermal dissipation rate is required (Ref 20-26). Different methods of studying TCR are mentioned in the literature, namely by analytical models or numerical simulations; however, the experimental studies approach is the most common (Ref 18, 27-29).

Besides, the employed distances between the cooling channels to the mold/insert interface in real HPDC dies and inserts, which can vary, also affects the TCR value. The correlation between these distances and the formation of thermal cracks on the surface of the die/inserts is complex. Factors such as thermal gradients, material properties, cooling system design, and casting process parameters influence crack formation. Thorough analysis, simulations, and practical experience are needed to determine the specific distances and optimize them to minimize the risk of thermal cracking.

This paper presents the study of heat transfer through the contact interface between a disk-shaped steel used for casting molds (1.2344) and disk-shaped steel used for AM-machined inserts (1.2709). The influence of (i) the distance between the cooling channel in the insert and the mold/insert interface, (ii) the water flow rate through the cooling channel and (iii) the contact pressure at the interface were studied. Heat transfer tests were carried out with water flows rates ranging from 215 to 425 mL/min, with contact pressure values between 0 and 30 bar. The thermal behavior of this assembly was evaluated in the transient and steady states. In the transient state, the system’s time constant and the temperature variation rates were assessed, while the actual heat loss through the water channel and the coefficient of heat transfer were determined in the thermal steady state. TCR values for this contact pair were also calculated. Thus, the heat transfer conditions in steady state were studied, as well as the possibility of locally modifying the cooling conditions during the thermal manufacturing cycle. This work aims to determine TCR values and other thermal properties at the mold/insert interface, under various working conditions. The obtained data will serve as boundary conditions in numerical models of casting processes, allowing for the optimization of the design of conformal cooling channels in a mold cooling insert.

In this study, two steel samples were prepared for testing in contact with each other to simulate heat transfer from a mold to an insert. One of the samples was produced, by AM, with a cooling channel to simulate an insert that is crossed by a cooling fluid. To simulate the mold/insert interface, two steel pieces, designated as A and B, were used. Sample A is a 1.2344 steel (DIN standard) commonly utilized in the manufacturing of die casting molds. It has a density of 7.8 g/cm³ and a thermal conductivity ranging from 25 to 30 W/m.K (between room temperature and 700 °C). Sample B is a 1.2709 steel (DIN standard) produced via AM (Laser Power Bed Fusion) and can be used in the fabrication of mold inserts. It has a density of 8.1 g/cm³ and a thermal conductivity ranging from 14.2 to 28.6 W/m.K (between room temperature and 1300 °C).

Sample B was produced with a pass-through channel to simulate a simple cooling channel. Figure 1 shows the technical drawing of sample B with a Ø 5-mm cooling channel and three Ø 2-mm holes for placing the thermocouples during the heat transfer tests. Sample A has the same design, without the cooling channel. The thermocouple holes were placed in the piece to measure the temperature along the thickness of the sample (thermocouple positions in the 0-21-mm thickness range of the disk: 4, 10.5, and 17 mm), and were separated by 120°. The diameter and depth of each hole were, 2 and 25 mm, respectively. Samples A and B contact surfaces were polished with P600 sandpaper to achieve a proper surface finish in the contact areas.

Each of the steels corresponding to samples A and B was characterized for its coefficient of linear thermal expansion (CTE). Thermal expansion analyses were conducted using a TA Instruments Q400, where a glass probe with a diameter of 2.75 mm made surface contact with the steel sample (size: 3 × 3 × 3 mm). Each steel sample was heated at a temperature ramp rate of 3 °C/min under normal atmospheric conditions, ranging from 40 to 250 °C. The method employed enabled precise measurement of the CTE, a critical parameter for predicting dimensional changes of materials at various temperatures. By comparing the outcomes obtained from samples A and B, insights can be gained into the distinct thermal expansion behaviors between the two steels and their possible impact on the A/B contact interface. This knowledge can enhance our comprehension of the materials' performance and durability in this context and facilitate further improvements in their design and application.

Heat transfer through the contacting samples A and B was tested. During these tests, water was used as a coolant. The cooling channel in sample B is displaced from its line of symmetry. In this way, depending on which of its surfaces is in contact with sample A, it is possible to obtain two different test configurations corresponding to different distances between the channel and the A/B contact interface (6.5 and 9.5 mm). The thermal properties of the two A/B systems were investigated by testing coolant flow rates of 215, 300 and 425 mL/min through the sample B cooling channel. Thus, it was possible to monitor how the hc and TCR value were affected by these parameter changes. Tests were conducted with an initial input power of 460 W. Figure 2 shows the experimental setup, used for this study.

Schematic representation of the heat transfer equipment used in these tests

The experimental setup for the tests comprises a heating resistor (3, in Fig. 2), which is the heat source. This heating resistor is a sleeve-style component that envelopes a copper cylinder (2, in Fig. 2) and provides an initial power of 460 W through it. It was thermally insulated with rock wool to reduce heat losses during tests. The copper cylinder, which conducts heat to the samples (4 and 5, in Fig. 2), is hydraulically driven, thus ensuring constant contact between samples A and B during the tests. Three different contact pressure were explored during the tests: 0, 15, and 30 bar. At the opposing end of the setup, a cold water stream was placed (7, in Fig. 2), with cooling fins (6, in Fig. 2), which act as the cold source of the system. Samples A and B were placed between the hot and the cold source and were wrapped with rock wool to minimize heat losses, ensuring the steady-state condition. According to the setup shown in Fig. 2, the configuration includes ten K-type thermocouples: one in the hot source, three in each sample (distanced by 6.5 mm from each other), and one in the cold source. To assess heat losses from the equipment, two K-type thermocouples were installed on the surface of the equipment, and a thermocouple was used to measure the room temperature. In sample B, two K-type thermocouples were also placed at the entry and exit of the cooling channel to assess the heat loss (T_IN to T_OUT). The temperature variation, at the different measurement points of the experimental device, was inferior to 2.5 % of the measured value. By recording these temperatures over the time, the thermal conductance, hc, and the TCR were determined according to Eq 1-3 (Ref 18).

As seen in Eq 1, the heating resistor applies an input heat flux to the system \({(Q}_{\mathrm{e}})\), but a portion of this heat is lost during the process \(({Q}_{\mathrm{o}})\). The remaining heat that is conveyed to the sample material is known as the effective heat flux (\({Q}_{\mathrm{E}})\):

By utilizing the data gathered from the thermocouples, as well as the effective heat flux, it is possible to calculate the \({h}_{\mathrm{c}}\). This involves dividing the heat flow by the temperature gradient at the point of contact, as demonstrated by Eq 2:

Equation 3 relates the parameters of apparent contact area (A), temperature difference at the interface (ΔT), and effective heat transfer rate \({Q}_{\mathrm{E}}\) to TCR. ΔT is determined by extrapolating the temperature profiles recorded for each material.

The TCR is a measure of the impact of contact on heat transfer and is defined as the reciprocal of h_c. By determining the TCR value under different test conditions, it can be used as a boundary condition in numerical simulations for further optimizations.

The heat released through the channel to the coolant water—actual heat loss (\({Q}_{W}\))—is determined by Eq 4, where m is the mass of water and depends on the water flow rate (kg/s), \({C}_{P}\) is the specific heat of the water, and ΔT is the temperature change of the water between the entrance and exit of the channel:

The actual coefficients of heat transfer from the steel system to the coolant water (\({h}_{\mathrm{W}}\)) are determined by Eq 5, where A is the area of thermal conductivity, and ΔT is the difference between the temperature average of the water inside the channel and the temperature average of the steel on the surface in contact with water:

To characterize the steels present at the A/B contact interface, thermal expansion analyses were conducted on samples A and B steels to determine their thermal expansion properties. Figure 3 displays the graphs depicting the dimensional variation as a function of temperature obtained during the thermal expansion tests of steels A and B, from 40 to 250 °C under normal atmospheric conditions (temperature ramp rate: 3 °C/min). The CTE values for each steel were determined using the graphical analysis software (TA Universal Analysis Software) associated with the equipment used to conduct the thermal expansion tests. The CTE values for sample A steel (used for molds) are shown in Fig. 3, which displays three distinct zones. On the other hand, sample B steel (used for inserts) shows a linear dimension change evolution, with the corresponding CTE values (Fig. 3). Sample B steel, manufactured via additive manufacturing, exhibits greater dimensional variation than sample A steel, and a slightly higher CTE within the service temperature range. These differences in dimensional variation between the two steel types cause fluctuations in contact pressure during the thermal cycles to which the mold/insert assembly is subjected, leading to variations in heat transfer conditions during the manufacturing process.

The arithmetic mean roughness (Ra) was measured on both contact surfaces for the two samples, which had the same surface finish (P600). Figure 4 displays a roughness curve from the steel disk contact surfaces of both samples A and B (Fig. 4a and b). The recorded profiles indicated that the Ra measured on the surfaces was identical for both samples, with a mean value of 0.05 µm. Despite having the same Ra (0.05 µm), samples A and B exhibited distinct roughness patterns. The surface of sample B, which was manufactured using AM, has a more uniform surface than sample A for the same surface finish.

During the heat transfer tests, the temperature profiles in the A/B system were measured at points T1 through T6, as indicated by the thermocouple positions shown in Fig. 2. Figure 5 illustrates the two types of heat transfer tests conducted, showing the temperature curves measured at position T6 (the point furthest from the heat source). The system was first thermally stabilized by using a flow rate of 215 mL/min before increasing the flow rate to 300 mL/min (gray curve) and subsequently to 425 mL/min (black curve). The study aimed to examine the effect of increasing the flow rate from 215 mL/min to 300 mL/min, and then to 425 mL/min, as well as the effects of the flow rate transitions, as shown in Fig. 5.

The heat transfer system's behavior was characterized in two different scenarios: during the transient state (when the flow rate was increased) and during the thermal steady state before and after the change in the cooling flow rate. Figure 5 shows that the flow rate transitions and subsequent thermal stabilizations to a new steady state are very fast. This study aims to help understand the effect of different flow rate changes on the thermal cooling cycle of casting molds, and how these changes can assist in adjusting it locally (in time and temperature).

This study investigated the impact of cooling channel location on the heat transfer properties at the contact interface between two different types of steel, steel A (1.2344) used for casting molds, and steel B (1.2709) used for cooling inserts produced by additive manufacturing (AM). Both transient and steady-state heat transfer tests were conducted to evaluate the system's thermal behavior.

In the transient state, a closer proximity of the cooling channel to the interface resulted in a faster cooling response of the mold. Specifically, an increase in flow rate from 215 to 425 mL/min led to an 9% greater increase in temperature variation rate compared to an increase from 215 to 300 mL/min (38%). Furthermore, increasing the contact pressure from 0 to 30 bar resulted in a 50% faster cooling response, while an increase in the coolant flow rate accelerated the cooling response by 123 to 167%. The actual contact pressure between the mold and the insert is determined by the type of screw tightening and the working temperatures, which in turn influence the CTEs of each part. These are critical factor once a higher contact pressure reduces TCR values and enhances the cooling response of the mold/insert assembly.

In the steady state, the TCR values were affected by the contact pressure between the two steel pieces, showing, for an increase of 30 bar, a decrease of 21% and 33% for cooling channels located at 6.5 and 9.5 mm from the interface, respectively. For the tested range, cooling flow rates did not significantly influence the TCR values. Irrespective of the test conditions, the TCR values are consistently lower when the cooling channel is closer to the interface (~28%), indicating more significant heat transfer.

This study provides comprehensive insights into the impact of each studied parameter on the TCR, which is a critical boundary condition for designing conformal cooling inserts produced by AM.