November 1-4, 1999

Cleveland, Ohio

Warren Bishenden, P.Eng., Supervising R&D Engineer

Rabindra Bhola, Analysis Engineer

Exco Engineering

Ontario, Canada

The quality of castings in high-pressure die-casting is a function of many interdependent parameters. Experience has shown that many defects can be traced back to poor die temperature in a few critical areas. It would therefore be very beneficial to have the ability to directly control these critical features – making them less sensitive to thermal related parameters.

Exco has developed a system to attain and maintain parts of a die cavity to a uniform preset temperature range during the production of castings. This was accomplished by using a Proportional Integral Differential (PID) control algorithm that regulates the water flow-rates to the die based on direct temperature feedback from within the die. Preliminary results demonstrated the effectiveness of this system. In addition to reducing scrap during casting production, valuable heat transfer data was recorded. This information will be used to conduct further research in the thermal behavior dies.

Pressure die-casting is a process in which a molten metal (usually Aluminum or Magnesium) is injected into a steel mold under high pressure (~10,000 psi) and allowed to solidify. When the part is sufficiently cooled to the point where it is rigid enough to retain its shape, it is ejected and the cycle repeats. The die serves two basic functions: to retain the desired shape of the casting and to remove the heat from the molten metal in a reasonable amount of time. There are many things that must be done right in order to efficiently produce good castings. Most of which are related to the fill pattern and the subsequent solidification of the casting. It is already an established fact that proper thermal management of dies yields a higher casting production rate and improved part quality. Improper die temperature causes a number of defects in the part, including shrink porosity, poor fill (knit lines), soldering, stuck castings, etc. And the die temperature is dependent on a number of process variables. These include cycle time, spray distribution and duration, waterline layout and flow rate, casting volume/geometry, metal temperature etc. Experience has shown that a majority of casting defects from transmission cases are due to one or two problem areas. And these problem areas are often related to improper temperature management.

The conventional approach to thermal management of dies is based on trial and error and the experience of the operator. But as industries become more competitive, there is an ever-increasing need to develop a more systematic approach to manage the temperature of dies. Conventionally, the designer establishes a water line layout, and the die caster run the die with maximum flow rates through each line. When significant thermal defects are produced, the waterlines are manually adjusted. The action of an experienced operator does improve the situation. However, it would be necessary for him to monitor the process frequently without knowing the effect on temperature until it is reflected in the scrap rates; at which point it is too late to correct the problems before they occur.

There are a number of commercial temperature control units designed to help manage the temperature of a die, most of which are hot oil or hot water units. Hot oils are circulated through the die to vary its temperature in a desired direction. But unless there is a probe within the die, its temperature cannot be accurately controlled. And this does not bring us any closer to operating the die consistently at a desired preset temperature. There are also a number of papers written on controllers based on direct temperature feedback from the die. Booth S.E.[4] varied the cycle time based on cavity surface temperature feedback, but this method will change the temperature of the entire cavity and cannot correct localized problems (which are most common). Also, varying cycle times has a very slow response time and may slow production to unacceptably low levels. There are a number of control schemes, which vary water line flow with on/off solenoid valves [1, 2 & 6]. In our quest to systematically understand and isolate the thermal problems of dies, we have conducted a number of experiments and computer simulations on a given die. Out goal is to determine if it is feasible to control a die’s temperature by varying the flow rates through the water lines. The following pages will discuss experimental and numerical results, which will provide more insight into automatic die temperature control. Also, a brief discussion a prototype die temperature controller and its performance is also included.

There are two main questions that should be addressed when designing a temperature control system which adjust water flow:

1. Can the waterlines remove enough of the die’s heat to significantly influence its temperature?
2. The temperature in a die varies with position and time, so which reading is most appropriate to represent the region that we wish to control?

To address point one: Flow meters and temperature sensors were installed at the inlet and outlet of all the waterlines of a prototype case die. After reaching steady state, the flow rates, inlet and outlet temperatures were recorded. This provided us with the necessary information to calculate the average energy that is removed from the die via the water lines. By measuring the casting ejection temperature, pour temperature and knowing the mass of the metal poured, we can calculate the total energy that enters the die per unit time. For this particular die operating at steady state, the water lines removed 78% of the total energy that enters the die.

Thermocouples were inserted in the die 1" away from the cavity surface to monitor its temperature. With flow rates set at maximum, the steady state temperature was recorded. This was repeated for water line flow set at minimum. Figure 1 show the effect of water flow in steady state temperature of the cover die. From these results it can be seen that cover temperature (1" away from the cavity) can be adjusted by more than 120° C (216F) by simply changing water flow rates. This represents the maximum influence that a controller can have on the temperature of the cover die. When the die ran at the high temperature extreme, soldering and stuck castings become common. And when the die run at the low temperature extreme, poor fill and knit lines become a problem. Therefore, this window of control is sufficient to allow the die caster to operate at an optimum temperature for maximum efficiency and high casting quality.

The temperature of a die varies dramatically with position and time in the vicinity of the cavity. We performed thermal simulation of the die (using Magmasoft) to examine the effect of water flow rates on the die temperatures. The temperature of two prescribed locations were recorded and plotted (figure 2). Point (1) is located at the cavity surface (0.1" inside the die steel) and point (2) is located 1" away from the cavity surface in the die steel. For time less than 30 minutes, the water was set at maximum flow (2 gal/min per line) and for time greater than 30 minutes the flow rate was reduced to .5 gal/min. It is evident that the temperature at locations (1) and (2) are related to each other. That is, as the average of (1) increases, (2) also increase by a proportional amount. This would suggest that location (2) could be used to represent what happens to (1). From these results, it was concluded that temperature at 1" below cavity surface can be used to represent temperature at the cavity surface and any change at the surface reflects a proportional change in temperature inside the die steel.

Based on the information found in the analysis and experiments, we designed and built a prototype temperature controller to vary water flow based on temperature feedback from within the die. Figure 3 is a schematic of the temperature controller. The basic idea behind this controller is to measure the temperature of the die at a given time and perform PID control based on these readings to regulate the coolant to the die. The thermocouples must be positioned to sense changes due to the water lines and as well, reflect cavity temperature. Since one thermocouple reading is insufficient to represent the temperature of the entire die, it is necessary to divide the die into separate zones which can be considered independent. Each zone can be defined as a region whose temperature is affected by a group of waterlines. This group of cooling lines can be considered as a single cooling unit and their flow can be regulated by a single valve with temperature feedback from a single thermocouple strategically positioned in that zone. For example, one entire slide or a "valvebody" in a transmission case die can be considered as one zone. For cases where thermal defects are even more localized, the size of the zones can be reduced accordingly.

The best way to test the controller is to determine if it can attain a prescribed set-point and to compare the controller’s test results with those without. Figure 4a shows measured temperature variation for two positions within a die under normal operations (i.e. water are on maximum flow and production is consistent). These results show that if the cooling is not tampered with, it is possible to get very consistent temperature during production. The only problem is that the die was running much too cold to produce good casting. Figure 4b shows the same process where the operator manually adjusts the flows as he see fit at a given time in the process. The problem here, is that when scrap parts are produced, it is already too late to correct and most of the time it is over corrected and the error in the process continues. Note the dramatic fluctuation in mean temperature as an operator tries to correct the problem.

After properly tuning the temperature controller for this specific die, and noting from previous runs which temperatures produces the best results, the die was run with the controller on. It automatically corrects flow as it sees fit to attain and maintain the required set-point value. Figure 5 shows the performance of the temperature controller for the bottom slide of the prototype case die. The optimal PID parameters are not yet obtained which accounted for the overshoot at 30 minutes. But after changing the proportional bandwidth, the result is a stable and consistent temperature of the bottom slide. The circled region is a result of having the shot tip stuck causing the die to cool during the down time. But it is interesting to note that the set point was re-established within three shots. This is because all water was automatically shut off as the die cools as a result of the down time, which reduced the startup scrap. More tests will be conducted to further demonstrate the effectiveness of the system to prove that it can quickly recover from down times and also maintain consistent die temperatures throughout production. We found it very beneficial to record the performance of the controller since we can directly correlate casting quality with the temperatures history of the die. Therefore if there is a problem in the process, the thermal history can be used to compare operating temperatures that produce good parts with temperatures that produced scrap.

1. There must be more that adequate cooling to the areas that we are interested in cooling. A die that runs hot with all the water on cannot benefit from this system
2. Zones must be defined as regions that have water lines that are common to each other and controlled as one unit.
3. Thermocouples should be placed at a location that would reflect the amount of cooling flow to a particular zone. It is recommended that the thermocouples be placed about 1" from the cavity surface and about the same distance from the nearest waterline.

1. Kiser, W.D., Sanders, S.D., Frost, "The llzro-Battelle Multichannel Temperature Controller" 7^th SDCE International Die Casting Congress, p5172, (1972).

2. Dent, B.K., Fifer, R., "Production Operation With The llzro-Battelle Die Temperature Controller" 7^th SDCE International Die Casting Congress, p5572, (1972).

3. Larkin, R.J., "Automatic Control of Die Temperature in Zinc Die Casting" 6^th SDCE International Die Casting Congress, p 62, (1970).

4. Booth, S.E., "A Die Temperature-Cycle Time Controller" 6^th SDCE International Die Casting, p54, 1970.

5. Peterson, A.P., "Thermocycling Control of Aluminum Die Casting Machines", 8^th SDCE International Die Casting Exposition and Congress, p B-t75-024, 1975.

6. Reddy, R., "Temperature Control of Die Casting Dies", ", 8^th SDCE International Die Casting Exposition and Congress, p B-T75-025, 1975.