In the extrusion process, good temperature control is important to achieve good process stability. There are two main types of temperature control: on-off control and proportional control. In on-off control the heater power is either full on or completely off.
The temperature versus time for on-off control of heating is shown here; the power vs. time is shown as well. When the measured temperature is below the setpoint, the power is full on. When it reaches the setpoint the power shuts off. After the initial increase from room temperature, the temperature will vary in a cyclic manner with a corresponding on-off cycling of the power.
The advantage of on-off control is that it is simple and the average temperature is right at the setpoint. The disadvantage is that the actual temperature always cycles and the actual temperature variation can be quite large, as much as 10-20 degrees C. The larger the extruder, the greater the temperature variation tends to get. Because of this, on-off control is not recommended in extrusion, except for non-critical processes.
Advantage on-off control:
• Simple
• Average temperature right at set-point
Disadvantage on-off control:
• Actual temperature cycles
• Temperature variation is quite large
In proportional control, the power is proportional to the temperature within a certain temperature region; this region is called the proportional band. The temperature versus time for proportional control is show here; the power vs. time is shown as well.
The advantage of proportional control is that the temperature can reach a steady value, as opposed to on-off control. The power level can adjust itself exactly to the level that is required to maintain the correct temperature. For good temperature control in extruders it is important to have proportional control of both heating and cooling.
Some commercial extruders combine proportional heating with on-off cooling - this results in poor temperature control when cooling is activated.
Advantage of proportional control:
• Temperature can reach a steady value
• Power level adjusted to correct level automatically
A limitation of simple proportional control or P-control is that the temperature can remain steady only as long as the thermal conditions around the extruder are constant. When there is an upset in the thermal conditions, such as a change in ambient temperature, the actual temperature will change and the P-control will not be able to correct it. In other words, in P-control there is no reset capability.
Disadvantages of proportional only control:
• Temperature will change when thermal conditions change
• There is no reset capability
In proportional control with integrating action, also called PI-control, there is reset capability.
The controller integrates the difference between actual temperature and setpoint and continues to act on the process until the difference is zero. When there is an upset in the process there will be a temporary deviation from the setpoint, but eventually the actual temperature will go to the setpoint again.
Controllers with derivative action react to the rate of temperature change. This allows the controller to react to a process upset more quickly. Controllers with proportional and derivative action are called PD controllers, controllers with proportional and integrating action are called PI controllers, and controllers with proportional and integral and derivative action are called PID controllers. PID control is commonly used in extruders.
For a control action, the controller has to be tuned to the characteristics of the extruder at typical process conditions. Tuning of a PID controller involves determining 1) the correct width of the proportional band and 2) the time constant for integrating and 3) the time constant for derivative action. Even a good controller will give poor control if it is not properly tuned.
As a result, careful attention should be paid to tuning controllers that require manual tuning. Nowadays, there are number of controllers that tune themselves automatically; these are called “self-tuning” or “auto-tuning” controllers. With these controllers one does not have to worry about manually tuning the controllers.
A relatively new method of control is fuzzy logic control or FLC. FLC is an artificial intelligence based technology, designed to simulate human decision making. It can be used in systems that use many variables to enhance process control. Developing a fuzzy logic application requires the generation of a knowledge base; this can be a time consuming process.
Generation of a knowledge base for FLC involves identifying:
• Process variables that are important in control
• Membership functions for each variable, such as high, low, and medium
• Fuzzy rules which define the knowledge of what to do about an observation, based on previous operating experience
• FLC is slowly starting to be used in the plastics processing industry. It has already been applied a number of times in injection moulding, fewer applications have been reported in extrusion. It has been shown, however, that FLC can outperform conventional PID control if the knowledge base is sufficiently developed.
