Tools for Real-Time Process Control

One of the most desirable aspects of plastics and composites is the ability to make net-shaped parts. The same process that creates the material also creates the structure. The penalty for this advantage is that the process of curing a thermosetting plastic or composite part is irreversible. Any part that is not properly processed represents a loss of part, material and the money and time required to make that part, although larger parts are usually repaired if possible. Proper shape becomes a controlled property in addition to the bulk material properties, such as mechanical (stiffness or strength), physical (density, void content, etc.), chemical (degree of cure or carbonization, chemical resistance), electrical (resistivity, conductivity), or any combination of these. 

No control is possible without some sort of preestablished plan of action. This plan can be rigidly followed throughout the cycle or it can be flexible, subject to change when the measured progress of the process indicates that change may be advantageous. The tools used to develop the plan for a process have been discussed earlier. This section presents some of the tools and methods used to follow those plans and provide flexibility where needed. 

It is not possible to discuss real-time control without a brief discussion of sensors and the measurements they represent. In traditional process control, the measurements and the properties to be controlled are identical. For instance, one controls the temperature of a fluid using feedback from a thermocouple. There is also generally a fairly predictable relationship between the measurement and the forcing function necessary to change that measurement. Except for unusually simple cases, that is not true of polymer processing. The multiple, complex properties to be controlled cannot be measured and are not always 

The main focus of this chapter, however, is on the tools that can use this sensor information for real-time control. These tools are still largely developmental because inferential control is relatively new to the composites industry and change is dependent on both technical merit and on other changes in the culture of the industry. 

---