The objectives of these experiments are two-fold. First, to characterize two heat exchangers (steam-water and hot water-cold water) by determining heat transfer coefficients as a function of Reynolds number and operation mode (co-current vs. counter-current). Second, to demonstrate how feedback changes the dynamic behavior of a system. In particular, we want to show that by using a simple approach like PI control, zero steeping error, a reasonable speed of response and acceptable damping can be achieved by proper setting of the control constants. In conjunction with lectures on control, this experiment can be used to demonstrate how model mismatch limits the performance of a feedback system and how advanced strategies like feed-forward, cascade and adaptive control can be used to improve performance.

A number of variables can be measured and manipulated:

• Temperature sensors are installed at numerous locations
   
• Flow rates can be manipulated at two points via automatic valves
   
• Flow rates are measured at two points
   
• Different configurations can be chosen by using manual valves
   

The experiment has three distinct sections. First, a controlled flow of cold water is heated in a steam condenser. After the water has been heated, it is passed through two heat exchangers in series. Both heat exchangers can be operated either in co-current or counter-current mode. Likewise, the sections can be analyzed separately or together as a system. All measurements and manipulated variables are connected to a laboratory computer employing a LabVIEW interface.

This experiment can be used to illustrate concepts developed in the design, heat transfer, and process control courses. The students are introduced to the use of heat transfer coefficients, design models, and feedback. Results demonstrate how design calculations and simple design models can be used to synthesize a feedback control system and how different control configurations can be used to improve performance. The experiment can also be used to show dynamic analysis and control may influence decisions made at the design stage.

• For the steam-water heater, determine the steam film coefficient, water film, and overall coefficient as a function of the operating conditions, e.g., steam temperature and water flow rate.
   
• For the hot water-cold water heat exchanger, determine the hot and cold water film coefficients as functions of temperature level and water velocity in the heat exchanger operated in co-current and counter-current modes.
   
• Measure and correlate heat-transfer parameters in the system using steam heating of tube-side water followed by countercurrent cooling of the flow. Measure dynamic response to a step change in tube-side water flow rate.
   
• Measure and correlate heat-transfer parameters for the upper and lower segments of the cooling-water heat exchanger operated cocurrent with the tube-side flow. Measure dynamic response to step changes in steam and in cooling-water flow rates.
   
• Optimize control of the exit temperature of process water from the steam heat exchanger using steam flowrate as the manipulated variable. If possible, also optimize control of exit temperature for process water from the water-cooled exchanger to minimize effects of disturbances in steam flowrate.
   
• Calibrate and compare the different flowmeters in the system. Control the exit temperature of process water from the cooling-water heat exchanger by adjusting the flowrate of process water.

There are several aspects of this experiment that can be controlled, through which the principals of process control can be studied. This experiment enables you to evaluate various modes of temperature control (at different positions with different control parameters, PID). There are two flow rates and the steam valve position which can be manipulated to control the outlet temperature of either coldwater inlet stream. These can be operated in either co- or counter-current flow.

The approaches to each control would vary. In each case, you would first determine the dynamics of the process in response to the changed (manipulated) variable. This can then be approximatedby a dynamic model (first- or second-order, wi th or without time-delay, etc.). Various "ideal" control parameters (P, I, and D) can then be estimated from any of several models (Z-N, C-K) available in the literature.

In each case, it is necessary to compare the dynamic response model to the measured response. This relation can be quantified by an evaluation of the measured response versus the model.This is further verified by a comparison of the response with control to the model with control.

Higher levels of dynamic analyses are available in this experiment. The integral windup (reset). can be set by the computer. Further, the set point can be varied sinusoidially (with pure "P" control) such that frequency response analysis can be performed.

In each of these cases, it is necessary to perform analyses between laboratory periods to prepare and plan the experiments to be performed in the next laboratory period.