Micromachined Mesoscopic Recuperators
Greg A. Whyatt
Project Description
This project involved designing, building, and operating a test stand for evaluating the performance of gas phase microchannel recuperators. In addition, a model was developed to aid in the design of recuperative heat exchangers and to allow understanding of the results of testing. Non-LDRD project funds were used to further develop and enhance the models and to design and build a recuperative heat exchanger for testing. The heat exchanger was then tested using non-LDRD project funds.
Technical Accomplishments
The heat exchanger test stand was constructed and demonstrated. A description follows.
Test Stand Setup
The test stand is shown in Figure 1. The installation of the OTT project recuperator #1 in the test stand is shown in Figure 2. The building nitrogen passes through a pressure regulator and then through a Brooks model 5863, 0-1000 slpm (standard liters per minute) mass flow controller. The mass flow controller sets the nitrogen flow entering the cold side of the exchanger. The nitrogen then passes through the exchanger; through a Watlow, 3 kW electric in-line furnace; and then through the hot side of the exchanger. The nitrogen is discharged through a tube-in-tube water-cooled exchanger and into a duct connected to the hood ventilation system. Chilled water is provided from a building supply.
Figure 1. Heat Exchanger Test Stand. On the left is the computer used for data logging. The squarish grey boxes on the wall and on the benchtop are the furnace power supply/ controller and the 3 kW electric furnace, respectively. Transmitting gauges provide differential and inlet pressures on both sides of the exchanger. The exchanger is to the right of the furnace (covered in insulation and aluminum tape). The silver cylinder to the right is a 0-1000 slpm mass flow controller connected to the house nitrogen supply. | 
Figure 2. OTT Recuperator #1 Installed in Test Stand. Tees at each connection allow for pressure taps and thermo-couples to monitor temperature, pressure, and pressure drop on each side of the exchanger. The device is insulated during operation. |
The temperature is measured at the inlet and outlet of both sides of the exchanger. The inlet pressures and the differential pressures are measured for both sides of the exchanger. The Dwyer model 603A differential pressure sensors have a 0-100-in. water range. The maximum operating pressure for the test stand is limited by the 35 psig maximum pressure on these gauges. A computer displays and logs the data for later evaluation.
The furnace is contained within an insulated box to reduce heat loss and prevent burns. While operating at furnace temperatures up to 700&$176;C, the box exterior was not hot enough to cause burns.
Collection of Data
Data were collected by setting the desired nitrogen flow rate and furnace temperature and then waiting for a steady state to be reached before recording a set of data. The nitrogen flow or furnace temperature would then be adjusted and the process repeated. In general, the dynamics of the system are not limited by the behavior of the exchanger, so transient responses cannot be studied accurately. Achieving a steady state required 15 to 45 minutes, depending on flow and temperature.
Evaluation of Models
The data from recuperator #1 were evaluated against model predictions. Agreement with predictions was good, as shown in Figure 3. Application of a reduced weighting for longitudinal conduction for metal outside the heat exchanger core improved the model, as shown in Figure 4.
Figure 3. Model Predictions vs Data before Weighting Edge Metal for Longitudinal Conduction. | 
Figure 4. Agreement of Model and Data after Weighting of Edge Metal for Longitudinal Conduction Correction. |
Similarly, the initial model estimates of pressure drop were fairly good but were improved by adjusting the assumptions for entrances and exits. This led to the agreement shown in Figure 5 for ambient temperature data.
Figure 5. Comparison of Pressure Drop Model to Ambient Temperature Data after Adjustment of Entrance/Exit Loss Coefficients. |
Conclusion
The test stand was set up and used successfully to test a recuperative heat exchanger. Good agreement between model predictions for pressure drop and heat transfer and measured data provide greater confidence in the ability to design microchannel recperators.
