NOTE: This experiment uses flame heating. Ensure that a fire extinguisher is on hand and that no flammable materials are near the experiment. Follow all standard precautions for fire safety.

1. Fabrication of sample for quenching (see photograph, Fig. 1)

1. Cut a small length (~24 mm) of 9.53 mm diameter copper rod. Drill two small holes (1.6 mm diameter) about halfway into the rod near the two ends. These holes will be the thermocouple wells. As the holes and thermocouples are relatively small, we can assume they have minimal effect on the overall heat transfer behaviour.
2. Use high temperature epoxy (e.g., JB Kwik) to affix high temperature thermocouple probes into the two holes. Ensure that the thermocouple probe tips are pressed into the center of the copper sample as the epoxy sets.
3. Set up a water container as a quenching bath. Insert a third reference thermocouple into the bath near where the sample will be quenched.
4. Connect the three thermocouples to a data acquisition system. Set up a program (in LabVIEW for example) to log transient temperature measurements to a spreadsheet.

Figure 1: a. Photograph of instrumented copper sample in cooling water bath. b. Heating copper sample.

2. Performing experiment

1. Position a Bunsen burner or chafing fuel canister next to the quenching bath. Light the flame.
2. From a safe holding distance, gradually warm the sample over the flame (to ~50°C recommended for the first experiment). The sample can be held by the thermocouple leads (Fig. 1b).
3. Start logging the thermocouple data to file, and immerse the sample in the quench bath. Hold the sample steady so that forced convection heat transfer is minimal. Stop recording temperature data once the sample approaches within a few degrees of the bath temperature.
4. Repeat this procedure for progressively higher initial sample temperatures (up to ~300°C). For cases above 100°C, observe the boiling behavior after quenching the sample.

3. Data Analysis

1. For the logged temperature measurements, record the average sample temperature at each time as the arithmetic mean of the two embedded thermocouple readings.
2. Calculate the sample cooling rate at each logged time j as = (T_s,j+1-T_s,j)/(t_j+1-t_j) (values will be negative). Here, t_j is the time of each logged reading. It may be helpful to smooth these cooling rate curves by performing a running average with a sample window of 2-3 readings.
3. Calculate the experimental heat transfer coefficients h with Eqn. 2 using the cooling rate from Step 3.2, and measured bath (T_∞) and sample temperatures (T_s). How do these heat transfer coefficients compare with predicted values (Eqn. 4, see Results)?
4. For a case with initial temperature below 100°C, use the initial experimental temperature measurement and numerically integrate Eqn. 2 to predict the cooling over time. Use Eqn. 4 to predict the convection coefficient at each time. Compare this curve to measured values. For numerical time step size Δt (e.g., 0.1 s), the temperature can be integrated as:
     (3)