|The iron greenhouse energy budget|
The tests trying to produce some replication of Willis's Iron greenhouse have been repeated using a multichannel thermocouple probe recorder measuring to 0.01°C (accuracy 1°C) and somewhat modified test setup.
The sensor has been modified to make it more responsive (less copper in the plate than in the original cone. Fine wire thermocouple used to reduce the heat conduction.
The insulated hot box uses much thicker insulation and the double sided grey sprayed copper plate temperature is monitored using another fine wire thermocouple. The temperature between the 2 IR windows and outside the window is now measured. A small fan is used to provide a continuous stream of ambient air between the hot box and the sensor to prevent conduction and convection effects upsetting the sensor reading. The heating voltage is maintained to 9.254 volts +-3mV ensuring constant power input to the hot body.
The object of the experiment is not to replicate EXACTLY the steel greenhouse thought experiment. For a starter the hot object is only "surrounded" on one side and the major loss of heat from the object is via conduction through the insulation. There is obviously conduction and convection occurring in the experiment which are prevented by using a vacuum in the thought experiment.
|Inside of Hotbox|
Grey painted Plate Insertion:What is expected is a significant increase in the hot body temperature when the grey plate is inserted. The Sensor should report a drop in temperature until the system has reached equilibrium It should then be at the same level as before the grey plate was inserted - I.e. the radiation from the hot box should be constant before and after grey plate insertion.
Reflective plate insertion
The rise in temperature of the hot body should now be significantly hotter than either no plate or grey plate (100% of forward facing IR should now be reflected back onto the hot body causing the temperature to rise until the additional energy balance can be restored.
The sensor should show a drop in heat for the time that the reflective plate is in position.
- During the test the voltage applied to the resistors heating the hot body is monitored and maintained within +-3mV of the nominal (giving a power variability of 0.065%)
- The test setup was run monitoring temperatures for about 10 hours.
- The recorded results were then analysed over a period when the ambient was most stable (after sunset).
- To allow for ambient vatiation the measured forced ventilation temperature in front of the IR window was smoothed and then subtracted from the hot body temperature.
The first temperature rise is with a reflective plate and the second rise is with a grey painted copper plate inserted between hot body and IR window. The low temperatures are when no intermediate plate is inserted.
The air outside the IR window plot (green) is the temperature as measured from the forced airflow. The corrected temperatures refer to measured temperature less the air temperature.
The temperature of the Hot Body shows results as expected - maximum temperature from reflective plate lower temperature from grey plate lowest temperature from no plate.
The measured temperature (= IR output) from the sensor does not agree with expected result..
With the grey plate not equalling the hot body temperature there would be expected lower IR emission so perhaps this could explain the lower temperature compared to no plate.
With the reflective plate the IR output does not go to zero. possibly the plate warms and the insulation is insufficient. The IR sensor needs to be improved - possibly an IR thermometer? But will these then read the temperature of the IR windows?
So the results show
A definite increase in hot body temperature if a reflective plate is used (5.5°C)
A definite increase in hot body temperature if a grey plate is used (3.5°C)
If you believe that backradiation or reflection cannot add energy to the hot body from which the radiation originates then these results alone disprove this belief.