Setting and controlling the frost point:

• Mixing of wet and dry flows
  The frost point is set by mixing wet CO2 flow and dry CO2 flow. To keep the chamber pressure constant while adjusting the humidity, the total flow rate must remain constant, therefore any increase in the wet flow must correspond to a decrease in the dry flow.
• Frost/Dew point of the wet and dry flows
  The frost/dew points of the saturated and dry flows are controlled by their temperature. In the saturator, CO2 out of the bottle is passed over liquid water or ice giving a saturated vapor pressure that depends on the temperature of the saturator. The dry flow is generated by passing bottled CO2 through a chamber that is held at -78.5 C (slurry of dry ice and ethanol). This flow is therefore dried to that frost point.
• Expansion through critical orifice
  Since the atmospheric pressure on Mar's is only 4-6 torr, the flow must be expanded to this pressure range. This is done by means of a sonic orifice. This expansion also decreases the concentration of water vapor. Currently we have two sizes of orifices, but others can be used to obtain different expansion ratios for control of flow rate, humidity, and pressure.

Example frost point data:

The first figure below shows how the frost point can be controlled by varying the the wet/dry flow rates, while keeping the pressure constant. The top plot is the frost point where each step has a remarkable constancy. This small settling time of humidity in the chamber is a result of using electro-polished tubing. The second and third plots are the dry and wet flow rates, respectively. Notice that a change in one flow rate corresponds to an equal and opposite change in the other which maintains the pressure at a constant level, as seen in the fourth plot.

This figure illustrates the settling time of both the frost point and the pressure after performing a step decrease in the wet flow rate. The frost point settled down within about 7 seconds after performing the switch, and the pressure was minimally affected as can be seen in the last plot.

This figure also illustrates the settling of both the frost point and pressure after performing a step decrease in the wet flow rate, but this time the frost point is really low to start with. It can be seen that there are two characteristic settling times for the frost point. The first can be attributed to the settling time of the chilled mirror hygrometer and the second to the humidity "capacitance" of the system.

This last figure shows how the frost point can be continuously changed in a stable manner, which has applications in simulating a Mars diurnal frost point variation. The bottom plot termed 'Balance' is the signal that comes from the hygrometer which lets one know that the frost point is indeed the actual frost point. If this number deviates significantly from ± 0.01 V, then one could not attribute the frost point signal from the hygrometer as the actual frost point. Here it can be seen that the frost point from the hygrometer is indeed the actual one. These data were collected with the system at 6 torr.