Time-dependent heat leaks

A special care is to be devoted to the control that all the parts of the apparatus have reached the desired temperature when parts remain at higher temperature, due to the high value of the specific heat, the cooling only by radiative exchange is usually impossible. To open a gas heat switch, several hours of pumping are usually necessary to reduce the pressure to a value suitable for the thermal isolation. An insufficient pumping leads to a time-dependent heat leak due to desorption and condensation of the residual gas at the coldest surfaces. 

The enthalpy of copper at nitrogen temperature is H77K = 6 J/g, so the total entropy of the sphere will be about 6 x 106 J. The time needed to cool from 77 K down to 4K is of the order of 4h. The total helium consumption from room temperature to 4.2 K would be about 6001. The temperatures reached in a test run are reported in Table 16.2. The expected final sphere temperature is about 20 mK. A comparison of MiniGRAIL and Nautilus cool down is made in Table 16.2. The high power leak on the sphere has been attributed to a time-dependent heat leak caused by the ortho-para conversion (see Section 2.2) of molecular hydrogen present in the copper of the sphere (see Fig. 16.5) (the Nautilus bar instead is made by Al). A similar problem has been found in the cool down of the CUORICINO Frame (see Section 16.6). 

For a sufficiently fine subdivision of the body, it is possible to find a suitably exact solution of the heat conduction problem. Several FEM software packages are available on the market. The ANSYS and the COMSOL systems are often used by calorimeter-developing corporations. The software can be used to define the time-dependent heat flux curves at any site of the calorimeter system and also to indicate the temperature field if this is necessary for the purposes of the discussion and for a better understanding including heat leaks of all kinds. 

In the event of failure of all the heat removal loops of the secondary circuit, the primary reactor water heats up and starts to evaporate. The heat-up time depends on conditions in the reactor and in the environment and ranges from several hours up to several days. Evaporated primary water passes to the air space of the reactor hall and condenses on the inner surfaces of the containment which conducts heat to the environment. There is also a quite substantial heat loss into the ground surrounding the reactor aided by the large siuface area of the reactor tank. Devices to assist condensation in the leak tight reactor hall are envisaged, as is return of the condensate to the reactor. 

The discussion of differential thermal analysis instmmentation is concluded with the description of thermal analysis under extreme conditions. It is mentioned in Sect. 4.3.2 that low-temperature DTA needs special instramen-tation. In Fig. 4.10 a list of coolants is given that may be used to start a measurement at a low temperature. From about 100 K, standard equipment can be used with liquid nitrogen as coolant. The next step down in temperature requires liquid helium as coolant, and a differential, isoperibol, scanning calorimeter has been described for measurements on 10-mg samples in the 3 to 300 K temperature range. To reach even lower temperatures, especially below 1 K, one needs another technique,but it is possible to make thermal measurements even at these temperatures. Usually heat capacities and thermal conductivities are obtained by heat leak, time-dependent measurements. 

Steam heat provided the fastest recovery time of any system because of the oversize source available in the boiler room. It offered a uniform mold temperature, as do all liquid systems, but is limited to about 350F (177C) maximum. Steam heat is also messy and requires good maintenance, or rusty pipes and leaks become all too common. Steam controls and the accompanying valves are expensive and many are not dependable. 

---