README_Thermocouple Tests

To ensure validity of the of the experimental thermocouple measurements from
the large temperature jump tests, the thermocouples were placed in an 
insulated vessel of hot water. The water was unmixed, the thermocouples were 
left to record the temperature as the water cooled to assess the difference 
between them.

In the attached data, the standard deviation has been calculated and there 
are two plots showing the error bounds from the mean average temperature 
reading.

This was performed as the most likely error will come from the data 
acquisition device itself. The DAQ device was an NI 9214 TC module plugged 
into a cDAQ-9178 chassis. The 9214 module specification sheet is attached. 
The largest error is likely to come from the cold junction, the module used 
has an isothermal terminal block to reduce this.

The quoted typical value of error (not the maximum value), at high resolution
sampling at 100C and 0C is 0.37 and 0.36C +/-. The standard deviation plot 
with the error bounding is within this value.

The data reported in the paper which relies on the temperature measurements 
is the enthalpy of reaction. The enthalpy value is plotted based on the rate 
of change and achieved using a best fit over many thousands of data points. 
The value itself is an estimate as stated but the error should be small due 
to the method taken.

The irreproducibility of the reported results will come from the salt itself 
as many factors could affect the onset of reaction - such as the purity of 
the water used to prepare the samples, or whether the salt has been left 
overnight before the test runs. To evaluate the error or uncertainty in any 
value would be incredibly small, but reproducing the result could still show 
a discrepancy. None the less the constants derived in the model - which 
aren't derived from temperature readings - still approximate the reaction 
rate well.