At 01:57 PM 8/4/2011, Jed Rothwell wrote:
I wrote:
Except that the data recorded in a boiler test is EXACTLY what you
see here, for crying out loud!
To be more specific, I mean that they record only one value for
temperature and one for the flow, even though these values probably
fluctuated measurably over the course of the test.
The boiler test might call for the inspector to record temperature
and flow every minute for 10 minutes. Or it might demand a graph of
automatically recorded data, or a high/low value. It does not; it
asks only for a single value for each parameter. This reduces
accuracy somewhat.
This argument is preposterous. It's true, which makes it no less
preposterous. A boiler is a very well-understood device, with
centuries of experience behind it. There are no questions about the
operation of a boiler; rather there are some parameters which have
been found to be indicative of proper function, and possible
problems, with the boiler. Long experience has shown that if there is
a problem, these simple measurements will show it.
Imagining that testing an experimental device with unknown operating
characteristics can be handled with a boiler form is what's preposterous.
Yes, if all the procedures for taking the measurements have been
designed by experts, and if fraud is not suspected, sure, it could be done.
