Michele Comitini wrote:
What about something like "Do e-cat's people know how to make correct experiments ? "
My answer: Yes, they know how to do correct experiments, but they do not know how to write papers in English describing these experiments. Or if they know how to write such papers, they have not taken the trouble. This is an important experiment, so they should take the time.
I have no problem with the procedures or choice of instruments. Only the presentation is substandard. If I had been there during the 18-hour test, observing the procedures they did, I would have:
1. Collected data every 5 minutes manually into a lab notebook: flow rate, flow meter odometer reading, inlet, outlet temperature, and power meter instantaneous and total kWh.
2. I would have done some manual checks with my own instruments to confirm their instruments, such as collecting the flow for 20 seconds in bucket and confirming that it is weighs about 20 kg.
3. From that I would have published a paper that sensible people would find convincing. The extreme skeptics would not believe it. People who imagine that conventional flow calorimetry might be wrong by a factor of 1,000 for no apparent reason would not believe it. But any HVAC engineer would know it is correct.
They could have done this, even without a computer or video cameras, using the laboratory techniques I learned in grade school in the 1960s. I mean techniques such as keeping a lab notebook and writing down the date, time and instrument values, and then setting a kitchen timer to remind yourself to write them down again in 5 minutes (or 10 minutes after a while). This is not rocket science. I am sure this is how Levi was trained to do experiments. Everyone his age and mine, who grew up and went to college before we had computers, learned to do it this way.
If they have a lab notebook, I hope I can persuade them to publish more data from it. They deserve to be criticized for holding back this kind of data. It is unprofessional. An inspector doing a boiler test is justified in writing down only one value. It does not even need to be a computed average or a thermocouple min/max reading. It can be "what the thermometer showed most of the time." One value is enough to prove the point. But it is customary for a scientist to provide more data.
They could have done this back then, and they darn well can do it now. What I outlined here is exactly what I proposed to do -- to augment & confirm their instruments with conventional, old-fashioned manual techniques, circa 1970. I spelled it out in more detail than I have here. Rossi said he did not want me to do this. He invited Krivit instead, and he showed him a set of procedures that are interesting, educational, and that an astute observer might learn a lot from. But these procedures prove nothing, for reasons I spelled out when Rossi described them to me.
The quick test that Rossi did for Krivit does have a legitimate use. If you already know the machine is real, and it works, that test is a convenient way to check how it is performing today, or with a new configuration. It is a way to do a quick experimental iteration. I expect this is why Rossi does it. It is similar to the way a programmer who is debugging a large program runs a quick utility to simulate a situation and force an intermediate program state, rather than running the whole thing from the beginning. The programmer sets up a "fake" situation, skipping over many steps, because the goal is to test a subset of the whole, and the programmer knows that the overall program exists and works, so there is no need to prove it. An observer from outside will not be convinced the overall program works, but the programmer is not trying to prove that. She is trying to tackle a specific problem that day in the shortest, most convenient fashion she can.
- Jed
