tomclarke wrote:Reading the comments (some from Defkalion) it is not clear what they will allow. Static calorimetry is simpler in terms of equipment than flow calorimetry, but much more difficult in terms of analysis. So just working out whether a test of this type is safe or not requires great precision. If there were enough extrenal control over placing of thermocouples, and composition of outer layer (between the inner and outer thermocouples, it would be OK. There is no reason this whole part of the experiment should not be supplied externally.
Anyway, on many grounds, starting with Rossi, I view Defkalion as engaged in PR, not science. So I expect some tests which seem impressive but remain unsafe.
You could be right I suppose, but I would hope that any participants would be credible evaluators and would speak their mind regarding test deficiencies/issues (and results) after the tests rather than allow themselves to become PR schills for Defkalion. If they did not have these types of credentials, I would not care what they or Defkalion said about the test results anyway.
“Test results may be published by the independent evaluators in the media they choose and in our site, jointly.” - DGT
It seems like we should not have to wait too long to find out.
tomclarke wrote:The tests sound great till you see what they do. They have two black box reactors with thermocouples on the inside and outside. One will be active, one inactive. They measure temp difference, power in.
You gotta be kidding me.
If that's what they offer as an independent test to prover over unity than it means that (like Rossi) they still have nothing in hand to show.
If I have read the Defkalion test parameters (within which tests will be agreed) right:
Drill down to the parameters within which testing will be agreed. They will not allow flow calorimetry. This is what would (easily) have proved Rossi's reactors real or false had he ever adjusted the flow rate for an output decently above both input and ambient (say 50C) and decently below 100C to eliminate uncertainty due to phase change.
The tests sound great till you see what they do. They have two black box reactors with thermocouples on the inside and outside. One will be active, one inactive. They measure temp difference, power in.
This relies in the thermal conductivity of the two reactors being identical. But there is no procedure for testing this. Faking the setup as described would be very easy.
The method is OK if very carefully checked, but it is not intrinsically as bomb-proof as flow calorimetry. That BTW is why Rossi clearly did not have anything. His methods (from the first 10 tests) could very easily have been adjusted for bomb-proof results. And many people told him how to do this.
These tests are particularly unhelpful because the results will depend on the internal thermal characteristics of the two (sealed) reactors. It looks as though people will be asked to assume that the two reactors are identical - else why have the control reactor. But that is not externally verifiable.
This type of calorimetry is commonly used, but with careful calibration of test vessel thermal conductivity before and after. Looks like Defkalion are replacing this step (which would be much more difficult to fake) by the two reactors, which are trivial to fake.
Still, I will await what really happens in these tests. With no expectation.
Which is exactly what I understood and said earlier, after which Crawdaddy gave me flack. Well maybe he should be the one reading the test procedures. So thanks for confirming what I thought.
They will not allow flow calorimetry. This is what would (easily) have proved Rossi's reactors real or false had he ever adjusted the flow rate for an output decently above both input and ambient (say 50C) and decently below 100C to eliminate uncertainty due to phase change.
This remains one of my fundamental sticking points that pushes me to non-belief.
Such a simple thing to address and end the Rossiworld Circus.
There can be no reasonable explanation why any of them refuse doing this. A simple closed loop test where they heat somebody's swimming pool. Easy day.
The development of atomic power, though it could confer unimaginable blessings on mankind, is something that is dreaded by the owners of coal mines and oil wells. (Hazlitt)
What I want to do is to look up C. . . . I call him the Forgotten Man. (Sumner)
An advantage of Defkalion's test is that they remove much of the volume of apparatus that had skeptics claiming there was a hidden pool of molten salt (or whatever) supplying heat over the course of the experiment.
If they get COP > 20 for 48 + 48 hours without a change of fuel, it should be easy enough to establish that there is an excess of heat.
tomclarke wrote:
These tests are particularly unhelpful because the results will depend on the internal thermal characteristics of the two (sealed) reactors. It looks as though people will be asked to assume that the two reactors are identical - else why have the control reactor. But that is not externally verifiable.
Nope! This is from the protocol document:
Following a parallel test run of both Reactors for at least 48 hours, the two Reactors will be switched for a second run (Reactor #1 empty and Reactor #2 active) in order to authenticate the same results.
Different thermal characteristics, different results. Same characteristics, same results!
Carl White wrote:An advantage of Defkalion's test is that they remove much of the volume of apparatus that had skeptics claiming there was a hidden pool of molten salt (or whatever) supplying heat over the course of the experiment.
If they get COP > 20 for 48 + 48 hours without a change of fuel, it should be easy enough to establish that there is an excess of heat.
tomclarke wrote:
These tests are particularly unhelpful because the results will depend on the internal thermal characteristics of the two (sealed) reactors. It looks as though people will be asked to assume that the two reactors are identical - else why have the control reactor. But that is not externally verifiable.
Nope! This is from the protocol document:
Following a parallel test run of both Reactors for at least 48 hours, the two Reactors will be switched for a second run (Reactor #1 empty and Reactor #2 active) in order to authenticate the same results.
Different thermal characteristics, different results. Same characteristics, same results!
I know. But they say they will be swapping powder etc, so we do not know whether the thermal characteristics between the thermocouples will be affected. And we suspect that the powder switching will not be open to inspection.
If the entire outer reactor, including thermocouples, is open to inspection at all times it should be fine. But will this happen?
This argument boils down to arguing that the CF experimenters have no idea how to do calorimetry. A much more authoritative figure than deadalus2u disagrees. I trust the opinion of Robert Duncan much more than some guy on the net.
Don't trust anyone, go look up the source material yourself. It is not difficult to compare good with bad experiments: if you do the errors and ommissions in the bad ones become obvious.
Ummm have you read the original Mizuno conference presentation? Please link it, I searched our university database and it does not appear to be available.
I do look up source material myself. I have been encouraging others to do the same for months. The evidence is compelling in the literature. There are dozens of papers with infinitely better apparatus, procedures, and measurements than those in the unpublished report you linked. See the Storms 2010 review for a leading reference.
I would be interested to read the Mizuno conference report (my databases do not have it available. I cannot judge how "well done" the replication is without the original research report.
Carl White wrote:An advantage of Defkalion's test is that they remove much of the volume of apparatus that had skeptics claiming there was a hidden pool of molten salt (or whatever) supplying heat over the course of the experiment.
If they get COP > 20 for 48 + 48 hours without a change of fuel, it should be easy enough to establish that there is an excess of heat.
tomclarke wrote:
These tests are particularly unhelpful because the results will depend on the internal thermal characteristics of the two (sealed) reactors. It looks as though people will be asked to assume that the two reactors are identical - else why have the control reactor. But that is not externally verifiable.
Nope! This is from the protocol document:
Following a parallel test run of both Reactors for at least 48 hours, the two Reactors will be switched for a second run (Reactor #1 empty and Reactor #2 active) in order to authenticate the same results.
Different thermal characteristics, different results. Same characteristics, same results!
I know. But they say they will be swapping powder etc, so we do not know whether the thermal characteristics between the thermocouples will be affected. And we suspect that the powder switching will not be open to inspection.
If the entire outer reactor, including thermocouples, is open to inspection at all times it should be fine. But will this happen?
I doubt!
So you agree that if the independent testers are allowed to monitor the surface of the calorimeter accurately and continuously (an obvious necessity) and to perform control experiments with heating elements of known power, then this experiment will be an example of "well done" calorimetry.
This argument boils down to arguing that the CF experimenters have no idea how to do calorimetry. A much more authoritative figure than deadalus2u disagrees. I trust the opinion of Robert Duncan much more than some guy on the net.
Don't trust anyone, go look up the source material yourself. It is not difficult to compare good with bad experiments: if you do the errors and ommissions in the bad ones become obvious.
Ummm have you read the original Mizuno conference presentation? Please link it, I searched our university database and it does not appear to be available.
I do look up source material myself. I have been encouraging others to do the same for months. The evidence is compelling in the literature. There are dozens of papers with infinitely better apparatus, procedures, and measurements than those in the unpublished report you linked. See the Storms 2010 review for a leading reference.
I would be interested to read the Mizuno conference report (my databases do not have it available. I cannot judge how "well done" the replication is without the original research report.
Carl White wrote:An advantage of Defkalion's test is that they remove much of the volume of apparatus that had skeptics claiming there was a hidden pool of molten salt (or whatever) supplying heat over the course of the experiment.
If they get COP > 20 for 48 + 48 hours without a change of fuel, it should be easy enough to establish that there is an excess of heat.
Nope! This is from the protocol document:
Different thermal characteristics, different results. Same characteristics, same results!
I know. But they say they will be swapping powder etc, so we do not know whether the thermal characteristics between the thermocouples will be affected. And we suspect that the powder switching will not be open to inspection.
If the entire outer reactor, including thermocouples, is open to inspection at all times it should be fine. But will this happen?
I doubt!
So you agree that if the independent testers are allowed to monitor the surface of the calorimeter accurately and continuously (an obvious necessity) and to perform control experiments with heating elements of known power, then this experiment will be an example of "well done" calorimetry.
Iit is much more complex than that. It would need to be known that the two reactors were (and stayed) with identical thermal characteristics between the thermocouples, which means internal access before and after the switchover. I rather suspect no internal access.
Otherwise, you could detect Q=20 reliably based only on external heat loss temp coeff estimates, which if reactors could be examined would be good enough. It would not be very reliable, so it depends on the measurements, but Q=20+ should be clear.
For this to be safe it needs to be certain what is total the input power to the inactive reactor, as well as the active reactor. Not just an upper limit but a precise measurement. Since power could be stored in a battery you would need a long enough run.
tomclarke wrote:
I know. But they say they will be swapping powder etc, so we do not know whether the thermal characteristics between the thermocouples will be affected. And we suspect that the powder switching will not be open to inspection.
If the entire outer reactor, including thermocouples, is open to inspection at all times it should be fine. But will this happen?
I doubt!
So you agree that if the independent testers are allowed to monitor the surface of the calorimeter accurately and continuously (an obvious necessity) and to perform control experiments with heating elements of known power, then this experiment will be an example of "well done" calorimetry.
Iit is much more complex than that. It would need to be known that the two reactors were (and stayed) with identical thermal characteristics between the thermocouples, which means internal access before and after the switchover. I rather suspect no internal access.
Otherwise, you could detect Q=20 reliably based only on external heat loss temp coeff estimates, which if reactors could be examined would be good enough. It would not be very reliable, so it depends on the measurements, but Q=20+ should be clear.
For this to be safe it needs to be certain what is total the input power to the inactive reactor, as well as the active reactor. Not just an upper limit but a precise measurement. Since power could be stored in a battery you would need a long enough run.
Sorry, I don't understand. Standard practice with this type of calorimetry is to run a control experiment to determine the temperature of the calorimeter at a given input power.
How could the internal thermal characteristics of the reactors make a difference to the measurements if control experiments determine the external temperature vs input power dependence?
How could the internal thermal characteristics of the reactors make a difference to the measurements if control experiments determine the external temperature vs input power dependence?
Because they could do something so the "control unit" looks much worse in the readings than an "empty reactor" would, thus manipulating the resiults.
One possibility I could see to make this a little more interesting would be a double blind test. So an independent 3rd party, inserts the "fuel rod" into any one of the two units (provided this is easy enough to do) and neither the testers, nor the Defkalion people know which one it is. Then the measurements are performed. That would rule out all sorts of means of error. Switching the fuel rod over to the control unit half way through the experiment would also help, IMHO. At least then some means of manipulation could be ruled out. Still not all of them, of course.
Crawdaddy wrote:Ummm have you read the original Mizuno conference presentation? Please link it, I searched our university database and it does not appear to be available.
So, Rossi has dead already? Or not yet?
What about Dr. Duncan? Where he speaks about good calorimetry done by Rossi?
Or I am wrong and this thread not about Rossi?
@Tomclarck
What complicate scheme are you discussing? Thermocouples, two reactors instead one, etc.
In "5 kW" Rossi had 1 water pump with mass flow about 2 g/s, one thermometer only in input pipe, delta T as I remember 70 deg till 100 deg and there was assumed that all input water converts in steam.
Where has Rossi shown another calorimetry scheme?
My 14 years old son already can solve such exercises. And little more complicated exercises too.
For example how Rossi can make fool trustful admirers.
How could the internal thermal characteristics of the reactors make a difference to the measurements if control experiments determine the external temperature vs input power dependence?
Because they could do something so the "control unit" looks much worse in the readings than an "empty reactor" would, thus manipulating the resiults.
One possibility I could see to make this a little more interesting would be a double blind test. So an independent 3rd party, inserts the "fuel rod" into any one of the two units (provided this is easy enough to do) and neither the testers, nor the Defkalion people know which one it is. Then the measurements are performed. That would rule out all sorts of means of error. Switching the fuel rod over to the control unit half way through the experiment would also help, IMHO. At least then some means of manipulation could be ruled out. Still not all of them, of course.
This type of calorimetry must be conducted using controls. Prior to the experiment the calorimeter must be calibrated with a known input power, there is no oportunity to "do something" to the other reactor core if standard practices are used. The most basic of the "mutually agreed upon" experimental protocols would obviously be calibrating the calorimeter. No competent person would ever agree to put there name on such a measurement without a series of control measurements, which are standard practice in this type of calorimetry.
This type of calorimetry must be conducted using controls. Prior to the experiment the calorimeter must be calibrated with a known input power, there is no oportunity to "do something" to the other reactor core if standard practices are used. The most basic of the "mutually agreed upon" experimental protocols would obviously be calibrating the calorimeter. No competent person would ever agree to put there name on such a measurement without a series of control measurements, which are standard practice in this type of calorimetry.
And how would calibrating the calorimeter prevent tinkering with the control reactor? Unless you can see what is inside, you cant know what is happening in there...