On LENR Forum, there is a thread on Shanahan’s critique of cold fusion experiments, and this post appeared by THHuxleynew:
I’ll give his last comment first:
PS – I don’t make these arguments often here, since I feel they are perhaps known by those interested in them, and strongly disliked by others. So I will not continue this argument unless new facts are added to make it worthwhile.
In fact, THH addresses an issue that I have never before seen raised. It is of limited impact, but it proposes a possible artifact that could afflict some experiments, that should probably be explicitly ruled out (or confirmed!)
Jed was arguing something familiar, common, and … incorrect, and THH nails that.
It may help to look closely at the strands of argument here:
I agree with Jed, he makes a number of true statements, but his point does not address mine. My point was that measuring salt balance does not determine the amount of entrainment, because entrained liquid can be either condensed (no salt) or non-evaporated (with salt). Condensation does not change the heat balance. But entrainment, in an open cell as we discussed here, does. Jed trying to argue that F&P can know there is no entrainment (and therefore no resulting change in heat balance) by measuring salt content. This is false.
He is correct. We will explain. Measuring the salt assumes that entrained liquid is unvaporized electrolyte, and the electrolyte is salty. However, that is not the only possibility!
Jed: Condensation is exothermic, so the heat lost to boiling is added back into the cell by condensation. You can test this by measuring the heat of vaporization in a cell with some condensation. It does not change from the textbook value. The null experiments by F&P all had condensation and they all produced the textbook value.
Jed is completely correct that condensation does not change the heat balance. However, this is missing THH’s point. The problem is not condensation alone, but condensation followed by entrainment of the condensed vapor (which would have no salt in it). The PF cell has a long, thin tube as a vent. if the vent is at a lower temperature than the cell interior, I would expect condensation to take place within it (heating it up).
This requires more than a little care to examine! It does seem possible that condensate (salt-free) could then be blown out of the cell. It would be, in boil-off cells close to the boiling point and would then evaporate outside the cell as it hits the unsaturated air. This water was not expelled as vapor, though, as it left the calorimetric envelope. If it is treated as having been vaporized, the heat of vaporization would then incorrectly enter the calculations.
And this cannot be ruled out by measuring the remaining salt. That would apply to “splash,” i.e., perhaps boiling or bubbling electrolyte that tosses it into the head space and then flow carries it out. The cell design militates against this as to any major quantity, but condensed electroyte might well be preferentially expelled. The devil is in the details.
The problem is that such results can be over-generalised. They only apply when conditions remain the same. The entrainment issue applies to unusual boil-off conditions. By definition the control, which does not have such extreme boil-off, will have different conditions, in a way likely to alter this result.
THH’s argument gets a bit iffy here. If the control is lacking an “extreme boil-off,” why? The point of the PF “simplicity” was that the boil-off time would be the experimental result. The loss of unevaporated water would indeed decrease the boil-off time, but only as an additional effect. That the boil-off is more rapid is a result, not a set condition. Presumably the conditions were set so that without XP, the boil-off times would be the same.
Jed: In a closed boiling cell with 100% condensation, the heat balance from vaporization is always zero. There is no heat lost to vaporization, because no vapor escapes.
I agree – but this is not relevant to the matter at hand which is discussion of F&P open cells in boil-off phase.
Both seem correct.
Jed: You are wrong about the salts,
I don’t believe you have shown that?
Jed often argues from conclusions based on evidence outside the argument. This then creates sprawling disagreements that never resolve. In this case, THH’s original point is very simple: the salt measurement does not definitely rule out liquid entrainment, liquid leaving the cell while unevaporated.
Jed: and you ignore the fact that they did several other tests to ensure there was no entrainment.
This is an offensive “you ignore” argument, common with trolls. Jed is not a troll, but … he’s not careful. He is very knowledgeable but has stated many times he doesn’t care about communicating clearly with skeptics. It’s unfortunate. Jed has paid his dues, to be sure, doing an incredible level of work to maintain the lenr-canr.org library (and he has been personally supportive to me in many ways). But we should keep him away from outreach to the mainstream! — Unless he is willing to develop better communication skills, dealing with genuine skeptics, and here, THH certainly resembles a genuine skeptic.
No – I point out that it is not possible to know which tests are done on which experiments, and note the danger of over-generalising results. That is addressing this fact, not ignoring it.
He is correct, and there are such dangers.
Jed: It would make no difference whether they did each of these tests every time: once every 10 tests would be fine. Note that they ran hundreds of cells, 16 at a time.
Only if the one in 10 included the (1 in 10 – I’m not sure?) cells that showed this special boil-off. We don’t know this.
This problem is addressed with random sampling and controls. I am not claiming Jed is wrong, only that his arguments are far less conclusive than he makes them out to be. Jed was correct, it is not necessary to verify every instance, but in doing that one would need to look out for possible sample bias.
THH is correct to at least suspect that rapid-boil-off cells would be more likely to entrain condensed water, which would again shorten the boil-off time. Obviously, one would want to see tests for expelled liquid, though that isn’t necessarily easy. I think measuring the heat of condensation on an external trap might be necessary. I’ve seen no descriptions of this.
THH is also not paying attention to the primary phenomenon, the rapid boil-off, treating it as an experimental condition, rather than a result. If there is the rapid boiloff changing the cell conditions, yes, entrained water could cause calorimetry error, but Pons and Fleischmann were not depending on the calorimetry at that point. The possible level of error could be estimated, and it is limited to the correction made to heat measurement for vaporized water. Looking at cell conditions, one could estimate the range of possible values.
Jed: They also tested closed boiling cells where the heat of vaporization plays no role (as I just said), and these cells also showed excess heat.
This kind of thinking fries my brain. Jed is arguing for the correctness of a conclusion (real heat, not artifact), which is the opposite of scientific process. There can be different artifacts in different experiments. What is needed is something that can be measured across all experiments, or at least most of them. We have that.
The heat/helium ratio. I remember when I started proposing measuring that with increased precision, there were arguments within the field that this was unnecessary, we already knew that helium was the ash.
However, if there is a single phenomenon that produces both heat and helium, in a consistent way, i.e., with a constant ratio, within experimental error, each measurement validates the other, again within the error bars. Ideally, helium should be measured in every D20 cold fusion experiment. At this point it’s too expensive, but that could change. It would kill all these arguments about various possible artifacts. If the heat/helium ratio holds in the experiment, the calorimetry was almost certainly correct, in spite of all the i’s not being dotted and the t’s crossed.
It has been pointed out that there is no end of possible artifacts, which is why the “they must be making some mistake” argument is so offensive. It’s pseudoscientific, proposing theory as creating a conviction of error. That makes sense when one must make some quick decision, but it makes no sense when one is examining experimental results to see if there are possible reasons to reconsider one’s beliefs.
(Cold fusion is not actually theoretically impossible, the arguments all require assuming a specific reaction and then calulating the rate for that reaction, which completely fails to be relevant if that is not the reaction.)
When Storms’ 2010 review was published in Naturwissenschafter, I winced when I saw the abstract: “reaction between two deuterons to make helium.” That was Storms opinion (generally rejecting multibody reactions, largely out of ignorance of the possibilities, and then thinking of two nuclei coming together, though, in fact, his theory is multibody, merely in a different way. It is not the simple two-body reaction that the abstract suggested.
That would be a different paper, with results and conditions we would need to look at afresh. Shanahan’s affect might be relevant here, or something else. Or perhaps this other sustem would be solid evidence. We would need to consider it. Either way, it does not change the arguments here relating to F&P open cell results.
Jed: Unless you have a scientific reason to believe there was entrainment, you should stop beating that dead horse. You have not given a single reason other than “maybe” “I suppose” “we can imagine” or “some scientists think they may eventually find a reason.”
This was way off. The explanation of how entrainment was ruled out was simply wrong. Testing the remaining salt does not show lack of liquid entrainment. I’m sure Jed can understand this, so, why not simply recognize that this particular argument has not — so far — been addressed.
That is where we disagree about the nature of skepticism. F&P posit some new effect (LENR) to explain anomalous results. It is they who must show there is no plausible mundane explanation – as they try to do – not others who must prove such an explanation.
Nevertheless, THH here takes on standard pseudoskeptical cant. “It is they who must show.” Must according to what? Someone can assert some evidence for something new, and can show evidence that they think supports it. There is no “must.” Both skeptics and believers fall into this trap. They become demanding, attached to a position, and the position of “wrong until proven true” or close equivalents, is pseudoskeptical. The moral imperative “must” deludes us. People need freedom to change their minds, we resist attempts to force us to accept based on coercive arguments. THH has the complete right to be skeptical, which is properly an agnostic position. He isn’t convinced yet, and he is the world’s foremost authority on whether or not he is convinced. Jed has the complete right to believe or accept whatever he wants … and to disbelieve skeptical arguments until and unless he is convinced.
The problem arises when one party or side attempts to claim the other is “wrong.” “Wrong” — like “Right” — is a complex judgment that does not exist in reality, and that gets into deeper ontology. The naive will think my statement preposterous!
Jed: Oh, and “condensation in a cell changes the heat of vaporization.” No, it doesn’t. Try it.
If THH said that, he misspoke. But I don’t think he said it. Rather this was Jed’s interpretation, and if so, the use of quotation marks was an error.
Condensation in the cell, as above, can affect open cell experiments by allowing entrainment not discovered from salt balance check.
This does not change the “heat of vaporization, which is a constant for a particular liquid. Rather it changes the correction made for vaporization, if and only if the liquid actually leaves the cell as a liquid, condensed, instead of as vapor. One would need to look at a particular experiment to see if this is relevant. I don’t think THH explained the problem well enough, I can see Jed continuing to think that it is the condensation that matters, and thus that THH is wrong wrong wrong. But that is not what THH is talking about. He is talking about the possibility of water leaving the cell as liquid instead of as vapor, having first been condensed inside and only then blown out. Thus the amount of water leaving the cell unvaporized would not be determined by measuring salt loss.
I don’t have the experience to say much more about this, about how much of an effect this might be. But I agree with THH on the primary issue, and it seems clear enough. Against this would only be argument from authority (they were experts and could not possibly make such a stupid mistake). Or other arguments that depend on there being a single effect without having actually shown that.
For closed cells we have other issues, and specifically, unless the calorimetry calibration is known independent of cell temperature distribution, ATER/CCS. But it does not help to mix up different cases – open and closed.
Each approach must be evaluated separately. Because of problems with confirmation bias and the file drawer effect, there are many problems in interpreting cold fusion experimental results. I remain satisfied as to the reality of the effect by the heat/helium reports, which actually point to a testable hypothesis, which has been confirmed by many, even though there is also room for improving the work, increasing precision, etc.
This is much more definitive than a pile of anecdotes, using varying experimental methods, showing heat but without being able to predict it. The multiplicity of excess heat reports is evidence, all right, but circumstantial. The correlation of conditions with results (such as loading ratio with heat) is supportive, but also subject to other possible interpretations. Heat/helium, by comparison, ices it.
We are discussing F&P’s open cell results. I’m not going to address directly here the question of whether condensation in the cell can ever affect the heat balance (by indirect means), it is not what I’m arguing now. Given more space we could however consider it. I’ve never stated or implied that condensation changes the heat of vaporisation.
THH is the clearest, best, and most civil of all the skeptics I have encountered in about eight years of discussing cold fusion. He, and people like him, are important to the progress of cold fusion, more important than “believers,” unless the latter are scientists practicing real science, where the goal is to prove oneself wrong. (I.e., that the hypothesis fails to predict results). Those have paid their dues, and it is actually their work that is of ultimate importance, not their conclusions as such.
Author: Abd ulRahman Lomax
See http://coldfusioncommunity.net/biography-abd-ul-rahman-lomax/ View all posts by Abd ulRahman Lomax
27 thoughts on “A new argument on evap calorimetry”
In the scenarios you and THH describe, the calibration would show false excess heat. It would not agree with the textbook heat of vaporization. As I described here, they did many calibrations with non-working cells, such as Pt-H, and in these tests they always measured the expected amount of heat.
In other words, the test to measure the salt remaining in the cell was not the only thing they did to ensure there was no problem. If the salt inventory had been the one and only method they used, you might have a point. Here’s the thing: THH wants to look at one aspect of these experiments at a time, raising one doubt in isolation, and then going to the next aspect while ignoring the first one. Yes, the salt inventory might have problem. Yes, the calibration tests with no heat might have problem. Yes, the cell geometry might be a problem. But each and every one of these things is addressed in the paper, and taken as a group of factors, mutually supporting, they make it extremely unlikely there is a problem.
Jed, you are missing the point. You don’t like a possible conclusion from the point, so you find *another* reason to reject it. However, the point made was narrow and correct. Measurement of the remaining amount of the salt does not rule out loss of fluid by entrainment. That is not complicated. The problem is not necessarily simple to address. Calibrations are one way, but then we would need to look at what was different in the calibrations. The full issue of possible artifacts is huge, and finding agreement difficult if every point is defended even when the reality of it is clear. There is a rather obvious way that the electrolyte could evaporate, condense within the cell where it is cooler, and then, this being near the vent, be blown out. No salt loss, but loss of water.
This problem would only appear significantly when the cell was actually boiling, I’d expect. I’d want to look at the calibration conditions, carefully.
The only point that I commented on in disagreement with you was in the claim that measurement of the salt remaining ruled out loss of fluid without evaporation. That is simply inadequate.
I had never before considered the possibility THH brought up. I had often thought that electrolyte loss would have shown an accumulation of salts outside, and I wrote this. But the device as constructed, as far as I can see, could operate as a still, which then blows the purified liquid out, certainly if it condenses within the vent. The electrolytic gases, hydrogen/deuterium would then create a flow to move that condensed water out.
That some other test or observation might rule this out is a completely separate point. If this point is clearly addressed in the paper, point it out! We have the Morrison-Fleischmann debate under study here. This issue is addressed in Stage Four Calculation in the Fleischmann reply to Morrison.
Pons and Fleischmann give two arguments about the issue of “whether parts of the cell contents may have been expelled as droplets during the later stages of intense heating” as raised by Morrison. The first is the “titration of cell contents,” which is the argument shown to be inadequate. The second is not so clear, but does show a clear problem: it depends on the behavior of cells “temperatures in excess of 90 degrees C ” But the conditions involved are at the boiling point, actually, probably, slightly higher.
This is actually similar to another problem that has been discussed, the difficulties of evap calorimetry if an effluent may be wet steam. It is not easy to measure steam quality, which is the actual issue here. If there is an error here, and there may well be as it appears to me now, it would affect the COP measured by Pons and Fleischmann, to the extent that it depends on estimates that assume full evaporation. I doubt that this would reduce the COP to 1.0, but … it’s a problem to be addresed. The HAD experiments were used to claim some of the highest COPs reported for the FP Heat Effect. Maybe they were so because of a problem that appeared at the boiling point.
In general, I was, with others, troubled by the sloppy conditions in a boiling cell. I think this requires far more care than the knee-jerk rejection by Morrison or the incomplete arguments of Pons and Fleischmann. The question I asked in starting the study of the debate was what is actually shown. This is not resolved by claiming that some problem is “extremely unlikely,” unless there is a specific analysis of each possible issue.
If we come up with issues here that cannot be resolved, we can cover the remaining dispute and take the matter to the CMNS community for more advice, for expert commentary.
Some of the issues we find in detailed discussion have never before been addressed. We can clean that up, or at least outline any remaining dispute. However, Jed, I so no argument there that titrating the remaining contents is adequate to determine whether or not there was entrained water; it was obviously based on an assumption that this water would be the electrolyte, as droplets thrown up from boiling action; the cell design does indeed make that unlikely. Then the *next issue* is whether or not calibration in other experiments under other conditions would be adequate to show that the condensed-the-expelled mechanism was not occurring. So far, on that, my sense is that it is not, as well.
This is one reason why I turn to helium measurements. Pons and Fleischmann rejected them as too expensive. That was penny-wise and pound-foolish, and I suspect that in saying that, there was another consideration that they did not want to state: the helium results they had (and did not publish, in spite of a promise, leading to Park going ballistic) contrdicted their idea of a bulk reaction, and they “did not want to fight on two fronts.” Pons and Fleischmann got caught up in defense, which damages scientific integrity and neutrality. It happens on all sides, and we need to be aware of this, present company included.
Welcome to Cold Fusion, the Scientific Fiasco of the Century,
“if it was running as expected, it wouldn’t need a special note”
For me, I am deeply skeptical of expectations. So for LENR results, where positive results mean something interesting and not predicted by conventional theory, I reckon expectations cannot be relied upon. In fact they cannot even be relied upon, however when results are unsurprising there are a greater chance that all of the contingent expectations are correct, though equally perhaps in this case the results are uninteresting.
So however eminent Fleischmann, or anyone else, surprising results that are not explicitly so do not for me make clear evidence. They are interesting, and merit further work to get clear evidence for quite a while. Whether that work has been done enough in this case is an issue.
You wrote: “So for LENR results, where positive results mean something interesting and not predicted by conventional theory, I reckon expectations cannot be relied upon. In fact they cannot even be relied upon, however when results are unsurprising there are a greater chance that all of the contingent expectations are correct, though equally perhaps in this case the results are uninteresting.”
I strong disagree. That depends on which conventional theory we are talking about. This is not a conflict between theory and experiment, or between Fleischmann and nuclear physicists. It is a conflict between old, well-established theories and some modern nuclear theory.
Cold fusion violates some conventional nuclear theory, but according to experts such as Schwinger, that theory is less well established than people realize. It covers only nuclear plasma, not the solid state.
In order for cold fusion to be wrong, older, more well-established conventional theories would have to be wrong, along with many conventional experiments. Essentially, everything going back to Lavoisier’s calorimetry of 1780 would have be thrown out, along with a lot of chemical theory, the work of Faraday, and the Second Law of Thermodynamics. This is far less probable than the likelihood that a new path to nuclear fusion has been discovered in the solid state.
(Lavoisier’s ice calorimeter measured the metabolic heat from guinea pigs, which is ~1.4 W, and CO2 production, for 10 hours. This demonstrated that the ratio of heat to CO2 was the same as combustion. These measurements had enough precision and confidence to confirm the heat from many cold fusion experiments.)
THH said:That is where we disagree about the nature of skepticism. F&P posit some new effect (LENR) to explain anomalous results. It is they who must show there is no plausible mundane explanation – as they try to do – not others who must prove such an explanation.
Nevertheless, THH here takes on standard pseudoskeptical cant. “It is they who must show.” Must according to what? Someone can assert some evidence for something new, and can show evidence that they think supports it. There is no “must.” Both skeptics and believers fall into this trap. They become demanding, attached to a position, and the position of “wrong until proven true” or close equivalents, is pseudoskeptical. The moral imperative “must” deludes us. People need freedom to change their minds, we resist attempts to force us to accept based on coercive arguments.
Abd has made this point about my use of must several times here. I agree with him, as a matter of communication such coercive language is unhelpful.
I tend to use it when (correctly or otherwise) I view the must as a logical connective, that all can agree. Abd for example uses must here when my phraseology was of choice: Each approach must be evaluated separately. Because of problems with confirmation bias and the file drawer effect, there are many problems in interpreting cold fusion experimental results. . Again for something that he thinks (and I concur) all can agree.
Abd and I often tend to argue along similar lines, and pat each other on the back because we understand each other argument. Both of us value clarity and the avoidance of emotional reactivity in debate – whilst acknowledging that we as humans are prone (ourselves) to have such reactivity. For me, that understanding is important because without it I’d be much less likely to correct my own fixed responses, and less open to that correction when it is pointed out by others. And clarity in debate is something that I positively value, often above the instrumental results of the debate.
In the LENR debate I have an advantage over those working in the field which is that I am less involved, and therefore less likely to have emotional reactions that I can’t see round. Less likely of course does not mean I avoid this, and I’m not making any specific point about others who though they work in the field may have excellent strategies for similarly avoiding the expression of emotional reactions in intellectual debate.
We see in the UK and the Brexit debate – also perhaps – though I’ve less information the US Trump debate political arenas where emotional reactions are mostly what matters. A dispassionate view of the matter, however correct, will not sway the political argument and I find myself, deeply saddened that political decisions should be made in such a passionate atmosphere where pragmatic and careful analysis of facts is understandably secondary to grand emotional gestures. I’m a fan of Walter Baghehot: http://www.bbc.co.uk/programmes/b09jdc7q
Right. If I point out a problem with the use of “must,” it does not mean that I don’t do it myself. “Must” can be a simple expression of logical consequence, an expression of our own thinking. Where it gets iffy is where it is describing a moral imperative, as if someone following a differing line of approach *must* be “wrong.” This is done more subtly with claims like “extraordinary claims require extraordinary evidence.” They obviously don’t. People make extraordinary claims all the time with no evidence! Or with weak evidence! What do we have with an extraordinary claim with *ordinary evidence.* We have a claim that I think is wrong. But what is *my* evideence, and just how extraordinary must it be, and who decides what is ordinary and what is extraordinary.
Ah. Here is how: if I don’t ordinarily think of it, it is extraordinary, whereas what I think is ordinary and requires no strong evidence, or even weak evidence, for that matter, since I am a logical and scientific person who would not beleive anything without evidence. My shit don’t stink, but yours does.
There is a study, the sociology of science, that looks at all this. Scientists are humans and “scientific thinking,” even for scientists, isn’t particularly normal, and people use the claim that others are not practicing the scientific method to reject the ideas of others, while not applying it to their own ideas. Very human!
SD: In order to be able to accept Shanahan’s ideas of calorimetry errors being the sum of the effects seen, I need to have a way in which those errors would correlate with the Helium measurements. I don’t see a mechanism that would do that. Those errors are certainly possible, but why would they result in Helium being sucked into the system in proportion to the magnitude of the error?
I agree that He correlation would be something else again, which is why I await new He correlation experiments with interest. The correlation data that exists is complex and does not convince me because I have not been able to work through all that complexity to find a clear non-artifactual correlation. In answer to SD’s question here I would say that such a mechanism could be indirect: property Z of the experiment correlates with artifactual excess heat and also with artifactual He. While I cannot either think of Z (off hand) given specific experiments and time I might well be able to so so. To rule out such a correlation would require a lot of analysis considering a wide range of artifacts.
There is a common theme here. Suppose there are artifacts A,B,C, and a whole set of experiments E. e is made up of multiple series of experiments using different equipment or conditions. Within each series we have comparability, and can note correlations. Across series, where different series may suffer different artifacts, it is difficult to be sure of correlations. It is also difficult to rule out artifacts on specific series from meta-analysis of the entire corpus E. I’m not saying it can’t be done, but the argument required is subtle and logically needs to consider all possible combination of artifacts on series. thus for example, maybe artifact A can be ruled out of series (ei) and artifact B ruled out of series (eb). I would not want to generalise those results.
If this sounds complicated it is because it is complicated. Meta-analyses are inevitably that and can generate subtle errors.
As I said above, I look forward to a new clearer set of He/excess heat results where correlations can be better analysed.
Tom – at times I think this debate is getting too complex when we’re talking about meta-analysis of multiple experiments. Too many variables, and it’s far too easy to misinterpret some offhand phrase. It’s maybe best to concentrate on the Miles experiment only and make things simpler.
In the Miles experiment, is it possible to put forward any logical combination of non-nuclear effects that would give the heat/helium correlation he ended up with? IIRC the flasks were simply numbered and sent off to be analysed for Helium, so the analysts didn’t know whether they came from a cell that produced heat (or how much) or whether it didn’t, or whether it was indeed a dummy slipped in for calibration purposes (add in a known number of atoms of Helium and see what was measured). Using a weak alpha source for a known time could produce that calibration flask if needed, after all – the analyst didn’t know what was supposed to be in each flask.
I can’t think of any such combination of effects that would produce that correlation unless we accuse Miles of fabrication of data. In order to invalidate Miles we need to specify either a logical sequence of experimental errors or say he fudged it. I don’t see a middle way there.
As regards notes on unusual things, when I’m running a test I know precisely what I should be seeing, and so notice anything that shows that is not what I expect. I was successful at Failure Analysis through doing exactly that, though that was in electronics and not in LENR. Wet spots where they shouldn’t be is definitely something I’d notice, so I’d expect both F+P and Miles to notice that. How could they miss it? I see your point that such observations (“it wasn’t wet”) are not in the experimental details, but I’d only expect to see a note that “it was wet, and we needed to fix that, so here’s the next design we used where it didn’t get wet”.
Still (maybe again through the FA experience) I do tend to concentrate on what goes wrong, since it was my job to make sure it didn’t continue to go wrong. I also see this as where your attitude and mine differ, since my reports only needed to specify what needed to be changed to fix the problem whereas yours seem to require extreme detail since you don’t know what someone will ask questions about years later. For me, the reduced failures at the end of the line and in the field were proof enough to the management that I had been correct.
Here, therefore, my viewpoint is that a competent experimenter would have noticed the entrainment if it had occurred. It would be an unexpected and unwanted observation, and for obvious reasons they’d have fixed it. I suppose the way we could find out would be to run cells of the exact design Miles did (since it was that experiment where the correlation is seen) and see if any produce water or electrolyte coming out where it shouldn’t. Is it normally OK but you get droplets out under high gas flow (but then again, I’d expect that to be written up in the experiment if it happened)?
If something anomalous was visible, then I’d expect it to be seen and noted. It could be that I’m more aware of things happening (that I didn’t predict) than most, but then that does also seem to be a trait of scientists in general. Maybe especially in the case of people who’d run a year-long experiment that was against theory. Looking at the lengths Miles went to to get good data, I really can’t see him missing a wet spot.
On the Brexit debate, it seems some dispassionate analysis would be useful. It’s pretty obvious that one aim of the EU negotiator has to be to make it less advantageous to the UK being out (since otherwise some other countries may decide to leave as well), but that since on trade alone the UK will actually be better off then this requires some severe conditions being laid on the UK. It’s also pretty obvious that if the punishment is too severe then the EU will suffer. It’s also obvious that the Northern Ireland problem is actually insoluble if Northern Ireland will accept neither a border between the UK and Ireland nor a border between Northern Ireland and Eire. Funny thing is that it seems that all the UK people really wanted was that immigration from the Eastern Bloc of the EU was controlled, and would have been happy with just that. Now, of course, it’s a collision of ideals and it will be difficult to achieve a solution that doesn’t damage everyone involved. Possibly Abd hasn’t been that interested in this (and it’s off-topic), but it’s a nice example of people sticking to principles that sound good but have predictable practical difficulties.
Still (maybe again through the FA experience) I do tend to concentrate on what goes wrong, since it was my job to make sure it didn’t continue to go wrong. I also see this as where your attitude and mine differ, since my reports only needed to specify what needed to be changed to fix the problem whereas yours seem to require extreme detail
Were I doing what you were doing I’d require not much detail as you. These experiments however are not doing what you were doing. In fact they are doing the opposite. You were checking for stuff going wrong, and putting it right. These experiments were looking for stuff going wrong, under the assumption that this indicates a nuclear reaction. So the positive results are not seen as something going wrong to investigate, but as a desired outcome.
In this context, I would subject any such results to intense scrutiny.
Tom – I think you maybe missed the precise point I was making here. When running any sort of test, I was constantly looking for anything that did not match what I expected at that point. Wet patches where they shouldn’t be would have been such an unexpected occurrence, and thus I propose that Miles would have seen it and would have noted it if it had happened. The nuclear/non-nuclear assumption is not necessary here – the cells will be producing gas either way and it will be coming out of the vent. Similarly, where I was looking for electronic signals, I would have noted also when things got unreasonably hot or any other unusual thing happening. Leaks from an electrolytic capacitor (thus a wet spot) would have been seen (as would a mark from dried electrolyte showing that such a leak had happened at some point).
I see that close observation as a characteristic of a successful investigator, whatever the discipline. Miles was obviously good at his job, and so I would expect him to see anomalies better than most. Specifying that he could have missed entrainment is saying that he was incompetent, and of course if you look at the care he took to get his measurements then that accusation seems somewhat far-fetched.
The question you should maybe consider is whether you would have noticed the entrainment if you had been in his place. I expect you would have done, and would have seen it as a problem and would have fixed it.
That observational acuity is something that I think runs across all disciplines. It seems unlikely that the cells were watched 24/7, so it’s possible to propose that the entrainment only happened at night or when people weren’t around, but that seems pretty far-fetched and unreasonable.
Though the specific observation (that entrainment didn’t happen) was not written down, I see it as unreasonable to suppose that it happened and wasn’t noticed. You and I would have seen it if it had happened, and I expect Miles or his co-workers would have seen it.
Entrainment is quite unlikely for Miles. We could look at the Miles/Jones interchange, through. Where entrainment of the kind described by THH could have happened would be with the HAD experiments that Pons and Fleischmann described in the paper that stimulated the Morrison critique, and for that, the proposed mechanism would not involve the loss of cell salts. Further, the water and vapor would be at the boiling point. The vapor would normally condense and that condensation would not be easily distinguished from entrained water that condensed inside the cell. It could escape notice, unless actually measured.
The “simplicity” of the boil-off experiments was in time-to-boil-away. They found accelerated boil-off with active cells, but I don’t think they published all the cell statistics. Still, if that was properly done — I have not completed reviewing the response to Morrison — faster boil-off would indeed indicate XP, with proper controls. What entrainment would cause could be an overstimate of the XP involved. And I haven’t seen a quantitative analysis of that possibility.
The salt titration was a red herring, addressing only one kind of entrainment, actually the most unlikely given the cell design, but the cell design might actually encourage the “distillation” entrainment, which could become significant with higher temperatures. I would think it would self-limit, because the internal condensation would heat those cell components and would stop at that location. It gets complicated.
Calibration should have been done, I’d think, with resistive heating and boiling off. I don’t see that they did this, they only calibrated up to 95 C.
Abd – I was spotlighting Miles, since it seems to me that finding a reason for a false correlation there (from systematic errors) would be most critical. I agree that the boil-off experiments could have had such entrainment more easily, but there it seems there wouldn’t be uncertainty that there was excess heat but only of the actual quantity of it.
In order to propose a reasonable systematic error, I think that a major qualification is that it would not have noticeable evidence and thus could reasonably have been missed by an alert observer. There can be nothing that would bring the thought “that’s odd”. It would thus have to happen in a location where it wasn’t visible, or be a slightly-larger occurrence of something that was expected.
As I see things, though, the value in these discussions is mainly to give a heads-up to Texas about what they might need to eliminate as a potential error. People who are convinced that either F+P’s or Miles’ experiments had errors will continue to think that, no matter what logical arguments are presented here. If however the discussions here turn up a way in which *something* could have happened without Miles noticing it, then that would be useful for the people in Texas to know and eliminate the possibility of it happening in their experiments.
Put it this way – looking back all the time tells you what you’ve stepped in, but looking forward avoids stepping in it.
For Miles, I’d need to look in detail at the experiments to comment. If they are closed cell but with significant unmeasured heat losses dealt with via calibration then CCS/ATER would be one possible artifact to consider.
Tom – report is at http://lenr-canr.org/acrobat/MilesManomalousea.pdf .
They were open cells; on page 16:
“7. The electrolysis cells were not sealed, hence the gases vented directly into the atmosphere. Rubber stoppers were always used for the cell tops. In later experiments involving helium analysis, the cells were sealed with silicon rubber and the gases were vented through an oil bubbler. There may have been contamination from both the silicon rubber and the mineral oil in
the bubbler. It may have been better to have vented the gases through a D 2 O bubbler and then through the oil bubbler. Sealing the cells minimized H 2 O and carbon dioxide (CO 2 ) contamination from the atmosphere but may have added acetic acid and other species from the silicon rubber.”
The problem with D2O purity is also mentioned there, with the initial supply of 3N D2O having more success than that from an alternative supplier.
To me, it seems that Miles responded to any criticism by trying to fix the process – the criticism was acted upon. See bottom of page 21 for a check on the recombination heat and a cross-check on the calorimetric calibration.
Still, read it and see what you think. I was impressed by the lengths Miles went to to get reliable data.
THH, I’m going to recommend that you stop speculating. The truly important Miles work was the heat/helium work, which was still technically open cell, as far as the calorimetry is concerned. The gases were collected outside of the calorimetry envelope. If you want to look at the experiments, I suggest creating pages to do so, you can do that.
What is important with Miles is that a consistent method was used to measure XP — or to estimate it. Given that, and given that his work became a correlation study, the helium findings generally validate the calorimetry, independently. It’s an indication that even if not every possible artifact is eliminated, those artifacts were not operative in a major way. Inferring “nuclear” from heat alone, when the measurements are at relatively low levels of heat and there are a multitude of possible artifacts, depends on an “argument from authority,” i.e., that the researchers are competent. They were competent, but this is a circumstantial argument and not adequate to clearly overcome a major prejudice against the possibility of anomalous heat. The general consensus in the literature is that there actually is a heat anomaly, and Shanahan actually acknowledges that, but asserts his alternative hypotheses (many of them, actually) without setting up or facilitating actual research and, over the years, arguing tendentiously. I congratulate him for being willing to criticize — as you know, I consider that essential to science — but not for all the comments about “believers” and “cold fusioneers.” I actually carefully reviewed Shanahan’s work and acted to preserve the page he had written on Wikipedia on cold fusion calorimetry, and that had no impact on his assessement of me. Here, I’ll say it: he is an ungrateful wretch, self-satisfied. But that doesn’t make him wrong on any given point, only that he is heavily biased. You are giving him more credence than he deserves … but it is also correct that the phalanx of scientists who responded to his JEM Letter did not accurately describe all of his objections, and some of his objections were so outrageous that it appears they did not actually examine them in detail, but just rejected them by their collective authority. I don’t approve of that, but they are also human.
Yes. Exactly. Simon, I would be greatly honored if you would accept Editor privileges here. You understand what I’m doing.
Abd – I’m honoured that you ask, but bear in mind that I’m chasing some crackpot ideas and that it’s maybe better that this site is not associated with the fringe stuff I’m doing. Apart from that, I don’t have a lot of time spare with the current workload. If I’m right, then that work (recycling the energy we already have) is even more important than LENR, and I need to be certain before dropping it or doing other things.
What you have done is to persuade *someone* to back the replication/augmentation of the Miles experiment series. It is useful to crowdsource ideas on how this may be improved such that it is proof against the sceptical attacks that will inevitably emerge, and in order to do that we need discussions on Miles and possibly McKubre and to not get diverted into other experiments where the results are more questionable. I’m however somewhat surprised at the dearth of people commenting here, where there’s such a lot of analysis of the field and the capability of affecting the advance when a suggestion is taken up. Maybe it’s the in-depth analysis of such suggestions here that scares people off. On other sites a line or two of reply seems around par for the course, so there’s not any space for any deep thinking.
The outlier in the Miles series is the Pd-Ce alloy, where heat was seen but no Helium. Possibly there was Helium generated but it remained locked under the surface. Reverse electrolysis may release it. The Pd-B alloys seemed good, but there’s an implication that the process for producing rods was less than ideal and that the swaging step produced a lot of dislocations. In general, do we want dislocations or not? Given that the Pd-B alloys were very hard, and thus difficult to deal with mechanically, it may be useful to produce foils by a sputtering process on a peelable/soluble substrate (sputtering retains the stoichiometric ratio of the target once the initial cleaning phase is done). It’s interesting that Miles works out energy per cc for the cathodes, and only uses the area in the mA/cm² calculation. Foils would be better, and if we don’t need the work-hardening crystal faults but only need the faults produced by the alloying atoms, then a sputtered foil would be ideal. Since there is also a problem that touching the prepared cathode stops it working (Na and K poisoning?) then a freshly-made foil produced by sputtering in a D2 atmosphere would probably avoid that problem, and the connecting wire can be attached by sputtering the foil onto it rather than needing to spot-weld afterwards. In order to find out if the work-hardening faults in the crystal structure are needed, a bit of work with a ball-pein hammer on the foil would provide enough relative movement of the crystals to give that type of structure. It’s also possible to bend the foil beyond its elastic limit a number of times (but not enough to make it crack).
One variable that doesn’t seem to be controlled is the crystal structure of the cathodes. Seems to be that only the percentage composition is normally considered, though I’m open to correction if I missed that. Where we’re talking about diffusion of the D through the lattice, crystal boundaries and number and type of dislocations (and their pinning) may be a critical point.
Maybe a bit off-topic for calorimeters, but that came from re-reading Miles.
Yeah, that’s my suspicion about the Pd-Ce alloy. It might create a surface layer that helium cannot penetrate and, yes, anodic erosion should then release the helium. In the history of heat/helium experiments, out of many, in only two was anodic erosion used, and both instances may have been more or less inadvertent.
Note that this was even better than a single test correlation. The flasks were sent to three different labs for helium analysis, in blind tests. Each of the labs confirmed the ratio. None of the people in the labs knew how much excess heat the cells produced (if any). None of the people knew what the other labs had found.
The chances that all three labs would make mistakes of the same magnitude in the same direction are so small we can dismiss them. That simply cannot happen.
As you say, this does not require any sort of meta-analysis. If THH does not understand how this was done, or what the significance was, or why it is extremely unlikely these results were mistaken, THH is incompetent.
This seems a little off to me. Three different labs were used at different times, if I’m correct. The labs generated data, they did not “confirm the correlation.” Miles interpreted the data as showing correlation (correctly, in my opinion. That is basically math.) MIles estimated the probablity of chance correlation, given his inclusion of the outliers, at one in 750,000. It would be stronger if he had designed the protocol to exclude such ab initio. I.e., the Pd-Ce cathodes would be “additional investigation,” not part of the main study, decided before getting helium results. Miles wanted to study Pd-Ce later but wasn’t funded. I don’t wonder. Those results are quite interesting!
We lose a lot when we judge results as “positive” or “negative.” Results are results, no matter what we think about them.
THH does, if I’m correct, understand the significance. So did Huizenga in the second edition of his book. Heat/Helium is the strong suit of cold fusion experimental results. It is a replicable experiment that does not depend on having a “reliable device.” So-called “negative results,” no heat, actually help confirm correlation. Obviously one needs some excess heat results as well, but it doesn’t have to be many. The more the better, obviously, and the more precise the estimate of the ratio.
I don’t know what exactly they are using in Texas. The Miles approach could simply be replicated with better calorimetry and more precise helium measurements, plus anodic stripping to capture all the helium. I suggested all this because it was highly likely to not waste the research funding.
One of the big problems with cold fusion research is that replicators changed the experiments. In writing my paper for Current Science, I was asked for eye candy, a chart or picture. I wanted to create something that reported all the experimental findings. However, I realized, after a little effort, that there were many problems. Data is not reported consistently. And protocols varied greatly. So at this point what we have are “general confirmations” that are not easily comparable. I remain with the position that the preponderance of evidence is that the heat/helium correlation is a real phenomenon, but … I also recognize the problems. Miles alone, by the way, was much better in that the same basic protocol was followed. However, even he changed things. For example, he tried a Pd-Ce cathode and created the strongest outliers, two experiments with heat and no reported helium. I can easily imagine a cause for the result, but … in correlation studies, one must not arbitrarily exclude data or one could, by cherry picking, consciously or unconsciously, create a correlation where there is none or where it is weak. Miles did not exclude any of the tests, to my knowledge. (there was one more outlier, a cell where there had been a power failure and the calorimetry was suspect. So “measured heat” and “no helium.” Thus … the complete Miles series is 33 measurements, with 12 no-heat cells, none of which showed helium, and 21 heat cells, of which 18 showed the elevated helium, with a reasonably consistent value for the ratio, given the experimental imprecision.
There is a correlation. I find that unmistakeable from the collection of all experiments. The ratio remains with some level of controversy.
What I find is that reasonable skeptics agree on the value of a “new, clearer set” of results. I have talked with Duncan but have nothing to report yet, except for his general optimism and that the work is proceeding.
Personally, I can’t propose any logical way in which in Miles’ experiment the systematic errors in heat measurement could correlate with the (double-blinded) Helium measurements. Any entrainment problems would not produce that correlation, though they may change the measured number of MeV per Helium atom produced.
Entrainment with salt would be noticed by the mess it left (and LiOD is caustic so would be noticeable if you touched it), and entrainment of condensed water would most likely be noticed – wetness where it shouldn’t be wet is pretty obvious. Here I’d expect the people doing an experiment to notice and take note of anything that was unexpected, and where nothing unexpected happened I would expect no specific note that it was running as predicted. People don’t get to be “world leading” by ignoring anomalies.
In order to be able to accept Shanahan’s ideas of calorimetry errors being the sum of the effects seen, I need to have a way in which those errors would correlate with the Helium measurements. I don’t see a mechanism that would do that. Those errors are certainly possible, but why would they result in Helium being sucked into the system in proportion to the magnitude of the error?
I can see THH’s reasoning for “if it wasn’t written down, there is no data”, and also Jed’s reasoning for “if it was running as expected, it wouldn’t need a special note” (for me, I don’t make special notes when all is as-expected, but note odd stuff even if it seems non-relevant). For me, though, these possible causes of errors do not explain the correlation, and so are of less importance unless a way is proposed in which they can produce that correlation. Take the ideas of possible errors into account with the new experiments, but we can’t go back and fill in the missing data from the historical ones – we have what we have.
As I see it, if the heat/Helium correlation can be explained by a non-nuclear systematic error that is logical, then the LENR story can be put to bed by the mainstream scientists as experimental error alone. Without that, though, the correlation remains the solid basis of justification for further experimentation. Discussion of other sources of error is useful in improving the next experiments, but unless those errors together produce the correlation then they don’t impact the experimental reality of LENR.
Meantime, we’ll still have people who believe LENR happens and those that believe it can’t. Better, maybe to look at the evidence.
Ah, Simon, such a pleasure. The most obvious — and incorrect — cause for the correlation would be heat causing seals to leak. Against that are facts:
1. The strongest reason. “Heat” in these experiments does not mean “hotter” generally. when the cell temperature is allowed to vary — I have been strongly recommending against that! — the increased temperature in these experiments is no more than a few degrees C at most. In other work, cell temperature is held constant by fluid flow and the heat necessary to keep the fluid at the same temperature is measured. I have been recommending, in isoperibolic calorimetry, that an internal heater be used to maintain elevated temperature. Storms did that with some recent work and it has the advantage that higher temperature is known to create a stronger reaction (in the range up to boiling, but Storms did not boil and boiling introduces complications better avoided.)
2. In some experiments, helium rose above ambient. That was so in the Case work and in Apicelle at al (2005). Apicelle et al is notable in that the XP was not particularly large. However, they obviously had a small headspace, allowing the detection of lower levels of helium. This was also one of two experiments where conditions caused the release of all the helium, thus the ratio went to the theoretical value, within experimental error. There are deficiencies in both those experiments, but … I have never asserted them as “proof.” Merely as evidence confirming a general claim, and suggesting further research, which is all we need!
3. That leakage would point to the theoretical value is an astounding coincidence! The non-recovery experiments (all but two) have generally shown roughly 60% of the expected helium. That the only two recovery experiments (that dissolved the outer layer of palladium) pushed the ratio to 100% within experimental error is a very strong finding. But we need a more accurate value, because so far the data still allows certain other theories of the effect other than deuterium conversion to helium. (They would be expected to produce a number outside the error bars, but not necessarily far outside them.)
You wrote: “The most obvious — and incorrect — cause for the correlation would be heat causing seals to leak.”
Yes, as you say, this is incorrect. Miles addressed this issue in his papers. He pointed out that some of the tests that produced anomalous heat were run at less input power with less overall power than the tests that failed to produce heat. In other words, the tests with no anomalous heat were sometimes hotter than the tests with anomalous heat.
People should not keep raising this objection. They should read what Miles wrote, so they will see that he already addressed this. Or they can read my summary:
“Could the heat itself have caused a tiny increase in diffusion, perhaps by changing the permeability of the rubber? No, because some of the dud cells were run at higher overall power than the heat-producing cells. Electrolysis power levels during dud runs were raised and lowered considerably, from a half-watt to two watts, with no measurable effect on the helium background. Furthermore, the rubber tubing was outside the cell at room temperature.”
By the way, skeptics often do not raise this argument. I mention it because it is the most obvious idea that could easily occur, and many people would not realize what you and I both mention about temperature and excess heat.
Those were single-blind tests, not double-blind. Double blind tests would involve Miles sending the flasks to a third party who would then send them to the three labs in random order. Miles sent them direct. He knew which flask came from which test, and which test produced anomalous heat.
Sorry. Yes. Not double-blind. His procedure should be adequate, unless he is suspected of chicanery, in which case double-blind would be better. (And he would deposit his heat results with a third party as well, before the helium results are available.)
None of Mile’s critics alleged a problem with this. Mostly they either did not understand or ignored the correlation and only criticized the calorimetry and helium measurements separately.