home *** CD-ROM | disk | FTP | other *** search
- Path: sparky!uunet!olivea!decwrl!mips!pacbell.com!tandem!zorch!fusion
- From: 72240.1256@compuserve.com (Jed Rothwell)
- Newsgroups: sci.physics.fusion
- Subject: From Gene Mallove
- Message-ID: <920721182712_72240.1256_EHL54-1@CompuServe.COM>
- Date: 21 Jul 92 19:55:55 GMT
- Sender: scott@zorch.SF-Bay.ORG (Scott Hazen Mueller)
- Reply-To: Jed Rothwell <72240.1256@compuserve.com>
- Organization: Sci.physics.fusion/Mail Gateway
- Lines: 65
-
- To: >INTERNET:fusion@zorch.SF-Bay.ORG
-
- Gene Mallove's Response to Tom Droege's posting:
-
- I think Tom has jumped to many unwarranted conclusions in his
- analysis of Takahashi's initial results. Jed has addressed many of these
- incorrect conclusions in his response, with which I fully concur. But I have
- a few comments of my own.
-
- Takahashi has clearly seen anomalies that we have not yet seen --
- healthy indications of *temperature elevation* in his cell as a function of
- time (sweep). This has nothing to do with postulated convection variations in
- calibration. Both his high and low end cell temperatures rise. As long as we
- can trust his thermocouple and his ambient temperature drift correction,
- he's got excess heat. This is the main point of his work. I agree that it is
- better to also have a coolant loop inlet-outlet delta-T measurement, and that
- is what we and he have recently been doing, with the assistance of Tom's
- wonderful thermoelectric delta-T gadget. Takahashi apparently is getting
- excess heat with such measurements; we have not yet seen it
- unambiguously.
-
- I agree that at Takahashi's high current input there is something like
- 0.07 C delta-T (electrolyte-to-coolant) per input watt to the cell. And I
- agree that his number would be extremely difficult and unreliable to model
- because of the complexity of convective heat transfer. The problem in
- Tom's analysis, however, is the comparison of this 0.07 with another
- number, 0.61 C/watt, which was obtained by Tom in an entirely different
- way. Tom seems to think that 0.61 C/watt contradicts 0.07 C/watt. It
- doesn't, because 0.07 is right and 0.61 is wrong.
-
- As I understand Tom's analysis, he takes the *initial* slope of the
- temperature decline from Hi to Lo and measures it. I will assume that he
- has measured it correctly and that it is 34 C/hour or 0.0094 C/second. He
- then uses this number multiplied by the thermal capacity of the cell (575cc)
- *(1 cal/C cc) * (4.1 joules/cal) to to get an *initial* heat outflow of 22.1
- watts. That may be a fine estimate of the initial heat outflow. But then
- Tom's analysis breaks down when this initial slope-derived number (22.1)
- is divided into the total temperature difference drop from 33.5 to 20 C.
- That results in the 0.61 C per watt. But this makes no sense because the
- slope of the power curve is changing throughout the 45 minutes to an hour
- to reach the lo-phase equilibrium.
-
- Basically, one can't use a complex time history of temperature decay,
- which may be strongly dependent on the convective heat transfer history,
- to infer anything about the initial high-phase cell resistance. Dividing
- (33.5-20 by 22.1) gives a meaningless number. So there is no basis, in my
- view, for comparing 0.61 with 0.07.
-
- On the basis of this, in my view, bogus comparison, Tom suggests
- that stirring must be changing drastically as the current is shut down. He
- winds up concluding that Takahashi has not calibrated properly because of
- this supposed changed stirring. Let me point out that that Takahashi did
- not calibrate with a pure resistance heater. He used a dead cathode and
- calibrated with electrolysis fully under way, so his cell was always stirred
- That's what he emphasized in his MIT talk. So it seems to me that on that
- misunderstanding alone Tom's entire argument collapses. I do not believe
- Tom's assertion that the conduction resistance of the calorimeter increases
- that much when the power drops.
-
- A final point: both the Bow, NH, experiment and Takahashi's new
- one have time history data for heat outflow as the cell switches from Hi to
- Lo. There is no need to do computations based on curve slopes, we have
- the raw measurements in hand.
-
- - Gene Mallove
-