GE's Boys, Levy, Fuller and Niemi conclude:
“5 Burnout heat fluxes were measured with heat generation around the entire periphery of the rectangular channels. Burnout occurs at the corners and the values there are correlated by a McAdams type equation. Values one third of those obtained in circular pipes were measured.” (Page 140)
Discussion by S. J. Green and B. W. LeTourneau of the Bettis Plant,
“It would seem overly conservative in most cases to apply these data directly to the coolant channels in nuclear reactors. In conventional reactor fuel plates, the fueled region is of slightly less width than the coolant channels. While there is a significant amount of gamma heating in the extreme ends of the fuel plates and in the side plates, the geometry and the heating rates are normally such that the heat flux through the narrow edges, and the portions of the wide sides nearest the corners, is much less than that opposite the fueled region.
For this reason, the rectangular channel test sections used for electrically heated burnout tests [28, 29] at the Bettis Plant have been designed with a reduced metal cross section in the corners, such that the heat flux at the narrow edges and on the part of the wide faces nearest the corners was about 20 percent of that in the main part of the channel. With this type of channel (and at pressures form 2000 to 600 psia) the burnout heat flux was found to be in fact higher than that for a round tube at the same conditions, and the burnouts did not occur preferentially at the corners.”
Here is the lengthy dissertation:
GETR Levy Heat Transfer
S. Levy R. Niemi R. Fuller
Heat Transfer to Water in Thin Rectangular Channels
ASME Transactions
Journal of Heat Transfer, Volume 81, Series C, Number 2, May 1959, pages 129-140, discussion pages 141-143.
www.lindahall.org
Contact at ASME: McAteer (212) 591-7033
Contact at Linda Hall Library: Sabina Marsh, extension 793
The following attachment contains the document(s) you requested from the
Linda Hall Library Document Services Department. The attachment is a
MIME attached .pdf, and you will need a .pdf viewer, such as Adobe
Acrobat Reader in order to view the document. If you do not already
have access to a .pdf viewer on your computer, you can download one free
of charge from Adobe's website:
http://www.adobe.com/products/acrobat/readstep.html
Terms and Conditions
The contents of the attached document may be copyrighted. If the
document is protected by copyright law, the following restrictions
apply.
You may print only one paper copy, from which you may not make further
copies, except as may be allowed by law.
You may not retain the electronic copy after printing out a paper copy.
You may not make further electronic copies or convert the file into any
other format.
You may not cut and paste or otherwise alter the text.
You may forward the document to the end user in order for them to print
the single copy, but you may not save or print the document before
forwarding.
Permission obtained from the Rightsholder or the
Linda Hall Library
Document Services*
Phone: 816-363-4600 Toll-free: 800-662-1545 Fax: 816-926-8785 email:
docserv2@lindahall.org Web: www.lindahall.org
*Also providing AIAA Dispatch, EiCompendex, CSA Materials, and Ei
EnCompass services
Levy, Fuller and Niemi conclude:
“5 Burnout heat fluxes were measured with heat generation around the entire periphery of the rectangular channels. Burnout occurs at the corners and the values there are correlated by a McAdams type equation. Values one third of those obtained in circular pipes were measured.” (Page 140)
Discussion by S. J. Green and B. W. LeTourneau of the Bettis Plant,
“It would seem overly conservative in most cases to apply these data directly to the coolant channels in nuclear reactors. In conventional reactor fuel plates, the fueled region is of slightly less width than the coolant channels. While there is a significant amount of gamma heating in the extreme ends of the fuel plates and in the side plates, the geometry and the heating rates are normally such that the heat flux through the narrow edges, and the portions of the wide sides nearest the corners, is much less than that opposite the fueled region.
For this reason, the rectangular channel test sections used for electrically heated burnout tests [28, 29] at the Bettis Plant have been designed with a reduced metal cross section in the corners, such that the heat flux at the narrow edges and on the part of the wide faces nearest the corners was about 20 percent of that in the main part of the channel. With this type of channel (and at pressures form 2000 to 600 psia) the burnout heat flux was found to be in fact higher than that for a round tube at the same conditions, and the burnouts did not occur preferentially at the corners.”
Responding to the Bettis remarks, the GE investigators address the burnout data as follows:
“With regard to the burnout data it is very hard to see how the local heat flux can be three times as large as the average heat flux. Conduction effects along a 28-mil wall could never produce such extreme peaking. It is also important to note that the proposed data apply to test reactors as noted in the first sentence of the introduction and not to power or propulsion reactors referred to by the discussers.”
Remarks from Leyse:
Within 6 months of my employment during 1960 at General Electric’s Vallecitos Atomic Laboratory, I recognized the error in the General Electric burnout correlation for nuclear test reactors and I did not use it in studies that were aimed at increasing the power level of the General Electric Test Reactor (GETR). In their response to the Bettis engineers, Levy, Fuller and Niemi state, “… it is very hard to see how the local heat flux can be three times as large as the average heat flux. Conduction effects along a 28-mil wall could never produce such extreme peaking.”
Indeed, in 1960 I recognized that it is not at all hard to see how, with an unrelieved corner, the local heat flux can be three times as large as the average heat flux. As observed by Green and LeTourneau , Levy, Fuller and Neimi did not provide a reduced metal cross section the corners.
The heat generation within the volume of the corner adds very substantially to the corner heat flux and conduction effects have nothing to do with it. In fact, with the relatively low thermal conductivity of stainless steel, heat conduction away from the corner hot spot is not sufficient to overcome the intense peaking of heat flux at the corner. Thus, it is obvious that the GE test section will always burn out at a corner.
Furthermore, the remark by the GE investigators that their work applies to test reactors in contrast to power reactors is specious at best, and more accurately, it is unambiguously absurd.
In 1960 I introduced the Vallecitos staff to the Bernath correlation for burnout heat flux for capsules for the GETR, and I must say it had a chilly reception. However, in the safety analysis for increasing the power level of GETR, publication APED 5000-A, July 1965, I reference three separate burnout correlations: Bernath, McAdams and Mirshak. If I had been forced to apply the GE burnout correlation, the power level increase would not have gotten off the ground!
In an internal memo, Interpretation of Burnout Margin Requirements for GETR Capsules, May 16, 1967, J. E. Morrissey of GE’s Irradiation Processing Operation (IPO), declared, “APED 5000 A uses the Bernath correlation.” Morrissey does not mention the GE burnout correlation, it is ignored without evaluation.
No comments:
Post a Comment