At 03:26 PM 10/13/2011, Higgins Bob-CBH003 wrote:
Hi Mark,

I will consider such a drawing.  However, the present diagram is not
geometrically correct - the internal unit is rotated in the
cross-section so as to highlight the fins.  What is needed is a proper
drawing from the pictures.  I just don't know what useful insight would
be obtained from spending the time on that.

I don't think anything's to gain. The only thing I'd change is to mark the pressure regulator as speculative.

I think it would be more useful to draw a speculative cross-section of
the headers of the heat exchanger to have a proper discussion of the
heat flow.  It is my contention, that because of the high secondary
flow, that the heat from even primary hot water would not cause a
significant error temperature rise in Tout.  This is because the heat
conducted through the brass from the primary input would have to travel
along the brass shell past the flowing secondary water for at least an
inch through a cross-section of about 1/4".  The secondary water tube in
that section can be considered nearly a perfect sink because of the high
flow, and almost all of the heat from the primary will terminate in that
water - which is where it is supposed to terminate anyway.

I started on that : http://lenr.qumbu.com/111010_manifold_001_h1200.jpg
(but misunderstood the "nut" that my observer said the thermocouple was connected to -- so I didn't draw the attached pipe segments).

See (eg) my 2-resistor models at
http://www.mail-archive.com/vortex-l@eskimo.com/msg52539.html
(Horace Heffner first suggested the resistor model, but I used it in a different way).

21-resistor model
http://www.mail-archive.com/vortex-l@eskimo.com/msg52547.html

He gave me some clear pictures of the manifold :
http://lenr.qumbu.com/111010_pics/111010_1_crop.jpg
http://lenr.qumbu.com/111010_pics/111010_2_crop.jpg
http://lenr.qumbu.com/111010_pics/111010_3_crop.jpg
http://lenr.qumbu.com/111010_pics/111010_4_crop.jpg


Regards, Bob Higgins

Reply via email to