DeTomaso Mailing List: July 2001, Message #65

(scroll down)

> Guy,
>
> So you believe in a round earth AND perpetual motion machines?
>
> I don't think I said air won't go down hill, there just has to be enough
> pressure to force it where it needs to go. And all of your excellent
> examples demonstrate that. No argument.
>
> But in #4, what happens after all the air is out of the radiator and the
> return tube?

I agree, if that hose fills up with water, it may be useless.  While it is
unlikely a column of air will support a column of water from underneath,
 the water would typically fall past the air, making bubbles that rise
through the water column, I don't know if there is a scientific term for
this ) , at some point an amount of water in that hose would be sufficient
to offset the pressure of the air in the radiator, that is it's weight would
be the same as the weight of the water in the header tank pressing down
(up?) on the air in the radiator.  If the air can get to the horizontal
section, they you'd have a column of water on top of a column of air ( or
more strictly, alongside each other ) and the air might "bubble" past the
water.  Then again it might not due to surface tension or other "non-linear"
factors.  Maybe I wasn't clear about this earlier, if so I apologize.

Also, it's not clear to me what the intent is once the water pump is
turning, the flow would tend to be toward the radiator.  It is probably not
insignificant that the tech bulletins call for the hose to be attached at
the very top of the header tank, but why?

I don't have this arrangement on my car, so I cannot volunteer to replace
the lines temporarily with clear plastic to observe what goes on.  But given
that this arrangement was "formalized" on the later cars with a metal tube
from front to rear, I hesistate to believe it simply doesn't work.

Guy D.


>
> In our mental exercise for today, we now add some air back into the
> radiator; let's say we gently blow some air in the other vent cap at the
> top of the radiator. For simplicity, let's assume the header tank cap is
> off. Water will be displaced out of the radiator and air will start down
> the return tube. But since the water pressure in the return tube will be
> equal to the pressure at the water/air boundary in the top of the
> radiator we have an equilibrium situation and the air bubble is trapped.
> The only way to get the flow started is to suck the water out of the
> return tube, which reduces the "back pressure" so that there is a
> pressure differential that will push the air back to the header tank.
> Sort of the reverse of getting a syphon started. Or, as I suggested
> below, if there is enough temperature difference (and therefore density
> and pressure) between the water on either side of the bubble, then flow
> is possible. But since the return line should be cooler, the flow should
> be from the header tank through the return to the radiator which will
> not get rid of the bubble.
>
> Richard
>
> Guy Dellavecchia wrote:
> >
> > Richard, Chuck, et al,
> >
> > Where does the idea that air cannot go "downhill" in a tube come from?
> >
> > Example 1:  I often refill transmissions, Pantera and others, using a
> > siphon.  Air in the hose I use is no factor b/c gravity pushes the much
> > heavier oil into the trans and the air gets pushed out of the way.
Often
> > the hose takes a circuitous path from the oil jug to the trans.  Uphill,
> > downhill, doesn't matter.
> >
> > Example 2:  Take a loop of hose, a garden hose perhaps.  Hold one end 5
ft
> > off the ground, hold a u-shape at the middle 4 feet off the ground, hold
the
> > other end next to the first end, but 3 feet off the ground.  Can you not
> > pour water into the five foot high end and have it eventually come out
the
> > other end at 3 feet off the ground, the water having pushed the air
downhill
> > at the start and at the 4 foot high u-shape?
> >
> > Example 3:  Fill the header tank on your Pantera.  Open the vent on the
> > radiator, coolant comes out, yes?  Now, if you connected a hose to that
> > vent, ran it (downhill) to the back of the car, but set the hose on the
> > ground (at the back of the car), coolant would still flow out, yes?  Now
> > start raising the end of the hose at the back of the car, does the
coolant
> > stop?  I say it will continue to flow "uphill" as long as the the end of
the
> > hose is lower than the fill level in the header tank.
> >
> > Example 4:  Picture a completely empty Pantera cooling system.  No vents
on
> > the radiator.  Start filling it from the header tank.  Coolant fills the
> > radiator only to the top of the upper metal pipe.  Top of the radiator
is
> > filled with air, like pushing a glass upside down into water.  Now
connect a
> > hose to the top of the radiator.  Let the air out of the radiator.
While
> > doing so, do you expect the air to stop when the hose is pointed
downhill?
> > While venting the air out through this hose, if you suddenly connected
it to
> > the header tank, would you expect the air to stop?
> >
> > Next let's debate if the world is flat or round.  I have dibs on round.
> >
> > Guy D.
> >
> > ----- Original Message -----
> > From: Richard Barkley <rbarkley@earthlink.net>
> > To: Multiple recipients of list <detomaso@realbig.com>
> > Sent: Sunday, June 24, 2001 1:51 PM
> > Subject: Re: Cooling
> >
> > > Guy,
> > >
> > > Good analysis. Except if the water flowed as you described in the
return
> > > tube with the engine off we'd have a perpetual motion machine. Unless
> > > there were an uphill path all the way from the air bubble in the
> > > radiator to the water tank (so the bubble could rise to the top), the
> > > bubble would stay trapped.
> > >
> > > To convince yourself of this, draw a tank with a loop of hose going
> > > down, around and coming back to the tank in the shape of a big U. The
> > > end of the U away from the tank represents the air trap in the
radiator.
> > > Now draw a horizontal line anywhere across the U. If the fluid in the
U
> > > is at the same temperature (and thus the same density) the pressure at
> > > each set of horizontal points on the U will be at the same pressure
and
> > > no fluid will flow. If it did, we'd have a perpetual motion machine.
On
> > > the other hand, if parts of the U are at a different
> > > temperature/density, then there will be a pressure difference and
> > > corresponding flow. This is the basis of convective flow. In our
> > > application I would expect that the return tube would cool much more
> > > quickly than the main tubes. That would make that side of the tube
> > > "heavier" and water would flow out of the tank toward the radiator
> > > convectivly.
> > >
> > > So, either way (running or cooling off) the water would flow to the
> > > radiator in the bypass tube. But contrary to my previous thoughts,
this
> > > won't get rid of the air bubble, it would just displace water already
> > there.
> > >
> > > As you suggested, the bypass tube would work as a vent on the initial
> > > fill as long as the tank end remained above the water line. But this
> > > seems like the only way it would work. It also seems like way too much
> > > bother for the factory to go to. So now I'm stumped. Maybe there's a
> > > missing piece of info here.
> > >
> > > Richard
> > >
>
> Sincerely,
> Richard Barkley
>
> To Ski or not to Ski, that is the question!
>
> Richard Barkley
> (310) 373-6695 (home), 813-2432 (work)
> E-mail: richard.barkley@trw.com (work), rbarkley@earthlink.net (home)
> Mammoth Condo 2BR+loft/3Bath - Horizons 4 #186 (760) 934-6758
> Condo web address: http://home.earthlink.net/~rbarkley
>