My 68 MGB/GT has an electric fuel pump that is mounted underneath the
chassis just forward off the rear axle on the right side. It is below the
fuel tank. I've never heard of a vapor lock problem with MG's. Could this
be because of the fuel pump being at a lower level than the tank? If so,
should the TR6 PI fuel pump be mounted down low under the chassis to cure
vapor lock as you mentioned? By the way, I've never suffered vapor lock,
but doesn't vapor lock typically occur when a car is sitting in the heat
(like when you stop to go to the store) as opposed to while driving?
Thanks, Greg
----------
From: Michael D. Porter [SMTP:mdporter@rt66.com]
Sent: Sunday, January 17, 1999 4:05 AM
To: Anthony Gordon
Cc: Hutmacher, Greg; triumphs@autox.team.net
Subject: Re: Triumph PI System info
Anthony Gordon wrote:
>
> The vapor lock problem seems to be well known. The solution is to
do one
> of two things. Buy or construct the cooling loop system that
allows the
> unused fuel returning to the tank to conduct heat away from the
Lucas
> winscreen wiper motor adapted to become a high pressure fuel pump
(how do
> they screw up so many things ...?), basically, the tube wraps
around the
> pump body about 10 times (I think TRF or Rimmers show one in their
> catalog).
Two other possibilities--the first is relatively cheap--provide a
priming pump to feed the inlet of the pump. Second, more expensive,
is
to replace the Lucas pump with a Bosch model (quite a common swap).
Some buses in the late `80s and early `90s used diesel engines
converted
for use with methanol (big mistake), and a principal concern was
vapor
lock. In fact, the fix was a radiator with a couple of fans on the
return line to the tank. Even so, the heat of vaporization of
methanol
is even lower than gasoline, I think, and the fuel heating occurred
not
so much from the pump as from the fuel passing through the engine's
fuel
rail (effectively raising the temperature to near that of the engine
since that is essentially internal).
Nevertheless, the principle of a small finned section of return line
might drop the fuel temperature sufficiently to make vapor lock less
a
problem, as would a similar section before the inlet of the pump.
I wonder, however, about the cooling coil around the pump. At least
in
the case of the Bosch pump, fuel circulates through the pump and
motor
internally, so the temperature of the fuel would be raised passing
through the pump, anyway. Moreover, wrapping coils carrying return
fuel
around the exterior of the pump simply raises the temperature of the
fuel going back into the tank, essentially continuing to raise the
temperature of the fuel as long as the pump is running.
It seems far more logical to keep the pump outside the car where
it's in
the airstream, and, if need be, provide some sort of heat sink for
the
motor case. Further, if it's below the tank just enough, head
pressure
in the tank will help feed the inlet. After all, vapor lock is
common
enough in hot climates with carbureted cars--and it almost always
occurs
on the inlet side of the pump, causing the pump to starve and lose
prime. Same principle should apply with a high-pressure electric
pump.
The best cure, to my mind, would be one which insured flow into the
pump
inlet.
> Second route: replace the Lucas unit with a low pressure
> (18-30lb/sq in) unit from a European import or a domestic car.
The high
> pressure from the Lucas pump isn't actually needed (so I've been
told, but
> since my stuff isn't here and on the car, maybe I should wait).
Nope--that won't do--as I understand it, the injectors require pump
pressure to open, and the injectors don't begin to open until pump
pressure is at least 50-55 psi. Pumps used on digital or analog
injection don't need to open the injectors, since the injectors are
electrical solenoids. The pressure in such systems need be only high
enough to keep the fuel rails filled and to minimize pulsations in
the
system from injector opening and closing.
Beyond that, the fuel calibration is very likely dependent upon fuel
pressure, since it's entirely mechanical, and in such a system, rate
of
flow is affected by pressure. I'd have to see the innards of the
metering unit to be more positive about that, though.
Cheers.
|