land-speed
[Top] [All Lists]

RE: Dry sump oiling HP gains

To: "'Ed Weldon'" <23.weldon@comcast.net>, <land-speed@autox.team.net>
Subject: RE: Dry sump oiling HP gains
From: "John Staiger" <lsrvette@yahoo.com>
Date: Fri, 20 Oct 2006 12:01:38 -0400
Additional thoughts from an engineering point of view...

I apologize if this has been discussed already -- however I did not see this
thought covered in the emails I received.

With reference to the engine bottom end as an air pump -- one train of
thought when running a dry sump is not to focus on reducing (if this is even
possible) the air resistance as this is insignificant in the long run -- but
to focus on reducing the matter that is moving with the air... Since oil has
mass, and mass has weight, when the piston moves air (up or down) it moves
molecules of oil suspended in the air (the worst case would be Foam) up and
down.  This adds a surprising large amount of mass to the rotating mass of
the engine. The [combined] mass must be accelerated and decelerated just
like moving the steel or aluminum parts. Although properly designed pans and
windage trays help -- there is still a lot of oil being picked up and flung
around just suspended in the air.  Multi-stage dry sumps and vacuum pumps
focus on removing every molecule possible (even if this means crating a
vacuum that pulls on the pistons) so that the rotating mass of the engine
does not; 1) hit the mass and have to accelerate it, and/or 2)have to pump
it (accelerate & decelerate it).

I believe this is the same subject, often discussed with reference to the
Aero dynamics of vehicles -- it's easier to make a [low hp] hole in the air
than to close it up...  

For impact, I am reminded of a thought introduced by the chief engineer of
the Thrust SSC & JCB diesel -- he told a friend of mine at BV this year that
the SSC had to accelerate something like a ton (yes 2K lbs.) of dirt when it
ran on Black Lake... more food for thought and debate...

-- snip --
1.  The air pumping losses from the pistons may be more than you think...




<Prev in Thread] Current Thread [Next in Thread>