[s-cars] What constitutes a leak?

[Tony Guttmann]

Dear Dave,

s noted in my original message, I also plugged the big hose that enters
the throttle body housing (i.e. the intake manifold an the opposite side
of the block to the exhaust manifold). Thus, as far as I'm aware I'm not
allowing the air to go down each intake runner.

Does this change your response?

Thanks,

Tony
On Wed, Aug 17, 2005 at 02:18:13PM -0400, djdawson2 at aol.com wrote:
> Tony,
> Your expectations for a "leak proof" system are unrealistic.  You are expecting to pressurize the intake tract, and have zero leaks... is what it sounds like from your description.  When you put a plastic bag over the exhaust pipe, you were on the right track, but drawing the wrong conclusion.
>  
> When you pressurize the intake, where does that air go?  Well, first through the turbo compressor housing, then the plumbing (BPV, and IC), and finally the intake manifold.  At that point, the pressure goes down each intake runner.  If that particular cylinder's intake valve is open, the cylinder will fill.  BUT, if it isn't open, that cylinder will NOT fill... or fill very slowly.
>  
> Suffice it to say that air under pressure WILL (not yelling, just declaring the certainty on this fact) go into some cylinders.  Likewise, some cylinders WILL have exhaust valves open, or partially open.  As a result, you WILL certainly create a path for the air used for pressurizing the intake to eventually escape through the exhaust system.  This is not a fault, but simply a reality of testing in this fashion.  You could be super anal and remove both the intake and exhaust camshafts, thereby closing all valves, and reducing (not eliminating) all possible paths for air to "leak" past the cylinder head.  OR, you could remove your downpipe, and seal the output flange on the turbine housing.  Then you would  be able to expect zero, or very nearly zero pressure loss.  Bottom line... if the space shuttle were powered by a reciprocating engine, the test results would be the same... it would "leak"... or more accurately, simply have a path for pressure to escape the system.