Department of Energy Argonne National Laboratory Office of Science NEWTON's Homepage NEWTON's Homepage
NEWTON, Ask A Scientist!
NEWTON Home Page NEWTON Teachers Visit Our Archives Ask A Question How To Ask A Question Question of the Week Our Expert Scientists Volunteer at NEWTON! Frequently Asked Questions Referencing NEWTON About NEWTON About Ask A Scientist Education At Argonne Rate of Vacuum Leak
Name: Matt
Status: other
Grade: 12
Location: MN
Country: USA
Date: Summer 2011

I am an automotive diagnostic technician. A group of friends and I are discussing a phenomena with how on-board fuel control systems deal with "pirate air" (vacuum leak, air leak, un-metered air leak, etc). We are coming up on the idea that even though the size of the leak (area?) may remain constant the amount it flows does not. The intake manifold acts as a low pressure (air) reservoir. Let us say at idle there is 33kPa of pressure in said reservoir with a barometric pressure of 100kPa. As engine RPM is raised initially pressure will increase but as the raise RPM is stabilized pressure typically drops, lets say to 28kPa, the flow of the leak will now increase, correct? I understand there are other factors and am in no way fluent with fluid dynamics. But as a whole, is this theoretically sound?

Yep, the pressure difference across an orifice (the hole) is the primary determinant of gas flow, with the air flowing from high pressure to low pressure, as you have described. It is not a linear process though (e.g. if you double the pressure, you do not necessarily double the gas flow / leak).

If you want to read more, I suggest you look up "fluid flow through an orifice", which is essentially what you're talking about. There are lots of resources on this topic.

Hope this helps,
Burr Zimmerman

A fine question, Matt.

This is probably relevant to you:

Also and other valve companies. (They often give reference data on how much flow goes thru their valve when wide-open, as a function of the pressure on each side.) see Plot "For Air" on page 19 at the top (The curves are strikingly non-linear!) (never use water-flow curves to model an air-leak - they have completely different shape)

Some leaks are through relatively short, wide single passages. Like the throat of a rocket nozzle, although not necessarily so well streamlined. They tend to inject a jet-like stream of gas into their low-pressure side. The gas accelerates into this trap-door-to-vacuum as fast as it can, and the behavior is controlled by the mass and momentum of the gas. (It's a lot like freeway traffic funneling through the one open lane around an accident.) For hi-side to low-side pressure ratios greater than about 2.0, The peak speeds at the throat actually approach the speed of sound. The total flow is proportional to the hi-side pressure. (The more gas there is, the more can fall into this hole). So, surprisingly, the total flow is independent of low-side (downstream) pressure. Into 40kPa or 0kPa, the flow from 100kPa will be the same. It is said that pressure signals from the low side cannot propagate upstream and have any effect on the hi-side or flow rate or pattern.

Other leaks are basically made of multiple leaks in series, each with a pressure-ratio only a little over 1.0. A thick rubber-wall hose with a pinhole or hairline crack would fall into this category. A toilet-paper filter from manifold to air would be the ultimate example. I tend to call it a diffusive leak. Think of it as a salmon-compatible stream: so many intermediate surfaces for air-molecules to bounce off of, and so much passage-width, that the net speed is slower than sound and some molecules actually can and do go upstream during the leak, and the total flow can and does depend on downstream pressure, being proportional to the difference between upstream and downstream pressure.

Longish open-end tubes tend to be in a similar category , or a category in between.

In between jet-like and diffusive is the Venturi-like leak: a single throat with a pressure-ratio which is more than 1.1 but less than 1.8.

I am not so sure the flow varies just the way you are speculating. Some leaks will vary, others will not. I think the ones big enough to make a difference to the engine will often be jet like, and then they would add a relatively constant current of air regardless of manifold pressure. What is the typical big leak you see? "Pirate air" might imply it's often a negligently open-ended tube. I think these change with pressure, but slowly, by a percentage maybe half of the percent change in the pressure ratio.

Try thinking it through with the presumption that the leak-flow does not change (or changes < 20%) for 0-40kPa manifold pressure. Even less for 28 to 33kPa. See if that fits the engine behavior.

Jim Swenson

Hi Matt,

Yes, the amount of air leakage into an engine's intake system quite clearly is not constant. This air leakage depends on two factors. The first is the area of the holes where leakage is occurring, and second is the degree of intake manifold vacuum. Although the leakage path area is constant, the vacuum is definitely not, thus the volume of air leaking varies strongly depending on manifold vacuum.

Inlet manifold vacuum depends primarily on throttle position and engine RPM, and is highest at high RPM and closed throttle (such as when descending a hill in a lower gear). In this situation, the motor is acting as a pump, sucking air out of the manifold. But with a closed throttle, the only air entering the manifold to replace that being sucked out, is via the leakage you refer to. The result is a high manifold vacuum, and greatest air leakage.

Opening the throttle (even a little) under the above conditions will allow more air to enter the manifold, and this the degree of manifold vacuum will be reduced. Similarly, even without opening the throttle, reducing engine RPM will result in the engine sucking less air out of the manifold, and thus also reducing vacuum. Either way... by either opening the throttle, or by reducing engine RPM... manifold vacuum and hence air leakage will be reduced.

As you were already hinting at, air leakage at low throttle openings and higher RPM can be significant, and can upset the desired fuel mixture and cause "trailing throttle misfiring". This is why most fuel injection systems completely cut off all fuel being injected, whenever conditions of a closed throttle and greater than about 2000 RPM are detected.

Robert Wilson


Laws of dry friction (of a bicycle tire in contact with the road) br>
The properties of sliding friction were discovered by experiment in the 15th to 18th centuries and were expressed as three empirical laws:

Amontons' First Law: The force of friction is directly proportional to the applied load. Amontons' Second Law: The force of friction is independent of the apparent area of contact. Coulomb's Law of Friction: Kinetic friction is independent of the sliding velocity.

But, Amontons' 2nd Law is an idealization assuming perfectly rigid and inelastic materials. For example, wider tires on cars provide more traction than narrow tires for a given vehicle mass because of surface deformation of the tire.[citation needed] [edit] Dry friction

A highly pressurized bicycle tire will have less contact area with the surface of the street than the same tire at a lesser pressurized level and therefore will have less deformation, creating less braking friction. Under inflated tires have wider contact with the road and therefore the case of the idealization of Amontons' 2nd Law applies.

Sincere regards,
Mike Stewart

Click here to return to the Engineering Archives

NEWTON is an electronic community for Science, Math, and Computer Science K-12 Educators, sponsored and operated by Argonne National Laboratory's Educational Programs, Andrew Skipor, Ph.D., Head of Educational Programs.

For assistance with NEWTON contact a System Operator (, or at Argonne's Educational Programs

Educational Programs
Building 360
9700 S. Cass Ave.
Argonne, Illinois
60439-4845, USA
Update: June 2012
Weclome To Newton

Argonne National Laboratory