Exhaust Manifold Velocity

What was done:

As a follow-on to the exhaust manifold pressure drop testing that was performed I decided to use a velocity probe to assess the airflow velocity out of the exhaust manifold at a fixed airflow rate.


I still had available the AWE Tuning, Silly Rabbit Motorsport, and Audi S4 stock exhaust manifolds.

awe-tuning, srm, Audi B5 S4 exhaust manifolds
(L-R) AWE-Tuning, Silly Rabbit Motorsport, Audi exhaust manifolds

Test Procedure:

The velocity probe was placed perpendicular to the plane of the exhaust manifold outlet approximately intersecting the plane across the outlet opening.

exhaust manifold velocity probe test
Velocity probe at outlet to exhaust manifold

Due to the interaction of the airflow with the probe the readings varied slightly.  To try and minimize the affect of this variation 10 samples were taken for the AWE and SRM manifolds and averaged.  The higher velocity from the stock exhaust manifold caused greater variation and therefore 20 samples were averaged.

Only the middle primary was tested.

Here is each product on the bench as the tests were being conducted.

audi b5 s4 exhaust manifold velocity probe test
Stock Audi B5 S4 exhaust manifold
awe tuning exhaust manifold velocity probe test
AWE-Tuning exhaust manifold
srm exhaust manifold velocity probe test
Silly Rabbit Motorsport exhaust manifold

The Results

exhaust manifold velocity probe test results
Airflow velocity test results

Based on the respective diameters of the primary runners and the outlet diameter of the manifolds the results are not surprising.

Project: B5 PB550

A few more months have passed and another option has popped up, this one from Project: B5 – a turbocharger called the PB550.

Project: B5 PB550 Turbocharger
Project: B5 PB550 Turbocharger

Looking at this offering and comparing it to the Project: B5 RS6-X it is not immediately clear what distinguishes these two products.  I have tried contacting Project: B5 to find out what the differences are but have not had any reply back to my inquiries.  Looking at the two product write-ups it seems that the two differ in the following ways:

  • PB550 uses a K04 style billet compressor wheel.
  • RS6-X uses a double stack RS6 style compressor wheel.
  • The RS6-X has a clipped turbine, the PB550 does not.

At this point I have never seen any data on an RS6-X turbocharger or the PB550.  Based on the specs they are still in the region I consider a candidate system, but the lack of data and lack of communication from Project: B5 make it unlikely I would purchase one of these.

As for now the candidate list is:

  • BorgWarner K04 (great amount of data)
  • FrankenTurbo F21/MF (minimal amount of data)
  • ProjectB5 RS6-X/PB550 (no data)
  • SRM Billet K04 (no data)
  • TTE 550 (minimal data)


BoostManager Flow Calibration

After some troubleshooting with the folks from Aquamist to figure out how to record the voltage from their sensor on my LM-1 I set out to calibrate the Aquamist flow sensor in conjunction with the BoostManager water-methanol injection system.

aquamist 806-428 flow sensor
Aquamist Flow Sensor (806-428H)

The Aquamist sensor has three output ranges that can be set depending on how much water flow is expected to be generated by the system.  As it turned out my system was right on the edge between the middle setting, good for around 1200 mL/min., and the upper setting that can handle approximately 2500 mL/min.

Aquamist Water Injection Nozzle
Aquamist 1mm Nozzle with Integrated checkvalve

The decision was whether to get a full range reading, with less resolution, or lose the very top end but have better granularity over the rest of the detection range.

I decided to go with the middle setting that gave better resolution.  I don’t know how much affect spraying into the charge air will have on the flow rate, but I expect it to drop the rate some, which should serve to minimize the negative aspect of the range I chose.  Also I know I can simply set a maximum duty cycle for the pump lower than what it is capable of if I want to keep the flow readings within the reporting range of the sensor.

Snow Performance Tank and CoolingMist Pump
Snow Performance 2.5 gal. tank & Coolingmist Pump

I set up the BoostManager software to test the flow rate.  It has a handy feature that allows the pump to activate with just a ‘press of the button’ on the software.  The pump duty cycle can also be set from 0 to 100 in this ‘test state’ in order to check the flow rate at the different duty cycle settings.

Along with the BoostManager software the data display software for the LM-1 was open to allow me to record the output voltage from the Aquamist flow sensor.

BoostManager & Logworks Screens
BoostManager & Logworks Screens

To capture the water from the nozzles I removed the nozzles from the APR bipipe, tied them together, and placed them in a measuring container with a piece of aluminum foil over the top to contain water mist or splashes.

Flow calibration setup
Flow calibration setup

Here’s a picture of the nozzles directed into the container with the splash cover removed.

Aquamist nozzles into liquid container
Aquamist nozzles into liquid container

With the setup complete I ran the pump at several different duty cycles and recorded the output voltage for each test point.

The pump duty cycle versus flow rate is shown below.

Coolingmist Pump Duty Cycle Chart
Coolingmist Pump Duty Cycle Chart

Shown below is the flow sensor output voltage for the various flow rates.

Aquamist flow sensor output voltage
Aquamist flow sensor output voltage

With the sensor calibration table created and the flexibility of the BoostManager software to allow for the pump duty cycle to be varied based upon inputs from the boost pressure sensor and RPM, I will be able to create a tailored w/m injection rate table.

Eurodyne Boost Manager Water-Methanol Duty Cycle Table
Boost Manager W/M Table

Knowing how much fluid is being injected will help in building a table that is well matched to the amount of fuel that is being injected into the engine.