I've been conducting pressure, temperature and drag survey tests on the stock and modified RV-8 cowling for several months now and thought I'd share some results.
Need to throw out a teaser first - Dan Horton, Sonny Wiersema and I have instrumented our RV8 cowls with Piccolo tubes and co-axial dynamic pressure probes, which Dan elegantly crafted in his usual expert manner, connected to digital water manometers. Each of us has a different cowl configuration: Dan has a plenum with modified inlets and exit, Sonny has a Sam James and I have the stock RV8 (although I've modified it with an exit nozzle and variable exit flap now). The purpose of our testing was to compare the various cowl configurations and develop comparisons in pressure recovery and a metric for cooling efficiency. We will release our findings when we are complete with all the testing....look for it - we think you'll find it very interesting and helpful!
The purpose of this post however is to report (quantitatively) the effects of adding an exit nozzle (with variable flap) and oil cooler duct on Exit Drag.
Instrumentation:
I mounted a coaxial pressure probe, which measures the dynamic pressure, in the exit plane of both the stock cowl and with the exit nozzle installed. Additionally a special temperature sensor was installed in the exit flow in order to calculate the flow density needed to determine exit velocity.
I won't go into all the details of how the sensors were configured, calibrated, or installed as those details will be included when Dan and Sonny are done collecting their data. Suffice it to say the probes were calibrated and the installation and test procedures verified through nearly 15 hours of testing.
The manometer itself was calibrated and is very accurate from 0-20" H20 which is the range we are looking for.
Note: The flow through the cowl and out of the exit is anything but steady. Heat xfer and lots of flow obstacles make quantifying the flow thoroughly very difficult. The data gathered only quantifies the flow at those particular probe positions and cannot fairly be extrapolated to other areas. The numbers I present are only valid as comparisons from one configuration to another and not necessarily representative of the true flow state throughout the cowl or exit area.
Test Procedure:
Level Airspeed sweeps from 110mph-Vmax at a PA of 3500 ft, Recording pressures and temperatures in 20mph increments.
Additionally, GPS 4 course airspeed calibration data was collected simultaneously to ensure the data reduction was accurate.
Data: Exit Total Pressure (Pitot side of coaxial probe - A/C Static), Exit Dynamic Pressure (coaxial Pitot - coaxial static), Exit Temperature, Freesream Velocity (TAS), OAT, PA, Ground speeds, Track
Configuration
Superior Xp-360, carb, FP Wood prop
2 cowl configurations:
Stock - unmodified cowl exit, OC mounted low on left engine mount fed by a 3" dia SCAT, no OC exit duct. OC inlet blocked from 7 sq " to 3.5 sq ", Stock Inlets, stock baffles, Heater inlet blocked 50%. The entire cowl perimeter, including the two halves and prop flange have been sealed with baffle material to minimize leakage and ensure the maximum amount of air flows through the exit. Exit/Inlet Ratio = .75
Modified: Same as stock with addition of an exit Nozzle with variable exit flap which reduces exit/inlet ratio from as low as .38 with flap closed and up to .70 flap open.
Exit Flap Closed:
Exit Flap Open: Note the axial pressure probe
The exit nozzle takes advantage of the sloping exit ramp in the stock RV8 configuration reducing the exit area from 47"sq to 25"sq.
As most of us know, the exit is the system throttle. The larger the exit the more mass flow we will have through the system and the cooler our engine should be. Having a large exit, while providing better cooling will also increase the exit drag as the flow velocity will be significantly slower through a larger exit area. The drag is measured as the change in flow momentum in the longitudinal direction. Flow momentum = Mass Flow * the change in flow velocity from freestream (true airspeed). The higher the exit flow velocity the lower the change in velocity and therefore the lower the change in momentum = less drag. Unfortunately, the pursuit for less drag through smaller exit area will also reduce cooling capacity. For a fixed geometry exit a tradeoff between drag and cooling must be made. For this reason I chose to fabricate an exit flap allowing me to increase the exit area to nearly stock dimensions or reduce the area to less than half. This flap is controlled by a Ray Allen trim servo connected to a switch on my stick grip with position info displayed on the VP-50 electrical system panel.
The RV8 stock exit area, is designed for the worst case cooling condition of WOT climb at Vy which means too much mass flow exists during high speed level flight leading to excessive drag. During climb I open the flap providing full cooling mass flow while in level flight the flap is closed to reduce drag - not a new concept as cowl flaps have been around forever. Additionally, I applied the same concept of exit throttle to the OC. Most folks seem to prefer blocking the inlet and some believe this is more effective than blocking the exit. This idea doesn't match theory nor experimental data. If we block the inlet to the OC we will certainly reduce mass flow and temps will increase but with a drag penalty as the air will exit the OC slower than it entered and will not aid in increasing overall cowl exit velocity. Now if we throttle the OC exit with a nozzle we will both reduce the mass flow to increase OT AND increase OC exit velocity which again reduces drag. It matters not from a purely cooling perspective which side of the OC you block but if you want drag reduction throttle the exit and accelerate the flow. My OC throttle reduces the the exit to 4" sq with an inlet area of 7" sq. OT without the OC blocked typically remained in the 160-170 range. With the inlet blocked 50% OT were in the 185-195 range. Now with the OC exit throttled to 4"sq OT are in the 185-195 range but with greater exit velocity.
Pics of the OC exit throttle:
Results:
Using a fixed W/delta technique I measured a TAS increase of 6MPH with the exit flap closed and nozzle installed over the stock cowl configuration and a 7MPH increase from the exit flap open condition (the exit flap produces some drag itself when open).
Here is the data:
Cooling: I'll provide more info on cooling when the full test is completed but for now I'll say the exit flap works as designed in that during climb when flap is open my Avg CHTs are just slightly higher than stock (even with the flap open the exit area is a bit smaller than stock) but during WOT level flight they avg low 300s with exit closed. A WOT climb at Vy with the exit flap closed results in CHTs rising about 20-30 deg hotter than with it open as expected.
Thanks to Dan for fabricating and sending the pressure probes and many discussions with he and Sonny along the way. The full test will be full of good information. I tried to keep the math out of this posting but if there is interest I can submit more detail on how the data was reduced.
Need to throw out a teaser first - Dan Horton, Sonny Wiersema and I have instrumented our RV8 cowls with Piccolo tubes and co-axial dynamic pressure probes, which Dan elegantly crafted in his usual expert manner, connected to digital water manometers. Each of us has a different cowl configuration: Dan has a plenum with modified inlets and exit, Sonny has a Sam James and I have the stock RV8 (although I've modified it with an exit nozzle and variable exit flap now). The purpose of our testing was to compare the various cowl configurations and develop comparisons in pressure recovery and a metric for cooling efficiency. We will release our findings when we are complete with all the testing....look for it - we think you'll find it very interesting and helpful!
The purpose of this post however is to report (quantitatively) the effects of adding an exit nozzle (with variable flap) and oil cooler duct on Exit Drag.
Instrumentation:
I mounted a coaxial pressure probe, which measures the dynamic pressure, in the exit plane of both the stock cowl and with the exit nozzle installed. Additionally a special temperature sensor was installed in the exit flow in order to calculate the flow density needed to determine exit velocity.
I won't go into all the details of how the sensors were configured, calibrated, or installed as those details will be included when Dan and Sonny are done collecting their data. Suffice it to say the probes were calibrated and the installation and test procedures verified through nearly 15 hours of testing.
The manometer itself was calibrated and is very accurate from 0-20" H20 which is the range we are looking for.
Note: The flow through the cowl and out of the exit is anything but steady. Heat xfer and lots of flow obstacles make quantifying the flow thoroughly very difficult. The data gathered only quantifies the flow at those particular probe positions and cannot fairly be extrapolated to other areas. The numbers I present are only valid as comparisons from one configuration to another and not necessarily representative of the true flow state throughout the cowl or exit area.
Test Procedure:
Level Airspeed sweeps from 110mph-Vmax at a PA of 3500 ft, Recording pressures and temperatures in 20mph increments.
Additionally, GPS 4 course airspeed calibration data was collected simultaneously to ensure the data reduction was accurate.
Data: Exit Total Pressure (Pitot side of coaxial probe - A/C Static), Exit Dynamic Pressure (coaxial Pitot - coaxial static), Exit Temperature, Freesream Velocity (TAS), OAT, PA, Ground speeds, Track
Configuration
Superior Xp-360, carb, FP Wood prop
2 cowl configurations:
Stock - unmodified cowl exit, OC mounted low on left engine mount fed by a 3" dia SCAT, no OC exit duct. OC inlet blocked from 7 sq " to 3.5 sq ", Stock Inlets, stock baffles, Heater inlet blocked 50%. The entire cowl perimeter, including the two halves and prop flange have been sealed with baffle material to minimize leakage and ensure the maximum amount of air flows through the exit. Exit/Inlet Ratio = .75
Modified: Same as stock with addition of an exit Nozzle with variable exit flap which reduces exit/inlet ratio from as low as .38 with flap closed and up to .70 flap open.
Exit Flap Closed:
Exit Flap Open: Note the axial pressure probe
The exit nozzle takes advantage of the sloping exit ramp in the stock RV8 configuration reducing the exit area from 47"sq to 25"sq.
As most of us know, the exit is the system throttle. The larger the exit the more mass flow we will have through the system and the cooler our engine should be. Having a large exit, while providing better cooling will also increase the exit drag as the flow velocity will be significantly slower through a larger exit area. The drag is measured as the change in flow momentum in the longitudinal direction. Flow momentum = Mass Flow * the change in flow velocity from freestream (true airspeed). The higher the exit flow velocity the lower the change in velocity and therefore the lower the change in momentum = less drag. Unfortunately, the pursuit for less drag through smaller exit area will also reduce cooling capacity. For a fixed geometry exit a tradeoff between drag and cooling must be made. For this reason I chose to fabricate an exit flap allowing me to increase the exit area to nearly stock dimensions or reduce the area to less than half. This flap is controlled by a Ray Allen trim servo connected to a switch on my stick grip with position info displayed on the VP-50 electrical system panel.
The RV8 stock exit area, is designed for the worst case cooling condition of WOT climb at Vy which means too much mass flow exists during high speed level flight leading to excessive drag. During climb I open the flap providing full cooling mass flow while in level flight the flap is closed to reduce drag - not a new concept as cowl flaps have been around forever. Additionally, I applied the same concept of exit throttle to the OC. Most folks seem to prefer blocking the inlet and some believe this is more effective than blocking the exit. This idea doesn't match theory nor experimental data. If we block the inlet to the OC we will certainly reduce mass flow and temps will increase but with a drag penalty as the air will exit the OC slower than it entered and will not aid in increasing overall cowl exit velocity. Now if we throttle the OC exit with a nozzle we will both reduce the mass flow to increase OT AND increase OC exit velocity which again reduces drag. It matters not from a purely cooling perspective which side of the OC you block but if you want drag reduction throttle the exit and accelerate the flow. My OC throttle reduces the the exit to 4" sq with an inlet area of 7" sq. OT without the OC blocked typically remained in the 160-170 range. With the inlet blocked 50% OT were in the 185-195 range. Now with the OC exit throttled to 4"sq OT are in the 185-195 range but with greater exit velocity.
Pics of the OC exit throttle:
Results:
Using a fixed W/delta technique I measured a TAS increase of 6MPH with the exit flap closed and nozzle installed over the stock cowl configuration and a 7MPH increase from the exit flap open condition (the exit flap produces some drag itself when open).
Here is the data:
Cooling: I'll provide more info on cooling when the full test is completed but for now I'll say the exit flap works as designed in that during climb when flap is open my Avg CHTs are just slightly higher than stock (even with the flap open the exit area is a bit smaller than stock) but during WOT level flight they avg low 300s with exit closed. A WOT climb at Vy with the exit flap closed results in CHTs rising about 20-30 deg hotter than with it open as expected.
Thanks to Dan for fabricating and sending the pressure probes and many discussions with he and Sonny along the way. The full test will be full of good information. I tried to keep the math out of this posting but if there is interest I can submit more detail on how the data was reduced.