Aerodynamics for turbocharged and supercharged RX8s
10-06-2015, 09:37 AM
#1
Registered
Thread Starter
Aerodynamics for turbocharged and supercharged RX8s
In a previous life I was an air force engineer and did a lot of aerodynamic work. I'm therefore putting some thought and effort into the aerodynamic aspects of my 13B REW conversion, for example ducting air for cooling, front splitter design and rear wing design.
What I am doing is overkill for anyone doing an engine swap but I thought my work might be interesting reading and help others. I therefore intend to put write articles for the main parts of the forum covering:
General aerodynamic considerations.
Lift and drag reduction measures suitable for a fast RX8.
Cooling, including my measurement and assessment of pressure distributions over RX8 bonnet/hoods (I've already posted an initial assessment on the UK club's site, copied below, and want to update it with further measurements before posting in the main section).
Rear wing positioning and design (I want to design and make my own wing and have already started an analysis of air flows over my R3's boot/trunk).
Please will anyone who has carried out their own analysis or design tell me what they've done?
What I am doing is overkill for anyone doing an engine swap but I thought my work might be interesting reading and help others. I therefore intend to put write articles for the main parts of the forum covering:
General aerodynamic considerations. Lift and drag reduction measures suitable for a fast RX8. Cooling, including my measurement and assessment of pressure distributions over RX8 bonnet/hoods (I've already posted an initial assessment on the UK club's site, copied below, and want to update it with further measurements before posting in the main section). Rear wing positioning and design (I want to design and make my own wing and have already started an analysis of air flows over my R3's boot/trunk).Please will anyone who has carried out their own analysis or design tell me what they've done?
10-06-2015, 09:38 AM
#2
Registered
Thread Starter
I’ve been doing a fair amount of DIY aerodynamic work for my project car and thought I’d share my results so far.
Background
I’m redesigning my project car’s engine support systems (engine cooling, oil cooling, charge air cooling, air inlet, fuel system etc) and decided to rekindle my past experience and education in all things aerodynamic by being scientific rather than just copying someone else’s work, particularly as there are so many horrible arrangements out there.
I’ll only explain the aerodynamic issues involved where I think they’re particularly relevant. For a general primer on car aerodynamics the internet is your friend, for example the article here.All you need to know for now is that air passing round and over a car generates drag and lift (the opposite of downforce) that we want to minimise and these forces vary with the square of our car’s speed (eg doubling your speed quadruples their value).
My recent investigations
I analysed the pressures over my R3’s bonnet and bumper in order to find the best places to take in and exhaust air for my various needs, particularly if my existing carbon bonnet grilles were suitable locations (albeit still needing some carbon cutting) - see images below for library images of my bonnet. To see what those needs are then read on, otherwise if you just want to see my results then jump to the next section.
For more background, a car’s cooling systems take in fast air and slow it right down, a bit like holding a flat plate into the airstream, and generate typically one third of a car’s drag. This drag, and to a lesser extent lift from higher engine bay pressures, can be reduced by:
• Using ducting to take air in and out of the cooling matrices (radiators, intercoolers and coolers).
• Slowing down only as much air as needed (most cars only use some 1/3 of the air hitting a cooler’s inlet, the rest creating drag as it’s slowed down and left to find a way around the car). In broad terms, a cooler should have ducting that takes air from an inlet ¼ of the cooler height at its height ahead of it then exhaust through an outlet 115% of the inlet size at one height behind the cooler.
• Exhausting air from the cooling matrices smoothly into the air flow around a low pressure area around the car (most cars, like the RX8, dump air from their radiator into the engine bay then under the car, where it bounces around creating drag and lift).
• Ensuring that the exhausts do not cause separation or other unwanted consequences.
• Limiting the air passing into the engine bay.
This aerodynamic investigation fits in with my general aims, which are to:
• Take inlet air from the lowest point on the front bumper, which is where the RX8 takes its radiator air, for the cooling systems and turbo air inlet.
• Take radiator and intercooler exhaust out of the bonnet at an angle close to that of the bodywork, preferably at the location of the grille in my carbon fibre bonnet so that I can keep it without too much modification. However, it seems that I will, after all, have to duct a large part of the radiator exhaust into the engine bay and over the engine
• Run some cooling air over the turbo and between the turbo and inlet manifold, exhausting it out of either the rear grille vents in my bonnet or the existing side exhausts (there mostly for styling).
• Make the most of my limited options for the oil coolers, although after much head-scratching I’m coming to the conclusion that space limitations mean that I may have to stick with their current location. Note that dumping exhaust at an angle in the front wheel well isn’t the worst option but the existing ducting is pretty crude and far from ideal; for example, the outlet slats are too small, probably a compromise to reduce soiling from the wheels, which is why owners who track cars or live in hot climates open them up to reduce oil temperatures (see the mod here.).
Test Method
I used plastic tubing to take static pressures from various locations on the my R3, setting the open end at right angles to the local air flow, to Magneholic differential pressure gauges set in a jig with my phone. I then set my phone video to record and drove my car on local roads, motorways and private roads at speeds of up to 80 (private road, honest) in 10 mph intervals and analysed the videoed results at home. Weather conditions and road traffic limited the time that I could spend testing and much of what I did was too inaccurate due to passing traffic, winds and inadvertent climate control use (it raises the cabin pressure easily by 20-40 Pa!). Furthermore, my initial results did not produce the smooth curve that I’d expected so I reran with screw-in restrictors to damp down pressure fluctuations and hence reduce estimation errors; in most cases this gave results very close to my initial readings and only confirmed the accuracy of my earlier runs. The pictures show the gauges and a tube fixed in place; I generally ran with 4 hoses at one time.
Results
The chart shows the pressure distribution measurement on the centreline by distance from the bonnet’s front edge. The floor tray at the bottom of the bumper is at -60 and the bonnet ends at 102. Cp is the ratio of the pressure difference expressed as a proportion of the dynamic pressure at that speed; positive values represent a positive force on that surface and negative values represent a force in the opposite direction (for our bonnet that means low pressures and lift).
This pressure distribution is broadly similar to that of the RX7 and other curvy coupes as shown below.
Other results were:
• The values of Cp at the rear of the bonnet showed small but consistent increases in Cp with speed that reflected changes in the shape of the air flow with speed.
• The values of Cp were consistent across the bonnet until falling away slightly within 150 mm of the bonnet edges.
• The undersurface of the bonnet had pressures the same Cp at all points and varying from 0.10 at 20 mph to 0.14 at 80 mph
• The side vent behind the front wings had a Cp of a steady zero at all speeds.
• The air vents at the base of the windscreen had values of Cp between 0 and 0.02, showing that the base of the windscreen is a reasonable place to take air for the interior.
Application of results to my Project Car
Existing Main Bonnet Vent. The main vent in my carbon bonnet is shaped to match the curve of the bonnet front lip and sides. Due to restrictions under the bonnet, its ductable area can be approximated to a rectangle 300 mm long (240 mm if the rear carbon slope is retained) and 600 mm wide starting at 28 cm along the bonnet. The chart shows that this ductable area is within a useable area of low pressure and calculations give it an average Cp of approximately -0.37. The grille outside this ductable area has bonnet pressures much lower than under-bonnet pressures with a Cp of 0.49 in the right direction. The grille is therefore a reasonable location to exhaust air and should be cut forward if it has to be lengthened.
Existing Rear Side Vents. The rear side vents are effectively over the rear corners of the engine bay. They can be approximated as rectangles 190 mm long (150 mm if the rear carbon slope is retained) and 125 mm wide beginning 78 cm along the bonnet. Although they were better than I’d expected, my results confirmed the general truism that grilles at the rear of a bonnet are a poor place to vent the engine bay. Their location near the edges of the bonnet gave them a small negative Cp of -0.03; a similar grille on the bonnet centre line would have a Cp of 0 from a small negative pressure (Cp of -.08) across the front half and a positive pressure (Cp of 0.09) across the rear half. We are more interested in the difference between the bonnet and engine bay pressures, and here the grilles give a more pleasing differential Cp of 0.16 in the right direction (0.13 for a rear centre vent), even if considerably lower than the ducts at the main grille.
Side Vents. The side vents have a lower pressure than the engine bay with a Cp of 0.12 in the right direction. They can therefore be used to take hot air from the engine bay but are a poor second to the bonnet grilles, not only because of their lower pressure differentials but also because the possible exit area is smaller and the route there is convoluted.
Turbo Compressor Air Inlet. The negative pressures across the bonnet show it is no place to take air for the compressor air inlet. The engine bay itself is more encouraging, but this still leaves the main problem that its high temperature isn’t helpful. The air intake should therefore be taken low on the front bumper, its size to be determined by further analysis.
Background
I’m redesigning my project car’s engine support systems (engine cooling, oil cooling, charge air cooling, air inlet, fuel system etc) and decided to rekindle my past experience and education in all things aerodynamic by being scientific rather than just copying someone else’s work, particularly as there are so many horrible arrangements out there.
I’ll only explain the aerodynamic issues involved where I think they’re particularly relevant. For a general primer on car aerodynamics the internet is your friend, for example the article here.All you need to know for now is that air passing round and over a car generates drag and lift (the opposite of downforce) that we want to minimise and these forces vary with the square of our car’s speed (eg doubling your speed quadruples their value).
My recent investigations
I analysed the pressures over my R3’s bonnet and bumper in order to find the best places to take in and exhaust air for my various needs, particularly if my existing carbon bonnet grilles were suitable locations (albeit still needing some carbon cutting) - see images below for library images of my bonnet. To see what those needs are then read on, otherwise if you just want to see my results then jump to the next section.
For more background, a car’s cooling systems take in fast air and slow it right down, a bit like holding a flat plate into the airstream, and generate typically one third of a car’s drag. This drag, and to a lesser extent lift from higher engine bay pressures, can be reduced by:
• Using ducting to take air in and out of the cooling matrices (radiators, intercoolers and coolers).
• Slowing down only as much air as needed (most cars only use some 1/3 of the air hitting a cooler’s inlet, the rest creating drag as it’s slowed down and left to find a way around the car). In broad terms, a cooler should have ducting that takes air from an inlet ¼ of the cooler height at its height ahead of it then exhaust through an outlet 115% of the inlet size at one height behind the cooler.
• Exhausting air from the cooling matrices smoothly into the air flow around a low pressure area around the car (most cars, like the RX8, dump air from their radiator into the engine bay then under the car, where it bounces around creating drag and lift).
• Ensuring that the exhausts do not cause separation or other unwanted consequences.
• Limiting the air passing into the engine bay.
This aerodynamic investigation fits in with my general aims, which are to:
• Take inlet air from the lowest point on the front bumper, which is where the RX8 takes its radiator air, for the cooling systems and turbo air inlet.
• Take radiator and intercooler exhaust out of the bonnet at an angle close to that of the bodywork, preferably at the location of the grille in my carbon fibre bonnet so that I can keep it without too much modification. However, it seems that I will, after all, have to duct a large part of the radiator exhaust into the engine bay and over the engine
• Run some cooling air over the turbo and between the turbo and inlet manifold, exhausting it out of either the rear grille vents in my bonnet or the existing side exhausts (there mostly for styling).
• Make the most of my limited options for the oil coolers, although after much head-scratching I’m coming to the conclusion that space limitations mean that I may have to stick with their current location. Note that dumping exhaust at an angle in the front wheel well isn’t the worst option but the existing ducting is pretty crude and far from ideal; for example, the outlet slats are too small, probably a compromise to reduce soiling from the wheels, which is why owners who track cars or live in hot climates open them up to reduce oil temperatures (see the mod here.).
Test Method
I used plastic tubing to take static pressures from various locations on the my R3, setting the open end at right angles to the local air flow, to Magneholic differential pressure gauges set in a jig with my phone. I then set my phone video to record and drove my car on local roads, motorways and private roads at speeds of up to 80 (private road, honest) in 10 mph intervals and analysed the videoed results at home. Weather conditions and road traffic limited the time that I could spend testing and much of what I did was too inaccurate due to passing traffic, winds and inadvertent climate control use (it raises the cabin pressure easily by 20-40 Pa!). Furthermore, my initial results did not produce the smooth curve that I’d expected so I reran with screw-in restrictors to damp down pressure fluctuations and hence reduce estimation errors; in most cases this gave results very close to my initial readings and only confirmed the accuracy of my earlier runs. The pictures show the gauges and a tube fixed in place; I generally ran with 4 hoses at one time.
Results
The chart shows the pressure distribution measurement on the centreline by distance from the bonnet’s front edge. The floor tray at the bottom of the bumper is at -60 and the bonnet ends at 102. Cp is the ratio of the pressure difference expressed as a proportion of the dynamic pressure at that speed; positive values represent a positive force on that surface and negative values represent a force in the opposite direction (for our bonnet that means low pressures and lift).
This pressure distribution is broadly similar to that of the RX7 and other curvy coupes as shown below.
Other results were:
• The values of Cp at the rear of the bonnet showed small but consistent increases in Cp with speed that reflected changes in the shape of the air flow with speed.
• The values of Cp were consistent across the bonnet until falling away slightly within 150 mm of the bonnet edges.
• The undersurface of the bonnet had pressures the same Cp at all points and varying from 0.10 at 20 mph to 0.14 at 80 mph
• The side vent behind the front wings had a Cp of a steady zero at all speeds.
• The air vents at the base of the windscreen had values of Cp between 0 and 0.02, showing that the base of the windscreen is a reasonable place to take air for the interior.
Application of results to my Project Car
Existing Main Bonnet Vent. The main vent in my carbon bonnet is shaped to match the curve of the bonnet front lip and sides. Due to restrictions under the bonnet, its ductable area can be approximated to a rectangle 300 mm long (240 mm if the rear carbon slope is retained) and 600 mm wide starting at 28 cm along the bonnet. The chart shows that this ductable area is within a useable area of low pressure and calculations give it an average Cp of approximately -0.37. The grille outside this ductable area has bonnet pressures much lower than under-bonnet pressures with a Cp of 0.49 in the right direction. The grille is therefore a reasonable location to exhaust air and should be cut forward if it has to be lengthened.
Existing Rear Side Vents. The rear side vents are effectively over the rear corners of the engine bay. They can be approximated as rectangles 190 mm long (150 mm if the rear carbon slope is retained) and 125 mm wide beginning 78 cm along the bonnet. Although they were better than I’d expected, my results confirmed the general truism that grilles at the rear of a bonnet are a poor place to vent the engine bay. Their location near the edges of the bonnet gave them a small negative Cp of -0.03; a similar grille on the bonnet centre line would have a Cp of 0 from a small negative pressure (Cp of -.08) across the front half and a positive pressure (Cp of 0.09) across the rear half. We are more interested in the difference between the bonnet and engine bay pressures, and here the grilles give a more pleasing differential Cp of 0.16 in the right direction (0.13 for a rear centre vent), even if considerably lower than the ducts at the main grille.
Side Vents. The side vents have a lower pressure than the engine bay with a Cp of 0.12 in the right direction. They can therefore be used to take hot air from the engine bay but are a poor second to the bonnet grilles, not only because of their lower pressure differentials but also because the possible exit area is smaller and the route there is convoluted.
Turbo Compressor Air Inlet. The negative pressures across the bonnet show it is no place to take air for the compressor air inlet. The engine bay itself is more encouraging, but this still leaves the main problem that its high temperature isn’t helpful. The air intake should therefore be taken low on the front bumper, its size to be determined by further analysis.
The following users liked this post:
Warrior777 (08-02-2019)
01-09-2016, 04:35 PM
#5
Registered
Thread Starter
I attach a draft write-up an example of the ones I'm doing. I've not checked it yet so please see it as a work in progress.
All comments welcome.
All comments welcome.
The following users liked this post:
Warrior777 (08-02-2019)
02-09-2016, 09:01 AM
#6
Registered
Join Date: Jul 2012
Location: Melbourne, Australia
Posts: 39
Likes: 0
Received 0 Likes
on
0 Posts
Awesome read. And here I was, thinking about opening up the front grill where the license plate sits to allow more air to go into the radiator!! I have much to learn!
02-09-2016, 12:02 PM
#7
This is what I like about RX-8s.. they attract engineery types and stuff like this happens.
Question: if the pressure difference across the stock side vents is minimal, does it follow that the airflow through them is small? I ask because 0 pressure differential could imply that flow out of the side vent and flow over the side vent (as in... outside air) are moving at roughly the same speed (that is vehicle speed). But I'm an aerodynamics noob.
Question: if the pressure difference across the stock side vents is minimal, does it follow that the airflow through them is small? I ask because 0 pressure differential could imply that flow out of the side vent and flow over the side vent (as in... outside air) are moving at roughly the same speed (that is vehicle speed). But I'm an aerodynamics noob.
02-10-2016, 05:37 AM
#8
Registered
Thread Starter
IMO the air flow through them is indeed small, due to the relatively low pressure differential between the engine bay and grille exterior, the convoluted route air has to go through to get there and the small inlet in the engine bay wall.
I looked at improving air flow through the grilles but decided that it was too much effort for the gain involved and opening up the bodywork would affect stiffness for the front suspension and crash crumple zones.
IMO the grilles are a just a styling feature.
I looked at improving air flow through the grilles but decided that it was too much effort for the gain involved and opening up the bodywork would affect stiffness for the front suspension and crash crumple zones.
IMO the grilles are a just a styling feature.
Thread
Thread Starter
Forum
Replies
Last Post
