Exploring the Low Pressure Zone Behind a Car: What Causes Drag?

[noagname]

I watch a lot of mythbusters and they constantly talk about the low pressure zone behind a car. I'm confused how does that create drag?

 

[AtomicJoe]

Well I would say that because the pressure is higher at the front than at the rear then a force pushes the car back, thus if you can reduce that drag you can increase the speed.

 

[rcgldr]

As as car travels through a volume of air, at the back end of the car, it introduces what would be a void if the air didn't accelerate to fill in that "void". The air has momentum and viscosity, so it can't instantly fill in that moving "void", and the end result is a low pressure area that accelerates the air mostly forwards and somewhat inwards. At the front of the car there's a higher pressure area, but a lot of the air flow will separate and flow around the car, so the increase in pressure at the front of the car is not as much as the decrease in pressure at the rear of the car.

The pressure difference, higher in front, lower in rear, creates a net backwards force on the car called drag. It also causes the affected air to be accelerated forwards, which is why you see leaves on a road being blown forward after a vehicle passes by.

If the car is streamlined, such as a tapered tail, more of the acceleration at the rear of the car is inwards instead of forwards, and this reduces the drag. This is why land speed vehicles like the ones run at Bonneville salt flats have long tapered tails. The front isn't as much of an issue, so they just have rounded noses.

 

[boneh3ad]

Well I would say that because the pressure is higher at the front than at the rear then a force pushes the car back, thus if you can reduce that drag you can increase the speed.

This.

As as car travels through a volume of air, at the back end of the car, it introduces what would be a void if the air didn't accelerate to fill in that "void". The air has momentum and viscosity, so it can't instantly fill in that moving "void", and the end result is a low pressure area that accelerates the air mostly forwards and somewhat inwards. At the front of the car there's a higher pressure area, but a lot of the air flow will separate and flow around the car, so the increase in pressure at the front of the car is not as much as the decrease in pressure at the rear of the car.

The pressure difference, higher in front, lower in rear, creates a net backwards force on the car called drag. It also causes the affected air to be accelerated forwards, which is why you see leaves on a road being blown forward after a vehicle passes by.

If the car is streamlined, such as a tapered tail, more of the acceleration at the rear of the car is inwards instead of forwards, and this reduces the drag. This is why land speed vehicles like the ones run at Bonneville salt flats have long tapered tails. The front isn't as much of an issue, so they just have rounded noses.

Sort of. The shape of the front plays a large role in how high that pressure in the front is (and over what area it is acting). You also didn't really explain the reason for the drop in pressure correctly, especially how a streamlined rear-end helps. With a streamlined rear, the flow can more easily remain attached to the car, and while the flow in the back will still be lower in pressure than the front, at least it will be fairly reasonable. If the back is blunt, it creates a pressure gradient so large that it actually causes the flow to "separate" from the surface of the car and essentially form a large circulating bubble of air that has very low pressure relative to the rest of the vehicle resulting in a much larger drag force. This same phenomenon (separation) is what leads to stall on airfoils.

 

[rcgldr]

The shape of the front plays a large role in how high that pressure in the front is (and over what area it is acting).

But for sub-sonic land speed vehicles a rounded nose is usually good enough, but a large tapered tail section is common in some classes of vehicles. Land speed motorcycles often look like stretched torpedoes.

With a streamlined rear, the flow can more easily remain attached to the car, and while the flow in the back will still be lower in pressure than the front, at least it will be fairly reasonable. If the back is blunt, it creates a pressure gradient so large that it actually causes the flow to "separate" from the surface of the car and essentially form a large circulating bubble of air that has very low pressure relative to the rest of the vehicle resulting in a much larger drag force. This same phenomenon (separation) is what leads to stall on airfoils.

At the macroscopic level, drag is related to the rate of energy being added to the air by a moving object, most of which is related to how much the air is accelerated forwards by the vehicle. The tapered tail introduces the "void" gradually into the air, allowing the air to fill in that "void" by accelerating mostly inwards and less forwards, which reduces the net forwards acceleration of air, which reduces the drag. That gradual introduction also keeps the flow reasonably attached as you mentioned.

In the real world, except at very flow speeds, the flow will detach somewhat and transition into turbulent flow, for anything the size of a car and moving at 100 kph or so. The reason a blunt rear end results in so much reduction in pressure isn't directly due to the turbulent vortices, but instead to the fact that the air further aft of those vortices is accelerated forwards to fill in the "void" aft of those turbulent vortices, and the pressure reduction corresponds to the rate of forwards acceleration of air. For a vehicle with a blunt rear end, using "wings" on the sides of the car near the rear to direct the flow inwards into that "void" can help reduce drag, but not as much as using a tapered tail.

 

[boneh3ad]

But for sub-sonic land speed vehicles a rounded nose is usually good enough, but a large tapered tail section is common in some classes of vehicles. Land speed motorcycles often look like stretched torpedoes.

That is certainly true. It has a role but it isn't a huge one. In fact, because of the fact that the front end has a grill and is taking in air, it is both necessary to have a blunt nose to a degree and the fact that there is penetration means it isn't a true stagnation point, so that helps out as well. It is certainly the case that having a pointed nose isn't going to give you much on any normal car.

At the macroscopic level, drag is related to the rate of energy being added to the air by a moving object, most of which is related to how much the air is accelerated forwards by the vehicle. The tapered tail introduces the "void" gradually into the air, allowing the air to fill in that "void" by accelerating mostly inwards and less forwards, which reduces the net forwards acceleration of air, which reduces the drag. That gradual introduction also keeps the flow reasonably attached as you mentioned.

In the real world, except at very flow speeds, the flow will detach somewhat and transition into turbulent flow, for anything the size of a car and moving at 100 kph or so. The reason a blunt rear end results in so much reduction in pressure isn't directly due to the turbulent vortices, but instead to the fact that the air further aft of those vortices is accelerated forwards to fill in the "void" aft of those turbulent vortices, and the pressure reduction corresponds to the rate of forwards acceleration of air. For a vehicle with a blunt rear end, using "wings" on the sides of the car near the rear to direct the flow inwards into that "void" can help reduce drag, but not as much as using a tapered tail.

Be careful not to equate turbulence and vortices. I realize that isn't likely what you are doing here, but I feel like that needs to be pointed out for anyone reading this. They are certainly related, but they aren't the same.

Regardless, the way you explain this "void" is somewhat strange. If you had a truly tapered body such as in a solar car, there will be no separation. In reality, for an average car, there isn't a separation bubble until you leave the body of the car, as modern cars are generally aerodynamic enough to prevent that until the air passes over the read edge of the trunk. In general, the flow isn't even going to be very turbulent on most cars until pretty far back on the body. The whole forebody has a strong favorable pressure gradient which both opposes any sort of separation and retards the growth of T-S waves, which are the dominant mode of laminar-turbulent transition in the case of an average car. If you look at videos of wind tunnel testing you can confirm that the flow remains attached until it leaves the body and that it is not turbulent for most of the way over the body.

I don't ever remember associating the form drag with the "turbulent vortices" behind the car, but rather the separation bubble, which is an area of flow that is generally slowly circulating but fairly stagnant compared to the rest of the air moving over the car (in the car's reference frame). That is the same phenomenon you are describing, just in a different reference frame. In the inertial frame, that bubble would seem to be dragged by the car.

 

[Ranger Mike]

see post dated Dec31-08, 10:58 AM by
Ranger Mike title of post is Drafting
may help explain

 

[rcgldr]

Regardless, the way you explain this "void" is somewhat strange.

"Void theory" - at one moment a solid object is occupying a specific space within the air, and at a later moment that solid object has moved on and is gone from the space it once occupied, leaving what would otherwise be a void if the air didn't fill in that space once occupied by the solid object.

We probably went beyond what the OP was asking for anyway, since he just wanted to know why the low pressure aft of a car correlated to drag.

 

[Ranger Mike]

Imagine that you are in a room that has an opening you wish to walk through. The Opening is covered by two curtains. These two curtains are hung in such a way that they overlap. In order to walk between them you must push each out off the way of your path. This requires some effort ( work). Add to this the fact that these curtains are relatively long and heavy. Now consider that any time they are disturbed, it takes some amount of time for each to settle to a state of rest.

This is what we have to deal with when racing. There is a curtain of air that weights 14.7 pounds per inch. A race car at speed has to push its way through this curtain of air. As the front of the car exerts pressure on the curtain ( high pressure) it moves the curtain " out of the way" and as it passes through the curtain, it takes a small amount of time for the curtain to return to its original state of rest it was in before you pushed on it. The area in back of the race car at speed now has a low pressure area that sucks in debris, water particles, dirt,dust because it has low pressure. That's why you see the back side of the over the road semi tractor trailers pretty dirty form dust road crap and the like...low pressure...This is also why you see some semi trucks with a wing on top of the cab. It is an effort to reduce drag.

Any time you can stream line or smooth out the air flow ( caused by pushing the curtain out of the way,) moving through the media ( air) and smooth the transition of reconnecting the separated streams ( think of the two curtains pushed out of the way) you will reduce drag. So vortex strakes ( 2% reduction of drag) , boat tail plates ( 6% reduction) etc... on the rear portion of the trailer are an attempt to smooth out the transition and thus reduce drag.
btw, 90% of parasitic drag is pressure drag, 8% is friction drag and 2% is roughness drag.

 

[AtomicJoe]

But for sub-sonic land speed vehicles a rounded nose is usually good enough, but a large tapered tail section is common in some classes of vehicles. Land speed motorcycles often look like stretched torpedoes.

What really matters is the frontal area of the vehicle, ie the shape it would punch out it if it drove through a a wall.

The problem with the long nose is another type of drag, I think it is called viscous drag, that is the air dragging on the long nose.

For the same reason long tails are not much good either.

I do note that I do not see racing cars with long tails!

 

[AlephZero]

I don't ever remember associating the form drag with the "turbulent vortices" behind the car, but rather the separation bubble, which is an area of flow that is generally slowly circulating but fairly stagnant compared to the rest of the air moving over the car (in the car's reference frame). That is the same phenomenon you are describing, just in a different reference frame. In the inertial frame, that bubble would seem to be dragged by the car.

I used to own a hatchback that did a nice demonstration of that, when driving in rain.

A horizontal stationary "line" of water would form part way up the rear window. Clearly it was being blown upwards, by the circulation from the air leaving the top of the car. The rest of the rear window would be completely dry, since the flow pattern stopped any rain reaching it.

The "line" moved vertically up and down the window depending on the car's speed, in a range from about 40 to 70 mph.

I now have a different car, same size body and engine but considerably more fuel efficient (better than 50mpg at a steady 70 mph on a level road). I now get a uniform spray of water all over the rear window. Presuably they tweaked roof shape so the "bubble" circulation flow (like a typical textbook drawing of flow over a backward-facing step) has been replaced "random" turbulence, and less drag.

 

[boneh3ad]

In a situation like that, AlephZero, even a minute change in shape can make all the difference. Changing that expansion angle as the air leaves the back of your car by making the roof slope back a little bit more or the back slope down little bit less would have noticeable effects.

The other way to keep the flow attached is by tripping it to turbulence before it encounters that adverse pressure gradient. Turbulent boundary layers are much more robust in that respect. That is the reason for the dimples on golf balls. It is an interesting problem.

 

[RandomGuy88]

The other way to keep the flow attached is by tripping it to turbulence before it encounters that adverse pressure gradient. Turbulent boundary layers are much more robust in that respect. That is the reason for the dimples on golf balls. It is an interesting problem.

When tripping the boundary layer you have to be careful about where you trip it though. A turbulent boundary layer is thicker and grows much faster so it is possible that a turbulent boundary layer can separate fairly easily. This happens on airfoils and wings if you trip the flow to far upstream.

 

[rcgldr]

What really matters is the frontal area of the vehicle, ie the shape it would punch out it if it drove through a a wall.

A good tear drop shape can be several times thicker than a cylinder (moving sideways so that the air foil is a circle) and still have less drag.

I do note that I do not see racing cars with long tails!

The rules don't allow it anymore and in the case of high downforce cars like Formula 1, (900 hp, 1300 lb car with driver), the drag factor is more than double a typical street car, but it doesn't matter, since downforce is the primary goal.

http://en.wikipedia.org/wiki/Automobile_drag_coefficient

The pre-downforce racing cars like Mercedes Benz 154 and 196 had somewhat long tails. As already mentioned the Bonneville land speed type motorcycles look like lengthened torpedoes, like this BUB streamliner:

http://www.knfilters.com/images/press/ccarr1.jpg

 

[boneh3ad]

A good tear drop shape can be several times thicker than a cylinder (moving sideways so that the air foil is a circle) and still have less drag.

That is because a teardrop shape won't experience von Karman shedding. You could compare a circular cylinder to an ellipsoid cylinder and still have a very nice drag reduction because of the change in the separation location without having an incredibly elongated tails. At some point you can elongate it too much because the added viscous drag will be greater than the savings from form drag. You only need to make the back end aerodynamic enough to prevent (or delay) separation.

 

[rcgldr]

You only need to make the back end aerodynamic enough to prevent (or delay) separation.

Then why the very long tails on streamliners that go 350 mph versus the relatively shorter wing chords on mach .8 civilian jet aircraft? The explanation I recall for the long tails is so that the air flow is mostly perpendicular to the vehicle's direction, outwards as it separates at the front, inwards as it converges along the tail, with the long tail increasing the ratio of inwards versus forwards flow of the air. These tails are much longer than the minimum needed to prevent separation of flow. Again look at this BUB streamliner:

http://www.knfilters.com/images/press/ccarr1.jpg

 

[RandomGuy88]

Perhaps the tail is so long because that extra length is need to accommodate internal features such as propulsion.

For a body as streamlined as the BUB most of the drag is probably skin friction. But the drag coefficient of the BUB is actually quite high compared to a lot of aircraft.

 

[rcgldr]

Perhaps the tail is so long because that extra length is need to accommodate internal features such as propulsion.

It's a hollow tail section, the only thing going through there in the case of the BUB is the exhaust pipe. The engine, drivetrain, and rider are located between the tires. Their website claims it had the lowest coefficient of drag ever tested at the wind tunnel site they use: 0.08 (look at the specs), which is high compared to aircraft, but better than all but a few experimental vehicles. Also the BUB needs to generate some amount of downforce to make sure it doesn't become airborne at 350+ mph.

http://www.seven-streamliner.com

Better than the Sunraycer (CD = 0.117):

http://en.wikipedia.org/wiki/Sunraycer

and in the same league as the experimental Nuna solar powered vehicle (CD = 0.07):

http://en.wikipedia.org/wiki/Nuna

 

[noagname]

Thanks everyone that really answered my question.

 

[boneh3ad]

It's a hollow tail section, the only thing going through there in the case of the BUB is the exhaust pipe. The engine, drivetrain, and rider are located between the tires. Their website claims it had the lowest coefficient of drag ever tested at the wind tunnel site they use: 0.08 (look at the specs), which is high compared to aircraft, but better than all but a few experimental vehicles. Also the BUB needs to generate some amount of downforce to make sure it doesn't become airborne at 350+ mph.

http://www.seven-streamliner.com

Better than the Sunraycer (CD = 0.117):

http://en.wikipedia.org/wiki/Sunraycer

and in the same league as the experimental Nuna solar powered vehicle (CD = 0.07):

http://en.wikipedia.org/wiki/Nuna

Cars don't usually need a drag coefficient that low. For the BUB, it is an ultra-high speed vehicle so they need to reduce drag as much as possible. For any Average Joe driving their Ford Fiesta down the highway, there isn't going to be that big of an advantage in streamlining it further, after all, drag scales with the square of velocity. Solar cars need a tiny drag coefficient because they run on solar power, which means they don't have very much horsepower.

Honestly, I can't say why the BUB has such a ridiculous length. The designers must have had something in mind. I would imagine that the $C_{p}$ distribution on it was just very nice so they went with it. The important thing is that they could easily have had zero separation with a much less elongated body. There were other motives involved.

 

[rcgldr]

Honestly, I can't say why the BUB has such a ridiculous length. The important thing is that they could easily have had zero separation with a much less elongated body.

Yet the long lengths are common for these type of vehicles, in spite of the fact that it means they can't be driven if crosswinds are significant.

http://www.goldwingracingproject.info

http://www.streamliner.com

 

[boneh3ad]

That is because the reason clearly isn't only about separation control. There has to be something else at work. It could be to try and get internal components fit in correctly or exhaust ducted correctly. I don't honestly know as I don't work on those vehicles to know all the needs. All I know is that you don't need such a long afterbody to reduce or eliminate separation.

 

[AlephZero]

I wonder if directional stability is the reason. The big difference between a high speed car and a plane is the wheels.

If the CP is in front of the driving wheels it will tend to yaw. That's not a big problem for a conventional RWD road car, but 350 mph means 10x the aero side forces compared with 100mph and 3.5x less time for the driver to correct it.

Some commercial passenger plane designs turned out to be unstable in yaw when flying straight. The fix was usually to make the autopilot fly them consistently yawing to one side (a fraction of a degree is enough to kill the problem), otherwise the passengers in the back rows started to get airsick from the side to side oscillations even if they were too small to be felt consciously.

 

[rcgldr]

I wonder if directional stability is the reason. If the CP is in front of the driving wheels it will tend to yaw.

Note the goldwing streamliner has it's rear (driving) wheel near the back end of the vehicle, without much tail section remaining. I'm not sure where the CP needs to be relative to the steering and driving wheels for the vehicles to remain stable. Both yaw and roll stability would be affected by the CP. Some downforce is used for pitch stability and to keep the vehicles from blowing over or going airborne.

That is because the reason clearly isn't only about separation control. There has to be something else at work.

A long gradually tapered tail gives the air more time to react (less acceleration) towards the receding tail section of the vehicle as it passes through a volume of air, so there is less reduction in pressure.

 

[cjl]

Then why the very long tails on streamliners that go 350 mph versus the relatively shorter wing chords on mach .8 civilian jet aircraft? The explanation I recall for the long tails is so that the air flow is mostly perpendicular to the vehicle's direction, outwards as it separates at the front, inwards as it converges along the tail, with the long tail increasing the ratio of inwards versus forwards flow of the air. These tails are much longer than the minimum needed to prevent separation of flow. Again look at this BUB streamliner:

http://www.knfilters.com/images/press/ccarr1.jpg

Because the cars are wider than the thickness of a jet airplane wing?

For flow separation, the relevant parameter is the thickness to chord ratio, not the length alone. On airliners, the thickness to chord ratio is around 11-13%, so the dimensions of the streamliners look about right. From the dimensions I can find online, the bub streamliner appears to have a thickness to chord ratio around 9%, which is a bit longer than most airliner wings, but not a whole lot. It's certainly within a reasonable range. Keep in mind that airliner wings tend to be a bit thicker because of strength, as well as fuel capacity (fuel is held in the wings). Thin wings carry less fuel, and tend to be heavier due to structural concerns.

AlephZero: I'd love to see a citation on that yaw instability. To the best of my knowledge, no passenger plane is unstable in yaw (though some of them are very lightly damped in some modes), and flying with a slight sideslip would cause a rather monumental increase in drag, which is very undesirable (obviously).

 

[AlephZero]

I was thinking particularly about one of the McDonnel Douglas planes (but I forget which one, sorry). I've seen a reference in a book on control systems which said there were effectively three equilibruim positions close together, two stable ones either side of zero yaw and an unstable one at zero.

The Piper Navaho was very reluctant to fly with zero yaw as well, though I don't know if it was actually unstable or just underdamped, But as one of our company pilots used to say, the animal of the Navaho is the snake, so how would you expect a plane named after them to fly?

 

[boneh3ad]

A long gradually tapered tail gives the air more time to react (less acceleration) towards the receding tail section of the vehicle as it passes through a volume of air, so there is less reduction in pressure.

That's not how fluid mechanics works. The overall reduction in pressure would likely not be much (if any) different; only the pressure gradient would change noticeably.