[Sutripta]

Is any research done for the case the car is not in still air, but in a wind. Not galeforce. In one of the cars (Alto K10 first model) internal cabin noise would increase significantly in such cases.

For us laymen maybe a discussion on
Teardrop shape.
The Kamm tail.
Porsche with all its tails, and active measures.
Mitsubishi rally cars with its tiny 'fins'

The theoretical shape which gives a Cd of 1.

Sutripta

[PGA]

Originally Posted by Jeroen

What is it that actually causes drag force? Is it similar to an aircraft wing, simply put high pressure at the front / low pressure at the back?

Originally Posted by MegaWhat

You are spot on in terms of what causes drag. For the car, it is primarily due to high pressure at the front and low pressure at the back. The high pressure at the front is caused as the car essentially rams into the air, while the lower pressure at the rear is caused as the air leaves the car in a separated state. It is hence important to contour the hood to minimise this "ram" effect and minimise separation (drag generating regions) at the rear. Apart from this, there is also friction drag generated as the air rubs over the surface of the car. It depends on surface finish, contamination and overall surface area that the air rubs across. But it is the pressure drag that is the primary contributor - similar to rowing a wide oar through water.

Thanks a lot Megawhat for a wonderful effort. My two cents in trying to simply a bit.

Total drag of on a body travelling through a fluid is sum of its induced drag and parasite drag comprising of form drag and skin friction drag. Whereas Induced drag is maximum at low speeds and tends to go down with increase of velocity, components of parasite drag tend to increase at square of velocity. The air compression being spoken of ahead of an object is form drag and drag due to movement of air over the car will be induced drag. I am sure Jeroen knows the diagram below as good as back of his hand.

Majority of times, its a tradeoff between Cl and Cd based on what are you designing. Lift and drag are in effect two components of the same force.

For a car trying to manage Cl is not desired unless its an F1 or one is trying to break a land speed record. Its Cd that you always focus on. One almost never encounter a laminar flow on normal road be it India or Autobahn because of obvious reasons such as traffic, trees, buildings, curves and weather etc. So whatever happens to CL while working around with Cd is managed by putting spoilers or vortex generators or such devices.

Lastly some very interesting bits about Cd of various items, cars and aircraft, scroll down to examples.

https://en.wikipedia.org/wiki/Drag_coefficient

[Hemantteen]

Thanks for very informative thread. In the Indian car scene, as some Bhpians noted that aerodynamics may not play a vital role but I think if implemented properly it can reduce drag to a small extent, thereby increasing a tiny bit of mileage and a little improvement in nvh. Fuel savings may not be much but will build up over a period of time.

[MegaWhat]

Originally Posted by Sutripta

Is any research done for the case the car is not in still air, but in a wind. Not galeforce. In one of the cars (Alto K10 first model) internal cabin noise would increase significantly in such cases.

Excellent questions Sutriptada! I'll try answering to the best of my ability. I may have to look up a bit more for details and will get back later.

Noise is basically a reflection of how chaotic the flow is. You’ll notice that noise increases whenever there is large scale flow separation - or vortices – around a car; especially near regions of high velocity (because local velocity of air increases near the roof and hood of the car). A car is usually designed to give minimum drag when it is travelling straight and air is coming at it head on. The volume of air that experiences flow separation will increase significantly if the situation changes - like in the example you mentioned. That happens because there is significant cross flow - wherein air impinges on the car at an angle instead of head on resulting in more flow separation, more drag and more noise. I have setup a run to see the effect of this on the Tesla. I’ll update it here when the numbers are out.

Originally Posted by Sutripta

Teardrop shape.

The teardrop shape is widely known to give the minimum drag possible. Essentially, it disturbs the air the least as it moves through it. Below is an image from NASA describing the Cd of different shapes.

Cd for various shapes

The airfoil here is the “teardrop” shape. If you cut an aircraft wing perpendicular to it’s wingspan, you’ll see that the cross-section looks something like that as well, but slightly curved – to generate a lot more lift and rather low drag. This shape varies across aircraft, depending on their desired operational envelope, but in general, it is streamlined.

The reason the teardrop shape provides minimum drag is:

1. The front is rounded, so the air doesn’t “ram” into a flat face, instead is guided along backwards gradually.
2. The rear is shaped to keep the air attached to the surface and not separate until the end of the object itself.

Essentially, we keep the flow attached throughout the length of the teardrop resulting in minimal drag due to pressure difference between the front and the back.

Originally Posted by Sutripta

The Kamm tail.

This is interesting. Wikipedia explains the concept rather well.

“The ideal shape to minimize drag is a teardrop. However researchers including Kamm found that abruptly cutting off the tail resulted in minimal increase in drag. The reason for this is that a turbulent wake region forms behind the vertical surface at the rear of the car. This wake region mimics the effect of the tapered tail in that air in the free stream does not enter this region, therefore smooth airflow is maintained which minimises drag.”

Essentially, if we take a teardrop shape and cut out a tiny part of it’s sharp tail, replacing it with a sharp vertical face, the air approaching and leaving the object still behaves as if it is interacting with a teardrop shape. The flow is able to do that because the sharp tail of the teardrop is actually replaced by a stable pocket of separated air. So the air still sees a teardrop shape in a way. Just that the tail end is now made up of air instead of whatever object the rest of the shape is made of.

Even in the case of the Model S, it sports something like the Kamm back. You can see that there is a stable pocket of separated air behind the car (refer image below). Now, even if we design the Model S to fill some part of that region with a sharp tail, the flow will still follow almost the same shape. There will be some reduction in drag, but it will be marginal as compared to the complications of making a shape as sharp as that and having such a long car.

Stable pocket of separated air behind the Model S

To understand why, imagine you are driving down the road shown in image 1, which is a typical Y-intersection. You are approaching from one arm of the Y in the direction of the black arrow at 100 kmph. You see the intersection ahead, there’s nobody on the road, and you end up reaching the intersection and going along the straight arm without any issues. That is similar to what a fluid element will behave near the end of a teardrop shape.

Image 1 - Y-Intersection

In the second case, for some reason, the Y-intersection is truncated (image 2) and there is no warning. The road designer expects you to take a sharp right turn and a sharp left turn to reach what would otherwise be the original Y-intersection. But if you are approaching the intersection at 100 kmph, that right-left manoeuvre is probably not going to happen. In most probability, you will overshoot the road and end up right where the original intersection was supposed to be. That is similar to how a fluid element will behave with a Kamm back (essentially, a truncated teardrop). It does not “see” that the shape is missing.

Image 2: Truncated Y-intersection

Originally Posted by Sutripta

Porsche with all its tails, and active measures.

The 911 was designed in a way that air stays attached to the car almost till the end. This led to low drag but a lot of lift and instability at high speeds. So to get the rear wheels planted, the engineers started looking at what we know as the ducktail spoiler. I think it became rather iconic and hence led to the crazy variety of spoilers they sport.

I’ll need to look into the active measured in a bit more detail before I can comment on it. Will check it up. But essentially, spoilers spoil the flow and reduce lift of the car; whereas a rear wing actively generates downward lift. It does not reduce the upward lift generated by the car, but counters it.

Originally Posted by Sutripta

Mitsubishi rally cars with its tiny 'fins'

What do you mean “tiny fins”? An image will help.

Originally Posted by Sutripta

The theoretical shape which gives a Cd of 1.

I’d like to know why you thought of the number 1? Cd can be more than 1 as well. Based on wiki, it seems Usain Bolt has a Cd of 1.2.

Originally Posted by PGA

...Lift and drag are in effect two components of the same force...

For a car trying to manage Cl is not desired unless its an F1 or one is trying to break a land speed record. Its Cd that you always focus on...

Thanks a lot for that plot PGA. Indeed, there is only one aerodynamic force acting on the car. We split it into components of lift and drag to easily understand their effect and be able to easily design things to manage them.

There has been a lot of focus on drag to reduce the fuel consumption that comes with it. But at the same time, I don't think we can ignore the aspect of lift. In India, probably yes, we could keep it on the backburner. But for cars expected to run at 190-250 kmph such as on the autobahn, the tendency to lift will be high as well unless actively analysed and accounted for.

[v1p3r]

Fantastic thread, @MegaWhat! I've run a race team and worked with CFD (STAR-CCM+ on high performance computing), and I must say your explanations and examples are spot on. An aerodynamics 101 course would be lucky to have you as an instructor.

I could never get the hang of complex aero - just meshing nearly drove me insane. But aero is now possibly more important that engines are. Adrian and Sebastian will confirm that.

[Sutripta]

Thanks for taking the time and effort to answer. Since I don't do it myself, really appreciate it in others.

Originally Posted by MegaWhat

The teardrop shape is widely known to give the minimum drag possible.

What would be the Cd of a teardrop moving in reverse?
And that of the shape of two teardrops with heads superposed. Essentially both ends being 'tails'.

Raindrops - do these differ in shape depending on size?

If you cut an aircraft wing perpendicular to it’s wingspan, you’ll see that the cross-section looks something like that as well, but slightly curved – to generate a lot more lift and rather low drag.

Aerofoils after we have understood the basics. Till then laminar flow, supercritical etc will have to wait.
And then we'll come to heat generated on by fast moving objects in air.

The 911 was designed in a way that air stays attached to the car almost till the end. This led to low drag but a lot of lift and instability at high speeds. So to get the rear wheels planted, the engineers started looking at what we know as the ducktail spoiler. I think it became rather iconic and hence led to the crazy variety of spoilers they sport.

The Porsche spoiler had the added task of getting enough air through the intercooler. And for Porsche function before form (except for overhung rear engines!)

What do you mean “tiny fins”? An image will help.

Lookup Mitsubishi EVO vortex generator for both images, and lots of words. Some from Mitsubishi themselves.

I’d like to know why you thought of the number 1? Cd can be more than 1 as well.

A Cd of a square plate is more than one.
It is dimensionless. Normally one would like to have an easily manufactured item to have a known value (Cd in this case). And if possible, one chooses that value to be 0 or 1, or some value related to the very fundamentals of physics. So what gives Cd = 1.0?

Sutripta

[v1p3r]

Originally Posted by Sutripta

Raindrops - do these differ in shape depending on size?

Yes. Larger drops get buffeted by air resistance and broken into smaller drops. Past a point (I think it is 2 mm) they turn spherical.

Originally Posted by Sutripta

The Porsche spoiler had the added task of getting enough air through the intercooler. And for Porsche function before form (except for overhung rear engines!)

Porsche's engineering and development teams use the ratio of rear axle lift to Cd as a measure to improve their aerodynamic performance. Some of this is solved as a result of their basic vehicle architecture.

If I recall right, a lot of race cars will have overall Cd over 1, and massive negative Cl. LMP and touring cars especially tend to be very heavy on downforce, while F1 cars are slightly more balanced with the advantage of being infinitely adjustable for every track.

Some good reading material wrt race cars -
https://www.mech.kth.se/courses/5C12...raghi_2010.pdf This guy is technical director at Tatuus, the chassis manufacturer for a bunch of series including Formula Renault.
http://www.fem.unicamp.br/~phoenics/...304.092016.pdf This guy has written what many consider the beginner's bible for race car aerodynamics. A must have for any motorsport engineer, along with Milliken.

[V.Narayan]

Truly one of the best of Team BHP. I must thank you for your detailed explanation, simulations and those helpful charts and diagrammes. For those who really want to understand aerodynamics in layman's terms you will not get a better laid out essay than this. People think of aerodynamics as only smoothening out the curves and (hopefully) reducing drag but as @MegaWhat explains aerodynamics can use tiny fitments to alter wind forces dramatically in order to significantly improve traction or reduce trailing drag. This article deserves a second read. Rated it a well deserved 5 stars.

My only grouse with aerodynamics in cars is that it has led to lower and lower cars which become more and more difficult to get into requiring half a yoga asan pose to manage it. And all this for improving speed in the 125+ kmph range which you don't get to experience for half or much more than half of your driving hours. Aerodynamics really comes into play meaningfully only after 175 kmph. That is why you see helicopters looking like dragonflies with nary a concern for streamlining.

[MegaWhat]

Originally Posted by Sutripta

What would be the Cd of a teardrop moving in reverse?
And that of the shape of two teardrops with heads superposed. Essentially both ends being 'tails'.

Raindrops - do these differ in shape depending on size?

I don't have exact numbers right now, but a teardrop moving in reverse will have a Cd similar to a sphere. With a teardrop having their heads superimposed, I think it'll be close to the teardrop shape itself. Will have to run a simulation or something for comparative numbers, there isn't much information available online on this.

Raindrops, yes the shapes do vary with size and the break into smaller drops as they fall. Here's an image for reference from Wiki:

Originally Posted by Sutripta

The Porsche spoiler had the added task of getting enough air through the intercooler. And for Porsche function before form (except for overhung rear engines!)


Lookup Mitsubishi EVO vortex generator for both images, and lots of words. Some from Mitsubishi themselves.

Oh yeah, that makes things more interesting for Porsche. I don't know why they have this multitude of wings and spoilers. Most of them will behave similarly, but there will be some differences in the performance of each of them. I'm guessing it's just a variety to suit the buyers taste. Most of them will get the essential job done.

Coming to the EVO, Vortex generators (VG) are like tiny wings. They generate high pressure on one side of the fin and low pressure on the other. So at the tip of the VG, part of the high pressure air tends to flow directly to the low pressure side, resulting in a tiny region of swirling flow - a vortex. Like I mentioned in the write-up, a vortex increases drag, but it helps in mixing of the flow. So the vortex generated by the VG helps in mixing high energy flow (located away from the cars surface) with low energy flow (located near the cars surface). This increased energy within the flow (near the cars surface) allows it to stay attached to the cars surface for a longer distance thereby reducing drag a little bit.

VGs are tricky things. They work well, but only under specific flow conditions, like a specific speed. Variations of the speed will reduce the effectiveness of VGs. They're popular on aircraft because most of the time, the cruise speed is almost constant. But there's a lot of speed variation for a car and I'm not sure how effective these will be at all speeds on the EVO. But I'm sure they'll work at a particular range of speeds. This range will vary depending on the cars shape and the VG design.

Originally Posted by Sutripta

A Cd of a square plate is more than one.
It is dimensionless. Normally one would like to have an easily manufactured item to have a known value (Cd in this case). And if possible, one chooses that value to be 0 or 1, or some value related to the very fundamentals of physics. So what gives Cd = 1.0?

Oh so you mean like a standard. I don't think a standard has been defined for Cd. But a cube has a Cd around 1.05, maybe that can help as a standard.

Originally Posted by v1p3r

Porsche's engineering and development teams use the ratio of rear axle lift to Cd as a measure to improve their aerodynamic performance. Some of this is solved as a result of their basic vehicle architecture.

If I recall right, a lot of race cars will have overall Cd over 1, and massive negative Cl. LMP and touring cars especially tend to be very heavy on downforce, while F1 cars are slightly more balanced with the advantage of being infinitely adjustable for every track.

Some good reading material wrt race cars..

That's good to know about the Porsche, thanks! I think it's because most of the lift in their case is generated by the rear half of the car.
Indeed, race cars play around those ballpark numbers. It's neat how the aerodynamics of an F1 car can be tailored to the track. Thanks for the links!

Originally Posted by v1p3r

I could never get the hang of complex aero - just meshing nearly drove me insane. But aero is now possibly more important that engines are. Adrian and Sebastian will confirm that.

Thank you for the kind words V1p3r. Means a lot coming from someone experienced in this field. It'll be great if you could add any points that you think I may have missed.

Meshing has become quite automated now. I think CCM+ also has a rather easy to use mesher. I'm just saying so you can try your hand on CFD again .

Aero I think goes hand in hand with the engine. For maximum punch, you need both. Miss out on any one and you'll end up with a bland performance and a day that says "can do better". After all, good aero only helps translate the engine performance to on road performance effectively.

Originally Posted by V.Narayan

My only grouse with aerodynamics in cars is that it has led to lower and lower cars which become more and more difficult to get into..
..That is why you see helicopters looking like dragonflies with nary a concern for streamlining.

Thank you sir for your very kind words. I'm glad you found this useful.

I agree about the low seating positions! It's a big cause of back ache in my family and I might have to get a tall boy sometime soon.

With the helicopter, I think the downwash from the rotors dominates the flow around the helicopter, so much so that there's no point considering streamlining for the oncoming flow. Crazy pieces of technology these things. A lot of aerodynamic research gets focused on the rotors, tip vortices, tip deflection and rotor noise.

It also comes down to cost and complexity. For example, the extremely popular Boeing 737 does not have a door to cover the wheels after they have been retracted. Throughout the flight, at speeds of 800 kmph, these wheels are open to the air, albeit retracted of course. There is a cost to pay in terms of drag. It's just cheaper as compared to designing and making retracting doors of these things.

[indranilmuk]

Great article and answers to many questions which comes to our mind whenever we see a new car being launched. I remember Honda claimed in India on their City to have that arrow shape design. Thank you so much.

[sgmuser]

Superb! Great primer on Aerodynamics!

Somehow, I do not believe in what Porsche has said about Aerodynamics. Speed alone can not help to manage the effects of Aerodynamics. There are many other factors, as covered here!

The Scorpios used to waver and wobble especially in the rear a lot and induce lots of uncertainty for the driver. Especially in the speed upwards 60 kmph on wards. I had a 2006 model with CRDe buth with leafsprings. Even though it feels a bit better with coil springs in the rear in the later version, still, I always thought it was a wasted opportunity by Mahindra in terms of CD. I used to research and analyse the boxy design, compared with many foreign models and always notice all the SUVs, has a tapering in the top rear. It has other issues as well contributing to this, but, I believe this is one of the biggest factor contributing to the instability in the highways. I am not a Mechanical engineer, feel free to educate me! Thanks! Image source: Mahindra website.

[frozen.ash]

Wow, MegaWhat, you have put up a very interesting and highly educational read. A million thanks for putting in all the time and effort to compose this descriptive thread. There is so much to know about automotive aerodynamics, and you have stirred the Have to say I am truly pleased with your style of explanation: starting from way too basic, and slowly building up on the details. Just the way I like it. Wish you had been an author/ publisher back when we were studying Fluid Dynamics in college! Cheers!

[Sutripta]

Originally Posted by sgmuser

The Scorpios used to waver and wobble especially in the rear a lot and induce lots of uncertainty for the driver.
...
It has other issues as well contributing to this, but, I believe this is one of the biggest factor contributing to the instability in the highways.

I would say the MM Scorpios problems were suspension design related rather than aerodynamic related. Rear seat passengers would often get seasick!

Taliking of Scorpios, Ford UK had a model called the Scorpio. (1970s IIRC). Was touted as an aerodynamic marvel. Was found to be unstable on British highways. (Testing was apparently done in Germany). The problem in this case was aerodynamic + suspension related, and the fix was surprisingly cheap - a pair of small plastic panels had to be added. The actual cause was the amount of time/ money spent on finetuning the aero design which went down the drain. And the egg on face for Ford's engineering prowess.
Viper and MegaWhat might have more details.

Sutripta

[RBalaM]

Hi,

Thank you for posting such a neatly explained article on aerodynamics. Very well written. Got to learn few things from it. Also Would like to add few points, correct me If I am wrong.

Along with lift and drag, we have side force acting on the vehicle. So totally there are 3 forces and 3 moments.
3 forces
Drag - pressure drag (maximum contributor)+ friction drag
Lift - (- ve lift indicates down force)
Side force
3 moments
Pitching
Rolling
Yawing

All these factors determines the aerodynamic property of a vehicle, but the major concentration is on drag and achieving a cd value.

The iso image apart from showing the wake, can also provide us a picture on how balanced a wake is behind the vehicle. As achieving a balanced wake is equally important in enhancing the aerodynamic capability of a vehicle.

[MegaWhat]

Originally Posted by sgmuser

The Scorpios used to waver and wobble especially in the rear a lot and induce lots of uncertainty for the driver.

I'd agree with Sutripta here. I think it's a lot more to do with the suspension than aerodynamics. Especially at 60 kmph.

Originally Posted by RBalaM

Hi,

Along with lift and drag, we have side force acting on the vehicle. So totally there are 3 forces and 3 moments.
3 forces
Drag - pressure drag (maximum contributor)+ friction drag
Lift - (- ve lift indicates down force)
Side force
3 moments
Pitching
Rolling
Yawing

All these factors determines the aerodynamic property of a vehicle, but the major concentration is on drag and achieving a cd value.

The iso image apart from showing the wake, can also provide us a picture on how balanced a wake is behind the vehicle. As achieving a balanced wake is equally important in enhancing the aerodynamic capability of a vehicle.

You're correct RBala. Any object moving through air experiences a force acting along each coordinate axis and a moment about each axis. But for a car, side force becomes important during turning and when there is a natural wind blowing sideways at the car. In the case where there's no or less wind and the car is going along a straight road at decent speeds, the side force is close to zero and only the force acting along the car and upwards (or downwards) become more significant. That's lift and drag. That's the situation I'd say 65-70% of the time. Normally, we design to be efficient at the condition that the car is going to operate in for the maximum time, and then we just verify that the forces along the other axes are still manageable. We look at the other axes in detail only if the forces seem out of control. This validation can be done by CFD or a wind tunnel test. Do correct me if I'm wrong here.

The section on "results - effect on traction: rear axle vs front axle" actually includes the effect of pitching moment as well. I didn't mention it to keep the intro simple .

An aerodynamic moment can be thought of as a torque that the air imposes on the car. Something like the air trying to rotate the car about it's center of gravity, and about all three axes. Pitching moment is one where the hood tends to go up or down and the trunk goes down or up accordingly (clockwise or counter clockwise essentially). Rolling is sideways and yaw is side to side. Now again in the case of a car going straight and no wind, there will barely be any roll or yaw. Besides, the suspension usually takes care of it as it is not as strong as the pitching moment. Although they do have some impact and can even be used for a benefit, these moments are more critical for aircraft than passenger cars. Maybe they play a big role on trucks though.

A balanced or stable wake is important as well just as you say. An unstable wake is a reflection of an unstable flow. Which means that the lift and drag values aren't stable either and fluctuate rapidly. That can affect handling adversely and is best avoided. The spoilers on our production hatchbacks help with this. They ensure that the flow does not fluctuate and separates at the same point at all speeds - keeping a stable wake, stable Cl and stable Cd.

Thanks for bringing up these aspects!!