For those who’s eyes glass over from all the fancy physics, let’s try an analogy.
Can a person on water skis go faster than the boat pulling them? Of course they can. I’ve done it, as has most anyone who has ever gotten beyond merely getting out of the water. You push down hard nearly perpendicular to the motion of the boat and you will find yourself skittering right up alongside the boat at the end of your taut line.
You see, it’s not MERELY the wind and the sail which allows a sailboat to exceed the wind-speed, but a combination of the wind, sail and KEEL which allows that to happen. Merely analyzing wind and sail alone won’t get you within a bernouli-hair of completing the equation.
Not really. There are not two separate causes of lift (deflected air and the pressure differential) - these are two ways of thinking about the same effect; they go hand-in-hand. You can’t have an airfoil generating lift from the pressure differential unless it’s pushing on (and deflecting) the airflow. Similarly, you can’t have deflection without a pressure differential. They’re two ways of describing lift. So it’s wrong to say that some proportion of the force comes from the pressure difference and some other proportion comes from air deflection.
I like the deflected airflow explanation because it’s intuitive to everyone. And despite what Stranger was saying, I don’t see any reason why a sail is not an exact equivalent of an airplane wing. What I don’t like about the pressure differential explanation is that it is completely unintuitive, at least to me. The typical explanation that I learned in Physics 101 way back in 1978 or 79 was that we were taught the Bernoulli effect and why it works, then were taught that the air has to go over the top side of a wing faster so that it can meet the air on the underside at the same point where they were separated. I didn’t question it then, but the question is - why does it have to meet up? Does it have a date? And we know from wind tunnel photos that they don’t meet up, that the air on top actually arrives at the trailing edge first. But why does it do that? I still don’t understand, and I doubt that very many people who don’t design airfoils for a living understand either.
So we have a nice, intuitive, correct explanation that’s not very useful for calculations. And we have a bastardization of a correct explanation which almost no one understands, and the usual bastardization is incorrect anyway.
djailer, thank you for that helpful analogy. That does get to the heart of the matter.
CurtC, well said. The problem with the deflected air explanation is the knowledge that the top of the wing deflects at least as much air as the bottom. That is counter intuitive to the layperson, who can visualize the air bouncing off the bottom of the wing, but not clinging to the top of the wing as it races past. Even though the Coanda Effect has probably been witnessed in the sink many times.
I recall hearing somewhere about the assumption the two streams have to meet.
As for how the top airflow outraces the bottom airflow, consider this. As the wing slices through the air, there is a blunt edge on the leading edge that is piling up air as it slams into it, kinda like running your hand through snow. That leading edge is called the stagnation point. As the air flow separates to go around the wing, the air going under the wing is kept kinda piled up (higher pressure), while the air stream going over the wing spreads out more (lower pressure). This occurs either because the shape of the wing halfs is different (look at the wing profile), or if the wing is symmetric (some sport planes), the angle of attack is such that the effective shape is different.
The curve of the top of the wing creates a zone where the freestream air is not flowing directly, so the air along the wing has less air in the way (lower pressure) and thus races along the wing (i.e. speeds up), clinging to the wing top edge. As that air reaches the back of the wing it has turned so it aims somewhat downward as well as rearward, and thus as it leaves the wing it is deflected down. The bottom of the wing is either essentially flat, leaving it in the freestream air and thus keeping pressure higher, or else the angle of attack keeps it a bit like that hand through the snow, keeping the pressure higher. This slows the air flowing under the wing and pushes the air slightly down.
So, lower pressure air flows faster, higher pressure air slows down. Ergo, the stream over the top of the wing gets there first, and is deflected down more than the air under the wing, creating the downward momentum that keeps the plane up in the air.
I like how this explanation avoids the Bernoulli principle, seeing as that only applies to pipes. First the wing air is low pressure, then it speeds up. Not the inverse.
It has a pretty good description of how lift is created and talks about some of the topics of this thread, like how the streamlines must meet error possibly began, and how Bernoulli equation can be derived from the Newtonian explanation. It has pictures showing the streamline separation (faster on top and outracing the bottom) and a picture showing how air is deflected downward from the top of the wing.
My one quibble is in describing what causes the pressure to decrease and thus the air to speed up over the wing.
It took me multiple passes at this section to finally realize he is saying the same thing I am. The air streams start parallel. The wing interrupts the parallel flows, so the flows have to distort to go around the wing. The distortion on top of the wing is greater. As the air is squeezed upward, the hump on top causes there to be a pocket behind that hump that the freestream of air is not flowing into. However, the air coming over the wing sees that pocket as a low pressure zone, and expands that direction, which speeds up the flow.
This only works in laminar flow. Stall occurs when the flow coming over the wing separates from the wing rather than flows along the wing. Stall comes from turbulence. Laminar flow maintains streamlines, turbulence is mixing them up (like a blender). In smooth laminar flow the streamlines want to cling to the wing to flow through the low pressure zone, which pulls them down, and thus directs the trailing air flow downward in addition to backwards. In turbulent flow, the streamlines no longer stay smooth, but jumble, which fills the low pressure zone without the clinging and thus there is no deflection of the streamlines, no air directed downward, and so the plane falls out of the sky.
That is why wing geometry is special. The smooth shape is to maintain laminar flow. The wrong shape and turbulence occurs. The wrong speeds or angle of incidence, and turbulence occurs. A flat plate wing can get you airborn, but is much more susceptible to turbulence, over a far greater range of conditions/angles/speeds. Ergo, not good for flying.
Referring to wiki again, and the topic of this thread, sailing:
My quibble here is the description of the sail being “pulled”. Pressure doesn’t work that way. Pressure is by definition a push. However, what causes forces (and thus motions) is pressure differentials from one side to another.
Let us examine to hypothetical sails, identical in shape. We will assume the conditions are possible - it’s a hypothetical.
On one sail we have wind hitting one side of the sail and standard atmospheric pressure on the back. The wind causes an increase in pressure on that side, and the pressure difference is establishes Pw - Pa = Push1.
Second sail has the same wind hitting one side (same angle), but has wind flowing past the other side. The flowing wind reduces pressure on that side of the sail (let’s not argue cause and effect - Bernoulli says both exist together). That means the pressure on the backside of the sail is Pb < Pa.
So Pw - Pb = Push2, but Pw - Pb > Pw - Pa (Pa is larger than Pb, so the pressure difference is less)
Ergo, Push2 > Push1. Boat sails faster. Same result, but all from pushing, no pulling.
By the way, the Bernoulli Principle does not only apply to pipes, but does only apply to closed volumes. In other words, you can define a volume using imaginary boundaries instead of actual boundaries, but only if you can prevent fluid flow through that boundary. Laminar flow is good for this because the streamlines are consistent and even. As long as you set the boundary condition sufficiently far from the obstacle (i.e. wing), you can define an arbitrary line that is the boundary and ignore what happens above it.
When fluid flow encounters an obstruction, one of three things can occur. The fluid flow can divert, back up, or speed up.
If the flow is in a constrained volume (i.e. pipe), then it cannot divert and must go one direction.
If the flow is forced (i.e. pumped, fan driven, etc) into the channel/pipe, it cannot back up, because pressure behind keeps pushing forward.
That leaves one option for getting past an obstruction - speed up. That is why a Venturi speeds up fluid flow. The mass flow rate is forced and fluid cannot back up or divert. Forcing the fluid to move forward faster means less of the fluid is bouncing off the sides, ergo pressure drops. That is why Bernoulli says that speed increases in flow cause lower pressure.
However, in airplane wings, diversion is a possibility, which is why pressure decrease drives velocity and not vice versa.
My Newtonian way of thinking about that is that the top side of the wing is curved smoothly so that the airflow follows the curve, which then means that the top side of the wing gets to deflect air downwards in addition to the air deflected by the underside. That approximately doubles the lift.
A sheet of plywood used as a wing would deflect air on its underside, but the sharp edge at the front would cause the airflow to separate there and the top wouldn’t throw the airflow downwards.
I’ve been told that for a well-designed modern airfoil, the contribution to lift of the downward deflection of air flowing over the top can exceed 65%.
CurtC, that is essentially correct. The laminar flow clings to the contour and follows it, which bends the streamline, and thus gives downward motion. A sheet of plywood at low angles of attack would have the same characteristic, but at higher angles of attack the edge causes conditions where the flow doesn’t want to cling and instead becomes turbulent. That is what stall is - flow separation at turbulent conditions.
It’s related to the Coanda Effect.
When a fluid flow speeds up, the pressure drop (per Venturi Effect). If the flow is surrounded by more fluid flow, then the surrounding fluid shifts inward from all sides evenly. If you enclose the flow on all sides (i.e. a pipe), then the lower pressure is consistent all around, so still symmetrical flow.
But if you put a rigid object on one side only, then the flow on all other sides shifts in to maintain pressure, but on the object side the surrounding fluid is not present to shift in, so the fluid stream and the object shift together. If the object is hard mounted, then the flow stream shifts, and you see a bent flow - Coanda effect. But if the object is not hard mounted, and has higher pressure on the underside than the upper side, you get the object shifting as well. Still the same cause, Coanda effect, but the stream is bent less because the object moves.
However, in aircraft, the smooth surface allows the coanda effect to keep the flow laminar and therefore turn the flow smoothly along the surface. A sharp surface/bumpy surface would induce turbulence, and thus send the streamlines into a tangled mess. No more flow cling to wing, no more turning of flow direction.
Xema, that is correct. More of the lift is coming from the top of the wing surface turning air flow downward than from the bottom of the wing “deflecting” air. It’s not just the pressure differential from bottom to top, it is the coanda effect turning the direction of the stream.
Good Lord. No disrespect to the contributors above, but clearly I’ve failed in the essential Straight Dope mission, namely, conveying the nut of the answer. You guys are all over the place. All you need to know about sails-as-airfoils is that the passing wind creates a zone of low pressure on the convex side of the sail, allowing the relatively higher pressure on the concave side of the sail to propel the boat forward. The boat isn’t being magically pulled, it’s being pushed. What confuses people is that when the boat runs with the wind, the air pressing on the sail is the only process at work, and thus the boat can go no faster than the wind can push it. In contrast, when the boat is close hauled, there are two processes involved: (1) the air pressing on the concave side of the sail, and (2) the reduction of pressure on the convex side. This greatly increases the net force pushing the boat forward.
I’m purposely omitting a great many subtleties about keels, vectors, apparent wind and the like since they obscure rather than clarify the answer. Let’s make matters simpler still. If we take the wind out of the picture altogether and posit a total vacuum on the leading (convex) side of the sail (and for amusement neglect the drag of the boat in the water), the boat will obviously be pushed ahead at high speed by the pressure of the air on the sail’s trailing (concave) side. In this scenario it won’t occur to anyone to ask why the boat is sailling faster than the wind because there’s no wind. Clearer now?
I hold to my view that this explanation, lacking as it does any mention of the attendant deflection of air, will be found unsatisfying by most thoughtful people. I see no meaningful advantage to omitting the aspect of the problem that provides the commonsense action-reaction “Newtonian” explanation, and considerable disadvantage.
Wow - I’ve been a fan of The Straight Dope for a long time, but never knew about this forum until I stumbled on this today.
There are some very interesting questions and concepts here. I think I can help as I’ve been working with these very problems for a long time.
First… air can’t “pull” anything (except in the sense of skin friction - which we should ignore for now). Air applies positive pressure on a surface - always. Sometimes that pressure is higher, sometimes lower. Sometimes even well below ambient pressure. In this last case we talk about it producing a partial vacuum and “pulling” on the surface, but the reality is that it’s still pushing on the surface, but with much less force than the air that’s pushing back from the other side. On a wing (or any airfoil) the air on bottom is pushing up with more force than the air on top is pushing down. The difference in those two forces is the net lift.
Now, some people are going to say “differential pressure doesn’t explain lift - deflecting airflow (i.e. conservation of momentum) does”. Well, both explanations are 100% true and completely in agreement with each other. They are simply two different ways of solving (or looking at) the problem. One is no more right than the other.
As to how/why a sailboat can sail faster than the wind, the best intuitive answer I can give is that it’s a lot like squeezing a watermellon seed between your thumb and finger. The sail and keel are being “squeezed” by the water and air - and the boat shoots forward as a result - just as the watermellon seed can shoot out much faster than your thumb and finger move together.
Perhaps the more interesting question is whether a boat can actually sail to a direct downwind point (via tacking) faster than the wind gets to that same point. The answer is Yes.
More interesting still, is the question of whether it’s possible to make a wind powered vehicle that goes DIRECTLY downwind faster than the wind. It is.