Friday, April 29, 2011

Some thoughts on water flowing over a paddle- by Robert Stehlik

Quickblade's Jim Terrell recently came out with an excellent video breaking down the SUP stroke with high tech video analysis.   If you have not seen it, watch it in the window below (double click for full screen view).
The video analysis clearly shows that on longer raceboards at cruising speeds, good paddlers plant the paddle and move past it.  It shows Rob Rojas in slow motion, planting the paddle, applying power with the shaft bending and no visible "slippage", the blade merely rotates at the waterline.   With his forward momentum, he actually pulls the blade out in front of the spot he planted it.  To me this clearly shows that once the blade is planted, the water is compressed against the face of the blade and there is very little "slippage" or backward movement.  This made me think more about how a paddle blade moves through the water.  I'm not a scientist or paddle designer but just want to share some of my thoughts.

When people talk about paddle blade design they usually explain how the shape and design of the blade moves through the water, visualizing how the blade face gets pulled backward through the water.
In reality, during an efficient race stroke, there is very little movement of the blade once it is planted.  The water is compressed against the face of the blade and if the force is applied at the right time in the right dosage, there is very little slippage or movement of the blade backwards, it is effectively planted in the water, not moving through the water.
So, most of the movement of the blade through the water occurs when the blade is sliced down into the water during the catch and when it is pulled out of the water during the release.  In both cases, the paddle moves sideways, or tip first with water rushing past both sides of the blade.
When you think of it this way, most of the water flowing over the paddle is not flowing over or past the face of the blade but moving sideways, during the catch and release.  When designing a paddle the concern should be to make the sideways movement during catch and release as smooth and efficient as possible as this is the way the paddle travels through the water the most: slicing into and out of the water.
Instead of looking at the face of the blade, look at the edge/ tip and side profile of the paddle as that is the direction the blade moves through the water mostly.  It seems to me that a thin, flat blade should be most efficient slicing through the water sideways, while paddles with big spines, concaves, curves or other features designed to "catch" more water will only disrupt a clean sideways entry and exit.  It seems that a mild dihedral close to the waterline will not disrupt the waterflow during entry or exit much but maybe Kialoha is onto something with their completely flat, relatively thin blades.
It also means that it does not matter whether you put a sticker on the face or back of the blade, the sticker edge might cause a tiny bit more friction on either side during catch and release, but should not make a difference during the power phase.

So what's up with those "magic" golf ball dimples on the face of the Quickblade elite racer
blades?
Some people seem to think they are designed to "hold" more water.

According to the Quickblade website FAQ section:
Q -
What is the difference between the Elite racer and the Magic?

A -The Magic paddle is identical to the Elite racer with the difference of the blade surface.
Q -What is the special feature on the QB Magic paddle?
A -The QB Magic paddle has dimples on the backside of the blade similar to a golf ball. This has proven to be most effective for very strong powerful paddles that can drive it with force.
I'm kind of puzzled by this.  The dimples are on the face of the blade, not the backside (I guess it depends how you look at it).I think they mean to say it works well for powerful paddlers. I would think that the dimples would be most effective at reducing friction when the paddle is moving through the water sideways, much like dimples in a golf ball make it travel farther with less friction(I think that's why they have dimples anyways, correct me if I'm wrong).Do they work?  I can't tell the difference and neither can anyone I talked to that tried both but it's one of those things that if you believe in it, it works.  It's called the placebo effect.  At $10 more than the non-dimpled elite race paddle the dimples are cheaper than a holographic power band bracelet and might actually do something (although it also makes the blade about an ounce heavier).
Just don't put a fat paddle edge guard on a Magic blade, that would just be silly.
If the purpose of the dimples is to reduce friction during catch and release, the dimples really should be on both sides of the blade, not just on the face of the blade, as the water rushes past both sides equally.  And if it really reduces friction, why not cover the bottom of the board with dimples, too?





If you are into the science of flow and friction, you will enjoy this scientific explanation posted by ehrawn on standupzone.com:
The text is easier to read in the comment section below, where I posted the same text again.

Here is a good illustration of what's happening with a golf ball.
http://www.aerospaceweb.org/question/aerodynamics/q0215.shtml


"Since the laminar boundary layer around the smooth sphere separates so rapidly, it creates a very large wake over the entire rear face. This large wake maximizes the region of low pressure and, therefore, results in the maximum difference in pressure between the front and rear faces."

In comparison, lets look a flat plate. First, perpendicular to the flow:
http://en.wikipedia.org/wiki/Drag_coefficient


"fluid approaching the object is brought to rest, building up stagnation pressure over the whole front surface." Negative pressure builds at the back because of the eddies that form along the edges which adds to the effective drag. The flow around the plate will be turbulent even at very low velocities.

But parallel to the flow:
http://www.roymech.co.uk/Related/Fluids/Fluids_Drag.html


You'll notice that there is no advantage to creating turbulent flow earlier on plate, like there was with the sphere. You won't get less wake or any other advantage because pressure drag is negligible to skin friction.

So lets talk numbers. Drag Coefficient depends on the Reynold's Number which is a function of velocity, but, common estimates for the drag coefficient of sphere is about 0.47. A flat plate perpendicular to the flow is about 1.17. A flat plate parallel in laminar flow is about 0.001, and in turbulent flow about 0.005. http://www.engineeringtoolbox.com/drag-coefficient-d_627.html

Drag force depends on the drag coefficient and a "frontal area" which we can just assume is the same for all cases: for the parallel plate, it's actually the side area, if you get what I'm saying. As such, the drag force for the parallel plate is 2-3 OOM less than the other two scenarios. So then you get into the question about laminar vs turbulent. Crunching some numbers:

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

L is the length of the blade. Assuming I can get the paddle in the water in about 0.5 sec, in sea water, your looking at a Re of about 1.1 * 10^9. Re > 10^6 is so we are way turbulent.

Essentially, what I'm trying to argue is that I don't think there is an advantage to making the flow turbulent on catch and release. I'm guessing that a quick catch will get you turbulent even without the dimples. Even if there was and effect caused by the dimples, it would be so minuscule compared to rest of the stroke, that it wouldn't be noticeable to the paddler. Look at the mechanics of a stroke. As soon as the blade enters the water, the perpendicular flow is going to be all that the paddler feels. At the release, there's going to be so much cavitation at the back of the paddle, as soon as you stop applying forward pressure, the built up pressure on the face and suction on the back will pull the blade into the void…not really a void but a low pressure volume. Again, it will already be turbulent flow, so dimples aren't going to do anything.

Of course, the focus of my masters was mechanics. I haven't studied fluids since my undergrad, so, Robert, how'd i do?

1 comment:

  1. Here is a good illustration of what's happening with a golf ball.
    http://www.aerospaceweb.org/question/aerodynamics/q0215.shtml


    "Since the laminar boundary layer around the smooth sphere separates so rapidly, it creates a very large wake over the entire rear face. This large wake maximizes the region of low pressure and, therefore, results in the maximum difference in pressure between the front and rear faces."

    In comparison, lets look a flat plate. First, perpendicular to the flow:
    http://en.wikipedia.org/wiki/Drag_coefficient


    "fluid approaching the object is brought to rest, building up stagnation pressure over the whole front surface." Negative pressure builds at the back because of the eddies that form along the edges which adds to the effective drag. The flow around the plate will be turbulent even at very low velocities.

    But parallel to the flow:
    http://www.roymech.co.uk/Related/Fluids/Fluids_Drag.html


    You'll notice that there is no advantage to creating turbulent flow earlier on plate, like there was with the sphere. You won't get less wake or any other advantage because pressure drag is negligible to skin friction.

    So lets talk numbers. Drag Coefficient depends on the Reynold's Number which is a function of velocity, but, common estimates for the drag coefficient of sphere is about 0.47. A flat plate perpendicular to the flow is about 1.17. A flat plate parallel in laminar flow is about 0.001, and in turbulent flow about 0.005. http://www.engineeringtoolbox.com/drag-coefficient-d_627.html

    Drag force depends on the drag coefficient and a "frontal area" which we can just assume is the same for all cases: for the parallel plate, it's actually the side area, if you get what I'm saying. As such, the drag force for the parallel plate is 2-3 OOM less than the other two scenarios. So then you get into the question about laminar vs turbulent. Crunching some numbers:

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

    L is the length of the blade. Assuming I can get the paddle in the water in about 0.5 sec, in sea water, your looking at a Re of about 1.1 * 10^9. Re > 10^6 is so we are way turbulent.

    Essentially, what I'm trying to argue is that I don't think there is an advantage to making the flow turbulent on catch and release. I'm guessing that a quick catch will get you turbulent even without the dimples. Even if there was and effect caused by the dimples, it would be so minuscule compared to rest of the stroke, that it wouldn't be noticeable to the paddler. Look at the mechanics of a stroke. As soon as the blade enters the water, the perpendicular flow is going to be all that the paddler feels. At the release, there's going to be so much cavitation at the back of the paddle, as soon as you stop applying forward pressure, the built up pressure on the face and suction on the back will pull the blade into the void…not really a void but a low pressure volume. Again, it will already be turbulent flow, so dimples aren't going to do anything.

    Of course, the focus of my masters was mechanics. I haven't studied fluids since my undergrad, so, Robert, how'd i do?

    ReplyDelete

Thanks for posting a comment on Zen Waterman, your comment will show as soon as we have a chance to screen it for spam. Mahalo, we hope to see you on the water!