Two heads - drag forces compared

[VGB]

Vince,
I don't understand why the ambient air itself would generally be turbulent. Is that the case when the air is still (i.e no wind) as well?
This quick search on turbulent ground air seems to say it takes relatively high winds for the air to become turbulent.

How much wind causes turbulence?
The stronger the wind speed (generally, a surface wind of 20 knots or higher is required for significant turbulence), the rougher the terrain and the more unstable the air, the greater will be the turbulence.

Since fluid dynamics effects are so nonlinear I would tend to think that the superposition of skin drag and form drag effects might be a bit chancy but is seems that is what is done in most of the papers that I see that are describing drag effects on fly lines.

Evidently that is also an accepted approach in studying drag effects on inclined cylinders in air or water.
superposition_claim.jpg

Gordy

It appears that my response to this post has been removed entirely.

Allegedly, Heisenberg said:

When I meet God, I’m going to ask him two questions: why relativity? And why turbulence? I really believe he’ll have an answer for the first.

Wind is not the sole cause of turbulence, it is a complex subject and some of the causes are discussed on the Ambient Air turbulence thread.

viewtopic.php?f=11&t=3862

Regarding superposition, it is noteworthy that it is an approximation. The fidelity of aerodynamic modelling is proportional to the size of your budget and even complex models can have large error bands when subjected to verification.

[Torsten]

Hi Gordy,

The tilt angle on the fly leg for Paul's and Lasse's distance cast seems to be around 9 degrees so using a normal velocity of 40*sind(9)=6.2 m/s would give a calculated Reynolds number of 500 to 700 using the other estimates given in your original post.

9 degree are too high for flow calculations, because the line was accelerated upward to the flow.

Approximations are OK! This thread was meant to answer one of Bernd's questions, see post #1.
The form drag is here a small side effect, drag force is dominated by the skin friction.

For the skin friction coefficient, I've found a nice alternative paper, see [1]

Greetings,
Torsten

[1] WHITE, Frank M. An analysis of axisymmetric turbulent flow past a long cylinder. 1972.
https://citeseerx.ist.psu.edu/document? ... a69dbe26ba

[VGB]

Hi Torsten

Approximations are OK! This thread was meant to answer one of Bernd's questions, see post #1.

Approximations are fine as long as they are presented as such.

Regards

Vince

[gordonjudd]

9 degree are too high for flow calculations, because the line was accelerated upward to the flow.

Thorsten,
How does the acceleration of the line in the tangential direction impact the expected drag on a tilted length of line traveling at some Vx velocity?

Tracker gets an angle of around 9.6 deg for the tilt angle of the trailing line in this cast of Paul's. No idea if the line was being accelerated at that point.

tilt_angle_Paul's cast.jpg (36.43 KiB) Viewed 1043 times

Gordy

[gordonjudd]

For the skin friction coefficient, I've found a nice alternative paper, see [1]

Torsten,
Using the equation given in that paper by F.M. White for the expected skin drag coefficient for your case B conditions gives:

Cdt_white for Re_t=5.251e+07 and L/d=3.287e+04 is 0.0049'

Surprisingly close to the .005 value measured by Gaddis.

Gordy

[gordonjudd]

If the skin drag is producing turbulent flow on the upstream side of the line then that would reduce the form drag to the Cd value expected for the critical flow Regime regardless of the much lower Reynolds number you would get from using the diameter of the line and the normal velocity?

Torsten,
That supposition appears to have been born out in the experimental studies described in Svein Erstal's Doctoral Thesis available at "An Experimental Study of. Hydrodynamic Forces on Cylinders and Cables in Near Axial Flow".

In section 2.3 of that thesis he concludes:

The drag coefficient typical drops from Cd = 1.2 in the sub-critical regime to Cd=0.4-0.6 in the critical regime before increasing again to Cd= 0.8 for trans-critical flow. Since this change is explained by transition from laminar to turbulent boundary layer upstream of the separation point, the boundary layer due to longitudinal flow should be of importance. If the boundary layer on the cylinder is turbulent due to longitudinal flow, the trans-critical value of Cd = 0.8 should probably be used even in cases where Re2d is in the sub-critical regime.

Since the longitudinal flow on a tilted section of fly line meets that criteria (turbulent flow stemming from the high Reynolds numbers associated with the line length and high tangential velocities expected for a fly line) I am going to start setting Cd=0.8 in my future form drag calculations.

Gordy

[gordonjudd]

9 degree are too high for flow calculations, because the line was accelerated upward to the flow.

Thosten,
I can see that there would be an upward force on a ds differential length of line if that section was being accelerated in the tangential direction, but would think that before the loop is formed the relevant acceleration of the line is in the horizontal direction not in the tangential direction.

After the loop is formed the fly leg would be shortening and thus has a velocity component in the tangential direction. However, how much net acceleration of the fly leg in the tangential direction would depend on the tug of war between the rho_l*v_loop.^2 force related to the momentum change of the fly leg and the opposing force produced by drag. Thus some tangential acceleration of the fly leg could result as the fly leg length shortens, but I would expect that value to be much smaller than the horizontal acceleration of the line imparted by the caster up to MRF.

Thus especially for heavier line with high linear mass density values the 9 degree tilt angles measured by Tracker in Paul and Lasse's distance casts are believable. Lighter lines traveling at 40 m/s would of course have smaller tilt angles.

It would be good to take some photos of the tilt angles observed in different casts to see if the angle expected from drag calculations are borne out in practice.

Gordy

[Torsten]

Hi Gordy,

this thesis really a good work, I've already achieved it, a great reference. For the external effects like the ambient air turbulence I've found now also a pretty nice paper, will write about it later.

The tilt angle can be misleading if you do video analysis. Actually you have to analyse the line velocity components of the line (at some reference points) and then you need to project them on the line shape to get the tangential and normal velocities. Should be possible with the Tracker software too.

Greetings,
Torsten

[VGB]

I certainly would have been concerned about read across due to this, the testing and simulation is not on a representative shape:

[gordonjudd]

Actually you have to analyze the line velocity components of the line (at some reference points) and then you need to project them on the line shape to get the tangential and normal velocities.

Thorsten,
Unfortunately we generally do not have markers on the trailing line to measure the true x and y velocities along the line. However, I would think that a reasonable first order approximation to the real velocities would be to measure the nominal tip velocity up to around MRF while the length of line from the rod tip to the fly remains constant.

As you know the expected tilt angle where the upward force from form drag on a ds incremental length of line just offsets the downward force from ro_l*g and skin drag is sensitive to the velocity of the tilted line. Since the velocity is changing, the line may not get to a steady state tilt angle like it is expected to do when a line is trolled in water or towed at a constant velocity in an airborne situation.

For the external effects like the ambient air turbulence I've found now also a pretty nice paper, will write about it later.

I think that air turbulence is a bit of a red Herring for most of the situations where I want to be casting. On days with little or no wind the velocity of that turbulence will be very small compared to the velocity of the line.

Thus when the Reynolds number for the line length and tangential velocity of the fly leg is high enough to produce turbulence (as will generally be the case) then I think using Erstal's expected value of .8 for the form drag coefficient is a reasonable thing to do for our calculations.

Gordy