Constant Acceleration

[VGB]

I was also accused of using "grade school" logic. This may have been true as I may have been trying to explain what a taxonomic structure was in single syllable words.

Maybe they though taxonomist was someone that stuffs animals?

[Stoatstail50]

Well, whatever, I learned a lot about basic cast mechanics doing all that stuff, I was new and it was a privilege to be included in the discussions on here.

What I also discovered was that the understanding of basic mechanics in the anglersphere was not based on good science but was rooted in wierd homespun theories concocted by dominant casting personalities who were largely unable to accept having their self declared authority and well publicised, but often erroneous, ideas challenged.

Good times

[Paul Arden]

I learned a lot about the way people think and the different approaches taken to try to solve problems. It gave me some interesting insights. I think by looking at this very small example of how a committee works, we can understand why the world is in such a state!

[gordonjudd]

I follow the mathematical argument for the existence of jerk, but the "so a caster can't do it" conclusion feels suspect.

Few,
As you know angular jerk is just equal to the rate of change of the angular acceleration. Consequently there will always be some jerk profile associated with a cast that will depend on how the caster rotated the rod.

Generally there will be some "noise" when you take discrete derivatives of measured data in a program like Tracker and those ripple levels tend to increase with each additional derivative calculation. Lots of luck getting ripple free velocity or acceleration graphs from digitized data in Tracker. Fortunately Matlab has some built-in spline smoothing routines and means of taking derivatives of the resulting fits that can be used to produce a representative jerk profile for a cast.

As shown earlier here is the measured angular acceleration profile that Maxine McCormick produced while false casting 50 feet of line.

accel_profile.jpg (37.93 KiB) Viewed 198 times

You can see that she had a slight linear increase in her acceleration profile at the beginning of the cast followed by a rapid nonlinear increase in the the applied acceleration starting about .37s before RSP1 (t=0 point on graph). Her acceleration profile had no sudden changes that would produce a spike or bump in the expected jerk profile as shown below.

jerk_profile.jpg (41.04 KiB) Viewed 198 times

In contrast the jerk profile for a cast employing constant acceleration would be zero. A cast with a linear acceleration change over some time range would have a flat value over that range as shown in the jerk profile of Maxine's up to -.4s.

I suspect the jerk profile for most casters would not be nearly as smooth as Maxine's, but never the less it could be calculated with the right software.

Gordy

[Few]

Thanks very much, Gordy, for the thoughtful reply. Maxine is a marvel, no doubt about it. I've been thinking about putting some retroreflector stickers on myself and my fly rod, making a video, and seeing if I can get Matlab to track them without it becoming a nightmare. Thanks for the reminder that the numerical derivatives will generate a lot of noise. If only I could devise a way to smooth the velocity and acceleration graphs in Matlab, and then transfer the results back into my stroke to improve my casting.

I fear my comments and questions have been a little out of synch with the intent of this thread. I was intrigued by the comments about what amount of jerk (time derivative of acceleration) is physically possible. Not surprisingly the thread's focus is very much on casting (!) and not so much on whether a human can generate a particular magnitude of jerk, or a particular jerk as a function of time. I'm trying not to derail the discussion further. Rest assured I am completely on board with the idea that jerks should not be allowed near fly rods if nice loops are the goal!

Thanks again for the help.
Few

[Graeme H]

G’day Few,

There’s a free video based physics software called Tracker that works well it’s only on the Mac OS though.

Cheers, Graeme

[NM]

G’day Few,

There’s a free video based physics software called Tracker that works well it’s only on the Mac OS though.

Cheers, Graeme

It's available for Windows and Linux as well Graeme. I am using the Windows version and it works great.
Cheers,
Nils

[Few]

Thanks very much! Just a day or two ago I figured out that's what some casters here were using. I've downloaded it but only scratched the surface. Do you have to manually, frame-by-frame, identify points to be tracked or is there some way of streamlining that process? I'm on a Mac , for what it's worth. I was hoping someone might have created a "how-to" video on casting or some similar activity to get me started. I mentioned Matlab because that's something I have a long history with---not with video tracking, though.

Few

[Few]

By the way, Gordy, I just noticed the 300 frame per second rate associated with the pretty graphs you posted. Yikes! I'll not be reproducing that rate with anything I have on hand or can afford. I think the eFly website says their gyros generate data at 30 pts per second (and I don't have 1,500 Euros burning a hole through the pocket of my fishing vest anyway). My iPhone will do 120 fps at HD resolution, or 60 fps at 4k resolution. Fortunately, I can't generate a "microsecond wrist" so I'm hoping I can learn something even from those data. An advantage of being limited by casting skill rather than available technology!

Few

[Graeme H]

G’day Few,

There’s a free video based physics software called Tracker that works well it’s only on the Mac OS though.

Cheers, Graeme

It's available for Windows and Linux as well Graeme. I am using the Windows version and it works great.
Cheers,
Nils

Excellent! I didn’t know that and I’ll look for it now.

Thanks