Excellent FP on modeling

[Graeme H]

Merlin,

That time will come. We have been talking about a model for aerodynamic lift in an inclined section of line lately and we're yet to see evidence supporting that model. I and others have posted evidence against it (now removed from the thread), but it would be refreshing to see something coming back the other way.

Cheers,
Graeme

[Merlin]

That is not a model Graeme, but a calculation. It does not belong to any model of a line flight to my knowledge. To me the check is a thing of the past, I checked it quite a long time ago when Gordy performed his tests.

Your videos raise question about its use: in a tethered case, the inclined section does not seem to belong to the loop. In a untethered case it does apparently. We can also wonder whether the way tails are attached to the line influence the observations, they are not completely free. I should try to make my own.

Merlin

[Paul Arden]

Perhaps in addition to casting a chain we could cast something that offers a HUGE amount of drag to see if we can generate some visible or measurable lift? I've just tried casting some fritz chenille but I only have a very small section. If we had a long piece maybe we can actually get a situation where we can see lift or DOG. And in this particular case it would be quite easy to cast it from an angled tube! Instead of chenille is there some material that anyone can think of that would produce even more drag?

Incidentally there is no need for anyone to take any of this stuff personally. Goodness knows, it's not going to shake the world, but it would be nice to know, right? Let's keep our heads right here. Lift either does exist or it doesn't - and if it does then it needs quantifying. If we can do the right experiments then we will know and all live happy fulfilled lives afterwards, whatever the result.

Cheers, Paul

[Graeme H]

I've done this experiment and filmed it Paul. I'll post it in another thread.

Cheers,
Graeme

[Paul Arden]

This is great, man. I can see now that you are coming to fish how are brains are working together. We certainly won't be needing walkie talkies if we separate, instead we can use telepathy.

I will buy some PVC tubing for a repeatable starting position at the beginning of the forward cast. Three or four of these, stuck together with duck tape, suspended off a jungle vine.

Cheers, Paul

[Graeme H]

You'll be able to by proper joints with the PVC ...

[Walter]

To make it short (see post #2 in the “Thank you” thread), a model is just a tool and must be fit for purpose. In my case they are built to allow comparing situations, identify trends which can be understood (in the sense of physics), and not to make accurate predictions.

https://www.sfu.ca/~vdabbagh/Chap1-modeling.pdf

Merlin,

Without making predictions how do you verify the accuracy and/or applicability of the model? Currently we have two (actually threeish and probably more) models of a fly line in flight.

The first is that there is no lift associated with a fly line in flight. What we see is described by Newton’s laws, in particular N1. Can we make a prediction based on this model? How about if we cast in a horizontal plane? This model predicts that a horizontal cast will not be pushed sideways and that’s what we see happen. What if we cast an upside down loop? This model predicts the line won’t auger into the ground and that is borne out by observation. It tells us that in the bouncing bomb cast the rod leg will get lifted off the ground. It does.

The other model(s) say that there is lift and that this lift is produced by many possible sources - aerodynamic forces in an inclined section at the bottom of the loop. Momentum flux (sorry but I’ve never been interested enough to actually understand how and where this momentum flux exists in the cast and how it generates lift). Some sort of paddle wheel effect at the loop. Aerodynamic effects in the fly leg.

Let’s look at the case where the lift is generated purely by aerodynamic effects either in the inclined section of the loop or somewhere else. This model predicts that a sideways cast will deviate from its course similar to a hook or slice in golf. We have no evidence of that. Observation tells us this doesn’t happen even in very long casts. It tells us that if we cast at a downward trajectory the loop will auger into the ground. It doesn’t. I’m not sure what it tells us about the rod leg in the bouncing bomb cast so for me that would fall under limitations of the model. If someone has a more sophisticated version of the model that makes a prediction in this case I would be interested to see it.

The other, non aerodynamic, causes for lift? I have to admit that I find what has been presented is somewhat vague and hard to understand. What do these versions predict? So far they only seem to predict that when a reasonable experiment involving fly line casting is put forward to prove or disprove the accuracy of the model that the model or assumptions will quickly change in an effort to negate the value of the experiment or that some sort of “read across” from things like falling bead chains and springs or string shooters will be introduced to dispute the results of the experiment.

Models need to make predictions.

My two cents worth.

[Torsten]

James,

Now, unfortunately in the case of the 'drag induced lift' paper by Gatti-Bono, she has chosen the wrong primary causation. So in effect she's taken a minor (or negligible) parameter, i.e. aerodynamic lift, and built the whole model around it. As such she's neglected the overriding factor - the line's trajectory.

In general I think this paper fails to prove or show the conclusion that drag induced lift has any influence on distance, but I think it's something for a separate topic, where I'll review this paper.

I'm sure she's an intelligent woman and if someone early on would have asked her - 'Caroline - why don't my side casts veer off in the direction of the rod-leg' I'm confident she'd have instantly recognised the error in her assumptions.

That casts veer off in any direction is not something they seem to claim and it's to me a wrong expectation.

My subjective speculation: in my opinion is that unlikely, because drag forces are unevenly distributed and when we're talking about the transition between top / bottom leg only.

However, the peer review team almost certainly wouldn't contain any fly-fishing physicists who would know that side-casting was even a 'thing'.

I can't imagine that you'll find many fly fishermen among peer reviewers. As I've written in another thread - a peer review is no warranty for quality and that the overall publication makes sense.

It appears that you, Gordy and Merlin cannot see this flaw also. You treat trajectory as a minor parameter because you've already 'fudged' your ideas/models not to need it. You then look for confirmation in the most obscure of places, rather than looking at actual fly casts.

"It appears" *for you* in your imagination maybe - the rest of the above paragraph is completely false. First: we're individuals, that means I don't have to agree with Gordy or Merlin, second just because I've not recently written about that implies not that I'd treat the trajectory as "minor parameter". The opposite is true and I've written many years ago how important that is. This goes back to the first bead chain experiments, I was the first motivating that.

Can I suggest you take a long hard stare at Bernd's Dangly Bits (BDB). Now I know Bernd isn't a physicist but his picture of the progression of a fly cast is insightful and absolutely matches reality - as it should because he drew it after watching what happens with real fly-casts. What he saw was a very complex trajectory with the fly line going up, down, forward and to the side. In it you'll see why the fly leg lifts.

Well there is nothing wrong about this. I'm just not a fan of mixing everything together, but rather to subdivide a problem.

[Walter]

In general I think this paper fails to prove or show the conclusion that drag induced lift has any influence on distance, but I think it's something for a separate topic, where I'll review this paper.

From the paper in question (Section 3: Example and Conclusions):

This note explains a fact observed by fly casting experts,namely, that a climbing loop is advantageous in distance casting.How to generate a climbing loop through control of the castingstroke is left as a 共considerable兲 exercise to the reader

[Merlin]

Hi Walter

Your last post deserves a long answer; I shall try to make it as short as possible. We are not on the same page about models for sure. I cannot clearly identify the models you mention and who runs them. I know a couple of models in the academic range: Gatti Bono and Wang.
Caroline Gatti Bono and Noel published a document in March 2003 explaining how their model was built. At that time, there was no lift discussion yet, the aim of the paper was to try validating the model, but in a limited range when we see the experiment. The paper is “Numerical model for the dynamics of a coupled fly line / fly rod system and experimental validation”. Here is an interesting graphic:

GB model experiment.JPG (29.97 KiB) Viewed 4317 times

If someone has done better that that, feel free to publish it. A controversy rose when Noel wrote a paper on lift at the same period of time “What are the drag forces on the loop and how do they slow its fall”. This paper is limited to a half circle loop. Then CGB and Noel published a paper “Effect of loop shape on the drag induced lift of fly line” in 2004, reviewing aerodynamic forces on 4 loop shapes, including the inclined one, with the conclusion that you mention in your last post. At no time there was a check with the earlier model. Comments and conclusions of these two papers are what you call “read across” I believe, they just use reasoning to provide conclusions, without any check of their relevance.
Today we might hope to be able to run such a sophisticated model and implement the conditions of the cast performed by Graeme and analyzed by Dirk. Then we could see if the loop nose has a down up trajectory, but this is just a dream.

The second model I know is the Wang one, here is an illustration:

Wang model.JPG (30.24 KiB) Viewed 4317 times

Dr McGlinn uses also some similar modeling in his paper which was discussed on the old Board. None of these authors validated their model against an experiment. In the case of Wang, he considered a critical damping case (no counterflex) incidentally. Nobody is perfect, you know.

I guess that the description and comments you make of different models is based on reasoning and not on actual model runs. To my knowledge there is no “lift or no lift” models, these are mental concepts leading to various opinions and speculations today.

A few words about my own models. I have three main ones. The conditions are always the same: vertical plane, parallel legs, loop already shaped. Two are “horizontal”, they do not include gravity nor lift or fall. One of them is able to consider line taper and was built for simulating Lee’s dual casts. The third one is a snap cast model which includes gravity and reaction from the collision with the ground. It was built to simulate a video from Graeme. Are they accurately predictive? No. Are the observed trends simulated? Yes (sometimes not like in the case of the Long Spey lines, but the subject stalled, and understanding as well). Thanks not to ask me any prediction about a potential lift. I calculated the same type of things than CGB or Noel with specific spreadsheets, but I do not call that type of calculation a fly line flight model.

Incidentally, there is no fly line flight model in use for line design as far as I know: Bruce explained the way SA was designing lines and from private conversations they are not the only ones to do like that.

It would be wise not to refer to imaginary models, which would avoid inflating controversies. We have a problem of understanding which may never be solved because the “best” models need powerful software for solving let’s say 6 ODE’s with a suitable converging mathematical methodology. That looks like a big black box, and when you change an input you do not have the explanation for the output. We know the GIGO effect of black boxes: garbage in, god spell out.

Does calculation on loop confirm a lift effect? No, since this calculation is not considering the full line.

Is the string shooter confirming a lift effect? On the string shooter yes, elsewhere it is not yet proven (it’s all in the rod leg) and IMHO it is unlikely for a fly line. But maybe I am wrong and that someday someone will prove us that the line is taking support from the air to keep itself aloft, in suitable conditions, needless to say.

Graeme gave an explanation for the down up motion of the loop nose in his video, but this is an interpretation, not a demonstration. Better than nothing maybe, for sure this refers to reality and not modeling and frankly speaking, we have no mean to check that with a model today, even not the beginning of an explanation with a scheme showing acting forces.

Merlin