Need some help Don't really know how to describe it.

[TheFunkBomb]

Good afternoon...

Let me tell you about me, what I'm trying to do and why.

I have always had a very scientific mind. Biology and Chemistry mainly. I haven't gotten into Physics but it's time. I believe in the scientific process of limiting variables in order to prove a theory.

I am also a fisherman who makes fishing lures. There are a ton of us out there but most of the amateurs stick to amateur processes. The basic process is to turn a lure (on a wood lathe), rig it up and test it out. If it has the desired properties, great. If not, it's back to the drawing board. This is very time consuming, old fashioned, and you have no guarantees of good results.

I have tried to improve the method by using CAD and a duplicator. This allows me to make minor changes to a previously failed lure and to see the effects of these changes. This has also gotten me shunned in the fishing lure world. You probably didn't even know there was a fishing lure world. :D Many don't see the point. They don't see how the improvement of computers can improve the quality of rather ancient designs. They believe I am making something more complicated when it doesn't need to be. Their loss.

My new idea is to continue the CAD idea but to make very different changes. I want to combine different densities of wood. I want to include materials that no one could turn before. I want to make designs that are so far out there that will make these people drop dead at their lathes. I want to take precise control of my lures and be able to make them behave as I want instead of being happy with mediocre results.

Unfortunately, when I opened this new door, I opened infinite variables which is disastrous to the scientific method.

What I am looking for is a way to take a 3d computer object and test it without actually carving it first. A simulated test tank. Or... at least an understanding of the physics that goes on when a lure is retrieved.

I really don't know how to describe it better without droning on and on.

Here is the deal... I just want to use science to help me. I want to go beyond what anyone else is doing. I want to prove to the world that I am onto something new and exciting.

If you're at all interested, just respond. I have ideas and I want to bounce them off like minded people who are filled with curiosity and the same desire to move forward.

Thank you,

Patrick

 

[Danger]

Welcome to PF, TheFunkBomb.
(My fingers get tired typing names like yours, so henceforth I will refer to you as 'Funky'. )
While your basic idea is not bad, I see one insurmountable obstacle (unless you have a super-computer). You have to model not only the lure, but the fish as well. I don't know how you can do that. It seems to me that, in this case, the old-fashioned way might be best. Past experience, common sense & trial-and-error might be more practical. I'm not putting your idea down, but it might not be within reasonable reach.

 

[TheFunkBomb]

Welcome to PF, TheFunkBomb.
(My fingers get tired typing names like yours, so henceforth I will refer to you as 'Funky'. )
While your basic idea is not bad, I see one insurmountable obstacle (unless you have a super-computer). You have to model not only the lure, but the fish as well. I don't know how you can do that. It seems to me that, in this case, the old-fashioned way might be best. Past experience, common sense & trial-and-error might be more practical. I'm not putting your idea down, but it might not be within reasonable reach.

Your point has merit. Fish can be unpredictable. It takes years (even decades) to understand the biology. Why do they show up at spot A and they're not at spot B? Why are they eating this and not that. This is where the scientific method fails because the fish itself has so many variables.

Fortunately though, after years of past experience, you start to tune the ideas. While it may not be exact, you get a better understanding of why something happens. I can determine that the fish will be here because of the amount of food, the water temperature, the wind direction. So while it isn't an exact science, you can make educated guesses.

Basically, I want to make educated guesses. Up until now, there has been a lot more guessing than education in the lure making field. I see this as counter-productive and leaves too much up to luck.

I'm not so concerned about the fish reaction in a sim. I have a pretty good idea of what I'm looking for. I know what triggers a fish to bite but what I want to see is how to maximize the triggers. I want to look at my sim results and figure out ways to make it a reality. I also want to look at videos of the natural forage and find ways to impart those actions into a lure.

There is a lot I want to accomplish and unless I have a thousand monkeys at a thousand lathes, my Shakespeare will never be complete.

All I really need is a simulator.

 

[FredGarvin]

It sounds to me like you need a good CFD package and the expertise and hardware to run it. You are better off with the build it and try it method. The only thing I would say is to see if you can somehow use lesser materials for your prototypes so you can quickly get them into testing. Once you have a geometry that you like, you move to making it out of the real stuff. That is exactly how a lot of complicated aerodynamic/fluid mechanics is still done. The situations are so complex that useful simulations are either extremely computationally intensive or they are impossible.

 

[TheFunkBomb]

It sounds to me like you need a good CFD package and the expertise and hardware to run it. You are better off with the build it and try it method. The only thing I would say is to see if you can somehow use lesser materials for your prototypes so you can quickly get them into testing. Once you have a geometry that you like, you move to making it out of the real stuff. That is exactly how a lot of complicated aerodynamic/fluid mechanics is still done. The situations are so complex that useful simulations are either extremely computationally intensive or they are impossible.

I was afraid of this answer. Unfortunately, lesser materials are out for prototyping because wood is the cheapest and very hard to substitute for in this application.

Let me tell you guys what I'm trying to do and maybe you can give some suggestions.

Usually, a person comes up with an idea, chucks some wood(usually a lighter wood like cedars) on the lathe and turns a body. They fit a lip on, drill holes for hooks and weighting and then test it out. If it doesn't work right, they may try a different lip or a different weight position.

Lead is used for the weighting and while lead is great, it has some drawbacks. Lead is heavy and must be used sparingly in the right position. Sometimes the positioning is terrible for long distance casting.

My idea comes from CAD. Since I am using CAD and a duplicator, I know exactly how my lure is going to look when it is finished before I even chuck the blank. This gives me the power to tinker with the previously inaccessible insides.

What I want to do is to take a lure "blank" and slice it into thin layers. By splitting my CAD drawing into the same width as the blank layers, I can do cool stuff like hollowing out the inside of the lure.

This will allow me to do really awesome things like fitting heavier wood inlays (such as Maple) into each layer and then laminating the layers back together and turning the laminated blank. What I'm hoping to accomplish is a more natural weighting system instead of just a slug of lead.

This hollow chamber (which I've started calling a cocoon) has no set shape. I am no longer held to the archaic rules of light wood and heavy lead for a surface swimmer.

Not only can I fit different wood inlays, but I can also add things like rattles, un-turnable materials like foams and acrylics. Or I could fit in a magnet system like they have on plastic lures that allow ball bearings to shift to the tail end for long casts but once they slide forward, they're stuck by a magnet.

If you haven't noticed, I'm quite possibly insane. I'm just trying to jump start this "industry" into thinking on a 21st century level.

 

[FredGarvin]

I understand what you are doing. What I am saying is that simulations like the ones you are looking for are not easy nor cheap. If you really want to think 21st century, you have to think proper flow simulation.

So it is your contention that not only the geometry of the part is key, but the distribution of mass as well? I guess I still don't see why you can't simulate with lesser materials put together in a lesser fashion simply for the purposes of testing. Even making the test article larger than the actual part can be an option. If you are not able to entertain the idea of having a consultant do some CFD work for you, then where does that leave you? You have reached the limits of what a CAD program can do for you. You need to move to the next level or get very creative with the old ways.

 

[frankinstein]

Unfortunately, when I opened this new door, I opened infinite variables which is disastrous to the scientific method.

Are there infinite variables or just allot of variables and how constrained are the issues you want to test in your sim?

I worked for a company that developed simulations of electrical, hydraulic and pneumatic circuits for weapon systems. These sims were mainly for training and troubleshooting purposes. When I first started at the company their approach was very limited since they would run into the infinite variables road block. But after careful inspection the rules for these circuits were actually constrained by causal properties. In effect process could be broken down to cause and affect events that constrained to a limited number of paths. Of course which paths a system could take on was unpredictable and that is where most of the programmers would give up at this stage. But again careful examination demonstrated recursive rules that were parameter based could be executed for each and every scenario possible. So inputting the contingencies simply meant looking at the system from a very high level and apply the recursive rules based on the dynamics of the system at any given time. No need to pre-conceive or code every possible scenario just understand what each element is supposed to do based on fundamental behaviors. With that said the system acts out what it needs to do based on a limited set of data for a particular scenario branch, all other branches are ignored.

If you can break down you system into very high level processes and orient contingencies based on rules and not data specific processes you may be able to develop a system that can emulate what you would like to know about your lures.


Frank

 

[mshinavar]

genetic algorithm...'nuf said.


not quite sure how you could program this into your GA, but if you could get the GA to have an understanding of flow dynamics, that could for the most part solve the fluid dynamics. also, you can enter in fish biology tendencies, really enter in any factors you want.

luckily, you may not need a super computer. yes, there are almost infinite possibilities, however GA's are artifically smart enough to realize bad ideas, and discount them. think of like an inversed geneology tree. you have throusands of inputs (in this case) and 1 final "best" solution. when the GA finds a bad trait, that whole branch is cut off and prevents further "bad" offspring.

additionally you could place priority on the different inputs, say you want the GA to design something more targeting the biology and character traits of the fish, and have the fisherman reel a little harder (higher drag penalty)

or you could just make a lure that looks like what the fish normally eats, worked in the past ;)

 

[gmax137]

So, am I correct in saying, your question is, if you draft up a lure in CAD you want to predict how it will fly thru the air when you cast it, and how it will wiggle and move thru the water when you reel it in?

If that is your question, I think Mr Garvin's response is right - very hard analysis using expensive computational tools. And really only will produce useful results if you understand the fluid dynamics yourself (so you can ask the computer the right questions and also understand the answers it provides).

Consider a more empirical approach - maybe you can build yourself a small tank and push water past your prototypes to see how they behave. Like a wind tunnel, only using water.