Mick and Emory:

Guide spacing may seem like an art to some degree, but I'm a true believer in phyiscal modeling and I've modeled some complex problems that I'm certain exceed that of a fishing blank and these models were predictive and found highly accurate when actual tests were performed. I may be wrong, but I have the impression that when I mention the word "formula," some folks are thinking a simple algebraic formula with fixed constants. But clearly, that's not what I'm talking about here. In fact, if folks had taken a peek at my applet and tried it out, they would have immediately seen that I've tried to distill out the major 1st order factors that impact guide spacing and make them inputs into the system of equations. The program doesn't compute these. The builder does. And based on these simple inputs it gives an approximate guide spacing to begin with. IMO, It is -as- good as any lookup table, in my experience, and -faster- than trying to find it in a lookup table.

It doesn't model the dead spots since I wrote the software in about an hour, but then again, guide spacing is so subjective that accuracy is NOT really important to rod performance. I don't think the lookup table model them either and after 95% optimization, (even before that), it's no longer the rod; it's the angler. So a starting point is a starting point. Sure, there's a dead spot near the top. That's easy to model and I have some ideas that will only eat an extra 50 million floating point cycles or so to compute the standard spin rod guide spacing, or 100 million FP cycles for a casting rod w/ Fuji New Concept spacing. Since your PC is probably over 1 GHz and has a comparable floating point unit, the computation costs an extra 0.1 seconds or less.

But the idea of software isn't to replace a look up table based yet again on empirical data. It's purpose is to give builders a faster way to say "what if" and get a spacing that should work fairly well based on physical principles. I use it and have found it works extremely well for keeping line slap to a minimum for any number of guides. But as a builder, it isn't me that tells me how many guides someone wants or the type of guides. A lot of times, the customer has said to me, "... uhhh, ya know my buddy Jake has a custom rod with fancy threads... can you make me one? " I tell him sure. Then he says that Jake has 10 guides on his boat rod. Can he get 12 guides on his rod?

Also, I've had the same type of request on a surf perching rod. I've had requests to take a 10 ft fast taper fly blank for 8 - 9 wt line and put 15 spin guides on it because the other guy on the beach had 12 on his. I can recommend that a spinning rod doesn't need so many guides, but the customer wins in the end and I have to make it work. In those cases, if the lookup table doesn't have the spacing because 15 guides exceeded the optimum, I can eyeball it or run the software. The guy will end up with 15 guides on a spin rod, 8 of them Fuji Alconite concept single foot mini's to keep weight down. Does the rod catch fish? It certainly does, and it has extreme vanity value too. :-)

As builders, I like the idea of making the best rod we can. But often, that's really subjective and constrained by what a customer wants. So I try not to stress too much over the guide spacing and make it good enough and fast to use. I have taken some tips here so far and will read the article later. But 1) I'll look into adding something to model flat spots at the tip, and 2) I'll look at dynamic load distribution and hoop stress models to see if there's a better fit for guide spacing that significantly changes spacing over a modified geometric taper model I use now.

As for manufacturers and their blanks, I haven't had as many issues with blank inconsistencies. But the makers are in a unique position to actually issue recommended guide spacings for various common guide brands since they know about flat spots. If any manufacturers are out there reading, it's something I'd support and provide input to.

James,
A model is just that a model. A model is not an actual blank. The problem is that every blank is a little different. For some of their high volume blanks a manufacturer may build them on 10 or 15 different mandrels and every one may be built by a different person who builds it slightly differently.
But let's back up a step. The optimum postition of the guides is important mainly for line control and stress distribution. But this positioning can actually be off from optimum by quite a bit without having a really major effect on line control or the amount of stress. If a given pair of guides are off, too far apart, by say 10% from what would be the optimum position the stress will only increase by roughly 10% and in most cases this 10% will also not have a dramatic effect on line control. The number and position of the guides is also somewhat a function of the size and type of guide used, for example high frame verses low frame.
It is also a function of how the rod will be used. In my judgement the number, size and weight of the guides is much more important for any rod that will be used to cast. A few too many guides can have a very significant effect on the performance of the rod. Too many guides will lower the rods resonant frequency which lowers the tip velocity which means that it will not cast as far or will require more effort to cast a given distance. But the size, weight and number of guides used on a heavy boat rod that will not be used to cast is not nearly as important.
There are a lot of what are basically trade offs. If I understand what your software allows the user to do I am not sure you are making the right trade offs. I would argue that the software should tend to lead the user toward the fewest and smallest guides that are consistant with how the rod is going to be used.
As far as the modeling itself is concerned. If you are modeling the guide spacing that is one thing but if I understand you seem to have suggested that you are modeling the blank or rod. If you are attempting to model the blank I do not think that you will be able to get by with algebra as you suggest. The characteristics of a blank are going to be a function of a several things that do not lend themselves to being modeled with algebra. I think that you are probably going to have to use partial differential equations to model several things like the sectional stiffness and sectional weight.

James,
Your comment above that all vendors follow standard descriptions......for tip flex, medium flex, full flex, or fast, moderate fast, fast, extra fast action all follow a similar taper.......uh.....big opening here I think. I think some advertise their rods as fast action when they really are not.....just for the fluff. Just oversimplified I suppose, but with all the anomalies of each manafacturer and the grade of material and his process, good/fair/or bad......I can't see a computerized calculation system helping me place guides on any given rod. But if I thought it worked.......I WANT A COPY OF IT!
Jim

OK, so my degree is in creative writing, not physics or math, but I do write a lot of software in my day job. This thread is fascinating.

How about this approach:

1) hang a static load off the tip-top--and just to keep it interesting, use enough pennies to deflect the rod 1/3 its length (CCS)

2) measure the distance from the floor to the blank at fixed intervals--say, every five or ten inches

3) compute the curve of the blank under load using polynomial regression, working from the fixed points you observed in step 2

4) minimize the area of the curve between the blank and a hypothetical line segment running taught between two guide locations--accounting for guide height, just for fun

5) "walk" one of the two guides until the area beneath the curve achieves the ideal compromise between tightness of line fit (least area) and number of guides (fewer is better)--> this last point is a fun challenge, as the area minimization tends to increase the number of guides, while the principles of good design would want to minimize weight/friction, hence minimize the number of guides

The benefits of this approach:

- you're using a true representation of the blank's curvature under load, not an arbiturary/logarithmic curve
- you also get your CCS data in the same step (hopefully helping CCS to become a standard)
- you're accounting for guide height, which is an interesting fringe benefit

The pitfalls:

- you've optimized guide placement for the rod when it's "loaded" (per CCS definition), but not at other deflection points
- you've assumed that the goal of good guide placement is to make the line follow the blank as closely as possible

Can anybody tell me how a calculation of the rod's resonant frequency might help eliminate my need for the polynomial regression analysis?

For what it's worth, I'm working on the software to drive a Web site devoted to the above--and would love your comments. So far, I tend to achieve about a 3-5% reduction in the area beneath the curve using this algorithm compared to my typical guide placement approach (which mirrors your Java applet, James).