Bamboo Tips - Tips Area
Stress Curves

< Home < Tips Area < Stress Curves

In it's most basic use a stress curve shows you how close a split cane rod is to breaking with the weight and length of line you specified. Garrison believed 200,000 ounces per square inch was a good, safe upper level. In reality  you can  go up to  220,000 or 230,000 without any  problems.  Garrison  himself  went  up  to 220,000 on his lighter rods.  Garrison believed  that  below  the 140,000 point the bamboo stopped flexing.

Some basic information on stress curves:

The X axis is rod length, with the tip on the left and the handle on the right.

The Y axis is ounces per square inch. This sounds like a pressure, as in PSI, but it is really a stress measurement.  The square inch refers to the area of the cross section of the rod at that point.

The higher the stress curve goes, the more the rod will bend, taking the  curve as  a whole.  I  say this  because a  reading of 200,000 near the tip, where the rod is thin in diameter is going  to bend more than a reading of 200,000 near the handle where the bamboo is thicker. But, if you compare a whole stress curve of one rod to the whole stress curve of another rod, the rod with a higher reading at the same point along the length will bend more at that point.

A stress curve means nothing if you can't relate it to something. The only way you can start to see how a stress curve can help you is to graph up the stress curves of real rods and cast them. In my case I didn't have a bunch of rods to try. I had to make them. I now have a bunch of experimental rods lying around, but fortunately I was able to sell a few of the better experiments. I was also lucky in accidentally making a rod I really like (a Cattanach taper) on my third attempt, and being able to contrast that to a rod I really didn't like (a Garrison taper).

As you get deeper into stress curves you can begin to pick out certain  characteristics  that tell  you what  kind of  action the  rod has, or will have if it hasn't been made yet. A Garrison rod, which I consider to be slow, has a well rounded "hump" near the tip and a fairly slow drop off as it goes towards the handle. Let's see if I can do this with ASCII art.

|
|             *  * 
|          *           *     *      *    *
|         *                                     *      *      *      *
|        *                                                   
|       *                                                           
|      *
|     *
|    *
|   *
|  *
_*______________________________________________

The rod that I'm always raving about, the Cattanach 7' 0" 4 wt, I consider to be fairly fast. It has a stress curve like this:

|
|             *   
|          *     *
|         *          *  
|        *                 * 
|       *                       *     
|      *                                *                         *
|     *                                        *     *     *            *
|    *
|    *
|   *
|  *
_*______________________________________________

The blip near the handle is the Cattanach hinge, and it greatly enhances roll casting. Don't forget to put it in. I did and the rod I made was a terrible roll caster. It isn't as necessary on longer rods, but on shorter rods it's definitely needed.

A Paul Young Para 15, what is described as a parabolic action looks like this:

|
|             *  *                             *  *
|          *         *                       *        *
|         *             *                 *            *
|        *                  *           *                *
|       *                        *  *                       *
|      *
|     *
|    *
|   *
|  *
_*______________________________________________

Looking at this rod, and never having cast one, I would guess that it would feel fairly slow, due to the enhanced bending near the handle, but able to throw a lot of line, due to the stiff mid section. I would also guess that it could roll cast really well.

There are as many variations of stress curves as there are rodmakers. This covers the slow, fast and parabolic actions.  (Darryl Hayashida)

Being somewhat new to the craft of rodbuilding, I have taken it upon myself to try to understand fly rod stress charts.  Now, not being an engineer, frankly, they are 'stressing' me out.  Please excuse the pun but it has been a long day.  If any one can lend advice in this matter it would be greatly appreciated. ie: what a slow, medium, or fast action rod stress curve looks like, what a slight kick upwards in a stress curve indicates etc.  Or if someone can point me in the right direction as far as understanding them (web sites and the such) it would be greatly appreciated.  (Robert Cristant)

The stress curves don't mean a d--n thing to me, I find the dimension graph tells me a lot more. Look at what the actual shape of the rod is. You can use a rod, or rods, you know as a benchmark to compare other rods against.  (John Channer)

We've had lively discussions of this topic.  I think we are disinclined to launch into it again. You should check the list archives for those months.

Start by reading Darryl Hayashida's simple description (above) of how to read stress curve charts.  (Frank Stetzer, Hexrod, Taper Archive, Rodmakers Archive)

I have heard a lot of you talking about stress graphs over the past couple of years.  I only got a single reply yesterday, with a good suggestion to look at the Hexrod web site.

However, I am still looking forward to a good explanation about how to read those stress curves, because they are with all the tapers I looked at in the Rodmakers archive.  What do those things tell me?  What difference do they make?  Why are they there?

I also asked about ferrule sizes for those tapers.  What do you guys do to determine ferrule size when looking at the tapers in the archives?  Do you look at a dimension at approximately the midpoint of the rod, and double it to get a close diameter?  (Jason Swan)

Darryl Hayashida has a good explanation of stress curves, which explains how most of us use them to judge and compare actions of rods from their tapers.

Like everything else, different people use different rules for choosing the ferrule size. Some round to the nearest 64th, some round UP to the nearest 64th (never down).  Obviously the more cane you remove  to fit the ferrule, the weaker that spot on the rod becomes.   (Frank Stetzer, Hexrod, Taper Archive, Rodmakers Archive)

You are approximately correct in how I determine ferrule sizes.  I build the blanks, measure them and fit the appropriate ferrule.  (Bill Lamberson)

To get a understanding of stress curves try this. Pick a taper, get a calculator, get the stress graph of the same taper. Start subtracting one station from the other, look at the corresponding point on the stress graph. Repeat this procedure for every dimension from the taper. You  will start to see a pattern occur that corresponds with the amount of change from one station to another and the amount of stress in the graph.

Find several rod tapers to do this with and you will get a better understanding of what stress graphs are all about. If you want to find a good rod to make. Forget the graphs and ask.

Use Frank’s Hexrod to get the ferrule size I need.  (Adam Vigil)

I've been pondering Garrison's stress curves again - typical flight of fancy on my part, when the mundane things are going on enough to keep me busy, but the brain is on over drive....  How did he come up with the numbers for a given dimension?  I realize that for a given dimension, the stress number is going to be higher for a smaller dimension, but is there a stress number per decimal inch for bamboo?  I tried dividing the stress for a given dimension by the given dimension, and for a given set of dimensions, the stress value doesn't agree from one dimension to another.  So, how did he figure the stress for a given dimension?  (Mark Wendt)

The dimension at any one point, per Garrison, is based on the total bending stress placed on that location.  The total bending stress is a function of the total mass placed in front of that point:  the amount of varnish, the number of guides, ferrules, bamboo, wraps etc.; and their location from that point.  Change any of these components and you end up changing the stress, which then changes the diameter at that point. 

You seem to want to know how the diameter is calculated.  Its based the formula for bending stress (I isn't an engineer, so if I screw this up somebody please step in):

rod diameter = [M / (s x 0.1203)]../p>

AOL will probably screw up those characters, but its supposed to read that the diameter is equal the cube root of the total moments at that point divided by total bending stress at that point divided by 0.1203.  The constant 0.1203 is specific for a hexagon, and would be different for a square or a pentagon.  I can go on of you like, such as how the moments are calculated, and the stress in calculated, etc., but most rod makers seem to get turned off at this point.  (Kyle Druey)

Here! Go here and Read to your heart's content and your mind's satisfaction. ;o)  (Martin-Darrell)

I have created a program to calculate stress curves, using Garrison stress curve math. In my program I start off with a graph, and a straight line. At each station I can click on the line and drag it up or down. When I drop it the taper diameters are recalculated and displayed. It works well for me, but it isn't very user friendly. One problem (among many) being that if I don't go from tip to butt in order I get a divide by zero error. Perhaps I need to start off with a straight taper instead of a level line...

It works well for me because I have become stress curve oriented in designing tapers. I start off designing the stress curve, then I look at the diameters and make the rod.  (Darryl Hayashida)

Garrison stress curves take into account the rod from the point in question to the tip. It does nothing with the rod from the station to the butt. It also assumes the point you are looking at is fixed. But the way I look at it is I'm not looking for total "truth" in the stress curve, I am looking for something that will give me a similar shaped curve for a similar action. So far Garrison stress curves seem to be doing at least that much.  (Darryl Hayashida)

HOW DO YOU KNOW WHEN YOU HAVE THE BEST ROD DESIGN?  (Jerry Foster)

I guess you know when you've achieved a good rod design program when you can trust the graph enough to allow you to use the original rod you like as a point of departure to design another either by saying this #6 is a great rod but too heavy so I now want it as a #3 or #4 and the program gives you the exact figures which I know has been done OR and this is where I think the work is tricky you say I like this rod but the butt is too thin or thick or whatever but I still want the same action and length and line weight. 

If you look at it from this point of view I don't think using stress curves as we understand them would work. You'd need to compensate for the increase in the butt along the entire rod instead of looking at each point independently.

On the surface this seems like it can't be done, if you alter the butt yet want the same action and line weight and length the Garrison style stress chart would show a very different graph even though the rod may in fact be similar or the same to cast but that's a shortcoming in the Garrison math. Garrison is using point by point bullet ballistics where as I think it needs complete recalculation of every point above and below any alterations using what I guess is rocket ballistics. You simply can't work some problems out using the wrong math. Possibly to achieve a WC 7'#4 with a thicker butt but retaining the same casting action you'd need make the changes at the butt but the trick to keeping the action might be to alter only the stations 15 and 20 at the tip or something like that. Maybe to achieve the same casting action the actual casting length of the rod would only be the last 35 inches towards the tip which is greatly reduced in section for example.

Using Garrison math and comparing the chart you'd say you've created a dog but it might not actually be the case when it came to casting the rod. What I'm suggesting is a complete departure from the Garrison stress curve as we understand it, thinking in Garrison stress and it's graphs only confuses the issue for what I'm suggesting.

How many times have you heard the comment that the rod cast completely differently to what you'd  expect from looking at the graph? Maybe what I'm asking really is impossible but maybe it's just a question of rethinking the approach rather than tinkering with the same formula. Maybe it's been done already. All I'm saying is with a few exceptions there don't seem to be a hell of a  lot of really different rods about, there are exceptions of course but I think to really break new ground it will take a different perspective on rod design. You can design boats that win America's Cups without building them to see how they go. The wing keel on the first boat to take the cup from America was designed 100% on computer before it was even a mold for the fiberglass model. Fluid dynamics are pretty tricky.

You can design sails on computer, air dynamics are pretty tricky. Racing sails cost $1m, you can't do too much experimenting on actual sails at that price, same with masts. You can design nuclear bombs on computer, I'd think atomic physics is tricky even if the result is easy to estimate. None of these would be possible if everybody kept fine tuning existing ideas, they all take a completely new perspective. I worked in Brisbane when the boat Ben Lexan designed that won the cup all those years ago but I had to run the place in Perth for a few weeks while the boat was being built. The boat was built in a large shed right across the road from where I worked and I often spoke to Ben. I wasn't allowed to see the boat of course but Ben did tell me what he did. Basically he threw the text book on hull design in the bin and with his by today's standards pathetically antiquated Apple he started again from basics. He couldn't have done it by just fine tuning what was already being done apart from start with the basic premise that you must displace enough water with the hull to keep afloat.

Now it seems hard to believe things weren't always done that way but much as I hate to say it because it's so hackneyed, he was told "it couldn't be done, all you can do is improve on existing design". Had he not been backed by a guy who wound in goal for about 10 years for being too free and easy with his public company money nobody would have given Ben the chance to try his concepts were so different to conventional theory. To look at racing hulls now they again all look the same but hull design is now improvements on Ben's ideas. Bamboo is natural and it's prone to variations sure but it's also a very high tech design as it happens.

So,  you'd know when you have come up with a good casting rod when the graph of the altered rod matches the original you have decided to use as a reference point 100% regardless of where and how the rod was altered to get there. Then you build it and see what happens. (Tony Young)

Jerry, I think another question to ask is what do you want the rod to do?  A good rod IMO is one that provides the desired performance in the most efficient manner.  There are no doubt numerous designs that would accomplish a given performance criteria, although there may only be one optimal design that does this most efficiently (it might not be well suited to ones personal taste, but from an engineering perspective it is most efficient).

I think you must first resolve what you want the rod to do, then you can go about modeling and designing it.  Do you want maximum line speed?  Quiet and soft presentations?  A diverse range of line speeds with similar presentations?  Others?  Once the performance criteria is decided then you must determine how the rod must bend to deliver that kind of performance.  For example, a quick taper with larger butt sections can generate tremendous line speeds, so do certain parabolic type tapers... but these two bendforms do it in a completely different manner.  Each has a completely different bending profile and generates the line speed with different efficiency.

Once we determine the performance criteria and develop how the rod should bend, then use more math to design the cross-sections that well enable the rod to bend accordingly.  Then, hopefully, we will end up with a GOOD rod.   (Kyle Druey)

I must confess that I am a pretty empirical rod builder.

I happen to be very fond of Jim Payne's designs, or at least of the interpreted versions of Jim Payne tapers that I build.

Because I like to have moderately quick rods, I lighten off the tips a bit and possibly give the lower butt profile a couple of thousandths here and there.

I find that, having measured and recorded these tweakings over a few rods, I am able to produce a pretty repeatable and predictable rod. Certainly not my design,   and probably not really much like what the owner of a real Payne would recognize as a Payne either, giving the vagaries of planing forms, of glues, of rings and bindings  and of varnish, as  well as  the variability of one-by-one manual production.

But I really do wonder whether I would be any better off if I tried to design rods from scratch, and whether the profiles would in fact be much different than those derived by empiricism. After all, how many tip impact factors can there be, assuming that differences  have to be gross enough for us to be able to translate them into bamboo with a plane, and have the subtleties survive gluing and varnishing.

It continually amazes me how much commonality there is between rod profiles when you take a half-hour or so and just examine the rod tapers in the taper archive, and ask the question again - do we really produce new rods or are we simply piddling in well-paddled puddles?

I really think that the time spent, by me at any rate, in buggerizing about with math that I don't really understand and trying to reinvent the firestick would just equate to less time spent doing what I like to do, building rods!  (Peter McKean)

That is my point. I'd imagine pretty much all rods are modifications and tweakings of other rods. That's fine but it's a dead end. It's a nice place to be for sure, nice and comfortable with great scenery but a dead end nevertheless.

It's just like a rock climbing crag that's been climbed on for 100 years like some of the British grit (that word again) stone. There are thousands of individual climbs there and every once in a while somebody convinces himself he's found a new route. Nobody can remember anybody ever having done this route before but the actual chances when looked at in the cold hard light of day of it truly being a new route are in fact vanishing close to zero. It's all been done before. The climbs are hard but some climbers are harder and until something new comes along that's about that.

A US engineering student rock climber by the name of Ray Jardine invented a fast action spring loaded camming device for use as running protection rock climbing called Friends. At first nobody saw why these would be any better than the hexes people always used, they looked complicated and seemed somehow against the principal of simplicity so as a result Yvon Choinard rejected his business proposition to buy the patent and it went to a British company called Wild Country. I know for a fact they made about 20 million pounds out of it because I worked for them for a while in the Brisbane office. Anyhow after these Friends appeared on the market suddenly people stopped considering 5.8 and 5.9 in the US system as hard and everybody was climbing 5.10 even 5.11. Then Boreal came out with sticky rubber for use on friction boots and there was a sudden explosion of new routes of 5.12. Little advances that opened the way in amazing ways. People were doing stuff completely un thought of only a year earlier. It's slowed down again now because technology has stagnated. The hardest climb in Australia at one time was a route called Country Road. It was graded 24 which is about 5.12 or more US. When the guy did the first ascent it was considered as hard as a free climb could ever be. Period! Everybody was so convinced of this they climbed it with aid to speed things up and put a bottle at the top with a note in it saying they would pay the bearer of the note $10,000 on demand. They did this climb with old style boots and hexes. A few years later guess who got the bottle? I even tried to collect the reward but they welshed. What was thought un needed and undreamed of wasn't. There are millions of examples.

All things are compromises from the house you live in, the car you drive and the rod you fish. As a rule the more basic the technology the more compromise you have to accept. It's a great old house, looks great lot's of tradition but it's a pain to maintain and draughty as hell. The car is a classic, they aren't made like that any more mainly because the brakes are terrible and suspension is dangerous. All the rods we're messing with, with the exception of quad and pentas are old style rods and we just nip and tuck a bit. Don't get me wrong, I'm not on some kind of iconoclastic binge here, I like the classic rods a lot. I want to be cremated with my Driggs and Dickerson 7614 taper rods but I can't help the feeling we're sort of messing about a bit doing it because that's how it's always been done and there is no other way.

That is why I'm so convinced a new level of rod development program may be useful. Imagine if you could plot the rod you use as a benchmark. The graph be it a curve like a distribution graph or ideally a straight line along the x axis as it's the zero mark you want to work from.  This graph isn't the stress curve, it's THE ACTION itself mathematically described and graphically displayed.  Now, back to compromise.

Currently it's accepted you can get a taper using a #6 and turn it into a #4. The compromise is rod weight. That's not a bad thing but it's a compromise, a trade off.  You can keep the rod weight and probably line weight and alter the action. The compromise is possibly the line weight and probably the rod weight. How do you keep the action and line weight but alter something like a regular to a swelled butt? Can it be done?

The variables apart from rod length are as I see it:

Rod weight
Line weight
Action

Presumably in the case of the conversion of a #6 to a #4 the graph, a curve or a straight line, doesn't matter which should be the same for these two rods (#6 and #4) even though the line weights and rod weights are different.

The action is the same, the rod weight has been traded to achieve the line weight. I say traded because the weight isn't an issue. Had you started with a #4 and wanted a #6 the weight would have increased so weight is a compromise, sometimes good depending on which way it goes.

Say you wanted to increase the butt width but keep the action and the line weight.  You should still be able to make compromises to get this but it may be tricky unless you have the means of calculating changes at all points of the rod as any change is made.

I made the example of the WC 7' #4. Good rod, love the action but I want a swelled butt but keep the action AND remain a #4. I said previously that perhaps what it needs is to alter the butt and to keep things sweet you may only need to adjust the last few stations in the tip at some or various points. I have no idea if that is so, I'm just thinking aloud.

We tend to think along the lines of more or less straight tapers, even in the case of paras but lets not forget the Marinaro tapers which were a series of independent tapers acting together as a whole rod. It's this holistic approach I'm suggesting but taken further. Possibly  a Mariner taper is what you'd wind up with by using this as yet etherware  program I have in mind. Who knows?

This program would have the taper displayed. I'd think a straight line along the x axis. You adjust the butt for e.g. and the line moves away from the straight line you begin with but it wouldn't just alter at the butt as the entire rod action has just altered. This graph is the ACTION so the whole line would alter. You then alter other points of the rod to try to regain the straight line BUT keep the line weight.  Doing this would allow you to stay within two defined variables, you're keeping the action, keeping the line weight but adjusting the rod weight or it's distribution along the rod. Why would you bother? Because it's a new tool. Apart from a lot of unprintable things a leading hand I worked with used to say, one worth remembering was that if the only tool you own is a hammer you'll treat all problems like nails. (He wanted me to buy my own screw drivers) Once you have a new tool it's up to people's imagination as to where to take it. I'm just mentioning modifying an existing taper but what else would this allow?

I'm quite prepared to accept it wouldn't work, it's just something I've long thought would be nice because swelling butts is something I've done a bit of and it's be nice to do this with more confidence than trial and error and to see if a particular rod would be the worse for it without guessing but there would be a lot of other uses.

I know I'm a mental midget but there are mathematical giants on list and this is the place to air these unreasonable ideas.

If people don't like the idea of this, you don't have to use it. Lots of people do climb in gym boots and twisted strand ropes too.  (Tony Young)

I see your point, but WHY, if you like the WC 7'0" #4, would you want to build a swelled butt on it?  Apart from the cosmetics, surely all it does is to shorten the action and you'd hardly need to calculate that.

In academic terms, I can see  your point about the necessity of pure design work, but as implied in your rock climbing analogy, real progress demands more than pure design; it needs new materials as well.

I would have thought that that work is certainly being done in rod building, but using graphite and boron, not bamboo.

My original comment was aimed at people like me, who are not really very computer-skilled. I am not a flat-earther, and I am sure that the honing of the Math skills once again to include a little differential modeling is not too big an ask, but to me it's not a valid use of the spare time available.

Different thing for a person with your computer literacy, or whose real job is rods, perhaps, but not a  thing that I am interested in doing.

What are you doing over Christmas? Quiet one here, with a bit of fishing looming, I suspect. And the rivers are looking really good - on my birthday I took 8 fish, all on a dry fly,   the  biggest 5 lb. 3 oz, the next two within an ounce either side of 4 lb.

Wild browns, of course!  (Peter McKean)

To continue on the climbing theme, "because it's there (to be done," was the full quote).

I might like a swelled butt because it's nice to know there is wood enough when a 10 LB brown inhales the fly.

What I'm getting at isn't so much for things like swollen butts and tinkering with existing tapers necessarily, it's just that if there was a means of maintaining an action which is different to the original taper, think about that, you could then go on to create rods capable of more than conventional tapers are currently because you can approach rod design from a different angle.  The crux of the matter is not stress curves, it's action definition.

On the surface it would seem that shortening the length of the action would change the action of the rod but it might not if you could understand the action and plot it and see what happens when you modify various sections along the rod. It may take tapering severely half way along the tip with a massive swelling after that, you couldn't really tell unless you made a rod like this or used this etherware program.  Some rods simply may not allow such radical redesign but you may wind up with a rod that outperforms anything else once you use the new tool for development.

I will remind of Vincent's tapers. They were a series of tapers one above the other. The action must have had something going for it yet there aren't any around people are talking about but where do you start in trying to develop one? Trial and error I guess but why if there was a tool to do it?

I have a personal theory that it's taken all this time for graphite and plastic lines to reach what bamboo rods achieved 50 years ago as far as trout are concerned. It's just that graphite has taken the brutal power stroke and speed to do what bamboo does in an elegant way and that's a more important point than mere aesthetics.

As you know, I'm not what you'd call lightly built and I have quite a strong casting arm and I far and away prefer bamboo, not for any reason other than I like it.  That is really the only answer I can find for why in this day and age bamboo is still of any interest.

There are a thousand reasons you could think of like wanting to get back to simpler things and the desire to use a truly hand made rod and the rod feels alive etc. etc. etc. but at the root of it all the bamboo rods still have to have something going for them that is definable and it's the elegance of the casting and the playing of the trout that cinches it IMHO.

If they were inferior they'd be curios which interestingly the monster British rods are. They're made by hand, anti modern etc., all the things bamboo trout rods we're all making are but they're also useless so they're hung on walls and used as kindling as I know a few were in my house when I was a kid.

No, I think there is still a lot that can be done with bamboo but there needs to be a different approach to design. I must say again at this point, I do love the classics and I'm not suggesting they are in any way inferior but there is no way forward by using them and making modifications ad infinitum. Maybe that doesn't matter after all, I'm not sure but I sort of think so.  (Tony Young)

I've had questions about determining line weight using a stress curve. I determine the line weight by plugging in different line weights into the spreadsheet I use until the highest part of the curve - usually the tip section - is at or slightly above 200,000. Or if I am starting out from scratch I decide what line weight I want and put in diameters to get the same result - high part of the curve at or above 200,000.

You can also use Frank Stetzer's online hex rod program - plug in different line weights until you get the curve to touch 200,000.  It's found here.  Put in your rod measurements, try different line weights until you get the 200,000 mark, and that's your line weight.  (Darryl Hayashida)

I much prefer to use tapers that do not need to have anything subtracted for varnish. In my opinion what good is working in thousandths when you throw in something as inexact as an adjustment for varnish. Now, however I need to come up with an adjustment since the only 2 Thomas 8' tapers I have are taken over varnish and I need to use one. Is it .004 to .006 for the whole taper (as opposed to form settings) and what is considered to be the standard?  (Bill Walters)

I knew there was another thing I used stress curves for but it slipped my mind - If I use a taper measured over varnish and I know the line weight, I subtract the diameter measurements until the high point of the curve gets to be around the 200,000 mark.  It is usually around .004 less than the measured diameters. If you don't know the line weight, keep the curve looking the same and adjust to the line weight you want. Also a good way to make the same rod in different line weights.  (Darryl Hayashida)

What are the axis values on the graphs I see with rod tapers?    (Lee Orr)

The X axis represents the dimensional length of the rod, and the Y axis represents the stress at that dimension.  (Mark Wendt)

So does stress equate to the amount of flex at that point?  (Lee Orr)

The stress factor is based on the cross sectional area of the rod section, factored with the weight of the length of the line (in grains), the moments of the guides, ferrule(s), cane, and the tip impact factor.  That's a somewhat simplified aggregation of what actually goes in to calculating the stress factors.  I went through the Garrison calculations once by hand, and there is a lot of things involved in coming up with the final figures.  (Mark Wendt)

Could someone please point me towards a published article explaining the left column of the stress graphs of the Hexrod graphs?  Stress values F(b) tells me what about the taper I am considering?  Do higher values in the column mean stiffer feel in the action?  I’m a bit weak on engineering skills (in other words please keep it simple) so any help or insight would be appreciated.  (Mike Monsos)

Wayne Cattanach, originator of the hexrod computer program, has a nice description of the stress calculations and what they mean; its on Jerry Foster's web site.

And here is a posting to this list by Darryl Hayashida in 1997 which explains it well too.  (Frank Stetzer, Hexrod, Taper Archive, Rodmakers Archive)

The more something is bent, the higher the stress is. So a stress graph shows you where, relative to the load imposed by casting forces, the rod should be bending the most. There is a good overall article on tapers and stress here. I said "should be bending" because the classic Garrison stress curve is a bit of a fiction in that it doesn't take account of the fact that a rod bends under a load. We can see that under a big load the tip of a rod isn't actually bending much -- most of the bend is further down the rod. However what the curve does give us is some idea of relative flexibility of different parts of a rod and enables comparison to other rods.  (Mike McGuire)

I do have a good idea of how the action of the rod will hopefully be like by the graph.  I guess my question is if you had a similar graph curve with the only difference being the position of the curve on the chart (one higher and the other lower on the graph) what would a caster notice.

Thanks for the links I’ll get to reading them and maybe answer my own question.  (Mike Monsos)

If the rods are loaded in exactly the same way, the lower rod would appear to be stiffer (smaller deformation) or require a higher weight line to accomplish the same deflection.  (Frank Paul)

Just to stir the pot a little on stress curves…

I have seen the same interpretation of Garrison’s stress as stated in  comment from below:

…

I said "should be bending" because the classic Garrison stress curve is a bit of a fiction in that it doesn't take account of the fact that a rod bends under a load. We can see that under a big load the tip of a rod isn't actually bending much -- most of the bend is further down the rod.

…

My personal OPINION is that the comment misinterprets what is represented by the data.  The Garrison figures represent the stress only at the point/station when there is a force of 4G’s applied to that station.  If you had a 5” rod, then the stress at that point would be what Garrison calculated.  If you had a 10” rod, the 4G force would produce the stress calculated.  That 10” stress says nothing about the 5” stations stress or the tips stress.

At some point in ever casting stroke each individual station is perpendicular to the forces being applied.  At that point the stresses in Garrisons calculations apply.

In my OPINION, there are some things a Garrison stress graph can tell you.

1. If any of the stations show a stress value of greater than 300,000 then you have a rod that has a higher probability of breaking at that point.

2. If all of the stations show a stress of less than 100,000 than you will have a rod, or broom stick, that will have a difficult time turning over the line.

3. You can discern where you might feel the most flex in a rod and whether it might be considered slow or fast or parabolic.

If you want to calculate the stress along all of the stations at any point during the cast you have to go beyond Garrison and add the angular dimension to the force being applied at that moment.  This is what Max Satoh and Al Baldauski are adding to their programs.  But then of course you have to learn what that data means and what you can do with it!  I think I will just go back to my basement and  see if I can  finish version 2.1 of my CNC mill sometime this year.   (Ralph Tuttle)

You nailed it.  The stress curve only shows the stress at that singular point at that singular time with that singular 4G force.  You have to look at each x-y point as it's own stress point.  And remember also, the Garrison stress curves are based on one inch increments, so you are only seeing the calculated stress for those exact one inch increment cross sections.  We've not calculated for the cross sections in between those one inch increments.  With the computers of today, and the software we have available, we could make those increments much smaller, but back in the day when Garrison was doing his calculations, all he had was a pencil, paper and a slide rule.  Just took too long to do a stress calculation on a rod taper using smaller than one inch increments, and for most purposes, one inch increments sufficed for what he was looking for.  (Mark Wendt)

I will stand by my statement that there is fiction involved here. The stress at a point is calculated from the sum of the moments applied from further out on the rod towards the tip. Any basic physics book will tell you that the a moment or torque about a point is the product of the force applied times the perpendicular distance from the point to the direction of force applied. The fiction is in the assumption that this distance remains simply the distance along the rod to where the force is, ignoring the fact that the rod bends and shortens the perpendicular distance, diminishing the moment. To give Garrison his due, he did the best he could with the tools he had. To do it right, one has to solve the Bernoulli Euler equation for large displacement of a tapered cantilever beam. Doing it analytically, the only approach available to Garrison, this is a very hard problem, but modern iterative computational  methods can handle it, as Al Baldauski and Max Satoh have done. Having said all that I do generally agree with what you say that a stress graph can tell you.  (Mike McGuire)

Okay.  We have to look at this from a couple of angles.  To help understand what's happening in a stress curve, lets first approach this using a standard, non-tapered cantilevered beam (we'll get to the taper in a bit).

Stress shows up on a standard non-tapered cantilevered beam when a force is applied to one end of it, while the other end is anchored (hence we have a cantilevered beam).  Since the beam is non-tapered the stress will pretty much show up as a straight line, disregarding pressure points, like the anchor point.  Picture if you will, this straight beam, bent into a bow due to the force being applied at the end opposite the anchored end.  You'll generally see the force distributed equally along the length of the beam (in a theoretical world), and the stress curve for a given force will generally be a straight line on the graph, and also generally proportional to the amount of force given to deflect the beam.

Now, we complicate things, by making that non-tapered cantilevered beam, tapered.  No longer is the stress distributed evenly the entire length of the beam.  The thinner parts of the beam will bend more, and the obverse, the thicker parts will bend less.

Lets say we develop a straight tapered beam.  We start at the thinner end, who's cross section dimension is 1".  The beam is 10" long, and increases equally an inch every inch, so that the thick end has a cross section dimension of 10".  We anchor that 10" end so that it can't move.  Now, we apply a force at the 1" end, sufficient enough to put a measurable bend in the beam.  The stress the beam sees at the thinner end will be higher than the stress seen on the beam at the thicker end, because the beam does not bend as much at the thicker end.  Since we're using a linear taper to construct the beam, you'll theoretically have a smooth curve from the high stress, thin end of the beam, to the low stress, thick end of the beam.

We further complicate this tapered beam by not making it a linear taper.  Now we start to see a not so smooth curve in the stress curve graph, because now the non-linear tapered beam is *not* distributing the stress so smoothly down the length of the beam.  That's why you'll see bumps and dips in our stress curves of the tapers we use for rod making. Those bumps and dips are deviations of the stress seen due to variations away from a linear taper.

We see higher stresses on the thinner end of our beam, because for a given force applied to the thin end of the beam, the more the deflection of the beam, and the less deflection of the beam at the thicker end of the beam gives lower stresses due to the less bending.

You can bend a materiel only so far, and then you exceed it's modulus of elasticity, and when you do that, the material becomes plastic and will not return to it's previous unbent state, and if the stress is too high,  the material ruptures.  (Mark Wendt)

I thought I might as well put in my 2 cents about Stress Curves.

First Stress Curves are not stand alone graphs. To start to understand them you must first understand what items go into making them show what it is they show. Without this information they mean very little. Garrison and Carmichael, in their book only gave one rather confusing example of how they are calculated. If you look on page 239 Carmichael states that there are five steps or items needed to calculate Stress.

1. The LENGTH, SIZE and WEIGHT of the line hanging outside the Tip Top Guide plus the TT itself.

2. The Line inside the Guides, Tip to Butt, this is a very small value.

3. The weight of the Varnish and Guides. Garrison Drew a Graph page 245 of the values which apply to a 6 weight Rod ONLY.

4  The weight and location of the Ferrules.

5  The weight of the Bamboo itself.

The values for each of these items that are used in this example are given, hidden, in this chapter. Normally there would be a set of Specifications listed, however Carmichael did not list them this way, none the less they are all in this chapter.

For information:

As each of the values are calculated they are multiplied by a Impact Factor of "4" to account for the "Dynamic Deformation", see page 242. The order of importance, or which item adds the most Stress is, The line outside of the Tip Top, the Bamboo, the Ferrules, the varnish and Guides and the Line inside the guides.

For this example;

The Line is 50 feet of 6 weight Double Taper-Floating line which weights 490 grains for 90 feet. Actually Mr. Carmichael used only 49 feet of line for his calculations because the Tip Impact of 2.50 for the 49 feet was easier to calculate, page 243. The Taper is approximated initially in Diagram 11 page 249. then again in Diagram 17 page 257.

What all of this says is that each Stress Curve is the result of very specific values given for each curve. If you do not state the values then the curve has no reference points and very little value.

I can by selecting the line weight and length make different lines have EXACTLY the same Stress Curve,  but only one will peak close to 200,000 oz inches at 50 feet of line, this will be the correct line for the Taper given.

To be able to evaluate what each Stress Curve means in terms of rod performance takes much practice. To do a new Design from a Stress Curve, I don't know, I only know that I can't, don't, do it and I would think there are very few who do. To modify an existing taper is much easier.  (Bob Norwood)

I want to thank everyone for trying to educate me about stress curves and their importance in rod making.  I’m planing to reread all these a few more times and try to get them through my thick skull.  LOL,  I’m also going to spend more time rereading the article that Mike McGuire linked  here that seemed to be very understandable for me.  (Mike Monsos)

"The weight of the bamboo itself"  (item #5)

Excuse me, but please explain how "the weight of the bamboo", for a given piece of culm, "itself" is determined. I once built up a short section of the material intended to be used, and found its weight per cubic inch, much different than the unit weight given in the Garrison/Carmichael text. Then, after doing roughly 19 single pages of computations, wondered whether the correct value was used.  (Vince Brannick)

Sorry if their is any confusion, what I was referring to is the items listed by Mr. Carmichael on page 239 right middle of page "and the weight of the bamboo itself". 

The actual Graph of relative weight of bamboo he used is shown in Diagram 10 page 248. Below the Diagram he states that for simplicity the weight of .668 oz/cu inch will be used for the calculations relating to the Moments caused by the Bamboo. You might have determined a different value.

Is this what you are referring to?

Also I didn't state that the Rod Taper calculations here are done for Garrisons 212 Rod. The calculations Carmichael is doing here had in fact already been calculated by Garrison a long time before, Carmichael is just going through them just to understand the process himself.  (Bob Norwood)

One point referenced with regard to the .668 value indicating an average(?) of the values of the bamboo with/without pith suggests a sampling of each by Mr. Garrison. The nature of such sampling though isn't described, The discrepancy with Mr. Well's figure of course may be the difference between tonkin and calcutta. It would almost seem that much confusion may be avoided through a realization of the infinitesimal effect some adjustments in the calculations may actually have ~ for example the shear force(s) of a person standing on the end a diving plank, along the length of the plank, measured perpendicular to the anchorage as opposed to the shortened length measured from points along the arc generated. Kind of silly, methinks.  (Vince Brannick)

There really isn't a confusion with your quote, nor with Mr. Garrison's design procedure used in arriving with the cross- sectional dimensions at the five inch increments. Where (how) the .668 oz./in.(3) "for simplicity", was derived, is the basic question. With recognition of a possible large discrepancy between the specific gravity of Tonkin cane  and the  Calcutta variety  of an earlier time, H.P. Wells, in his 'Fly-Rods and Fly-Tackle', circa 1901, gives the weight of one cubic inch of "six strip hexagonal split bamboo" as .57378 oz. (specific gravity .9915). At this point, it's incumbent on me to admit 'disremembering' the actual results of my weight sample, (quite possible at age 88?), but my computation sheets shows that for the moments of the "bamboo itself", the value of weight per cubic inch used in the calculations was .574 ~ too close to Mr. Wells to not think that I may have scrapped my own results, in deference to an authority. In any case, a point of note is that the line size designation may not have any effect on the weight factor for the moments calculation(s) of the "bamboo itself" ~ but my question remains, how do others, "designing their own rods", and using Mr. Garrison's theory, determine that weight factor? 

BTW ~ if it's of interest to anyone, the computations were for a 6' rod, using a 2 wgt. line. Here are the are the  calculated dimensions arrived at: @1".036"; 5".065"; 10".083"; 15".096"; 20".109"; 25".119"; 30".130"; 35";.141"; 40".151"; 45".162"; 50".172"; 55".183"; 60".193" (at 35" a #9 ferrule, but strangely a #10 was used in the calculations). also, Mr. Garrison would draw a more convenient curve for the tip top. Incidentally, the rod was not made, so if anyone is so inclined, I'd surely appreciate learning about it.  (Vince Brannick)

Don't forget, the weight of the cane will change with the moisture content.  (Mark Wendt)

Bob Milward's book has the value 80 lbs/cubic foot (= 0.741 oz/ cubic inch) which is about 10% higher than Garrison/Carmichael' 0.668.  I can't see where Milward discusses this value. Vince, what did you find?

The online Hexrod lets you change the default density from 0.668 if you like.  (Frank Stetzer, Hexrod, Taper Archive, Rodmakers Archive)

I am trying to wrap my pea brain around stress curves.  Is there an Idiot's Guide to stress curves?  I seem to have a good feel for rods but, I don't know what I am looking at when it comes to stress curves.

Some questions I have are:

What do the numbers mean? And, how do they relate to rod action?  Do higher numbers relate to the blank being "stiffer."

I would like to look at a stress curve and be able to figure out the action.  Right now my looking at a stress curve is kin to a dog watching TV.   (Pete Emmel)

Put me on that list too.  (Tony Spezio)

About a 1 1/2 years ago Bob Norwood was kind enough to sent me a free copy of his "Fly rod Taper Library",  it contains 301 PDF pages of tapers and other useful information that would be helpful in learning to understand tapers and their stress curves.  (Don Green)

Stressful?

Your 'engineering gurus' will be able to explain the stress curves in engineering terms. Whether those are understandable to 'laymen' depends on the ability to "read reading". (Some people can read reading, but have trouble reading writing). Simply stated stress is the internal force(es) per unit of area due to external forces applied. Those forces may be applied by a pulling force, (tensile stress), or a pushing force, (compressive stress), and/or shearing stress, due to force(s) applied perpendicular to a horizontal axis. These stresses are computed by knowing the area (square inches) of the object under stress, and the amount of force (pounds/ounces,etc.) being applied. The formula is simply Stress = Force divided by Area. S=P/A, where P = force.

In the case of a flexing fly rod, however, several forces are acting simultaneously, and a different stress develops, known as Bending (or Flexure) stress. Flexure stress calculations aren't so easily computed and a Flexure Formula involving 'Bending Moments', 'Moments of Inertia', and Fiber Stress is required. Now I'll leave it to the aforementioned engineers to explain how that works. Once the stress values are determined, a curve can be plotted revealing the stress at specific points along the length of the rod. (Vince Brannick)

To continue Vince's comment,  the following is usually true (but not always).

1. The higher the local stress the more easily the rod deforms at that location along the rod. So high stress implies a softer deforming location - rod dimension is small.

2. The lower the local stress the more difficult it is to deform the rod at that location. So low stress implies a harder/stiffer deforming location - rod dimension is large.

3. Looking at a stress curve, one may see a higher stress in the tip with a lower stress level in the butt which implies a progressive taper, while a flat stress curve implies a uniform taper, and so on.

4. There are many stress curve shapes that relate to how a rod deforms when loaded statically and dynamically. Take a look at Hexrod or RodDNA and you will be able to get a sense of how taper and stress curve relate.

Hope this adds a little to the discussion.  (Frank Paul)

Aw, c'mon Vince.  You're talking technical stuff.  We can't have that on the rodmakers list.  We have to talk about rodmaking...

Taking off from Vince's explanation, it helps to actually compare the actual taper dimensions to the same points on the stress curve.  Garrison's stress curves were generated with a "static" force applied to the rod (beam).  Static, in that the force did not vary.  That force put a bend in the rod (beam).  Since the rod is tapered from the butt section to the tip, sections of the rod are going to bend more or less, depending on their thickness.  The more bend, the more stress at a given point of the rod.  Also typically, you'll see a steep drop off in stress in the last 5 - 10" of the rod, at the tip.   Picture  a  rod in the cast - even when the rod is fully loaded, most of the bend in the rod occurs in the upper middle to the upper 1/3 of the rod.  The last 5 - 10" of the tip is more or less straight out.  Little flex, little stress.

Next injuneer...  (Mark Wendt)

"That force put a bend in the rod (beam)."

Actually, Garrisons model of a beam we call a fly rod does not bend. What appears as a very high stress point near the tip is in reality a very low stress point due to the shortening of the moment arm as the force on the tip is increased and the real fly rod bends. The force on the tip is a vector force meaning it has both magnitude and direction.. Example, mount a rod horizontally and suspend a weight from the tip. As the rod bends the tangent to the tip becomes more and more vertical. Keep adding weight and the force on the tip becomes one of pure tension.

Now that should confuse those who believe a Garrison model of stress is real.  (Jerry Drake)

Correct.  The stress curve would actually show what stresses were at any given point on the beam if the beam did not bend, theoretically.  Poor choice of words on my part, trying to explain stresses in laymen's terms.  (Mark Wendt)

You cannot really talk about Stress (the force applied) with out talking about Strain (the distance the object moved) Check this out.   (Dave Burley)

Don't you mean "since the rod is tapered from butt to the tip, sections of the rod are going to going to bend more or less, depending on their thickness. The more stress (STRAIN - DRB)  at a given point on a rod."

See this. It is a very good discussion, which I cannot send, I guess because of the diagrams.  Please check it out.

With no proof, I suspect the classic hanging chain or hanging cable which develops an equation for the tension along the chain or rope is a good place to start, even though the rod is loose at one end.  As it turns out a Bessel function represents the various oscillation functions.

Check out the graphs.

This will start to get us away from Garrison's static measurements and toward a more realistic motion( vibration/oscillation) of a bamboo rod and stresses and strains along the rod.

Time for an injuner or physics major to step in.  (Dave Burley)

I hadn't said that because I didn't really say much about the modulus of elasticity of the material.  Stress and strain are not the same thing.  Strain brings in a bunch of other things to the equation, such as plasticity, deformation, yield points, ultimate strength and rupture.  Since we are talking about a somewhat static load, which does not stress or strain the material in question to the point where some of those variables kick in, I didn't include them in the discussion.  (Mark Wendt)

To add to the confusion, the dynamic process of casting (or playing a fish) involves not just stress and strain, but also strain release.  At any point in a cast, a part of the rod can have applied stress, resultant strain, or strain release going on.  And, in terms of applied stress, don't forget the stress your hand is applying to the rod during casting.

I think most of us judge a rod (ignoring cosmetics) by its casting and fishing characteristics.  David Dziadosz gave the link to an excellent discussion by Bill Harms.  If understanding stress curves helps you understand better how a rod will behave, good!  If you use other methods, good!  The dynamics of casting and fishing make any method approximate at best.  (Tim Anderson)

Darryl Hayashida had a nice explanation of the basic stress curve shapes.  You can read it here.

If you look in the archives there are lots of discussions of this topic.  (Frank Stetzer, Hexrod, Taper Archive, Rodmakers Archive)

I think it's great that he comments on his experience with the "hinge". Too many people want to take it out these days. I would agree it's important for roll casting.  (Jim Lowe)

I don't have anything to add to the deeper thoughts on stress curves.  I wasted a lot of time on that only to find that it's all relative at the end of the day.  A stand alone stress curve is of little value without comparing it to the stress curve of a rod you've actually fished.  If that's a reasonable conclusion, then stress curve analytics add very little to a simple plot of the diameter of two tapers.  For example, a stress curve will tell you that a thinner mid  section  will  bend  more  than  a  heavier one...like you didn't already know that.  From my perspective, stress curves are of limited use.  Or...I'm just not smart enough to understand and apply the convoluted things to a fishing stick in a meaningful way.

If you want to add significant value to a taper chart, add trend lines to the tapers.  That's the genius of what Bob Norwood has done with a trend line extending from the tip top through the butt of the rod.  Deviations above and below the trend line are relatively easy to understand.  Bill Harms and Tom Whittle take that a step further with Vincent Marinaro's trend lines drawn through each section of a taper (Chapter 8 of Split & Glued).  The Norwood and Marinaro trend lines can be drawn with a pencil, ruler and graph paper.  No engineering degree required.

After wasting a lot of time on this subject when I should have been making rods, I've learned to compare tapers in four steps as follows:

1.  Compare the diameter of the tapers

2.  Compare the diameter of the tapers with trend lines drawn from the tip top through the action length

3.  Compare the diameter of the tapers in sections with trend lines drawn through each section

4a. Compare the Garrison style stress curves (so you can say you did if someone asks)

4b. Compare the more accurate bent stress curves (no one will believe you)

4c. Compare deflection curves (if you need ad copy for a marketing campaign)

If you'd like to see all that in an Excel spreadsheet, download FlexRod from my blog and play around with a few tapers.  But like I said, the first three steps can be done like a pro with a piece of graph paper.  Step four just adds mystery to an otherwise boring taper conversation.  That's probably why some of the best performing tapers were designed long before stress curves were the trademark of an expert.

Here's a link to FlexRod.  (David Bolin)

I'll try to keep this as much non-injuneer speak as possible, so as not to be labeled "too technical" for  the list.

A couple of things that a lot of people overlook while gandering at stress curves is where the curve actually bends, and how much it bends.  Those very significant factors can give you a good idea as to what the action of the rod is going to be.  If the stress curve bends quite a bit, and the curve is held high as it gets to the butt section, you would typically have some type of parabolic rod.  If the stress curve bend is situated more towards the tip, you are generally going to have a faster action, tippy rod, since the lower extremities are stiffer, and don't flex as much.  I generally use the stress curves as more of a comparison tool, comparing a known taper's stress curve either to another known taper, or to an unknown taper to get a general idea of what the rod is going to feel like in my casting.

Bumps and dips in the stress curve can also tell you a lot of things - ie, ferrule placement, swelling of the butt, the "hinge" effect and a host of other things.  (Mark Wendt)

The very simple answer is that a stress curve is a magnified mirror image of the taper. Where there is a high spot in the stress curve there is a corresponding slow spot in the taper, but a very slight change in taper shows up as a large change in the stress at that point. If you look at the taper graph of the rod, you will see areas where the taper is steeper than others, the less steep places will show up as a high spot in the stress curve graph, a steeper taper will show as a lower part of the stress graph. A straight taper will show a smooth stress graph and a taper with many changes in the slope of the taper will have a stress curve graph with many peaks and valleys in it. I'm a blasphemer and don't care for stress curves at all, I think they overcomplicate something that is fairly simple so I just look at the graph of the taper itself.  Steep rises in taper are stiffer areas, flatter sections bend more. Any steep rises will tend to keep the flex out in front of that spot until you get enough line out to overcome that area's resistance to bending, then the rod will flex down to the next steeper rise in the taper. A straight taper wil evenly flex more as more line is worked out, but can lack the "character" of a rod with a compound taper that has subtle differences in the slope. Hope this helps a little.  (John Channer)

I was looking for some writings by Wayne on Garrison's math and Hexrod. Maybe it was what Wayne had said, once upon a time, that it was for rod comparison. I found a writing by Bill Harms on Todd's Tip Site, that I thought is very good for all! Thanks Bill! I didn't see when that was written.

I think Hexrod makes a great starting point in designing your own tapers. I found what I like through Hexrod. Thanks to Frank and Wayne. In the last few years others have came up with their own programs for rod design, and that's fine. It keeps things interesting! However, it still comes down to how does the rod feel when casting/fishing! There's a lot of different factors that can change the way a rod feels when casting/fishing! Line, guide placement, adhesive, the list goes on and on! So, when I read negative comments about Garrison's math, I want to scream! If it hadn't been for Hoagy and Garry, well you know where I'm going! Maybe I'm too dumb to understand the dynamic forces on a rod in a graph, but Hexrod has been a good tool for comparing/designing tapers. Do we really need a pole vault pole to jump over mouse turds?? In Bill's writing he says the math is not essential for rod making, there's a lot of good tapers available to the rodmaker. When I started in 1998 I read and reread everything I could find about Garrison's math till one day it clicked! I could begin to understand what I was looking at in a stress graph!  (David Dziadosz)

Big thanks to everybody for their input.  I have a basic understanding and a good base to continue my venture into stress curves.  I am not going to start designing rod tapers.  I feel there are more than  enough existing tapers to keep me busy.  At least now armed with a little knowledge I can start looking at tapers as a starting point for what a rod's action should be.

I am partial the Wayne Cattanach tapers.  I like the feel of his tapers.  Before all the help you have given me, I would look at a taper and try to match one of Wayne's tapers since I know what his rods cast like.  I will probably keep leaning toward his tapers.  But, now if somebody wants a parabolic, slow action, or hinge for a decent roll cast I have a starting point for research.

Its still easier and more fun to ask the list for taper opinions.  (Pete Emmel)

And, Of Course Everyone knows(?) that this is a CANTILEVER beam we have here?  (Vince Brannick)

And the Garrison analysis is based on small beam linear deformation/stress theory for purely elastic materials.  :>)))  (Frank Paul)

I was trying to find some of the text I had  . .on stress curves but apparently it is somewhere . .anyway . .I was offer this as an insight . .Stress curves are like any other tool . .some believe and use them  . .others don't . .I think for the most part it is a reasonably accurate method to predict the outcome of a rod taper . .if the input is reasonable . . My grin for today is this . .Hexrod was originally written on a CoCo . . A Tandy Color computer with 16k of memory . .and the code  . .basic  . .was saves and ran off a cassette recorder . .a high speed connection then was 180 baud . . so go figure . .years later there have not been too many offered  options that are as widely used . . I think it was the CoCo that made the difference.  (Wayne Cattanach)

You are sure right - when I got your original loose leaf book it had the code written in BASIC and ran on DOS  - the software was on one of those 5&1/4 inch flex discs in the back of the book. That was my first effort many years ago at comparing tapers for stress and static deformation. It seems we have come a long way thanks to your early effort to make the code available to any rodmaker. Thanks for your efforts.  (Frank Paul)

OK, some of you experts check my reasoning.  I am looking at the Wayne Cattanach Sir "Darryl" 7042.  Looks like a fast action rod since the 10" mark peaks at over 200k then drops.  Also looks like there is a hinge for assistance in roll casting.  Pulled the curve off RodDNA if you want to check my reasoning.

Now would a taper that goes to 140k and levels off for the entire length be a parabolic with a fairly slow action?  (Pete Emmel)

You're on the right track, but have a little farther to go.  The WC Sir D is a fast action rod.  A rod which goes to 140 then levels off would be more of a progressive actioned rod.  Have a look at the two distinct high points on the Para 15 for an idea of what a "parabolic" action does stress curve-wise.  (Harry Boyd)

Did Garrison use 50 feet of line for determining his stress curves?  (Greg Reeves)

In the book his example for developing a stress curve for a 212 taper he used 50'. He also said you could develop a stress curve for a rod of any given line/casting length for the intended fishing conditions. So pick the line length to fit your conditions.  (Don Schneider)

In RodDNA you can vary the 'LineCast' for any rod you have entered and it will show you the resulting stress both graphically and in the Values section.  I don't know about Hexrod but I assume it will as well.  (Ralph Tuttle)

Yes you can do all all that in both programs.  I guess what I was ultimately thinking was that I usually use 200k to 225k for my high stress value in the tip region to determine the line weight a desired taper will cast best and wasn't really thinking about the length of cast.  I just thought that he used a standard length when doing his calculations.  (Greg Reeves)

Yes Garrison used 50 ft of 6 wt line to set his Stress values of 200.000 +- 20,000. He also use a weight of 490 grains for 90 feet of 6 wt line. He drew the first stress curve and forced the taper to follow it. He did a lot of things because he knew what he was doing and how Stress relates to Tapers.

If you use less than 50 ft of line, the stress caused by the line gets less until at about 30 ft of line, the stress of the line and the stress caused by the other 4 items are about equal. You can use the 30 ft length but you have to set different value of stress and tolerance, but I don't think it will be as accurate as the 50 ft value.

Actually it was Hoagy Carmichael who wrote the chapter on the taper calculations. To me he wrote it as a student would do a problem given to him by his teacher, except he never defined the given values for the problem in one specific place, but they are all there if you look.  (Bob Norwood)

Garrison's Stress.. There are a lot of new rodmakers on the list, would anyone like to have a discussion about Stress curves?? (Use, misuse, and abuse).

If it's too embarrassing to respond to me online, then send me a PM. (Jerry Foster)

Relax, this is not going to be even an essay!

Even if it could well be a book.

Which would be unecessary, as anyone can design a perfectly good rod mainly without any reference to stress curves other than being aware of their existence and the shape of the curve when the stresses are equal throughout the stressed beam.

Then it all really gets complex.

A rod is not a cantilever beam like, say the ones that keep the dress circle up in a theatre. Those are designed not to bend, a rod is, albeit within limits (limits frequently exceeded, we call it abuse).

Worse, the bit where its clamped, ie: the thick end, moves.

Worse again, the moving forces are controlled by a wayward, whimsical, opinionated animal, famous for changing its mind at the drop if a hat and without warning.
You still wanna design rods?

Lets briefly look at Garrison, who was a scientific or design engineer, which is not the same as being a mechanic, one uses a slide rule, the other a hand tool (now a computer, probably in both cases very often!).

He would immediately have constructed a straight line stress curve then looked to see what shape the taper should be to provide it, so far correct and the shape of the taper per was, is, and always will be correct, it’s physics.

That last point was and is the most important thing I'll ever say about rod design, although its not a rule, it’s a perfectly good theory, but there may be reasons to ignore it. You can only do that however, without creating a pig’s ear, if you understand it.
If youre gonna break the rules you gotta understand 'em good.

At this point we must inspect history. Garrison was building rods which would hold great lengths of frequently soggy double tapered silk line in the air. Nowadays we do nearly the opposite of that.

Therefore, Garrisons assumptions will not all be correct for modern tastes.

If you want to punch a fast sinking head hard into a headwind, to take an extreme example, using the very extreme double haul that one has to in these circunstances, you would be deeply eccentric if you chose a Garrison taper to do it with. This is not a critiscism, he was'nt designing for that condition.

I suspect taht Garrison tried to keep the stress line on the graph as straight as possible, but early realised that if it was parallel with the y axis the rod was going to feel like a bit of a noodle, so he had to tilt it, a bit.

These days we have to tilt it a bit more.

The snag is, the stress strain theory as applied to cantilever beams only works for very limited degrees of deflection, and nobody seems to be sure how much deflection Garrison was using in his calculations, he may have been using hardly any, in fact.

This is unhelpful if we are considering a rod bent over as far as one dare. If you look at a rod bent in such a way (use an old glass one!) you will see that most of the rod, possibly as much as the top two thirds is under absolutely no stress at all, in fact its straight down the line, all the stress is on the butt, in a smallish area some way above the handle, usually.

I repeat, in the fully bent position, at the point of breakage/permanent deformation most of the rod is under no stress at all. Provided you've not built in stress raisers anywhere it then breaks or sets at that point, not at the tip, where most people think the maximum stress is all the time.

If that were the case, by the way, tips would be perpetually breaking.

You may by now be ahead of me in deducing that stress curves are not immensely useful. Worse, they are over-intellectualising the issue to no very great purpose.
This sort of thig is fashionable, the MBA degree intellectualises many issues which you can either do or you can't, there are many scientists about who have'nt a clue, we used to call them fools with bits of paper. And, no, it was neither sour grapes, or elitism.
I digress, its one of my many faults.

Whilst you may think I have rubbished stress curves, if you are trying to learn about rod design you do need to understand what they are doing and saying.

Some few of you may have read recent posts on hinges, for instance. Well, the bigger the hinge effect the higher the stress figure at that point, and if you fiddle about with the graphs you can see that it does'nt take much of a hinge to create a very worrying stress raiser at that point in the blank.

You can't use the garrison graphs, or the hexrod ones, for this either, because, as I said, they don't show the stress curves for a rod bent to within an inch of its life. You have to create more stress by applying more strain, and, as far as I can recall since its some time now since looked at it, I think you can do this by increasing the "tip impact " figure.

Or, just take my word for it that any change in taper of any suddenness or deviating radically from the basic curve is going to cause problems, not maybe, not whether, just when!

I'm going to stop at this point, as promised! (Robin Haywood)

Interesting essay.

Just one point – a rod has to deal with casting as well as catching (which is of course the ultimate goal).

The angle at which the force is applied is critical to the location of the stress. You refer to the stress in the middle and top of the rod being zero when the rod is fully bent. If this situation is induced by the angler fighting a large fish on a long rod close in to a boat, for example, the bending in the rod, and thus the stress, is going to be most likely in the tip. In this situation (and similar ones resulting from injudicious removal of flies stuck in trees, rocks etc.) a hinge in the tip is not going to be your best friend. (Stephen Dugmore)

Good points Robin. It's good to understand the shortcomings of stress evaluation also. More? I would like to hear about your evaluation of tapered bamboo spring theory, (unloading the rod) which might be more relevant. However, since we don't have one in place a levered stressed beam will have to suffice. Stress is a tool; each can choose to use it or not. It is not the end-all either in rod design/evaluation nor in the search for absolute truth. But it does have its place. From what you have stated so far it is difficult to discern if you are dismissive because of factual faults or you have never tried to evaluate how they could be of use.

Good stuff. (Jerry Foster)

If by loaded bamboo spring theory, you are talking about rates of recovery from flexing then really, its to do with the constant for the material and the only way you can really change it is by hollow building. Cane is not homogenous, so, if you remove the less dense and therefore less stiff, middle you change the figures for the rest of it.

There is no absolute truth. Trouble is there are many absolute truths, I wonder how many we know about?

Nobody seems to know what presumptions Garrison made, let alone why and how he made them, we all know what they must have entailed considering, but not what they were.

This is all getting a horribly like quantum mechanics, is’n’t it.

And, of course, rod design theory pretty soon takes you out of straight line science into philosophy. But you gotta know the science first, lest you make a pillock of yerself.

I have tried to evaluate the use of stress curves for something more than a refernce point/point of interest/basic principle as outlined, but most of my findings I quoted, the main one being that I'm not sure that any more detail should really be necessary to a rod designer.

Rod design, at some point down the line means having opinions, views, preferences and prejudices, the more experience you have the less loony those things tend to be.

This, however, is a world away from ill or mis-informed rank amateurs and novices, with no understanding of basic principles indulging in wild guesses.

As i keep saying, you would'nt buy their product in any other walk of life, so why buy their rods?

I don't think I covered your first point well, possibly because i may have failed to fully comprehend the question.

I apologise, would you like to elaborate the question a bit? (Robin Haywood)

What is your evaluation of tapered bamboo spring theory? (Stephen Dugmore)

I haven't read the publication, but I really wasn't interested in a discussion but an algorithm. The math set to music so we can use it to develop tips that finish the cast and don't snap back. (Jerry Foster)

That's easy, the answer is very light tips.

But they break easily, although not whilst actually fishing.

its a balance, like life. (Robin Haywood)

My apologies for starting another conversation about stress curves...but im still not able to get my head around the "practical" everyday use of graphed stress curves of fly rods. I'm wondering how the graphs themselves speak about the action of the rod. 

Question: Can I look at the graphic representation of a rod's stress points and by looking at the graph know, before I make the rod, that the road will be generally a faster (fast) or slower (slow) action? I know that slower and faster are objective terms. That is part of the question though. Do graphic representations of a rod's stress at specific points bring any subjectivity into the discussion? Is that the (a) reason for their existence, they're purpose, or am I missing the whole point? 

For example: I want to make a rod with a fast action. Does high stress peaking at the 10 in mark dropping exponentially at the 20 in mark until inch 45, at which point stress raises slightly and then levels off at 70 in mark, indicate a fast taper?  (Cattanach 7 ½ 2 pc 4 wt)

Or....

For example: I want to make a rod with a slower (slow) action. Does high stress at the 10 in mark that is maintained---more or less up and down--to the 48 in mark and then drops slightly from there indicate a slower taper? (Payne 7ft 1 in) 

Or....?

Thanks for any insight. (Jim Sabella)

You have a very valid question. And yes, when used correctly, stress curves and rod tapers can supply much info. However, each stress curve should come with a set of SPECIFICATIONS and without these specs you can’t understand what the rod is or what the stress curve is trying to tell you.

Garrison in chapter 14 of his book gave a set, but they were very hard to find and most folks that read the chapter missed them and were not even aware they were required. Which means that most rod taper and stress curves do not have them, and therefore it makes reading the graphs next to imposible.

Do you have the Specifications for the rod or stress graph you are looking at and would like to understand? If you have them, please send then so that we can all have a look at trying to figure out the rod.

PS:   It’s OK if you don't have any, just pick out a rod you would like to learn about and we will start from there. (Bob Norwood)

Because of my engineering background I’ve been very interested in the work that you and others have done to calculate the deflection and resulting stresses in bamboo fly rods.  And I think I’ve finally come to a point where I might have something to add to the effort.
From the beginning I’ve been concerned about a couple of fundamental issues, primarily related to the equations that have been selected and the underlying assumptions that they carry with them.  I’m not too terribly concerned about things like the fact that E is assumed to be constant (Bob Milward has demonstrated that it’s not), or that bamboo is assumed to be isotropic (it’s undoubtedly orthotropic).  But I AM concerned about the fact that the Strength of Materials equations that are most often used were all derived under the assumption that the deflections are small (at least it’s assumed that they’re not large enough to effect the geometry of the problem).  Since Garrison’s calculator was limited to a slide rule he didn’t have much alternative but to accept the limitations of these simplifying assumptions, but our situation is thankfully quite a bit different today.

When I was still working for a living I had access to some truly wonderful structural guys whenever I needed them.  These guys were very experienced and always had the best, most up to date analytic tools that were available.  For anything but the most elementary problems we routinely reached for the FEA (Finite Element Analysis) packages, built up a 3-dimensional digital model of the problem, and let the computer whur.  These tools allowed us to analyze complex multi-layer orthotropic composite structures, structural ‘sandwiches’ that were built with a thick sheet of syntactic foam that was sheathed with layers of various composites, and all sorts of other problems that could never have been approached with classic or conventional techniques.  It became routine for us to execute the initial design with a set of 3-D tools like SolidWorks and then export the design to ProE where we could perform detailed dynamic stress analysis, thermal analysis, CFD, and lots of other wonderful stuff.  I’ve often thought how great it would be to have access to those resources today.  Unfortunately, those tools were terribly expensive and the learning curves were steep and high.

But now I’m thinking I might have found a reasonable alternative.  I’ve been playing with a 3-dimensional FEA software package that I recently got for free (that’s right, it’s Freeware) and I’m thinking it might have some applicability.  The algorithms assume that; 1) E is constant and stress is linearly proportional to strain, and 2) the material is isotropic (although a layered structure like a bi-metallic strip, can be analyzed).  But large deflections are allowed.  The package includes a Non-Linear Solver that increases the load in incremental steps and iterates an intermediate solution on each pass to accommodate the changes in geometry caused by the ‘large’ deflections.  The final results of the deflection calculations are then used to calculate the stress tensors at specified nodes and/or elements. 

I don’t have any 3-D design software that can create the input to the FEA software so I’ve had to create the FEA model manually. 

Thankfully, the relatively simple geometry of a fly rod makes this task fairly straightforward – I’ve developed an Excel spreadsheet that calculates all the node locations and helps to create the element specifications and assign all the node coincidences (as a test I created a model for an 8ft rod that had 703 nodes and 324 elements, a bit cumbersome by manual standards but quite small by normal FEA standards).

So far, I’ve loaded my fly rod model into the FEA tool, specified all the boundary conditions, and debugged the whole mess to verify that everything works as intended.  My next step will be to evaluate the results.  One of my first planned tests will be to analyze a blank that I currently have on my bench and then compare measured deflection to calculated deflection.  I recently made some deflection measurements on several of my rods and I was surprised to find how similar all the deflection curves turned out to be, so I don’t think deflection measurements alone will be adequate to validate the FEA calculations, but it should be a good first step.

If you, or anyone else on the List would like to collaborate on these efforts just let me know – I’ll be happy to describe my tests, share my results, and/or answer any questions you may have about what I’m doing.  This is actually turning out to be an interesting project and I’m hopeful that it might produce some useful results. (Bryan Coles)

I would heartily suggest you contact Bob Norwood for a copy of his Rod Analysis program. Very reasonable and will greatly de-mystify Stress Curves and the design of bamboo fly rods.

If you purchase his program, be sure to print out ALL of his white papers and read the exhaustive work Bob has done _ worked for me.  (Bill Scheider)