PLEASE NOTE: This is the Archived Sexyloops Board from years 2004-2013.
Our active community is here: https://www.sexyloops.co.uk/theboard/

Estimating the Effective Mass of a Fly Rod - What is the rod mass divisor factor?

gordonjudd
IB3 Member Level 1
Posts: 2214
Joined: Mon Jul 10, 2006 12:14 am
Location: California
Contact:

Estimating the Effective Mass of a Fly Rod - What is the rod mass divisor factor?

Post by gordonjudd »

mo is a strange characteristic in fact.

This thread is intended to shed some light on what the effective mass (m0) is all about with the goal of making it a bit less "strange."

In coming up with the effective spring constant (K0) and effective mass (m0) values that are used in the frequency vs added tip mass equation you find that K0 is not equal to the measured spring constant value of the rod (for small deflections) and that m0 is much smaller than the mass of the rod. It is not surprising that mapping the two dimensional vibration characteristics of a long tapered fly rod would be different than the values found for the “equivalent” one dimensional model, but I have always thought there should be some way to convert the true mass of a blank to the equivalent m0 value used in Grunde’s and Merlin’s model of the cast.

What is the physical significant of the effective mass term? The best description is given in this MIT lecture on the subject by Dr. Lewin at M.I.T. He explains how the added m0 term can be used in a mass/spring oscillator model to account for the kinetic energy that is associated with the mass in a real spring at about 32:00 in that video.

Here is the excerpt from French’s book that he mentions in his lecture.
Image
Image
Image

The relative displacement in a cantilevered beam is not a linear function of the length as is in a spring so the energy analysis for the mass in an oscillating fly rod is much more complicated than the procedure described above for a spring.

Merlin found a method that can be used to estimate the divisor for a stiff cantilevered beam that can be used as a starting point for a fly rod. We both had some dimensioning errors where this was discussed in another thread, so I hope that these equations will be more accurate.

To simplify this analysis I will express the frequency in terms of its radians/per second (omega) counterpart. The equivalent frequency can be determined by f=omega/(2.*pi).
The (omega frequency).^2 of the first mode for a vibrating beam is equal to
Omega.^2=(1.875^4)*EI/(L.^4*rho*Ao) (1)
where:
E is the Young’s modulus in N/m.^2
I is the area moment at the butt in m.^4
L is the length of the cantilever in m
Rho is the mass density in kg/m.^2
Ao is the area at the clamped end of the rod.

The total mass of a uniform beam is just rho*A*L so as Merlin pointed out equation (1) can be re-written as:

Omega^2=1.875.^4*EI/(L.^3*m_rod). (2)

The (omega)^2 of a spring/mass SHO is equal to:

Omega^2=k/m0 (3)
Where:
k is the effective spring constant in N/m.
m0 is the effective mass in kg

Given that the deflection at the end of a uniform stiff beam for a tip load of N Newtons is:

deflection=N*L.^3/(3*E*I) meter gives a corresponding k factor of
k=3*E*I/L.^3 N/m. (4)
Substituting that value into eqn (3) and setting the result to the same omega.^2 value in eqn(2) we find:

3*E*I/(L^3*m0)=1.875.^4*E*I/(L^3*m_rod) (5)

Cancelling terms you find:
m_rod/m0=1.875^4/3=4.11
Thus as Merlin noted in the other thread the divisor used to convert the mass in a stiff, uniform, cantilevered beam to the m0 value is around 4.

For one of my saltwater blanks I have found that m_rod/m0 ratio was equal to 11.8. Thus the equations for the uniform beam are going to have to be modified to get a value that makes sense for a tapered rod.

My thanks to Merlin for correcting my other derivation that ended up with an incorrect additional L term regarding the ratio of m_rod/m0. I think we now agree on how this is derived and can use this as a starting point to get a reasonable value for a flexible fly rod.

Gordy
"Flyfishing: 200 years of tradition unencumbered by progress." Ralph Cutter
User avatar
Merlin
IB3 Member Level 1
Posts: 798
Joined: Sat Mar 27, 2010 3:30 pm
Contact:

Post by Merlin »

Gordy

We just have two means to estimate mo for fly rods: frequency measurements with various loads or calculation (which mimics the same testing procedure).

Since mo represents the "mass in motion", it is related to the deflection profile of a rod, not to its speed (I checked that with the models).

I thought we could find some dimensionless scale by looking after the "factor 4 or something like that", but after trying with cane rods, it does not seem to work. Bad luck, otherwise we would have found an "universal scale".

Merlin
Fly rods are like women, they wont´play if they're maltreated.
Charles Ritz, A Flyfisher's Life
gordonjudd
IB3 Member Level 1
Posts: 2214
Joined: Mon Jul 10, 2006 12:14 am
Location: California
Contact:

Post by gordonjudd »

Since mo represents the "mass in motion", it is related to the deflection profile of a rod, not to its speed (I checked that with the models).

Merlin,

I know that you and I are probably the only people on the planet who are interested in understanding what the m0 value is all about, but I think you will find the following simulations involving the mode shape and mass distribution of a fly rod and how they impact the divisor factor related to get mo from the measured mass of blank to be illuminating.

I am always leery when a predicted value fits so well with a measured value, but I hope this simple analysis shows that French's method for computing the effective mass in a spring based on the relative kinetic energy in the moving mass of a spring can be applied to getting a similar value for the distributed mass in a fly rod.

Your insight into this problem is well taken, and is based on the fact that the relative 1/2mv^2 K.E. of a point in a distributed mass will be equal to the square of its relative deflection distance compared to the square of the max deflection at the end of the rod.

Using that fact to give a "relative" velocity factor, we can then use the linear mass density in the tapered rod to come up with a good estimate of how the integrated m(s)*d(s).^2*ds energy in the moving mass of the rod compares to the K.E. expected if the total mass of the rod was moving over the +/- max deflection distance in the same amount of time given by the oscillating frequency of the clamped rod.

I find it is always wise to debug a program by using it to first look at a simple case with a known result before jumping off to look at the more complicated variations of its input parameters.

Essentially this problem revolves around integrating the mass_l(s)*deflection(s)^2*ds kernel function to come up with the sum of the relative kinetic energy in the moving mass of a real world spring. For a linear spring that function is straightforward since the deflection(s) function is just a linear function of s and the linear mass density (mass_l(s)) has constant value.

When you run that simple case you get this result:
Image
You can see it predicts the same factor of 3 divisor that French got with his closed formed expressions, which indicates the numerical solution will give an accurate value for the expected divisor factor.


To apply the technique to a uniform beam requires knowing the mode shape for the first oscillation mode. Fortunately Haun provided that plot for a uniform beam in his thesis and showed it has this deflection vs normalized length (s=x/l) parameter shown in the blue curve below:
Image
Running that shape and a uniform mass density function through the program you get this result for the divisor factor.
Image
The resulting divisor factor of 3.97 is reassuringly close the factor you obtained with Dunkerley's approach described in the first post.

Now we can use the same approach for the more complicated case where both the deflection profile and the linear mass density are non-linear functions. Thanks again to Haun we find this deflection profile for a linearly tapered rod looks like the blue curve below.
Image
Grunde’s MOI method will provide this linear mass density for a linearly tapered rod. For this run I used the green curve that is marked as "density for tapered ring'' curve below. That has smaller second order term, and is more representative of the taper profile in a full flex rod. The red/blue curve for the Dan Craft rod is more like the taper you would see in a tip flex rod.
Image
Using the above deflection and mass density profiles for a tapered rod we get this integrated relative KE result.
Image
The resulting divisor factor of 11.58 is scary close to the 11.8 value I measured for a full flex saltwater rod. Considering it probably had a slightly different mode profile and linear mass density curve than the one used in the simulation the .22 difference is likely a lucky choice, but it shows this technique predicts the higher divisor factors we have measured for fly rods.

I think adapting French’s approach to predict the expected divisor for a tapered fly rod is good enough to show that m0 is related to the K.E. associated with the moving mass in the rod.

Gordy
"Flyfishing: 200 years of tradition unencumbered by progress." Ralph Cutter
Eugene Moore
IB3 Member Level 1
Posts: 93
Joined: Wed Jul 06, 2011 1:27 am
Location: Missouri
Contact:

Post by Eugene Moore »

Gordon,
I haven't found a single equation to describe the taper and mass distribution along a tapered hollow shaft.
My approach has been to use the diameter values at discreet points to construct flags describing the area of the blank.
By constructing the flags I can calculate the area and based on material density establish an average thickness. This can then be applied to the flags allowing a mass calculation for any position along the flag. This can then be viewed from the rotation axis position and a mass distribution be obtained. The accuracy and depth of resoultion depends on how many dia's you wish to measure along the blank. It will not however locate changes in thickness applied at discreet blank locations.
It will also allow the construction of blank spring deflection rates at discreet locations along the taper. In essence constructing the deflected blank shape and the polar moment of inertia for dynamic analysis.
In theory.
At least that's where I'm headed until someone can point the errors of my ways.
Eugene Moore
gordonjudd
IB3 Member Level 1
Posts: 2214
Joined: Mon Jul 10, 2006 12:14 am
Location: California
Contact:

Post by gordonjudd »

I haven't found a single equation to describe the taper and mass distribution along a tapered hollow shaft.

Eugene,
Take a look at Grunde's and Magnus' method for measuring the MOI of a rod. In the appendix Grunde explains how he derived the linear mass density of a rod from his mass and C.G. measurements.
Image
Image

That is how the blue piecewise straight line approximations to the red expected quadratic mass density curve for a linearly tapered rod were derived from Grunde's data for the Dan Craft rod as shown below:

Image

The green curves in that plot show the actual and quadratic fit to the section data that Dr. Spolek measured for a Loomis blank. You can see that a quadratic fit is very close to the actual density profile except for the spike you get because of the increased mass in the the ferrule

Dr. Spolek found that all of the rods he sectioned had nominally linear tapers although the taper on the inside of the blank might be a bit different than the outside and thus the wall thickness can change slightly as you go to towards the tip.

Here is a plot of his section data along with the linear fits for a Loomis rod that typifies that general design characteristic.
Image

Note that the linear taper in the radius of a constant thickness tube would have a cross-section area variation that would vary proportional to the radius squared. Thus the expected linear mass density variation in a linearly tapered rod will have a quadratic variation as shown the the plots above.

The changing slope that comes out of Grunde's linear variation for each section of the rod will give a good approximation to that quadratic curve in a four piece rod. It will not be so good for a two piece rod however, so you may need to use some diameter data to tweak the linear estimates.

Gordy
"Flyfishing: 200 years of tradition unencumbered by progress." Ralph Cutter
gordonjudd
IB3 Member Level 1
Posts: 2214
Joined: Mon Jul 10, 2006 12:14 am
Location: California
Contact:

Post by gordonjudd »

TCR factor is around 20, but I find the same (19) for a Garrison, 8 feet cane rod on the "slow butt side". My rods are in the range 13 to 15
.
Merlin,
As you have found the divisor to get the m0 effective mass from the total mass of a blank is very sensitive to the rod taper, and how that impacts the linear mass distribution and the bending profile of the rod.

Calculating the bending profile for a given taper design is a very complicated process that is still on my "to do" list in learning how to solve the 4th order differential equation involved. The only mode shapes I have seen for a tapered rod were given in Haun's thesis, so for now I have to live with one size fits all in regards to the bending profile I am using.

That is probably not a big problem as the deflection squared profiles for a uniform beam and a tapered rod are quite similar as shown below:
Image
I would expect the difference in the bending profiles for two different tapered rods would be even smaller.

Because it has such a non-linear response, you could expect the differences in two different linear mass density profiles to have a big effect on the resulting m0 divisor case. If you use the density profile for the red curve in the previous post instead of the green curve that was used in post #3 you get this result for the divisor calculation.
Image

Thus reducing the mass density near the tip of the rod has a big effect on the divisor factor. That rather modest change in the linear density curve caused the divisor factor to increase from the 11.6 factor for a full flex rod to the 17.0 factor shown above for a tip flex rod.

Send me the mass density profile you get for one of the Garrison bamboo rod profiles, and I will see what the divisor for the non-linear taper he used in his rod becomes.

Gordy
"Flyfishing: 200 years of tradition unencumbered by progress." Ralph Cutter
User avatar
Merlin
IB3 Member Level 1
Posts: 798
Joined: Sat Mar 27, 2010 3:30 pm
Contact:

Post by Merlin »

Gordy

I shall be able sending you the data for a Garrison rod during this week end I think.

I look at mo slightly differently today. It is for sure easier to consider blanks only first. Taking hardware under consideration is more difficult, but it can be done in a second time.

Since you clearly put into evidence the role of the moving mass and linear mass density, I checked my documents and found that I already had that relationship before, but failed to identify it clearly. Now I am sure we can write the following:

mo = Flex Factor independent of material * average linear mass density of the rod blank

In other words: mo = FF * rod (blank) mass / rod length.

The beauty is that FF can qualify the flex profile of any rod (cane, glass, graphite). If you consider the flex pattern as the definition of action, then FF is THE action parameter. FF is purely related to the flex profile of a rod. It does not depend on other parameters, its dimension is meters. Practical application gives a scale which could lie between 5 and 10 centimeters, I have to document that further on.

This scale is applicable to a blank. A rod maker can get mo directly from measuring the loaded frequencies for its blank. Then FF is easy to calculate.

If you want to estimate FF from a finished rod, a possible way is to conduct the frequency measurements to derive mo for the finished rod, then to deduct a contribution from hardware, to finally get an estimate of mo for the blank. Then FF is again easy to calculate, but one needs guidelines about hardware contribution to mo of the finished rod.

Merlin
Fly rods are like women, they wont´play if they're maltreated.
Charles Ritz, A Flyfisher's Life
Bill Hanneman
IB3 Member Level 1
Posts: 710
Joined: Thu Oct 28, 2004 12:54 am
Contact:

Post by Bill Hanneman »

The beauty is that FF can qualify the flex profile of any rod (cane, glass, graphite). If you consider the flex pattern as the definition of action, then FF is THE action parameter. FF is purely related to the flex profile of a rod. It does not depend on other parameters, its dimension is meters. Practical application gives a scale which could lie between 5 and 10 centimeters, I have to document that further on.

The beauty is that AA can qualify the flex profile of any rod (cane, glass, graphite). If you consider the AA as the definition of action, then AA is THE action parameter. AA is purely related to the flex profile of a rod. It does not depend on other parameters, its dimension is degrees. Practical application gives a scale which could lie between zero and 90 degrees. One does not have not documented that because it is obvious.

Why don't you simply develop the conversion factor?
Bill
Eugene Moore
IB3 Member Level 1
Posts: 93
Joined: Wed Jul 06, 2011 1:27 am
Location: Missouri
Contact:

Post by Eugene Moore »

Bill,
That's BS
I can define an infiinite number of blanks with the same AA.
Each will have it's own set of defined tapers and cumulative effects.
The problem with single linear equations based on a small data sampling is the location of the tapers. Especially those nearest the rod tip which contribute greatest to the net result.
I can comfortably design a blank with a stiff or soft butt and design the taper at the tip to be any value you wish to measure to mis-lead whoever you wish.
Eugene Moore
User avatar
Merlin
IB3 Member Level 1
Posts: 798
Joined: Sat Mar 27, 2010 3:30 pm
Contact:

Post by Merlin »

Bill

AA may be a suitable guideline for people building rods from blanks. For design, mo and the like are much more valuable. They allow designing rods for a series. I am conviced of that capability since commercial rod series has been extended thanks to it.

Merlin
Fly rods are like women, they wont´play if they're maltreated.
Charles Ritz, A Flyfisher's Life
Bill Hanneman
IB3 Member Level 1
Posts: 710
Joined: Thu Oct 28, 2004 12:54 am
Contact:

Post by Bill Hanneman »

Merlin,
AA may be a suitable guideline for people building rods from blanks. For design, mo and the like are much more valuable.

Thank you, and it appears it is. Of course mo and the like could be more valuable for rod designers. However, from the tone of our previous discussions, I don't get a very strong impression that anyone has been seriously considering it lately.

If you will go back to our previous discussion you will note Grunde’s work on the harmonic oscillator had been essentially forgotten and relegated to the dust bin until I brought it up in an attempt to test its application for the benefit of “people building rods from blanks".

Rod design was never a consideration as blanks are essentially the “given”. Thanks to your helpful responses, I believe I have learned the limitations of my approach.
---------------------------------
Eugene Moore
That's BS
I can comfortably design a blank with a stiff or soft butt and design the taper at the tip to be any value you wish to measure to mis-lead whoever you wish.

You may indeed be the greatest rod designer on the face of the earth, but I fear you deserve a failing grade in tact.
I can define an infinite number of blanks with the same AA.
Each will have it's own set of defined tapers and cumulative effects.

Good for you. Is that an accomplishment? There are already thousands of blanks on the market which have the same AA, and, I am certain each has its own set of defined tapers and cumulative effects. Big Wow!

And, pray tell me, what are you, the great rod designer, doing to help the angler choose a rod he might like? I would opine, "Nothing!". Instead you leave it to advertising department to describe their actions as Fast, Moderate, or Slow—without ever defining how fast is fast, or how a 5wt rod differs from a 6wt. :D

So, be happy, go design your rods. When put on the market, the CCS will be around to tell the angler just what it is you are offering. And, it will be doing in such a manner he can easily compare its performance with any rod of the competition. Or, is that what ad departments are afraid of? :D

Bill,
User avatar
Merlin
IB3 Member Level 1
Posts: 798
Joined: Sat Mar 27, 2010 3:30 pm
Contact:

Post by Merlin »

Bill

I never forget the model of the SHO, I think I may be the one who got it first (by 2003, if not, sorry for my ego, if so, sorry for Grunde's). It still helps me a lot in rod design understanding, but it is not the whole story, for sure.

Again, the fact that you restrict your perception of the adequacy between a rod and a caster to a limited domain is detrimental to the use of your scaling system. On the other side, it may make the day for rod builders from blanks.

Merlin
Fly rods are like women, they wont´play if they're maltreated.
Charles Ritz, A Flyfisher's Life
Eugene Moore
IB3 Member Level 1
Posts: 93
Joined: Wed Jul 06, 2011 1:27 am
Location: Missouri
Contact:

Post by Eugene Moore »

Bill,
Misdirecting again.
The point is that of those rods most will be poor casting and fishing instruments. Some will be fair, others will be excellent.
Based on the AA solely no one can judge the capability. The same goes for ERN and all other static measurements. PPF has some small merit but adding additional weight to the rod tip overwhelms the blank inertia and prevents rather than enhances resolution.
Until you consider how the mass is distributed along the rod length static measurements are factual but not informative.

I may choose to be tactful or deliberate. This forum, to my knowledge is voluntary. Courtesy is also something which, though appreciated is not mandatory as you've repeatedly shown.
Eugene Moore
User avatar
Bobinmich
IB3 Member Level 1
Posts: 606
Joined: Fri Sep 14, 2007 4:09 pm
Location: Rochester, MI
Contact:

Post by Bobinmich »

Gosh I get tired of saying this. You will say, "why don't you just quit saying it Bob and you won't get tired anymore!" I suppose you are right but I am going to say it one more time.

A plot of ERN vs MOI would tell you almost everything you want to know to compare different rods. Naturally not everything, but just about. Certainly more than the manufacturers subjective rating for you using the rod how he subjectively thinks you should. And it is one hell of a lot less complicated than all this other stuff.

Bob
User avatar
Merlin
IB3 Member Level 1
Posts: 798
Joined: Sat Mar 27, 2010 3:30 pm
Contact:

Post by Merlin »

Bob

I think Eugene and I talk about design, while Bill and yourself talk about blank or rod selection. So we communicate with some difficulty.

I can also keep saying that rod designers do know why they chose a particular line for a particular rod. There is nothing subjective here. If your casting style is close to their, then you agree with the line to rod fit, if you are at odds, then you disagree strongly.

Merlin
Fly rods are like women, they wont´play if they're maltreated.
Charles Ritz, A Flyfisher's Life
Locked

Who is online

Users browsing this forum: No registered users and 2 guests