Bamboo Tips - Tips Area
Glues - Stiffness

< Home < Tips Area < Glues < Stiffness


Would anyone care to offer an opinion on the differences in stiffness attributable to the various adhesives used  in rodmaking?  Resorcinol, Urac, Epon, Titebond II, Gorilla Glue, Hide, Melurac, Nyatex or any others. It sounds as though this is a possible way to fine tune tapers.  (Phil Rundhaugen)

    Well, HERE'S a can 'o worms if ever there was one!  We're gonna have some fun with this one.

    My vote goes to Resorcinol as the next best glue, overall, to a good, luthier's hide-glue.  But I wouldn't use the former 'cause I can't stand the purple railroad tracks, and I wouldn't use the second one because it's stinky, and about as user-unfriendly as can be, and under some conditions, is "reversible."

    So that leaves me with my practical compromise of Epon.  But you asked about differences in stiffness attributable to the glue line, and for that, you'd probably have to go with the Resorcinol.

    How 'bout it, fellas?  (Bill Harms)

      I agree with Bill Harms and use Epon exclusively for rods.  Resorcinol is the best glue that I ever used for rods except I can't stand the purple glue lines either. I also use Gorilla glue for the splices in nodeless construction, however I add just a tint of water to the bamboo splice before gluing since Gorilla glue seems to need some moisture to do a real good job. Did use Nyatex for quite a while and when the shelf life took over I switched to Epon which in my humble opinion is the best.  (Jack Follweiler)

      PS:  I have some rods that were made with Resorcinol in the early '80s and they are just as straight and stiff as the day I glued them.

      Perkins L-100 or URAC, kicked with Ammonia Chloride (or whatever the white sparkly stuff is) instead of the walnut flour stuff.  Resorcinol is really a fine glue, but it has too much baggage for me.  Also, the worst sets I've ever had in tips were in a rod I glued up with resorcinol.  It probably wasn't the glue's fault, but I'm so happy with the alternatives I'll probably never find out.  (Brian Creek)

      I should have asked for ranking, as in stiffest to softest.  (Phil Rundhaugen)

        Yep, that's really the issue here.  I'd still put my money on the Resorcinol to provide the "stiffest" glue-line, but of course, I have no data.  Various glues have "creep," but I swear I don't know how we could measure it in fly rod construction.  I guess that issue has been tested by manufactures themselves under their own conditions, but it would take a good engineer with a good lab to put all those glues to the test as we use them in making bamboo laminates.

        Zimny!!  Where are you when we need you?  (Bill Harms)

    I think that Nyatex makes them come out on the fast side, but I have no idea how it compares with most other glues.  (Jeff Schaeffer)

    I have done two rods with URAC, about a dozen with various polyurethanes, and quite a few with Epon.  Oh,  and one with an acid-catalysed polyvinyl acetate glue.

    I  cannot see any difference in stiffness, and if I could, I suspect that the rod-to-rod differences inherent in the kind of hand building that I do would be more significant than variations due to glue properties.

    Currently, for a number of reasons, not all of them good ones, am using Epon, and it's fine!  (Peter McKean)

    I haven't used every glue listed but I would say that the glue that comes out with the stiffest glue lines(of the glues I've used)  and thus the stiffest blank is URAC, Melurac, Resorcinol, TB II, Gorilla Glue. I believe you can use as a basis how hard the cured glue gets in the pot. Hard brittle cured glue is pretty stiff. Glue that stays flexible is not.  (Marty DeSapio)

    We are going to make 5 pentas for Roscoe this year and we are using a different glue for each one. (Epon, ProBond PU, Melurac, Resorcinol, and URAC). Now you have me thinking that we need to make a few more!

    Anyway, I have felt that if the glues I use for more than 5 rods each that URAC was stiffest, then Epon, then Gorilla Glue, and Titebond II was softest. These are subtle changes, but hopefully we can figure out once and for all. If anyone has any suggestions for the experiment, please let me know. So far the idea is this:

    • Saw strips from 4 or 5 culms.
    • Mix strips randomly among the blanks.
    • Heat treat all strips the same.
    • Use same spiral stagger for each.
    • Send off for glue ups.
    • Use same components, guide spacing, wraps, etc....

How would you suggest to go about varnishing. Maybe impregnating would be the way of ensuring equal finish on each rod. Maybe two coats of the Mike's Stuff?  (Bob Maulucci)

      I would suggest making your casting comparisons before varnishing, so as to eliminate one variable.  If the glue differences are subtle, any variations in varnish film thickness could skew your results.  The rod is still functional without varnish, and as Forest Gump would say, "One less thing to worry about."  (Tim Preusch)

        An interesting topic. The stiffest feeling rods I have built have been glued with URAC. I think Marty is right, the harder the glue dries, the stiffer the rod. The question I would have is if there is a downside to using a softer glue. Lets say you could wave a magic wand and get a rod together with no glue, just cane. The stiffness of that rod is "X".  I don't think there is any question you can get "X plus" stiffness with a URAC glued rod, and maybe a little bit of "plus" with resorcinol. I've never used Nyatex, but everything else seems to fall into a average stiffness category, which make me think that as long as a glue is good enough to hold the rod together, the stiffness of the cane must establish a baseline, and the glue becomes unimportant. I have never handled a rod I thought was on the "minus" side because of the glue. Just a theory. I will be very interested to cast the penta experiment rods.  (Tom Smithwick)

          I'm thinking that all of these variables are controlled by the combination of elements that we use to construct rods. Modifications in dimensions and heat treating will compensate for any variables in stiffness that come with the varnish, guides, ferrules, and glue. Of course, this requires the modification of existing tapers if you go that route. Remember that many of these older tapers are modified from the start by the fatter plastic lines we slap on them. I think jumping around from glue to glue may serve to modify tapers to your liking but I like the concept of the taper being the basis for the rod and trying to make the components more of a constant. The live bamboo portion of the rod is the part I struggle with the most.  (Dave Rinker)

          This is an interesting topic.  I don't think that the glues we use are typically being stressed anywhere near their limits in casting. Consequently, I doubt that there is much difference in the stiffness contributed by the glue itself.  In theory, a harder glue should make for a stiffer rod, but detecting that difference is another matter.

          I agree with Peter that there is probably more variability in the rods themselves, by virtue of being handmade, than there is in the glue.  I would add that there probably more variability in the bamboo itself than is contributed by the glue.  Bob Milward found as much as 18.9% variation in the modulus of elasticity within a single culm between the butt and tip ends.

          Another source of variation in glues is the thickness of the glue lines.  I started out using resorcinol, and noticed that it seemed to add about 0.005" to 0.006" to the thickness of a rod.  That's a line weight!.  I use ProBond now and can't detect any measurable increase in thickness when I glue up. So if you don't measure rods after glue up and scrape/sand them to the design dimensions, I suspect that the thickness of the glue lines contributes much more to the stiffness of  the rod than the physical properties of the glue itself do.  (Robert Kope)

            I think much of this increase in thickness is more a phenomenon of the  walnut-shell filler than the resin itself. Why not dispense with the walnut-shells and "kick" the resin with straight formaldehyde. I believe that's what Orvis did.  (John Zimny)

              John Palmer, a Toronto rodmaker, now long deceased, told me that he used a diluted formaldehyde to kick the resorcinol.  I tried it on a couple of rods.  The results were fine, but I couldn't stand the smell which lingered for days.  These days I use URAC 185 post cured in an oven @ 180 to 200 degrees F. and allowed to cool to room temperature.  (Ted Knott)

              It's about time you joined this discussion.

              I agree that the glue line thickness was caused by the walnut shell filler. Eliminating the walnut shell is one solution.  However,  I suspect  that the walnut-shell filler was added originally to insure an adequate glue joint thickness.  Eliminating it could produce glue-starved joints.  Even if it doesn't compromise the glue joints, you still have the smell, the mixing, the short working time, and the prominent glue joints.  My solution is PU.  (Robert Kope)

          Theoretically, a PMQ with the glue seam in the horizontal plane and guides on the apex will have zero glue effect because the bending stresses at the axis (and glue line) are zero.

          The equation is:

          Stress = M../p>

          where M = bending moment,

          y = distance from the flexural axis,

          I = area moment of inertia of the flexed beam.

          So the "glueless rod" could be approximated.   (Tim Preusch)

          I made two identical quads this year, one with URAC (Borden CR-591) and one with TB II. The TB II rod was different, slower even though the rods were both made from the same culm. It was slow compared to the past URAC rods I have made with that taper, but I suspect it was not just the bamboo. The URAC glued rod seemed to be the usual taper/feel.   (Bob Maulucci)

            My three identical PMQs:

            TB II - "softest" but best casting

            Resorcinol - "fastest"

            ProBond - "in the middle"  (Joe West)

        Differences of only .002, caused by either thicker glue or thicker varnish, has a noticeable effect on the stress curves for a rod.  I ran a stress curve on a version of the Sir D, 7' 4 wt Hex 2 piece, and then added .002 to the diameter at each station.  Here are the results for stations 10 through 30:

        Sta  Stress +.002          Stress
        10  177488.9041    190396.2805
        15  147126.5652    155422.4058
        20  119281.3623    124602.6311
        25  115351.6498    119746.5262
        30  108939.9888    112464.5441

        As you can see, there are some relatively significant changes to the stress curve, ranging from almost 7% reduction at station 10 to a reduction of just over 3% at station 30.  I would interpret this as the added thickness of the rod making its action stiffer in the tip sections by the stated percentages.  (Claude Freaner)

          But would varnish affect the stress curve the same as .002 more cane?  (Neil Savage)

            Good question, Neil!  My response is "I don't know...".  Mostly depends on the characteristics of the varnish.  I'd guess, though, that if we're only talking .005 to .010  of  varnish,     as     long     as    it     doesn't     crack (over-stressed!), then it probably is contributing to the Moment (stiffness) of the rod.  I also base this conclusion on reports from list members who say that lots of intermediates will significantly stiffen a rod's action...

            Probably a question we'll never be able to answer with scientific proof...

            BTW, a .002 extra diameter caused by "thick glue" probably DOES affect the stress curve the same as .002 more cane, since it "pushes" the cane to the outside, effectively adding more cane.  (Claude Freaner)

              You're probably right that we wouldn't be able to figure this one out for certain.  But I'm thinking that since the stresses on a bamboo rod are essentially those of compression/tension, it's really the cross section of cane fibers that are affected.

              So, even though varnish may increase the measurement of a rod may by, say .006" I would imagine that, since the molecules of varnish are so flexible, they wouldn't participate much in the overall compression/tension.  That is, they WOULD have to undergo compression and tension,  of course, but would do so easily and probably without measurable resistance.  The thickness of the varnish would sort of "go along for the ride" without contributing any stiffness.

              But, we're not done yet.  There's the accumulated weight of the varnish film (whatever it might be), and that weight would become part of the stresses the rod must now support.  So, it would seem to me the varnish would actually detract from a given rod's stiffness.  And isn't that what we find when we load a rod up with varnish?  (Bill Harms)

                Good points, Bill.

                I went back and did a bit more playing with the numbers.  Here are the same two columns I presented before, plus I mad another run in which I doubled the weight for the varnish and guides.

                Sta  Stress +.002         Stress            Varn*2
                10  177488.9041    190396.2805    195141.052
                15  147126.5652    155422.4058    160967.99
                20  119281.3623    124602.6311    130224.879
                25  115351.6498    119746.5262    126103.983
                30  108939.9888    112464.5441    119167.665

                If we assume, as you suggest, that the varnish is flexible enough that it doesn't contribute to the stiffness because of it's added diameter, then the numbers in the third column do suggest that the only thing extra varnish does is slow the rod down a just a little bit.  Glue thickness will have a much more definite effect than the amount of varnish.  (Claude Freaner)

                  Since my post of last evening, I received an interesting email from Darrol Groth.  He reports that, using the "common cent" method, his rods show a small, but noticeable improvement in stiffness after varnishing.

                  My thoughts on that (to be filed in the "for whatever it's worth" bin) are that a reasonably thin film of varnish probably penetrates and bonds with the outer fibers somewhat, and this may have contributed to the positive difference that Darrol noticed.

                  But, I also think that at a certain point, more varnish will simply add weight and detract from the rod's stiffness.  So, it's not the added dimensions of varnish, but the added weight that becomes a negative factor.

                  Just thoughts.....  (Bill Harms)

                    Darrol was kind enough to send me the info also.  Perhaps we can find someone who can accurately measure the weight of a rod that is already varnished and has it's guides on, measure the diameter at all stations on the rod and then use the common cents method.  After than, add a couple more coats of varnish, reweigh, remeasure diameters, and redo the common cents measurements.

                    That should give us a fairly accurate assessment of the effects of varnish thickness...

                    Any volunteers (you do the work, then send me the data and I'll do the analysis and Hexrod runs...)?  (Claude Freaner)

              The extra .002 the glue adds will push the power fibers to the outside. The total mass is not all that different and the softer mass of the pith remains. Is the location of the power fibers so critical? What if they were of the same mass and density and on the inside? Isn't the mass of the power fibers he critical issue rather than where they are located?  (Rich Jezioro)

                The strength of a rod (its relative ability to resist forces of  bending)  is a  function   of cross-section mass, but also the shape and dimensions of that mass.  (Square or rectangular sections flex differently  from, say, round ones.)

                Additionally, a rod's strength is a function of how far the outer surfaces are located from the center point.  The relevant stresses, as they affect how "strong" a rod may be, are those of compression and tension, and these accrue, most significantly, on the outermost surfaces.  Conversely, the closer one goes toward the center line of a given section, the less significant will be the effect of compression/tension -- until at the very center (regardless of the nature of the material located there), the effect is almost nil.

                For example, if you were to build a solid blank of given cross-sectional mass, and then built another blank of the SAME CROSS-SECTIONAL MASS, but somehow left the center hollow, displacing the same amount of mass outward to form a rod of broader, thin-wall facets, you would have a far "stronger" rod in this second instance.  That's to say, you would have the same cross-sectional mass in the solid rod as in the hollow-but-larger rod, but the two will be vastly different in dimensions, and so, also in their ability to resist the forces of bending.

                So (back to the initial issue), if my theorizing is correct, displacing the available amount of cane outward .002", by virtue of the glue will create outer surfaces that are farther away from the center line.  And this would have the effect of causing the rod to feel somewhat stiffer.  Remember, increasing all the stations in a taper by .005" is to move up nearly one full line size, and I think glue could create that effect just as easily as more cane. (Bill Harms)

                Nope.  The movement of the power fibers further out from the center of the beam, in effect makes a wider web in the beam, therefore increasing the stiffness of the beam.  In beam theory, the web of the beam actually contributes very little to the overall characteristics of the beam, though it does create some.  What makes, or breaks (pun intended) the characteristics of the beam, is the distance away from center the flanges (power fibers) are located, the thickness of the flanges (power fibers), and to some extent, the shape of the flanges.  The web (pith, glue, etc) is more or less along for the ride, except that it separates the flanges from the center.  Take a look at some of the beams their using in new home building.  They use chipboard for the web, and 2 x whatever for the flanges.  They are strong enough to build a suspended floor that will hold the weight of a car, and then some.  (Mark Wendt)

          You are making an assumption here that I question. That assumption is that the effect of the additional thickness of the glue is the same as if it were an additional thickness of cane. I am not user that those are equal. Pushing the cane to the outside is not the same as adding more cane. More cane means more power fibers.

          Someone, I don't remember who, built a rod with the pith in locations other than the inside. I don't remember what difference that made. Does anyone remember?  (Rich Jezioro)

            Mark explained it quite accurately earlier when he wrote "Nope.  The movement of the power fibers further out from the center of the beam, in effect makes a wider web in the beam, therefore increasing the stiffness of the beam.  In beam theory, the web of the beam actually contributes very little to the overall characteristics of the beam, though it does create some. "

            To visualize this effect a bit easier, take a common wooden yardstick, hold it flat and try to bend it.  It will bend quite easily.  Now turn it 90 degrees and try to bend it.  You will likely be unable to bend it enough to actually measure the bend.  The reason it won't bend much (stiffer) in the vertical plane is because the fibers (wood in this case) are much farther from the centerline, therefore stiffer.

            Someone, I don't remember who, built a rod with the pith in locations other than the inside. I don't remember what difference that made. Does anyone remember?

            Having a different substance (pith) at the extreme distance from the centerline might have an effect, depending on the modulus of elasticity of the pith vs. that of the power fibers part of the cane and the proportion of the strip that is pith vs the proportion that is all power fiber.  If the strip is thick enough, it begins to act as a composite laminated from two different substances.  (The resulting stress can be calculated quite easily for this, given the dimensions, and the modulus of elasticity.)  This would probably only have an effect in the butt section because in the tip section of the rod there are hardly any pith areas left - it's all power fibers.  (Claude Freaner)

            Working with beams, for a given web thickness, the farther you move the webs from center, the stiffer the beam.  The web of the beam, as it approaches center diminishes in it's input to the overall beam stiffness because as it approaches center, it less and less of the tension and compression forces acting on the beam.  (Mark Wendt)

            I have made some experimentation on building inside-out and asymmetric rods, see the page here.

            The difference between traditional and inside-out was surprisingly minimal, therefore inside-out building could be used to modify easily the taper just by planing it thinner (you don’t take more power fibers than pith cane). The asymmetric rods has been  purely experimentation to look for the possible beneficial effect of the additional "spine" to the function of the rod. I think that I have build PHY Driggs River Special with 8 (eight) different ways (hex, penta, quad, inside-out, asymmetric x3, different kind of ferrules) during last three years -  and enjoyed the results. (Tapani Salmi)

              When the outer stiffer bits are disposed nearer to the center of the piece then their effect on the stiffness of the resultant laminate is slightly reduced.  But it IS slight, and bearing in mind the variations in stiffness inevitable in using a material which is anything but homogenous and wildly inconsistent from culm to culm to boot, the variations from those latter sources could easily outweigh that effect.  Strictly, it is bad practice, unless you have some other reason for wanting to do this.  I can't think what that reason may be, but that means I just don’t know and enlightenment would be most interesting.  (Robin Haywood)

          It looks like your plan is a good one and to introduce more complications to deal with second order variables would add immensely to the complications. But what you will be needing is a subjective way to measure performance. I would propose something like the much maligned  Common Cents Tests to give numbers rather than opinions.

          I was thinking of suggesting to include a hex rod in the mix, but in a fair test it would fare badly compared to the pentas.  (Bill Fink)

            I would think that a deflection test of some sorts would be a great idea. Could we do that in Roscoe?  (Bob Maulucci)

              I think that your idea of using some type of soak impregnation is a good one.  Just as something to throw out as a solid number, perhaps VPM's should also be measured, though I would think they would be similar. But that, of course, is the point of experimentation.  (Carl DiNardo)

              Yes, it's a simple test (Originated with plastic  rod builders.  What would  you expect?). I'll take it on myself to bring to Roscoe the bare essentials. Would be interesting to measure line sizes for other rods as well. It's a revealing and humbling test.  (Bill Fink)

          Was watching the glue differences with a lot of interest. Wonder if you have introduced so many variables in the rods that the "test" might be meaningless or @ least subject to some errors.

          If I might suggest,  you make up a number of coupons of cane perhaps each of them 6" long or so to negate the node question. You could take a series of coupons from the culm. If they were split about 0.200" wide, perhaps 12>16 or so from each section of the culm. Each coupon would have the pith/enamel removed with a thickness planer and milled to about 0.180". The coupons would be glued back to back [ or front to front ] with different glues. Each glued coupon then would be subjected to a flex test of some type. In doing so, variations of HT, tapering, reel seats, guide spacing, ferrule weight etc. would be removed from the equation. If a number of coupons were used - maybe 4 for of each type of glue, repeatability should be somewhat assured.  (Don Anderson)

            That's a good idea Don. I think short sections would be good for testing, but I don’t have much in the way of fancy measurement tools here.

            However, I don't think there would be that many variables in the rods. They would be all made from interspersed strips from the same culms, heat treated at the same time, same guide spacing, same hardware, and maybe one coat of Tung Oil to finish it off temporarily. I am doing all the blank making, so the strips should be dead one exact with the hand Mill and the stop  in place. I think deflection and casting tests are definitely in order.

            If nothing else, we will have some nice pentas to fish with this fall and for a long time to come.  (Bob Maulucci)


Just curious but has anyone come up with any concrete data on the effect of the glue on the action of the rod? For instance, I use Epon/Versamid which was made to flex when fully cured. Used for laminating bows. Therefor, one might assume it to have less resistance to bending than other epoxies such as Nyatex.

Have heard great things regarding Nyatex but chose Epon simply because it didn't require refrigerating, had good shelf life and long working time, etc. Just wondering how two rods from the same culm and with the same taper, one glued with Nyatex and one with Epon would compare.  (Wayne Kifer)

    I can't say about the comparison.

    I can say though that Nyatex really doesn't need refrigeration. Just don't keep a can of either part open longer than needed. Spoon some out and hammer the top shut. My first cans were quart cans of each part and they lasted for six years before I got scared and bought new stuff. The next quart of each lasted about four years before I ran out. The pint cans I'm using are two years old at this point. No special treatment on any of it. Apparently Nyatex was made to flex as well. My understanding is was made to glue flock to rubber for the seal around the window of a car door. That bends a bit as well.

    Whatever it was made for, it works very well IMHO.

    Anyone want to do some kind of comparison of the bending of a particular taper glued with Nyatex and Epon, opposed to the bending of the rubber around a Buick window?

    I know I'd be interested!  (Mike Shay)


Site Design by: Talsma Web Creations

Tips Home - What's New - Tips - Articles - Tutorials - Contraptions - Contributors - Search Site - Contact Us - Taper Archives
Christmas Missives - Chat Room - Photo Galleries - Line Conversions - The Journey - Extreme Rodmaking - Rodmaker's Pictures - Donate - Store