Relative Flexibility and Weight of Top and Down Tubes
John Clay
John Clay
John Clay
I intended to change all "Weight Reduction" headings to "Weight Change". I see that I missed a couple. The + and - signs will make more sense if you take care of that mentally.
On Friday, March 30, 2018 at 10:44:54 PM UTC-4, John Clay wrote:
Alistair Spence
--
--
You received this message because you are subscribed to the Google
Groups "Framebuilders" group.
Searchable archives for this group can be found at http://groups.google.com/group/framebuilders (recent content) and http://search.bikelist.org (older content).
To post to this group, send email to frameb...@googlegroups.com
To unsubscribe from this group, send email to
framebuilders+unsubscribe@googlegroups.com
For more options, visit this group at
http://groups.google.com/group/framebuilders?hl=en
---
You received this message because you are subscribed to the Google Groups "Framebuilders" group.
To unsubscribe from this group and stop receiving emails from it, send an email to framebuilders+unsubscribe@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.
Ken Cline
The torsion constant [https://en.wikipedia.org/wiki/Torsion_constant] for a tube is 2*pi*(d_o^4 - d_i^4), where d_o and d_i are the tube's outer and inner diameters. Multiply this by the material's shear modulus and integrate over (multiply by) length to determine a tube's torsional spring constant. I assume that shear modulus is, like Youngs modulus, very similar for all types of steel.
John Clay
Thanks to everybody who's offered to help.
John Clay
Tallahassee, FL
On Friday, March 30, 2018 at 10:44:54 PM UTC-4, John Clay wrote:
Jon Kendziera
--
Jon Norstog
Since you asked...
The torsion constant [https://en.wikipedia.org/wiki/Torsion_constant] for a tube is 2*pi*(d_o^4 - d_i^4), where d_o and d_i are the tube's outer and inner diameters. Multiply this by the material's shear modulus and integrate over (multiply by) length to determine a tube's torsional spring constant. I assume that shear modulus is, like Youngs modulus, very similar for all types of steel.
Andrew R Stewart
Rochester, New York
USA
--
You received this message because you are subscribed to the Google
Groups "Framebuilders" group.
Searchable archives for this group can be found at http://groups.google.com/group/framebuilders (recent content) and http://search.bikelist.org (older content).
To post to this group, send email to frameb...@googlegroups.com
To unsubscribe from this group, send email to
For more options, visit this group at
http://groups.google.com/group/framebuilders?hl=en
---
You received this message because you are subscribed to the Google Groups "Framebuilders" group.
John Clay
Y'all might want to wait until the work is checked before saving the chart. It's going to get expanded and improved as well.
John
On Saturday, March 31, 2018 at 10:08:32 PM UTC-4, Thursday wrote:
"Thursdfay"An easy way to compare steel sections is by using the "section modulus" which is based on the section's rotational moment of inertia. It is pretty easy to calculate and often available in tables from the manufacturer.jn
On Sat, Mar 31, 2018 at 2:45 PM, Ken Cline <cl...@frii.com> wrote:
Since you asked...
The torsion constant [https://en.wikipedia.org/wiki/Torsion_constant] for a tube is 2*pi*(d_o^4 - d_i^4), where d_o and d_i are the tube's outer and inner diameters. Multiply this by the material's shear modulus and integrate over (multiply by) length to determine a tube's torsional spring constant. I assume that shear modulus is, like Youngs modulus, very similar for all types of steel.
--
--
You received this message because you are subscribed to the Google
Groups "Framebuilders" group.
Searchable archives for this group can be found at http://groups.google.com/group/framebuilders (recent content) and http://search.bikelist.org (older content).
To post to this group, send email to frameb...@googlegroups.com
To unsubscribe from this group, send email to
For more options, visit this group at
http://groups.google.com/group/framebuilders?hl=en
---
You received this message because you are subscribed to the Google Groups "Framebuilders" group.
To unsubscribe from this group and stop receiving emails from it, send an email to framebuilder...@googlegroups.com.
John Clay
For each tube I combined the butted ends into a single unit. I then independently calculated the cantilever beam mode deflection of the butted and center sections due to the given load, and simply added them. That is certainly not a correct or accurate way to find the actual deflections but for similar tubes and small deflection differences I think it allows correct ranking and a reasonably accurate quantification of the relative differences. I won't post the results yet but it didn't change the tube-flexibility ordering.
In searching around for guidance on how to treat this problem properly I ran into this: http://www.eng-tips.com/viewthread.cfm?qid=360332
I particularly enjoyed the good natured, if friendly-snark seasoned, commentary on engineering in the computer age offered by Greg Locock, about halfway down the page. If you're going into engineering (and if I could get a do-over), pick a specialty, dive as deep as you possibly can and never stop. Otherwise it evaporates and at best all you can say is "I once kinda knew how to do that". At this point I'd rather cope a tube than spend weeks, or more probably months, trying to resurrect some long lost skills necessary to proper resolution of this problem.
Somewhere on this list must be a structures specialist who can set up the proper model....without the use of a canned FEA package. Feel free to offer your assistance!
John
On Saturday, March 31, 2018 at 7:58:01 PM UTC-4, Jon Kendziera wrote:
That's a useful table and confirms most of my observations and calculations. A suggestion: In addition to torsional stiffness how about considering out of plane bending? A smaller factor in down tube stiffness, but more important for seat and top tubes. I suspect the results will be similar.
On Sat, Mar 31, 2018 at 6:45 PM, John Clay <nice.c...@gmail.com> wrote:
Alistair is checking the spreadsheet equations. There's also been an offer by someone who's far ahead of me with spreadsheet design/creation to improve and refine its capabilities as a tool for others. Once all of that is complete I'll make it available to anyone who wants it.
Thanks to everybody who's offered to help.
John Clay
Tallahassee, FL
On Friday, March 30, 2018 at 10:44:54 PM UTC-4, John Clay wrote:
John ClayTallahassee, FL
--
--
You received this message because you are subscribed to the Google
Groups "Framebuilders" group.
Searchable archives for this group can be found at http://groups.google.com/group/framebuilders (recent content) and http://search.bikelist.org (older content).
To post to this group, send email to frameb...@googlegroups.com
To unsubscribe from this group, send email to
For more options, visit this group at
http://groups.google.com/group/framebuilders?hl=en
---
You received this message because you are subscribed to the Google Groups "Framebuilders" group.
To unsubscribe from this group and stop receiving emails from it, send an email to framebuilder...@googlegroups.com.
Alistair Spence
Alex Wetmore
Thanks John. I made a similar spreadsheet 10ish years ago when we did the planing tests for BQ, but have since lost it. Your numbers match what the two rules of thumb that I put into my head:
- 7/4/7 OS rides like 9/6/9 standard, but is about 100 grams lighter
- 7/4/7 standard is about 150 grams lighter and almost twice as flexible
- 8/5/8 standard is about 80 grams lighter and somewhat more flexible
I think in terms of 9/6/9 standard as "the standard" because there that was the norm for so long, and anyone my age (mid-40s now) or older knows what those bikes feel like.
I also did the math using cantilevered beam calculations, but I'm not entirely sure that it is fair.
The other number that is interesting to add is dent resistance for heat treated vs not numbers. This showed me how they picked the amount of heat treatment to add for Verus HT (8/5/8) and Platinum OX (7/4/7) to get dent resistance than matched 9/6/9. It's probably worth doing that with tubing that is being made now though...
alex
Sent: Sunday, April 1, 2018 6:37 PM
To: John Clay
Cc: Framebuilders
Subject: Re: [Frame] Re: Relative Flexibility and Weight of Top and Down Tubes
John Clay
In addition to the summary tables below I've attached a copy of the spreadsheet so anyone can see how the calculations were made and modify or refine as might suit. I used torsion as the proxy for "overall" flexibility. That seems reasonable since the goal is a relative ranking of tube stiffness. The butts were modeled as steps located at mid-taper. Most of the tubing dimensions came from True Temper. Thats reasonably close to reality and it made the calculations simple & quick, and the results good enough for the purpose.
Top tube, relative stiffness calculation results: Lavender shaded columns are comparisons wrt 25.4/969 tubing. Aqua columns are against 25.4/747 tubing. Same drill for the DT section that follows except it's against 28.6 tubing.
I seem to recall, but haven't been able to find it again, Jan mentioning that for he & his crew planing occurs at higher power levels; something in the range of 500 or 600 watts. And the frames that plane for them have thinner or thinnest wall tubing. That suggests to me that planing might not be available to those of us who operate at far more modest power output levels. I'd be interested to know if my recollection is correct, and if my hunch seems sound as well. If so then planing may not be feasible for those of us who operate at far lower power levels.
| Assembly Weight | Weight Change (g) wrt 969 Conventional | % Weight Change wrt 969 Conventional | % Weight Change wrt 747 Conventional | Angular Displacement (deg) | Relative Increase in Torsional Flexibility wrt 969 Conventional | Relative Increase in Torsional Flexibility wrt 747 Conventional | |||
| 25.4/747 TT | 213 | -74 | -26% | 8.0 | 39% | ||||
| 25.4/858 TT | 250 | -37 | -13% | 17% | 6.6 | 16% | -17% | ||
| 25.4/969 TT | 287 | 35% | 5.7 | -28% | |||||
| 28.6/747 TT | 241 | -46 | -16% | 13% | 5.5 | -3% | -30% | ||
| 28.6/757 TT | 265 | -22 | -8% | 24% | 4.8 | -16% | -40% | ||
| 28.6/858 TT | 283 | -4 | -2% | 33% | 4.6 | -19% | -42% | ||
| 28.6/969 TT | 324 | 84 | 13% | 52% | 4.0 | -31% | -50% |
Down tube, relative stiffness calculation results.
| Assembly Weight | Weight Change (g) wrt 969 Conventional | % Weight Change wrt 969 Conventional | % Weight Change wrt 747 Conventional | Angular Displacement (deg) | Relative Increase in Torsional Flexibility wrt 969 Conventional | Relative Increase in Torsional Flexibility wrt 747 Conventional | |||
| 28.6/747 DT | 241 | -84 | -26% | 5.5 | 40% | ||||
| 28.6/747 DT Compass | 226 | -98 | -30% | -6% | 5.9 | 48% | 6% | ||
| 28.6/858 DT | 283 | -42 | -13% | 17% | 4.6 | 16% | -17% | ||
| 28.6/969 DT | 324 | 35% | 4.0 | -28% | |||||
| 31.8/747 DT | 268 | -56 | -17% | 11% | 4.0 | 1% | -28% | ||
| 31.8/757 DT | 295 | -29 | -9% | 23% | 3.5 | -12% | -37% | ||
| 31.8/757 DT Compass | 286 | -38 | -12% | 19% | 3.6 | -10% | -36% | ||
| 31.8/858 DT | 315 | -9 | -3% | 31% | 3.3 | -16% | -40% | ||
| 31.8/969 DT | 361 | 93 | 11% | 50% | 2.9 | -28% | -48% |
John Clay
Tallahassee, Florida
satanas
As for planing needing 5-600 Watts, I'm not convinced. IME, when things are right the difference is apparent even at low power. However, 1) I don't have a power meter, and 2) as there's no scientific definition of "planing" we might all be talking about different things.
Later,
Stephen
John Clay
Hi Everybody,
I've gotten a number of interesting off-list-only responses to this thread. While I enjoy and welcome private communications, most of these have included material that would be useful to any number of folks on the list. So rather than keeping your bits of information and experience between the two of us, please post them within the public thread where they will have far wider benefit, might catalyze a broader discussion of the subject at issue and even bring up other related (or not) subjects.
Thanks,
John Clay
Tallahassee, Floridap
On Friday, March 30, 2018 at 10:44:54 PM UTC-4, John Clay wrote:
Ewen Gellie
Mark Bulgier
John, thanks for this, very useful.
I made a similar spreadsheet for Davidson when I worked there, probably about 1990. I assume we get the same results, since math is math. I didn’t take a copy of that spreadsheet with me when I left though, because they were touchy about their intellectual property…
If I could suggest the smallest change, I would say don’t call them out as top tubes or down tubes. Some TTs are 28.6, and some DTs are 25.4, so just say the diameter and let the builder decide where each tube goes. Also it might be more conceptually easy to follow if you gave the stiffness of the tube rather than its reciprocal, the flexibility. Then both arrows point the same way – as diameter or gauge increases, so does the stiffness. That might just be me though – maybe some people think more in terms of deflection rather than stiffness.
Mark Bulgier
Seattle
John Clay
Mark wrote: "If
I could suggest the smallest change, I would say don’t call them
out as top tubes or down tubes. Some TTs are 28.6, and some DTs
are 25.4, so just say the diameter and let the builder decide where
each tube goes. Also it might be more conceptually easy to
follow if you gave the stiffness of the tube rather than its
reciprocal, the flexibility. Then both arrows point the same
way – as diameter or gauge increases, so does the stiffness.
That might just be me though – maybe some people think more in
terms of deflection rather than stiffness."
I agree with you on both counts! Good suggestions. There are a few other revisions that would be worthwhile too. I'm not sure I'll get around to making them but, maybe.
I started out with 969 conventional as the comparison baseline since it seems to be the defacto historical standard, but having the other table with thinnest wall conventional as baseline adds clarity and is a little more straightforward.
This should facilitate rational changes in tube selection recipes particularly when there's a frame of known tubes from which one wants to deviate.
One day somebody is going to generate a comprehensive tubing & geometry adjustable FEA model that spits out flexibility and frequency response for an entire frame, and exercise physiologists will be able to determine what range of those values are optimum for a given rider (which will vary as a function of effort level). In the mean time, this should help folks make informed design decisions and determine whether or not the characteristics ascribed to a particular frame make sense. For example: It would seem contradictory that frames from thinwall OS tubes “plane” if 969 or 858 conventional don't.
MNF
--
--
You received this message because you are subscribed to the Google
Groups "Framebuilders" group.
Searchable archives for this group can be found at http://groups.google.com/group/framebuilders (recent content) and http://search.bikelist.org (older content).
To post to this group, send email to frameb...@googlegroups.com
To unsubscribe from this group, send email to
For more options, visit this group at
http://groups.google.com/group/framebuilders?hl=en
---
You received this message because you are subscribed to the Google Groups "Framebuilders" group.
To unsubscribe from this group and stop receiving emails from it, send an email to framebuilders+unsubscribe@googlegroups.com.
Alex Wetmore
Mike, I think those are good concerns about modern ultralight tubing.
At the same time it is a hard comparison to make because the ultralight tubing usually has different heat treatment and properties than the heavier stuff. I'm most familiar with True Temper's catalog (RIP), but they used significantly different heat treatment for Verus (plain 4130, 9/6/9), Verus HT (heat treated 4130, 8/5/8), Platinum OX (7/4/7) and S3 (down to 5/3/5). The fatigue properties are going to be very different across those tubes.
I've long lost my spreadsheet, but I did the math for dent resistance of these 4 tubesets and found that they were all very similar (and I don't think that was an accident). Dent (yield) is different than fatigue too of course and doing FEA on tube fatigue is out of my skill set.
There are a lot of 7/4/7 bikes that are multiple decades old now that have also been ridden hard, so there should be some good empirical evidence.
I actually think that 9/6/9 bikes ride great for me, as long as they aren't built with oversized tubing. It's great having Surly, Soma, and others bring steel bikes back to the mainstream over the last 15 years, but I think they did a dis-service to them by making oversized tubing the norm. I also like oversized bikes with lightweight tubing, and it is a good (but with more expensive materials) way to save 100 grams or so on the frame.
alex
Sent: Monday, July 9, 2018 9:56:23 AM
To: John Clay
Cc: Framebuilders
Subject: Re: [Frame] Re: Relative Flexibility and Weight of Top and Down Tubes
For more options, visit this group at
http://groups.google.com/group/framebuilders?hl=en
---
You received this message because you are subscribed to the Google Groups "Framebuilders" group.
M-gineering
On 7/9/2018 7:05 PM, Alex Wetmore wrote:It's great having Surly, Soma, and
> but I think they did a dis-service to them by making oversized tubing
> the norm.
with high amplitudes designed for aluminium frames. And now with
diskbrakes, suspension 'corrected' forks etc
--
mvg
Marten Gerritsen
Kiel Windeweer
Netherlands
Dan Chambers
Eric Nichols
- Torsion (twisting) resistance is related to the fourth power of tube diameter
- Bending resistance is related to the square of the tube diameter.
John Clay
Eric,
Thanks.
I've forgotten an awful lot but are you sure about that? The units of planar and polar second moments of area are both length to the fourth power; the two deflection modes being proportional to the inverse of their relevant area moments.
The magnitude of the two moments, for the same tube, would be different due to the different arrangement of the area geometry wrt the two, orthogonal reference axes.
In any event, the purpose was to allow an FBer to know that this tube is a lot (or little) more (or less) stiff than that tube. And the stiffness context was “planing” or “liveliness”. And I didn't want to overheat my brain in trying to relearn years worth of engineering education that I've never used and long forgotten!!
I made a Section Modulus comparison (same metal mass but in a straight gauge arrangement) and it didn't contradict the tube ordering indicated by the torsion analysis.
John
Eric Nichols
Cheers
Eric
maskdesigner
A 52cm frame with a 140 lb rider isn't going to have the same feel as a 62cm frame with a 180 lb rider if they are made from the same tube set.
It's even more confusing if you reverse the weights.
If I had classic bicyclist proportions the charts above could be applied pretty easily but my interest in frame building is largely borne from the fact that nothing off the rack fits me and since I have short legs, the frames that come close are designed for wee little men while I walk around with a 50" chest.
I think the answer for me is OS .8/5/8 or OS .9/6/9 but I wish someone made OS .9/5/9 to make it a bit easier to weld.
If anyone knows of a tube set like that, please let me know.
Andrew R Stewart
have a different feel. And neither rider will ride the other's bike for any
real efforts or valid comparisons. The two are just too different in size,
rider weight placement and how each rider will evaluate the feel.
I suggest that you are placing too much focus on numbers. Be they scale
weight or lab measured flex (torsion, beam...). There's far more to a well
riding bike then these two factors. The longer I've ridden bikes I make for
myself the less focused on weight I get and my flex concerns go more to
speed mans wobble then sprinting feel.
I just ran some math about tube weight differences due to only a .1mm
difference of wall in the center section of a common double butted tube. I
got 24gm. So for the TT and DT that less then 2 oz. of weight. This is less
then a typical pee:)
I do understand the concerns about welding thin walls and there a 10% change
can be a difference if your skills are not spot on.
To close, at some point you either have to trust your builder or make a
frame your self and see how it feels. Andy (who is tempted to say "either
pee or get off the pot" but understands to many references to pee might be
in bad taste)
Andrew R Stewart
Rochester, New York
USA
From: maskdesigner
Sent: Saturday, December 08, 2018 11:10 PM
To: Framebuilders
Subject: Re: [Frame] Re: Relative Flexibility and Weight of Top and Down
Tubes
M-gineering
On 12/9/2018 5:10 AM, maskdesigner wrote:
> I think the answer for me is OS .8/5/8 or OS .9/6/9 but I wish someone made OS .9/5/9 to make it a bit easier to weld.
>
> If anyone knows of a tube set like that, please let me know.
and still being able to remove the mandrell
maskdesigner
A normally proportioned human can go to a better bike shop and try many flavors of tubing and geometry to see what they prefer so they really don't have much need for the numerical comparison and at the same time, the people who need it the most (because nothing fits them) get left out of the equation.
No one needs a prediction for what they can practically test unless they are comparing to gain confidence in a different prediction they can't test.
It's just frustrating to see something so close to being useful.
satanas
Later,
Stephen
John Clay
Here’s a rough swing at ranking the relative flexibility results of various TT/DT combinations for otherwise identical frames. I’m not expert in this and I hope I don’t embarrass myself but for better or worse you can find the table I made, here, as posting an image of it to the thread failed:
https://www.flickr.com/photos/21624415@N04/50641844997/in/dateposted/
The table uses the percentage change in deflection per unit load of the individual tubes, sums the values of the various pairs and divides the result by two (based on the rough assumption that each tube contributes half to the overall flexibility of each pair) to yield a percentage deviation from the 969 conventional diameter baseline. I don’t know if that’s academically correct but it seems to me that it ought to be reasonably close; close enough to be able to rank the various combinations from most flexible to least.
Anybody who can contribute more knowledge about this subject or approach than I is welcome to do so. This was a pretty quick draw from the hip this morning; I hope I haven’t grossly violated any first principles but it's entirely possible!
Note that I may have not used the RH butting/belly dimensions for the 25.4, 747 tubes; I don’t think I did that in the original spreadsheet and didn’t take the time to check or revise.
John Clay
Tallahassee, Florida
John Clay
Dan Chambers
Still viewable and functioning at the Web Archive: http://web.archive.org/web/20050215103730/http://anvilbikes.com/tubeCalc.php
John Clay
Dan Chambers
--
--
You received this message because you are subscribed to the Google
Groups "Framebuilders" group.
Searchable archives for this group can be found at http://groups.google.com/group/framebuilders (recent content) and http://search.bikelist.org (older content).
To post to this group, send email to frameb...@googlegroups.com
To unsubscribe from this group, send email to
framebuilder...@googlegroups.com
For more options, visit this group at
http://groups.google.com/group/framebuilders?hl=en
---
You received this message because you are subscribed to a topic in the Google Groups "Framebuilders" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/framebuilders/QaOg8c-17eE/unsubscribe.
To unsubscribe from this group and all its topics, send an email to framebuilder...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/framebuilders/c8168b7f-1634-4b2c-bb8b-74061cad4155n%40googlegroups.com.