derailer question
F. Hayashi
Will old style friction downtube shifters (like Campy NR shifters) coupled
to older rear derailers (like old Campy, Simplex, etc) work with an
8-speed cassette?
I'd like to put together a ep the wheels compatible with my other bike.
Thanks
+-------------------------------------------------------------+
| Fumitaka Hayashi - hay...@u.washington.edu |
| http://macrophage.immunol.washington.edu/~fumi/index.html |
| Aderem Lab - Dept. of Immunology - University of Washington |
+-------------------------------------------------------------+
TBGibb
In article <Pine.A41.3.96a.98010...@dante01.u.washington.edu>,
"F. Hayashi" <hay...@u.washington.edu> writes:
>Will old style friction downtube shifters (like Campy NR shifters) coupled to
>older rear derailers (like old Campy, Simplex, etc) work with an 8-speed
>cassette?
I am getting good results using a 1970s vintage Huret derailer (originally 5
speed) with a 7 speed cassette. But when I look at the limit screws I'm not
sure I'll get away with 8.
Tom Gibb <TBG...@aol.com>
Peter Guyton
F. Hayashi wrote in message ...
>Will old style friction downtube shifters (like Campy NR shifters) coupled
>to older rear derailers (like old Campy, Simplex, etc) work with an
>8-speed cassette?
>
>
>Thanks
>
>+-------------------------------------------------------------+
>| Fumitaka Hayashi - hay...@u.washington.edu |
>| http://macrophage.immunol.washington.edu/~fumi/index.html |
>| Aderem Lab - Dept. of Immunology - University of Washington |
>+-------------------------------------------------------------+
>
>
seven speed setup without a problem. Like Mr. Gibb said, it really depends
on the limit screws. The Simplex, appeared as though it had enough extra
play to handle an 8-speed (maybe) but unfortunately I never tried it. I
think it will really depend on the make of derailleur, the vintage and the
amount of extra "room" the manufacturer left. The position at which the
dropout hanger puts the derailleur may also be a factor.
Regards,
Peter Guyton
F. Hayashi
On Sat, 3 Jan 1998, Peter Guyton wrote:
> I used an older Simplex rear derailleur (SLJ 5500 - all aluminum) with a
> seven speed setup without a problem. Like Mr. Gibb said, it really depends
> on the limit screws. The Simplex, appeared as though it had enough extra
> play to handle an 8-speed (maybe) but unfortunately I never tried it. I
> think it will really depend on the make of derailleur, the vintage and the
> amount of extra "room" the manufacturer left. The position at which the
> dropout hanger puts the derailleur may also be a factor.
> Regards,
> Peter Guyton
So, how about a 'modern' derailer with old friction levers?
Do old friction levers pull enough cable to move, say a Shimano 105sc rear
derailer, across 8 cogs with campy spacing?
I think this combination could get me shifters and a rear derailer for
less than the price of a used derailer... or one STI/ergo lever.
Hoyt McKagen
M. McMaster wrote:
Mark McMaster wrote:
>
> Loading the wheel shortens the bottom spokes from
> their original static (albeit highly tensed) state, so how can you say
> they are not being compressed?
Because compression is contradictory to tension by convention (the signs are different
in testing machines and analysis) and reality (the diff between pushing and pulling).
IOW they cannot be in compression and tension at the same time. As I noted, you are free
to do an analysis starting with the baseline anywhere you like and you will get the same
results EXCEPT for it not matching what's really going on.
> So yes, we are essentially talking about a reduction in tension in the
> bottom spokes, but that does not mean it allows the rim to carry the
> loading to the top of the wheel, at least not much of it.
The rim may be weak radially but it is many times as stiff as a spoke in compression or
tension. So, since it is constained laterally and radially by the spokes, it can carry
considerable load in compression, all of it, easily, with ordinary human riders.
> The high static spoke tension is carried by the rim primarily by the
> circumferential stiffness and strength of the rim, rather than by its
> radial stiffness, which is quite low.
True
> For this reason, the spoke
> tension must be reasonably well balanced around the rim for the wheel to
> retain its strength and stiffness.
Not true; I have ridden them and had spokes loosen til they rattle and I have had spokes
broken too. The wheel will continue to work, up to the point where it looses lateral
support and potato-chips. (No, I haven't ever had one of those: I fix wheels before they
get that bad.) Yes, I know a wheel under proper tension may p-chip also, but that still
a function of lateral strength, as affected most often by poor torsional strength in the
rim). Of course the point you make is still valid, as the rim will perform non-uniformly
if not properly tensioned, but that it will still perform does reflect on the
compressive strength of the rim
> the spokes on the bottom of the
> wheel pull less downward on hub than before loading, rather than the
> spokes on the top of the hub pulling more upward. In this way it is
> correct to say that the load on the wheel is carried by spoke
> compression.
You are correctly stating what happens but not labeling the spoke's state to match. If
it is in tension, it cannot be in compression, by definition. IOW, that answer has zero
semantic content.
> The tension increases that result from increased rim radius are
> distributed around all of the wheel outside of the flattened load
> affected zone, including large portions of the rim below the hub.
True
>... does not imply tension increases carry the load ...
It does, see my measurements below.
> net load on the hub from the small tension increases from all directions
> around the hub is very small - certainly not enough to support the load
> on the wheel.
According to my experiment the total increase in strain, outside the contact patch, is
equal to the total decrease in strain in the contact patch area. That is consistant with
the loaded energy state of the wheel being constant. See my measurements, below.
> Given the wide disparity in the radial stiffnesses of the rim and of
> the spokes,
The compressive strength of the rim is the key point, not its radial stiffness.
(For example, if a steel cable and a
> rubber band were to be run side by side between two hooks, and then the
> hooks pulled apart to load them in tension, the steel cable, being much
> stiffer than the rubber band, would carry much more of the load the
> rubber band.)
True, but not explicit to the case at hand.
> > > The tension change supporting the load is carried almost exclusively by
> > > spokes at the bottom of the wheel in the region around the ground
> > > contact point.
My results show that every spoke changes tension. If you and Jobst missed that or
ignored it (I believe the latter in his case) you were simply being sloppy
experimenters.
> They are capable of carrying compressive load precisely _because_ they
> never lose complete tensile pre-load. As stated earlier, any load which
> acts to shorten the spoke is a compressive load, even if the original
> length of the spoke is achieved through a static pre-load.
Tension and compression are contradictory phenomena and cannot co-exist in the same
simply-loaded member.
> > This is self-contradictory, as above you said the upper spokes had an additional
> > increment of tension. Those increments 1) relate to the load on the wheel (as it is
> > overall flexible) 2) have components which operate in the right direction, 3) add up to
> > the right magnitude.
>
> Those increments 1) relate to the load on the rim, which is only a
> small portion of the load on the wheel, 2) have components which operate
> in the right direction on the top of the rim, but the wrong direction on
> the portions of the bottom of the rim outside of the flattened zone, 3)
> do not add up to the right magnitude.
I stand by 1) and 3). In 2) the forces are redirected through the rim.
> I have made sufficient measurements to demonstrate to myself that the
> finite element analysis in "The Bicycle Wheel" by Jobst Brandt is
> correct.
Please post those, if you can.
My impression is that Brandt himself stated the number of nodes in the analysis may have
been insufficient.
> And your measurements or calculations?
I set up a bike frame last night with the bike in a propped-up exercise stand (so the
headstock was vertical) and loaded the front wheel with dead weights set on a post
inside the steering tube. I used a come-a-long to unload the wheel while rotating it
between stations. A dial indicator was clamped to the hub inside the wheel and the tip
rested on the rim between two spokes. I took measures all the way around, at every 10
deg (36-spoke wheel), starting at the bottom, IE the indicator-bottom position is #0.
Load was about 50 lbs (cinder blocks and an auto flywheel). `Plus' deflections are
increased tension/greater radius and `minus' are decreased tension/smaller radius.
The indicator was zeroed after each move, the load applied and the reading taken, then
wheel unloaded read again. The indicator re- zero'd on each reading with good agreement
as a rule, but some striction in it made some readings difficult. In those cases, I
loaded, zeroed, read, unloaded as depending on direction of deflection, one method would
drive the indicator and the other relax it; I found that readings were only
problematical when the indicator was relaxing, probably due to internal friction against
the smaller motions.
Overall, I am confident of these readings to .0001. The table below shows the results:
Pos Def Pos Def
0 -.0045" (-.0028) 19 .0004
1 -.0025 (-.0015) 20 .0005
2 -.0005 (-.0001) 21 .0004
3 0 (.0004) 22 .0003
4 .0002 23 .0003
5 .0003 24 .0003
6 .0003 25 .0004
7 .0002 26 .0003
8 .0002 27 .0003
9 .0002 28 .0004
10 .0003 29 .0003
11 .0004 30 .0002
12 .0003 31 .0004
13 .0003 32 .0004
14 .0004 33 .0003 (.0003)
15 .0004 34 .0002 (.0005)
16 .0003 35 -.0009 (-.0001)
17 .0003 36 -.0028 (-.0008)
18 .0004 0 -.0042 (check in)
Numbers in () are done with the tire deflated: I took only those readings shown that
way.
Taking this at face value, we have a small portion of the wheel deflected to a smaller
radius and the entire rest of the wheel increased. As I noted, the sums of each were
almost identical, well within 1%, and I think the same thing would be true deflated if
I had finished it out. Since I didn't weigh the load or measure the spoke tension you
cannot say what strain produced these deflections, but given that the wheel was a
commercially made unit which has been used but is still round and as far as I know
never trued, it must have approximately equal spoke tension all around. You may note
some anomolies around positions 7-10 and perhaps 30. I don't know if that is an artifact
or whether the wheel actually becomes slightly triangular; maybe that is the bulge you
mentioned. However, you can see a trend to more deflections at the top of the wheel than
the sides, which is consistent with a semi-flexible rim and increased support from the
top.
You would need to know your test weight and intitial strain to calculate whether each
sum is equal to the test load, hence whether the load is shared between top and bottom
or whether (as I suspect) the deflections at the bottom are not as a result of applied
loading; IE, the hub hangs ENTIRELY from the top spokes. However, if we assume 50 lbs
load, 200 lbs initial tension and 11" spoke length, ALL of the minus deflections don't
add up to nearly enough, a total of only 19 lbs or so or very approximately half of the
load, and of course the strains in the rest also account for about half. <=The calcs are
very suspect though and I would like to see better figures. What I think will happen is
that properly calc'd out, both totals will equal the applied load, hence only one
phenomena can account for carrying it.
Why would I think the above flattening of the rim isn't related to the appied load?
BECAUSE WHEN I LET THE AIR OUT OF THE TIRE, IT CHANGES!! <=Correct, the profile of
deflected rim is different, that is MORE ROUND, even though the load is the same ... and
you'd better try this before you argue about it. That means the (inflated) rim's greater
deflection is an interaction between tire pressure and tire configuration in the contact
patch. I should leave explaining this finding as an exercise for you to figure out, but
IAC, it puts the nails in the coffin of your theory.
Hint: by putting air in a wheel you have increased circumference and radius (same as in
the sensing element of a pressure gauge, or a bare inner tube, and true of composite
structures like tire/rim also). The interaction is mediated by the connection between
tire and rim at the bead, which under pressure exerts a net radial outward force, a
summing of all radial vectors of forces tangental to the tire cross-section. Fact is,
the wheel grows about .002" dia with 70-80 lbs, try it! But in the contact patch, tire
config changes and so do those vectors, producing a net loss of the above radial force.
Hence, the spokes in the contact patch zone distort the rim inwards ... since the
distortion is a result of the above, it is not a function of load carrying. That is only
natural, as, since the tire is also not capable of carrying a compressive load (it is
also specialized only in tension), it can't `push' upward on the rim anyway there. And I
think that leaves tension in the tire, from above and the sides, to generate the
load-carrying on the rim,and the spokes above and to the side, to carry it to the hub.
Yes, I realise my accounting is imperfect, but it points to something you and Jobst
don't even address, so QED, babay.
Hoyt
--
Belfab- http://www.freeyellow.com/members/belfab/belfab.html
Penpals- http://www.freeyellow.com/members/batwings/penpals.html
Best MC Repair- http://www.freeyellow.com/members/batwings/best.html
Camping/Caving- http://www.freeyellow.com/members/batwings/caving.html
=> If you die, babay, we split your gear <=
Hoyt McKagen
Mark McMaster wrote:
>
> Loading the wheel shortens the bottom spokes from
> their original static (albeit highly tensed) state, so how can you say
> they are not being compressed?
Because compression is contradictory to tension by convention (the signs are different
in testing machines and analysis) and reality (the diff between pushing and pulling).
IOW they cannot be in compression and tension at the same time. As I noted, you are free
to do an analysis starting with the baseline anywhere you like and you will get the same
results EXCEPT for it not matching what's really going on.
> So yes, we are essentially talking about a reduction in tension in the
> bottom spokes, but that does not mean it allows the rim to carry the
> loading to the top of the wheel, at least not much of it.
The rim may be weak radially but it is many times as stiff as a spoke in circumferential
compression or tension. So, since it is constained laterally and radially by the spokes,
it can carry considerable load in compression. To find the limit of that, you need to
test a section of rim just long enough to span two spokes; don't be suprised if a
steamlined section of good alloy would go 3-4000 lbs.
> The high static spoke tension is carried by the rim primarily by the
> circumferential stiffness and strength of the rim, rather than by its
> radial stiffness, which is quite low.
True
> For this reason, the spoke
> tension must be reasonably well balanced around the rim for the wheel to
> retain its strength and stiffness.
Not quite true; I have ridden them and had spokes loosen til they rattle and I have had
spokes broken too. The wheel will continue to work, up to the point where it looses
lateral support and potato-chips. (No, I haven't ever had one of those: I fix wheels
before they get that bad.) Yes, I know a wheel under proper tension may p-chip also, but
that still a function of lateral strength, as affected most often by poor torsional
strength in the rim). Of course the point you make is still valid, as the rim will
perform non-uniformly if not properly tensioned, but it does reflect on the compressive
strength of the rim that it will still perform.
> the spokes on the bottom of the
> wheel pull less downward on hub than before loading, rather than the
> spokes on the top of the hub pulling more upward. In this way it is
> correct to say that the load on the wheel is carried by spoke
> compression.
You are correctly stating what happens but not labeling the spoke's state to match. If
it is in tension, it cannot be in compression, by definition.
> The tension increases that result from increased rim radius are
> distributed around all of the wheel outside of the flattened load
> affected zone, including large portions of the rim below the hub.
True
>... does not imply tension increases carry the load ...
It does, see my measurements below.
> net load on the hub from the small tension increases from all directions
> around the hub is very small - certainly not enough to support the load
> on the wheel.
According to my experiment the total increase in strain, outside the contact patch, is
about equal to the total decrease in strain in the contact patch area. That is
consistant with the loaded energy state of the wheel being constant. See my
measurements, below
> Given the wide disparity in the radial stiffnesses of the rim and of
> the spokes,
The compressive strength of the rim is the key point, not its radial stiffness.
(For example, if a steel cable and a
> rubber band were to be run side by side between two hooks, and then the
> hooks pulled apart to load them in tension, the steel cable, being much
> stiffer than the rubber band, would carry much more of the load the
> rubber band.)
True, but not explicit to the case at hand.
> > > The tension change supporting the load is carried almost exclusively by
> > > spokes at the bottom of the wheel in the region around the ground
> > > contact point.
My results show that every spoke changes tension unless it is passing the points of
inflection. If you and Jobst missed that or ignored it you were simply not doing good
experiments.
> They are capable of carrying compressive load precisely _because_ they
> never lose complete tensile pre-load. As stated earlier, any load which
> acts to shorten the spoke is a compressive load, even if the original
> length of the spoke is achieved through a static pre-load.
Tension and compression are contradictory phenomena and cannot physically co-exist in
the same simply-loaded member. That's from hands-on work with testing machines, and any
engineering textbook covering materials testing will inform you of that too. It's
axiomatic too that a member in tension cannot carry more load (which requires that it
increase its energy state) by becoming shorter (when it loses energy). What kind of
engineer are you?
> > This is self-contradictory, as above you said the upper spokes had an additional
> > increment of tension. Those increments 1) relate to the load on the wheel (as it is
> > overall flexible) 2) have components which operate in the right direction, 3) add up to
> > the right magnitude.
>
> Those increments 1) relate to the load on the rim, which is only a
> small portion of the load on the wheel, 2) have components which operate
> in the right direction on the top of the rim, but the wrong direction on
> the portions of the bottom of the rim outside of the flattened zone, 3)
> do not add up to the right magnitude.
I stand by 1) and 3). In 2) the forces are redirected through the rim.
> I have made sufficient measurements to demonstrate to myself that the
> finite element analysis in "The Bicycle Wheel" by Jobst Brandt is
> correct.
Please post those.
> And your measurements or calculations?
I set up a bike frame last night with the bike in a blocked -up exercise stand (so the
headstock was vertical) and loaded the front wheel with dead weights set on a post and
collar in the steerer tube. I used a come-a-long to unload the wheel while rotating it
between stations. A dial indicator was clamped to the hub inside the wheel and the tip
rested on the rim between two spokes. I took measures all the way around, at every 10
deg (36-spoke wheel), starting at the bottom, IE the indicator-bottom position is #0.
Load was about 40-50 lbs. `Plus' deflections are increased tension/greater radius and
`minus' are decreased tension/smaller radius.
The indicator was zeroed after each move, the load applied and the reading taken, then
wheel unloaded and read again. The indicator zero'd on each reading with good agreement
as a rule, but some striction in it made some readings difficult. In those cases, I
loaded, zeroed, unloaded, read, as depending on direction of deflection one method
would drive the indicator and the other relax it; I found that readings were only
problematical when the indicator was relaxing, probably due to internal friction against
the smaller motions. Overall, I am confident of these readings to .0001. The table below
shows the results:
Pos Def Pos Def
0 -.0045" (-.0028) 19 .0004
1 -.0025 (-.0015) 20 .0005
2 -.0005 (-.0001) 21 .0004
3 0 (.0004) 22 .0003
4 .0002 23 .0003
5 .0003 24 .0003
6 .0003 25 .0004
7 .0002 26 .0003
8 .0002 27 .0003
9 .0002 28 .0004
10 .0003 29 .0003
11 .0004 30 .0002
12 .0003 31 .0004
13 .0003 32 .0004
14 .0004 33 .0003 (.0003)
15 .0004 34 .0002 (.0005)
16 .0003 35 -.0009 (-.0001)
17 .0003 36 -.0028 (-.0008)
18 .0004 0 -.0042 (check in)
-Sum -.0112 +Sum .0099
Numbers in () are done with the tire deflated. See below.
We have a small portion of the wheel deflected to a smaller
radius and the entire rest of the wheel increased. The sums of each were
fairly close. Since I didn't weigh the load or measure the spoke tension we can't
say what strain produced these deflections. You may note some anomolies
around positions 7-10 and perhaps 30. I don't know if that is an artifact or whether the
wheel actually becomes slightly triangular; maybe that is the bulging you mentioned.
However, you can see a trend to more + deflections at the top of the wheel than the
sides, which is consistent with a radially flexible rim and increased support from the
top.
You would need to know your test weight and intitial strain to calculate whether each
sum is equal to the test load, hence how the load is shared between top and bottom.
Amazingly, the deflection at the bottom is not directly applied load!! It's the
misconception that the rim is supported by air pressure or that the tire pushes on it
that causes a lot of miscommunications here. Why would I think the flattening of the rim
isn't related to the applied load? BECAUSE WHEN I LET THE AIR OUT OF THE TIRE, IT
CHANGES!! <=the profile of deflected rim is different, that is MORE ROUND, even though
the load is the same ... and you'd better try this before you argue about it.
Why does it happen? Putting air in a wheel increases rim circumference, same as in the
sensing element of a pressure gauge or a bare inner tube. The interaction in the wheel
is mediated by the connection at the bead, which under pressure exerts a net radial
outward force, a summing of all radial vectors of stresses tangental to the tire
cross-section. In this case, the rim shrinks about .002" dia with 70-80 PSI let out of
it. We have by inflating it actually made the spokes much tighter! But in the contact
patch, tire config changes and so do those vectors, producing a net loss of the above
radial force. And sure enough, if those vectors were enough to stretch the rim bigger,
relaxing them will locally reduce it: the spokes in the contact patch zone distort the
rim inwards, losing energy that flows away sideways through the rim as a compression
wave at the local speed of sound, storing the energy removed from the shorter spokes
into the others, with a slice of that left over to compress the rim. How about that? The
tire never did `push' on the rim; as a structure specialized in tension it also can't do
that. That flattening was caused by the tire relaxing the local extra radial stress
instead. Something of the same thing happens at the top, due to overall tirewall
flexibility: it gets slightly taller vertically and the radial tire vectors increase,
stretching the upper rim slightly outward. That is probably the reason for that top lobe
in the data. QED, babay.
TBGibb
In article
<Pine.A41.3.96a.98010...@dante11.u.washington.edu>, "F.
Hayashi" <hay...@u.washington.edu> writes:
>So, how about a 'modern' derailer with old friction levers?
>Do old friction levers pull enough cable to move, say a Shimano 105sc rear
>derailer, across 8 cogs with campy spacing?
I don't know where my head was on my first comment but my oldest boy has an old
Raleigh Professional frame with the original Campy friction shifters and and an
Ultegra 8 speed rear derailer. It has all the cable pull it needs. The
cassette is a Shimano but there is still room on the limit screws.
Tom Gibb <TBG...@aol.com>
Rick Denney
I've run NR shifters and rear derailleur on eight-speed cassettes with
no problems at all. It requires a careful touch, but it works.
On Fri, 2 Jan 1998 10:09:20 -0800, "F. Hayashi"
<hay...@u.washington.edu> wrote:
>Will old style friction downtube shifters (like Campy NR shifters) coupled
>to older rear derailers (like old Campy, Simplex, etc) work with an
>8-speed cassette?
>
>I'd like to put together a ep the wheels compatible with my other bike.
>
>Thanks
>
>+-------------------------------------------------------------+
>| Fumitaka Hayashi - hay...@u.washington.edu |
>| http://macrophage.immunol.washington.edu/~fumi/index.html |
>| Aderem Lab - Dept. of Immunology - University of Washington |
>+-------------------------------------------------------------+
>
Rick Denney
Take what you want and leave the rest.
Propeloton
Any slant paralleogram-ie., indexing rear der travels less far per mm of cable
than a friction rear der so friction levers, when on the biggest cog will be
all the way backwards when on the biggest 8 speed cog-Campag did make an 8
speed friction set of shifters that had a bigger barrell on the right side to
aleviate this problem.
Peter
ProPelotom
Boulder
Mark McMaster
Thank you for making measurements of a changes in spoke length on a
loaded wheel. I'm sure you made better measurements than I would be
able to. Interestingly, the pattern of the spoke deflections you
measured are quite similar to the results predicted by the finite
element analysis in "The Bicycle Wheel" by Jobst Brandt.
So let's look at the results of your testing. The first observation
is that there may be transcription error in the table, since there are
measurements listed for spokes in positions 0 through 36, which would
mean a 37 spoke wheel. I assume that it was actually a 36 spoke wheel,
with one of the numbers repeated somewhere.
Looking at the numbers, several things that have been previously
discussed are apparant. Firstly, there is the pronounced "flat" section
at the bottom of the rim, demonstrated by the shortening of the half
dozen or so spokes in that region (I assume that position 0 is bottom
dead center). Also, there is there is an increase in radius around the
rest of the rim, which produced a not perfectly uniform, but otherwise
well distributed increase in spoke length around the rest of the rim. I
think we can go by the assumption that the spokes are linearly elastic,
so that tension changes are proportional to length changes, so I'll use
the measurements of length change interchangeably with (relatative)
tension changes.
In order to determine which spokes support the load on the wheel, let's
look at the change in force that each spoke exerts on the hub, which is
where the bicycle applies it's load to the wheel. When the wheel is
loaded, each spoke exerts a change in force on the hub, but it is only
the vertical component of that force which acts to support the hub (for
example, spokes that are horizontal to the hub, perpendicular to the
load, do not provide support for a vertical load on the hub). The
vertical component of each tension change of each spoke is equal to its
total tension change times the sine of the angle from the horizontal (I
am assuming that position 0 is at the bottom of the wheel, which is at
an angle of -90 deg. from horizontal, with respect to the hub). I put
the numbers in a spreadsheet, and calculated the vertical component of
each spoke tension change:
Pos. angle Deflection Vert. Def.
0 -90 -0.0045 0.004500
1 -80 -0.0025 0.002462
2 -70 -0.0005 0.000470
3 -60 0 0.000000
4 -50 0.0002 -0.000153
5 -40 0.0003 -0.000193
6 -30 0.0003 -0.000150
7 -20 0.0002 -0.000068
8 -10 0.0002 -0.000035
9 0 0.0002 0.000000
10 10 0.0003 0.000052
11 20 0.0004 0.000137
12 30 0.0003 0.000150
13 40 0.0003 0.000193
14 50 0.0004 0.000306
15 60 0.0004 0.000346
16 70 0.0003 0.000282
17 80 0.0003 0.000295
18 90 0.0004 0.000400
19 100 0.0004 0.000394
20 110 0.0005 0.000470
21 120 0.0004 0.000346
22 130 0.0003 0.000230
23 140 0.0003 0.000193
24 150 0.0003 0.000150
25 160 0.0004 0.000137
26 170 0.0003 0.000052
27 180 0.0003 0.000000
28 190 0.0004 -0.000069
29 200 0.0003 -0.000103
30 210 0.0002 -0.000100
31 220 0.0004 -0.000257
32 230 0.0004 -0.000306
33 240 0.0003 -0.000260
34 250 0.0002 -0.000188
35 260 -0.0009 0.000886
36 270 -0.0028 0.002800
Note: Spokes below the hub which had increases in tension exert a
negative vertical force change on the hub. Spokes above the hub which
had increases in tension and spokes below the hub with decreases in
tension exert positive vertical force change on the hub. Spokes running
horizontally from the hub (positions 9 and 27) exert no vertical force
on the hub.
You have added up the tension increases around the rim. But the change
of radius of the rim does not by itself support the load of the wheel.
Tension increases due to radius increase are distributed around most of
the rim, so many of the tension changes on one spoke are counterbalanced
by tension changes on the opposite spoke, i.e. many forces from spoke
tension increases are cancelled out. To determine how much of the spoke
tension increases actually contribute to supporting the wheel load, we
can sum the vertical components of each tension increase as it acts on
the hub (which will be positive above the hub and negative below the
hub), and any net vertical force is the contribution of tension
increases in supporting the load. Adding up all the vertical forces
(deflections) from spokes that increased tension (positions 4-34) yields
a net upward force of 0.00225" on the hub. Compare this to the sum of
the vertical forces (deflections) from spokes that that were de-tensed
(or "compressed", in my semantic, at positions 0, 1, 2, 35 & 36), and we
find a net upward force of 0.011182" on the hub. This shows that
tension changes that support the hub are primarily the de-tensioned
spokes at the bottom of the wheel, by a ratio of about 5:1 (83.2% of
load carried by decreasing tension in the bottom spokes, 16.8% by
increasing tension in the top spokes).
If in fact there is a transcription error in your data, and you
included an extra spoke tension reading above the wheel (the 37th spoke
mentioned above), than there would be one less spoke in the above
calculations pulling from above, and the contribution of the upper
spokes to supporting the load would be less by that amount (reduced to
about 13-14% of the total).
Of course the distribution and magnitudes of tension changes will be be
affected by relative changes in the stiffness of the rim and the number
and stiffnesses of the spokes. On a very stiff rim with few spokes, the
ratio of load carried by spokes on the bottom vs. the top load will be
lower (but never less than 1:1 even with a perfectly rigid rim), and
with a flexible rim with many spokes, the ratio will be higher.
Mark McMaster
MMc...@ix.netcom.com
Hoyt McKagen
Mark McMaster wrote:
>
> able to. Interestingly, the pattern of the spoke deflections you
> measured are quite similar to the results predicted by the finite
> element analysis in "The Bicycle Wheel" by Jobst Brandt.
I never disputed that, and in fact most of the difference we have is in terminology, but
the result of that completely obscures what is really going on.
> mean a 37 spoke wheel. I assume that it was actually a 36 spoke wheel,
> with one of the numbers repeated somewhere.
Somone else mentioned that too. Toss any number not near the zero point and the result
is about the same.
> where the bicycle applies it's load to the wheel. When the wheel is
> loaded, each spoke exerts a change in force on the hub, but it is only
> the vertical component of that force which acts to support the hub (for
> example, spokes that are horizontal to the hub, perpendicular to the
> load, do not provide support for a vertical load on the hub).
I stopped with the first mistake. Someone else said that too. The problem with it is
this: whatever loaded equally all those spokes outside the contact patch of that wheel
DIDN'T use some function of the angle going either way; it loaded and unloaded them
totally without regard for that ... so you can't use the function of the angle either!
<good hard-cal'c #s snipped for wrong premises>
<wrong conclusions snipped>
See ya,
Mark McMaster
Hoyt McKagen wrote:
>
> Mark McMaster wrote:
> >
> > able to. Interestingly, the pattern of the spoke deflections you
> > measured are quite similar to the results predicted by the finite
> > element analysis in "The Bicycle Wheel" by Jobst Brandt.
>
> the result of that completely obscures what is really going on.
>
> > with one of the numbers repeated somewhere.
>
> is about the same.
>
> > loaded, each spoke exerts a change in force on the hub, but it is only
> > the vertical component of that force which acts to support the hub (for
> > example, spokes that are horizontal to the hub, perpendicular to the
> > load, do not provide support for a vertical load on the hub).
>
> this: whatever loaded equally all those spokes outside the contact patch of that wheel
> DIDN'T use some function of the angle going either way; it loaded and unloaded them
> totally without regard for that ... so you can't use the function of the angle either!
The increased radius of the rim outside of the flattened zone is mostly
just a side affect of the flattening. This is why the spokes increase
in tension uniformly (symmetrically) around the rim. To know what
affect any spoke with a change in tension has on supporting a vertical
load, you _must_ know the vertical component of that tension change, not
just the magnitude of the tension change. If you believe that a tension
increase _anywhere_ on the wheel can support a vertical load, than you
are saying that the spokes that are horizontal to the ground also act to
support a vertical load, which would mean that a spoke can support a
load that is perpendicular to its axis. It seems odd that you believe a
spoke can not support a load in compression, and yet believe that it can
support a perpendicular load.
As shown by your measurements, the increase in rim radius will increase
spoke length (and tension) outside the flattened zone quite uniformly;
and you claim that all these tension magnitudes can be simply summed to
determine their part in supporting a vertical load. But what if I took
an unloaded wheel, and simply inserted a wedge or other spacer into rim
at the joint to increase its circumference. Surely this too would
result in a uniform increase in radius and uniform increase in tension
through out all the spokes; yet the wheel is supporting no vertical
load. So we can see that simple adding all the spoke tension increases
does not give a measure of an external load on the wheel. You mention
yourself that inflating a tire on a rim changes the rim diameter (and
subsequently changes spoke tension), even though the wheel is supporting
no load.
From this we know that if tension is changed uniformly around the rim,
the wheel will still be in equalibrium (no net loads). By applying a
(vertical) load to the wheel, we disrupt that equilibrium only in the
small flattened zone. Since this zone is small, the rest of the rim is
substantially in equilibrium. However, we can see that the tension
changes in the flattened zone are quite large, and vastly surpass the
small change of equilibrium in the rest of the wheel. It is the
relatively large changes in tension in the flattened zone that do the
job of the supporting the load.
> <good hard-cal'c #s snipped for wrong premises>
Correct premise was used for table, to find the vertical components of
spoke tension changes (the only part that can support a vertical load).
> <wrong conclusions snipped>
Correct conclusions, just not agreeing with your pre-conceptions of
load mechanics.
After looking at a set of hard data on spoke tension changes, and
having discussed the concepts of relative stiffnesses of the rim and
spokes, compressive loading of pre-tensioned members, the line of action
of load vectors and vector components, and force equilibrium within the
wheel, I don't see how I can be any clearer in explaining spoked wheel
mechanics. If there are some concepts that you still do not grasp, well
I've tried my best.
Mark McMaster
MMc...@ix.netcom.com
Hoyt McKagen
Mark McMaster wrote:
>
> The increased radius of the rim outside of the flattened zone is mostly
> just a side affect of the flattening. This is why the spokes increase
> in tension uniformly (symmetrically) around the rim.
It is purely such, IMHO. The length of that zone must reflect rim stiffness, but at some
point with current rims the effect is unlikely to go very far around the wheel. The rest
of the wheel seeks an equilibrium position like water finding a surface level.
To know what
> affect any spoke with a change in tension has on supporting a vertical
> load, you _must_ know the vertical component of that tension change, not
> just the magnitude of the tension change.
That would be true IF this were a rigid structure; as it happens none of us believe that
and it therefore makes me wonder why you are referring to it as if it must be.
If you believe that a tension
> increase _anywhere_ on the wheel can support a vertical load, than you
> are saying that the spokes that are horizontal to the ground also act to
> support a vertical load,
Yes, by having their loads on the rim translated by its own mechanics. Some of each
spoke's effect is felt by the whole wheel, or it wouldn't go out of round when one
breaks.
>It seems odd that you believe a
> spoke can not support a load in compression, and yet believe that it can
> support a perpendicular load.
I'm only pointing to the results, which show a load can produce the same tension changes
w/o regard for the spoke position. That implies a mechanism which is not sensitive to
position, and that mechanism must operate both directions in the same manner.
As I mentioned earlier today, you should consider the horizontal components of those
spoke loads nearest the contact patch against the local rim geometry. IOW, if you draw a
series of isoceles scalene triangles with the obtuse vertices downward at the center of
tire load application, and the remaining acute angles at the ends of successively
wider-spread pairs of spokes (local to the distorted zone), will the horizontal
components, times the appropriate function of angle for each triangle, add up to the
applied load? My answer would be yes. The big surprise in that is that the spokes
nearest the load zone support the wheel by pulling almost perfectly in the `wrong'
direction.
>
> As shown by your measurements, the increase in rim radius will increase
> spoke length (and tension) outside the flattened zone quite uniformly;
> and you claim that all these tension magnitudes can be simply summed to
> determine their part in supporting a vertical load.
Not quite. I think I am claiming that they are ~equal to the strains which go the other
way. I have begun to think it isn't necessary for either sum to equal the applied load,
and that does reflect more consideration. In fact, since the wheel distorts overall from
being loaded, some extra energy must be stored in it, and I wouldn't be surprised if the
absolute sum of both total strains (plus, of course, that energy stored in the rim as
additional compression) is greater than applied load, because not only does the energy
from the ones with reduced tension go there, but so must be the work done by loading.
Sorry if my position seems flexible; I only started with the idea that compression and
tension are radically different, hence JB's notions must be improperly based. I hadn't
thought to fully explain my own concepts, but if you had to pin me down on this, I will
stick to what is within here.
> an unloaded wheel, and simply inserted a wedge or other spacer into rim
> at the joint to increase its circumference. Surely this too would
> result in a uniform increase in radius and uniform increase in tension
You would have to measure the CHANGES in relation to load, yes, and not merely in how
you set up the wheel?
> Correct premise was used for table, to find the vertical components of
> spoke tension changes (the only part that can support a vertical load).
Applies to a rigid wheel only.
> If there are some concepts that you still do not grasp, well
> I've tried my best.
Now, Mark, you are evidently a fine person with a lot of knowledge and quite good
communications skills, and even though I have been pretty brisk, I have enjoyed this a
lot; my sincere thanks for your part in it. The problem we are having is merely that we
cannot find a verbal model which conveys all of the details. All that said, the grasp of
concepts or the failure of that still goes both ways.
Hoyt
--
Belfab- http://www.freeyellow.com/members/belfab/belfab.html
Penpals- http://www.freeyellow.com/members/batwings/penpals.html
Best MC Repair- http://www.freeyellow.com/members/batwings/best.html
Camping/Caving- http://www.freeyellow.com/members/batwings/caving.html
=> Josehp Lucas, Prince of Darkness <=
Alan Gabrielli
In article
<Pine.A41.3.96a.98010...@dante01.u.washington.edu>, "F.
Hayashi" <hay...@u.washington.edu> wrote:
> Will old style friction downtube shifters (like Campy NR shifters) coupled
> to older rear derailers (like old Campy, Simplex, etc) work with an
> 8-speed cassette?
>
I couldn't get NR to work well with a 7-speed freewheel, so I'm guessing
the answer is no.
Alan Gabrielli
Visiting Professor of Chemistry
Georgia Tech