Cables stretch and contract with [air] temperature
I thought it'd be interesting to quantify this, with specific attention to rear mech gear cables.
Cables do indeed stretch with temperature, but (and
@C R has got in with the key 'so what' while I started the morning's maths
)
Cable Pull
The OP has an 8sp drivetrain (Claris) and the (rear gear) cable pull (ie per click) is
2.8mm. (Note that 9sp = 2.5 and 10sp = 2.3 (Shimano))
Thermal Expansion and Contraction
(
https://www.certex.lv/en/technical-...t/properties-of-extension-of-steel-wire-ropes )
The coefficient of linear expansion (∝) of steel wire rope is 0.0000125 = (12.5 x10^-6) per oC
Change in length Δ(mm) = ∝ x L(mm) x 'T'(oC)
where:
∝ = coefficient of linear expansion
1800 = original length of cable (mm)
So ~20 microns (mm^-3) per degree of temperature change (o C)
So an extreme variation, say 20 degrees C, will result in a change of cable length of about
0.4mm
I guess the indexing error tolerance is sufficient to cope with such expansion/contraction due to temperature.
Elastic Stretch
Bicycles cables do not remotely approach their elastic limit so all stretch due to tension is elastic.
Elastic Stretch = (W x L) / (E x A)
· W = Applied Load ( kN )
· L = Cable length ( mm )
· E = Strand Modulus ( kN/mm²)
· A = Area of Cable = (D2 x pi) / 4 (where D= Dia of cable mm)
So for me: Elongation of cable at the RD is
3mm - seems quite a bit, doesn't it? But this aligns with the distance (about 30mm) one has to push the STI lever sideways (1:10 lever ratio) before the force applied (through another 20+mm) pulls the cable shifting the RD and then engaging the next ratchet in the STI mechanism.
BUT. Each shift up the cassette (assumes the same force required on the lever) stretches the gear cable the same amount so the actual (additional) cable pull is still 2.8mm each time, and thus the indexing works.
Assumptions/estimates:
W = 200N so 0.2kN (NB 1:10 ratio of STI lever, assumes a finger pressure of 2kg wt (20N) changing up the cassette)
L = 1800mm (RD cable)
A = 1.13mm^2 (1.2mm dia gear cable)
Typical value for E is (from
https://www.s3i.co.uk/wire-rope-technical.php )
· 1x19 = 107.5 kN/mm² (wire rope with 19 strands)
Finally an 'aside':
"Newly installed cables can seem to elongate, requiring readjustment. While it is generally agreed that inner wires actually stretch very little - if at all - housings and linings may compress slightly, and all parts may generally "settle in". Lightweight assemblies such as those used on bicycles are more susceptible to this phenomenon." (
https://en.wikipedia.org/wiki/Bowden_cable )
From limited experience, I'd say it does.
More so with the "crossover gears".
What experience are you drawing on? Slight indexing misalignment in very hot temperatures?
Why would this make a difference particularly when cross-chaining? (if that's what you mean by "crossover gears")