(Question 1 i guess): in term of reading this graph, the reason why out of plane bending (OOPB) is mention for CS is even those they are responsible for less % of the bar for CS compared to torsion, the CS take in more OOPB based on % of total?
Question 2: for CS the OOPB would be lateral - left to right?
Anyway, looks like this article agrees with Fairlight that a horizontally ovalized CS could be the way to go.
About Fairlight, they have gone even more all-out with ovalized tubings in their newest Secan, with not just the top tube but now the whole triangle as well:
āIn the centre of the seat stay, there is now an ovalized section - measuring 11 x 17mm. We also heavily s-bend the stays, which combined with the ovalization promotes flex (displacement), meaning increased comfort, especially over rough terrain.ā
āThe chain stays are extremely wide in the horizontal plane while narrow and flat in the vertical plane. Pedaling forces are horizontal and ground/rider weight forces are vertical so the shaping of the chain stays provides good power transfer but also high levels of comfort; especially when combined with the new flat ovalized seat stays.ā
Eyeballing the graph and bending look like the dominant mode to me. Either way you can conclude both are important.
That my interpretation as well.
This is where I want to set the record straight on ovalized tubes: You donāt magically tune one characteristic without affecting another by squishing a tube.
When you ovalize a tube, it gets stiffer (bending) in the major axis and less stiff in the minor axis. However, you greatly reduce the area moment of inertia (torsional stiffness): (link)
If you look at the ātorsional stiffness constantā of two tubes and their area moments:
19mm round => I = 13,000 mm^4
25x12 oval (19mm tube squished) => I = 7000 mm^4
The ovalized chainstay is half the tosional stiffness. By ovalizing a tube, you made something stiffer in bending but less stiff in torsion: two steps forward, two steps back.
You might get some marginal improvement in a characteristic you want, but you are just as likely to make it worse. You wonāt know unless you back it up with lab testing.
Taking a step further back, the dominant characteristic of the rear triangle is its geometry, not its individual tube shapes. When you cold set a frame, you can visibly see the thin part of the chainstay bending.
Taking the whole truss structure in context, the chainstay dimple area experiences the most force and has the thinnest geometry. That area will dominate the characteristics of the rear triangle.
In the front triangle, there is a better argument for ovalized tubes:
It allows you to have the strength of a larger diameter tube at the joints, with the torsional compliance of a smaller diameter tube
It can allow for more welding space on the BB and HT
Yeah i was trying to figure out if one force is more dominant than others and should be prioritized but thought they were very similar tooā¦
Which leads to this point
I was wondering how much torsion would you sacrifice but didnt know how to calculate, and that seems likeā¦a lot
Great to know!
I wasnāt focusing on this because I thought ovalizing a maintube might be more risky (no scientific backing) but very insightful to know too!
For the top tube, you said ātorsional complianceā instead of being risk-adverse about lack of torsionā¦Is this because the top tube is bigger so itās less of a concern?
The reason why I say torsional compliance is because that should be your motivation for ovalizing a tube: to make it less stiff. Otherwise, you would leave the tube as-is!
The rational should go like this (using arbitrary tube diameters):
I want the strength of a 31.8 toptube at the headtube junction
I still want the torsional compliance of a 25.4 tube
I ovalize the middle of the 31.8 tube to 25.4
Ovalizing a 31.8 tube to 25.4 mimics the torsional stiffness of a 25.4 tube, but has the strength of 31.8 at the headtube joint (the weakest spot). The tradeoff is weight. You could achieve similar results with thicker butts or a gusset.
Some companies argue that the thin top tube helps vertical compliance, which is true to some extent. Look how much this toptube flexes:
However, the vertical/frontal load is an order of magnitude stiffer than the torsion, which you feel all the time. So if you make the toptube more flexible in bending, you inevitably make the frame less torsionally stiff. Nothing comes for free.
The real reason for ovalizing tubes is about aesthetics, which is very important. We all want to ride beautiful machines.
ah ok when you said āIn the front triangle, there is a better argument for ovalized tubesā i thought it would be noticeably more straight forward but there seems to be considerable drawbacks still!
A few years ago when prototyping a new frame model, ovalising the DT (vertically) at the HT joint was the difference between the frame failing vs. passing lab testing in a certain configuration.
So tube shape can definitely make a difference.
Haha now i got the context, really need to click on the user profile before more
And since I got you here, may i also ask if ovalizing a tube would make it more āfragileā or sensitive for brazing? I already only have ~40mm of butts left at each end (no other choice for my short top tube) so Iām a bit on the fence about ovalizingā¦
Do you know what the ovalization accomplished that resulted in the test pass? Iām curious if it allowed better joining or if itās more structurally sound in front end impact or ??
I donāt have specific data for it, but I reckon it has to do with the fact that the tube is stiffer along the larger diameter of the ovalisation. @Daniel_Y mentions diameter vs. stiffness in his āWhy Steel will alwats be Realā YouTube video. Canāt remember if tube shaping (not just diameter and butting) is mentioned in that video.
My thinking would be it allowed for some extra flex on the minor diameter which would have pulled some stress away from the joint and softened the transition between tube and headtube. The round tube would have been much stiffer all the way up to the toe of the weld, concentrating the forces.
The tube was ovalised vertically at the HT end and kept round all the way to the BB.
Maybe Iām misunderstanding what you mean here, but I donāt understand how this would add flex.
So if the tube is ovalised. It is able to flex in the plane of the minor diameter easier than if it was still round. It is also stiffer in the major diameter. With the tube able to flex a bit easier it helps spread the stress,away from the joint. Flex is a great way of distributing loads on a structure but it has limits. I am asduming that the frame failed in a BB side load flex test. If thats the case, a āsofterā DT would help spread that stress that failed on the round ( stiffer) DT. It probably needs diagrams to explain it better.
Ok. Thatās where the misunderstanding is.
It was not the pedaling forces test that failed. It was one of the frontal load ones. Realising the test type was not specified, hence my confusion.
A long time ago I had a frame design fail ISO testing multiple times at the headtube (front load test, not BB load). When the headtube was switched out from an 853 (heat-treated) to a 631 (exactly the same tube just without the heat-treatment), it then started to pass the same test.
Not sure what my point is but it sort of ties in with the idea that a slightly āsofterā tube will help dissipate forces over a bigger area.
Very sound thinking. I initially fell into another variation of the āstiffer is betterā trap and thought what enabled passing of the test is longer (and thus stiffer oval side
Which brings me to this almost mind blowing point
Would never have thought of this since the predominant thinking is geo and construction matters so much more than material, but this is serious food for thoughtā¦Just to confirm, all else the same except just that head tube?
Geo, construction and material are all important. The material does matter and designing to the materials strengths and being careful of its weaknesses is the key.
Making things stiff just concentrates the stress into smaller areas leading to failure. Steves conclusion is on the right track. Itās about managing teh stress to spread it over a bigger area or conversely not building a structure that focusses it.
The structural engineers I work with in my day job more often than not specify lower yield structural steel members as they found that they help manage the stresses better in a lot of cases over the mega strong steels.
