The importance of frame compliance over rigidity

[Brucey]

.....in a traditional double diamond frame shape, pretty much impossible......

the front part of a double diamond frame is not fully triangulated so it can have an influence on ride feel. By contrast some frame designs (like the Raleigh 'X' frame) are fully triangulated and are noted for their strength and relatively poor comfort.

[slowster]

My very limited understanding of the science/engineering, is that the down tube will deflect. The question is not whether it will deflect, but whether the amount of deflection will be significant, and if it is, to what extent that can be modulated by varying the stiffness in the vertical plane of the top tube.

For the record, the down tube specification is below. Where the down tube meets the head tube it is both vertically ovalised and has a gusset.

853 DZB - 34.9mm - 30/40 opposed oval - 1.0/0.8/0.5/0.8 - with gusset
The tube starts life as 34.9mm round tube but is ovalized at both ends to become 30 x 40mm. The ovals oppose each other; the 40mm horizontal oval at the BB shell adds lateral stiffness, where as the vertical 40mm vertical oval at the headtube resists the braking and ground forces from the most highly stressed area of the bike. The tube has double zonal butting, which means an extra butt at the headtube end for strength. The butt profile is 1.0/0.8/0.5/0.8.

https://fairlightcycles.com/wp-content/ ... s-V1.2.pdf

[Brucey]

the DT, being 34.9mm, is oversize so would be a fair bit stiffer than in a traditional steel frame, to which the ovalisation only adds.

[slowster]

The vertical ovalisation at the head tube tube junction, plus a gusset, suggests that area has to be strengthened and made very stiff to cope with the higher bending forces resulting from the greater deflection of the horizontally ovalised 1" head tube. That suggests to me that the rest of the down tube will deflect significantly, presumably particularly the central 0.5mm section.

To be clear, I am not making any claims for the bike's compliance - I have never ridden one, but I find the concept interesting.

[Brucey]

The vertical ovalisation at the head tube tube junction, plus a gusset, suggests that area has to be strengthened and made very stiff to cope with the higher bending forces resulting from the greater deflection of the horizontally ovalised 1" head tube. That suggests to me that the rest of the down tube will deflect significantly, presumably particularly the central 0.5mm section.

To be clear, I am not making any claims for the bike's compliance - I have never ridden one, but I find the concept interesting.

one of the reasons they might have ovalised that tube is the same reason as caused them to spec. a 1.0/0.8/0.5mm wall thickness tube, ie. that the service stresses on the weld might be too high otherwise.

[531colin]

This is complex, isn't it?
I should use a stiff seat tube to react pedalling loads, but a flexible seat tube to flex under the rider's weight on the saddle over a bump.

I never did solve the "stiff seat tube" conundrum. If you ovalise it at the bottom, its tricky for front mechs. (remember them?); because the point where seat tubes fail is close to where the front mech. sits. Maybe it needs ovalising in the middle to allow it to flex more due to the rider's weight on the saddle?

I should have a lot of seatpost exposed to get maximum leverage and therefore maximum advantage from seat tube/seatpost flex.

Spa's Elan has a sort of "mast" where the seat tube extends above the top tube and seatstays.
The design of the Elan is hugely influenced by the need to get the handlebars at a comfortable height ( for ordinary people, rather than gymnasts) with the restriction on the number of headset spacers imposed by a carbon fork steerer.
I got the bars up by using an extended head tube (extended above the top tube). But too much head tube above the top tube "looks wrong", so the front of the top tube is higher than ideal, and in order to get the standover I wanted, I ended up dropping the back of the top tube as well. I left the seat tube length as it was (giving rise to the "mast") simply because I can't stand the idea of having a cupboard full of un-labelled frames where nothing measures the "size" of the frame.

Whereas I'm pleased that lots of folk (as well as me) think the Elan "rides nicely" , I'm more than a little irritated that i don't really know why it rides nicely. Perhaps the necessity of designing in a tall front end and a dropped top tube has turned into comfort bonus.

[531colin]

My very limited understanding of the science/engineering, is that the down tube will deflect. The question is not whether it will deflect, but whether the amount of deflection will be significant, and if it is, to what extent that can be modulated by varying the stiffness in the vertical plane of the top tube.

For the record, the down tube specification is below. Where the down tube meets the head tube it is both vertically ovalised and has a gusset.

853 DZB - 34.9mm - 30/40 opposed oval - 1.0/0.8/0.5/0.8 - with gusset
The tube starts life as 34.9mm round tube but is ovalized at both ends to become 30 x 40mm. The ovals oppose each other; the 40mm horizontal oval at the BB shell adds lateral stiffness, where as the vertical 40mm vertical oval at the headtube resists the braking and ground forces from the most highly stressed area of the bike. The tube has double zonal butting, which means an extra butt at the headtube end for strength. The butt profile is 1.0/0.8/0.5/0.8.

https://fairlightcycles.com/wp-content/ ... s-V1.2.pdf

Wow. In steel, that is some downtube. I have to admit i struggled working my way through Fairlight's bumpf, but i think they use that downtube on all sizes. Just for comparison, the downtube on my middle size titanium Elan is about that size, and titanium is about one third less stiff than steel.
Again, I'm struggling to remember Brucey's "rule of thumb".....i think it went that the difference in length of any tube is about 20% between the smallest and biggest size of any particular bike, and if you take a tube of length "X" and a tube of length X+20% and load those tubes with the same weight, then the longer tube will deflect twice as far as the shorter tube. (I apologise if I have that wrong.....I can't do the maths, but i have a suspicion that Brucey can.....and i can't "bookmark" stuff in a computer, and Slowster can!)
If thats the case, I really struggle to understand why anybody would use the same diameter tube on ALL SIZES of a bike?
Its not as if we need to source lugs for different tube diameters.
.

[Brucey]

the simple cantilever bend is slightly over +70% deflection for a +20% length increase. In practice, the difference between the smallest and largest frames is more than 20%, so the biggest frames could reasonably be expected to be about half as stiff as the smallest, if they are constructed in the same way.

[slowster]

Wow. In steel, that is some downtube.

The Genesis Croix de Fer has/had a 31.9mm down tube, but with a 31.8mm top tube. Fairlight use a 28.6mm top tube (again ovalised) for its largest frames.

On the subject of the relationship between tube length and deflection, I had not previously thought much about the length of the down tube. Measuring one bike I find it is about 10% longer than the top tube. Interestingly, Dave Moulton reckons that when varying the bottom bracket height, it is its effect in changing in down tube length and chainstay length, and thus their stiffness, that has the most impact.

http://davesbikeblog.squarespace.com/bl ... eight.html

http://davesbikeblog.squarespace.com/bl ... d-now.html

[Orbit531C]

Bearing in mind CJ's comments above about the different steels being comparable in stiffness, I really don't see how a 631 fork is stiffer than an "R" fork.

Bear in mind that I did not claim that it was. I was repeating Dave Yates' comment stating that 853 was too stiff for an audax fork, and that 'R' blades were a better choice. I don't know if he has made any comment comparing them with 631 forks, although I think that 631 was often his favoured/recommended choice for most customers for a frame.

Although I linked to Dave Yates' posts about this on YACF, it is something which he has commented on before in a magazine article which I once read (possibly in Arrivee). The background was that having been a racer/time trialist, he had started to do a lot of audax rides, and he completed a Super Randonneur series (200km, 300km, 400km and 600km) and Paris-Brest-Paris. He made a frame/fork in 853, but found the forks uncomfortably harsh during the SR series rides, and presumably switched them for something else before PBP.

I don't understand how 853 would be stiffer than 'R' blades, or 631, or any other steel alloy, if all other things are equal, given as you say the Young's Modulus is the same for all. And yet if Dave Yates says so based not only on his experience and knowledge as a framebuilder, but also as a rider doing PBP etc., I am not going to suggest he is wrong.

Maybe there is more to it than Young's Modulus. I think Brucey touched on this in this post - viewtopic.php?p=944790#p944790.

Dave Yates' description of his experience of 853 forks is curious. It would be interesting to know if the forks he used before and then went back to were raked to the same 45mm (biggest) rake that 853 blades are factory supplied with. As heat treated steel, 853 like 753 cannot be cold set by the framebuilder. Therein may lie the source of his different experience. Riding audaxes he may have preferred to use mudguards, and a 45mm fork rake is quite close for 'guards...Did he have a different rake on his 631 or R blades that he set himself in his workshop?
My 853 blades are beautifully twangy and flex over the rough lanes, but as they are brazed into a 11/8" steerer, a-b comparisons can't be made with the other (1") forks I ride.

[brumster]

The Reynolds fork on my double marathon framed Thorn tandem visibly flex on less than smooth road surfaces.

[531colin]

The Genesis Croix de Fer has/had a 31.9mm down tube, but with a 31.8mm top tube............

Is that for all sizes?

Spa D'tour has 5 sizes, top tube is 25.4mm in the smallest, 31.8mm in the biggest
down tube 28.6 in the smallest 34.9 in the biggest.
middle size (54cm) has 28.6 top tube and 31.8 down tube.

I'm not saying I'm right, but I am staggered at the differences between the different bikes, and i don't understand why people don't vary the tube diameters much more.....the smallest riders are getting the stiffest frames.....by far!
D'tour is designed for touring and carrying 4 panniers. Only the smallest D'tour uses the traditional inch top tube inch and eighth down tube as was used for lugged frames

[Brucey]

if the down tube contributes the bulk of the stiffness in a given design, the variation in stiffness between sizes might be a bit less than expected.This is because the down tube length varies relatively little between frame sizes.

[UpWrong]

I am going to say where I think the "comfort" comes from in a traditional steel frame; this is a frame with a one inch top tube and steerer, and inch and eighth seat and down tubes. (the steerer is thicker wall than the top tube)

When the front wheel goes over a bump, the fork, the steerer, and the front of the top and down tubes flex, so that the frame "nods its head".

A lot of people observe that an old-fashioned curved fork in an old-fashioned frame flexes more than a straight fork.
Theres nothing wrong with the observation, but theres everything wrong with attributing the flexing to what you can see, the curve in the fork. Look at my picture in a previous post. If the fork is to flex at its curve, the lever is only the fork offset, so between 40 and 60 mm. In fact, the fork flexes fore and aft at the tip, like a pendulum, and the head tube nods, because the lever is somewhere around 350mm.

If the front wheel, acting through the lever of the fork, can flex the frame's main triangle, then the back wheel can flex the rear triangle, and possibly the frame as well.

My bold. Thorn ovalised the seat stays to facilitate flex of the rear triangle in the vertical plane on their Audax Mark 3. Makes sense but others don't seem to have copied the idea.

[531colin]

the simple cantilever bend is slightly over +70% deflection for a +20% length increase. In practice, the difference between the smallest and largest frames is more than 20%, so the biggest frames could reasonably be expected to be about half as stiff as the smallest, if they are constructed in the same way.

if the down tube contributes the bulk of the stiffness in a given design, the variation in stiffness between sizes might be a bit less than expected.This is because the down tube length varies relatively little between frame sizes.

Given a choice of the biggest frames being half as stiff as the smallest or the biggest frames being more than half as stiff as the smallest, then I guess more than half as stiff is an improvement.

But I would expect the largest frames to carry a larger, heavier rider than the smallest frames.
So isn't it a reasonable design objective to make the biggest frames stiffer than the smallest, rather than less stiff?