On 3/6/2022 2:21 PM, AMuzi wrote:
> On 3/6/2022 11:53 AM, Frank Krygowski wrote:
>> On 3/6/2022 10:58 AM, AMuzi wrote:
>>> On 3/5/2022 8:45 PM, Frank Krygowski wrote:
>>>> On 3/5/2022 2:25 PM, AMuzi wrote:
>>>>> On 3/5/2022 12:25 PM, Frank Krygowski wrote:
>>>>>> On 3/5/2022 11:02 AM, AMuzi wrote:
>>>>>>> On 3/5/2022 9:11 AM, Dieter Britz wrote:
>>>>>>>> On Sat, 05 Mar 2022 02:50:18 -0800, Andre Jute wrote:
>>>>>>>> [...]
>>>>>>>>> My own bikes are all set up to understeer. The
>>>>>>>>> consequential
>>>>>>>>> predictability and stability of line almost regardless
>>>>>>>>> of road surface
>>>>>>>>> makes the  essential difference on downhill
>>>>>>>>> descents at
>>>>>>>>> speed on bad
>>>>>>>>> roads when I'm leaving the road racers behind with
>>>>>>>>> white
>>>>>>>>> brackets of
>>>>>>>>> fear beside their lips.
>>>>>>>> [...]
>>>>>>>>> Andre Jute Ah, them were the days!
>>>>>>>>
>>>>>>>> How do you set a bike up for understeer? I don't know
>>>>>>>> what under-
>>>>>>>> or oversteer means.
>>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>> Standard front geometry should be neutral[1] and
>>>>>>> designers
>>>>>>> know the system well.
>>>>>>> https://dclxvi.org/chunk/tech/trail/image/trail.jpg
>>>>>>>
>>>>>>> In that case, when turning the wheel side to side, the
>>>>>>> frame will neither rise nor fall.
>>>>>>>
>>>>>>> In an understeer design like this:
>>>>>>> http://www.yellowjersey.org/gcdl1.html
>>>>>>>
>>>>>>> The frame rises when the handlebar is turned, which is to
>>>>>>> say the bicycle travel in a straight path unless coerced
>>>>>>> otherwise. This may be desired for unpaved roads and/or
>>>>>>> heavy cargo loads. Put another way, the
>>>>>>> rider/bicycle/cargo weight must be lifted to turn the
>>>>>>> handlebar.
>>>>>>>
>>>>>>> The inverse, oversteer, makes a bicycle less stable
>>>>>>> (='more responsive') and rider effort is needed to
>>>>>>> keep it
>>>>>>> in a straight path. When the fork turns, the frame falls:
>>>>>>> http://1.bp.blogspot.com/-WJp92Vd-dMI/TtlTUAlSArI/AAAAAAAADLE/WmujhKVkXGg/s1600/sanrensho2.jpg
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>
>>>>>>
>>>>>> Sorry, I think "the frame will neither rise nor fall" is
>>>>>> incorrect. If the front end geometry generates any "trail"
>>>>>> at all, the frame must fall.
>>>>>>
>>>>>> Look at http://www.yellowjersey.org/gcdl1.html again.
>>>>>> Draw a
>>>>>> line representing the steering axis down to the ground.
>>>>>> Next, draw a line from the tire contact point directly to
>>>>>> (and perpendicular to) the steering axis. I suppose we
>>>>>> could
>>>>>> call that the 'lever arm' of the contact point.
>>>>>>
>>>>>> That 'lever arm' slants upwards from the contact point to
>>>>>> the steering axis. If you 'swing' it to the side by
>>>>>> turning
>>>>>> the handlebars, it rises relative to the bike. IOW, if you
>>>>>> clamped the top tube in a fixed position and height, the
>>>>>> 'swing' would cause the contact point to rise. Or,
>>>>>> switching
>>>>>> reference frames, if you leave the tire normally on the
>>>>>> ground, the frame will fall.
>>>>>>
>>>>>> I don't have a bike with a head angle as slack as the
>>>>>> black
>>>>>> one in the photo, but I just measured my touring bike. As
>>>>>> with every other bike I've checked, turning the bars
>>>>>> causes
>>>>>> the frame to drop a bit.
>>>>>>
>>>>>> See "Wheel Flop" at
>>>>>> https://cyclingtips.com/2018/11/the-geometry-of-bike-handling-its-all-about-the-steering/
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>> "Wheel flop is similar to trail in that it is
>>>>>> determined by
>>>>>> the combination of head tube angle and fork rake,
>>>>>> however it
>>>>>> is concerned with how the position of the front axle
>>>>>> changes
>>>>>> as the handlebars are turned. In almost all instances, the
>>>>>> height of the front axle is lowered when this happens"
>>>>>>
>>>>>
>>>>> Experiment report:
>>>>>
>>>>> Using a four-foot aluminum level I put one end on a table
>>>>> and the other end on the front of the top tube of a medium
>>>>> (56cm) Gunnar Road with Michelin 23 tires. Turning the
>>>>> handlebar, I observe the bubble doesn't move.
>>>>
>>>> OK, help me understand.
>>>>
>>>> Again, ISTM that if we marked the tire-to road contact point
>>>> with chalk, then rotated the bars 360 degrees, that chalked
>>>> point would describe a circle.  The circle would be
>>>> inclined
>>>> to horizontal by (90 - head tube angle), with its high point
>>>> directly forward. The marked point on the tire would be
>>>> farthest below the head tube's height with the bars straight
>>>> ahead. It would get closer to the head tube (or IOW the head
>>>> tube would drop) at any other steering angle.
>>>>
>>>> Is there some other factor I'm not visualizing that would
>>>> apply a contrary action, to cancel out this effect? So far I
>>>> can't think of one.
>>>>
>>>> I checked my Cannondale again. I held a meter stick
>>>> vertically, using my fingers to pinch it to the top tube. I
>>>> can easily feel the relative motion. At a 45 degree steering
>>>> angle (admittedly, used only for balancing at super slow
>>>> speeds) it seems the frame drops between one and tow
>>>> millimeters. At lesser steering angles the motion is almost
>>>> imperceptible, but as I visualize the geometry, it seems it
>>>> must be there.
>>>>
>>>> One thing I just noticed: The tire contact point actually
>>>> changes as the steering angle increases. Judging by the
>>>> spokes' position, at 45 degrees steering, the contact point
>>>> has moved forward about 10 degrees. Does this somehow affect
>>>> things?
>>>>
>>>
>>> On a 360-degree fork turn I have no idea but you're
>>> probably right. For normal range, as you note 45 deg left
>>> or right, any height change is between zero and negligible.
>>
>> I just did more geometry work. For now, I'm treating the
>> front wheel as a pure circle, i.e. a disc of zero thickness,
>> or a tire of zero width. (I'm not sure what difference tire
>> width will make, if any.)
>>
>> You're certainly right that the height change is small, at
>> least for small steering angles. The obvious limit is zero
>> height change for zero steering angle. But ISTM that as
>> steering angle increases from zero, there must be _some_
>> drop in frame height. Here's some of the geometry.
>>
>> The contact point of the tire on the road occurs at the back
>> of the "trail," and the trail value is often easy to look up
>> for a given bike. Alternately it's easy to calculate from
>> head angle H and fork offset.
>>
>> Let's call the trail value T.
>>
>> Spinning the fork all the way around makes the contact point
>> trace a circle. Its radius is R=T*cos(90-H)  and that circle
>> tilts upward 90-H degrees. That upward tilt is the basis of
>> what I've been saying.
>>
>> Steering angle S causes the contact point to swing a bit
>> upward along that circle. I'll give details if desired, but
>> using two projected views and some fairly simple trig, the
>> change in height as that contact point swings up (i.e.
>> closer to the top tube) is the amount the frame drops. Call
>> it D.
>>
>> I get D = R * (1-cos (S)) * sin(90-H)
>> or
>> D = T * cos(90-H) * (1-cos (S)) * sin(90-H)
>>
>> As I see it, for any conventional bike this is going to
>> cause the frame height to drop just a bit. It's not much,
>> because the circle tilts up only about 17 or 18 degrees, and
>> for most riding S is small. But it must be there.
>>
>> Plugging in an extreme steering angle of 45 degrees with a
>> 73 degree head angle and 60 mm trail, this gives a drop of
>> 4.91mm. Plugging in the same values but with a more typical
>> 10 degree steering angle gives a drop of 0.25mm which is
>> small indeed, but not zero. (The controlling term is (1 -
>> cos(S)), which is very small for normal steering angles.)
>>
>> BTW, I just held the meter stick against the top tube of my
>> ancient Raleigh commuter/grocery bike. It too exhibited a
>> drop as I turned the handlebars. It's almost undetectable
>> for small steering angles, but it's there. It doesn't seem
>> to be as large as 5mm, but then I don't know the bike's head
>> angle or trail. (Yet.)
>>
>>
>
> That's all well considered and well written.
>
> Generally if one can readily observe rise or fall through 90 degrees,(45
> each side) something is wrong/ damaged.
>
> Outliers such as classic roadsters are obvious in their rise without
> measuring tools. The effect on handling is apparent and dramatic
> immediately to riders inexperienced with the design.
>
Well, I don't have a roadster to measure, but I still don't see how a
rise is possible.

I couldn't find details on roadster geometry, so I scaled this image:
https://georgehahn.com/the-raleigh-dl-1-my-dream-bike/
I get a head tube angle of about 66 degrees and a trail about 107mm.

Putting it through my equations above, that yields a negligible drop of
0.6mm at at steering angle of ten degrees, and a large drop of 11.6mm at
a steering angle of 45 degrees. The equations and geometry don't predict
a rise.

If I've made a mistake, I'm curious what it is.

BTW, I tried for a bit to measure the drops using a dial indicator
clamped to my saw table. It's devilishly hard. Too many degrees of
freedom, and it requires constantly fiddling with the bike to seek the
top of the top tube. Pinching a vertical meter stick to the side of the
top tube seems the most reliable way to get at least a sense of the
vertical motion.

--
- Frank Krygowski