Interested in the *engineering* behind frame design.

RE: Interested in the science behind frame design.

so having said that, what external forces do they take into consideration when I bike is tipped over, breaking, accelerating, and maybe going up a dip or landing a jump...? What I mean is, in terms of racing, a stiffer chassis gives them more control over damping and spring rates in the suspension right? So, in turn, a stiffer chassis is better (excluding tortional stiffness)? but where does the balance come in where excessive material is used thus increasing wieght which is ofcourse has a negative effect on breaking, acceleration and tire wear.

Also, I would imagine too much stiffness in the "tortional" when the bike is in high lean angles, would be bad or does it not matter since the forces acting on the bike are normal to the reaction force (friction force + reaction force to gravity)?... how much is too much and how much is not enough? how is the target derived??? Rather confusing :S

In our business we want less than 0.1mm of deflection in the supporting stringers of the conveyor we are going to install our beltweigher in.

This is for 2 main reasons.

1. Our machiens rely on very precisely set roller height in the supported idlers and deflection will fuck with this and create weighing error.

2. Large deflection with a dynamic load will occour some what randomly. This means that the rigid structure which we rely on woudl be moving (minutely) all the time and this leads to an unstable zero in weighing.


I would be very surprised if there was ever a 'to stiff' frame. Think about it. The body position of the rider the and the particular track surface the bike is traveling on cornering or not / angle are going to create impossible to define deflections.

A frame with zero deflection you can use the suspension to tweak the ride of the motorcycle right? Imagine you set the rear shock JUST RIGHT in a high deflection frame, you take a different line the next session out, hit a new bump, the frame deflects differently and your rear shock setting is no longer right, you fly off the track impale yourself on a fans dag wood dog. YOU'RE DEAD. THAT'S RIGHT DEAD ALL BECAUSE YOUR FRAME WAS UNPREDICTABLE.

Large deflection in complicated machines = unrepeatability / unpredictability. The last thing you would want in a motorcycle... or a weighing system for that matter.

So like all things uber tech these days they would be going for the lightest possible way to produce the most rigid frame possible.

Possible the most intelligent thing I have done all day. /profit.

Humbug said:

But isn't deflection predictable? Its just undampend springing isn't it?

Picture this scenario. The bike is at full lean on a corner. The road sudenly jumps in hieght, lets say 3cm. As the front wheel runs over it, the tyre will deform, then slowly lift the wheel up towards 3cm directly vertical. the front fork I would imagine would compress abit causeing the front wheel to change line of track... as its travel is not perpandicular with the road.

at the same time it will then have to lift the head stem verticaly 3cm, while the rear wheel is still 3cm lower. so that should introduce a tourque on the frame wouldn't it? And that would be un damped wouldn't it?

I wonder if i was able to explain that right? :S

Click to expand...

In a perfect world the shock would fully absorb the bump and the head stem would remain static I think. There would be serious torque forces on the frame constantly. Which is why they would deisng them for as little deflection as possible.

Deflection in a 50x100x3mm SHS beam is predictable. You have it bolted here and put X weight there and it will deflect the same amount each time.

When you have the amount of physics and complexity involved in a motorcycle frame they need to use 3D engineering software to model the whole frame and crunch numbers. They can then work out how much forces from different angles / directions will affect the frame. I very much doubt they build frames to deflect as part of the motorcycles function. Too many variables. If you build something and want it modeled as I've mentioned you are looking at paying a company that does it. It will cost tens of thousands and months. SRS BZNS

Humbug said:

how can the forces be irrelivent if stiffness and overal wieght contribute to overall handling? If there are zero forces. there is zero deflection right? ofcourse, if the forces on the road are too great, you would have viabration comming from the road, through the chassis, causing traction loss from lack of dampening etc right?

so in the design of a frame, lets say a crash, it should be able to protect more delicate parts like the engine/petrol tank etc right?

therefor street frames are probably overkill in size strength and arnt required for a nice smooth race track?

I have no idea what I am on about :S

Click to expand...

Other way around, because you're not concerned with absorbing bumps etc so much on a race track (no speed humps and rider doesn't give a shit about comfort) you can make the frame stiffer, which makes the bike handle better, but also means that the rider has to be better to catch it if anything does go wrong.

Also, nobody is going to design a bike with crash in mind, the reason that the frame is out so far as possible is to maximise torsional rigidity with minimum material, additionally, frame members are very deep to prevent a longitudinal bending mode and the reason theres a big space in the middle and relatively (left to right) thin open structure is to make sure that there is lateral compliance so it doesn't put you in orbit the first time you hit a bump on your side.

Oh, and for your example of the bike on its side earlier, yes, thats pretty much the point, but you want to avoid torsion because the stiffness isn't constant because of suspension travel, as the suspension travels upward the moment arm that it applies to the bike chassis shortens resulting in a higher torsional stiffness. So the best logic is to prevent torsion and allow translation.