Flywheel 2 Wheels or 1?

So something more like this (?):
image

I could be wrong, but the consistency between shots (in other words, the repeatability of a trajectory) doesn’t depend on the configuration of the flywheel (to an extent) but rather the software running it.

I like to think about problems in the context of physics. So from my limited understanding and generalizations of the problem –– how “straight” the ball will go depends on a few things. Are you referring to the flatness of the arc of the shot? That depends on the exit angle and exit velocity. In general, all trajectories of projectiles only depend on the horizontal force vectors acting on the projectile (for the most part it would be air resistance but that’s negligible). Gravity is the only force acting on the vertical component and it’s constant. This diagram illustrates it better.
image
Most other forces are negligible but unless you have an “umbrella” hood (refer to MJS robot), the main thing controlling the consistency of shots is the software powering the motors and the mechanical reliability (for lack of a better word) of the flywheel.

3 Likes

It should probably be clarified that this is due to the inertia of the added wheel, and depending on the mass distribution of the wheels and the radius, it’s not impossible to have 3 wheels with a smaller MOI than 2 wheels, but of course this would mean that the 3 wheels have a smaller radius/mass distribution than the other 2. Its just a little thing but I wanted to point it out

1 Like

You’re completely right in your analysis of the standard projectile motion of a ball. However, what I was referring to was the side-to-side straightness- not the flatness of the arc or anything like that. Like you said and I said earlier,

The side-to-side “straightness” so to speak of the ball is determined by the force components of the flywheel. In a double flywheel, if one flywheel is going faster than another, it will veer in one direction and not be straight, while in a single flywheel, it’s primarily based on the material consistency of the hood, and any oscillatory motions in the flywheel (those caused by vibrations and an unbalanced wheel, for instance). But these factors are fairly constant (and can be remedied with good build quality and are about as much of a problem in a double flywheel anyways), so it stands to reason that single flywheels are slightly more consistent, but even then with a good enough PID loop you can probably control the speeds of the double flywheel pretty accurately (I doubt you can do this with vex’s meager sensor setup, but in the industry it’s entirely possible if you needed to launch something I suppose). Anyways this is what I was referring to in my comment and what you mentioned about mechanical reliability. as I said earlier none of this is really relevant this season because long range shots are not looking very viable, but it’s still good to know ig

6 Likes

Ah gotcha, that makes more sense. I would agree, but I think there would be even more consequences than just the “straightness” of the shot which only adds to your argument (including more friction and resistance against the motors and the ball when shooting).

You could also mechanically link the flywheels with gears so the motors can’t turn at different rates without fighting against each other. That still wouldn’t make logical sense given the other options. Double flywheels simply aren’t good in the context of Vex.

4 Likes

Not a bad idea, would be like how chain keeps the wheels going the same speed on a drive. Single flywheel would still be more efficient though like you said

1 Like

I don’t mean a double fly wheel I meant single shaft I just think one wheel would be to thin and could give the ball freedom to move side to side

Not really (as far as I’m aware at least). I know the balls this year aren’t perfectly spherical but think of it like a parabola touching an X-axis where the ball is the parabola and the x-axis is the wheel. That’s essentially the amount of contact on the ball. It’s at the very most the width of one wheel. Either way, the friction on the wheel is more than enough to prevent slippage granted you have enough compression.

image

1 Like

A parabola tangent to an axis doesn’t do justice at demonstrating the complexity behind the contact of the ball to the wheel… no offense.
Deformation, whether in the flywheel or the hood, is necessary (a parabola tangent to an axis implies that it connects at a singular infinitesimally small point). If you are trying to apply a non-zero force through a singular, infinitesimally small point, you’re going to break some physics lol (P=F/A is just one equation example of where things would go awry). Anyways, the faster the flywheel is spinning, the more tractive force it needs to move the ball (tractive force is static CoF*Normal force, essentially, but because of the previously mentioned issue with assuming a force applied from a singular point- which is what the application of the static friction coefficient essentially implies- you end up having to deal with some really not so fun math: https://en.m.wikipedia.org/wiki/Frictional_contact_mechanics [Read the section on the solution of rolling contact problems]).

The reason why compliant wheels (and rubber band rollers) are so prevalent this year is because unlike a cube (which, when treated as a rigid body, has infinite contact points along the same plane), a wheel treated as a rigid body touches the rigid ball at exactly one point (because we are assuming there is no viscoelasticity between two rigid bodies). Yet, low viscoelasticity would be actually be ok if the ball was perfectly spherical (it would actually be beneficial to an extent because it would reduce viscoelastic hysteresis, comparable to how train tracks have near perfect steel tracks and steel wheels so they are able to use essentially completely rigid wheels to achieve extremely high efficiency- which is what actually makes them more efficient than cars because cars deal with large amounts of viscoelastic hysteresis (rolling resistance).

Alas, change up balls are not perfectly spherical, and even if they were intended to be perfect spheres with low viscoelasticity, I would not trust manufacturing tolerances to the point that I wouldn’t include some form of additional deformable material between the wheel and the ball (whether on the hood or the wheel itself, unless the ball itself is sufficiently viscoelastic [like in nbn for instance, where viscoelasticity was less necessary because the balls already had it covered], as even trains spray compressed air mixed with sand in front of the wheels when they need a little extra traction when accelerating. So because of this, any viable design this year will use rollers with some kind of deformability (whether they be flaps, compliant wheels, rubber band rollers, etc) which quite frankly, renders a model that simulates a completely rigid contact scenario rather pointless.

7 Likes

Oh I guess I mistated my point. @64811AOverHeat was saying that he thought the flywheel needed more than one wheel for full contact on the ball. I was arguing that when you think of the ball in the context of a round surface contacting a flat surface (hence then parabola even though it’s a MUCH more extreme case) is synonymous with the flywheel situation. Note, this is an extreme example only to highlight the issue. Not that there is only one point of contact.

I’m not disagreeing with you, I think deformations do occur. As far as you presented your stance, I pretty much agree with all of it. I still think my oversimplification was warranted though.

2 Likes

Eh. Newtonian mechanics is oversimplified as is ig. Lagrangian method ftw

1 Like

I think what @64811AOverHeat is trying to say is have one axle, and 2 wheels on that axle. the two wheels are spaced apart to create a groove for the ball to go into. both wheels would contact the ball on opposing sides.

2 Likes

Thank you that’s exactly what I meant

I’ve never seen this done, and I don’t have a lot of experience with flywheels so I can’t say how well this would work, but my intuition tells me not well since the ball would be contacting only the corner of the wheels, and somehow I feel like that would cause issues. might just be me though.

Ohhh bruh :man_facepalming: I’ve seen this done before, I couldn’t say objectively if it works well or not.

I’m going to try something like it but with gears

1 Like