From the graph I see that, with a starting speed of about 2550 RPMs, it spins over 66 minutes.
Also, about 1900 RPMs are needed to reach 1 hour.
How fast can you spin it with just one finger whirl?
Do you think that a 1-hour top with 1 finger snap is possible (perhaps made of platinum)?
A very interesting engineering question. Increasing launch speed or rotor density may not be the best approaches.
There are 5 basic but somewhat conflicting lines of attack for any top. For a single-twirl finger top on a pedestal, from launch to fall...
1. Increase launch speed:a. Practice, practice, practice!
b. Optimize stem diameter and taper.
c. Reduce air resistance.
d.
Reduce AMI.
2. Reduce high-speed spin decay rate — mainly but not solely by attacking air resistance:
a.
Increase AMI.
b. Streamline the shape with special attention to colliding flows.
c. Reduce max radius.d. Reduce surface area.e. Reduce surface roughness.
3. Reduce low-speed spin decay rate— mainly but not solely by attacking tip resistance:
a.
Increase AMI.
b. Optimize tip and base materials — possibly at the expense of wear rate.
c. Reduce total weight.
d. Reduce tip radius of curvature.
e. Increase base radius of curvature if concave.
4. Reduce critical speed (remember, mass and density cancel out when uniform):
a. Reduce CM height above contact.
b. Increase AMI
per unit mass.
b. Reduce TMI
per unit mass about the CM.
5. Eliminate precession and especially wobble:a. Practice clean vertical launches at max effort.
b. Eliminate unbalance.
c. Insure that tip contact is exactly on spin axis.
Iacopo already does nearly all these things, and arguably better than anybody. But there are crucial trade-offs to be made with single twirls — especially around AMI, max radius, and critical speed — and they'll have to be played just so to max out single-twirl spin times.
Note the nearly exponential shape of Iacopo's spin decay curve (SDC). At every point, the slope is the deceleration rate, and in quiet sleep, it's proportional to
total resistance and inversely proportional to AMI.
That's the good news about AMI: From launch to fall, the greater the AMI, the less the top slows down for a given total resistance.
The bad news: At some point, AMI starts to limit single-twirl launch speed. I say this from long experience. And the more you struggle against it, the harder it is to get a clean vertical release with no wobble or precession. Practice definitely helps, but only to a point.
Note also that the SDC slopes go from very steep at launch to very shallow at fall. Hence, you add significantly more spin time by shaving an RPM off critical speed than you do by adding an RPM to launch speed, regardless of the air and tip resistances involved.
To me, that makes critical speed reduction (item 4 above) a prime target, as it extends the tail of your SDC. Iacopo already does the most important thing here: Reduce CM height above contact. That leaves size and shape.
It's easy to show that for
any shape with known AMI and TMI formulas and uniform density, you can always scale the absolute size by max radius R and the shape by one or more key proportions. You then find that critical speed is
inversely proportional to sqrt(R). Probably true, at least roughly, for any realistic top shape.
So bigger is better at the arguably more important low-speed tail of the SDC but generally worse at the high-speed head, where the braking is mostly aerodynamic. (I take the von Karman disk as good guidance here.)
So where's the sweet spot in size — considering that bigger also means greater AMI, density and proportions being equal? I think you'd have to find that sweet spot to max out single-twirl spin time, and that would require lots and lots of guess-and-check testing.
Note that critical speed has nothing to do with dissipation. And that absolute AMI and TMI play no role, as mass and density completely cancel out in a uniform flywheel. Only AMI and TMI
per unit mass count here, and these properties depend
only on size and shape, not on rotor density.
All of which leaves me wondering, what combo of mass, size, and mass distribution optimizes a single-twirl top for spin time? Not sure we've seen it yet.