Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[Iacopo]

Best spin times
Top A (no fairings, on lens) .................. 185 s
Top B (lower fairing only, on lens) ......... 188 s
Top C (upper fairing only, on lens) ......... 203 s
Top D (both fairings, on lens) ................ 333 s
Top D (both fairings, off lens) ................ 290 s

As I expected, the top without fairings has the worst time. 

But there are also results which I didn't expect in this experiment.
One fairing should be sufficient to cut the Von Karman flow and improve the spin time significantly, but in this your experiment both the fairings were necessary. Why ?

I suspect that the reason is the large spokes.  These spokes work like a fan and they probably have a large air drag relatively to the flywheel. 

The fairing attached below the top, (top B), seales the bore, reducing the Von Karman flow under the top, (how much is the clearance between the flywheel and the table when the top spins on the lens ?  With too much clearance the ground effect is little.)
But the spokes are still exposed to the air. 

The fairing above the top, (top C), isolates the spokes from the air above, still this could be not sufficient to interrupt the pump mechanism, because the air can enter below the top, then it can rise passing through the bore of the top, until finding the spokes, and at this point the air is ejected sideways by the spokes working like a fan.  The fairing in fact is not in contact with the flywheel, there are openings between them, and the air can be ejected through these openings.

Maybe for these reasons both the fairings are necessary for to reduce efficiently the pump effect;
with both the fairings, the spokes are more efficiently isolated from the outside, and, as for air drag, they become like non existent.

Was the cake good, at least ? 

[Jeremy McCreary]

Was the cake good, at least ? 

Chantilly cake -- yum! (In a PM to Iacopo about this experiment, I mentioned that the fairings were cut from supermarket cake and pie containers.)

But there are also results which I didn't expect in this experiment.

That makes 2 of us, though perhaps for different reasons.

These spokes work like a fan and they probably have a large air drag relatively to the flywheel.

Absolutely! We already have strong experimental evidence that the fully exposed spokes in Tops A and B were major sources of air resistance. Also agree that hiding them even partially from the surrounding air under an upper fairing had much to do with the significantly longer spin times turned in by Tops C and D.

One fairing should be sufficient to cut the Von Karman flow and improve the spin time significantly, but in this your experiment both the fairings were necessary. Why ? ... how much is the clearance between the flywheel and the table when the top spins on the lens ?  With too much clearance the ground effect is little.

Potential roles of von Karman-like flows: I try keep this diagram of a pure von Karman swirling flow in mind when imagining the flow around disk-like tops. Remember, this flow pattern's generated by a solid disk spinning in otherwise still air.



But the flywheel here wasn't even close to a solid disk, as its inner/outer radius ratio was 85%. Guessing therefore that the flow around the flywheel in Top A was mostly not von Karman-like. Perhaps not in Top B, either.

Was there still some centrifugal pumping action with no fairings? Surely, but likely with a very different geometry — perhaps a cross between your 2 recent diagrams below...



Hence, covering one flywheel face with a flat fairing made the flow over that face much more von Karman-like, not less. Moreover, when such a flow was present, there was nothing on that side of the flywheel to disrupt it. Also, unlike the von Karman case, edge-effects were likely important given the flywheel's (axial length)/(outer radius) ratio of 29%.

Potential ground effect: First, please remind me of the experimental evidence we have for a pure ground effect -- i.e., a clear-cut aerodynamic benefit obtained solely with a small ground clearance (no upper or lateral shrouds or fairings involved). Link to the post?

At release and critical speeds of ~1,000 and ~100 RPM, respectively, all of my test tops had Reynolds numbers of 5.0e4 and 5.3e3, resp. Hence, all operated largely in the laminar regime (save for any turbulence generated by the spokes and inner gear teeth). Fair to say, then, that the exposed side of each fairing had (laminar) boundary layer thicknesses of 2.0 or 6.3 mm, resp.

But when sleeping on a flat surface, Top D's lower fairing had a uniform ground clearance of 12 mm -- at least twice its underside boundary layer thickness at any speed. Aerodynamically speaking, then, Top D would not even have felt the table's presence in my experiment.

In the engineering of disk-like rotors inside tightly fitting shrouds, all it takes to keep the shroud from slowing down the disk is an air gap thicker than the disk's boundary layer at operating speed. Not sure what opportunity this leaves for a beneficial ground effect under a top in the laminar regime.

That said, will try to test a Top D variant on a flat surface with a ground clearance under 6 mm.

[Iacopo]

please remind me of the experimental evidence we have for a pure ground effect

Look at replies #5 and #8 in this thread.
I had the best ground effect with clearances from 1 to 5 mm.  Your top is larger and moves more air so in your case the optimal clearance could be larger, but I am not sure.
I believe that the ground effect works because, for the air to be ejected outside from the bottom of the spinning flywheel, first the air has to reach the bottom of the top, but the only possible entrance for the air is the clearance between the flywheel and the ground, so in the same clearance there are two flows of air in opposite directions, which obviously disturb each other, especially with a narrow clearance, this is what weakens the Von Karman flow and reduces the air drag. 

In the engineering of disk-like rotors inside tightly fitting shrouds, all it takes to keep the shroud from slowing down the disk is an air gap thicker than the disk's boundary layer at operating speed. Not sure what opportunity this leaves for a beneficial ground effect under a top in the laminar regime.

A shroud surrounding the whole top is another way to take care of the Von Karman flows, which is more effective probably because it works with the flows both above and below the top.

We already have strong experimental evidence that the fully exposed spokes in Tops A and B were major sources of air resistance.

covering one flywheel face with a flat fairing made the flow over that face much more von Karman-like, not less[/i].

I agree that your space station top is something more complicated.
I would say that the pump effect and the associated air drag come from

- The Von Karman flow around the flywheel, and/or the fairings.
- And, above all, the spokes, which work like a fan.

Even if the fairings increase the Von Karman flows, on the other hand they isolate the spokes, and in the overall costs/benefits balance
there is obviously advantage because your top spins much longer in this way.

This is how the air could move in you top C, (one fairing above the top).
The spokes eject the air outside and they can do so because there is an inlet for this air, the clearance under the flywheel and the bore in the top; this air is sucked from the spokes/fan and ejected sideways. 
So one fairing on the top is not sufficient to break the pump effect of the spokes, and top C isn't much better than top A, (no fairings).



But when the second fairing is applied, (top D), the bore in the top is sealed from below, and there are no more air inlets for the spokes, which can't work anymore. The big pump effect of the spokes is killed. The Von Karman flows instead are increased because of the fairings but in the whole the pump effect and the air drag are much lower and the spin time quite longer.

At this point, finding the optimal clearance between the flywheel and the ground could weaken the Von Karman flow under the top, with further reduction of the air drag.
 

[ortwin]

Once my "Easy Listening" suspension top is running stable, with reproducible results, balanced and with reliable RPM readings available, I hope to  make some measurements with/without fairings that contribute towards the questions that come up with the topic here. I think it is suited well for this task.

[Iacopo]

I hope to  make some measurements with/without fairings that contribute towards the questions that come up with the topic here. I think it is suited well for this task.

I am sure it will be interesting.

[Jeremy McCreary]

@Iacopo: Thanks for pointing me to your ground effect experiments. You clearly demonstrated a benefit for the 2 tops tested, especially with a 3 mm air gap beneath their rotors.

Q1: Which top did you test in Reply #5 and what's its diameter and metal?

Meanwhile, from your Reply #5...



OMG, look at that Simonelli top collection in the background!

Q2: Would you be willing to give us a video tour when you have the time? Pretty please?

[Iacopo]

Q1: Which top did you test in Reply #5 and what's its diameter and metal?
Q2: Would you be willing to give us a video tour when you have the time? Pretty please?

It is the Nr. 15, lead, 52 mm.

I made a little stand for my tops, I keep them on the table in my bedroom. 
I am drawing my new top in this moment.  I had practically so little free time during the last three months, but now I will have the next one or two months free for my tops, a breath of fresh air !



Each one of these tops can be seen in my YouTube channel, so a photo should be enough here;
from left to right, and from below to above, they are the Nr. 6, 9, 18 in the white egg, (first row), 8, 10, 12, 13, 14, 15,(second row), 20, 38, 29, 30, (third row).  All the other metal tops have been sold.  The eggs and the spheres are materials I could use for new tops, there are two nice malachite eggs which I will certainly use, sooner or later.

[Jeremy McCreary]

It is the Nr. 15, lead, 52 mm.
I am drawing my new top in this moment.  I had practically so little free time during the last three months, but now I will have the next one or two months free for my tops, a breath of fresh air !

Thanks! A very happy turn of events -- and one that can only work to our benefit. Excuse me while I drool over the photo some more...

[ta0]

The website on your screen looks familiar . . . 

Yeah, nice collection of tops! 

[Jeremy McCreary]

... will try to test a Top D variant on a flat surface with a ground clearance under 6 mm.

Looks like that's gonna take a while. For now, a much easier follow-up on my last experiment (Reply #117)...

Top E: Top A (no fairings) with hub shortened to reduce flywheel ground clearance.

All changes, Top A to Top E
1. Flywheel ground clearance (G): 22 to 13 mm on lens, 12 to 3 mm off
2. Scrape angle (θ_max): 15° to 9° on lens, 8° to 2° off
3. CM-contact distance (H): 25 to 16 mm
4. Total mass (M): 137 to 134 g
5. TMI at tip (I_1t): 4.5e-4 to 4.0e-4 kg m²
6. Measured critical speed (ω_C): 95 to 62 RPM

Unchanged
1. Flywheel, spoke, stem, and tip assemblies, max radius 84 mm
2. No fairings
3. Release speed (ω₀): 1,010±10 RPM
4. AMI (I₃): 7.1e-4 kg m²

Best times on lens, from 1,010±10 RPM to first audible scrape
Top A, H = 25 mm, G = 22 mm .................. 185 s
Top E, H = 16 mm, G = 13 mm .................. 254 s*

* Time for Top E includes a 1 s credit to compensate for its much smaller scrape angle.

Top E's impressive 37% spin-time gain here has little to do with aerodynamics. In this comparo, CM height H went from Top A = 25 mm in to Top E = 16 mm. Measured critical speed changed accordingly, from ~95 RPM to ~62 RPM, resp. And it takes Top E a full 60 s to cover this 33 s critical speed gap. So on the lens, aerodynamic effects could have contributed no more than 9 s of Top E's 69 s of added spin time.

Now for a purely aerodynamic difference: Top E on and off the lens, always stopping the clock at 100 RPM to eliminate the scrape-angle difference...

Best Top E times from 1,010±10 RPM to 100 RPM
Top E, G = 13 mm .................. 185 s
Top E, G =   3 mm .................. 183 s

Since there were no offsetting factors to explain this null result, it's fair to say that reducing G from 13 to 3 mm in Top E had no beneficial aerodynamic effect.


Ground effects: In airplanes, a beneficial aerodynamic "ground effect" reduces airspeed and thrust requirements at altitudes under 1 wingspan or so. Helicopters enjoy a beneficial ground effect all their own, and race cars use theirs to improve cornering by adding an aerodynamic downforce on their tires.

The underlying mechanisms in these 3 cases are completely different, and none apply to spinning tops. But in Replies #5 and #8 of this thread, Iacopo clearly demonstrated yet another beneficial ground effect: In 2 different classic Simonelli tops, reducing G with no other change prolonged the time needed to spin down from 1,600 to 1,500 RPM by up to 7%, with the greatest benefit at G = 3 mm.

Question is, why did Iacopo see a small but significant ground effect at G = 3 mm when I found none? Two reasons, I think:
1. My rotor had a big hole in the center for air to flow through, and his had none.
2. His tops were completely streamlined in all directions and operated entirely in the laminar regime. Mine, on the other hand, had spokes, gear teeth, and sharp edges capable of generating vortices, turbulent wakes, and other flow complications.

Take-home lesson: Arguably, the 2 most counterintuitive areas in all of classical mechanics are rigid body and fluid dynamics, and here we're trying to tackle both at once! Imagining the air flows our tops stir is tricky business. Imagining how those flows affect the air resistances our tops encounter is even trickier.

Note on play value: The more muscle needed to start a top by hand, the more wiggle room needed to avoid scraping during the twirl. With its 2° scrape angle on a flat surface, Top E had this problem in spades. Though Top E had decent play value on the lens, I view Top A as the keeper here for its superior ease of use.

[Iacopo]

Best Top E times from 1,010±10 RPM to 100 RPM
Top E, G = 13 mm .................. 185 s
Top E, G =   3 mm .................. 183 s

This is what I meant when I suggested that tops with a bore do not benefit from the ground effect.
Your experiment confirms it.

But if you can use your top with the fairings, I am sure you will see the difference, even more than me, because your top is large and light and very sensible to aerodynamics.   

[Jeremy McCreary]

More ground effect tests

Sheesh, now I understand why von Braun decided to spin his space stations in a vacuum! Back now to fully faired space station Top D — this time with the width G of the uniform "air gap" beneath the lower fairing ranging from 2 to 20  mm at constant CM height.

Top D: Spoked flywheel top, max flywheel radius R = 84 mm, flat fairings above spokes and below flywheel. When vertical on a flat surface, G = 12 mm.



Base: Small, deeply concave lens on right below, lubed with skin oil.



Bottom shroud: Clear round disk mounted around lens as shown. Flat part of shroud has max radius R + 6 mm.



Variables
1. Air gap thickness beneath bottom fairing (G): 2 - 20 mm

Constant across all time measurements
1. Release speed (ω₀): 1,010±10 RPM
2. Final speed (ω_F): 120±1 RPM
3. CM-contact distance (H): 25 mm
4. Total mass (M): 169 g
5. AMI (I₃): 8.2e-4 kg m²
6. TMI at tip (I_1t): 5.2e-4 kg m²
7. Measured critical speed (ω_C): ~110 RPM

Best Top D times from 1,010±10 to 120±1 RPM, on lens over shroud unless otherwise noted
G = 2.0 mm .............................. 303 s
G = 5.2 mm .............................. 310 s*
G = 8.4 mm .............................. 310 s*
G = 11.6 mm ............................. 319 s*
G = 20 mm, on lens sitting on table ..... 317 s*


Ground effect? Still no evidence for a beneficial aerodynamic ground effect in this top. In fact, just the opposite, as times only grew as G increased from 2 to 20 mm. Must be aerodynamic, as nothing else changed.

So why did Simonelli Nr. 29 below see a clear and beneficial ground effect? Starting to think that flywheel profile matters.



Was the shroud's radius large enough? Probably, but planning one more test with a much larger shroud radius at the same G values. Just waiting on the material.

Hula-hooping instability: Top D was happy to sleep quietly from release to fall at G > 5.2 mm. But at smaller air gaps, it insisted on hula-hooping with decreasing amplitude from release to ~300 RPM. Why? Guessing some kind of periodic instability in the restricted air flow beneath the top. The larger, shallower lens used in all previous tests (above, left) couldn't contain the hula-hooping reliably — hence the lens change.

Will this study ever end? Every time I think it's done, it just keeps going. Which reminds me of a song...

[Iacopo]

G = 2.0 mm .............................. 303 s
G = 5.2 mm .............................. 310 s*
G = 8.4 mm .............................. 310 s*
G = 11.6 mm ............................. 319 s*
G = 20 mm, on lens sitting on table ..... 317 s*

You made a large hole in the bottom shroud, so the air fueling the Von Karman flow under the top can come directly from there, if the bottom shroud stays higher than the lens.  Maybe when the bottom shroud is in a lower position, (gap 20 mm), its hole is sealed by the lens, and you have a bit of ground effect, but when you rise the bottom shroud for to reduce the gap, a gap forms between the lens and the bottom shroud, so you lose the ground effect.

[ortwin]

Once my "Easy Listening" suspension top is running stable, with reproducible results, balanced and with reliable RPM readings available, I hope to  make some measurements with/without fairings....

Just a quick result from "Easy Listening" on concave mirror: no fairing 8:45Directly after this run I applied a foil to the upper side: upper fairing 8:50

So really about the same. Actually I am surprised that the time did not become smaller.Maybe I am with my top geometry accidentally in the cross over region where benefits and costs of a fairing cancel themself out more or less?Methodically it would be better not to compare the full spinning times, but the times of the spin down from 500 RPM to 450 RPM for example

[Iacopo]

So really about the same. Actually I am surprised that the time did not become smaller.Maybe I am with my top geometry accidentally in the cross over region where benefits and costs of a fairing cancel themself out more or less?Methodically it would be better not to compare the full spinning times, but the times of the spin down from 500 RPM to 450 RPM for example

It is 1% difference, very little.  I agree with your observations;  with the fairing, you probably have the advantage of the ground effect and at the same time the disadvantage of more surface for the Von Karman flows.  If you can test at higher speed the difference should become more evident, even 500-450 RPM should be better than full spinning times, if you can be accurate enough as for the timings.