iTopSpin

Current Posts => Collecting, Modding, Turning and Spin Science => Topic started by: James on June 10, 2019, 04:13:14 PM

Title: Top Nr.15 (38 minute spin duration!)
Post by: James on June 10, 2019, 04:13:14 PM
Hey ;D!

I recently finished making my 15th top. It has a brass flywheel, walnut core/spindle and a conical tungsten carbide tip (in a tapered brass holder *credit to Iacopo Simonelli for the great idea*) which is internal. It weighs 252g, 58mm diameter. Current best spin duration: 38 minutes 26 seconds (Personal best!).

Here are some pictures:


(https://i.ibb.co/3rMT7Bj/IMG-4227.jpg) (https://ibb.co/3rMT7Bj)


Above is Top Nr.15 ^


(https://i.ibb.co/cy4Htwd/IMG-4228.jpg) (https://ibb.co/cy4Htwd)


Above is Top Nr.15 (right) and Top Nr.11 (left) for scale ^


(https://i.ibb.co/yXNh9JC/F50-DC20-C-64-E2-4-CCB-A964-1-EA383-D126-F9.jpg) (https://ibb.co/yXNh9JC)


Above is the underside. You can see the conical tungsten carbide tip. ^

I tried to attach a time lapse of Nr.15 spinning for 30 minutes (lasts 30 seconds). In it, you could see how it takes a while for the precession to stop, and it wobbles significantly just before falling over at a very slow RPM. (It was too large of an attachment, could you tell me how to post a .mov file? thanks)

I will conduct some efficiency experiments etc in the near future hopefully, and will upload the results. It has a very narrow spindle at the top (it is tapered to 3mm and is hand-knurled), allowing me to start it at a very high speed. I believe I can start it at about 2,500 rpm (and it falls at about 120 rpm or maybe less) but I will check this with my tachometer when I measure efficiency.

Please give me your feedback, comments or ideas for improvement! Thanks for reading.
Title: Re: Top Nr.15 (35 minute spin duration!)
Post by: topper777 on June 10, 2019, 08:12:20 PM
Great work!
Title: Re: Top Nr.15 (35 minute spin duration!)
Post by: the Earl of Whirl on June 10, 2019, 10:24:53 PM
Once again, just so amazing!!!
Title: Top Nr.15 (38 minute spin duration!) Efficiency
Post by: James on June 11, 2019, 08:20:24 PM
I used my tachometer to take RPM readings every minute during the 38 minute spin. I started the top at 2650 RPM (though I think I could start it a bit faster), and it fell at 169 RPM (I think i can get the balance a bit better too). I calculated the efficiency as: RPM at end of minute / RPM at start of minute. Using that, I made this graph:


(https://i.ibb.co/M6qXLWD/Screenshot-2019-06-11-at-9-10-20-PM.png) (https://ibb.co/M6qXLWD)


I hope that isn't cropped. If it is, could you remind me how to fix it? I tried altering the size on the imb page, but it corrected itself to the automatic size each time.

Anyway, the graph should show that efficiency increases from 92%, peaking at about 94% during the middle of the spin, and then decreases to 90% just as it falls. Due to my cheap tachometer, there are some anomalous results, however the trend line in red shows the curve of best fit.

Please feel free to give me your thoughts or questions. Thanks for reading!
Title: Re: Top Nr.15 (38 minute spin duration!) Efficiency
Post by: Jeremy McCreary on June 11, 2019, 09:59:31 PM
Strong work on both the top and the efficiency measurements!

If aerodynamic drag and simple sliding friction at the tip were the only braking torques acting on your top, your efficiency vs. time curve wouldn't have a peak. So something else is going on -- probably at the tip.

Q1: Do you see any evidence that your tip is wearing down or digging into its supporting surface in the course of a single spin?

Q2: Would you be willing to post your raw (time, RPM) data points? If so, just paste them into a message as text if that's easiest for you.

I'd like to compare your spin decay and efficiency curves to the curves predicted by some simple models of air and tip resistance. The latter can involve contact processes beyond simple friction. Tip resistance might then have a weak non-linear speed dependence, as your efficiency curve suggests.
Title: Re: Top Nr.15 (38 minute spin duration!)
Post by: ta0 on June 11, 2019, 10:26:20 PM
That's a very impressive time! Congratulations!

Can you tell us again how you define efficiency and how you measure it?
Title: Re: Top Nr.15 (38 minute spin duration!)
Post by: Jeremy McCreary on June 12, 2019, 12:42:37 AM
Can you tell us again how you define efficiency and how you measure it?

I used my tachometer to take RPM readings every minute during the 38 minute spin. I started the top at 2650 RPM (though I think I could start it a bit faster), and it fell at 169 RPM (I think i can get the balance a bit better too). I calculated the efficiency as: RPM at end of minute / RPM at start of minute.

James and Iacopo calculate this figure of merit the same way now. It's basically a minute-by-minute spin retention rate and hence a rough measure of the average braking torque exerted on the top over a particular minute of spin-down.

The empirical spin decay curves (SDCs) posted here in the past by ta0 and Iacopo are nearly exponential over roughly the first 2/3 of spin time. In that approximation, the efficiency between time t1 and some later time t2 is

E12 = N2 / N1 = e-(t2 - t1)/T,

where Ni is the speed in RPM at time ti, and T is the top's "lifetime" in the same time unit used for t. No matter where you start on an exponential decay curve, the decaying quantity will drop by ~63% over the next lifetime T.

When t2 - t1 = 1 minute, and T is in minutes, the efficiency reduces to

E12 = e-1/T,

Hence, the efficiency is constant in a truly exponential decay.

For an approximately exponential SDC, you can use the 1st equation to find the lifetime T from any two data points (t1, N1) and (t2, N2), like so...

T = (t2 - t1) / ln(N1 / N2),

where ln() is the natural log. If the SDC were truly exponential, you'd get the same lifetime from every pair of points on the curve.

Oops, forgot the best part: The lifetime T would be a great figure of merit for comparing tops WRT spin decay rate. Even if an actual spin decay curve is only roughly exponential, the   calculated lifetime should correlate pretty well with actual spin time once release and topple speed differences are figured in.
Title: Re: Top Nr.15 (38 minute spin duration!)
Post by: Iacopo on June 12, 2019, 04:54:36 AM

Hi, James !

Another nice top, congratulations !

My most similar top to this your new one is my Nr. 20, and your spinning data are not very different from the mine.

This is the efficiency graph of a spin of my Nr. 20, the top was started at 2962 RPM and toppled down at less than 30 RPM.
I calculate efficiency as percentage of RPM retained relatively to one minute before.

(https://i.imgur.com/MvfWjl8.jpg)

38 minutes is a very good spin already, I offer you some suggestions if you want to further improve it;

-  Your spindle is quite tapered in the last upper part; you have 3 mm at the end of the spindle, but the most significant diameter is not that at the end of the spindle, but that at the height where you grasp it with your fingers, at about 10 mm from the top of the spindle.
I have about 4 mm diameter at that height in my tops.
Because of the quite tapered shape at the end of your spindle, probably you have more than 4 mm at that height, and, if so, I believe you can improve a bit the starting speed making it a bit narrower, at that height.
You say that your spindle is knurled but it seems smooth, looking at the photos, (maybe you knurled it later ?)

-  Lubricants are not all the same. I use a thin layer of WD40, which allows for 1-3 minutes longer spins compared to other lubricants.

-  Accurate balance also is important for longest spins, as you already know.

-  The tip of your top seems slightly dull, not sharp.  If you can make it sharper, the top will spin slightly longer.

-  The tip in your top is not deeply recessed.  The more you recess it, the more stable the top, and the top will spin longer.  You can recess it up to about half the thickness of the flywheel;  if you recess it more than so, your top will become a "spindulum", which will never topple down but it will stay upright even when not spinning.  If you recess the tip enough to make the top to topple down at about 30 RPM, instead of 169, you will have about additional 6 minutes of spinning.   

     
Title: Re: Top Nr.15 (38 minute spin duration!)
Post by: James on June 12, 2019, 05:58:57 AM
Thank you all for the comments and advice. I will try to incorporate them into my next top.
I believe I did knurl the spindle after those photos were taken. Some photos:


(https://i.ibb.co/Jd2qdwL/IMG-4253.jpg) (https://ibb.co/Jd2qdwL)


Above is a photo with magnification, revealing the imperfections in my hand-filing.


(https://i.ibb.co/Msxv1QB/IMG-4254.jpg) (https://ibb.co/Msxv1QB)


Above is without magnification - what you would see normally.

As Iacopo thought, the diameter of the spindle where I hold it is about 4mm diameter. It is true, the tip could be more recessed. I was worried the hole I was drilling for the brass holder would break through the wood on the other side if I went to deep. I might try a non - replaceable tip for my next top, allowing me to have a deeper recession. The tip could be sharper. I was originally more conical, but as I balanced it with diamond filers, it became more rounded. I will also try to fix this next time, and I will eventually try with this top as well using diamond burrs. I believe I am experiencing some wear during the spin, both in the tip and base. Both tip and base are the exact same material. Supposedly, they are tungsten carbide gauge pins (I bought from amazon). However, over time, brown spots appear. This only occurs when I do not sharpen/grind them for some time, and they have no oil coating.

A photo:


(https://i.ibb.co/s3Y6zTW/IMG-4257.jpg) (https://ibb.co/s3Y6zTW)


Above shows the spots on 'tungsten carbide'. I am not sure if this is meant to happen, or if it is an indicator that the material is not tungsten carbide. I also calculated it has a density of 7.6g/cm^3. This is very different to the density of carbide which, according to google, should be around 15 g/cm^3. Thoughts?

Minute   Rpm      Efficiency
0   2650   0.923   92.3
1   2445   0.924   92.4
2   2260   0.926   92.6
3   2093   0.929   92.9
4   1945   0.931   93.1
5   1810   0.932   93.2
6   1687   0.932   93.2
7   1572   0.934   93.4
8   1468   0.933   93.3
9   1370   0.938   93.8
10   1285   0.935   93.5
11   1202   0.939   93.9
12   1129   0.942   94.2
13   1063   0.940   94.0
14   999   0.937   93.7
15   936   0.940   94.0
16   880   0.942   94.2
17   829   0.941   94.1
18   780   0.919   91.9
19   717   0.934   93.4
20   670   0.934   93.4
21   626   0.942   94.2
22   590   0.941   94.1
23   555   0.939   93.9
24   521   0.939   93.9
25   489   0.941   94.1
26   460   0.937   93.7
27   431   0.933   93.3
28   402   0.935   93.5
29   376   0.931   93.1
30   350   0.934   93.4
31   327   0.927   92.7
32   303   0.908   90.8
33   275   0.931   93.1
34   256   0.918   91.8
35   235   0.911   91.1
36   214   0.907   90.7
37   194   0.897   89.7
38   174   0.000   
            
38:26:00   169   (Falls over)

Above is my raw data for the 38 minute spin (efficiency graph)   

I have tried taking screenshots of the graphs so less is cropped:


(https://i.ibb.co/pz889Lw/Screenshot-2019-06-12-at-10-55-55-AM.png) (https://ibb.co/pz889Lw)


Above is spin decay (RPM vs time)


(https://i.ibb.co/p1gcVx9/Screenshot-2019-06-12-at-10-57-22-AM.png) (https://ibb.co/p1gcVx9)


Above is efficiency of Top Nr.15

Thanks for reading!
Title: Re: Top Nr.15 (38 minute spin duration!)
Post by: Iacopo on June 12, 2019, 08:29:17 AM
I also calculated it has a density of 7.6g/cm^3.

In fact this is far too low a density for tungsten carbide.
I calculated the density of my tungsten carbide rod and I found it is 14.3 g/cm3.
Also the carbide I have does not rust.
The your seems more like HSS steel, or carbon steel.
But you can find easily tungsten carbide rods on Amazon and they are not expensive..

This also could explain the difference between our spin times, because with carbide contact points the top spins longer than with steels.

Title: Re: Top Nr.15 (38 minute spin duration!)
Post by: James on June 12, 2019, 03:16:58 PM
Thanks for the reply. I will look on amazon for some real tungsten carbide. Since the tip of this top is replaceable, I could make a new tip with a carbide point for use in this top Nr.15 as well as those in the future. Also, Iacopo, I watched one of your videos a while ago about tip and base materials. I remember you saying that HSS and tungsten carbide together had the least wear. Do you think a tungsten carbide tip on an HSS base would be good? Also, have you ever tried using a base made of ruby/sapphire? If you could use this material, do you think it would be good in combination with a sharp carbide point?
Title: Re: Top Nr.15 (38 minute spin duration!)
Post by: Iacopo on June 12, 2019, 04:59:30 PM
Do you think a tungsten carbide tip on an HSS base would be good? Also, have you ever tried using a base made of ruby/sapphire? If you could use this material, do you think it would be good in combination with a sharp carbide point?

No, it doesn't work well in that way.
To do things right the spinning surface should be made of tungsten carbide, in any case;
the tip instead can be made of carbide, or HHS, if you prefer it;
with the HSS tip the contact points will last longer.
With the carbide tip instead the spin times will be about 10-15 % longer than with the HSS tip.

Tops with a recessed tip wobble without mercy, (nutate), when the spinning surface is weared out, so it is important to reduce to a minimum the wear of the base, (unless you are ok seeing your top to wobble persistently while spinning).
If your base is made of steel and not carbide, I believe that it will take not many spins before you will see your top to wobble spontaneously because of the worn out base.

I made and tried a ruby base, long time ago, and it behaved poorly.
If it worked as well as I hoped, I would be using ruby for my bases, but I don't;
certainly ruby is hard, and harder than tungsten carbide, but I read somewhere that wear is related not only to hardness, but to toughness too.  So I suppose ruby fails on the side of toughness, (it's hard but brittle).
Tungsten carbide is the most wear resistant material I know at present, for our needs.

Title: Re: Top Nr.15 (38 minute spin duration!)
Post by: Jeremy McCreary on June 12, 2019, 06:33:54 PM
James, Iacopo: At what time, if any, in the graphs above did your tops start precessing? Wobbling?

Are videos of these tops available? Of their entire spins?

Even ballpark times would be helpful. Thanks!
Title: Re: Top Nr.15 (38 minute spin duration!)
Post by: Jeremy McCreary on June 13, 2019, 01:16:40 AM
Above is my raw data for the 38 minute spin (efficiency graph)

Thanks for the data. Some prelim results...

Based on surrounding trends, there may be problems with the times or speeds reported at times 19, 33, and especially 38:26 minutes. So I ignored the last. The rest of your empirical spin decay curve is very close to the exponential

N(t) = N0 e-t / T,

where first measured speed N0 = 2,650 RPM and lifetime T =  14.52 min. The correlation coefficient R² = 0.99800 indicates a near-perfect exponential fit.

But if the measured spin decay curve had been exactly exponential, the efficiency would have been constant (a flat 93.3% for the N0 and T values above). Instead, you and Iacopo both got convex-upward efficiency vs. time curves. Given the differences in your tops, that's probably trying to tell us something about the braking processes involved. Hard to imagine that it's an aerodynamic signal, though.
Title: Re: Top Nr.15 (38 minute spin duration!)
Post by: James on June 13, 2019, 01:30:01 PM
Jeremy,

I don't have footage of the 38 minute spin, but i do have a time lapse of a 32 minute spin of this top. In this video, it takes 9 minutes for precession to completely disappear and wobble starts to appear at about 22 minutes in. This wobbling starts very subtly, then quickly increases until the top falls. I believe I started the 38 minute spin a bit faster, and the balance was a bit better. So, I think in the 38 minute spin, wobbling would also have started about 8 minutes from the end (30 minutes in) because the better balance would allow it to spin slower before wobbling starts . Next time I do a spin duration test, I will look out for when wobbling starts.
Title: Re: Top Nr.15 (38 minute spin duration!)
Post by: Iacopo on June 13, 2019, 03:57:59 PM
James, Iacopo: At what time, if any, in the graphs above did your tops start precessing? Wobbling?

Are videos of these tops available? Of their entire spins?

Even ballpark times would be helpful. Thanks!

I am sorry but I don't have those data, nor a video.

I can say that, in tops like these, there is generally some nutation and precession at the start, which disappear in a few minutes, until the top spins in sleeping position.  The wobble that appears at the end of the spin is "unbalance wobbling".
There is not nutation nor precession at the end of the spin, in these recessed tip tops, or, at least, I never noticed them, (unless the top was already nutating/precessing before, never having reached the sleeping position).

One curious fact about efficiency and wobbling is that not always a wobbling makes the top less efficient;
in my tops with recessed conical tip, this is still true, and efficiency is better when the top does not precess and does not nutate.
So I strive to spin the top as vertically as possible when I want a very long spin with one of my tops.
Anyway I noticed that the countrary happens with tops with a ball tip spinning on a flat surface:
efficiency is better when the top precesses instead of sleeping, (at parity of RPM).
Curious, isn't it ?
This is something you can experiment by yourself.
Title: Re: Top Nr.15 (38 minute spin duration!)
Post by: Jeremy McCreary on June 13, 2019, 06:00:07 PM
I don't have footage of the 38 minute spin, but i do have a time lapse of a 32 minute spin of this top. In this video, it takes 9 minutes for precession to completely disappear and wobble starts to appear at about 22 minutes in. This wobbling starts very subtly, then quickly increases until the top falls. I believe I started the 38 minute spin a bit faster, and the balance was a bit better. So, I think in the 38 minute spin, wobbling would also have started about 8 minutes from the end (30 minutes in) because the better balance would allow it to spin slower before wobbling starts.

Ignore the smoothed red curve in your efficiency vs. time plot and look just at the data points. Allowing for noise, you can make out a rising segment on the left, a falling segment on the right, and a more or less flat segment in between.

Based on your observations of spin-down behaviors, sounds like the flat segment might have occurred during sleep. If so, the upward and downward segments could reflect changing tip resistances during behaviors other than sleep.

Iacopo's efficiency curve for his Nr. 20 has the same structure. It must mean something.

Next time I do a spin duration test, I will look out for when wobbling starts.

That would be helpful. Getting a good spin decay curve (SDC) is the first step toward understanding the air and tip resistances in play during spin-down. I think we'd get some valuable insights from annotated SDCs showing when sleep, precession, wobble of any kind, and combinations of these behaviors were present. Wish I could collect good SDCs on my LEGO tops, but air resistance and low material densities bring nearly all of them down in under 2 minutes.

Collecting SDC points every minute on the minute gives you a feel for how the spin decay rate is evolving without plotting the data. But the most important thing in collecting a good SDC is to record the right times for the speeds actually measured -- even if the times aren't exactly on the minute.

Since the speed measurements take time, there will always be some slop in time-speed registration, but the misregistration should be minimized as much as possible. The big dips in your efficiencies at 19:00, 33:00, and 38:26 could be real if top was bumped or the tip hit an irregularity on the table. Otherwise, time-speed misregistration is a good suspect.

This might help: Calculate lifetime from successive measurements instead of efficiency (formula above). Then you can collect data points whenever convenient. The lifetimes will automatically correct for differing time steps between speed measurements.
Title: Re: Top Nr.15 (38 minute spin duration!)
Post by: Iacopo on June 14, 2019, 04:38:05 AM
Based on your observations of spin-down behaviors, sounds like the flat segment might have occurred during sleep. If so, the upward and downward segments could reflect changing tip resistances during behaviors other than sleep.

I believe that the downward segment at the left is mainly due to high air drag.
Then this way "efficiency" is calculated may be misleading because it may lead to believe that there is an increase of the total friction at low speed but this is not true, because we have the lowest total friction at the lowest speeds.   

Like James, I too observe continuous changes of the efficiency of the top during the spin.
my tachometer is good enough and I can say that these continuous changes are real.
These continuous and unpredictable changes are due to the tip friction.  Air drag too changes, at the decreasing of speed, but gradually, and always in the same way. 

Here are the data of the tip friction, (obtained making the top spin in deep vacuum conditions), of 19 spins of  my Nr. 27, (156 grams):
the top has an external carbide tip and was spun always vertically so there was nearly no precession at the start of each spin.
 

(https://i.imgur.com/KfaTjfM.jpg)

(https://i.imgur.com/cEfgI3x.png)


The 19 spins are numbered chronologically, the spin Nr. 1 was the first spin performed, when the tip was new and very sharp.
The first two spins are the ones with the lowest friction, telling that a sharp tip has lower friction that a dull tip.

The spins with a red number were performed without lubricant, the ones with green, violet and blue numbers are spins with some different lubricants.  It is evident that the tip friction is lower with the lubricants.

There is a continuous and random variability of the tip friction. 
There are more important and frequent changes in heavy tops than in light tops, especially without lubricant.

This is the tip friction of a light, (77 grams), top on lubricated spinning surface;
With light tops spinning on a lubricated surface there are not significative up and downs but the tip friction decreases gradually with speed.

(https://i.imgur.com/UHoEjwH.png)
 
I believe that the unpredictable changes of the tip friction are related to wear. 
In fact there is wear in the contact points;
this is the photo of a dimple in the carbide spinning surface caused by wear, at the microscope; the dimple is about 0.1 mm large and 0.03 mm deep, and is the result of 50 hours of spinning of a top with a conical tip.
These dimples are responsible for making tops with a recessed tip to wobble spontaneously, (nutate), when they spin on them.

 (https://i.imgur.com/0Sv3bv1.jpg)
Title: Re: Top Nr.15 (38 minute spin duration!)
Post by: Jeremy McCreary on June 14, 2019, 01:43:48 PM
Love discussions like this -- especially when they bring in real data on real tops!

I believe that the downward segment at the left is mainly due to high air drag.

Could well be. I'll have more to say about this in a coming post. The spin decay curve (SDC) data you provided some time ago on your Nr.25 under normal air pressure will figure prominently.

Like James, I too observe continuous changes of the efficiency of the top during the spin.
my tachometer is good enough and I can say that these continuous changes are real.
These continuous and unpredictable changes are due to the tip friction.  Air drag too changes, at the decreasing of speed, but gradually, and always in the same way.

Totally agree (i) that air resistance isn't to blame for the sudden variations in efficiency, and (ii) that varying tip resistance (sliding friction + rolling resistance + active wear processes + ??) is likely the main culprit. But still worry about artifacts due to time-speed misregistration. The efficiency calculation could be quite sensitive to such errors.

Q1: How exactly do you assign times to your speed measurements? (For both Iacopo and James.)

My tachometer readings can take many seconds to stabilize. Certainly couldn't take them every minute on the minute and expect the speeds I get to correspond to exact minute marks.

That's why I keep pushing lifetimes (formula several posts back) instead of minute efficiencies to track the evolution of total resistance. Lifetimes correct for variations in measurement intervals automatically.

There is a continuous and random variability of the tip friction.
There are more important and frequent changes in heavy tops than in light tops, especially without lubricant.

Interesting and potentially important observation.

I believe that the unpredictable changes of the tip friction are related to wear.
In fact there is wear in the contact points;
this is the photo of a dimple in the carbide spinning surface caused by wear, at the microscope; the dimple is about 0.1 mm large and 0.03 mm deep, and is the result of 50 hours of spinning of a top with a conical tip.
These dimples are responsible for making tops with a recessed tip to wobble spontaneously, (nutate), when they spin on them.

(https://i.imgur.com/0Sv3bv1.jpg)

That's quite a hole! Totally agree that active wear processes are likely contributors and perhaps even the main contributors.

Imagine all the "contact shenanighans" -- stick-slip, transient jamming, hopping, rapid variations in contact area and normal force, etc. -- that might go on with a conical tip spinning in a conical hole -- especially in the presence of precession or wobble!

Here are the data of the tip friction, (obtained making the top spin in deep vacuum conditions), of 19 spins of  my Nr. 27, (156 grams): the top has an external carbide tip and was spun always vertically so there was nearly no precession at the start of each spin.

(https://i.imgur.com/KfaTjfM.jpg)

The 19 spins are numbered chronologically, the spin Nr. 1 was the first spin performed, when the tip was new and very sharp.
The first two spins are the ones with the lowest friction, telling that a sharp tip has lower friction that a dull tip.

The spins with a red number were performed without lubricant, the ones with green, violet and blue numbers are spins with some different lubricants.  It is evident that the tip friction is lower with the lubricants.

Early on, your lubricated vacuum runs showed minor, long-period variations that could well be within measurement error. The big short-period spikes and dips didn't show up until lubricated run 13, when a hole had presumably already been started. But without lubrication (which would accelerate hole formation and also promote stick-slip behavior), they showed up in much earlier runs.

All of this points to the kinds of contact shenanighans I mentioned earlier.

This is the tip friction of a light, (77 grams), top on lubricated spinning surface;
With light tops spinning on a lubricated surface there are not significative up and downs but the tip friction decreases gradually with speed.

(https://i.imgur.com/UHoEjwH.png)

A linear tip resistance (braking torque) vs. speed curve with a negative slope implies an exponential SDC (speed vs. time curve). These tip resistance vs. time data points are nearly linear with a clear negative slope over a wide range of speeds. If simple sliding friction were the only contact process involved, you'd have a tip resistance vs. speed curve that's both linear and flat -- i.e., with zero slope.

But you would expect this kind of tip resistance vs. speed curve from speed-dependent viscous friction, and that points to the lubrication. Active tip wear and other contact shenanighans might also add some speed dependence.

Q2: How do you calculate your braking torques again?


Title: Re: Top Nr.15 (38 minute spin duration!)
Post by: James on June 14, 2019, 02:31:03 PM
Thanks for all of this really interesting information! For such a seemingly simple motion - spinning - there is a lot more going on than you would expect.

Q1) I usually use my tachometer starting about 5 seconds before the minute mark, allowing it some time to stabilise.
Q2) I divide the RPM reading of 'minute X' by the PM reading of 'minute X-1'. For example if at minute 8 rpm is 1000, and at minute 9 rpm is at 800, I divide 800 by 1000 to get an 80% efficiency for that minute.

I hope that answers your questions. If not, could you elaborate slightly?
Title: Re: Top Nr.15 (38 minute spin duration!)
Post by: Iacopo on June 14, 2019, 04:04:19 PM
Imagine all the "contact shenanighans" -- stick-slip, transient jamming, hopping, rapid variations in contact area and normal force, etc. -- that might go on with a conical tip spinning in a conical hole -- especially in the presence of precession or wobble!

The shape of the dimple is not conical, you can se it better here:

https://youtu.be/kXDKc8vOz3M?t=157

I will answer Q1 and Q2 later when I have a bit more time.


James, how many hours did it take to make your last beauty ?
Title: Re: Top Nr.15 (38 minute spin duration!)
Post by: James on June 14, 2019, 04:23:12 PM
If I were to take a guess, I would think It took about 10 hours. The base probably took another 8 hours. However, due to me being in school, I only have access to a workshop for about 3 hours a week. So It took me around 6 weeks. If I wasn't rushing to finish the top before my school ends, I would have spent longer on it.
By the way, I finished my GCSEs today! I am taking A-level DT for the next 2 years. As well as this, I have been working on a model rocket car with a group of friends in an attempt to break the world record for the fastest model rocket car. In addition, I will be working on a safe making competition and a can sized meteorological-engineering project. So I will be very busy. I will try to work on tops in my spare time, however, I will have less of it so production may slow down a bit.

How long does it normally take you?
Title: Re: Top Nr.15 (38 minute spin duration!)
Post by: Jeremy McCreary on June 14, 2019, 07:48:16 PM
Imagine all the "contact shenanighans" -- stick-slip, transient jamming, hopping, rapid variations in contact area and normal force, etc. -- that might go on with a conical tip spinning in a conical hole -- especially in the presence of precession or wobble!
The shape of the dimple is not conical...

Still, imagine all the shenanighans!

Like your diagrams showing how tip radius of curvature affects the size and shape of the lubricant film.

Your tops clearly benefit from lubricant, and I see that you prefer WD40. My heavier tops benefit, too, but the only lubricant I ever use now is skin oil from my face (per Alan's recommendation). Works pretty well.

Tried silicone lube once. Took days to get that *$^#%@! stuff off my hands.
Title: Re: Top Nr.15 (38 minute spin duration!)
Post by: Jeremy McCreary on June 14, 2019, 11:18:18 PM
For such a seemingly simple motion - spinning - there is a lot more going on than you would expect.

Yes, one of my favorite things about tops! Articles about them still appear in physics journals. And a model of top behavior that includes air resistance (usually the main contributor to spin decay) has yet to be published.

Q1) I usually use my tachometer starting about 5 seconds before the minute mark, allowing it some time to stabilise.

Thanks! Sounds like your times and speeds are likely to be well-aligned. Which means that the sudden efficiency spikes and dips you and Iacopo see are likely to be real. All the more interesting, no?

Q2) I divide the RPM reading of 'minute X' by the PM reading of 'minute X-1'. For example if at minute 8 rpm is 1000, and at minute 9 rpm is at 800, I divide 800 by 1000 to get an 80% efficiency for that minute.

Sorry, Q2 was for Iacopo. His tip friction graphs are really braking torque vs. speed curves, and I was wondering how he estimated the torques from his SDC data. There's more than 1 way. Efficiency and braking torque aren't the same thing, but the greater the torque over a given time interval, the lower the efficiency.
Title: Re: Top Nr.15 (38 minute spin duration!)
Post by: Iacopo on June 15, 2019, 01:42:32 AM
Q1: How exactly do you assign times to your speed measurements?

I timed the lapses every 50 RPM lost in the first graph and every 200 RPM lost in the second graph.

Q2: How do you calculate your braking torques again?

Here is a sample:

I timed 1'36"2 for my top Nr. 27 to go from 1000 to 900 RPM.
100 RPM lost in 96.2 seconds.

Deceleration is:
100 RPM = 10.47 rad/sec, (lost in 96.2 seconds).
10.47 : 96.2  = 0.109 rad/sec, (lost in 1 second).
Angular deceleration is  0.109 rad/sec2.

Torque is:
Moment of inertia x angular deceleration.
The moment of inertia of this top is 0.0000643 kg-m2.
0.0000643 x 0.109 = 0.00000701 Newtonmeters = 7.01 millionths of Newtonmeter.

Title: Re: Top Nr.15 (38 minute spin duration!)
Post by: Iacopo on June 15, 2019, 08:24:11 AM
I have been working on a model rocket car with a group of friends in an attempt to break the world record for the fastest model rocket car...

How long does it normally take you?

I didn't know about these model rocket cars, I looked for them in the web and.. wow.. they are really fast !!
You are involved in very interesting projects.

For the same top as the your I think it would take me about the same time like you.
But for more complex tops it takes me more time.
The most complex one I made at present is the Nr. 39 and it took me about 150 hours to make the full set, (top, base, case and accesories).
62 hours for to make the Nr. 20, (top and base).
About 4 hours for to make the Nr. 27, which is very simple, (top alone).
Some things I do are quite time consuming; varnishing, and polishing, for example.
Doing things with accuracy in every detail generally takes time.

Title: Re: Top Nr.15 (38 minute spin duration!)
Post by: James on July 27, 2019, 10:35:36 AM
For anyone interested in this top, there is now a video about it on my youtube channel: HiPer Spinning Tops. It has better shots of the spinning top than I have been able to post here and other pieces of information.

A link: https://www.youtube.com/watch?v=THPeeyesEIM
Title: Re: Top Nr.15 (38 minute spin duration!)
Post by: ta0 on July 27, 2019, 11:33:07 AM
Wow! It was difficult not to think I was watching a Iacopo video!
You have learned well from the Master!
Excellent work!
Title: Re: Top Nr.15 (38 minute spin duration!)
Post by: James on July 28, 2019, 02:49:36 AM
Thank you very much! I owe a lot to Iacopo as he helped me when I started making them. As well as that, as you can see, my tops are inspired mainly by his.
Title: Re: Top Nr.15 (38 minute spin duration!)
Post by: Iacopo on July 28, 2019, 09:25:45 AM
For anyone interested in this top, there is now a video about it on my youtube channel: HiPer Spinning Tops. It has better shots of the spinning top than I have been able to post here and other pieces of information.

A link: https://www.youtube.com/watch?v=THPeeyesEIM

Just out of curiosity;  did you make something particular for driving the motion of the camcorder in the sequence starting at 0:19 ?
Or you simply moved it by hand, along the edge of the table ? The motion seems very fluid.
Title: Re: Top Nr.15 (38 minute spin duration!)
Post by: James on July 29, 2019, 11:54:26 AM
Im glad you thought it was fluid  ;). Yes, I pushed it by hand. It took me about half an hour to get the shot I wanted. I may invest in some better filming gear in the future. At the moment I am limited to my iPhone, a cheap tripod and lenses harvested from broken optical equipment and lasers.
Title: Re: Top Nr.15 (38 minute spin duration!)
Post by: Iacopo on July 29, 2019, 02:03:06 PM
Im glad you thought it was fluid  ;). Yes, I pushed it by hand. It took me about half an hour to get the shot I wanted. I may invest in some better filming gear in the future. At the moment I am limited to my iPhone, a cheap tripod and lenses harvested from broken optical equipment and lasers.

I too tried in the past and I know it's not very easy, in fact I nearly always use the tripod and I avoid to move the camcorder for my shots.  You made a good sequence.  I thought that maybe you made some sort of rail for the camera.

 
Title: Re: Top Nr.15 (38 minute spin duration!)
Post by: James on July 29, 2019, 09:35:44 PM
I have tried using round pencils under a book so the camera and tripod can slide, however the motion is difficult to control. I would have tried making a rail but most of my tools were in England. I also don't have the appropriate materials. It may be a project for the future.
Title: Re: Top Nr.15 (38 minute spin duration!)
Post by: Jeremy McCreary on July 30, 2019, 10:12:51 AM
I have tried using round pencils under a book so the camera and tripod can slide, however the motion is difficult to control. I would have tried making a rail but most of my tools were in England. I also don't have the appropriate materials. It may be a project for the future.

Now you've given me today's LEGO project: A motorized remote control phone or Go-Pro camera dolly with adjustable height and a manual pan-tilt mount. Similar things have been done before, and I have all the parts!

Could be on rails using a LEGO train as a platform, but I think a car on steerable rubber wheels would be steadier and more flexible WRT the dolly's path.