It recently occurred to me that a clay ball that perfectly resembles a golf ball, would probably make for a very effective sling projectile. Being round, I think it would be easy to make them with a two-sided mould, they would stack nicely in a kiln, and they would also be quite durable compared to bi-conical projectiles.
I'm also rather curious as to how far they could be thrown, with the better aerodynamics as a result of the dimples.
Has anyone tried making these?
Do you think it's a practical idea?

dont see why it wouldnt work, wonder what the difference in weight would be.
but for me id make biconicals over spheres

What if you made a biconical with dimples...? Is that a stupid idea? ::)

It would be interesting - Making your own out of clay would be a lot cheaper than buying tons of golf balls, and more environmentally friendly as well (Especially if you dug the clay out of your own yard!) Making a regular golf ball mold would be considerably easier than a dimpled biconical, but it'd be cool to see how it'd turn out.

Captain_Twine wrote on Sep 7^th, 2020 at 9:48pm:

What if you made a biconical with dimples...? Is that a stupid idea?

I've been thinking the same thing.
I have a batch I made quite a while back, but haven't tested them yet.
If the dimples do work with biconicals, then could you imagine how far lead glandes with dimples could go?
If not, what about a very small lead golf-ball?
Much to experiment with...

Mersa wrote on Sep 7^th, 2020 at 8:58pm:

wonder what the difference in weight would be.

Just made a clay ball about the size of a golf ball, and its almost 100g when wet. (I also accounted for shrinkage) 

Are you trying to make ammo for a maximum distance?? Or just experimenting??

Mersa wrote on Sep 7^th, 2020 at 11:09pm:

Are you trying to make ammo for a maximum distance?? Or just experimenting??

Mostly just experimenting, but a few practical things might come out of it. I've heard that a dimpled golf-ball hit by a professional can fly about twice as far as a smooth ball, so I think dimples could be of benefit for maximum distance projectiles. 
The clay golf-ball idea is just something that I think would be good for target practice.

I believe this is a combination of a few factors actually.
Dimples reduce the turbulence on the surface of the ball .
The ball is struck at a lower angle not 45 degrees, a large amount of backspin and the ball Creates lift. This is why golf balls go further than theoretical maximum in a vacuum. They use air as an advantage. I would like to see a comparison between a golf ball and your clay copy.

biconicals with dimples work really well.
not sure of they work better than without dimples. But some years back - a friend gave me a couple of golf cross balls and one did the rounds of about 4 countries and an equal amount of members for people to try and make moulds out of.

Not sure which country it ended up in, but I never got it back :-)

I have tried modelling golfballs in the past - but they tended to crash my cad setup.

As I now have amuch more poweful machine - I'll give it another go.
Plus I know a few more modelling tricks now :-)

Once the basic golf ball base in made - you can stretch it into a biconical.

The actual maths to cover a biconical in dimples is probably beyond me :-)

But I will give standard golfball type another go.

As far as weight goes - it's about the same size as the skulls - so around 60-70gms when dry. Withmy modelling clay anyway.

As my new 3d printer is now finished being modded (sapphire pro 2 corexy), it'll be a decent test for that too.

lol - oh yeah 16gb ddr4 and golf balls are working :-)

Openscad just needs a stupid amount of resources.
Got some really interesting patterns as well.

here's the first interesting looking one.

Going to keep fiddling with dimple size and depth and positioning.
I think golfball dimples are shallower and smaller.

But rendering them is no longer a problem :-)

---

getting some weird patterns.

---

the problem is I can keep messing with number all bloody day lol

Normal golfballs have between 300-500 dimples.
My biggest problem is preventing too many overlays, which ends up as grooves rather than dimples.

I'll play with the math a bit and see if I can find a different way to place the dimples.

Looking good though. :-)

Or you can just go to thingiverse and use one that someone else has already done :-)
https://www.thingiverse.com/thing:1618447/files

Think I'll print it out as a golf ball first :-)

Certainly looking promising!
A couple of ideas came to mind.
1: If you can't make a model with that many dimples, you can just do half the number of dimples. When one makes the clay balls, they can press it once, then rotate the ball, and press it a second time to get the full number of dimples.

2: Instead of having dimples in the mould, one could use a separate dimple board to role the clay balls on afterward. (If that's a better idea)

This might be of some interest.
As the drag coefficient for turbulent flow is less than that of laminar, I think it's safe to assume that these bodies would benefit from having dimples on their surface.

Bottom image shows why really pointy glandes are not preferable for maximum range.

---

Look for chocolate molding molds.  There are also golf balls

https://m.de.aliexpress.com/item/32820619333.html?trace=wwwdetail2mobilesitedetail

Archaic Arms wrote on Sep 8^th, 2020 at 4:24pm:

This might be of some interest. As the drag coefficient for turbulent flow is less than that of laminar, I think it's safe to assume that these bodies would benefit from having dimples on their surface. Bottom image shows why really pointy glandes are not preferable for maximum range.

I haven't seen this before, but always wondered why this was the case. Thank you for educating me!


Jaegoor wrote on Sep 8^th, 2020 at 4:25pm:

Look for chocolate molding molds.  There are also golf balls

That's a good idea.

I tried it years ago.  I did not find a significant difference.  Biconical cross golf balls are really good to shoot.  I always wanted to make a form of it.

Pointy glandes are for war or hunting, not max distances.
Everything has its place

Mersa said : "Pointy glandes are for war or hunting".

Why ? It's possible but ...  :-?

@albion
once you've got a cad model - then it's easy to make the mould.

having compared an actual golf ball with the model I found. The dimples are a lot shallower and do overlap on a normal golf ball.

So I'll go back to my script and have another fiddle :-)
Golf cross ball mould for jaegoor - check :-)

Jaegoor wrote on Sep 8^th, 2020 at 5:13pm:

Biconical cross golf balls are really good to shoot.  I always wanted to make a form of it.

That's a really good idea!
Also because of it's shape, the dimples would turn out better on the sides of the projectile, when made from a 2-sided mould.
It would also sit better in my split-pouches...

@CA... golf dimples should be super easy in OpenScad... it’s just a nested for loop to generate rings of spheres... then you subtract those from the biconical. You can tweak the depth and size of the dimples easily too.

It does seem like it would be more difficult to get the mould to release though unless you are using a silicone mould

nah - talcumn powder or flour works a lot better. the press moulds - as long as you dust them properly release really easy.
even this skull shape comes out fairly easy - and these were made form playdough.

---

putting even dimples on a glande shape is actually really bloody complicated.
as the size of the circle of dimples and the number changes for each 'loop'.
And my maths does not include any knowlege of geometry - I still have not the foggiest diea what sin and cosin are or what they do.

It's a lot easier making a dimpled sphere and then stretching it into a glande shape.

And looking at a golf ball, it would appear that overlaps aren't that important.
Golf balls actually have more of a wrinkled skin appearance - rather than obvious dimples. That's a lot easier to do.

I will take the dogs for  a walk and let the little grey cells do some background processing.

CA, I may join you in your OpenSCAD quest for a dimpled sphere. and I agree it isn't as simple as a for loop of rings of subtracted spheres. The dimples are close packed on a golf ball. A few nested for loops might do it, but I'm not convinced, I think the needed equation and/or algorithm to get the close packing is a bit more complex than that.

Curious Aardvark wrote on Sep 10^th, 2020 at 12:49pm:

putting even dimples on a glande shape is actually really bloody complicated. as the size of the circle of dimples and the number changes for each 'loop'. And my maths does not include any knowlege of geometry - I still have not the foggiest diea what sin and cosin are or what they do. It's a lot easier making a dimpled sphere and then stretching it into a glande shape. And looking at a golf ball, it would appear that overlaps aren't that important. Golf balls actually have more of a wrinkled skin appearance - rather than obvious dimples. That's a lot easier to do. I will take the dogs for  a walk and let the little grey cells do some background processing.

I’ll try to take a crack at it maybe this weekend. You’re right that it would be difficult without an intuition on sin and cos :)

Those skulls would be rather intimidating if they were made of lead (or some other heavy material), with a whistling hole in each eye...

I’ve used clay balls as ammo for most of my time slinging- I live in an area with high clay soils.

I just dig a spade full of two of topsoil, make it into mud, knead well to make it sticky, roll into balls and leave to bake in the sun or slowly bake in the oven.

They’re easy, quick and hold together really well. My favourite thing about them is they break down next time it rains making them very eco (and lawnmower blade) friendly. I haven’t done dimples yet...

Those golf cross balls look fantastic!!! Can you tell more about how they are to sling? I was thinking of getting some and putting some threaded rod through the length of them to up the weight a little...

*Edit... I cleaned up the code for readability and fixed the relationship between the divots and the ball.  There was no point in leaving bad code (not that this is good... just less-bad), so I replaced my original post with the fixed code below...

try this script CA...  :)
you can scale the ball to get an elongated glande, but the divots will also stretch.  It's a start anyway...
========================================================
//=============
// Parametric Golf Ball
//   Written for Curious Aardvark
//   www.slinging.org
//
//
//=============


//User-Settable Parameters
//========================================
ball_dia = 42;

divot_dia = 5;

divot_depth = 1;

//divots around the equator of a golf ball = 30
numdivots_ring = 30;
// number of rings of divots between the equator and the pole of the ball... total number of actual rings is 2*numrings+1
numrings = 6;


//Calculated Parameters - Do not change
//========================================
ball_rad = ball_dia/2;
divot_rad = divot_dia/2;
//virtual surface radius where the divot beads go
divot_surface_rad = ball_rad+divot_rad-divot_depth;

//angle between divot beads
angle_increments= 360/numdivots_ring;

$fn = 50;


module main()
{
    difference()
    {
        positives();
        negatives();
    }
   
    //This next line is just for visualization of how the divots are made... should be commented out most of the time
    //%negatives();
   
    module positives()
    {
        sphere(d=ball_dia);
    }
   
    module negatives()
    {
       for(j=[0:numrings])
       {
           working_ring_rad = divot_surface_rad*cos(j*90/(numrings+1));
           ring_distance_from_center = divot_surface_rad*sin(j*90/(numrings+1));
           numdivots_ring = floor(3.14159*2*working_ring_rad/divot_dia);
          angle_increments = (360/numdivots_ring);
          rotate([0,j*angle_increments/2,0])
          for(i= [0:angle_increments:360])
          {
              //divots on right side
              translate([0,ring_distance_from_center,0]) rotate([0,i,0]) translate([divot_surface_rad*-cos(j*90/(numrings+1)),0,0]) scale([1,1,1])sphere(d=divot_dia);
              //divots on left side
             translate([0,-ring_distance_from_center,0]) rotate([0,i,0]) translate([divot_surface_rad*-cos(j*90/(numrings+1)),0,0]) scale([1,1,1])sphere(d=divot_dia);
             
          }
      
      }
      //cap the two ends with a single divot
      translate([0,divot_surface_rad,0]) sphere(d=divot_dia);
      translate([0,-divot_surface_rad,0]) sphere(d=divot_dia);
    }
}


scale([1,1,1]) main();

---

IronGoober wrote on Sep 10^th, 2020 at 3:05pm:

CA, I may join you in your OpenSCAD quest for a dimpled sphere. and I agree it isn't as simple as a for loop of rings of subtracted spheres. The dimples are close packed on a golf ball. A few nested for loops might do it, but I'm not convinced, I think the needed equation and/or algorithm to get the close packing is a bit more complex than that.

Now are you convinced IG? ;D

That looks pretty good! Better than I expected for sure. But I wonder if you could fully get the close packing of dimples (one dimple surrounded by 6 others, all touching) using this method.  I think you'd have to play with the spacing and rotation of each ring.

I'll take a swing at it when I get back from my vacation. Certainly did better than I expected.

If you have nothing better to do with your time than reinvent the wheel in an unnecessarily complicated way, then I suppose you can do that IG, but there’s absolutely no need to do that. This  design is parametric. If you want the divots to touch, just change the parameters.  You can either increase divot_depth or increase divot_dia to get that result.

Increasing divot_dia will make the divots wider but not deeper. Increasing divot_depth will increase both the diameter and the depth of the divots. 

...by the way,  I think this conversation may have moved into the “too-much maths“ category, but when you try to do this your way, you will eventually realize that you can only pack circles perfectly touching 6 other circles on a planar surface (if all of the circles are the same size). When the surface you are packing is a sphere, the geometry doesn’t work out the same.

If you look at a real golf ball, a handful of the divots are smaller than the others because of this.

right been doing some research.
Actually watching a video of a guy making dimpled fan blades - who actually did the research, I've just nicked it  :-)

So - depth of golf ball dimples is 0.254mm ie: really bloody shallow.

Random patterns work better than regular patterns and it doesn't matter if the dimples overlap.
What is important is simply that the surface disrupts the flow of air.
An irregularly 'dimpled' surface is all that is requaired.

Which alll makes things a lot easier :-)

Larger shallower dimples will have the same result as larger numbers of smaller ones.

So noc - while a great script (albeit formatted by a programmer and really bloody difficult to read :whistle: )  - it's too regular :-)

But with with larger shallower dimples should be a good place to start.

lol - see this is where my -'what the hell does this do if I add some random numbers and rinse and repeat' method is better for this kind of thing.

You're script won't do large shallow dimples. The larger the dimples the fewer it generates and making them really shalow, just leaves pinpricks.

There probably needs to be a 'resize' or 'scale' in there for the dimple spheres. To keep the same surface area covered when you move them further away from the central sphere. 

Yes, I just noticed that I wrote it to NOT overlap the dimples... here's a fixed version where I added something called "overlap_constant" that allows the number of dimples to be unevenly spaced around each ring.  Try this version out instead and play around with the divot dia and overlap constant relationships and see if that gives you what you are looking for...
==================================

//=============
// Parametric Golf Ball
//   Written for Curious Aardvark
//   www.slinging.org
//
//
//=============


//User-Settable Parameters
//========================================
ball_dia = 43;

divot_dia = 7;

divot_depth = 1.25;

overlap_constant = 1.5;

//divots around the equator of a golf ball = 30
numdivots_ring = 30;
// number of rings of divots between the equator and the pole of the ball... total number of actual rings is 2*numrings+1
numrings = 6;


//Calculated Parameters - Do not change
//========================================
ball_rad = ball_dia/2;
divot_rad = divot_dia/2;
//virtual surface radius where the divot beads go
divot_surface_rad = ball_rad+divot_rad-divot_depth;

//angle between divot beads
angle_increments= 360/numdivots_ring;

$fn = 50;


module main()
{
    difference()
    {
        positives();
        negatives();
    }
   
    //This next line is just for visualization of how the divots are made... should be commended out most of the time
    //%negatives();
   
    module positives()
    {
        sphere(d=ball_dia);
    }
   
    module negatives()
    {
       for(j=[0:numrings])
       {
           working_ring_rad = divot_surface_rad*cos(j*90/(numrings+1));
           ring_distance_from_center = divot_surface_rad*sin(j*90/(numrings+1));
           numdivots_ring = floor(3.14159*2*working_ring_rad/divot_dia+overlap_constant);
          angle_increments = (360/numdivots_ring);
          rotate([0,j*angle_increments/3,0])
          for(i= [0:angle_increments:360])
          {
              //divots on right side
              translate([0,ring_distance_from_center,0]) rotate([0,i,0]) translate([divot_surface_rad*-cos(j*90/(numrings+1)),0,0]) scale([1,1,1])sphere(d=divot_dia);
              //divots on left side
             translate([0,-ring_distance_from_center,0]) rotate([0,i,0]) translate([divot_surface_rad*-cos(j*90/(numrings+1)),0,0]) scale([1,1,1])sphere(d=divot_dia);
             
          }
      
      }
      //cap the two ends with a single divot
      translate([0,divot_surface_rad,0]) sphere(d=divot_dia);
      translate([0,-divot_surface_rad,0]) sphere(d=divot_dia);
    }
}


scale([1,1,1]) main();

Curious Aardvark wrote on Sep 11^th, 2020 at 11:56am:

lol - see this is where my -'what the hell does this do if I add some random numbers and rinse and repeat' method is better for this kind of thing. You're script won't do large shallow dimples. The larger the dimples the fewer it generates and making them really shalow, just leaves pinpricks.

Whatever you have to tell yourself to avoid learning basic geometry ;D

To get large, shallow dimples, you will have to increase the number of rings too. I could make the script smarter and have it automatically figure out the right ring spacing, but I get the feeling that you and IG will both just insist on doing this your own way anyway. Well... I demonstrated that you CAN use nested for loops to dimple a ball anyway. If you want to use a different approach, I won’t stop you :)

@noc - that looks aloty better - but I think we'll have to agree to disagree on what 'readibility' means in terms of openscad :-)

I hate indents and like to keep one operation on one line, not stick each seperate command on a seperate line. And I hate really long variable names :-)

Not a programmer :-)

But the golf ball looks good :-)

well until you put divot depth at 0.25 - at which point it all goes to pot :-)

What we ned are reallly SHALLOW and wide 'divots'.
basically you need to reshape the sphere into a more flying saucer shape. to give maximum amount of 'sphere' in touch with the main sphere.

So why does everyone else use radius and not diameter ?

If I use diameter I know how big a sphere/cylinder is without having to think about it.
If you use radius you have to mentally double everything to get the actual size.

Just seems more work than necessary. 

I don't know about you noc - but I keep wanting to turn it into a deathstar :-)

Ok CA... I will play around with the parameters and see if I can give you what you want.  The requirements keep changing. First it was a dimpled ball, then the dimples had to touch, then the pattern needed to be random, and then wide and shallow. Assuming that it stops changing, I can probably modify my script to give you what you want... unless what you really want is to just do it yourself... in which case I should stop “helping” :)

Curious Aardvark wrote on Sep 13^th, 2020 at 3:27pm:

I don't know about you noc - but I keep wanting to turn it into a deathstar :-)

You'd need one heck of a sling for that, CA. ;D

Any update on this?
Personally I think this would make for the great practice ammo once fired, and could fly far as well.
Perhaps a different route should be taken? Such as manually making a mould directly from a golf ball, or taking a smooth clay ball and rolling it on a dimpled sheet?

No, stop, my Brian hurtings  xx

I haven’t touched it in a while. The code is all there if anyone else wants to take a crack at it :)