Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

How do you create a perfect sphere?

  1. Jul 13, 2013 #1
    the formula for the surface area of a sphere is SA = 4 (pi) r2, with pi = 22/7 and r = radius of the sphere.

    for example the SA for earth with a radius of 6,378 km is 510,065,600 km2

    what would the radius be in order that for you to lay a grid of perfect square on the surface of the sphere?
  2. jcsd
  3. Jul 13, 2013 #2

    Vanadium 50

    User Avatar
    Staff Emeritus
    Science Advisor
    Education Advisor
    2017 Award

    There are no perfect spheres in nature.
  4. Jul 13, 2013 #3
    but how do you create one?
  5. Jul 13, 2013 #4

    Vanadium 50

    User Avatar
    Staff Emeritus
    Science Advisor
    Education Advisor
    2017 Award

    You don't.
  6. Jul 13, 2013 #5


    User Avatar
    Gold Member

    No, not really. Close but no cigar.
  7. Jul 13, 2013 #6
    Here's the closest you'll probably get to a perfect sphere. The video has a lot of superfluous, albeit interesting information. How the sphere is made is in the video around the middle.

  8. Jul 13, 2013 #7

    D H

    User Avatar
    Staff Emeritus
    Science Advisor

    And even if you do manage to make a perfect sphere, you can't lay a grid of perfect squares on it. Squares live on planes, not spherical surfaces.
  9. Jul 14, 2013 #8


    User Avatar
    Gold Member

    pi = 22/7 will not get you a “perfect sphere”, 21.991148575128552669238503682957/7 will get you closer.

    How do you perfectly match a square on a circle? You are trying to do the impossible...
    View attachment 162790

    A square on the Euclidean plane has an internal angle of 90° at all four vertexes.
    View attachment 162791

    But when you apply a square on a sphere you get 120° internal angles.
    View attachment 162792
    This is not a “perfect square”...

    You can use deltoidal icositetrahedron, but your “perfect squares” are lost again.

    View attachment 194677

    The best thing to do is triangulation, and high resolution on vertexes (=many). It will get you close, but never perfect.

    View attachment 194678
    Last edited by a moderator: Apr 14, 2017
  10. Jul 14, 2013 #9


    User Avatar
    Staff Emeritus
    Science Advisor
    Gold Member

    To determine if a sphere is perfect, you need some way to measure it. That measuring device has some uncertainty associated with it (all such devices do). That means the uncertainty of the measuring device has to be greater than perfect. How is that possible? It isn't.
  11. Jul 14, 2013 #10
  12. Jul 14, 2013 #11


    User Avatar
    Gold Member

    Agree lisab, the only “perfect” certainty in nature is Heisenberg's uncertainty principle... :smile:
  13. Jul 14, 2013 #12


    User Avatar
    Science Advisor
    Gold Member

    Even non-spherical shapes are not easy.

  14. Jul 14, 2013 #13
    thanks everyone for helping :)

    so even if the radius of sphere transforms pi (22/7) into a perfect number, i won't get a perfect sphere?

    if the radius is 3500 km, the surface area would be 154,000,000 km2...these could be perfect square each measuring 1km on each of its four sides? so if i wrapped these into a sphere, some of you saying i cannot do so?

    i was thinking of that whether on a curved plane or flat plane, if the grid defines perfect squares on these planes then whether you walk on the curved plane or flat plane the distance would be the same. but some of you said if i transform a flat plane into a curved plane, then the angle of the square changed?
  15. Jul 14, 2013 #14


    User Avatar
    Gold Member

    No worries mate :wink:

    Nope, 7 x pi is not 22 but a “never ending mess of imperfection”...

    Ever heard of polar bears and this guy?? :smile:


    And with 1 km2 you will get trouble everywhere else too. See the bend in the water horizon in northern Wisconsin?


    Trust me, same mess everywhere! :biggrin:


    Yup, the Euclidean plane does not work on spheres. However if you make your “perfect square” tiny, let’s say 1 cm2, the ‘distortion’ is of course less. But you will never make a “perfect sphere” from “perfect squares”... impossible.
  16. Jul 14, 2013 #15
    If someone's definition of pi is 22/7, then maybe they don't have such strict criteria as to what constitutes a "perfect" sphere.
  17. Jul 15, 2013 #16
    could you please elaborate your version of "pi" ?
    much appreciate the knowledge i could gain from you :)
  18. Jul 15, 2013 #17
    22/7 goes wrong already in the third decimal place, which is why it's not a popular approximation. 3.14159...never goes wrong, it just never ends. It keeps getting closer and closer to pi without ever landing precisely on it.

    When I worked in the machine shop we rounded pi off to 3.1416, which was as close as we ever had to be for any practical purpose. 22/7 or 3.1428, might easily have gotten us into a situation where parts didn't fit what they were supposed to fit.
  19. Jul 15, 2013 #18
    thanks for sharing your shop experience, i just didn't realize that in our physical world that machines parts would need to be in the third digits or more after the decimal ...but good to know :)

    the video of the si-28 sphere is interesting :)
  20. Jul 19, 2013 #19
    if i have a ruler that measure 1m on a flat surface, if i bend it to measure a ball, the 1m does not change. or i lay a grid system with many perfect squares of each side being 1m on a flat surface, the length of the sides on these perfect squares will change their sides other than 1m each?
  21. Jul 20, 2013 #20

    Jonathan Scott

    User Avatar
    Gold Member

    If you do that on a ball, a shape with four equal "straight" sides cannot have right angles at all four corners. If you try to force the corners to be right angles, the last corner will not meet.
Know someone interested in this topic? Share this thread via Reddit, Google+, Twitter, or Facebook

Similar Discussions: How do you create a perfect sphere?
  1. How you doing Russ? (Replies: 25)

  2. How do you pronounce it? (Replies: 11)

  3. How do you think? (Replies: 4)

  4. How do you think? (Replies: 29)