Improvements on method from Part 1:

Apparatus:
Part One had a margin of error of 40mm (though this was practically reduced, as I decided whether the ball had made a direct enough contact with the string to count), but Part 2 involved many calculations and different computer programs to find results that are accurate to 0.1 of a degree. Also, given that I have done similar test before, I have found other small ways of increasing consistency in tests (through knowing which balls to use), along with the robot (ball launcher) settings and racket position.

Around 1000 graphs were produced and analysed during How Spin Works Part 2!
  •       5 Regail brand table tennis rackets were used as tested rubber surfaces.
  •       3 star (top quality) ITTF Approved Double Fish table tennis balls were shot at the rackets.
  •       1 Double Fish Ball-Launching Table Tennis Robot fired the balls towards the racket with topspin.
  •       1 Casio camera (capable of filming at 240fps) was used in recording the data.
  •       1 90-degree brace to hold the rackets at a 90-degree angle.
  •       1 rubber band to hold the rackets to the brace.
  •       1 Metre rule was used as a scale for data analysis in Vernier Logger Pro and Vernier Video Physics.
  •        1 protractor was used to measure the throw angle (supported with calculations).

General Procedure

1. First, with an old bat, the robot was set up on full topspin and the equipment was moved around to find the optimal position for the equipment in order to record accurate data. An outline of the equipment was drawn on the table (similar to a shadow board) to ensure the equipment could be moved back into position if it was bumped or moved.
2. Once the positions were established, there were two separate tests to conduct; one for each playing characteristic.
3. All of the rackets were tested at first (without treatment) to ensure they all had similar results in both power and spin.
4. Each week, one day before tests, each racket would have exactly 50 shots played with it, to accelerate the degradation of the rubber. Immediately afterwards, the two rackets to be cleaned would be lightly wiped over with a wet sponge.
5. The tests were repeated at exactly 3:30pm every Sunday, every week.
 
​Variables:
  • Racket 1 was cleaned once a week and kept in a case when not in use (testing or 50 shots).
  • Racket 2 was kept in a case, but was not cleaned.
  • Racket 3 was cleaned, but not kept in a case.
  • Racket 4 was the control, and was not cleaned, nor was it kept in a case.
  • A fifth racket was kept in a different position, which was subject to approximately 1 hour of sunlight a day. This racket was not cleaned, nor was it in a case.
The independent variable was therefore the cleaning method, and the dependent variables were power and spin.

The Power Test

​The power test was a simple bounce efficiency test. The ball was dropped from a level platform with 300mm elevation from the top of the racket, and the rebound was filmed from two angles (to eliminate parallax error) and repeated ten times to find the average rebound height (outliers omitted). This was repeated for each racket.

The Spin Test

​The spin test was a much more involved test. First, the robot was set up to full topspin, while the racket was attached to a 90-degree base by rubber band. The robot would then fire ten shots towards the racket, while the camera filmed its rebound. The rebound was then analysed to calculate the throw angle. Note the throw angle (a table tennis term) is the angle of reflection (as opposed to the angle of incidence, which was kept constant). A racket which has an angle of reflection greatly affected by incoming spin (due to friction – i.e. more spin) is said to have a high throw. The ten rebounds were then averaged out (outliers omitted) to find the average angle of reflection.

Calculation Procedure:

The final results for each test followed this procedure:
  1. The video (with locations points recorded) was uploaded into Vernier Logger Pro.
  2. (Power): The scale was then set by the height of the ruler equalling (x) amount of pixels. The rebound height in pixels was then automatically calculated by the software to give a result in millimetres.
  3. (Spin): The position-graph would have its axes manipulated so that the x-and y-axes were equal (so the ratio was 1:1 – giving a scale representation of the actual rebound) and the angle of incidence was recorded with a protractor, as was the angle of reflection. The angle of incidence was then subtracted from the angle of reflection to find the throw angle (the increased angle due to spin).
  4. To confirm the results of the spin test, the coordinates of several points were recorded. The distance between points were then calculated through Pythagoras’ Theorem, and then by the use of Trigonometry (the Cosine rule – (side, side, side)), the angles of the triangles were found.
  5. The ten angles of the rebound were then averaged (outliers omitted) to find the final result for each bat every week.