It is impossible to measure the real tension in a racquet exactly. In this project, however, the two middle main strings where monitored with special load cell (more below) that could determine the tension with high accuracy while the racquet could still be used to play with. The results were compared with RacquetTune. This project has been discussed here and here.
The tension in a strung racquet
The racquet was strung with a Kirschbaum Proline II 1.25 mm (polyester) string at 23 kg. The tension was monitored until the strings broke after five weeks (800 hours) and 8 hours of playing:
One of the main reasons for doing this test was to see how well RacquetTune was able to monitor the tension drop. Since the load cell and the app were measuring two different things (two strings vs an overall tension) the absolute value would probably be different, but the trend should hopefully be the same. So taking the same data as the figure above and including the RacquetTune results gives the figures below. Both curves show the same data, but in the first one the x-axis is the time in log-scale, which makes the beginning more clear and also makes it easier to compare the trends.
This is a very satisfying result and shows that RacquetTune can be relied on to monitor the right trend of the racquet. That the tension of the two main strings is lower than the overall tension is to be expected. A further illustration of the good correlation can be given by plotting the two tensions vs each other:
The set up
The load was measured with specially designed load cell that sits in the racquet. This made it possible to detect a tension variation of 0.01 kg. It was calibrated in the racquet with another load cell/scale with a +- 20 g accuracy.
The metal bar is made of high grade (airplane) aluminum to get a linear stress – strain curve. It weighs 15 g, so the increase in racquet weight is acceptable and since it sits close to the center of gravity it adds very little to the swingweight and balance. The racquet is a Wilson kTour Team which is very suitable since it has grooves in the shaft that can support the tension bar.
The bar has 4 strain gauges applied symmetrically on both sides and coupled as a wheatstone bridge to minimize the effect of temperature and uneven load. They are supplied with 10 V which gives a signal of about 10 mV for a 40 kg (20 kg per string) load.
To calibrate it was placed it in the racquet and the racquet in a string machine. A piece of string was inserted into the bar and connected to a reference load cell/scale sitting in the middle of the racquet. The other side of the load cell was connected to the tensioner of the string machine. The string machine was therefore only used to apply the tension, while the load was measured more accurately. Different loads from 16 kg to 44 kg (8 – 22 kg per string) were applied, it was kept for at leasts 30 s to settle, and the output voltage was measured. 25 measurements were made and fitted to a straight line that gave the relation between output voltage and tension.