Tension loss experiments / 1-piece vs. 2-piece and others

Discussion in 'Badminton Stringing Techniques & Tools' started by s_mair, Jul 12, 2021.

  1. s_mair

    s_mair Regular Member

    Joined:
    Oct 10, 2013
    Messages:
    5,366
    Likes Received:
    4,152
    Location:
    Germany
    Part I: 1-Piece vs. 2-Piece stringing pattern

    Preface

    One topic that pops up on BC on a regular basis is the question, if certain stringing patterns perform better or worse than others in terms of tension loss over time. A statement that comes up is that using 1-piece patterns have a bigger tension loss than 2-piece patterns (or vice versa), last seen in this thread: https://www.badmintoncentral.com/fo...-more-i-can-do-to-reduce-tension-loss.188003/

    And so far, there has been a lot of opinions, impressions, myths and discussions, but very few to zero actual facts that might shed some light on the question. That’s the reason I’ve started my little experiement.


    Experiment setup

    Two identical rackets (Victor Super Waves 36) were strung with the same string (Adidas Wucht P68, 0.68 mm gauge) directly back to back by the same stringer (myself) using the same machine (StringMaster Deluxe with WISE tension head), the same settings (12 x 12 kg / 26.4 x 26.4 lbs., no pre-stretch) and the same process flow details (no added tension before tie-off knots, no double pulls).
    The only difference was the pattern used.

    Racket #1 (white strings): 1-Piece Haribito Professional “Simon Mod”
    Racket #2 (red strings): 2-Piece pattern “AK-style”.
    So basically a standard Yonex pattern with @kakinami’s special twist in terms of knot positions

    The reason for using these two patterns was simple – those are the patterns that I am most comfortable with and hence have the lowest risk of making mistakes or delivering a weaker result due to the lack of experience with the used pattern. I’ve strung both rackets to the best of my abilities without any rush, so both jobs took about 30 minutes.

    Rackets.jpg

    Measurements

    Since I do not own any gear that can measure the actual dynamic tension of the stringbed (like this one for example: https://www.stringway-stringing-mac...-stringlab-2-stringbed-test-badminton-tennis/), I went with the frequently used measurement of the dominant stringbed resonance ping frequency, measured with the app CarlTune on an iPhone 12.

    The measurements were done once every day for one month, the first one was done right after finishing the string job.

    Both rackets have been stored right next to each other throughout the experiment and both rackets have been played simultaneously during warm-up and always “against each other” by players at the same playing level. Players have switched the rackets halfway through warm-up.

    It’s of course possible that one of them has seem a couple of shots more or with a slightly different distribution of hitting power, but it should not be enough to make a real difference in the overall results.

    The storage location of both rackets was changed several times during the second half of the experiment period. Reason was that I got a bit bored and wanted to see how much influence things like temperature and humidity changes might have on the ping readings. And let me spoiler already that this has produced some very interesting results that go beyond the original target of this experiment. More on this later on.

    Side notes:
    This is not a direct measurement of the string tension but it should still produce usable date for the comparison which is needed. I have done ping measurements for every string job over my first years as a stringer and I would say that I have gained serious experience in how to do this correctly (in case anyone needs more information on ping measurement, just click here: https://www.badmintoncentral.com/fo...d-frequency-to-monitor-string-tension.124011/).

    It turned out that with the current version of CarlTune, it has become a lot more difficult to get a usable reading since the app has become a lot more sensitive and jumpy. However, I think that I was able to get usable results on all measurements. I would estimate that there is an inaccuracy of round about +- 2 Hz involved in every measurement. I did my best to middle out the readings by experience.


    Results

    Here are the graphs for both rackets. L1 to L4 show the moments when the rackets have been moved to the different storage locations and conditions (T: temperature, RH: relative humidity):

    upload_2021-7-12_10-29-3.png

    The original spreadsheet with all raw data can be found here:
    https://www.dropbox.com/s/h71b3zlwcgm13wq/Tension Loss.xlsx?dl=0


    Conclusion

    Both pings have behaved pretty much identically, no matter the external influences.

    There is no significant difference to be seen in terms of tension loss over time between 1-piece and 2-piece stringing patterns.


    Looking at the data points, racket #1 had a slightly lower starting point (1127 Hz / 1130 Hz) and an ever so slight lead in the end (1109 Hz / 1107 Hz). Even if we ingored the fact that this is most likely within the range of measurement inaccuracy, those 5 Hz “loss” would translate to a difference of 0.17 lbs. if we follow the rule of thumb that 30 Hz difference equals 1 lb. in tension. So completely negligable.

    Based on the data gathered in this one experiment, it doesn’t matter if a racket has been strung with a 1-piece or 2-piece pattern when it comes to the amount of tension loss afterwards. I don’t see any reason why results would be noticeably different with using other 1- or 2-piece patterns, as long as they have a comparable amount of slack string running around the outside of the frame.

    There will surely be differences if we compare string jobs from different stringers and/or different machines, tension ratios, rackets or the strings used.


    Summoning @kwun, @kakinami, @thyrif, @Dekkert, @speCulatius, @Super85, @Antxn, @Mark A, @khoai, @akatsuki2104, @Alex82
     
    #1 s_mair, Jul 12, 2021
    Last edited: Aug 13, 2021
    DarthHowie, I3ugs, emjay and 18 others like this.
  2. s_mair

    s_mair Regular Member

    Joined:
    Oct 10, 2013
    Messages:
    5,366
    Likes Received:
    4,152
    Location:
    Germany
    ...and now for the nerds

    If we look at the data and beyond the initial question of this experiment, we see quite a lot of movement in the ping frequencies under different storage conditions. And some of them were surprising. I have made a second graph only containing the values measured after moving the rackets to different storage locations.

    I have also added three more data points where I covered the stringbed with a soaked towel over night and then letting it dry up in the stokehold again for another day. Since those values would have been distracting for the first part of the experiment, I have excluded them there.

    upload_2021-7-12_11-0-8.png


    Temperature and humidity
    We see that the climatic conditions appear to have a significant effect on the ping frequency. And let me remind you again that we are not really measuring the stringbed tension or stiffness, but merely the dominant resonance frequency which is influenced a lot by the diameter and other properties of the string.

    In the first half of the whole measurement when the rackets were stored at the same place with fairly constant climatic conditions, there have been very small day to day variations. However, there is a lot more movement as soon as the rackets were moved around.

    What was really stunning to me was the switch to the cool and humid basement. I had expected the frequency to become higher due to the cooler temperatures and the shrinkage of the string, but the numbers showed that the exact opposite was true.
    Moving the rackets to the warm and dry stokehold for a day, the values jumped back up again significantly which meant that it was not a random behavior or a misreading the day before.

    After discussing this discovery with my trusted consultant @speCulatius, we came up with the following possible explanations:

    1) Thermal expansion/shrinkage of the frame
    2) String absorbs humidity and swells, leading to an increase in diameter

    Since carbon fiber compounds show rather low thermal expansion/shrinkage, we eyed more towards humidity playing a major role here.


    Plastic and water absorption

    Different types of plastic material can absorb different amounts of water. According to Adidas promo material, the core of the P68 string is made of Polyamide (also known as Nylon) and according to https://www.iket-horb.de/2019/12/12/quellen-polyamid-feuchtigkeit/, (only available in German, sorry) PA can absorb around 3% of its own weight and tends to swell in size while doing so.

    As said before, the diameter of the string is a major factor when it comes to its resonating behavior. So if the string really absorbs water and grows in diameter, this would be a possible explanation of the seen ping frequency changes.

    I decided to maximise the effect before the end of the experiment by putting a wet towel on the stringbed over night to allow the string to absorb as much water as possible. And the results were stunning as the ping dropped a whooping 50 Hz and jumped back up again after a day back in the stokehold. I cannot tell for sure since I haven’t done any diameter measurements, but there is a strong indication that the change in thickness is the main reason for these ping changes there.


    So does high humidity causes tension loss?

    Unlikely. The only thing that this experiment has shown, is that the ping frequency drops noticeably with a raised humidity and pretty much completely recovers as soon as it gets dry again. Even if it might sound like the tension has dropped as well, the stringbed didn’t feel any different when pushing it with my hands – even after basically drowning it in water.

    As said before, the absorbed water most likely causes the Nylon fibers to swell and increase the diameter of the whole string. And this causes the lower ping frequency. The stringbed tension has not changed.

    But since the human perception is very flawed, I can easily imagine that a lower pitched hitting sound might be perceived by the player as a lower tension, even if the effective tension has not changed. Although it is unclear if the change in humidity causes a change in the repulsion behaviour of the string.


    Bottom line

    There are a number of simple factors that might lead to the impression, that stringjob A loses more or less tension over time than stringjob B. The impact of the stringing pattern appears to be somewhere between negligable and nonexistent.

    And since this experiment has turned out to be a lot of fun, I guess I will repeat it to see how much difference the use of pre-stretch will have on the tension loss. :)

    All you crazy BC stringing nerds out there, your remarks, comments and follow-up discussions are of course highly welcome!
     
    #2 s_mair, Jul 12, 2021
    Last edited: Jul 12, 2021
    DarthHowie, Goggles, I3ugs and 14 others like this.
  3. akatsuki2104

    akatsuki2104 Regular Member

    Joined:
    Nov 15, 2007
    Messages:
    400
    Likes Received:
    222
    Location:
    Europe
    As the sound moves/vibrates through the air, could the humidity and density of the air cause the ping frequency to change? Thus, as you said, that it's not really the tension that has dropped but just a sound difference.
     
  4. s_mair

    s_mair Regular Member

    Joined:
    Oct 10, 2013
    Messages:
    5,366
    Likes Received:
    4,152
    Location:
    Germany
    Nah, that's not really likely. The wet-towel-measurement was done with having the normal ~50% humidity around. Only the strings were soaked.
     
    akatsuki2104 likes this.
  5. kwun

    kwun Administrator

    Joined:
    Apr 24, 2002
    Messages:
    41,044
    Likes Received:
    2,066
    Occupation:
    BC Janitor
    Location:
    Santa Clara, CA, USA
    Brilliant. absolutely brilliant.
     
    DarthHowie, kakinami and speCulatius like this.
  6. Mark A

    Mark A Regular Member

    Joined:
    Sep 2, 2005
    Messages:
    7,170
    Likes Received:
    695
    Location:
    St Helens, UK
    Nice experiment and write-up, Simon ;)

    The results don't surprise me in the slightest - how much tension is lost on the last string when it's tied off? The clamp might jump back a couple of mm... the string used for the crosses is five metres long. By the time the knot-loss has propagated through that lot, it's going to be literally tens of grams per string. And if the cross tie-on knot is under active tension, there's no loss there at all.

    Yet one more confirmation that patterns make sod-all difference - just get the string in the frame, with good practices, and play with it.
     
  7. khoai

    khoai Regular Member

    Joined:
    Sep 30, 2008
    Messages:
    263
    Likes Received:
    89
    Location:
    US
    Oops, double post.
     
    #7 khoai, Jul 12, 2021
    Last edited: Jul 13, 2021
  8. Sumanth99

    Sumanth99 Regular Member

    Joined:
    Oct 19, 2018
    Messages:
    1,949
    Likes Received:
    977
    Occupation:
    Software Engineer
    Location:
    Hyderabad
    Written like a research paper, if there are journals for related studies I am sure this will be valued with high regard.
     
    kakinami and s_mair like this.
  9. s_mair

    s_mair Regular Member

    Joined:
    Oct 10, 2013
    Messages:
    5,366
    Likes Received:
    4,152
    Location:
    Germany
    I’m totally with you on this. Stringing needs a lot less fairy dust and magical formulas as it needs proper and consistent craftsmanship.

    It’s so tempting to overthink things and to get completely lost in the details instead of focusing on the basics.
     
  10. Mark A

    Mark A Regular Member

    Joined:
    Sep 2, 2005
    Messages:
    7,170
    Likes Received:
    695
    Location:
    St Helens, UK
    I'm just waiting for somebody to say that the test wasn't fair because the strings weren't the same colour.
     
  11. kakinami

    kakinami Regular Member

    Joined:
    Jan 26, 2005
    Messages:
    1,039
    Likes Received:
    703
    Location:
    somewhere
    Nerd <3

    Sent from my LG-US998 using Tapatalk
     
  12. kakinami

    kakinami Regular Member

    Joined:
    Jan 26, 2005
    Messages:
    1,039
    Likes Received:
    703
    Location:
    somewhere
    The test wasn't fair because the strings were not the same color. ( sorry I am American we spell it wrong ) =P

    Sent from my LG-US998 using Tapatalk
     
    flyingcords, s_mair and speCulatius like this.
  13. visor

    visor Regular Member

    Joined:
    Dec 7, 2009
    Messages:
    16,401
    Likes Received:
    2,001
    Location:
    Vancouver, BC
    Nicely done! How about playability of the rackets as determined by your players, in terms of power, feel, accuracy etc? Which one do they prefer?

    Sent from my SM-G988W using Tapatalk
     
  14. s_mair

    s_mair Regular Member

    Joined:
    Oct 10, 2013
    Messages:
    5,366
    Likes Received:
    4,152
    Location:
    Germany
    They (which was my usual practise buddy and myself by the way...) didn't feel any significant difference. At least during those couple of minutes warm-up, both were doing their job (the rackets I mean). There was no aha moment when switching between them.

    My buddy had a little aha moment though when he hit his first smash and the shuttle almost punched a hole through my calf. He never believed me, that a higher tension can have a positive effect on power when you have a matching swing for it. So I see a certain chance that he will go up from his usual 22 lbs. next time for his other rackets as well. :)
     
  15. Mark A

    Mark A Regular Member

    Joined:
    Sep 2, 2005
    Messages:
    7,170
    Likes Received:
    695
    Location:
    St Helens, UK
    At least you acknowledge the fact that you spell it wrong.
     
    kakinami likes this.
  16. flyingcords

    flyingcords Regular Member

    Joined:
    Jan 17, 2015
    Messages:
    235
    Likes Received:
    65
    Location:
    france
    I am quite sure that s_mair is thinking about re-doing all the test using 3 pieces of string taken from the same reel ;)
     
  17. s_mair

    s_mair Regular Member

    Joined:
    Oct 10, 2013
    Messages:
    5,366
    Likes Received:
    4,152
    Location:
    Germany
    That was such an easy bait I laid out there, I was sure that the BC Sherlocks would be jumping right at it. :D

    For the pre-stretch yes vs. no experiment I'll go with Gosen GT5 strings from the same reel.
     
    speCulatius likes this.
  18. Mark A

    Mark A Regular Member

    Joined:
    Sep 2, 2005
    Messages:
    7,170
    Likes Received:
    695
    Location:
    St Helens, UK
    If you're doing prestretch vs no, you can pre-empt my objection by adding a hand pre-stretch.

    I always pre-stretched by hand because 34 + 10% would have been... I was about to say "mental", but "even more mental" would be more accurate.
     
  19. s_mair

    s_mair Regular Member

    Joined:
    Oct 10, 2013
    Messages:
    5,366
    Likes Received:
    4,152
    Location:
    Germany
    If I had three rackets available, I would include a hand pre-stretched one as well. So for this time, there will only be none and machine pre-stretch to be compared I'm afraid.
     
  20. Antxn

    Antxn Regular Member

    Joined:
    Feb 7, 2016
    Messages:
    106
    Likes Received:
    11
    Location:
    Singapore
    Nice experiment!

    Sent from my VOG-L29 using Tapatalk
     

Share This Page