@TW Professor @TW Staff I was looking in the TWU TWISTWEIGHTS Table (http://twu.tennis-warehouse.com/cgi-bin/twistweight.cgi) and I see different TW values for for the RF97 rackets than the results I myself found for these rackets (both the old and new models were 13.0.) I was wondering if you could share the method you use to compute TWs? I pivot the racket side to side (for TW) and front to back (for SW) from a 10 cm axis and subtract the two and that's my TW. I use the SW calculator to input my parameters. My scale is 500 g +/- 0.01 so I multiply the weight by x10 when I input it. I divide the resulting SW by 10 to get my result in tenths. I prefer to not use a hang string at the top of the racket because it just introduces more error in calculating TW and SW. If all I'm measuring is the distance from 10 cm to COM I think there is less error. Of course my scale and timer introduces some error too. Thanks in advance, Irvin

A really excellent idea. Thanks for the idea, It is simple enough to make a bifilar pendulum to measure TW with little or no friction. Wow I'm going to play with this. You're my hero!!! EDIT: I wish there was a LOVE button. Not for you but for your idea.

Hah, in a roundabout way you can thank TW Professor. I read about it first in this article which was predominantly about the book he co-authored The Physics and Technology of Tennis

This might take a couple more weeks to answer as the TW Professor is out of the office for the holidays and work. Thank you for your patience! Thanks, Brittany, TW

I wanted to know how TW Professor calculated TW. I have a method but wanted to know how they came up with the figures. EDIT: Getting the TW readings down to 1/100th kgcm^2 accurately is pretty hard. An RDC only get the TW down to +/- 1 kgcm^2 which is not very accurate at all my method gets it down +/- 1/10th.

Thanks Irvin. So I guess there is not a good way for someone to calculate an accurate twist weight. I have a frame spec'd out to what I like and have a few old frames I would like to make the same, but they aren't on the twist weight chart TW provides.

I'm pretty sure the method I use is pretty accurate. See post #15 in this thread https://tt.tennis-warehouse.com/index.php?threads/need-help-customising-ps97.578751/#post-10828460

First I use an ice stick / hobby stick and cut it into four pieces as you see marked. I actually cut the two long pieces a little shorter. Then glue it together as you see in the bottom with the rounded corners inside to reduce friction. I sanded down one edge so a utility wire can be placed in the stick 1 cm from the top. I marked it with a 1 so it's always placed in the same position. I use a measure to place the top of the stick 9 cm down from the butt of the racket so the pivot is at 10 cm. I then mount the racket on some cup hooks with the utility flag wire through the stick with the stick on the top/bottom bevel so the stringBed swings parallel to the swing path on the 10 cm axis and time the swing for 10/20/30 complete swings. This is going to be the TW time. I then rotate the stick around the racket 99 degrees so it's on a side bevel and the stringBed moves perpendicular to the swing path and time 10 swings. This is the SW time. After a little practice timing you can get very consistent results. I use a scale that weighs in 1/100th g and input the results in TWU's SW calculator. I multiply the weight by 10 so 336.78 g becomes 3367.8 g. Then I decide the result by 10 so 3658 Kgcm^2 becomes 365.8. I do this so I can narrow down the TW to tenths. Same thing for SW, then subtract the SW reading from the TW reading to get the actual TW and not a combination of the TW and SW. Hope that helps.

This thread came to mind this morning, it looks like we still don't have info on how TWU calculates, but I was thinking their specificity of 1/100th may be a result of averaging several passes.

I read this article and tried to measure the twistweight of one of my racquets using the formula and method described in it. The numbers I plugged in seemed fine, but I got a really weird result of 4.2 kg cm^2. I was considering the constant they use (40.28), and was wondering if you knew what it meant. According to my calculations, its units should be sec^2/cm which I also found weird. Also, all the units I used were metric. I considered that maybe different units should be plugged in but those also gave rise to weird results. Basically, if you could help me understand this formula better that'd help. I'm really confused right now with the results in getting.

I wish I could help, but I only found the formula from what appears to be a credible source. I've never done the work myself. I have done the estimation formula but that only accounts for racquets with uniform mass distribution, which are less common today than then.

It may help you to look at the estimation formula at the bottom as they may inform if your values are correct as they use some of the same variables. If I remember right I used values on the chart provided and worked backwards to verify the values I used were right.

The formula described here: http://www.tennisindustrymag.com/articles/2005/02/a_new_twist_on_the_twistweight.html

That is a terrible method. More often than not you have to add a 2 oz bolt near the center two string at the head to get the balance point between the two point you're using to swing your racket. And the string you use to swing the racket on must be parallel to the racket's centerline.

Fair enough. I did have to use a couple large coins at the tip, but that shouldn't change the twistweight by much.

Because the TW is calculated on the centerline and inertia is mdd mass is insignificant is the distance is low. You can actually compute the inertia of the coins and subtract that out if you want.