Announcement

**stumphges** · 08-27-2011, 09:54 PM

Hi Jim,

Thanks for the feedback!

Good questions that require long answers. So here goes, in two parts:

1) Studies seem to look only at groundstrokes, where the upswing is straight forward, pushing mains down.

What is the influence on/ best string set up to develop spin on the seemingly-more-complex service swing?

First of all, the effect of a string setup that is particularly good for spin is less noticeable on serve than on groundstrokes. The reason why is kind of complicated.

We know from lab tests that Luxilon Alu Power, for example, provides about 30% more outgoing spin (after it hits the racquet) on a simulated groundstroke in the lab than a typical syngut (think Prince synthetic gut original or Gosen OG Micro).

Having read the more technical studies, you know the following, but for those TP subscribers that haven't...

The way that groundstrokes are simulated by Cross and Lindsey and the guys at ITF is like this: You fire a ball at a clamped or suspended racquet at a certain speed. That fired ball also has lots of backspin on it, say 3000 rpms. Why backspin? Because a ball that bounces off the court with topspin is only topspin from the "court frame of reference". From the racquet's frame of reference the ball is approaching the stringbed with backspin on it. It's hard to get your head around this, but that's how it works.

Anyway, the ball comes in with backspin and hits a stationary racquet that is angled to the line of the incoming ball such that it simulates the impact angle of a topspin groundstroke. When all these parameters are set, we can use equations to figure out exactly what kind of groundstroke we're simulating. For example, we could set the incoming ball speed, ball spin and impact angle such that it simulates a shot with these characteristics:

Incoming ball speed: 30 mph (after bounce)
Incoming ball spin: 3000 rpm topspin

Racquet speed at impact location: 70 mph
Racquet swing angle: 40 degrees upward
Racquet face angle: 90 degrees (vertical)

With these parameters we'd get:

Outgoing ball spin with syngut: 1300 rpms
Outgoing ball spin with Alu Power: ~1690 rpms (+30%)

So with Alu Power we get 390 rpms of extra spin - 30% more outgoing spin - on our shot as compared to the same shot hit with nylon.

But notice one thing. When a ball comes in with topspin and we hit it so it goes back out with topspin, we're changing the direction of the spin. If we look at it from that perspective - how much we're changing the spin - the effect of copoly strings is much less:

Syngut:

incoming spin: -3000 rpms (because topspin coming in is backspin to the racquet)
outgoing spin: 1300 rpms
Change in spin: 4300 rpms

Alu Power:
incoming spin: -3000 rpms
outgoing spin: 1690 rpms
Changing in spin: 4690 rpms

We're still getting 390 more rpms, but the dhe difference in how much the spin has changed from incoming to outgoing is only 9% (4300 vs. 4690). The difference between syngut and Alu Power, in terms of outgoing spin, which is what we care about with groundstrokes, is 30% - very impressive. But the difference between the strings in terms of how much the spin has changed from incoming to outgoing is only 9% - not so much.

One of the reasons tennis researchers had so much trouble with the idea that copoly was giving players so much spin on the court is because, until 2006, they were doing their lab tests without incoming spin.

Instead of firing a ball with a realistic 3000 rpms of topspin they would fire a flat ball and watch it rebound off an angled racquet. When they did this the difference between a syngut and copoly string would be so small that they questioned its significance.

Let's say we fired a ball with zero spin onto an angled racquet strung with syngut and the outgoing ball bounced off with 1300 rpms of spin.

The change in spin is 1300 rpms. (0 to 1300 rpms)

Now we fire a ball with zero spin onto the same racquet at the same speed but with Alu Power instead. We know that the change in spin will be about 9%.

So we can estimate the change in spin to be 1378 rpms. (0 to 1378 rpms)

It's easy to see how the difference between strings seemed much less significant under test conditions where there was no incoming spin. The change of spin is not so great, with incoming spin or without incoming spin. But when the total change in spin is very high - like we see when trading topspin groundstrokes - the difference in outgoing spin is very significant.

In the first article in the TP series on spin we saw that the big breakthrough in understanding copoly came when a Japanese researcher, Y. Kawazoe, used a 10,000 fps camera to capture the actual string-ball interaction. But Kawazoe also did something else that made a difference. He had read the ITF studies on zero spin rebounds that showed very little difference between strings. He didn't buy it, but he didn't really know why they were getting those results either. He suspected it might have something to do with how they were setting up their lab experiments.

So instead of filming impacts with clamped racquets he filmed actual shots out on the court. There, he had incoming topspin and outgoing topspin. So when he measured the outgoing spin he was seeing that big difference in outgoing spin, not the small difference in the change of spin.

That's a very long-winded way to explain why, on serve, where the ball is hanging in the air with zero spin, the difference in outgoing spin is not so noticeable. We might get 9% extra spin with Alu Power on serve, rather than the extra 30% we'd expect on our topspin groundstrokes.

The second part of your first question:

What is the influence on/ best string set up to develop spin on the seemingly-more-complex service swing? For example, on this site's great slo mo of Fed's serve, you see his forearm motion causing the strings to go horizontally across the back of the ball (so depression-rebound would be much like that of a forehand), yet at the same time his arm/ shoulder/ body are throwing the racquet upwards, so the topspin component comes from the cross-strings hitting at an angle (I assume -- correct me). How the heck does that work?

I think you're right in interpreting the film to show that, on kick serves in particular, the cross strings are coming across the ball and interacting with it in a way that the main strings do on a topspin groundstroke.

And I don't think it's just on serve that this happens. Both Nadal and Federer get significant sidespin/spiralspin on their topspin forehands, especially on balls hit below the hips and balls hit above the shoulder. You can see on John's high speed footage that on low balls these guys are hitting up and across the ball in such a way that they put outside-in spin on it. And on high balls they put inside-out spin on it.

Nate Ferguson told me that Roger's gut mains rarely get stuck out of line. But that the Alu Power Rough crosses will get stuck up from hitting inside-out forehands. If you think about Fed's "fall-away" inside out forehand, what Bailey calls a lateral transfer, if he's hitting it at shoulder height or above his racquet is coming across the ball from outside to inside. His cross strings would be hitting the ball and getting pushed "up" towards the tip of the frame. If they snapped back in time they would put extra sidespin and/or spiralspin on the ball (the proportion of each being a function of what location on the ball is struck).

But we know from Lindsey's friction experiments that copoly doesn't slide on gut as well as gut slides on copoly. So Fed's gut mains - sliding on copoly crosses - rarely get stuck out of place, even though he's hitting topspin groundies continuously. But his copoly crosses, when stretched in the same way mains usually are on these shots where he is coming across the ball as much as over it, don't snap back as well because of the greater friction encountered by copoly trying to slide on gut "rails".

As long as his copoly crosses snap back into line he'd be getting maximum sidespin on these types of shots. But if they are getting stuck out of place he'd be losing spin. Nadal's RPM Blast is probably better for sidespin and spiralspin because the mains and crosses are the same material, so sliding is equal in both directions. Maybe more importantly, he gets a consistent response on these shots because the strings always come back into line, so on the next shot they are ready to slide and snapback again. Federer might hit one sick sidespin forehand, and then the next might not bend as much because his strings have got stuck.

**stumphges** · 08-27-2011, 11:10 PM

Response to Jim's second question:

2) Re: Open/Closed string patterns. The old bromide is: Open for spin velocity, closed for control. But does that really make sense with the new, hybrid/ loose strings? i.e. if you want "more strings rebounding' to paraphrase the article, doesn't a closed pattern do that -- thus more strings would be pushing up the back of the ball yielding more topspin?

This is a really good question and I think it could either way - depending on the string tension.

Usually, top players would increase tension if they switched from an open pattern racquet to a closed pattern. They would want the stringbed stiffness to be the same. If they did this the resistance to sliding - friction - might end up being equal between the closed and open pattern, but maybe not. The open pattern, even with each string being at a higher tension, has fewer string intersections. So if each main string had only 18 cross stings to slide across, rather than 20, they might slide more easily, even if they are tighter.

The cross stings being tighter - and therefore stiffer - might also do a better job acting as "rails" in this situation. In the last article in the series, I mentioned Cross and Lindsey's hunch that gut might be poor cross because its fibers stuck out and grabbed the main as it tried to slide. Another reason that Lindsey mentioned is that the gut, being much softer than copoly, would deform in front of the sliding main string, forming a little gulley with a hill in front of it. So as the main string tried to slide across the gut, it would be pushing that little hill as it went.

However, if the cross string is tighter and stiffer, that hill would presumably be smaller and resist the progress of the sliding string less. I guess this is the 3D aspect of string-on-string friction.

And that brings us to string stiffness in regards to the main strings. In his experiments with extremely open string patterns (http://twu.tennis-warehouse.com/lear...dstiffness.php), Lindsey found that extremely open patterns increase spin dramatically, but only at high tensions.

Depending on the type of string, he found that in 16x19 patterns, sometimes spin would be better at 30 pounds and sometimes better at 60 pounds. But with the extremely open, 16x10, pattern, spin was awesome at 60 pounds, but miserably low at 30 pounds.

His explanation is that - at 30 pounds in the very open pattern- the strings moved too far, and didn't have enough time to snap back before the ball left the strings. We would expect the string-on-string friction to be the lowest in this 16x10 pattern strung at 30 pounds. The strings would be pushing against each other the least because of the low tension. And the angle of the weave would be very flat because of the small number of strings. But in this case the friction was too low.

The loosely strung mains could slide so easily and so far on the crosses that they couldn't get back fast enough, and didn't have enough energy, to torque the ball and apply extra spin on the rebound.

The 16x10 pattern was the best for spin at 60 pounds because there was very little inter-string friction, and the worst for spin at 30 pounds because there wasn't enough stiffness to go with the super-low friction.

In his latest experiment - on lubricated strings (http://twu.tennis-warehouse.com/lear...pinandlube.php) - Lindsey found much the same thing. Strung at 60 pounds, all string types in a 16x19 pattern generated more spin when WD-40 was applied. The reduced friction led to more spin. The mains slid and snapped back really easily and had enough stiffness to be even more efficient in applying spin-inducing torque.

With a 16x10 pattern strung at 60 pounds WD-40 improved spin. But when Lindsey applied the lubricant to the 16x10 pattern at 30 pounds, spin was absolutely horrible. The strings were so slippery and so pliable that they slid way too far to have enough time and energy to snap back.

Another thing that happens with more string movement is higher launch angle. Even when making the strings more slippery results in more spin - like lubricating a 16x19 pattern - the launch angle is increased. And that effect will vary depending on how much spin is incoming: the more topspin on your opponent's groundstrokes, the higher the launch angle off your freely moving strings. This is the effect that Ferguson talked about in the first article in the series when he mentioned how hard it is to volley when you have copoly in your own racquet.

Ferguson also mentioned that Sampras is stringing RPM Blast at 70 pounds in his 98 inch Babolat. Very, very, bone-crushingly tight for such a stiff copoly. But probably Pete is getting a predictable response - not too much string movement - on volleys, and still getting enough string movement and snapback to juice up his groundstrokes.

And I guess that's what it comes down to: getting a string setup that works for your game. So, all things being equal, open patterns should give more spin at a given tension, as long as the tension is high enough so the main strings have enough stiffness to slide and stretch tangentially, but still snap back in time to spin the ball with that extra torque. Sliding is a good thing, but too much sliding is not.

And "too much" is probably player-dependent. Pete apparently likes moderate string movement - high string stiffness - so his volleys are still controllable. Nadal wants it looser so he gets maximum snapback and spin with decent pop on his groundies.

And "how stiff?" is also a function of how fast we swing. Nadal strings at 55 pounds, which apparently is the right tension for his 80-90 mph forehand swings. At those swing speeds the energy of ball-string impact is about right to slide and stretch his mains enough, but not too much, for maximum spin.

But for an intermediate player, who might swing around 65 mph on a forehand, a lower tension might give optimum stiffness. 55 pounds for the intermediate might be too stiff - the strings won't bend enough to really get the snapback mechanism loaded fully.

And there's one more factor that comes into play when matching tension to swingspeed. I mentioned in the last article that on extremely steep swings slippery strings might be too slick to fully grip the ball. But string stiffness also comes into play here. Stuart Miller at ITF told me that at these extremely steep angles a softer string will be better because the lesser amounts of ball-string friction will be able to push and slide a softer string.

Finally, you mentioned more main strings being better because then you've got more strings snapping back with that spin-enhancing torque. Barry Phillips Moore - the old pro who played with spaghetti strings and later invented the 14x16 Hi-Ten racquet that Woodforde used - is leaning in this direction. He's working with designs with lots of main strings and few crosses. Pretty much the same direction that Lindsey has gone with his testing of 16x10 patterns.

But I'm not sure we want "more strings rebounding" so much as more strings to grip the ball. The energy that goes into the strings is finite. It can bend fewer strings farther or more strings less far, assuming the resistance to bending and sliding (friction plus string stiffness) is the same in each case.

However, more main strings might grip the ball more quickly than fewer. Recall from Part III that that's one of the hypotheses for why textured strings might be more spin-enhancing - that higher ball-string friction helps the strings re-grip the ball during the snapback phase. So more main strings might be better from that perspective.

Even better than 16x10, then, might be 18x10 or 20x10.

Or 20x6, which is pretty much spaghetti strings. But to do this with a woven racquet you'd have to string the mains really, really tight. Like 90 pounds or more, according to Phillips-Moore.. He uses really thick poly and special frame designs that can withstand those tensions without snapping. (This is also an ideal application for Kevlar strings, which are three times as stiff as poly and grip the ball very well.)

Actually, it's possible that the gut mains in spaghetti-strung racquets were too loose. Strung with poly at 90 pounds, a spaghetti racquet would probably be more spinny, and launch the ball at a lower, more predictable angle. The inventor of spaghetti strings, Werner Fischer, and Barry Phillips-Moore both said that, had spaghetti strings not been banned, the design would have been refined.

Copoly is probably better and more versatile than the spaghetti strings that made it, briefly, to the tour. String it tight, like Sampras, and you get excellent spin, but still enough control on volleys. String it loose like Volandri and you get excellent spin with soft feel and power closer to syngut. String it in the middle like Nadal and you get optimum spin, excellent and predictable control and decent pop.

And, while a 20x10 pattern might be fantastic for topspin, on those sidespin shots the 10 crosses would be a liability. Probably, a different number of strings would be impacted on each shot, which would mean a variable number of sliding strings, variable amounts of spin, and a capricious launch angle from shot to shot.

**stumphges** · 08-28-2011, 12:20 AM

A caveat to the above post on string pattern and spin, which wouldn't fit above because of the 10,000 character post limit (oh man, I need to get back to work):

In the above post responding to Jim's second question, I'm using an interstring friction/string stiffness model to predict how different stringbeds will respond at different tensions. This model is grounded and based on the experimental work of Crawford Lindsey and Rod Cross. But there are examples of setups that function in ways this model can't predict or account for.

For example, Lindsey found that both Alu Power and RPM Blast generated more spin at 60 pounds than 30 pounds. Using the friction/stiffness model we could explain that result easily by saying that at 30 pounds the main strings were too loose and didn't have time to snap back quickly enough.

But, in the same experiment, he found that a much softer copoly, Polyfibre TCS, generated more spin at 30 pounds than at 60 pounds. TCS at 30 pounds is about 20% less stiff than Alu or RPM Blast at the same tension, so why would it be able to snap back quickly enough at low tension when those stiffer strings couldn't?

I don't know. Either this model isn't complete or it's flawed in some way. In either case, more research will provide the answer. It's also possible that the answer is staring me right in the face. Anybody see it?

(In the meanwhile, Polyfibre TCS 16 looks like a good string if you want to try copoly at very low tensions.)

**jimlosaltos** · 08-28-2011, 09:21 AM

Thank You !

Originally posted by stumphges View Post

A caveat to the above post on string pattern and spin, which wouldn't fit above because of the 10,000 character post limit (oh man, I need to get back to work):

(

Whew!

Greatly appreciate all the effort in providing such a detailed answer.

I think you've got a Part's IV and V to your series written already with these posts !

Very interesting stuff. My biggest fear, though, is that all this will be going through my head when I try to hit the ball -- my concentration is already questionable.

Thanks again. / jim

Announcement

Copoly effect on serves?

Copoly effect on serves?

Comment

Comment

Comment

Comment

Who's Online