Home ⇒ golf ⇒ shafts ⇒ Current Article

All About Spines

Loose ends

Other technical issues

What causes spine?

Actually, the name "spine" originally came from the belief that, since most steel shafts were welded tubes, the seam created an asymmetry in stiffness. In the 1990s, Apollo was alone in making its steel shafts without a seam, and they based an advertising campaign on this issue of directional uniformity. Wishon and Summit debunked this in their 1992 book; turns out the Apollo seamless shaft had even more stiffness asymmetry than the TrueTemper welded shaft in the study. So the welding seam isn't the only cause of spine, and not even the major cause.

That said, let's look at the four major processes for manufacturing golf shafts, and see what sort of spine we can expect from each:

Welded steel tubing - Today, just about all steel shafts are made by bending sheet steel into a tubular form and welding it. As noted above, the major manufacturers of steel shafts seem to have figured out how not to let the seam interfere with directional uniformity of stiffness. While I have heard of some steel shaft models that have up to 4cpm of spine, I have never seen a TrueTemper shaft with more than 2cpm, and the vast majority do not exceed 1cpm. And TrueTemper manufactures the vast majority of the steel shafts used today.^[1]
Extruded metal tubing - From time to time, we see shafts (usually aluminum) extruded seamlessly. Theoretically, they should be possible to be made more directionally uniform than welded steel tubing. In practice, they don't do any better -- at least in part because steel is already so good.
Sheet-wrapped graphite - (This is also referred to as "flag-wrapped".) This approach involves taking a sheet of woven carbon fibers and wrapping it around a rod called a "mandrel". In more detail:

The fibers may be woven together, or may be unidirectional -- all fibers parallel. In either case, they are held together in a sheet by being impregnated with a resin. Since the fibers are going to be resin-impregnated anyway once the shaft is made up, the sheet is referred to as pre-impregnated, or "pre-preg".
Several sheets of pre-preg, with the fibers in different directions, are wrapped onto the mandrel in multiple layers. For instance, fibers oriented down the length of the shaft are for stiffness and spiral fibers 45º from the shaft axis are for torque. The number of layers of each kind (and the properties of the fibers) will determine the bending and torque profile of the shaft.
Sheets that are wrapped onto the mandrel have a beginning and an end. The discontinuity at the beginning and end mean that the shaft is slightly stiffer where the extra layer is. Hence there is a potential for a spine; some directions have more fiber layers than others. In fact, sheet-wrapped graphite shafts do tend to have more spine than other types of shafts. But there are ways of mitigating this, for example:

Design the wrap so the beginning and the end occur at the same position around the shaft. That way, no one direction has more nor less material than any other. This is possible, but requires careful control to get it to line up just right.
Use more, thinner layers instead of fewer, thicker layers. That way, the difference represented by a single layer is smaller.
Design the wrap so the edge of the pre-preg is spiraled around the tube. That way, no one direction gets the whole impact of the discontinuity; it is distributed around the shaft.

Filament-wound graphite - These shafts don't use a pre-preg. Instead, the individual fibers (filaments) are wound in a spiral onto a mandrel. Actually, they are more woven than wound; spirals in both directions are applied simultaneously and woven together. The process involves a pretty intricate machine to weave the fibers. After they are wound in place, they are resin-impregnated and cured. The result is generally remarkably free of spine. There are other disadvantages of filament-wound shafts, but consistency of directional stiffness (and consistency of specs in general) is a definite advantage.

You can expect most brands and models of shaft to have spines in the 0-2cpm range, except for sheet-wrapped composite models. Of course, there are exceptions both ways to this rule; there are low-spine sheet-wrapped shafts, and a few disappointingly "spiny" shafts from the other processes. So it pays to know the brands and models you work with.

When you're in the range of 1cpm spines, further improvement is somewhere between quality control and the inherent limits of the manufacturing process. But that's probably OK. I have seen no evidence that further improvement buys any additional advantage, whether or not the spine is aligned.

Cut shafts

The question sometimes comes up, "Will the spine/NBP directions be the same in a trimmed shaft as they were in the raw shaft?" The answer quite often is "no". This raises a couple of questions: why does this happen, and what should you do about it?

The answer to "why" follows pretty closely the reasons for spine in the first place. Think of why spine occurs, then ask why we would expect it to be in the same direction -- or, for that matter, in a different direction -- over the length of the shaft.

The cause of the biggest spines is discontinuities in sheet-wrapped shafts, typically resulting in a difference in the number of layers of pre-preg as you go around the shaft. If the pre-preg seams are straight up and down the shaft, then the spines that they cause should be the same over the length of the shaft. But often the seams spiral down the shaft, either by design or inadvertently. When that happens, the resultant spine is a composite of the spines over the length of the shaft. By trimming from one end or the other, you remove some directional components of this composite -- and change the direction of the overall resultant.
Most of the causes of smaller spines are related to quality control or manufacturing processes. For such shafts, the answer depends on the specific defect or cause involved. And it isn't a big deal; most such shafts have small enough spines that alignment is not an important process.

OK then, what to do about it? The answer is pretty simple. Don't worry about actually marking the spine carefully until the shaft is trimmed. Then mark it and align to that mark.

But there's a bit of a hitch here. If the size of the spine is more than about 3cpm, then the frequency of the club will depend on the orientation of the shaft. If you have any reason to believe that the spine will be sizeable, it is worth finding the spine early so you know what direction to test the flex as you trim. Then, once the shaft is trimmed, do a careful spine location that you will use for alignment.

Makes sense?

How much spine is "negligible"?

There are still a few clubmakers around who believe that there is no such thing as a negligible spine. To them, if you can find a spine, it is important to align it. And if you can't find the spine, then you're in big trouble -- because you know it's there and you won't be able to align it. I've watched these guys FLOing a shaft and muttering curses when it refuses to wobble in any orientation.

I'm amused by their antics because they are treating a very good situation as if it were very bad. But then I'm convinced that "negligible spine" is an important concept. By negligible spine, I mean a level of spine below which you can't get a measurable improvement in performance by aligning the spine. Solid, controlled experimental work has not yet been done to find this level, so there is still a remote possibility that there is no such thing as neglible spine. But, based on the sort of engineering considerations that got us to this point in the article, let me argue the position that alignment doesn't matter for spines below a certain level.

But first, let's look at the arguments (I originally wrote "psychologies", and probably still believe that) supporting the "no such thing" side:

The perfectionist. I know a lot of clubmakers who build to much tighter specs than necessary "because I can". These are people who get a frequency meter that measures to 0.1cpm so they can do a better job of frequency matching. They get a digital swingweight scale not for the convenience, but so they can match swingweight to a tenth of a point. And they insist on aligning a spine if they can measure it -- and are often troubled if they can't detect it. To them I say, "If your time to do this is compensated, either by money or personal satisfaction, then it is worth it to you. But don't confuse 'because I can' with 'it matters in terms of tangible results'."
The golf mystic. I know and like a few people who believe that everything about golf is a mystical, magical experience. To them, a misaligned spine is "bad karma", and will be punished by bad shots. I have learned not to discuss golf technology with those people; we will never agree. As a scientist and engineer, I believe that spine effect is a physical phenomenon. There is nothing magic about it. A misaligned spine causes problems because of the forces on or motions of the shaft that it causes, not because of some mystical karma associated with shaft alignment. If you got this far in my article, you are at least somewhat interested in the scientific point of view, so let me just dismiss the mystical argument as not scientific.
The bio-chemical analogy. Ecological studies sometimes conclude that there is no acceptably low level of the studied chemical that is environmentally safe. That is cited as an existence proof that the "no acceptable level" conclusion has a place in science. And it does! But all the no-acceptable-level results I've heard of involve a cumulative effect -- continued exposure over long periods of time. What is cumulative about a golf club? This argument has no relevance here.

Ultimately the effect of a misaligned spine is a less-than-ideal force or motion. These are things that can be measured. For instance, SST and Butler found that a misaligned spine can cause off-center hits. Now please follow this reasoning:

There is some acceptable degree of off-center, some amount of "miss" that will not produce any measurable effect on the shot or the feel. Some might say 1/4". Almost everybody would agree that a 1/10" miss of the sweet spot will not affect a drive.
Let us assume for a moment (the assumption turns out to be true) that the off-centeredness of the hit is monotonically related to the size of the spine and the degree to which it is misaligned. ("Monotonically" means that, any time you increase the spine or the misalignment, the off-centeredness increases.)
Given that assumption, the worst possible alignment of a spine is still acceptable if the spine is so small that its effect is less than 1/10" off center for the hit.

That meets my definition of negligible spine.

Let's look again at what a misaligned spine does to the shot result. We already listed the possibilities under "theories". The physical phenomena involved in all the theories is either:

A force that is not in the same direction as the bend of the club, causing out-of-plane motion of the clubhead, or
A torque tending to rotate the club, when the bend is not aligned with the spine.

Both of these phenomena can be quantified, and we can set an acceptable level on the result. Let's try that.

Out-of-plane forces - I have done a ball-park calculation^[2] of how far off-center the hit might be due to an out-of-plane force. The assumptions were a driver with a fairly soft shaft (230cpm). The golfer swings the club in a way to produce a maximum bend of 3" during the downswing. (That's quite a lot of bend. Less bend would give less off-centeredness for a given level of spine.) One assumption contrary to fact is that the shaft is in the worst possible alignment for the entire downswing, (That will tend to give a very "safe" estimate, since the shaft does not stay in one orientation during the downswing. The level of negligible bend would be higher if we assumed something more realistic.) Here is the result:

A 10cpm spine results in a 1/4" out-of-plane deflection of the clubhead. This is on the edge of measurable performance or feel for most golfers.
A 4cpm spine results in a 1/10" out-of-plane deflection of the clubhead. It is unlikely that any golfer would detect this.

So, at least as far as out-of-plane forces are concerned, the threshold of negligible spine is probably 4cpm or higher.

Just for the record, the calculations showed that the off-centeredness did vary monotonically with the size of the spine at any given degree of misalignment, and also varied monotonically with the degree of misalignment for a give size of spine. So our earlier assumption is correct.

Torque - The theories that operate here hold that the NBP should be aligned to some direction, so that bend occurring at impact (or perhaps slightly before or after) does not create a torque that might rotate the head away from a square clubface. To see how big a torque would be significant, we want to compare (A) the torque due to a misaligned spine with (B) the torque already being exerted by the golfer on the grip to square up the clubface. If torque A is a sufficiently small percentage of torque B, then the spine is negligible for practical purposes.

How small is "a sufficiently small percentage"? Well, a lot of the torque exerted by the hands on the grip is predetermined by anatomy. As the wrist cock releases from 90º to 0º, the club naturally rotates the clubface -- from in the swing plane initially to towards the target at impact. If you try to prevent this from happening -- if you try to release the wrist cock without releasing the rotation -- you could actually hurt yourself. Relatively little of that rotation is "discretionary", something that the golfer can control, either consciously or unconsciously. So it is unreasonable to demand a very small percentage like 1%; we will use a higher percentage.

In support of this decision, remember that we use pretty much the same swing to square up all our clubs, driver through short iron. Where we have different moves or thoughts for different clubs, the rotation of the hands to square the clubface is not part of that difference. Yet the clubheads themselves have vastly different moments of inertia. (MOI of a clubhead is its resistance to being squared up.) There is a 2 to 1 difference between a driver's and a 5-iron's moment of inertia. So there is substantial evidence that even a 50% variation in square-up torque isn't a problem.

Given these consderations, a figure like 10% is probably close to the mark for "negligibility", but let's be conservative and pick 5%. That is, if the torque due to a misaligned shaft is no more than 5% of the torque needed to square the clubface, then the spine can be considered negligible

Again, I have done a ballpark calculation^[3] of the torque needed to square up a driver, and the torque due to a misaligned shaft in the vicinity of impact. The assumptions in this case are:

The same soft driver shaft as before (230cpm).
A modern 460cc 200g driver head.
All the uncocking and rotation in the last 100msec before impact (consistent with strobe photos of good swings, and reinforced by ShaftLab data).
A bend at impact of 1.5" (that's a very large bend, according to ShaftLab data; smaller bends show less effect from a misaligned spine, thus a bigger negligible spine).

The result is that a 4cpm spine (actually a little less, 3.6cpm), when misaligned as badly as possible (whatever that means^[4]) will provide an undesirable torque of 5% of the torque the golfer is already exerting. So the level of negligible spine for torque is about 4cpm.

Bottom line: It looks like a 4cpm spine is somewhere around the threshold of "negligible", whether the problem created by spine is off-center hits or torque due to a bending shaft. The calculations leading to this conclusion are rough; they might be off by a factor of two either way, but probably not that much. I'm pretty confident that the number for each effect is somewhere between 2.5cpm and 6cpm. And most "consensus" estimates are also in that range.

What I do

You may well ask, "OK then, what do you do about spine alignment in clubs that you build?" Here's what I do:

To begin with, I want to make very clear that I don't think spine alignment is a fine tuning for performance or feel. It is damage control! Spine is a shaft defect. The best answer, IMHO, is to minimize the defect to begin with. Spine alignment is a last resort, IMHO, to minimize the damage that the defect does.

I go out of my way to order shaft models that I know to have negligible spine -- then I just don't worry about aligning. My personal belief is that the threshold of negligibility is probably 3-5cpm, and I treat a spine of less than 3cpm as something I can ignore. Some ways I use to choose shafts:

Some manufacturers are a lot better about minimizing the spine than others. I try to keep track of this, both from my own experience and reports in the forums where I hang out.
Most steel shafts and filament wound graphite shafts have almost no spine at all, certainly much less than most sheet-wrapped graphite shafts. There are exceptions. A few [relatively small market share] brands of steel shafts have noticeable spine. Some sheet-wrapped graphite manufacturers have negligible spine.

When dealing with a shaft that I am not sure is spine-free (for practical purposes), I determine the spine by FLO, because it's quicker than Differential Deflection and, unlike feel finders, locates the true spine. Then I determine the size of the spine by comparing the frequency in the two FLO planes.

Sometimes I have to deal with a non-negligible spine. For instance, the customer may really want some shaft model and for whatever reason I don't say no. When that happens, I align the shaft with the spine in the heel-toe plane. Reasons for that alignment:

It assures an in-plane restoral force during big bending early in the downswing, because that large early bend is heel-toe.
The stronger of the two forces is fighting toe droop at impact.
Interestingly, that results in the NBP in the target plane at impact -- as recommended by the usual "conventional wisdom". That's because of the 90º relationship between spine and NBP. So, even if the first bullet item is not a good reason, I am able to automatically hedge my bet.

Here is an interesting issue: if there is a significant spine, then the frequency at the spine is significantly different from the frequency at the NBP. So which of the two frequencies do I use for matching? Unlike most clubmakers who align spines, I trim to the spine frequency. The reason is that's where most of the bending occurs, if you align the spine in the heel-toe plane. You want the frequency (or stiffness, if you match by deflection) to be effective during bending, and most of the bending occurs in the heel-toe plane.

Rules implications

Let me finish by indulging in a little rant about the Rules of golf.

Appendix II of the Rules (the part that details the rules of clubs) says

Bending and Twisting Properties
At any point along its length, the shaft must:

bend in such a way that the deflection is the same regardless of how the shaft is rotated about its longitudinal axis; and
twist the same amount in both directions.

That rule has been there for a long time. I find almost identical wording in my 1987 rulebook, and it was old then.

Think about it. The rule says that shafts should not have spine. If it is impossible to meet that rule precisely, then testable limits to spine need to be set -- and enforced. The USGA has done this with other specs (e.g. - coefficient of restitution).

I think the USGA has dropped the ball here. They have caved in to the shaft manfacturers on one side (who don't want to be held to standards) and Dick Weiss on the other (who threatened lawsuit unless the USGA legitimized spine alignment). The decision they adopted is:

Manufacturers of clubs may orientate or align shafts which have spines for uniformity in assembling sets or in an effort to make the shafts perform as if they were perfectly symmetrical. However, a shaft which has been orientated for the purpose of influencing the performance of a club, e.g., to correct wayward shots, would be contrary to the intent of this Rule.

This is unenforceable, as it requires knowing the intent of the clubmaker. It also allows "supershafts", assuming that the supershaft effect is real. (As I noted earlier, there is no public information supporting that it is. But it might be.)

Soapbox time! The USGA needs to find its spine. They should:

Conduct test to determine at what level spine alignment affects performance.
Set and enforce a limit on manufactured shafts below this level, to assure that spine alignment has no effect on the performance of a club.

I view this as "protecting the game", which they are supposed to do. It would mean that golfers would not have to worry about buying a spine-aligned club. It would reduce the influence of special equipment in winning, as opposed to simple skill at the game. I expect such a proposal to be opposed by:

Shaft manufacturers. Those shaft manufacturers who don't meet the specs would see their costs increase. Those who already manufacture low-spine shafts would lose their competitive advantage.
Club manufacturers. The OEMs operate by calling for competitive bids for shafts, and choosing the manufacturers by price. This is yet another spec the would need to control. Additionally -- and more bothersome -- it will probably increase the price of the low bidders (who likely achieve their current low cost in part by not controlling spine).
Custom clubmakers. They currently offer spine alignment as value added, either as a competitive advantage or an extra-price option. This will be gone.

The support should come from the golfing public -- if they knew enough about the issue to generate support. Unfortunately, they do not understand spine alignment except as a buzzword, nor do they have any sort of effective lobby. The USGA should be their lobby, but it seems to have forgotten that over the last couple of decades.

Given the USGA's recent history, I expect the opposition to be successful.

(If you want to take me to task based on some political view about regulation vs free market, please first look at my article on the role of rules in sports.)

Notes:

As of 2006, TrueTemper's share of the steel shaft market was 82%. The company's Chad Hall is quoted in GolfWrx Blog, "Retail market share information is tough to substantiate in the golf industry, but we estimate our worldwide steel shaft market share at approximately 82%. Tour usage in iron shafts is at 98%. The wood shaft category on tour is much more highly fragmented with the leader averaging around 25% each week."
The ball-park calculations involved using the FLO spreadsheet from another article, and measuring the deviations in early oscillations.
This time, the ball-park calculations are fairly extensive. You can find them in the appendix if you're interested.
There are several different theories of spine effect that depend on the torque resulting from bending a misaligned shaft. Those theories differ as to what they consider properly aligned. One says NBP to target, another says NBP to center of gravity. This calculation doesn't take sides; it just computes the torque assuming an amount of bend in the worst possible direction for the shaft -- however it may be oriented.

Last modified -- 3/2/2008