Do-It-Yourself Shaft Puller

Dave Tutelman -- November 27, 2011

What it is and why

This article is about a Shaft Extractor, a tool that helps a clubmaker remove a shaft from a clubhead. The featured tool is one you can build for yourself. The original idea (including the key components) comes from Ron Blanchard, and I modified his initial prototype to be easier to build and more practical to use.

Traditionally a shaft extractor has been called a Shaft Puller -- even though it could be reasonably argued that it pushes the clubhead off the shaft more than pulling the shaft out of the clubhead. But let's begin at the beginning, and see the procedure of removing a shaft from a clubhead in such a way as to save both components for reuse.

  1. You clamp the shaft in a vise. Very securely, because you are going to use considerable force pulling the head off.
  2. Heat the hosel with a torch, to break down the epoxy that holds the clubhead to the shaft. If you are prepared to heat it for a long time, even a heat gun will do the trick.
  3. Once the epoxy bond releases, pull the clubhead off the shaft tip. This takes a lot more force than you might imagine, because the epoxy is still mechanically attached even if no longer adhering chemically.
With a steel shaft, you can do the pulling by hand. (Be sure to use a thick glove. That clubhead is hot!)

You can't do it that way with a graphite shaft, for several reasons:
  • The carbon fibers are held together with a resin, either epoxy or something very similar. If the heat breaks down the epoxy cement (as it should) it is also close to breaking down the resin that holds the shaft itself together. So you want to use less heat and more force for a graphite shaft.
  • Following up on the previous point, the force needs to be straight along the shaft axis, and applied at the clubhead's hosel. No twisting. No off-center force. Any of these could damage the shaft (especially since it is on the verge of being compromised by heat). And even if you are not interested in saving the shaft, twisting it during removal can leave chunks of broken-off shaft in the hosel; you will need to ream it out. (Don't ask how I know this.) BTW, you are interested in saving the shaft. Today's composite shafts are usually more expensive than the clubheads.
So, for graphite shafts, you need a tool that will exert a lot of force on the hosel, straight along the shaft axis, to push the clubhead off the shaft. Here is a typical commercial shaft puller. This one is made by Ed Mitchell's Steelclub company, one of the best suppliers of heavy-duty clubmaking tools.

The shaft is clamped horizontally just above the label. You can see this is a hefty clamp with a T-handle, capable of resisting a lot of force.

There is a second screw drive with a T-handle, this one moving a carriage horizontally. Mounted on the carriage is a metal "fan" with four slots in it. The shaft fits loosely in one of the slots, and the metal surface mates up against the hosel. (There are different slots because shafts come in different diameters. Each slot has a slightly different width.) Let's refer to the piece with the slots as a "claw", because it grips both sides of the hosel like a claw and pushes the clubhead forward.

When you turn the T-handle for the horizontal screw drive, the carriage moves the claw forward to apply force to the clubhead. There is enough room around the clubhead to move a torch to heat the hosel; this is an important design feature for any shaft puller.

You may have noticed that there is no stand for the tool. This shaft puller was made to be clamped in a bench vise. Other designs have a flat mounting plate to be screwed to the workbench.

Models like the one shown sell for about $200. A full professional model (made to be permanently mounted to the workbench) can easily run over $300. And there are even more expensive models, as well as designs based on a hydraulic jack rather than a screw thread.

These prices are easily affordable by a shop that does dozens or hundreds of reshafts a year. But the hobbyist or sideline business may be looking for a less expensive tool -- or looking to trade sweat equity for a lower price. This article presents the plans for a good DIY (do-it-yourself) shaft puller.

(Before I continue, Golfsmith and GolfWorks recently introduced budget shaft pullers for about $70. I haven't used either one, but they look useful. If they do the job, then they are cutting the dollar value of your sweat equity, should you choose to build one yourself. This one cost me about $30 in parts, roughly half the $70 of the budget puller. The difference in cost is not nearly as compelling as when the alternative started close to $200 and went up from there.)

Prior DIY shaft pullers

Tom Flanagan's design, and my version

A bunch of years ago (probably the late 1990s) Tom Flanagan, a California clubfitter, came up with a DIY shaft puller based on a bench vise. The plans are posted on Clubmaker Online's resource pages. The picture to the right is linked from that site. The bench vise is the chassis of the instrument, the shaft is clamped to the fixed rear jaw, and the movable front jaw is the carriage. The vise is assumed to have removable jaw inserts; they are removed, and replaced with parts fashioned from thick aluminum stock.
  • The shaft clamp is pars A and C.
  • The claw is part B.
  • There is a drop-on insert, part D, with a narrower slot than the claw; it is inserted for pulling narrower shafts.
I liked TFlan's design, and built my own version (shown at left). I had found a solid aluminum bench vise at a yard sale, and picked it up for a dollar. I figured aluminum would be easier to drill and tap, which would give me a better clamp. Turns out it did, but still not good enough.

The clamp only acted on a little over an inch of shaft, which was not enough to prevent slipping. The tremendous force exerted by the claw just overwhelmed the clamp force, unless the epoxy was really well broken down. The vise could exert enough force, but the shaft clamp could not properly resist it.

I bought a drill-press vise from Harbor Freight, intending to somehow mount it to the rear stationary jaw, but I never got around to it. (But we shall see that my having the vise on hand was a good thing; it was there for the next design.)

Ron Blanchard's shaft puller

In March of 2010, Ron Blanchard emailed me:
This may be worth sharing with Shop Talk or other appropriate forums at some point.
I have just finished a prototype shaft puller for my own use (my third different design, actually) and I thought it would be useful to have you put an eye on it and give me some feedback.
His design was based on two tools from Harbor Freight, a favorite hardware vendor for inventive clubmakers.
  • He clamps the shaft in a 4" drill press vise, using commercially available auxiliary jaws for protecting shafts in a vise.
  • The pulling thread and carriage are an automotive pulley puller (part #66868), with the pulling arms removed from the carriage. (The arms are removed by unscrewing a pair of knurled bolts.)
(Click on the thumbnail at left for a picture of Ron's shaft puller.) Ron mounted both subassemblies to a piece of wood, using light aluminum angles as brackets to hold the pulling spindle. The cap of the spindle presses against the vertical edge of the vise, just below where the shaft is clamped. He uses a socket wrench on the hex head of the spindle to crank the carriage away from the vise, which generates a huge pulling force between where the claw holds the clubhead and where the vise holds the shaft.

The most important feature that makes this a superior shaft puller is the very high threads per inch (TPI) of the threaded spindle. More TPI means you can exert more force on the clubhead (the reasoning is given below). Most vise threads are of the order of 4 to 8 TPI. The vise I used for my Flanagan shaft puller was at the high end of the scale, at 8 TPI. The spindle in Ron's prototype has 18 TPI, meaning it can exert more than twice the force of my Flanagan-design shaft puller.

Simple machines, and why more TPI gives more force

(You can skip this if you're not interested in the physics of it.)

Screw threads are one of physics' "simple machines", the most elementary of which is the lever. The lever trades motion for force; you can use a lever to exert more force, but as a result can not move the pushed object as far. Since the total work done is the force times the distance through which it moves, this does not violate conservation of energy. The work (energy) is the same; you have simply traded distance for force. See the tutorial on torque for another way of looking at the lever.

The next simple machine is the inclined plane. When you push something up an incline, you don't have to use as much force as if you lifted the object straight up. Just like a lever, you are trading distance for force. If you are lifting something that weighs 100 pounds, you would have to exert a force of 100 pounds to lift straight up. But you get a "mechanical advantage" from an inclined plane. Mechanical advantage? That is a technical term, meaning the force multiplication you can get from a simple machine. For an inclined plane, the mechanical advantage is equal to the length of the plane divide by it's height. Suppose you have a 30-foot-long incline that rises 6 feet. It would have a mechanical advantage of 5 (30 divided by 6).  So if you were pushing the 100-pound object up that incline, you would need to exert only 20 pounds of force, not 100 pounds. And again, energy is conserved; you only exert 20 pounds, but you push it 30 feet to achieve a rise of only 6 feet.

Now let's look at a screw thread. It is just an inclined plane wrapped around a cylinder. A single turn of the screw moves the screw forward by 1/TPI of an inch, but the edge of the screw travels the full circumference of the cylinder. So the mechanical advantage of the thread is the circumference divided by the "pitch of the thread" (which is 1/TPI).

For instance, let's look at an everyday 1/4-20 screw thread. That designation means a diameter of 1/4 inch (radius of 1/8 inch) and 20 TPI. The mechanical advantage of the thread is the circumference (pi times 1/4") divided by 1/20 of an inch. Doing the arithmetic, it comes out about 16. That's a pretty hefty mechanical advantage.

We can generalize this to a formula for mechanical advantage: circumference divided by thread pitch.

Mechanical advantage =

 = 2πr * TPI

So we can see that we can get more pulling force from the screw if we have more threads per inch.

But we also have the shaft radius there. How does that affect things?

Turns out it doesn't! Remember, we are exerting force through the handle of a vise or spindle. That handle is a lever, the most basic simple machine. We exert force on the long handle, and that force is applied at the threads. The lever, being a simple machine, has its own mechanical advantage: the handle length divided by the radius of the threads. When we compute the total mechanical advantage -- that of the lever times that of the screw thread -- the shaft radius cancels out.

Total Mechanical Advantage =
Handle length

2πr * TPI  =   Handle length * TPI * 2π

Bottom line: We get pulling force from our design proportional to the crank handle length and the threads per inch.

The Blanchard-Tutelman shaft puller

(Click on thumbnails for the full-size image)

Design considerations

Ron Blanchard had asked for my comments, which resulted in a substantial email exchange. It was almost a year later (January 2011) that we fully understood what was good about the idea and needed to be preserved, what needed to be improved or fixed, and how to go about it. Here was our list.
  • Good - High threads per inch for a very powerful pull.
  • Good - Drill press vise made a very secure shaft clamp. (This was needed even more than for the usual shaft puller, because of the extra pulling force provided by the high TPI.)
  • Bad - Mounted on a wooden board. I'm an electrical engineer according to my degree, and I'm used to prototyping electronic circuits on "breadboards". This mechanical breadboard was a good way to start, but had a couple of problems as the base for a real-life shaft puller:
    • The base occupied enough space around the clubhead that it somewhat restricted access for the torch or heat gun.
    • Wood can be burned, and we were using heat that could easily burn the board.
  • Bad - The claw was far enough back on the carriage so that the carriage restricted torch access a bit, and might even threaten to interfere with the clubhead itself.
  • Bad - The spindle occasionally hung up on the holes in the thin brackets that held it in place. A redesign of the spindle supports was necessary.
I already had the vise on hand, and the Harbor Freight store in nearby Brick Township was having a sale on the pulley puller, so I built a shaft puller that addressed the shortcomings of Ron's first prototype. Here is a photo of my shaft puller in action. The configuration of the working parts is the same as Ron's prototype, even though it may not appear so. The differences are aimed at the specific shortcomings of the prototype:
  • The wooden board - The whole assembly is now mounted on a length of heavy-duty aluminum angle. Much narrower, and non-flammable. Instead of clamping it to the bench, you can mount it in a bench vise.
  • Clearance for the claw - The claw has been turned 180 degrees. The business end is now at the outer end of the carriage, and the carriage is not in the way of the torch nor the clubhead. (This introduces a potential stability problem, which we deal with below.)
  • Brackets for the spindle - Our first thought was a bushing in the bracket. Plastic bushings were out; they melt at the temperatures we are using. And brass bushings are expensive and harder to work with. We thought of using thicker aluminum for the brackets, and rounding the edges of the spindle hole. That gave me an idea! Eye bolts already had holes with rounded edges. On top of that, they had little material except for the eye itself (the "bushing") to get in the way of a clubhead or torch. So I did it with eye bolts.
You can permanently attach Wilton multi-grip vise jaws like Ron used. I chose to use a standard rubber shaft clamp, the sort that every clubmaker has lying around. They fit the 4" vise perfectly.

Detailed drawings

I just got some neat CAD software (TurboCAD Deluxe) and finished the exellent video training supplied with it. I thought I'd do a set of drawings of the Shaft Puller as my first real TurboCAD project. Here it is. If you click on a thumbnail, you'll get the page as a PDF file.

The cover sheet has a quick overall rendering, plus the bill of materials for the project.
Three rendered views show the tool from different angles: southwest dimetric, northeast isometric, and southwest from below to show the bottom.
A working drawing ("wireframe" style) with dimensions. This shows how to drill the aluminum angle that will become the main beam of the shaft puller.

Note that the holes for the eyebolts and the mounting holes for the vise are not on the same centerline. That is intentional; we don't want the bolt heads of the vise mounting bolts to interfere with the spindle where it pushes against the vise.

Figure your own positions for the vise mounting holes. The dimensions in the drawing work for my vise. But mine is several years older than the one Harbor Freight now sells, and is slightly different. So the holes may need to be in a different place. The important thing is that the spindle pushes against a vertical surface of the vise about 7" from the front end of the aluminum angle.
Three simple parts on this sheet:
  • The claw is made from aluminum angle, drilled and slotted according to this drawing.
  • A stock 1/4" eye bolt does not have a large enough eye to pass the spindle through. It has to be bent open until the spindle can pass through and turn freely. The eye bolts I found at Lowe's had larger eyes than the ones at Home Depot, but either needed bending.
  • I use a drop-in collar to fit the slot of the claw more closely to the shaft diameter. The part shown has two slots, suited to the two most common shaft diameters. The shaft diameters accommodated are .335" and .370".
The spindle is the Harbor Freight Pulley Puller with the arms removed. Without the arms, there are three parts: the spindle itself, the carriage, and an end cap. The carriage and end cap have to be removed during assembly, to get the spindle into the eye bolts.

The arms had been held on with a pair of knurled bolts. I kept one of the bolts, which I use to attach the drop-on collar to the carriage, hand-tight, so it doesn't get misplaced when I'm not using the Shaft Puller.
The vise is simply carriage-bolted to the main beam.

The eye bolts need to be held in position, implying each eye bolt needs two threaded parts working against each other. We deal with this by threading the holes in the main beam, then snugging up a hex nut on each eye bolt to lock it in place.

The pan head cap screw must be tightened enough so the claw does not turn under pressure. Unlike the prototype, where the carriage was pulling the claw, this design is pushing the claw, an inherently unstable position. I have not encountered any difficulty due to instability, but it could happen.
If you encounter any instability -- any tendency of the claw to turn that can't be fixed by a tight pan-head bolt -- here is a suggestion from Ron that should fix it immediately.
  • With the claw properly fastened to the carriage, drill a 1/8" hole through the base of the claw and at least 1/16" deep into the carriage. (Slightly deeper is OK, but don't try for as much as 1/8".)
  • Remove the claw and slightly enlarge the hole in the carriage. 9/64" diameter should be enough, and 5/32" is the maximum you want.
  • Press-fit a 1/8" x 1/4" roll pin into the hole in the claw base. Reassemble the claw to the carriage, and tap the roll pin in until it is at least flush with the claw, and perhaps even slightly recessed . (The drawing shows it recessed.)
The roll pin will prevent the claw from turning, but the assembly can be taken apart just as easily as before.

I use this shaft puller as a workhorse in my shop. It is easy to use, and more effective than other pullers I have used. (The only exception is a custom-made hydraulic puller with a massive clamp that was way too expensive for me to ever consider.) Thanks for the idea, Ron. It's a great device!


Ron Blanchard came up with the idea of using the part from Harbor Freight as the spindle and carriage, pushing against a vertical plane of a drill-press vise. This is the key to the whole design.

Ken Doyle of Mindscape Australia was the instructor for the video course that taught me how to use TurboCAD. In subsequent email, he gave me more information and considerable encouragement. He also told me about the Foxit PDF viewer, which is considerably better than the usual Adobe Reader.

Last modified - Dec 3, 2011