Return to Index

Making Chuck Springs
George Langford, Sc.D.
February 2, 2009; copyright 2009
The Hjorth spring winder:
US Patent No. 861,238, 3 inches overall length.

Hjorth spring winder, US Patent No. 861,238
Jorth spring winding tool in use
The Hjorth spring winding tool both guides the wire onto the mandrel and also applies drag to aid in reducing the amount of springback, thereby reducing the difficulty of sizing the spring.

As shown in the image at left, the mandrel was first clamped in a chuck on the spindle of an Ultra-Rapid drill clamped in a bench vise, and then the free end of the wire was slipped between the jaws of the chuck.  While the spindle was being rotated, the flower-shaped wheel rode on the mandrel with its web (suitably selected to control the spacing between successive wraps of the wire) and the U-shaped guide clamped the wire to provide drag.  The wire was further guided by the little hole in the tool handle.  Naturally, one doesn't want any loops in the wire that might grab a finger ...

This wire is 0.020 inch in diameter, so the force to hold the tool was almost imperceptible.  Similarly, hardly any torque was needed to turn the mandrel.

The chuck jaws are shown below:
Millers Falls No.2 chuck jaws

The original springs were sized to fit the holes in each face of the jaw, but may have either been (a) too small an outside diameter, (b) too large an outside diameter, or (c) too many loops.  Any one of these errors would tend to cause the springs to buckle and become crushed between the jaws when they were tightened.  Or a user simply got careless while inserting a drill bit.  I measured these holes, first with a depth micrometer whose spindle thankfully is smaller than these holes, and then by finding the largest pin from the set shown below, left, that would fit into the holes.   I found that the depths of the holes were not uniform (ranging from 0.115 to 0.146 inch) but that the sums of adjoining pairs were better (0.252 to 0.261 inch). 

Therefore, I picked 0.25 inch as the maximum compressed height of the springs, and 0.140 inch as the maximum coil diameter.  The smallest wire I had was 0.020 inch.  That gave a spring design of 12 turns of 0.020 inch wire with a maximum outside diameter of 0.140 inch.
These are the wire-gauge pins:

Pin set, wire gauge
Here are the trial spring coils (wire diameter/mandrel diameter):
Trial springs, using Hjorth spring winding tool
The finished springs (about 3X actual size):

Finished springs in place
The first three trial springs above were too large in their outside diameters, but the bottom two springs, made on a 0.089 inch mandrel, came out OK.  I also made another set using the 0.093 inch mandrel, but with more drag, that also came out OK.  The trial springs all have longer tails than shown above - the rest is hidden because I poked them through the paper to hold them in place.

I didn't have a handy empty original chuck, but I found one that is "close enough."  When I pushed the jaws with their springs all the way to the small end of the chuck shell, the jaws came together without binding the springs, and the free lengths didn't change perceptibly.  The relaxed position of the three jaws and their springs at left is sufficiently wide (as shown at left) that they'll always be a snug fit in the actual chuck shell.

I did not attempt to square off the ends of these springs, which would have required that I first make a closed-coil spring and then stretch it to obtain the desired free length.  This would have left the spring with objectionable residual stresses and would have also wasted two or three coils that could not contribute to the compressibility of the spring.

The original patent is reproduced on the next page.

The drag adjustment not only makes the coils more closely wrap around the mandrel, but also reduces the residual stresses that make the spring fail to recover all of its original length after compression.  This is accomplished because the central portion of the wire's diameter remains elastic while the outside is being bent; the greater the amount of tension applied to the spring while it is being coiled, the smaller the size of the remaining elastic portion in the center of the wire.