I built this device in 2004.
Sam Geffner described a "combination rappel and prusik device" in 1961 (The Tech Explorer, 5(2), 13-15; reprinted in the 1961 Speleo Digest). Sam described two versions - a single-ended one and a double-ended one. He did not provide engineering drawings. This descender is based on his descriptions; however, I made a few modifications of my own. I will note these in the descriptions.
The frame is made from two pieces of 1-1/2" by 1/8" (32 by 3.8 mm) 6061-T6 aluminum. [Sam recommended using 3/8" (9.5 mm) steel, but did not specify if this was the width, thickness or both]. The front plate is 81 mm. long; the rear 115 mm. These are connected by two 1/2" (12.7 mm) long, 3/4" (19.2 mm) diameter 6061-T6 aluminum brake bars, held in place by 5/16" (8.2 mm) stainless steel hex-head cap screws with nylon-insert lock nuts. I threaded both brake bars, and arranged the bolts so that the friction of the rope would act to tighten the bars on the cap screws. A 5/8" (16 mm) hole drilled in the back plate provides a clip-in point.
The middle brake bar pivots on a lever made from 1" by 1/8" (25 by 3.8 mm) 6061-T6 aluminum. My lever pivots on a cylindrical area turned into the lower stainless steel nut [Sam used a lever that pivoted on a third bolt. His sketches do not clearly indicate how this lever clears the bolt for the lower bar. One sketch shows an L-shaped lever, but it is not aligned to provide clearance. Two other sketches show straight levers]. There is a third brake bar, 3/4" (19.2 mm) in both diameter and length, mounted on the end of the lever arm. This brake bar is threaded as well. Note that the hex-head cap screw passes in the opposite direction as the others - this is to ensure that rope friction acts to tighten this bar as well. I added an aluminum washer (cut with a hole saw) at the end of the brake bar to provide a positive engagement to help the brake bar resist lateral motion [Sam's did not have this feature]. A flat cut in the washer provides rope clearance for rigging.
The Geffner is an old design dating from the days of body rappels and prusik knot ascents. It is a simple device, easy to manufacture at home, but lacking in modern sophistication. In practice, it acts much like any other brake bar rig. The rope has a greater contact angle than it does with a single carabiner with brake bar, so it provides more friction. The designer can vary the friction by spacing the two fixed brake bars closer (more friction) or farther apart (less friction); however, getting them too close together (so that the rope is pinched between the moving and fixed bars) will cause the device to lock rather than run smoothly.
Mounting the lever on a separate pivot requires making the device longer; however, by doing so, one could choose dimensions that would allow the arm to move to either side of the device. This could pose a hazard, since the brake bars would try to unscrew on one side. If you try this, please pin the bars. An advantage is that one could align the lever with the frame for easier packing.
Sam describes using the device as an ascender. In 1961, the idea was worth trying. I cannot imagine why anyone would want to do that in this day in age, except as an academic exercise.
MiscDesc937.html/070318
©2007, Gary D. Storrick