3-D Sprags and Rollers
EVA Roller Wrench

Commercial Roller Wrench
Awards and Recognitions
Project Support
Project Personnel
Other Applications
Other Resources
More Information

wrench home

e-mail us

Neutral Buoyancy Laboratory Evaluations

During the Hubble Space Telescope Third Servicing Mission Neutral Buoyancy Laboratory (NBL) engineering development tests of NBL 98.2 (NBL 98.2-1 and NBL 98.2-2), we evaluated the extravehicular activity (EVA) three-dimensional (3-D) roller wrench.

The test setup, shown in Figure 1, consisted of a 24" wide by 36" high by 1/8" thick task board mounted to a test stand with foot restraints. The distance from the bottom of the test stand to the top of the task board was 84". The foot restraints were mounted parallel to the task board and pitched forward 15°. The task board, shown in Figure 2, consisted of four tasks which are described below. The work envelope containing the four tasks was centered on a line parallel to the floor that extended 55" from the centerline of the bolt pattern on the foot restraint. The board was placed 23" from a line perpendicular to the floor extending through the center of the same bolt pattern.

Figure 1 ¥ 3-D Roller Wrench Test Setup

Figure 2 ¥ 3-D Roller Wrench Task Board

For the NBL 98.2 wrench evaluations, the following tools were available (part numbers are in parentheses):

  • 3/8" Drive International Space Station (ISS) Ratchet (SEG33106927-303)
  • 3/8" Drive 3-D Roller Wrench (SSL-RW-17-0200-02) This wrench was locked permanently in the loosen (counterclockwise) direction.
  • 3/8" Drive 3-D Roller Wrench (SSL-RW-17-0200-03) This wrench was locked in a position such that the wrench could apply torque in both directions (i.e., a rigid bar or non-ratcheting tool)
  • 7/16" x 3" GFE Socket (10159-10084-01)
  • ISS Ratchet Cheater Bar (TBD)
  • Tape Measure

Task 1. Confined work site range-of-motion

Task Setup
A confined work site range-of-motion task was used to compare the ability of a suited subject to turn a double height 7/16" hex bolt when the range of throw of the wrench was limited to 15° and 30°. Two wrenches were compared, the ISS Ratchet wrench and the Roller Wrench. Since the ISS Ratchet and Roller Wrench have different handle neck sizes, the top set of limit pegs were used for the ISS Ratchet and the bottom set of limit pegs were used for the Roller Wrench. The test subject was requested to give an acceptable/unacceptable rating for both wrenches in turning the bolt in each of the two ranges of motion. Figure 3 shows the confined work site task in greater detail.

Figure 3 ¥ Confined Work Site Range-of-Motion

During the First Servicing Mission (SM1), Story Musgrave commented on how the struts at the Solar Array Carrier forward and aft latch work sites constrained his throw while using the Essex wrench.


  1. Set the ISS Ratchet Wrench in the loosen position with a 7/16" x 3" GFE Socket attached.
  2. Place the handle of the wrench between the 15° limit pegs.
  3. Turn the bolt 5 turns. The number of turns can be determined from the rod extending perpendicular from the boltÕs body.
  4. Place the handle of the wrench between the 30° limit pegs.
  5. Turn the bolt 5 turns.
  6. Repeat steps #2 - #5 with the Roller Wrench (SSL-RW-17-0200-02) and the 7/16" Socket.

Task 2. Maximum EVA free-floating torque

Task Setup
While free floating, the subject was asked to apply the maximum amount of torque possible using the ISS Ratchet Wrench in one hand while holding onto the handrail with the other. The resulting torque was measured with a torque sensor, as shown in Figure 4, and the torque data was sent to the surface through a cable and recorded on a computer.

Figure 4 ¥ Maximum EVA Free-Floating Torque

The work site around the center Multi-Use Lightweight Equipment Carrier-to-Radiator bolts (which nominally release at 15 ft-lbf) cannot be reached by the Remote Manipulator System, and it is undesirable to set up and take down a Portable Foot Restraint so the task may have to be done free-floating.


  1. Set the ISS Ratchet Wrench to tighten and place the square drive tang into the torque sensor with the wrench parallel to the handrail and with the handle to the right.
  2. Remove boots from foot restraints.
  3. While holding the handrail with one hand, apply maximum amount of torque using the wrench in the other hand.
  4. Repeat step #3 with the wrench in the loosen position and then in the tighten and loosen position with the handle to the left.

Task 3. Ability of an EVA subject to manually control high-inertia, low-friction interfaces

Task Setup
The goal was to rotate a double height 7/16" hex bolt (shown in the middle of Figure 5) connected to a 22:1 planetary gearhead and a 20-1/2" diameter, 1/2" thick copper plate in an attempt to simulate the deployment of the Solar Array-III (SA-III) which has an inertia of approximately 286 slug-ft^2. First the subject was asked to rotate the bolt from 180° (wrench handle parallel to the floor and to the left) to 90° (wrench handle perpendicular to the floor pointing towards the water surface) at approximately 5°/sec and stop the interface at the 90° mark. This simulates the SA-III panel deployment. The task was repeated at a quarter the angular rate or approximately 1.2°/sec to simulate engagement of the SA-III at soft dock. The test subject was asked to comment on their ability to control inertia using both the ISS Ratchet and Roller Wrench.

Figure 5 ¥ High-Inertia, Low Friction Interface Simulation

This was done in order to evaluate the need for a ratchet lock-out feature for manual operation of high-inertia interfaces. During SM1 Solar Array-I (SA-I) Primary Deployment Mechanism (PDM) retraction, Tom Akers had limited control of the SA-I, apparently because he was using a ratcheting wrench to drive a high-inertia, low-friction interface. He commented that he would move the wrench 10° and the interface would move an additional 30° before friction caused it to stop. In the event that there is a manual PDM retraction on the Third Servicing Mission, an evaluation must be done to determine if a tool should be available that does not ratchet thereby giving the crew member positive control of the array in both directions. Some involved with the development of Solar Array-II think that a non-ratcheting tool is the safest way of handling the PDM.


  1. Set the ISS Ratchet Wrench in the tighten position with a socket attached and with the handle parallel with the floor and to the left (180° position).
  2. Turn the bolt clockwise at 5°/sec. The rate can be determined using the movement of the rod extending perpendicular from the boltÕs body against the dial attached behind the bolt.
  3. Attempt to stop the bolt at the 90° mark.
  4. Repeat steps #1 - #3 turning the bolt at 1/4 the rate specified in step #2 (approximately 1.2°/sec).
  5. Repeat steps #1 - #4 with the Roller Wrench (SSL-RW-17-0200-03).

Task 4. Aft Shroud Door Latch work site simulation

Task Setup
There were two objectives of the task: determine the ability of a weightless suited subject to apply 60 ft-lbf of torque required to sheer the Aft Shroud Door Latches (ASDL) using the ISS Ratchet attached to a cheater bar; and evaluate the impact of a ratchet wrench on the useful throw by measuring the throw envelope while simulating shearing of the ASDL bolts. The torque sensor simulated the ±V2 Scientific Instrument ASDL bolts. It was centered 0.74" parallel from the surface of the plate to mimic the ASDL bolt's distance from the door surface and the female end of the sensor was 16.5" to the right of the vertical handrail (simulating the location of HR16 on the +V2 side and HR7 on the -V2 side) as shown in Figure 6. Torque data was sent to the surface from the sensor through a cable and recorded on a computer. The useful throw envelope was measured with a tape measure and recorded.

Figure 6 ¥ Aft Shroud Door Latch Work Site Simulation

Because of the pitch of the teeth on the ratchet wrench and because it is attached to the cheater bar, the useful throw of the wrench may be severely limited and the handle of the wrench may even move outside of the EVA work envelope before the wrench locks. Because of the roller wrench's almost instantaneous locking, the useful throw can be increased. There may also be a need for the roller wrenchÕs smaller throw over that of the ratchet wrench for EVA applications that require multiple throws of a long-handled manual tool. For instance, the large range of motion needed to drive or fail the ASDL bolts using a long handle may be especially fatiguing in a suit.


  1. Remove the palm wheel from the ISS Ratchet.
  2. Attach the wrench to the ISS Cheater Bar.
  3. Set the wrench to loosen and place the square drive tang into the torque sensor.
  4. Pull the wrench toward you until the wrench locks.
  5. The perpendicular distance from the edge of the task board to the end of the wrench handle will be measured and recorded.
  6. Apply the maximum amount of torque using the wrench.
Torque Data Acquisition System

Torque was measured with a 3/8" square drive socket extension reaction torque sensor. The sensor consisted of a full, four-arm Wheatstone Bridge circuit, with an elastomer coating to protect against mechanical damage, attached to a standard 3/8" drive socket extension. The gages were attached to a 100 foot, four conductor, shielded water resistant cable designed for submersion in 50 feet of water. The sensor, also designed for submersion in 50 feet of water, was covered with a special water resistant coating. The cable was routed to the surface, along the deck away from the waterÕs edge, and up to the hallway outside the control room and under the control room door. The pigtail terminations were plugged into the back of a strain gage meter with built-in transducer excitation and an isolated RS-232 communications link. An RJ11 4-wire serial line ran from the strain gage meter to a DB25 serial connector which connected to a DB25 to DIN 8 cable. The cable was plugged into the serial port of a Macintosh laptop computer. Power for both the strain gage meter and Macintosh PowerBook was provided through an AC power cord which was plugged into a Ground Fault Circuit Interrupter. The data acquisition system used is shown in Figure 7.

Figure 7 ¥ Torque Sensor Data Acquisition System

Accuracy of the torque sensor is 0.18% of the 600 in-lbf full scale value or 1.1 in-lbf. The accuracy of the strain gage meter is 0.005% of the recorded value. Pre-test calibration plots for both sensors appear below.

A paper summarizing the test was written:

  • Roberts, B., "Three-Dimensional Roller Wrench and Human Performance Neutral Buoyancy Laboratory 98.2 Test Results," SSL Document Number 99-001, March 1999. [ paper (878 K PDF) | presentation ]