Measuring and testing – Dynamometers – Responsive to torque
Reexamination Certificate
1998-11-05
2002-08-27
Fuller, Benjamin R. (Department: 2855)
Measuring and testing
Dynamometers
Responsive to torque
Reexamination Certificate
active
06439066
ABSTRACT:
FIELD OF THE INVENTION
This invention relates to sensors that measure torque applied to a shaft. It particularly relates to sensors having a sleeve torsionally engaged with the shaft and sensing means responsive to torsional strain in the sleeve.
BACKGROUND OF THE INVENTION
Many known torque sensors operate by responding to magnetostrictive effects resulting from strain in a stressed member or transducer. Some of these are in commercial production. Efforts have been directed toward using magnetostrictive effects to measure the torque applied to the steering wheel by the driver of a motor vehicle. One known design torsionally engages a sleeve having desirable magnetostrictive properties to a portion of the steering wheel shaft. Another design uses the magnetostrictive properties of a current production steering wheel shaft to eliminate the cost of attaching a sleeve to the steering wheel shaft. In a third known design magnetostrictive material is beam or vapor deposited on a steering wheel shaft. The known designs have not proved entirely satisfactory. Known methods of attaching a sleeve require processes that are not easily adapted to large volume production. The same is true of beam or vapor deposition of magnetostrictive materials. Efforts to use the shaft itself have suffered from the difficulty of obtaining shafts consistently having desired magnetostrictive properties. For measuring steering torque in an automobile the ideal sensor would be inexpensive and compatible with existing steering wheel shafts.
The expression “torsionally engaged” is used herein to describe engagement between a first element and a second element for transmitting torque therebetween. It includes engagement for transmitting torque by a rigid attachment such as a weld or adhesive joint or both elements being made of one piece of material. It also includes engagement by means that transmit only torque exemplified by a wrench socket engaging the head of a bolt. The expression “torsionally engaged” is used to cover a broad range of torque transmitting engagement means that may or may not transmit forces in addition to torque.
A torque sensor incorporating a sleeve of magnetostrictive material is described in U.S. Pat. No. 5,351,555 issued Oct. 4, 1994 to Garshelis. Particular attention is focused on the Garshelis patent because it is believed to offer the lowest cost sensor responsive to torque applied to a magnetostrictive sleeve. However, the invention is applicable to any torque sensor having a sleevelike transducer that is torsionally stressed when torque is applied to a shaft.
The Garshelis design provides a sleeve (“transducer”) permanently magnetized in its circumferential direction. Garshelis discusses attachment of the transducer to the torsionally stressed shaft and (column 15 beginning at line 7) describes requirements which must be met by the chosen method of attachment:
“proper operation . . . requires that there be no slippage between any of the components at their interfaces. . . . Somewhat less obvious, but no less important, is the requirement that there be no inelastic strain in shaft 8 in any cross section which includes the transducer 4. Thus, all strains associated with the transmission of torque must be fully recoverable when the torque is relaxed.”
and in column 16 beginning at line 5
“As already indicated, the transducer 4 and underlying shaft must act as a mechanical unit. Rigid attachment of the transducer 4 either directly or indirectly to shaft 8 is crucial to proper operation”.
In fact, attachment by adhesive bonding (using known adhesives and known designs) or interference fit (Garshelis' preferred method) do not satisfy the above quoted requirements. All known designs based on adhesive bonding or interference result in peak stresses exceeding the capabilities of the bond.
In column 16 beginning at line 5 and continuing through line 23 of column 17 Garshelis discusses three categories of torsional engagements between the transducer and the shaft. The categories are 1) salient point, i.e. splines, knurls, teeth etc. at the ends of the transducer mating with similar features on the shaft; 2) distributed, i.e. adhesive bonding or interference fit; 3) diffuse, i.e. welding or brazing the ends of the transducer to the shaft. The first “1) salient point” and the last “3) diffuse” work well but manufacturing methods for achieving these attachments are not easily adapted to automotive manufacturing procedures.
About friction or adhesive bonding Garshelis states (column 16 lines 37 through 41):
“This bonding limits the maximum measurable torque to a lower value than might otherwise be handled by the shaft 8 alone or transducer 4 alone, but is advantageous for other reasons as indicated previously.”
Accordingly, Garshelis expresses a known need for an “advantageous” process such as adhesive bonding that does not limit the maximum measurable torque to “a lower value than might otherwise be handled by the shaft 8 alone or transducer 4 alone”. Garshelis goes on to state (column 16 lines 41 through 47):
“Press or shrink fits can be used to obtain the desired circular anisotropy, and can provide very substantial gripping forces which as a practical matter will not be broken by expected torques on shaft 8. With clean, degassed (and perhaps deoxidized) surfaces, the effective coefficient of friction can rise without limit and act somewhat like a weld.”
Providing “clean, degassed (and perhaps deoxidized) surfaces” on the elements before they are joined by press or shrink fits is expensive and time consuming. It is difficult to assure such qualities in many millions of parts as required for automotive production. It is not stated in the Garshelis patent but it is believed that to achieve in a press fit an effective coefficient of friction that “can rise without limit and act somewhat like a weld” as stated in Garshelis the “clean, degassed (and perhaps deoxidized) surfaces” must be joined and heat treated at high temperatures in a suitable atmosphere for many hours. To obtain a shrink fit heat treatment is believed to be required both to achieve an effective coefficient of friction that “can rise without limit and act somewhat like a weld” and to cause the shrinkage required for a shrink fit.
Another method of achieving an interference fit between the transducer and the shaft is described by Garshelis with reference to FIGS. 14, 15 and 16. In this method the shaft is hollow and an expander is drawn through the shaft to expand it thereby providing the desired hoop stress. This process also is believed to be difficult and expensive to implement in mass production of steering wheel shafts.
The following numerical examples will clarify the issues related to attaching a sleeve by adhesive or interference fit (without heat treatment or other processes to achieve an effective coefficient of friction that “can rise without limit and act somewhat like a weld”). In column 10 lines 3 through 5 Garshelis cites the example of a shaft diameter of 0.5 inch (1.27 centimeters) and a transducer wall thickness in the 0.030 to 0.050 inch (0.076 centimeters to 0.127 centimeters) range. The wall thickness is important to achieve sufficient magnetic flux (Garshelis column 10 lines 24 through 31). From the well known fact that torque transmitted by a shaft is distributed as the third power of the radius it follows for the case of the aforementioned 0.5 inch diameter shaft that if the transducer and shaft have similar shear moduli (which is likely to be the case) 36 percent of the total torque will be transferred to the transducer in the case of 0.030 inch transducer wall thickness and 52 percent of the total torque will be transferred to the transducer in the case of 0.050 inch transducer wall thickness. A possible diameter of a steering wheel shaft of an automobile is 2 cm and it might be subjected to a maximum torque of 600 newton-meters (450 ft-lbs). Such a torque might be applied by a large healthy male driver after the wheel reached the end of its travel. At the one centimeter radius of the outer surface
Fuller Benjamin R.
Thompson Jewel V.
LandOfFree
Torque sensor does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Torque sensor, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Torque sensor will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2885246