To successfully apply cam indexing drives, consider more than just the indexing torque and bearing life. Watch out for these potential pitfalls and know what to do about them.
The plant superintendent for an automotive gasket supplier was perplexed as he stood beside the indexer service technician. Before them was a $350,000 tableau of precision die sets, fixtures, and structural steel frame interlaced with hydraulic piping and electrical conduit. In the center was a $9,000 cam indexer, thought to be perfect for the application: generous torque capacity, predicted bearing life into the next ice age, and accuracy exceeding the specifications.
When new, these rotary dial machines achieved station accuracy of ± 0.004 in., but now three were off by almost twice that amount. One machine was consistently off-station by more than 0.250 in. — at each of its eight work stations! Though the error on three machines wasn’t compromising part quality yet, the fourth could no longer produce good parts. So it sat idle, with zero output and zero revenue.
The technician examined the indexer for run-out, shaft end play, and backlash. The assembly appeared to be in good working order. Finally, after several hours, he found the answer beneath a layer of grease and dust: there were no positioning dowels between the indexer and machine base. The indexer was held in place only by four bolts with enough hole clearance to cause trouble. It no longer lined up with scribe lines on the base, having shifted within the bolt hole clearances.
Correcting the problem was easy. Properly aligned and held in position by tight dowels, the indexer would now provide many years of reliable service.
This story shows the importance of going beyond ordinary machine design practices when applying cam indexing drives. The repeated reversing loads inherent in these drives invite a gaggle of gremlins not found in continuous motion systems.
As a designer, stopping these gremlins means you must carefully select the drive components, evaluate any external loads on the indexer, and consider the operational effects of emergency stops and jogging.
Be sure to install gear reducers, motors, and torque limiting clutches so they will perform as expected under the repeated reversing loads. Finally, look beyond the machine acceptance tests and anticipate the eventual use — and abuse — of the indexer on the plant floor.
Accommodating external loading
Cam indexing drives have static and dynamic torque ratings, plus bearing load ratings. Unfortunately, suppliers usually publish only the predicted life for the cam follower bearings at rated cam speed. But, be aware of other hidden loads and ratings when selecting a cam indexer.
For example, an unwary designer might select an indexer that has excellent predicted bearing life based on a slow index time. However, the unit may be unable to withstand the forces of drill heads pushing down on the work stations, which are located at a distance from the dial center. In this case, a gremlin surfaces as a moment load on the dial that exceeds the rating of the output shaft support bearings, Figure 1.
Manufacturing processes such as rivet forming, stamping, pressing, die cutting, drilling, reaming, sawing, and some types of welding, produce moment and thrust loads and, to some extent, radial loads, depending on the orientation of indexer to peripheral equipment. Identify these external loads and work with vendors to select an indexer that satisfies both transfer time and loading requirements.
If a load exceeds the cam indexer rating, add backup structures to isolate the indexer from the load. For axial (thrust) or moment loading, back up the dial with anvils or support blocks that clear the bottom surface of the dial by 0.005 to 0.008 in. (depending on dial flatness) when indexing. When a load is applied during dwell, the dial flexes and bottoms out against the anvil, easing the load transmitted to indexer output shaft bearings.
Where an external load is applied tangent to the dial at a distance from its center, Figure 2, the resultant force on the cam follower is Fr = Ftan × (WR/PR), where Ftan is the tangent force, WR is the work station radius, and PR is the cam follower pitch radius.
Tangential loading generates a shear force on the cam follower stud and a compression force on the needle bearings in the follower head. Extreme force can cause stud failure, cam deformation, or reduced indexer life. To counter tangential forces, either clamp or shot-pin the dial so the clamp or pin takes the load. Either method greatly reduces or eliminates the force on the indexer.
Continue on page 2
Shot pins and cam relief
Designers often use a shot pin device to counter tangential forces. Actuated by a mechanical cam or pneumatic cylinder, this tapered pin engages the dial after it stops at a work station, Figure 3.
The use of shot pins opens the door for another gremlin. As the shot pin engages the dial, it rotates the dial slightly, causing stress in the cam followers. This occurs where cam followers are preloaded on the cam during dwell. The cam follower stud deflects under a tangential force exerted on the dial by the shot pin. In time, this causes brinelling of the cam surface and fatigue in the follower stud.
To accommodate shot pins in preloaded systems, the cam rib is ground 0.002 to 0.004 in. per side to permit free play of the followers as the shot pin moves the dial during dwell, Figure 4. The relieved portion is usually half of the dwell period length, and it is centered about the mid-point of the dwell. This leaves a short dwell segment at each end of the relief where the cam followers make a gentle transition into preload before entering the next index portion of the cycle. Common practice mandates at least 15 deg of preload dwell between the ends of the relief and the cam index portion.
Power transmission considerations
To ensure smooth, controlled indexing motion between stations, plus consistent station accuracy, consider factors such as cam shaft speed and gear backdriving.
Cam shaft speed. As the cam rotates, the highest displacement and velocity usually occur at the midpoint of the index part of the cycle. Because of this displacement, any fluctuation in cam speed causes an increased output torque. These torque fluctuations evidence themselves as erratic dial motion during indexing, and shuddering — or an audible grinding noise — as the dial nears a station. The repeated shock torque and vibration caused by this fluctuation diminishes the indexer life.
Manufacturers of cam indexers recommend holding the cam shaft speed nearly constant, within 3 to 5%, which can be achieved by conventional drives.
Most of today’s electric motors operate at speeds within the tolerance required for indexing. With dc permanent magnet motors, the speed remains near constant with a well-regulated drive controller. AC motors run well and at fairly constant speed with constant supply frequency. Servomotors with feedback devices provide excellent consistency in speed but are more costly than conventional motors.
Some designers favor fluid power systems to run cam indexers. Again, consistent shaft speed is important when choosing these components. Many high-quality hydraulic motors, with suitable flow controls, maintain constant speed for cam indexing. Air motors, however, are not well suited for use with cam indexers because of excessive speed fluctuation.
Gear backlash and backdriving. The relatively large inertia load of a dial tends to backdrive through the cam to the speed reducer as the load decelerates. Therefore, when selecting a geared speed reducer for the input side of the indexer, chose one with the lowest possible backlash. For this purpose, cam indexer vendors favor double-enveloping worm-gear speed reducers, with precision matched sets of worm and gear.
Such reducers that have single-stage ratios of 40:1 or higher are considered to be self-locking, meaning that they usually can’t be driven backwards. If self-locking gears are used, the gearset can lock-up during dial deceleration, imposing severe shock loading on the reducer and driven equipment.
Moreover, if a gear set is backdriven with the worm running at a speed below the manufacturer’s suggested minimum for its ratio, the indexer output motion can become erratic. This phenomenon, known as “stairstepping,” produces unacceptable camshaft speed fluctuations with the adverse effects described previously.
To reduce the likelihood of shock loading and stairstepping, work closely with suppliers of cam indexers and worm-gear speed reducers to determine the highest singlereduction ratio that can be used at the desired worm speed. Where stairstepping is likely to occur, mount a low-ratio (up to 6:1) reducer on the input side of the worm-gear speed reducer. This permits using a worm unit with a lower ratio and one that is not self-locking.
Continue page 3
Finally, size the reducer to handle indexer input torque with a minimum gear service factor of 1.5 for the application. If additional equipment such as toggle presses, cam devices, or conveyors will be powered off the indexer cam shaft extension, make sure the reducer can handle the worst case torque combination (not all devices generate peak torque at the same time) of these accessories in addition to the indexer’s required input torque.
Designers generally apply torque limiting (overload) clutches to indexing drives where the dial can hit — or be hit by — tooling, robot arms, press rams, and stamping platens. These clutches are intended to protect the indexer only, not the human operator nor the processing equipment tooling. The added expense of a clutch is easily justified by the potential cost in machine downtime if the indexer sustains damage in a crash with tooling or a transfer mechanism.
Determining which type of clutch to use, and its location in the drive, is equally important to sizing the clutch for torque. The two types of clutches commonly used with indexers are the friction (slip) clutch and the detent-release clutch. The former is better suited for use on the input side (cam shaft) of an indexer, and the latter for the output side.
Most engineers favor using a detent clutch mounted on the indexer output flange or shaft, which is seemingly regarded as being closest to the load and therefore more responsive in sensing an overload.
Detent clutches have rings or plates with pockets into which are seated detents — balls or tapered plungers. Torque from the driving ring is transmitted through the seated detents, under spring pressure, to the driven ring. The clutch disengages when there is sufficient torque to drive the detents out of their seats, causing the rings to separate. The trip torque depends on the spring pressure and can be adjusted in most units.
These clutches have a small clearance in the detent seats, approximately 0.0005 in., to let the detents move in and out. To determine the amount of clearance at a work station, multiply the detent clearance by the ratio of work station radius to detent pitch radius. For example, if a work station is centered at a 24-in. radius and the clutch has balls at a 3-in. pitch radius, the clearance at the work station radius is 0.004 in. [(24/3) × 0.0005]. This clearance, combined with the cam dwell runout, gives the work station positioning accuracy provided by the indexer-clutch combination. Some cases may require a larger diameter clutch (with detents at a larger radius) to obtain the necessary work station accuracy. In such cases, you may need a larger indexer to mate with the larger clutch even though a smaller unit would handle the indexing torque.
Using an output clutch also requires careful attention to external loading on the dial. In processes involving spot facing, cut-off saws, bending, and forming, be sure to secure the dial against tangential loading. A large tangential force can produce a torque larger than the clutch’s trip setting, causing it to release and the dial to break away. For thrust or moment loads — as in pressing, stamping, or riveting — back-up the dial from above or below, depending on the direction of force. Otherwise, the loads can cause the dial to rock on the clutch detents and compromise work station alignment.
Contaminants such as weld spatter, chips, and grinding dust impair the working of a precision detent clutch mechanism. Also, corrosive washdown solutions such as caustic and water, as well as some food and confectionery products, cause the clutch mechanism to rust or even seize. In such environments, installing a cap over the clutch will provide limited protection against contamination.
Avoid using detent clutches on the input side of an indexer that is driving a large inertial load such as a dial with fixtures. If an operator stops a cycle as the dial moves between stations, the clutch may disengage and the cam will no longer be restrained by the input drive components (motor and geared reducer). This loss of control lets the dial backdrive the cam as the dial decelerates. Due to the high ratio of displacement between cam and dial, the cam may be backdriven at a speed faster than the cam shaft design speed. The cam then free-wheels at high speed through dwell and coasts into the index part of the cycle, where it attempts to transfer energy back to the stopped dial. This causes a large enough shock load to damage the cam followers.
By contrast, a friction clutch is used to limit torque on the input side, often on indexers located in the adverse environments discussed previously. This type of clutch is fairly well enclosed and less susceptible to contamination. The inputside clutch is also suitable for applications that require extreme indexing accuracy. Because the clutch is not located between indexer and dial, there is no loss of positioning accuracy.
Continue on page 4
Indexer vendors usually furnish a friction clutch integral with the speed reducer that they supply with the indexer. The clutch is either a multiple-plate or conical type. In either case, adjust the trip torque setting by turning the nut loading the springs against the friction surfaces. Power is transmitted from the speed reducer through the clutch to the indexer cam (input) shaft.
Emergency stops and jogging
In normal operation, the cam profile controls dial motion as the cam rotates at constant speed. Cam indexers handle the load smoothly through the index period and bring it to a controlled stop as the cam followers enter the dwell part of the cycle.
Making a rapid emergency stop (E-stop) while the dial is between stations introduces another gremlin into the system: the excessive torque generated in stopping the dial more rapidly than it would otherwise. An E-stop is usually necessary when an operator removes protective guarding or penetrates a light curtain or, in some cases, to shut down a machine that is not running properly.
During an E-stop, the drive comes to an abrupt halt, subjecting the cam to a rapid change in velocity. The cam deceleration is usually unknown or, at best, has just been guessed. The deceleration rate of the dial, then, becomes a combination of:
• Deceleration of the dial imparted by the cam if it were running at constant speed.
• Unknown deceleration of the cam as it rapidly slows from operating speed to zero.
This compound deceleration rate can generate torque through the indexer that is several times higher than the indexing torque.
Where the E-stop torque exceeds the indexer rating, the unit can sustain damage, usually a shear failure of the cam follower stud or fracture of the outer bearing race. If the unit continues to run with broken followers, the fragments can gouge or knurl the cam surface, which impairs — or halts — indexer operation. Repeated E-stops, especially during the high-velocity portion (mid point) of the index, leads to elongation of cam follower mounting holes in the hub, which causes binding, backlash, and reduced accuracy.
When an E-stop is initiated to protect a machine operator, it is imperative that the dial be brought to rest as rapidly as possible. if an E-stop time is known, you can calculate the maximum torque generated during the stop based on the stop occurring at maximum dial speed. This torque is higher than normal, which requires that a unit be sized larger than the indexing torque would dictate.
An E-stop can cause a detent clutch to disengage, with the dial rotating out of control until friction or a crash brings it to rest. This dangerous situation can cause operator injury and equipment damage. For this reason, the friction clutch described earlier for use on the indexer input shaft may be a better choice in applications requiring indexer protection and E-stop capabilities.
Jogging the indexer during machine set up, tool changes, and production affects the unit in much the same way as an E-stop, but with added shock loading due to acceleration on each restart that occurs while the cam is in the index mode. As with an E-stop, the compounding of torques can produce an output torque in excess of indexer capacity.
Nearly all applications require jogging, even if only to debug a newly built machine. Simple provisions in the electrical control system can permit jogging without damaging the indexing drive. One such means is a secondary speed-regulating circuit parallel to the primary circuit. The secondary circuit runs the indexer well below operating speed so that it generates less torque during jog starts and stops. When the machine returns to run mode, the primary speed circuit is engaged and once again lets the cam run at the faster speed desired for indexing.
As discussed earlier, the most important distinction between indexing equipment and continuous motion systems is the repeated reversing action of the load during indexing. If backlash or clearance exists in any part of the drive, the reversing load works the mating parts back and forth slightly at first, and in larger amounts as the machine continues to operate.
Continue on page 5
Make sure the indexer is installed with dowels in addition to the usual bolts. At assembly, installers mount the dial on the indexer mounting flange pilot and fasten it with cap screws, Figure 5. At this stage, the cap screws are only partially tightened to allow machine set up and dial position adjustment. Once alignment is complete, the installer drills and reams indexer flange holes to match the dowel holes in the dial. A tight fit between dowels and holes ensures torque transmission with no play.
The dial fits tightly over the indexer pilot in accordance with the dial bore tolerance specified by indexer manufacturers. For this reason, use tapped jack screw holes in the dial. If removal of the dial becomes necessary, bolts can be threaded into these holes to evenly lift the dial until it clears the pilot.
Installers should mount the indexer on the machine base in like fashion, with dowels, but no pilot. Because torque generated by the indexer causes an equal but opposite torque on the machine base, these dowels should fit tightly so that the indexer can’t move relative to its bolts. This was the lesson that our gasket manufacturer learned at the beginning of this article.
Any sprockets, gears, pulleys, and other drive components that are secured to shafts only by key and setscrew will eventually begin to work back and forth on their keys. This causes knurling or enlargement of the keyway and breaking of the key edges. This wear eventually causes the key to roll over in the keyway so that no torque is transmitted through the failed connection.
To insure that rotating elements remain solidly fastened to shafts in indexing systems, use clamp collars or split bushings for a tight friction fit. If rotating elements will be installed on a flange, apply dowels and cap screws or bolts as described above.
Edward S. Novit is a field sales engineer for CN Sales Inc., a Chicago-based manufacturer’s representative for Ferguson Co.