Traditionally, screwdriving is not as fast as other assembly operations, such as riveting or ultrasonic welding. That’s because screwdriving is full of variables ranging from different types and shapes of fasteners to different materials and different drive styles.

Throughput is synonymous with cycle time—the time required to produce a part or complete a process. It is often governed by the slowest machine in a workcell. However, engineers face many challenges when attempting to boost throughput in screwdriving applications.

“The challenges are reducing the time required to drive the screw, improving the joint quality (eliminating stripped or partially driven screws), and reducing waste from lost or dropped fasteners,” says Kevin Buckner, design engineer at Design Tool Inc.

“Improving throughput is all about trying to get more parts processed now than before,” adds Phil Klingler, product group manager at ASG. Div. of Jergens Inc. “Being able to produce more parts in less time improves productivity and reduces costs. However, everyone struggles with throughput.”

“Throughput is always a point of discussion with manufacturers, because everybody has cycle times that they’re trying to minimize,” notes Boris Baeumler, applications engineer at DEPRAG Inc. “They often want to buy the least amount of screwdriving equipment as possible to achieve faster cycle times.”

Poor fastener quality causes numerous throughput issues. For instance, problems with thread cutting and thread forming create variances.

“Manufacturers are still learning that cost and quality are not the same,” notes Klingler. “This causes many errors that can’t be easily or quickly resolved.”

“The No. 1 challenge to boosting throughput is screw quality,” claims Jan Aijkens, general manager at DEPRAG. “Screws that are out of tolerance will cause [headaches]. However, this is [usually] less of a problem with a hand feeder, since the operator is right there and most jams can often be easily cleared.”

“Fastener quality is one aspect of assembly operations that’s often overlooked,” adds Buckner. “When a manufacturer is using automation or handheld autofeed screwdriving equipment, quality issues with fasteners, such as size variance, eccentricity problems between the head and shank, bent screws, or partially formed screws, can cause feeding problems and downtime.”

Often, the cheapest screws are not the best solution in the long term. “If you buy lousy screws, it may create [jams] in the feeder mechanism,” Aijkens points out. “Some tolerances and dimensions are critical. Equipment might not work, just because somebody saved a few pennies along the way.”

Poor product designs also cause many screwdriving slowdowns. “The biggest mistake that engineers make when attempting to improve throughput is not considering the screwdriving process when designing the components,” says Buckner. “Screws will often be [placed] in locations that prevent either handheld or automated equipment from being used, which makes improving throughput very difficult.”

Sometimes, all it takes is a minor alteration. If engineers are willing to make changes to the material, the fastener or the application, they can improve assembly speed.

“Adding features to a part, such as a side pocket or a recess that can be used to align tooling, can make a big difference in throughput,” notes Baeumler. “A design may feature a screw that has a square drive on it that doesn’t engage well with the bit. Changing [the head] to a Torx or a cross-recess design can help.”

According to Baeumler, fastener size also affects throughput. “Generally, the larger the fastener, the slower the cycle time,” he points out. “It’s usually easier to improve throughput with smaller screws.”

And, the type of tool being used can influence throughput. For instance, operators can typically drive a screw faster with pneumatic drivers vs. electric drivers. That’s because pneumatic tools tend to have higher spindle speeds.

Spindle Speed

All things being equal, the easiest way to improve throughput is to increase the spindle speed on screwdrivers. For instance, if torque requirements can be met, the speed of the driver can be changed from 1,000 rpm to 2,000 rpm.

One solution is to reduce the amount of stroke to the minimum for the given application. “By reducing the amount of transit time the spindle spends between the home position and making contact with the part surface, the faster the cycle times,” says Scott Graham, marketing manager at Weber Screwdriving Systems Inc. “With the opportunity to adjust the amount of clearance and bit strokes, we can optimize the cycle time without compromising functionality and offer jam-free operation.”

Another way to speed up or optimize cycle time is to run the motor at a high rate of speed until the screw head is about one turn away from seating. “[Our] standard DC controllers offer both analog and digital depth integration that will perform an accurate linear measurement of the screwdriving bit relative to the part,” explains Graham. “This way, we can achieve the quickest rundown and accurately slow down at the very end of the driving cycle for precise torque and angle control.”

However, spindle speed is not always a one-stop solution. “One of the biggest mistakes that engineers make when attempting to improve throughput in screwdriving applications is increasing the speed of the driver,” says Neil Maniccia, global product group manager for ASG Precision Fastening. “Increased speed can impact the clamp load in the joint, which can cause failures.”

“Just turning up the dial can be a big mistake,” adds Baeumler. “Some applications just don’t allow for that, such as plastic screws.

“Some people will go from 700 rpm to 1400 rpm, because they think it will double their cycle time,” warns Baeumler. “But, in reality, that ‘solution’ may create problems. Because there are so many other processes involved with screwdriving, increasing spindle speed is just a small factor.”

“We generally see harmful stroke speeds and bit pressures when someone is trying to optimize or otherwise trying to squeeze every millisecond out of a screwdriving process,” says Graham. “Screws have a certain preferred forward advance velocity and bit pressure that will ensure that the torque readings being taken are true to what’s actually occurring on the screw head. By over-speeding the advance speed of the bit, it’s possible to do damage to the screw threads or application components.

“By applying too much downward force, the screw thread contact and prevailing torque can be artificially raised, causing false torque readings which can manifest into unseated screws,” adds Graham. “By using the correct bit-speed-to-thread-pitch ratio, and just enough pressure to keep the driver bit engaged, you can better trust any torque reading taken and greatly reduce the risk of damaging the screw thread in the process.”

Automation Options

Most experts agree that some sort of automation is required to speed up the screwdriving process. Options include screw presenters, bowl feeders, poke-yoke devices and robotics.

“Improving throughput requires automatic feeding of the screw,” says Baeumler. “Feeders present fasteners quickly and properly oriented to the tool.”

There are several different ways to get screws to the driver. A simple, inexpensive solution is the shake-tray, where screws flow over a perforated metal tray that tilts back and forth.

The next step up is a screw presenter that holds about 100 screws. A vibratory track orients screws and brings them to the end of a track where an operator can pick them up with a magnetic bit and then drive them.

A blow-feeder is the next step. An operator dumps screws into a hopper or vibratory bowl. The tooling orients each screw, brings it out to the end of a track, separates it and drops it into a feed tube. A blast of air shoots the screw to the end of the driver.

The fastest and most complex solution is to use an X-Y robot, along with a multibowl feeding system tied into a multispindle screwdriver.

“Automating the entire screwdriving process using fixtured equipment designed to drive all screws in an assembly can greatly increase throughput by reducing the number of operators required,” says Buckner. “This also frees the operator to perform other tasks while the screws are being driven.”

Both handheld and custom-fixtured machines are available. “In handheld applications, autofeed screwdriving equipment addresses these challenges by delivering the screws to the driver more quickly and eliminating the need for operators to handle the fasteners,” explains Buckner. “In fixtured applications, multiple drivers can be utilized to reduce cycle time or assemble multiple components simultaneously to achieve higher rates.

“Typically, screw presenters are used when the fastener dimensions prevent the screw from being blow-fed, or when the application does not allow [the use of] autofeed driver components,” says Buckner.

Screw presenters place a fastener at a pickup point so that the operator can insert a magnetic bit and engage the screw. The operator does not have to handle fasteners, which reduces waste from lost fasteners and improves quality by eliminating the chance of a dropped fastener entering the assembly. As soon as the screw is removed from the pickup point, the presenter advances another screw, so the next screw is available as the screw is being driven, which reduces cycle time.

However, screw presenters aren’t ideal for all screwdriving applications. “We can’t see where any presenter could actually improve throughput compared to a blow-fed solution,” says Graham. “Generally speaking, feeding a screw directly to a driving head concurrently while driving eliminates the need for the screwdriving system to go to a natural position to pick up a screw and will always save time and increase throughput.”

Another option is to use poke-yoke devices, which can improve throughput by eliminating assembly mistakes. “Poke-yoke systems generally make checks and balances to verify . . . part placement and other quality control functions,” Graham points out. “This [reduces] failed parts due to operator error.”

However, Graham says poke-yoke processes can sometimes impact throughput in a negative way. “It depends if they can be performed concurrently with other functions,” he explains. “When they can’t, the overall process is generally slowed down.”