When implementing a robot into a dispensing line, there are many issues a manufacturer may face. These must be addressed, whether integrating robotic dispensing into an existing manufacturing line or designing an entirely new manufacturing line from scratch. For example, it is important to take a close look at what production workers are, or would be doing manually, and how those tasks translate into an automated setting. Ultimately, automating a dispense process will increase both throughput and process reliability. But robots are not without their limitations, and they must be implemented intelligently for manufacturers to reap their full benefits.
When looking at a robotic dispensing application, there are two main issues that need to be addressed: the type of dispenser a manufacturer is going to use and the type of robot that will control it. Assemblers upgrading an existing process may think they already have the first question settled. But many times incorporating a robot means upgrading your dispensing system as well.
Many manufacturers use manual or semiautomated syringe dispensers. An inherent issue with these kinds of equipment is imprecise shot size and excess flow after the air has been shut off. This may be acceptable in a manual process where operators can remove any excess adhesive, but when that same operation is automated, this can cause problems.
When dispensing cyanoacrylates, for instance, any excess adhesive leaking out from a joint will eventually crystallize into a light, chalky residue, or "bloom." This creates an aesthetic problem when assembling consumer goods like safety glasses or a makeup case with a mirror. In the context of electronics assembly, this can pose a functional issue.
To solve these kinds of problems, manufacturers can use precision diaphragm valves or positive-displacement valves that employ servomotors and small electromechanical controllers to control dispensing. Diaphragm valves can be particularly effective in robotic applications, thanks to their ability to incorporate precise shutoff and "suck back" features.
When dispensing two-part adhesives, the adhesive will start to cure as soon as the resin and hardener mix, which means an increase in viscosity. Most two-part dispensers operate using pressure and time-regulated dispense technology; however, with this approach, as the viscosity goes up, the adhesive volume dispensed will go down. Therefore, it is generally best to choose either a two-part adhesive with a long open time or implement a positive displacement dispense system. Positive displacement dispense systems are typically more expensive. But, because they dispense according to a set volume, as opposed to regulating pressure at a consistent level for a set period, they guarantee consistent shot volumes.
Types of Robots
Once a manufacturer has decided on a dispenser, it's time to focus on the type of robot. Critical factors to consider include the number of axes dispensed upon, the type of process flow, part placement, size, speed and cost concerns.
If an existing process is dispensing a single dot onto a part, then a one-axis robot, or vertical-axis slide station, may be suitable. Specific examples of this kind of application include medical tube bonding or needle assemblies that require a small drop of cyanoacrylate or light-cure acrylic adhesive to be dispensed onto a specific location.
In these cases, the part will be positioned under the robot, and the robot will locate to the same dispense position each time. The amount of adhesive will be uniform from part to part. Although this may seem like a basic application, increased consistency in the location and amount of adhesive dispensed can improve quality and decrease scrap rates.
A robot with two axes typically moves up and down and from side to side, increasing speed and accuracy when dispensing basic, straight-line patterns. Two-axis robots can be used for less complex applications like dispensing adhesive onto box lids-an important consideration when packaging is included as an integral part of an assembly line. They can also be used for "tacking," or creating simple, adhesive paths in temporary or nonstructural assemblies.
A three-axis robot generally moves up and down, from front to back, and from side to side. Most assemblies require three-axis dispense systems. For example, if you have a filter and need to dispense a gasket of silicone around the edge, you can implement a three-axis robot integrated with a pail pump and precision valve to place the gasket on your filter in a consistent location. A three-axis machine may also be the robot of choice for joining enclosure parts in a product like cell phone assembly.
The fourth, fifth and sixth axes are all related to rotation, and introduce expanded capabilities for a jointed arm. More complex, four-plus axes robots allow manufacturers to address angled, multiplane and rotating dispense requirements.
Typically, when you increase the number of axes on your robot, the purchase price and cost of repair also increases, while ease of programming decreases. For these reasons, it is usually best to determine exactly how many axes are needed and pursue that type of robot. Bear in mind, though, that in some cases a manufacturer may want to use a single robot for numerous applications that require differing numbers of axes, either immediately or somewhere down the road. If this true, it may be worthwhile to invest in a more complex robot, with an eye to the future.
Automated dispense systems can accommodate both batch and in-line flow processes. The two main types of robots used are benchtop and selective compliance assembly robot arm (SCARA) robots. Benchtop robots typically have an attached tabletop, and the movement of the table controls one axis while the other axes are controlled by a Cartesian gantry. SCARA robots are designed specifically for in-line processes, but can be mounted over a stationary table and used in the same way as a benchtop robot. Both robot types are comparable in terms of speed and precision. In most cases, cycle times and accuracy will be limited by the type of dispense system chosen, not the robot.
Correct part positioning prior to dispense is essential to the success of an automated system. In fact, the accuracy of adhesive placement is directly related to positioning precision. Unless it is equipped with machine vision, a robot will move to the exact same place every time. Therefore, if part position is not consistent, adhesive placement will be inconsistent.
Typically, the goal of automated assembly is to speed the manufacturing process. Because different operators' part change-out speeds may vary, simplifying the process will minimize differences in part change-out speed and overall assembly time. Concentrating on the design of the part fixture is probably the best way to decrease operator variability. This includes creating a fixture that makes it easy for operators to place and remove the part, while at the same time ensuring parts are correctly configured.
When automating an in-line process, parts must be positioned on a conveyor in a consistent manner. If the conveyor does not have a locating device to ensure that all parts are configured the same way, the robot must have a vision system to determine their location and orientation-which adds a great deal of cost.
In some cases, machine vision will be used to determine how the part is configured, based on various locating points. Once this has been done, the controller converts the program to dispense onto the part, based on the part's orientation. Other robotic systems will use machine vision to move the part to the correct configuration instead of translating the dispense program.
When using a batch process, maximizing the number of parts fixtured on a single tray is clearly the best way to go. However, while manufacturers want to have as many parts per tray, or batch, as possible, there must be sufficient space to allow operators to insert and remove parts on the tray.
It is also important to take the cure mechanism of the adhesive into account. For example, if a hot melt adhesive cures in 15 seconds, a limited number of parts should be placed on a single tray to ensure that components can be assembled before the adhesive sets. The same is true for a two-part adhesive with a short working time. By contrast, with a 3-minute-cure hot melt, the longer working time means more parts can be placed on a tray.
When short adhesive open times are a limitation, implementing an indexing table or an in-line process may provide the best solution. This allows the part to be moved to the next manufacturing step right away instead of waiting for the entire batch to be dispensed.
What Size Robot Do I Need?
In many ways, robot size and fixturing are closely related. To determine the size robot you need, consider how many parts can fit in a given work envelope while ensuring that the payload specifications of the tool head are not exceeded. In most cases, a larger robot is a more expensive robot, so manufacturing engineers will want to do a cost-benefit analysis when putting multiple fixtures on a single tray. Having more fixtures means freeing up an operator for longer periods of time; however, the cost of having a robot with enough reach to get to all those parts may not be justified.
Obviously, when dispensing onto larger parts that are independently fixtured, the situation is much more straightforward. A manufacturer simply needs to employ a robot that can reach the entire dispensing path.
When looking at robot payloads, it is extremely important to stay within a robot's rated limits. Many robots operate by stepper motor. If the payload exceeds the specification, the stepper motor may get off track, and the robot will either not move or dispense in areas other than where it is intended to dispense.
Remember that the mass of an end effecter can generate a substantial inertia on a robot arm, especially in high-speed applications. While the situation is not as dramatic as that in some heavy-duty material applications, like manipulating engine blocks, engineers still need to be careful. Hot-melt dispensing systems, for example, include a lot more hardware than a simple one-part, diaphragm dispenser, so dispensing robots for these applications need to be selected accordingly.
Integrating Automation into Existing Processes
When creating an automated dispense system, many components must be integrated into the process. Integration options frequently provide for a more efficient workplace, but bring with them additional cost and increased programming complexity. The first integration-dispensing equipment-is essential to the system. Other integration options may assist in system control, curing, inspection and safety.
A system can operate via programmable logic controller (PLC) or a PC. In both cases, the technology can also be used to control the rest of the manufacturing system. Curing systems can be automated as well. For example, integrating a light source such as a UV- or visible-light-cure wand may eliminate another step in the process, thereby decreasing operator time.
Automation must never reduce quality. Incorporating inspection equipment will ensure that a process is stable. Flow monitors that measure pressure changes in a dispense valve are another option. These devices may create a sound or turn off the program if an inconsistency is present.
Automating a process with fast-moving, heavy machinery often requires certain, basic safety measures. Common safety equipment can include part-present sensors that only allow a program to run when a part is in position. Light curtains that stop the program when they are broken ensure that robot arms and end effectors won't injure their human operators.
Once an automated dispense system has been designed and assembled, the final step is programming. While programming a robot may seem like a difficult task, training from the equipment manufacturer may simplify and speed the process. The latest generation of dispensing software incorporates many user-friendly features that make basic dispensing path programming a snap.
Keep in mind that the greatest challenge can be in the details when creating an effective dispensing path. For example, if you are dispensing a high viscosity silicone gasket you want to pay close attention to the knit line where the start and end of dispense come together. Some tactics used to minimize the knit line are to dispense at a slower speed and closer part distance at the beginning of the gasket, then at a faster speed at the end, with the dispenser shutting off before the robot stops traveling.
Similarly, programming tactics for dispensing through sharp turns may be dependent on the viscosity of the material. If you are dispensing a bead around a sharp corner and the viscosity of the adhesive is high, you may need to have your dispense needle closer to the substrate to ensure that it is being placed on the substrate in the correct locations.
While the dispense program is the main focus, needle-height calibration, system purge and initialization may also require programming. The distance between the part and the dispense needle is critical to the accuracy of the dispense pattern. Adhesive change-out and needle replacement are inevitable in an adhesive dispense system, although the frequency of these changes depends on adhesive usage and cure mechanism. For example, needle changes will be more frequent with a fast-curing cyanoacrylate than with an acrylic that only cures after exposure to UV light.
To ensure the needle is in the same position after each change, end users must program needle-placement procedures. One method is to program the robot to move to a specific position, then, once the robot is in place, locate the needle with the help of a reliable reference point, such as a flat region on a fixture. There are also dispensing programs that can adjust the dispense path depending on the location of the valve and needle tip, although these add quite a bit of cost.
Adhesive needle purging is considered standard procedure for most adhesive chemistries. If the adhesive has the potential to cure in a static needle, then frequent purging of the idle system is typically recommended. For example, if you are dispensing a cyanoacrylate through a needle, the potential for the adhesive to cure in the tip over time is fairly high. Assemblers can avoid frequent needle changes by programming the robot to dispense a small amount of adhesive every few minutes whenever the robot is sitting idle for a given length of time.