The injection molding industry is a diverse, competitive market where innovative products must meet demanding requirements. For many companies, these requirements lend themselves to cautionary approaches to new products, and a robust testing strategy to minimize operational failure. All of these practices can result in a slower time to market and higher development costs. To remain competitive, these companies are adopting new, model-based techniques that inform design choices with more accuracy than ever before.

A leading injection molding company recently developed a new machine that could meet strict performance requirements with smaller, cost-effective motors. The new design would save money with smaller motors, but it imposed new controller requirements to avoid costly machine damage. To accomplish this new design without cost or time overruns, the company used a model-based design strategy called virtual commissioning. With this strategy, a physics-based model of the new design could be used to inform the precise requirements for motor sizing, and could also serve as an effective testing platform for the new, complex controllers.

Without their own virtual commissioning expertise, the company turned to software and development services from Maplesoft and B&R. Using MapleSim, the modeling and simulation tool from Maplesoft, a dynamic model of the machine was developed. With services from B&R, the model was created, which provided the virtual basis for the machine’s motor sizing and control strategy.

Cost-Effective Motor Sizing for New Motion Profile

Using MapleSim, a physics-based model of the new design was created (Figure 1). The company provided B&R with the details of their new mechanism, including geometries, material specifications, and the proposed motion profile of the machine. With this information, the MapleSim model was created to test for the loading requirements of the motors. By importing the company’s existing CAD information into MapleSim, the model was set up and customized in a matter of days, allowing motor sizing to begin sooner than the company’s traditional processes would have allowed.

Figure 1: The MapleSim model of the new mold-closing mechanism included mechanism geometries, motor specifications, and other details to ensure a realistic simulation.

To ensure the accuracy of the MapleSim model, the customer provided data from their previous motion profile for this mechanism. By demonstrating the same loading requirements for the traditional motion profile, the model was shown to have high degree of realism, which gave the company confidence to move forward and explore their new, proposed motion profile for this mechanism.

Figure 2: The simulation results in MapleSim were able to show the motor requirements given the proposed motion profile, in addition to a 3-D visualization window for quick visual inspection.

The new motion profile was implemented in MapleSim, and the results can be seen in Figure 2. The simulation results automatically display a 3D visualization for a quick design check, and the subsequent motor loading requirements are displayed alongside the visualization. These results were formulated into a common Speed-Torque graph (Figure 3). In this graph, the machine’s duty cycle (green) is seen to be well within the company’s motor requirements (blue). This information assured the company that, at every point in the machine’s proposed motion profile, the motor would be operating within safe limits, and that the motor itself was sized without an unnecessarily large margin of error.

Figure 3: The Speed-Torque graph, as seen here, demonstrated that the required motor torque (green) would stay within operating characteristics of the specified motor (blue) throughout the machine's duty cycle.

Control Strategy Testing

The new machine would split the motor loads in half by using a dual axis closing mechanism, helping reduce the costs of a single, larger motor. A key requirement in this new design is that both axes remain in parallel during the opening and closing of the mold. If either axis becomes misaligned, the machine risks being significantly damaged. To prevent this damage, the control strategy must ensure axis alignment under a variety of conditions. To reduce the risks of machine damage, virtual commissioning techniques were used to test the controller performance against the physics-based model that was previously created.

To prepare the MapleSim model for virtual commissioning, it was refined to reflect the new dual-axis design (Figure 4). The new mechanism was duplicated and connected to a mold subsystem, which modeled the mold as a spring-damper system. This subsystem provided the required amount of force on the mechanism as it proceeded to close the mold and experience the pushback of the injection molding material.

Figure 4: By doubling the mechanism in MapleSim, the new closing mechanism was modeled in preparation for the control design in B&R Automation Studio.

The MapleSim model was then exported as a Functional Mock-Up Unit (FMU), which acts as a standalone, executable model that can be used in a variety of other design tools (Figure 5). The FMU was imported into B&R Automation Studio, where it functions as a variety of inputs (motor torques) and outputs (sensor data) that can run in realtime for control testing.

Figure 5: The Function Mock-Up Unit (FMU) acts a set of inputs and outputs that can be used by automation software for control design and realtime testing.

Once in B&R Automation Studio, the accuracy of the new model and controller were tested to ensure that they ccould provide meaningful results for the company’s goals. In the weeks following the motion profile testing, the company created a physical prototype of their new dual-axis mechanism. While this prototype would be able to provide them with some testing information, the key piece of information - controller robustness to keep axis alignment - would impose signification machine damage, making repeated testing an extremely expensive choice on a physical prototype. By using a virtual model, these risks can be avoided.

Figure 6: The MapleSim model, upon importing into B&R Automation Studio, demonstrated highly accurate realism when compared to a physical prototype.

The results of the realism testing can be seen in Figure 6, which show a very close fit between the model performance and the physical prototype performance. This accuracy gave the customer confidence to proceed with the controller robustness testing. By inserting a customizable friction block on only one of the axes, the model could be adjusted to simulate for a variety of conditions that would give rise to misalignment. The controller testing was performed using a variety of friction values, and the controller’s impact on axis alignment was shown to be successful at bringing the axis into alignment after frictions were introduced (Figure 7). By testing the controller against the virtual model, countless simulations can be run in a fraction of the time it would otherwise take with a physical machine, saving time and money to ensure a robust, reliable machine.

Figure 7: The results of controller testing showed that an environmental factor, such as uneven frictions in the axes, could be quickly accounted for by the controller.

Conclusion

Virtual commissioning can help companies explore, test, and validate new designs, all while reducing the time and money spent on product development. In this case, the company was able to create a cost-competitive injection molding machine using a new dual-axis design with smaller, inexpensive motors. By using MapleSim, the precise loading requirements of the motors were simulated, ensuring the company had selected the right motor for the job. To avoid the costs and delays of damaging prototypes for controller testing, they were able to ensure the controller’s robustness by using virtual commissioning. Taken together, these techniques have given this company a powerful new set of design tools, helping them continue to stay one step ahead in an already competitive market.