Equipment Optimization

Apart from their critical importance in control design and tuning, our physics-based models serve several other purposes. Two important applications are modeling for equipment design, and modeling for performance optimization.

Modeling for Design

What design features limit the current performance of my equipment, and what changes do I need to make to attain the desired level of performance?
A key goal of our customers in designing the next generation of process equipment is attaining substantial improvement in performance over that of the existing equipment. From this perspective, the starting point of the design process is framing the question, “What aspects of the existing equipment are limiting today’s performance?” Here, SC Solutions’ modeling expertise can provide meaningful answers. Our physics-based models incorporate the parameters (geometrical features and material properties) of the equipment that play a dominant role in determining performance, e.g., how heat transfer at the edge of a plate affects the temperature uniformity on a wafer. Models can help determine how the number and location of actuators and sensors affects achievable performance. Using a model of the existing equipment, it is relatively straightforward (and inexpensive!) to run simulations and perform exhaustive sensitivity analyses to these key system parameters that determine critical performance. These analyses provide the necessary guidance on making the modifications to the system that may be needed in order to achieve an improved level of performance. Performing such analysis in simulation rather than prototyping hardware not only is a significant time and cost saving, but is generally more exhaustive in terms of finding the best solution moving forward.

Modeling for Optimization

How do I maximize my performance, increase my throughput and guarantee reliability over a wide range of process conditions?
As in the case of equipment design, physics-based modeling is an excellent way to optimize equipment operation to improve performance and throughput. SC Solutions employs the latest advances in optimization techniques together with the physics-based models to perform comprehensive analyses, both in open-loop and in closed-loop with feedback control. These studies indicate how feedback and/or feed-forward control can maximize performance of the existing equipment, and how throughput can be increased. In addition, such simulations can provide insight into the robustness of the process to varying process conditions, which in turn may indicate how process reliability may be improved.

Where do I place my sensors and actuators in my system?
For multiple actuator and multiple sensor systems, our modeling and optimization capabilities can help address the important issue of the placement and number of actuators and sensors. These studies can help determine the locations of sensors and actuators which maximizes performance, e.g., achieves maximum temperature uniformity across a wafer undergoing thermal processing. This capability is especially relevant to many semiconductor processes where manufacturers have to meet stringent performance metrics while minimizing the number of sensors and actuators – the latter stemming from cost, physical access, and reliability considerations. In some applications, we have used our models as virtual sensors to supplement the limited number of actual sensors available.

How do I tune my controller to achieve an optimal balance of throughput, performance and reliability?
When it comes to controller tuning, whether it involves a PID controller or a more complex model-based controller, one of the main tuning challenges is to find an optimal balance between achieving performance and throughput versus achieving reliability of the controller over many process runs and a wide range of process conditions. Often, the process goals are in conflict, and a systematic analysis of the trade-offs involved is needed in order to specify the requirements. Use of simulations of the physical model of the equipment or process is a very effective way to conduct these trade-off analyses. Many ‘what-if’ scenarios can be tested rapidly and comprehensively in simulation. Together with SC Solutions expertise in control theory, the model-based simulation approach provides a powerful systematic tuning method that helps maximize equipment performance.

How do I select my filter parameters to reduce noise without impeding performance?
Sensors that monitor or control certain critical performance variables of a process are often noisy. The common approach is to filter out the noise. However, filtering introduces a certain amount of delay in the signal. While this delay is not an issue in slower processes, filter dynamics can significantly limit performance of fast processes. Examples of the latter would be rapid heating and point-to-point motions. For our customers, we select and design filters that provide sufficient noise reduction without limiting system performance, either as part of the controller design process, or as a stand-alone data acquisition solution. Examples include Kalman filter design, outlier detection, and fault-detection/fault accommodation. Model-based simulations are instrumental in providing insight for distinguishing system dynamics from filter dynamics.