New developments have enabled automatic response time optimization, or “autotuning,” to quickly establish excellent flow control in a new or recently changed process.
Optimizing Response Time to Improve Flow and Pressure Control
Every control device has a characteristic response to disturbances in a system and to setpoint changes. For some devices, the response time can be optimized based on the specific process and environment. New developments have enabled automatic response time optimization, or “autotuning,” to quickly establish excellent flow control in a new or recently changed process.
Flow control quality depends on response time
A flow or pressure control system must be able to sense changes within a process and respond quickly enough to mitigate them. For some simple processes, the control system may be as basic as, say, a good pressure regulator coupled with an orifice to maintain a desired flow rate. Most systems, however, are more complex and require a control system to respond to the changes in pressures, temperatures, etc., that will affect the flow rate.
A flow device’s “response time” can be defined in various ways. Figure 1 shows a common definition of the control response time, as the sum of:
the “dead time” (the duration between the setpoint request and when the device begins to respond), and
the “time constant” (the time required for the device to move 63.2% from the initial value to the new value).
Figure 1: A common definition of control response time.
In general, a faster device response will correct disturbances in a system more quickly and will enable the device to track setpoint changes more closely. Figure 2 illustrates how various control devices may respond to a pressure wave moving through a system. The initial effect of the disturbance is the same for all four control devices; however, the device with faster response time (blue) minimizes the effect of the disturbance and returns the flow to the setpoint more quickly.
Figure 2: When an upstream valve opens, a control device with faster response time (blue) can mitigate the disturbance more quickly and precisely.
The example in Figure 2 also illustrates that process disturbances have unique characteristics which the control device must accommodate. Some control devices provide a single “speed factor” to optimize response time, which may be insufficient to match the characteristics of the disturbances. A device which enables greater control over the response can be optimized more precisely to the process conditions.
Optimizing flow control: A tradeoff between speed and range
A control device’s response time can be optimized for a particular set of system conditions. The optimized settings will ensure good flow control as long as the process conditions remain similar to that optimized state.
But conditions can vary, and that variability can significantly change the system characteristics. In many cases, then, it may be preferable to optimize the response time to accommodate a range of potential conditions, rather than maximizing the response for one particular set of conditions. This approach will often provide more consistent performance.
When to optimize flow control response speed
Optimization can improve flow control performance in a variety of circumstances:
At installation, to optimize the response time for the particular system
When the system is used in a different way; for example, if the system pressure is increased or decreased significantly, or if the process gas is changed
When system plumbing is significantly changed; for example, if a volume or restriction is added or removed
When a control device is moved between systems or experiments
When an exact response is required to protect a sensitive process
When multiple devices need to provide the same response within a connected system
When the needs of the process change, perhaps requiring faster or more precise control.
Figure 3 shows an example of how optimization can improve flow control. In this example, a flow control device was set up while the inlet pressure was 30 PSIG. If the inlet pressure is increased significantly (yellow and gray), the control response is degraded, as evidenced by the transients after the setpoint change. If the inlet pressure is reduced significantly (blue), the response may become too slow to respond to disturbances. In all of these cases, optimizing the response time again for the new conditions can restore the level of performance.
Figure 3: Increasing or decreasing the inlet pressure significantly after flow control has been optimized can impact flow control stability. Re-optimizing the control response time can restore performance.
Autotuning flow control response speed
A control device can be optimized by manually adjusting its control parameters. Manual optimization, however, typically requires a degree of knowledge about control system analysis and design. Additional equipment may also be required to measure the response time. Thirdly, manual optimization typically requires a significant amount of time and iteration.
In response to these issues, new methods of “autotuning" have been developed, incorporating a knowledge base of control expertise to automatically optimize a flow control device. Autotuning enables users who may not have extensive knowledge of control design to achieve excellent control for specific process conditions.
During autotuning, the control device moves to a series of setpoints. For each setpoint change, the device determines system properties and optimizes control parameters. When the process is complete, the device response is adjusted to the optimal settings.
Autotuning is much faster than most manual optimization techniques. As an example, most Alicat Scientific flow control devices that incorporate autotuning can complete the process in 30–90 seconds, though some ultra-low flow devices may require longer.
Advanced configuration options
For most situations, the autotuning algorithm will determine the best response time without any user input. The function can, however, be further configured to support atypical process requirements or to meet specific control goals.
The most impactful of these options is how autotuning will address the tradeoff between maximizing speed for one condition vs. handling a wider range of process variability. In Alicat flow control devices, for example, users can choose from several “Speed” settings:
FAST is the default option which balances speed and versatility for most situations. FASTEST maximizes response speed, allowing a small amount of overshoot. VERSATILE or MOST VERSATILE will provide response speeds that accommodate wider ranges of conditions, but with the tradeoff of slower response for some conditions. The control system may not be able to respond to quickly changing conditions. GOAL is an additional Speed option for advanced users who need to achieve a particular response profile or who need to tune multiple devices to provide the same response.
Figure 4 shows the response of a mass flow controller that has been optimized using these various Speed options.
Figure 4: Result of various Speed configuration settings.
Suggestions for achieving the best optimization
Autotuning is most effective when the process conditions maximize the pressure delta across the valve(s) involved. Unstable control may result when a device operates at a higher pressure delta or higher common mode pressure than it was optimized for. If operating conditions are expected to include higher pressures than are available for optimization, a speed level that emphasizes versatility is preferable.
Autotuning is more sensitive to fluctuations in the environment than normal closed loop control. Most fluctuations will result in control loop gains that are smaller than they might otherwise be, as it is difficult to separate the effects of the disturbance from the response of the system. Ultra-low flow and other slowly responding devices will be much more sensitive to disturbances or other fluctuations in the system.
Conclusion
Control device optimization can provide the best device response for a particular process and environment. Automated optimization, such as Alicat’s Autotune function, makes it possible to measure current performance and improve it, without requiring extensive knowledge of control systems and parameters. Advanced functions such as autotuning enable a broader range of process engineers and technicians to achieve excellent flow control, even if the process or environment changes dramatically.
About The Author
Brian Clendenin is senior software engineer at Alicat Scientific, Inc., leading instrument firmware development for mass flow and pressure controllers since 2014. He has over 30 years of experience in signal processing and control, beginning at Raytheon Missile Systems. Brian has a bachelor’s degree in Engineering and Applied Science from the California Institute of Technology and a master’s degree in Aerospace Engineering from the University of Michigan.
Did you enjoy this great article?
Check out our free e-newsletters to read more great articles..
