Session: 09-03-01: General Topics on Engineering Education

Paper Number: 119724

119724 - Modeling Refrigeration Systems With Simscape and Matlab: a Component-by-Component Approach 

Modeling refrigeration systems has been growing in popularity as it has become easier and faster to model two-phase fluids. It has been embraced for component sizing, to reduce development costs, and by controls teams to develop next-gen systems without the need of physical testing. The modeling of refrigeration has also enabled academic pursuits of researching  new configurations and refrigerants by allowing a simpler path without the need for complicated fabrication or instrumentation. It can also be used in school or training to build a better understanding of each component and the refrigeration cycle as a whole. Simscape provides many solutions for modeling refrigeration systems easily from system-level analysis all the way to fully parametrized geometric components. The Simscape model (Figure 1) can then be used within Simulink to develop a control system around it to obverse its behavior.

In this talk, we will demonstrate a workflow in MATLAB and Simscape to build a refrigeration model component by component based on measured data.

-          Determine the pressure-enthalpy (P-h) diagram: System sizing starts with looking at P-h diagrams and deciding the conditions the system will be operating in. These initial values can be plugged into the system-level refrigeration cycle (Figure 2) to quickly get the model running and verify measurements are near the desired performance. This is most useful for system-level sizing, supervisory control design, and defining requirements, all without the necessity to model the component-level details of the refrigeration cycle.

-          Set up test harnesses and parametrize components: In order to model a closed-loop refrigeration cycle, it is crucial to parametrize and size each component correctly before assembling them into a loop. Each component is parametrized by inputting known geometric values and then creating a test harness. The test harness can define boundary conditions and then use the Parameter Estimator app to determine unknown values based on real-world or desired performance data.

-          Assemble and test subsystem models: Subsystems are useful for verifying that components will work together in the full system and are simpler to initialize and run by defining known boundary conditions. Control systems can also be developed using these to run quicker or focus on individual components.

-          Close the refrigeration loop: The components are assembled to close the loop and initial conditions are calculated. Multiple ways are demonstrated. Initial vapor quality is calculated using REFPROP or a script is used to measure the volume of the system and calculated with desired refrigerant charge. Compressor speed and expansion valve position can be easily modified to make the system meet desired performance values.

-          Implement closed loop controls: A simple PID controller is implemented to control the expansion valve based on superheat. The control system is contained within its own subsystem and can be easily modified or changed to another controller strategy.

Presenting Author: Andrew Greff Mathworks

Presenting Author Biography: Andrew Greff is a senior application engineer at MathWorks. He specializes in physical modeling using Simscape and focuses on thermal, fluid, and multibody systems. Before joining Mathworks, Andrew worked for GM and Stellantis developing advanced hardware and controls for engines. He obtained his PhD in mechanical engineering from the University of Alabama.

Authors:

Andrew Greff Mathworks