Simulink: Adding expansion valve and try to close the refrigerant loop but fail to converge with initial conditions

Hello!

I've been following a tutorial, trying to build up a closed refrigerant loop. The first attachment was the tutorial.

For my loop, the followings are my parameters:

evaperator inlet h4:271.32 kJ/kg P4:0.3 MPa

evaperator outlet h1:373.382 kJ/kg P1:0.3 MPa

compressor outlet h2:413.958 kJ/kg P2:1.5 MPa

condenser outlet h3:271.32 kJ/kg P2:1.5 MPa

heat source for evaperator: 4.7 kW and i calculated my mass flow:0.047kg/s

cross-sectional area for evaperator is 1.5131cm^2(hydraulic diameter 1.388cm)

cross-sectional area for condenser is 0.49265cm^2(hydraulic diameter 0.792cm)

pressure ratio:5

nominal shaft speed:3800

I made the condenser a pipe that can just take away all the heat flow, so i minus the enthalpy signals (evaporator inlet and compressor outlet) ,and product it with the actual flow rate to get the condenser heat source signal.

For the evaperator>compresser>condensor part (second attachment), it was great and i also followed the valve test step before i make it a closed loop.

But after i add the orifice(third attachment) and put in the initial conditions(just like the tutorial, initial pressure 1018.4kPa, vapor quality 0.2467 for condensor and evaporator), it failed to converge.

If i used the initial conditions "specific enthalpy" (evaperator h:271.32 kJ/kg P:0.3 MPa, condenser h:413.958 kJ/kg P:1.5 MPa), the mass flow rate will be around 0.02kg/s instead of 0.047kg/s and simulation will be asserted because internal energy will be greater than the valid maximum internal energy at the compressor outlet.

I have no idea how to make it work, it's been bothering me for days. My goal is to make the model like the tutorial but with my own P-h parameters. Please help me out thank you!

Answer 1

[Rishi]

Hi Brendon,

I understand from your query that you want to know why you are getting the error saying that convergence is failed with initial conditions.

The error you are encountering occurs when your model is too ideal and there is a sudden transition. In such cases, the Simscape solver needs to go through the ‘transient initialiation’ phase. The initial guess might be too far from the result, consequently failing to converge. You can find more about it in the below documentation:

https://www.mathworks.com/help/simscape/ug/how-simscape-simulation-works.html?searchHighlight=bql2kkl-2.html&s_tid=srchtitle_bql2kkl-2.html_1#bqoxlpd-1

To resolve the error, you can try making the following changes:

• Reduce the time step and see if it helps the solver to converge.
• Try decreasing the Consistency Tolerance parameter value in the Solver Configuration block.
• Consider adjusting component parameters to make the model more realistic and less ideal.

To learn more about Transient Simulation Issues, you can refer to the following documentation:

https://www.mathworks.com/help/simscape/ug/troubleshooting-simulation-errors.html#:~:text=and%20Techniques.-,Transient%20Simulation%20Issues,-Transient%20Initialization%20Not

Hope it helps.

Answer 2

[Yifeng Tang]

Hi Brendon,

I'm concerned that the way you are taking heat out of the condenser may generated unintended consequences and make the cycle unstable. It's true that in steady-state, what you are calculating is the exact amount of heat exchanger, but in transient, e.g. initializing, the flow may not have been established so the energy flux you get may make the state inside the condenser to an unrealistic point.

In steady, I would suggest that you:

• assume a reasonable ambient temperature at the condenser, implemented as a temperature source.
• use the same test harness idea to parametrize the pipe so it transfers the expected amount of heat for given inlet 2P state and flow

Then try to close the loop.