I am using the commercial library TIL to construct a Heat Pump cycle in DYMOLA. When integrating the components to form a closed cycle it fails to start (fails to reduce the DAE index). However when testing the components separately shows no such error. Can anyone help me to know the probable reason for such an error and suggest any method to overcome it.
Thanks in advance :)
Dear Aryan,
these kind of errors (in systems similar to yours) can be solved usually by introducing pressure losses around the affected positions. You might also want to consider to introduce a kind of mass/energy storage to your system as it will impove robustness and speed of simulation. Though I don't know what is inside the models you use in detail - probably there is already a storage included in the condenser model.
hope this helps,
Friedrich
Dear Friedrich,
At present I am using the basic TIL components with some calculated input parameters. Using an appropriate pressure drop model is possible with these components but I doubt if there is any option for mass/energy storage. Plz suggest a method for it. Will update you on status:)
Thanks
Dear Aryan,
I am not acquainted with the TIL library. However, I am working with hydraulic circuits in Dymola/Modelica. I agree with Friedrich about the pressure losses, but I think that in addition you wil need to set a reference for the pressure at some point (just as you would define a ground with 0V for electric circuit calculations). Otherwise the solver can only determine the difference between different pressures and this is what generates the strange DAE error message in my simulations.
Best,
Bram
Dear Aryan,
what is a bit unconventional with the TIL is the concept of pressure state calculation. As far as I know there only two position where the dynamic pressure is calculated (in your model the der(i) blocks, one for high and low pressure, respectively). All other pressures are depending statically on these high and low pressure states. I would expect that if you had no pressure loss at one of the heat exchangers between the pressure state points then this would result in a DAE index problem. Which components are affected by the DAE index problem?
Concerning mass energy storage: In principle the heat exchangers' balance equation will feature energy and mass storage (i.e. dynamic enrgy and mass balance are supported). What they probably not provide is a mean to outbalance larger density variation due to load changes. When the evaporator load is increased the mass in the evaporator will fall and this mass is stored usually in some storage device like a feedwater tank (large stationary applications) or a condenser hotwell (smaller, mobile applications). If you donot have this components or something similar providing this functionality then your system simulation will be a) difficult to initialize (aside from your DAE index problems) and b) difficult to control. In this sense the simulation is comparable to the real world! So you might want to check the library for some kind of separator model or two-zonal model
Hope this helps.
Friedrich
Dear Friedrich,
The point you suggested about the mass storage helped a lot. The initialization problem is solved by using a separator model (available in TIL library) at the condenser outlet. Though I can't understand why it is still showing errors when connected to the exit of evaporator which is a justified place to connect a separator (to ensure that only vapor enters the compressor). However the model works well for the steady loads(with open boundaries as shown in fig. attached). It further helped me to determine the actual state points. Now I am in a state to determine the pressure variations when connected to my system for heating purpose.
Thanks a lot ..!!
Hi Aryan,
can you post your wiring leading to the errors you mention and the error message?
The usage of the separator of course depends on its specific implementation... so you might want to check the documentation and/or the equation underlying the model.
Greetings, Friedrich
ps: consider voting for helpful asnwers... I think its a means in research gate to get some quality standard established :-)
Hi Aryan,
from your error message I learn that your initial values imply a) a backflow into the idealSeparator and b) an large pressure gradient in sim.dpdtPort. I suspect that the initial guess values for the pressures cause the difficulties here. You might also consider using a non-ideal separator having two serparate (non-isochor) control volumes for liquid and steam phase.
Hope this helps,
Friedrich
- Badges
- Science topic
- Similar topics
- Engineering
- Chemical Engineering