Hello there,
I have two questions in this area,
1. How to avoid the probelm ''Excessive temperature change in Fluent. I am unable to understand the cause for this error. When Hybrid Initilization is done, this problem is observed. The case is when convergent nozzle exit gas expands in open atmosphere.
2. Secondly, how to model the gas flow between two containers in Fluent connected through a nozzle. From fluent i could only understand the flow is simulated between inlet and outlet. But i want to study the behaviour of the containers as well.
Any leads in this direction?
Regards,
Naveen
Flow between two containers (closed vessels) can only be induced, if the containers/vessels are initialized at different pressure levels and if they are initially separated (e.g. by a diaphragm) and where this separation is removed at the start of the flow observation / simulation.
This can be simulated, if the flow conditions in your configuration are manually initialized with the corresponding conditions before starting the simulation. Then the pressure difference will drive the corresponding flow through the nozzle. Of course hybrid initialization can not be used for this purpose. The result will be a transient flow, which continues until the pressure between the two (closed) vessels will be equalized.
Hello Thomas,
Many thanks for the very precise answer. Could please info on how to initialize the conditions manually? and if there is valve how can simulate opening the valve? Is it dynamic meshing?
In Fluent for comparable simple initialization tasks, there is the so-called "patching" procedure. It should be accessible through the Fluent menus or the simulation tree on the left hand side of Fluent. But as I said - it is ment for only rather simple initialization tasks. Otherwise you need to write an initialization UDF in Fluent, which assignes the required values based on some (x,y,z) criterion.
In CFX this would be more easy, since CCL expressions using step functions or even more complex functions can be used in order to create the desired initialization. But you mentioned, that you are bound to Fluent so...
If it is required to simulate the (slow) opening of the valve, than this would require a dynamic mesh simulation, where the cross section in the valve would open in accordance to some law over time. If the nozzle cross section is opened rather fast (fast in comparison to the overall duration of the flow), than one could start the simulation with the fully openend cross section and the mentioned domain initialization. At least that would be a much quicker first guess on what might happen for your application.
Regards,
Thomas Frank.
Dear Thomas Frank ,
Thank you very much; it solved many doubts i had on simulation initialization and valve opening.
- Badges
- Science topic