Robust hsu D flash method #2202
Open
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
The SaturationAncillaryFunction::invert method is used to provide guesses for several flash routines. However, if the inversion fails, the entire flash routine will also fail. This commit makes the routine more robust. Because it requires so many arguments, they have been wrapped into their own struct.
The exrapolating secant solver is needed to prevent issues when the Brent solver cannot bound the provided density, but in some cases the ancillary function has no solutio and the extrapolation goes very far out of range. This commit adds a limit to the result.
Perform more clipping to prevent the ancilliary curve solution from giving us a bad starting point.
Finding the saturation curve can still sometimes fail, despite our best efforts. However, we can simply brute force our way to a temperature solution through Brent's method. This is slow, and we do not know the phase, but it is better than nothing.
|
This is an impressive PR, addressing a particular point that has caused me a lot of pain. Thank you very much. Are there any points to discuss before merging? @jowr Do you want to take a look? |
GCC does not seem to support brace initializers for structs with default values. This commit removes the brace initializations, which also makes the style more consistent throughout the code.
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Description of the Change
Providing inputs to CoolProp of D and U (or any of HSU) is usually stable, but sometimes can cause annoying issues. This is noted in issues #1965 and #2157 (as well as some of my personal projects). This PR attempts to solve these issues once and for all as an extension to the work done in PR #2173. This is done in 3 main changes:
Benefits
We should now always get a solution for D_HSU inputs.
Possible Drawbacks
It is possible that one of our fallbacks does not provide the user with a phase, instead setting it to iphase_unknown.
The fallback solver can be slow because of needing a ton of Brent iterations, after already doing thousands of other iterations trying to find the saturation curve (which is now allowed to go to many of thousands of evaluations). However, a slow solution is almost always better than no solution.
Verification Process
I ran large sweeps of inputs to ensure that no errors were observed. I also integrated the results with another code and it seems to behave acceptably. I made custom test codes for issues #1965 and #2157. See the attached C++ code.
coolPropTester.zip
Applicable Issues
Closes #1965
Closes #2157