equinor / pysand Goto Github PK

To ease integration with pysand-web, adding a display name to the material properties dictionary would be of great help (pysand.erosion.material_properties):

{ 'carbon_steel': {'rho_t': 7800, 'K': 2e-9, 'n': 2.6, 'angle_dependency': 'ductile', 'name': 'Carbon Steel'}, 'duplex': {'rho_t': 7850, 'K': 2e-9, 'n': 2.6, 'angle_dependency': 'ductile', 'name': 'Duplex'}, 'ss316': {'rho_t': 8000, 'K': 2e-9, 'n': 2.6, 'angle_dependency': 'ductile', 'name': 'SS316'}, 'inconel': {'rho_t': 8440, 'K': 2e-9, 'n': 2.6, 'angle_dependency': 'ductile', 'name': 'Inconel'}, 'grp_epoxy': {'rho_t': 1800, 'K': 3e-10, 'n': 3.6, 'angle_dependency': 'ductile', 'name': 'GRP Epoxy'}, 'grp_vinyl_ester': {'rho_t': 1800, 'K': 6e-10, 'n': 3.6, 'angle_dependency': 'ductile', 'name': 'GRP Vinyl Ester'}, 'hdpe': {'rho_t': 1150, 'K': 3.5e-9, 'n': 2.9, 'angle_dependency': 'ductile', 'name': 'HDPE'}, 'aluminium': {'rho_t': 2700, 'K': 5.8e-9, 'n': 2.3, 'angle_dependency': 'ductile', 'name': 'Aluminium'}, 'dc_05_tungsten': {'rho_t': 15250, 'K': 1.1e-10, 'n': 2.3, 'angle_dependency': 'brittle', 'name': 'DC 05 Tungsten'}, 'cs_10_tungsten': {'rho_t': 14800, 'K': 3.2e-10, 'n': 2.2, 'angle_dependency': 'brittle', 'name': 'CS 10 Tungsten'}, 'cr_37_tungsten': {'rho_t': 14600, 'K': 8.8e-11, 'n': 2.5, 'angle_dependency': 'brittle', 'name': 'CR 37 Tungsten'}, '95_alu_oxide': {'rho_t': 3700, 'K': 6.8e-8, 'n': 2, 'angle_dependency': 'brittle', 'name': '95 Alu Oxide'}, '99_alu_oxide': {'rho_t': 3700, 'K': 9.5e-7, 'n': 1.2, 'angle_dependency': 'brittle', 'name': '99 Alu Oxide'}, 'psz_ceramic_zirconia': {'rho_t': 5700, 'K': 4.1e-9, 'n': 2.5, 'angle_dependency': 'brittle', 'name': 'PSZ Ceramic Zirconia'}, 'ZrO2-Y3_ceramic_zirconia': {'rho_t': 6070, 'K': 4e-11, 'n': 2.7, 'angle_dependency': 'brittle', 'name': 'Zr02-Y3 Ceramic Zirconia'}, 'SiC_silicon_carbide': {'rho_t': 3100, 'K': 6.5e-9, 'n': 1.9, 'angle_dependency': 'brittle', 'name': 'Silicon Carbide'}, 'Si3N4_silicon_nitride': {'rho_t': 3200, 'K': 2e-10, 'n': 2, 'angle_dependency': 'brittle', 'name': 'Silicon Nitride'}, 'TiB2_titanium_diboride': {'rho_t': 4250, 'K': 9.3e-9, 'n': 1.9, 'angle_dependency': 'brittle', 'name': 'Titanium Diboride'}, 'B4C_boron_carbide': {'rho_t': 2500, 'K': 3e-8, 'n': .9, 'angle_dependency': 'brittle', 'name': 'Boron Carbide'}, 'SiSiC_ceramic_carbide': {'rho_t': 3100, 'K': 7.4e-11, 'n': 2.7, 'angle_dependency': 'brittle', 'name': 'Ceramic Carbide'} }

• Sand transport jupyter notebook
• Bringing it together notebook
• Readme update

Add CD functionality for automatic versioning and pushing to PyPi when merging PR's.

Implement something like codacy. Github Actions may be used?

(raw-zero)^exp. default exp=1.

4.63 from DNVGL RP-O501

Create a pseudo model based off DNVGL work for non-slam check valves. Inlcude warnings on sonic velocity.

The description of Hydro transport model currently reads:
Equinor sand transport model for horizontal pipelines Based on T. Søntvedt (1995) and R. Schulkes (2002) work in Hydro

This is not 100% correct, as the model appear to have been influenced by other test-data, and is not able to re-produce 1:1 lab measurements from the mentioned Schulkes report. New description should reflect this

At the moment there is a logger warning when geometry factors above 4 in erosion.py. In reality one could use higher GF for components to adjust for special cases like jets.

Extend Github action "test pysand" with a build test to test.pypi.org. Ref to tagreader.

Then another function to convert to mm/year

currently tested on version 3.6 on Travis. Add following to readme. Version number should be updated in setup.py classifiers. Possible to update setup.py classifiers from travis.yml?

https://shields.io/category/platform-support

Liquid density and liquid viscosity is used for horisontal sand transport models. Should be included in the fluid properties module.

When sand production or velocity input is not known, minimum relative choke opening should be 0.2

Update the bringing together jupyter notebook after implementation of issue #43.

• Warning on high velocities (> 70 m/s)
• Droplet erosion model (base-asdn g73 10)

K, n, rho_p, rho_t etc

Wait for issue #4 has been completed

Do not get travis checks on pull requests.

Add functionality for crushed proppants in erosion functions, e.g

crushed=false as input. if crushed=True: return 3xE

Dersom en av input-verdiene til validate_asd er negativ vil asd_step_emerson og asd_step_clampon returnere "None". I logger.warning-teksten til validate_asd står det at nan vil bli returnert. Foreslår at asd_step_emerson og asd_step_clampon endres til å returnere np.nan is stedet for None dersom validate_asd returnerer True

minimum choke opening based on actual sand production for plug and cage type chokes, ref DNVGL RP-O501

rho_p and material functions missing, need to check RP:

• straight pipe?

rho_p not part of model, need to check RP:

• welded joint, reducer, probes

rho_p missing in models using bend model, should be added to:

• choke gallery, manifold, flexible

Ductile should be default

.. to transport.py

After implementing pysand in IOC erosion calculations we have encountered several issues with NaN values. IOC accept NaNs as a valid result, however pysand throws an exception if a NaN value is passed as input. The result is that the IOC calculations fail and cannot progress past the timestamp containing a NaN and data starts to lag.

We have made some modifications to the pythonscripts in IOC to replace NaN values before they are passed to pysand, but maybe pysand should accept NaNs as this is a valid datapoint in python?

Fix readme

Implement type hinting:
https://peps.python.org/pep-0484/

Inputs like v_m, rho_m, mu_m, Q_s..
Maybe the simplest thing would be to say
if E < 0: return np.nan for all erosion models? This way, any negative input will throw np.nan...

• v_m should be above 5 m/s as per DNVGL RP-O501
• validate E_meas. Should be positive and float or pd.series etc...

ER probe included from version 1.2 (er-probe branch)

Jupyter notebook showing example use of choke opening calculations

Currently some inputs validation is done assuming integer or float types only. Should be updated to handle series input too.

Added parameters to erosion input validation from choke mode. needs to be tested.

Include as function within erosion.py?

• include ps.__version__ etc to indicate what version was used in the examples
• include example with using dataframes (time series for erosion calcs). May be imported as a csv.
• erosion from mm/year to mm/ton

As of now, requirements are stated in requirements.txt and within setup.py < install_requires.
Requirements should only be needed one (1) place.

Example:
v_m = np.arange(5,50,10) # Mix velocity [m/s]
D = 0.2

for hastighet in v_m:
st_pipe = pysand.erosion.straight_pipe(hastighet, D)

raises exception that 'v_m is not a number' (v_m = hastighet)

np.vectorize(std_step_emerson)(df['v_m'], df['GOR']) raises GOR is not a number exception when df['v_m'] is float and df['GOR'] is int data types. May be related to #63 ?

• D boundaries in RP-O501 (0.01 - 1 m) should be stated in the warning
• d_p boundaries in RP-O501 (0.02 - 5 mm) should be stated in the warning
• When D and d_p are negative, they should not give result
• alpha in probes should be limited to 10-90 deg (page 34 in RP-O501)

Following parameters gave no error:

• v_m to be positive (limitations exist in RP-O501: 0-200)
• rho_m to be positive (limitations exist in RP-O501: 1-1500)
• mu_m to be positive (limitations exist in RP-O501: 1e-6-1e-2)
• Q_s to be positive (limitations exist in RP-O501: 0-500 ppmV)
• R to be positive (limitations exist in RP-O501: 0.5-50)

All above checked for bend model in version 1.2

The check valve model has been added in v 1.6.3. The Jupyter Notebook on Erosion examples should be updated to reflect added model.

Convert to poetry package management

in setup.py file