A 'simple' and fast 1 dimensional computational fluid dynamics simulator made for simulating Ejectors for heat pump and refrigerator applications. It includes modules for simulating two phase supersonic nozzles, that are crucial parts of supersonic ejectors.
Using this package you can easily calculate mass flow-rates of motive and suction nozzles of ejectors with the user defined ejector geometry and fluid specification. The provided functions solve simplified differential-algebraic equations of the fluid flow (mass, momentum and energy equations). This package calculates and plots speed, pressure, quality, etc. profiles along the ejector axis. And lets you easily calculate ejector efficiency.
You can play with the primary nozzle and ejector geometry, and update the calculation results with any fluid from the Refprop (https://www.nist.gov/srd/refprop) library or the coolprop library. This makes it an ideal tool for a rough and fast ejector dimensioning. You can calculate and evaluate with ease dozens of ejector geometries within minutes. And you know it, that it replaces hours and days of heavy calculations on commercial software packages!
This module can be used for 1 component, one or two phase flow simulations. In the first version it applies the homogeneus equilibrium model (HEM) to calculate material properties. (For the thermodynamic properties it needs the RefProp 9 or newer to be installed - not tested for older versions)
For a short description of the method see the pdf in the Docs directory:
/Docs/
The english description of the method will be published soon, and you must reference to it if you use this package in any publication!
To see, how it works, look into the jupyter notebooks located in the directory: src/simpy_ejector/useCases/
Don't forget, that simpy_ejector requires either coolprop or refprop materials library.
If you use refprop, which is faster but needs a commercial refprop license, then you have to install the python-Refprop package ctREFPROP: https://pypi.org/project/ctREFPROP/. In windows you need to have the refprop.dll installed (for the development I have used version 9.1) and set the environment variable 'RPprefix' to the path of the 'refprop.dll' file. For example if your REFPROP is installed into "C:/Program Files (x86)/REFPROP/", then open a command window, and type:
setx RPprefix "C:/Program Files (x86)/REFPROP/"
(or set the RPprefix user-environment variable through windows menu)
In case you use the free coolprop library (see https://coolprop.org/ it is really cool), then you install it as
pip install coolprop
For the latest Documentation, Demo, Tutorials and Examples visit the project page on github: https://github.com/AdamBuruzs/simpy_ejector
Only tested on Windows. ( Linux installation manual is still to be done. )
pip install simpy-ejector
** If you use Refprop: **
And set the environment variable 'RPprefix' to the path of the 'refprop.dll' file. For example : "C:/Program Files (x86)/REFPROP/"
Test if refprop is setup and working correctly, start a python shell, and write:
>>> from simpy_ejector import refProp
>>> RP = refProp.setup("water")
>>> out = refProp.getDh_from_TP( RP, T=350, p = 100)
>>> print(f'water at 300K and 100 kPa : Density = {out[0]} kg/m3, spec enthalpy = {out[1]} kJ/kg')
You should get a result of 973.72 kg/m3 and 321 kJ/kg. If you get an error, then check your refProp installation, and 'RPprefix' environment variable settings.
** If you use coolprop **
>>> from simpy_ejector import materialFactory
>>> RProps = materialFactory.MaterialPropertiesFactory.create(material="water", library="coolprop")
>>> RProps.getDh_from_TP(T = 350, p=100)
You should get [973.727, 321.838]
After installation you can import the package. In the simpy_ejector.useCases.ejectorSimulator module you find a class that helps you very quickly setting up and performing a simulation of an ejector. So simply write into a python file:
import logging
import sys
logging.basicConfig(stream = sys.stdout, level = logging.INFO)
import matplotlib.pyplot as plt
import pandas as pd
from simpy_ejector.useCases import ejectorSimulator
from simpy_ejector import materialFactory
Then set the fluid (refrigerant), that you want to use in your ejector. For example:
fluid = "Water"
proplibrary = "coolprop"
RProps = materialFactory.MaterialPropertiesFactory.create(material=fluid, library=proplibrary)
Set the inlet states for the motive and suction nozzle:
## Example test case for a water ejector
Pprim = 1000 # kPa
Psuc = 300 # suction pressure kPa
hinPrim = 750 # kJ/kg
hinSuc = 2800 # kJ/kg
Then specify the ejector geometry:
# set up geometry parameters, and motive and suction nozzle states:
params = { "Rin": 1.5, "Rt": 0.29, "Rout": 0.87, "gamma_conv": 15.0, "gamma_div" : 6.0, "Dmix": 2.67,
"A_suction_inlet" : 16 ,
"mixerLen": 12 , "gamma_diffusor": 2.5, "diffuserLen": 10}
params.update({"Pprim": 2007, "hprim" : 365.5, "hsuc": 437.1, "Psuc" : 276.3 })
Use this picture to find the meaning of the parameters:
The parameter units:
| parameter | description | unit |
|---|---|---|
| Rin | primary nozzle inlet radius in | cm |
| Rt | primary nozzle throat radius in | cm |
| Rout | primary nozzle outlet/exit radius in | cm |
| gamma_conv | primary nozzle convergent part angle | degree |
| gamma_div | primary nozzle divergent part angle | degree |
| Dmix | mixer diameter (2*radius) in | cm |
| Pprim | primary nozzle inlet pressure in | kPa - kiloPascal |
| hprim | primary nozzle inlet specific enthalpy | kJ/kg |
| hsuc | secondary/suction nozzle inlet specific enthalpy | kJ/kg |
| Psuc | secondary nozzle inlet pressure in | kPa - kiloPascal |
| A_suction_inlet | primary nozzle inlet cross section area | cm2 |
| mixerLen | length of the mixer | cm |
| gamma_diffusor | angle of the diffuser profile in | degree |
| diffuserLen | length of the diffuser | cm |
set parameters of the mixing calculations:
params["mixingParams"] = {'massExchangeFactor': 2.e-4, 'dragFactor': 0.01, 'frictionInterface': 0.0,
'frictionWall': 0.0015}
create a simulator object:
esim = ejectorSimulator.ejectorSimu(params, fluid=fluid, proplibrary = proplibrary)
## a small numerical kick is needed by the motive nozzle to transition into the supersonic flow in the divergent part:
esim.dv_kick = 0.004 # in m/s
plot the ejector geometry:
ejplot = esim.ejector.draw()
This line will create for you an interactive matplotlib plot of your ejector
## calculate the primary mass flow rate:
esim.calcPrimMassFlow(plotCrit0 = True, chokePos="divergent_part")
This will iteratively calculate the critical inlet velocity, where the motive nozzle chokes. The critical speed and mass flow rate will be printed on the standard output
calculate the critical (= choked flow) solution in the motive nozzle:
## calculate the primary mass flow rate:
esim.calcPrimMassFlow(plotCrit0 = True, chokePos="divergent_part")
vcrit_tot = esim.params["vin_crit"]
print(f"critical speed motive nozzle {vcrit_tot}")
then the full solution in the motive nozzle
res_crit = esim.motiveSolver()
print(f"By the motive nozzle exit:\n {res_crit.iloc[-1]}")
You will see the chart with the motive nozzle flow, the velocity, quality and pressure plotted in the motive nozzle
solve the pre-mixing equations:
esim.premix(res_crit) # this sets the mixer, that is needed for the mixing calculation
print(f"suction MFR [g/s] {round(esim.mixerin['massFlowSecond'],3)}")
solve the mixer equations until the ejector outlet
esim.mixersolve()
Let it calculate the mass-flow rates and checks:
esim.massFlowCheck()
outletValues = esim.solMix.iloc[-1].to_dict() # .transpose()
average_h = (outletValues["MFRp"] * outletValues["hp"] + outletValues["MFRs"] * outletValues["hs"])/ (outletValues["MFRp"] + outletValues["MFRs"] )
print(f"Ejector outlet specific enthalpy {round(average_h,3)} kJ/kg, pressure {round(outletValues['p']*1e-2,3)} bar,"
f" vapor quality {round(outletValues['q_average'],3)} ")
esim.plotMixSolution(res_crit, esim.solMix, "simpy_ejector 1D Ejector flow solution")
You get an interactive matplotlib chart, so you can zoom in to see the values.
The pressure and specific enthalpy is printed at the ejector exit.
That's it, you have performed your first ejector simulation with simpy_ejector !
simulation result for an ejector (with a simulated normal shock wave)
Pressure profiles in a nozzle with a simulated shock-wave.
Pressure profile comparison with experiment (in this case without shock-waves):
The research leading to this repository was financed by the Austrian Research Promotion Agency (ffg) over the projects VWE, and ETHP