search

ibpsa / modelica-ibpsa

Modelica library for building and district energy systems developed within IBPSA Project 1
https://ibpsa.github.io/project1
145 stars 84 forks source link

Extension to FlowMachineInterface to allow polynomial performance curves #1361

Open PhOstmann opened 4 years ago

PhOstmann commented 4 years ago

To better match realistic fan performance maps, the FlowMachineInterface model could be extended in a way that allows the specification of polynomial performance curves. In order to keep the original model untouched, a PartialFlowMachineInterface should define the In- and Outputs. From this model the FlowMachineInterface can then be extended. As another extension of the new PartialFlowMachineInterface a new interface will be developed and tested in AixLib issue #909. The planned structure should maintain the functionality of the original FlowMachineInterface while giving the opportunity to implement the new calculation of the fan performance.

mwetter commented 4 years ago

@PhOstmann : Is the issue that you like to have eta=f(V_flow, y) rather than only eta=f(V_flow) where eta is the hydraulicEfficiency and the motorEfficiency?

Before switching to polynomials, did you check whether cubic splines (as are used now) wouldn't suffice. See for example

simulateModel("IBPSA.Fluid.Movers.BaseClasses.Validation.FlowMachineInterface(eff(nOri=3, per(pressure(V_flow={0,0.5, 1}, dp={1000, 750, 0}))))", method="dassl", tolerance=1e-6, stopTime=1.0, resultFile="FlowMachineInterface");

that shows good interpolation if you only provide few data points. This seems to be sufficient given the large sensitivity of the fan/pump performance with respect to conditions at the inlet and outlet (turbulence, asymmetric flow profile etc). Using data table would also ensure that all data can be provided in the same format.

PhOstmann commented 4 years ago

@mwetter That's indeed the underlying issue. In my test scenarios eta (especially the hydraulicEfficiency) would always stay constant while the speed changes according to the boundary conditions. While that can be a good estimate near the nominal point we want to investigate the behaviour under strong part load conditions, where the efficiency should not be constant. So it's rather an issue of calculating the efficiency than the pressure rise. I guess the issue title is misleading.