217 shows that even for an official example we could not manage to create appropriate curves for compressor/fan/turbine using ThermofluidStream.Processes.Internal.TurboComponent.dp_tau_const_isentrop.
From a user point of view the parametrization of a compressor/ fan/turbine is rather silly. For skew = 0, w > 0, m > 0, pr >= 1 the used caracteristic curve pr(m_flow) simplifies to pr = 1 + (w/w_ref)^2 - (m/m_ref)^2, which means:
A compressor pr >> 1 can be parametrized with w_ref << w, i.e. pick a low reference angular velocity to yield the correct pressure ratio for the real angular velocity
A fan pr >= 1 can be parametrized with w_ref <= w, i.e. pick a slightly higher reference angular velocity to yield the correct pressure ratio for the real angular velocity
In a next step one can adapt m_ref to yield correct curves: E.g. to parametrize a quadratic compressor with maximum pressure ratio of pr(w = 1000 rad/s, m = 0kg/s) = 5, and maximum mass flow rate m(w = 1000 rad/s, pr = 1) = 1 kg/s, one can calculate the reference angular velocity and reference mass flow rate using: 5 = 1 + (1000/w_ref)^2 - (0/m_ref)^2 -> w_ref = 1000/2 rad/s and 1 = 1 + (1000/500)^2 - (1/m_ref)^2 -> m_ref = 2 kg/s. However i dont think its really inuitive to choose w_ref = 500 rad/s, m_ref = 2 kg/s to yield the desired compressor.
In my opinion the characteristic curve for a compressor/fan can be parametrized way easier using e.g. an approach similar to #183 i.e. dp/dp_ref = (w/w_ref)^2 - (m/m_ref)^2
For turbines pr < 1 the characteristic curve is pr = 2^((w/w_r)^2 - (m/mr)^2), which makes it even harder to parametrize.
In my opinion another issue is, that for pr <= 1 a fan suddenly is treated like a turbine: The curves are continues but not differentiable and the behavior is not realistic at all.
From a thermodynamic perspective i would apprechiate to use the entropy function instead of assuming ideal gas with constant specific heat capacity. For a pressure ratio close to 1 const. cp might be sufficient but often its not in my opinion.
217 shows that even for an official example we could not manage to create appropriate curves for compressor/fan/turbine using
ThermofluidStream.Processes.Internal.TurboComponent.dp_tau_const_isentrop.From a user point of view the parametrization of a compressor/ fan/turbine is rather silly. For
skew = 0, w > 0, m > 0, pr >= 1the used caracteristic curvepr(m_flow) simplifies topr = 1 + (w/w_ref)^2 - (m/m_ref)^2, which means:pr >> 1can be parametrized withw_ref << w, i.e. pick a low reference angular velocity to yield the correct pressure ratio for the real angular velocitypr >= 1can be parametrized withw_ref <= w, i.e. pick a slightly higher reference angular velocity to yield the correct pressure ratio for the real angular velocitym_refto yield correct curves: E.g. to parametrize a quadratic compressor with maximum pressure ratio ofpr(w = 1000 rad/s, m = 0kg/s) = 5, and maximum mass flow ratem(w = 1000 rad/s, pr = 1) = 1 kg/s, one can calculate the reference angular velocity and reference mass flow rate using:5 = 1 + (1000/w_ref)^2 - (0/m_ref)^2 -> w_ref = 1000/2 rad/sand1 = 1 + (1000/500)^2 - (1/m_ref)^2 -> m_ref = 2 kg/s. However i dont think its really inuitive to choosew_ref = 500 rad/s, m_ref = 2 kg/sto yield the desired compressor.dp/dp_ref = (w/w_ref)^2 - (m/m_ref)^2pr < 1the characteristic curve ispr = 2^((w/w_r)^2 - (m/mr)^2), which makes it even harder to parametrize.pr <= 1a fan suddenly is treated like a turbine: The curves are continues but not differentiable and the behavior is not realistic at all.