Incorrect implementation of internal energy in WaterSSTP (liquid-water-IAPWS95)

[speth]

Problem description

Calculations of the internal energy in the WaterSSTP class (YAML model name liquid-water-IAPWS95) give thermodynamically inconsistent results, compared to using the identity u = h - P*v and du/dT = c_v.

Steps to reproduce

>>> s = ct.PureFluid('liquidvapor.yaml', 'liquid-water-IAPWS95')
>>> s.TP = 360, 101325

>>> s.enthalpy_mole
-281174223.8705604

>>> s.int_energy_mole + s.P * s.volume_mole
-101223397991.49594

>>> cv1 = s.cv_mole
>>> u1 = s.int_energy_mole
>>> dT = 0.01
>>> T = 360
>>> s.TD = T + dT, None
>>> cv2 = s.cv_mole
>>> u2 = s.int_energy_mole
>>> 0.5 * (cv1 + cv2)
69121.5991841279

>>> (u2 - u1)/dT
-256291643.85375977

System information

• Cantera version: 2.6.0 or main at 1ab81ac

Additional context

This was discovered as part of #1299, which implements Cantera/enhancements#114.