Hydrogen Properties Package

TEMPERATURE RANGE: 13.8K TO 12,000K
PRESSURE RANGE: 0.1 bar TO 10,000 bar

A Thermodynamic and Transport Property Package for Para- and Dissociated Hydrogen

TEMPERATURE RANGE: 13.8K TO 12,000K
PRESSURE RANGE: 0.1 bar TO 10,000 bar

This package was developed by Gary Chen and Samim Anghaie based on NASA/NIST databases and calculations in temperatures ranging from 13.8 to 3000 K. Higher temperature and pressure data is generated by extrapolation of the data base.

NASA and the National Institute of Standard and Technology (NIST) compiled hydrogen properties were used to develop a computer routine that can conveniently interpolate thermodynamic and transport properties of hydrogen in temperature and pressure ranges of 13.8 to 12,000 K and 0.1 to 10,000 bar, respectively. Linear interpolation (LI), natural cubic spline interpolation (NCSI), and multiple polynomial curve fitting with polynomial bridging between the curve fits were considered for development of the hydrogen property package. Since LI is the standard method for use in property acquisition programs, it was incorporated into this package. NCSI was used with the objectives of increasing accuracy and of representing the hydrogen properties as continuously differentiable functions. Analysis of polynomial curve fitting was performed using polynomial bridging between curve fits to maintain the desirable NCSI characteristics of increased accuracy and continuous differentiability and to maintain the computation (CPU) time comparable to LI.

Motivation for This Work

INSPI in conjunction with NASA is developing nuclear thermal propulsion (NTP) system analysis codes.  Hydrogen is considered for both the coolant and the propellant in NTP systems.

Hydrogen is a mixture of parahydrogen and orthohydrogen which differ by the direction of the nuclear spin, assuming that nuclear spin exists.  The composition of the mixture varies from 100% parahydrogen at near liquid temperatures to 25% parahydrogen and 75% orthohydrogen at near room temperature and above.

NASA has compiled a para- and dissociated hydrogen database from NASA and NIST databases and extrapolation modules.  The NTP system analysis codes require hydrogen properties as continuous functions of pressure and temperature (or enthalpy).  The objective is to develop a hydrogen properties acquisition code using the discrete data from NASA to approximate the properties as continuous functions of pressure and temperature (or enthalpy).  Different methods of curve fitting were implemented and evaluated for this code package.

Evaluation of Curve Fitting Methods

Accuracy comparison between LI and NCSI was performed by selectively eliminating block data points at a given pressure and temperature interval which contained linear and non-linear regions. When the results were compared to the removed data, NCSI had better accuracy in the non-linear regions. This difference in accuracy will be less when the data points that were removed are returned.

The CPU time required for NCSI depended on the calling code. An NCSI code tailored for CFD codes and a LI code essentially had comparable CPU times; thus, NCSI will give better accuracy and no additional CPU time cost. For a general NCSI code, LI used six percent less CPU time. A recommendation on which acquisition code to use depends on the preferred trade off between accuracy and CPU time.

Comparison of NCSI to curve fitting indicated that the respective results were essentially the same. The CPU time required was on the order of magnitude of LI's CPU time. The characteristics of the NCSI are maintained using polynomial curve fitting with polynomial bridging while avoiding the penalty of increased CPU time for a general NCSI code. For a CFD tailored NCSI code, curve fitting with bridging would not increase efficiency.

Acknowledgement

This work was sponsored by the National Aeronautics and Space Administration Lewis Research Center, Office of Nuclear Propulsion, under NASA Grant NAG3-1293.