I would use a logarithmic mixing rule, although that might not work 100% quite often it give good estimates.

eta_alloy=exp(conc(A)*ln(eta_A)+ conc(B)*ln(eta_B))

of course, the viscosity is temperature dependent. You might not find the viscosity of U and Zr in literature, so I would suggest estimating it to be in the order of several mPas, which is typical for many metals.

Hi Florian ,

Thank you for your reply, actually I want to find different thermophysical properties of molten metal fuel alloy such as density, surface tension, viscosity, etc. As you said, a logarithmic mixing rule is a very good choice for viscosity. In fact, I find some data for U-5Zr and U-15Pu-10Zr as shown in the figure. Once I can get individual properties of each component, I will solve this problem.

Your text book gives the kinetic viscosity, which actually is called kinematic viscosity.

from "https://en.wikipedia.org/wiki/Viscosity#Dynamic_and_kinematic_viscosity"

Dynamic and kinematic viscosity

In fluid dynamics, it is common to work in terms of the kinematic viscosity (also called "momentum diffusivity"), defined as the ratio of the viscosity μ to the density of the fluid ρ. It is usually denoted by the Greek letter nu (ν) and has units ( l e n g t h ) 2 / t i m e {\displaystyle \mathrm {(length)^{2}/time} } 📷:

ν = μ ρ {\displaystyle \nu ={\frac {\mu }{\rho }}} 📷.

Consistent with this nomenclature, the viscosity μ {\displaystyle \mu } 📷 is frequently called the dynamic viscosity or absolute viscosity, and has units force × time/area.

so, calculating the viscosity is probably not a big deal

Thanks for your kind reminder, yes, it is usually called kinemaitc viscosity.