[Julian121]

I run a transient simulation with air ideal gas as material but forgot to set reference temperature and pressure. Now when I examine entropy at any location most entropy contours have negative values.
Since default ideal gas material assumes reference specific entropy as zero, does this happen because of wrong reference temperature and pressure?
As transient simulations are time consuming, I decided to define a new variable based on two new reference temperature and pressure as:
S - S(ref)=c(p)*log (T/T(ref)) – R*log (P / P(ref))
Expression: Specific Heat Capacity at Constant Pressure*log(Temperature/289 [K])-R*log(Pressure/101325 [Pa])
Then, I defined the reference temperature as the mean value between inlet and outlet of a compressor and the inlet pressure as the reference pressure.
Now, when I use the variable to draw entropy, I get error vertex values for variable do not exist on geometry /turbo surface.
Does anyone know how to fix it?

[Opaque]

Unless you want to compare with previous results using different reference values, the absolute value of entropy is irrelevant in nearly all cases.

The provided value is relative to S_ref, so it can be anything and still be correct. Say S_ref = -10^15.. It may be a nuisance, but it is still correct.

On the message, I would create the initialize the turbo features after I had created the new variable to see if the turbo post-processor missed those variables.

Can you create a contour on an inlet, wall, or another non-turbo locator?

[Julian121]

Originally Posted by Opaque

Unless you want to compare with previous results using different reference values, the absolute value of entropy is irrelevant in nearly all cases.

The provided value is relative to S_ref, so it can be anything and still be correct. Say S_ref = -10^15.. It may be a nuisance, but it is still correct.

On the message, I would create the initialize the turbo features after I had created the new variable to see if the turbo post-processor missed those variables.

Can you create a contour on an inlet, wall, or another non-turbo locator?

It did not show on a non-turbo location either but I am comparing the results between different cases relative to the same S_ref=0, so I think it does not matter anymore.

[Carnot]

In the expression, R has differenct unit with Cp.
From CFD-POST Constants, the unit of R is [ J K^-1 mol^-1 ]
But Cp has [m^2 s^-2 K^-1]
I think it is needed to make the units same.
R should be divided by molecular weight of the gas.

[Opaque]

Originally Posted by Carnot

In the expression, R has differenct unit with Cp.
From CFD-POST Constants, the unit of R is [ J K^-1 mol^-1 ]
But Cp has [m^2 s^-2 K^-1]
I think it is needed to make the units same.
R should be divided by molecular weight of the gas.

Though the equation may look incorrect (missing the MolarMass), it is nearly impossible for the wrong equation to be evaluated in CFD-Post since it checks for dimensional compatibility between terms. It would have triggered an error indicating the mismatch between the units. Sure, if forced by faking the units with a dummy value then it becomes a user beware.

Over the years I have learned that R in the thermodynamics books means "any consistent R" since the equation is valid per unit mass or per unit mol; therefore, it is only when evaluating the expression that the units must be consistent.