Learning Forum

[jose monteiro]

mode linfo : K-e standart, scalable wall functions, transient , pressure based, simple, inlet tube adiabatic and walls are couple (aluminium and fiber walls), mass flow rate of 0.008 kg/s, that gives me a velocity of 750 and 0.5 approximately, (important: im getting the conclusions below because when i do contours of total temp in the inlet the values that i input are equal and if i do static the values decrease as mentionated but in the surface of the tube are correct since in that zone is no slip condition and adiabatic so the fluid rests adiabtic and in that case total and static temperature are the same, it is why i think is the problem above).

hello, im simulating the filling of an hydrogen tank, and when i check the static temperature it comes down to values below what i input(to 275 from 293 Temp constant that i input for example) and then whith some time like 40 seconds of real time simulated it comes close to what i input. My velocity decreases with the time as it should in a real experiment and i check my mach number and was above 0.5 in the first second of real time simulated. My questions are:

1- what is the number that fluent uses as limit to compressible flow, i have seen some sources saying it considers compressible above 0.3 of mach number, so between 0.3 and 1 is compressible but subsonic that is my case, so i expect differences exist between total and static temperature and on a incompressibles dont. 

2- if what i said above is true, how can we input static temperature inlet (on pressure or mass flow inlets) or its impossible. Because i want to use a temperature profile inUDF that i found on a article and i suspect it is static temperature because most of sensors measure static temperature. 

3- if my input was a pressure profile and subsonic how can i do to an UDF with absolute pressure since is a static pressure and fluent only uses stagnation for pressure subsonic inlets.

 

[Rob]

The 0.3M guideline is for high speed flows through things. If you are modelling something where the gas density varies enough to need to model due to temperature or pressure then ideal gas is a good start. Real gas is also an option but I'd approach with caution as it's less stable. 

The inlet temperature is the value on the boundary, what else happens to the flow in that region? 

[jose monteiro]

As i said im simulating a transient tank being filled with hydrogen and im using mass flow inlet so my temp inlet is total temperature. I read somewhere that for incompressible or slow moving gás total = static temperature. When i do contourns and a report in the inlet i see that right after simulation starts my static temperature throughout my inlet gás tube reduces below what i put for total temperature and during the filling with the decrease in Mach (0.6 beginig to 0.1 in the end) my static temperature slowly converges to the value i put in total temperature in the inlet. I also test another two mass flow rates at inlet and withe the decrease in mass flow rate, so small mach numbers at beginning, the static temperature that falls is not so low and it converges faster to the inlet total temp.

1- my question is if there is a way of put static values on the inlet instead of total for temperature since i believe most temeprature sensors are static temperature and i want to use a UDF for inlet temp.

2- Do ansys do anything diferent when solving the equations or in the values i input when the mach goes below 0.3 during my simulation.

3- what happen if i choose a pressure inlet and have the profile for my UDF of a absolute pressure. Is it wrong to input that?

[Rob]

You can set the temperature/pressure based on what the panels allow you to do: check the options. 

Remember, if the pressure at the inlet is higher than in the tank the gas will expand on entering, that may explain the cold spot? 

Fluent will solve the same equations regardless of the speed; it's whether you pick suitable settings to cover the whole domain that will effect the result.