Fluid_Aspect_Course_2_3_6
In the real world, a column of fluid under acceleration experiences a pressure rise from “top” to “bottom”. This is why we feel 14.7 psi of air pressure at sea level, because of the column of air above us in the 1-g acceleration of earth’s gravity. The amount of pressure rise is given by this equation:
dp = ρgh,
where dp is the delta-pressure (Pa) between top and bottom points (p bottom – p top), rho is the fluid density (kg/m3), g is the acceleration magnitude (m/s2), and h is the distance (m) between top and bottom points. Note that this formula gives pressure in (Pa) but GUNNS typically uses (kPa) so make sure to convert the units when appropriate.
In systems that are always in micro-gravity, like on ISS, this affect never occurs so it can be ignored. However in terrestrial liquid systems or liquid systems on an accelerating vehicle (like a launch vehicle), this is a very important effect. It mainly shows up in liquid containers at different heights relative to the acceleration direction. The height term in this effect is commonly called “pressure head”.
Here’s an example picture of such a “gravity-fed” terrestrial system from Wikipedia. The height of the tank above the reservoir causes flow between them:
There are two ways to model this, depending on whether your acceleration field is constant (such as a fixed structure on a planetary surface) or varies (such as on a moving vehicle):
This is the simpler of the two methods to implement. This would be appropriate for any fixed liquid plumbing system on a planetary surface, where the density of the liquid doesn’t change significantly (constant mixtures or pure fluids, etc.) Here, we simply build in the constant dp = ρgh effect into the liquid tank (GunnsFluidAccum) and pipes (GunnsFluidPotential) links, based on the constant height of each component relative to the acceleration (gravity).
For GunnsFluidPotential and similar links, calculate the constant dp = ρgh value by hand and add this value to its sourcePressure input data. The link creates a positive dp from its port 0 to 1. So for a positive dp, you’d have Port 1 be at the bottom end, relative to the gravity.
For GunnsFluidAccum and similar links, calculate the constant dP = ρgh value by hand, where h in this case is the height from the accumulator fill/drain port at the bottom, to the very top of the liquid level when the accumulator is at maximum fill. Add this dp value to its springCoeff1 config data.
This picture shows the GunnsDraw network drawing that represents the example picture above:
If the plumbing changes orientation relative to the acceleration field, or if the acceleration itself is changing, then this method is more accurate. This is basically any moving vehicle. You’d also use this if density changes a lot. Normally this won’t happen since this effect is only significant for liquid, and liquids tend to be incompressible (constant density) and we don’t model mixtures of liquids in GUNNS.
Here instead of putting constant dp values in the link’s config & input data, you use a GunnsFluidPressureHead spotter to dynamically drive the dp for each link. Use a separate spotter for each link that needs to have the effect. The acceleration vector probably resides in a separate dynamics model, outside of the fluid network. You need to supply this acceleration vector to each spotter. If the acceleration vector resides outside the fluid network, in some dynamics model, you can’t set this pointer in the drawing; rather you need to set the spotter’s config data in the input file. For example in a Trick sim where the network has variable system.subsystem.network, containing a spotter named link2accel, this would look like:
system.subsystem.network.netConfig.link2accel.mAcceleration = <acceleration vector's variable>Here is the same GunnsDraw network as above, but now we have the spotters in place:
