Tank "Slosh" factors

You should be able to do a simple F=ma calculation.

Assuming you know the mass of fuel being carried, derive a accel/decel rate and calculate the force required. That force divided by the area of your front tank wall should be the added pressure seen during braking. Knowing that, you should be able to design your tank structure accordingly.

Good luck.

The fuel tanks we use in stock car racing are filled with light weight open cell foam. It adds very little weight, has almost no effect on tank capacity, and almost totally eliminates slosh. I am not sure what the foam is made of, but there are several companies that manufacture and sell racing fuel cells. We use Fuelsafe cells.

I'm not sure if foam is an option, but it shouldn't be too hard to find.

Are you making a Auxiliary fuel tank for more storage?

or a surge tank to try and remove the air in the system if the tank is able to suck air for a short time?

Rapid Accel/decelok, how about a honeycomb inside the tank vertically mounted so the "straws" are up down orientation, that would minimise your slosh to being confined within each tube space.

Baffles are fine, but from my experiance in Off Road Racing, its the sideways that is more of a problem ie. in a turn, or successive turns around a track, where the baffles I use are longitudinal to the direction of the car in the normal forward driving.

Or the surge tank, where you just run a lifter pump from the main tank to the top 1/3 of the surge tank, and a return from the top back to the tank, and a drain from the bottom to your fuel system.

Some people with injection systems take the return line from the fuel rail back to the surge tank, others take that line back to the tank.

In the marine world we have standards that require baffles to be no more than 24 inches apart, or roughly 60 cm if you are metric. Limber holes and lighening holes not to exceed 30% of baffle area. Don't forget limber holes along top cornes to prevent bell jar bubble effect. Depending on application you may require inspection ports. In which case place one port per baffled compartment.

Last tank I was involved with was 15 feet long 5 feeet wide and 3 feet deep. 1000 gal capacity . But the same criteraia applies even for a 100 gallon tank. Suggest using plug weld technique to weld baffles in the middle. Easier than doing it from inside. You can use a shoe box top design. For marine work we need to test such tanks to 3 psi. What certification standard are you working to? ASTM?

All fuel tanks normally require certification before being used. That standards body should have the criteria published.

The race car fuel cell with internal honey comb is fine for small quantities. Becomes impractical for larger tanks. A baffled tank loses less volume than a honey comb from what I have heard.

Thanks a lot for the info guys...this has been very informative.

A few clarifications:

• It's an auxiliary tank (not a surge tank)
• For aircraft...fitted inside the cabin so it must meet some very strict rules
  
• It must be inspectable visually (that foam sounds interesting...especially if it could be removed for visual inspections.
• Cannot leak (double walled with indicators if fuel shows up between the walls.
  
• Weight is another priority (weight saved = more fuel = more range)

Several concepts are being considered but the "sloshing" tank is a low profile one that fits under the floor so the potential for fuel movement is much higher than in a vertically oriented tank. These tanks essentially gravity drain into the main tanks so some fuel movement is not a big deal since the fuel pumps do not feed directly from these tanks and the main tanks are always full as long as there is any fuel in these tanks.

To keep the design simple and inspection procedures intuitive we're hoping to design the tank (mostly panels and corner connections) in such a way that it will simply resist the internal pressures of whatever the fuel might do (think 9G-16G airplane crash).

Check in with an Air Force or Air National Guard base if there's one near you. Military aircraft use foam cells in their tanks to good effect. If you approach nicely, most of the 'old hand' Sergeants will share info readily. The problem may have already been solved.

We've brought up the foam before and it seems the concern is the difficulty it introduced to inspections. Military planes get a lot more regular maintenance than commercial/private craft so when a plane is on the ground for inspection (not making any money) the inspections have to be quick and efficient.

KKjensen, we've done some similar projects for Delphi, if you would like to conduct some simulation analysis on a aircraft, or a marine application the methodology would be similar, contact me at ctheisen@hoff.com if you would like to discuss further... see notes below, we used Fluent for the simulation analysis. thanks Chris

The scope of a previous project was to perform CFD simulations of flow and sloshing on a Delphi SGM18 power steering bottle. The following flow cases were analyzed.

1. Sloshing

a. Braking

b. Left turn

c. Right turn

d. Acceleration

2. Transient slug flow simulation

Transient Sloshing analysis

A transient sloshing analysis was performed on the power steering bottle. These simulations involved the application of a constant acceleration on the power steering bottle fluid depending on the vehicle event being simulated – braking, left turn, right turn and acceleration. During these simulations, the inlet and outlet ports will be capped and no flow will be assumed to go through them. The free surface of the fluid will be reported with the intent to keep the inlet port always wetted.

The acceleration that the bottle will be subjected to – 0.5G for acceleration and 1G for all other events. The fluid properties were defined.

Transient Slug flow analysis

Simulations were performed to evaluate the flow behavior of a large bubble/air-pocket entering the bottle from the power steering flow circuit. The goal of the simulation was to predict the motion and entrainment of the bubble in the flow and to see if the bubble reaches the free-surface or gets trapped in some chamber or by a baffle in the bottle. A steady-state flow simulation will be performed at the target flow rate. This was followed by a transient simulation and the bubble will be introduced as a vapor/air pocket via transient inlet condition. The volume of the bubble will be controlled to Delphi's specification.

This simulation was performed as a slug flow – only a single air bubble will be introduced and it will occupy the entire cross section of the inlet port.

The simulation will predict the general nature of the bubble motion. Detailed tracking of the bubble surface was not performed and hence exact simulations of the breakup and coalescence of the bubbles was not accomplished. Surface tension effects was incorporated in the simulation.

The CFD model will be created using tetrahedral elements with prism layers on the boundary for higher accuracy boundary layer calculations. Fluent will be used for the CFD analysis.

Data Required

The following information will be required:

1. CAD geometry
2. Design fluid surface level for the bottle
3. Size of air bubble for slug flow
4. Surface tension of power steering fluid at 170˚F

Thanks for all the info guys.

It looks like we're going with baffles. They lend themselves to stabilizing the side walls of the tank and help remove stiffeners on the outside of the tank. Two birds, one stone, the easy decision. It'll mean a lot more fastener holes to inspect and some cramped assembly but in the long run this seems to be best under the circumstances.

Happy posting!