How do we KNOW that the return to the sump is what caused the prob? Are

we talking his redesign or the original install?

 

We do NOT know that the return to the sump caused the engine out on Paul’s last flight.

 

My original plan was to return fuel to the sump, but after Paul's first

engine out to  also have the sump vent be capable of overflowing into

one of the wing tanks (both of which gravity feed into the sump).

 

There are sump tanks and there are header tanks.  I would define a sump tank as one that is at or near the lowest point in the fuel system, and is gravity fed from the main(s).  Any small tank in a system that has to be fed by a pump, I’d call a header tank.  Header tanks are a different issue.

 

Here is a basic summary of my views on sump tanks.  I state them as facts, but obviously it is my interpretation of the facts.

 

A sump tank has the advantage of eliminating the need for a pump from the main tank; and eliminating the need for valves from one tank or the other.  Assuming it feeds from the lowest point in mains, it also gives you a single point for collection of water and impurities that may get in the fuel, and a single sample drain.  These are desirable and important reliability and safety features.

 

The Velocity has a true sump tank which is full (or above) any time there is fuel in the main tank.  It is also a reasonably large tank; about 2 ½ gal in the RG and I guess 4 – 5 gallons on the FG.  A fuel pump at or near the exit from the sump tank; at a level at or below the level of the bottom of the tank, with a reasonably sized line (3/8” or larger) from sump to pump should never see less than at least 1 atmosphere of feed pressure, and therefore be highly resistant to any vapor lock issues.  

 

With the fuel exit from the sump a bit up from the bottom (room for collection of a quart or more of water and impurities) there should be no need for a filter between the sump and the pump that could result in flow resistance.  Put the filter (30 micron or less, I suppose) downstream from the pump to catch anything you don’t want going to the injectors.

 

I don’t believe there is an issue with returning the bypass fuel from the pressure regulator to the sump as long as the sump tank is a few gallons or more.  When the engine is running there is always a net flow from the main to the sump.  There is not an issue of the sump “overflowing”. 

 

If you have a smaller sump, or if it is reasonable and simple (reliable) to do so; return the bypass to the main tanks (tanks, plural; for gravity feed returning to one tank could result in level imbalance.)

 

If you are pumping fuel at a rate that far exceeds that burned by the engine, there is a potential for heating of the fuel in the sump from heat picked up in the fuel lines and rail in the engine compartment.  Size the pumps so that one pump gives you slightly more than the maximum burn you expect from the engine.  More than that; or running both pumps more than just during takeoff, is just circulating a lot of fuel through the engine compartment to no advantage, and causing heating of the fuel.  Fire sleeve on the fuel lines is a good idea, and provides some (probably not much, but some) degree of heat insulation.

 

Monitoring the fuel temp at the bottom of the sump could be a good idea and provide useful information about the amount of heating of the fuel in the sump. I’d expect the worst case to be long periods on the taxiway with a fully heated engine at low power. 

Do a power run-up prior to takeoff. (This assumes your cooling system design allows this.)

 

The reason for circulating fuel through the fuel rail rather than a dead-end system is to keep the fuel temperature from rising in the rail, and to force any air or vapor bubbles (mostly a possibility at or prior to startup) back to the tank.  I have found that some period after shutdown of the hot engine, I could hear bubbles feeding back to the sump for a second when turning the fuel pump back on.

 

In its simplest form (as in the Velocity) both mains and the sump are vented to the atmosphere for air to exit upon fill-up, and for air to enter as the fuel level goes down in the mains.  The vents must of course be above the highest fuel level and the lines not subject to filling with fuel that must be drawn upward for air to enter.  Keep in mind that the fuel level in the sump will be the same as that of the main tank; meaning that the vent line from the sump will have fuel in it up to the level of the of the fuel surface in the mains.

 

I realize that there have been or may be other anomalies in the vent system (leaky fuel caps, or whatever) that can affect the function of the vent system (as in Jim’s case); but that’s a separate discussion.

 

FWIW,

 

Al