Another fuel tank plumbing option

[nkemp]

Maybe your right.

I made a mistake once ... I thought I was wrong ...but I wasn't Just kidding!

If you take a look at the GM units they are rather interesting. On one I have, there are two tubes coming off the pump top, the larger feeds the engine. The smaller returns to the sump (But I'm not sure of the flow direction (to or from the pump)). That unit also has a float controlling another valve (rubber diaphragm). It is this unit that picks up the fuel from inside the sump. Trouble is that I don't know why or what those do or how the unit "works".

If you get a chance to look at one, do so. It is not like the good old days (1981) :-)

Nick
897

[DMCMW Dave]

It will never fill above the tank level (not counting sloshing of the gas). Physically impossible since less gas is returned than the gas removed.

I'd argue this as well - Assume the car is sitting still (to avoid the "slosh" factor). The baffle should match the tank level as the check valve will let in gas to keep it level. ANY gas returning from the engine will make the baffle more full as fuel cannot go backwards through the check valve.

I believe that this design makes it less important where the sump is physically in the tank. I was worried at first as I was never a fan of moving the baffle forward under the pump. The GM design should not have an issue as the sump is much less "leaky" and much deeper than the DMC or any of the tunacan designs.

[Bitsyncmaster]

Now you got me thinking. If the baffle level is just a fraction above the tank level then no fuel will flow into those one way valves. So I'm back to thinking it will not be higher than the tank level

[DMCMW Dave]

Now you got me thinking. If the baffle level is just a fraction above the tank level then no fuel will flow into those one way valves. So I'm back to thinking it will not be higher than the tank level

But the excess fuel from the engine will tend to "top off" the baffle.

[Bitsyncmaster]

But the excess fuel from the engine will tend to "top off" the baffle.

I still don't see it. Remember the pump is drawing out of the baffle and returning less than it draws. If the baffle level is ever above the tank level than no fuel enters the baffle from the tank. So my thinking is the baffle will not get higher than the tank.

[nkemp]

I'm sure flow rate through the pump and engine fuel consumption can be calculated and if not, measured... to determine how much fuel flows back to the tank. I suspect that even at high load the amount of fuel used is much less than pumped and as such the sump stays full. Until the sump is full, the pump (in addition to feeding the engine) is simply pumping from the tank into the sump. I also suspect that until the tank is empty (below the pickup) the sump is full to the overflow.

Maybe it is worth the thought process to create an equation...Infinite input to sump (in finite terms) -pumped out + returned = sump level. How's that for a crude equation:-)

Or think about what happens when the key is turned to run and the pump starts. It is building pressure and excess fuel is returned to the sump.

Nick
897

[Bitsyncmaster]

Maybe it is worth the thought process to create an equation...Infinite input to sump (in finite terms) -pumped out + returned = sump level. How's that for a crude equation:-)

Or think about what happens when the key is turned to run and the pump starts. It is building pressure and excess fuel is returned to the sump.

Nick
897

But there is your error. The input to the sump is not infinite. It would vary from large quantity with the sump low (below tank level) to zero at tank level.

[DMCMW Dave]

But there is your error. The input to the sump is not infinite. It would vary from large quantity with the sump low (below tank level) to zero at tank level.

The (gravity) pressure from the tank will keep the sump at a minimum at the tank level through the check valve assuming the check valve offers low input resistance. ANYthing coming back from the engine is excess and can't run out, so it's a net adder over the tank level. I'd also hazard a guess that at idle and even cruising speed (other than hard acceleration) the gas returning to the tank is 80% or better of what the pump is removing from the tank. This actually would be pretty easy to test with the injectors in jars and using small measured tank and return containers.

So - yeah, with a low tank under hard acceleration the sump may drop below full, but under somewhat static conditions it will always be close to full and should never be below the tank level (of course under acceration with a low tank the tank level will be very low at the sump - which is my original concern with using a more open-bottom flat baffle positioned towards the front of the car).

We'll need to build a clear tank for a demo....

[nkemp]

Let's look at this from a sequence of operations perspective:

To simplify, lets assume the tank is partially full and below the top of the sump and that the sump & tank are at the same level;

• we start the car and the pump starts.
• some fuel is removed from the sump
• the removed fuel is "immediately" replenished through the backflow valves so that the tank & sump stay at the same level
• excess flow occurs in the fuel lines and is returned to the sump
• this excess fuel adds to the fuel level raising it above the tank level
  
  • because the sump is the same level as the tank and it can't flow out via gravity

So the combination of inflow through the valves and the returned flow raises the sump level.

The use of "infinite" may have been the wrong word even though it was qualified with "finite".

[nkemp]

So - yeah, with a low tank under hard acceleration the sump may drop below full,

But the sump is "relatively" large and short of making hi G donuts for a while, it will take a bit to empty the sump.

We'll need to build a clear tank for a demo....

How about using some of that transparent aluminum they used in Star Trek?

Nick
897