Bosche type fuel injection return line 1

[thruthefence]

I'm converting a 'carb' car to EFI (Toyota) and need to add the fuel return line to the tank. Every one I have seen has entered from the top of the tank (along with the fuel pump line & fuel qty wiring). This may be an 'ease of manufacture' feature I suppose.

Any reason I cannot enter the tank below the fuel level?
I have a fitting that would be convenient to use, but it sits at about the 1/4 mark on the tank, which is tall & narrow, and about 10 gallons capacity.

 

[MiketheEngineer]

Obviously the return line "returns" fuel to the tank. Lets say you have a full fuel tank (or almost)- any chance of the "return" line fuel flowing backwards?? What if the car is pointed downhill, uphill, in a turn, etc.

Just some ideas I had.

 

[dgallup]

As long as the tank pressure head and return line pressure drop is not higher than the pressure regulator setting I don't see a problem.

 

[thruthefence]

I was wondering if the head pressure from a full fuel tank (against the pressure regulator output) would act as a 'restriction to flow. After posting the original question, I went & eyeballed the chassis, & the regulator itself (on the fuel rail) is almost to the top of the fuel tank. So I don't think it will be a problem fuel-metering wise.

this will surely "simplify and add lightness" to the project.

thanks, Guys

 

[MikeHalloran]

If you're converting _a_ car, it probably doesn't matter.

Good practice in many industries includes no penetrations in any surface other than the top of fuel tanks, probably because of historical problems with reliability of tank wall penetrations, and the consequences of leaks.



Mike Halloran
Pembroke Pines, FL, USA

 

[RossABQ]

I've been confronted with a similar situation on an old truck and I chose to put a return line on the filler pipe. The filler pipe can be easily removed from the vehicle, drilled, and a long tube run inside the filler tube that dumps down closer to the central part of the tank brazed in. Note that many older filler neck pipes were lead-coated on the inside, so remove the coating before you do any welding on it (toxic fumes).

 

[patprimmer]

I think the head of pressure does not add up as the difference in outlet height has the same effect as the fuel in the line has the same SG as the fuel in the tank.

My reasoning agrees with Mikehalloran. Potential for leak at the skin fitting.

Regards
Pat
See FAQ731-376 for tips on use of eng-tips by professional engineers &

http://eng-tips.com/market.cfm

for site rules

 

[dicer]

Just try it and see how it works.

 

[TStaples]

Fuel tanks are designed to have as few entry/exit points as possible. The main reasons are safety, and evaporative emissions, more so than any gains in mfg. efficiency. It is safest to have no fittings or lines on the sides or bottom of the tank that could be sheared off or cracked, in a crash. I'm assuming you have no evaporative standard to meet. Besides, getting rid of a carburetor, will do FAR more to reduce overall vehicle evap. emissions, than any other single modification. That, of course, is assuming it's installed correctly.

The only other concern I would have is if the vehicle is old enough to not have a carbon canister, and pressure could build in the fuel tank on a hot day due to fuel heating, would it be able to push fuel back up the line to the regulator. Even if it did, would this cause any particular harm? It entirely depends on your particular hardware and mechanization, but I doubt it.

-Tony Staples

http://www.tscombustion.com

 

[LionelHutz]

Last time I did that, I just took out the sending unit and brazed a second line in the top. I don't like the idea of dumping fuel in the top because, to me, it would seem that would cause more vapour build-up or vapour venting of the fuel.

I can't see how the fuel flowing when parked could matter. The injectors are closed so there fuel system is a closed loop from pickup to return fitting with the exception of there being a spring loaded valve (the regulator) in the middle of the loop.

 

[TStaples]

Lionel,

You are assuming that injectors don't leak when closed. Just the opposite is the case. ALL injectors leak, to some degree, when closed. It's supposed to be very small, but it's there. Injectors utilize a metal-on-metal seal of some type (tapered needle & seat, ball & tapered seat, flat plate over flat seal ring, etc . . .) On a properly functioning vehicle, it should take hours for the rail pressure to leak down after shutting off a vehicle, but the rail pressure will eventually dissipate through the injectors, into the intake.

The concern would be if the backpressure to the regulator would be sufficient to push fuel back into the rail, and keep re-supplying fuel that would then leak out through the injectors. Assuming you could effect enough leakage, and the leakage occured after shutoff on a cylinder where the intake valve happened to be open, then you could, worst-case, fill the cylinder with liquid fuel during shutoff. When you next restart the vehicle, there is a high probability of hydraulically locking the engine during crank. The liquid fuel won't compress, and you then damage a piston or connecting rod.

Ask me how I know.

-Tony Staples

http://www.tscombustion.com

 

[LionelHutz]

Seriously, there is little change that would happen. By design, the fuel pressure regulator is more or less a 1-way valve allowing fuel to flow towards the tank.

And if this is such a dire consequence of a badly positioned return line then why doesn't this type of failure occur more often due to the feed line, which must pick-up under the fuel level at the bottom of the tank???

 

[TStaples]

Lionel,

As part of my nearly 30 years of experience in the fuel injection industry, at one time I was the Engineering Manager for a company that manufactured regulators for alternative fuels. Similar to injectors, they leak at their metering seats too. Never, ever, assume a regulator is also a shutoff valve at system shutdown.

However, on a modern, FI car, the vapors from the tanks are routed to a vapor canister, for capture, and very little pressure is allowed to build up in the system.

The OP mentioned that he is converting a carb. Toyota to FI. The tank venting properties of older carb. equipped cars was sometimes less sophisticated. Some of these systems allowed a fair amount of pressure to build up before a (typically) fuel cap vent might allow venting. I don't know those details. My whole point is that converting is not as simple as it may seem. What's going on at the tank is different, and not trivially so, between a carb. vehicle, and a FI vehicle. I recommend erring on the side of caution when dealing with fuel systems, especially when you don't have the specialized knowledge of a fuel systems engineer.

-Tony Staples

http://www.tscombustion.com

 

[LionelHutz]

OK, I guess you missed my point that worrying about this new return line back feeding and causing all kinds of havoc seems pointless when there is already a fuel pickup that can cause the same problems.

Also, I'm pretty sure excessive built-up pressure in an "old school" tank could also overcome the needle in a carb and cause fuel to either fill the engine or dump all over the intake.

I do agree that the venting of the tank should be investigated and also that odd failures can occur, using either injectors or a carb.

 

[NickJ67]

When worrying about pressure build up in the tank when standing in hot weather, it makes no difference whether the return line is above or below the fuel level. I seriously doubt that enough pressure could build up to cause significant leakage back through both the regulator valve and closed injectors.

I've converted my 60's Triumph to EFI and used the tank drain (with short internal upstand to avoid getting all the sediment) to provide flood feed to the high pressure pump and the original fuel line as the return. It works fine. I had a fuel pressure gauge fitted a fuel pressure gauge fitted for quite a while and that showed that the system would hold the 3 Bar running pressure for about 2 hours after switch off before falling away. The leakage was not through the injectors, so must have been through either the regulator or back through the pump NRV. Doesn't matter either way as pressure returns instantly during the 2 second pump prime cycle. All parts were "recycled" junkyard rescues - good stuff Bosch.

Worth mentioning that on some cars, where the tank is low mounted it is sometimes necessary to use a lift pump and swirl pot to provide a reliable feed for the high pressure pump. Some people take the fuel return to the swirl pot rather than the tank although I'm not convinced thats a great idea on a road car that spends time in traffic as the fuel get pretty warm surprisingly quickly. On my car the whole tank warms significantly, although at 8.5 gallons (imperial) it isn't that much bigger than a swirl pot anyway.

Nick

 

[thruthefence]

Thanks to all for the input; maybe a little more info will help.

The car is a 1970 S2 Lotus Europa. The engine is a Japanese domestic market 1.5 ltr 4AGE 20 valve, with the 'swinging gate' airflow meter. Four throttles feeding from a common plenum. The fuel tank is a trapezoid, wide at the bottom, narrow at the top. The Fuel pump, ( a Bosch, with filter sock, from a 2.0 Subaru turbo) sits in a baffle pan of approximately one gallon in the bottom of the tank. It feeds to the top of the tank through a #6 (3/8") aluminum line & AN bulkhead fittings. I have the room (barely) to build a swirl chamber in the tank for the return line, but would like to avoid the complexity. The tank has a 'bung', in it's side, where the original fuel qty transmitter was installed, and I hope to install an aluminum disc, with bulkhead fitting, in this spot, with a #4 (1/4") aluminum tube emptying tangentially against the wall of the tank, which should minimize the 'churning' of the fuel. The 'bung' is approximately 1/3 up from the bottom of the tank.

 

[TStaples]

Nick,
I agree that the pressure shouldn't normally build up to too high a point. It really depends on the RVP of the fuel in the tank, and the temps involved. I have seen values of 15 psi and more on certain closed systems, but most most vent below that.

Thruthefence,

Thanks for the additional info. Great car, BTW. However, I would assume, since it's a Lotus, that crash isn't your highest priority. Given that, the side port should work fine as a return point. Just make sure there is some type of venting fuel cap.

-Tony Staples

http://www.tscombustion.com

 

[patprimmer]

I would certainly run a sizable open vent if this does not need to pass smog regulations.

So long as the injectors are not above fuel level when standing and there is no pressure from vapours, and the pump is not running, there is a limited amount of fuel that can leak. The sums are simple and I expect that it will be not that much more than one chamber full.

Aloww for some in each chamber and some time to leak and some evaporation losses and the risk seems small.

Oh hydraulic lock can also blow head gaskets (if your lucky) or the centre out of a spark plug (if your luckier still) or pull or you head bolts, threads and all, out of the block.

Regards
Pat
See FAQ731-376 for tips on use of eng-tips by professional engineers &

http://eng-tips.com/market.cfm

for site rules

 

[benta]

My opinion is that the return line should end in the baffle pan.
An acquaintance of mine had a problem car, which would stop whenever the tank was at the 1/4 level.
After numerous attempts to correct the problem (at great expense), the last desperate attempt was to break the tank open.
Due to a production mistake, the return line ended outside the baffle pan, which meant that when the fuel level dropped below the edge, the flow was high enough to empty the baffle pan and the car stopped.
A new tank, and the problem was solved.

Benta.

 

[WGJ]

Here in the UK we ran 'old school' carb systems up to 1993.
Venting was direct to atmoshere via a restrictor of about 1mm dia, with the vent continuing via a nylon line of about 4mm ID to a clipping point on the (typically) steel tank somewhere.
Prior to 1993, certainly during the early 80s, FI/EFI derivatives of carburetted models used pretty much the same arrangement as carbon cannisters generally weren't used because legislation did not mandate them.

So - return your fuel to the tank via a 6mm OD (at least) fuel line; enter it via the sender flange and take it down to the bottom of the tank. It always used to work fine.

Bill