RA28 Intank Idea (Carby to EFI fuel tank)

[DoctorDubb]

Ive done a search many a times in regards to surge tank, EFI fuel tanks and the like. With the added cost of having a set up with 2 pumps, a surge tank, 2 filters, more pipes everywhere and then having to mount the surge tank somewhere. I have been thinking about an intank solution. The reason that you need some sought of 'surge device' is that firstly to a stock carby tank you cant just slap on an EFI pump as they are not good at sucking up the fuel. Also they dont have any baffles that stop fuel running to the sides of the tank when you accelrate hard or corner etc. Meaning your EFI pump runs dry which is very bad for the pump and you lose power. Many people have talked about an intank solution but no-one has really gone into details with it.

The specific case is that i have an RA28 tank which is reasonably flat and wanted to look into a solution of having an intank pump (advantages and disadvantages are discussed below). But would apply to any Carby to EFI fuel tank situation (one that doesnt have baffles). I set out for the following criteria: to have a cost effective solution (rather then building a whole new alum. drop tank ~$700). As well as using the original tank as it fits nicely aswell as being 'more engineerable' too. Keeping the design simple as welding to a fuel tank means that the weld has to been 'fully sealed', so the more welds the more possible holes you have - Welding to the bottom of the tank was just asking for holes/trouble. Original sender fuel level gauge can also be used for simplicity.

Anyway, Here is the idea:








N.B The original fuel lines will just be blocked up and i will have to get a EFI tank petrol cap so that pressure doesnt build up.

Its like a upside swirl pot design but with a different approach. It *possibly* could be setup in a way so that there was a low pessure inside the cylinder and a higher pressure outside (in the fuel tank) so that fuel 'wanted' to stay inside (but im not that good with fluid mechanics). As after thinking even more about the advatanges (aswell as listed above) to this type of setup; less noise as in tank, no surge tank (fuel) in your engine bay (the only other place for a surge tank in the RA28 is near the stock muffler location), no fuel smell from having lines in your boot and possibly less hassle from the engineer/cops in this respect too.It seems the best way to go over an external surge tank. I have a few ideas of the actual cylinder with different holes designs and possible even like a swirl pot (alternating conical 3/4 circles if you know what i mean) but thought the idea listed above would give the best 'surgefree' design.

Any comments on wether this will be affective as an anti-surge design or will/wont work at all would be much appreciated. I realise its not an revolutionary idea but just trying to bring together ideas and comments at the same time aswell.
Cheers,
Steve

P.S WARNING: It is dangerous welding/cutting a fuel tank - if you decide to try this at least wash it out many times with soapy water as well as running exhaust gas from you car while welding/cuting as a precautionary measure. And the normal disclaimer- i take no responsibilties for the safety/intgerity of this design, use your common sense when working with cars.

[The Witzl]

This idea has merit. It's pretty simple, and not too costly.

I've been thinking about an intank solution for the RA28 for some time now, and had come up with a relatively similar idea to this but have never given it any serious thought.

As an alternative to your idea of returning fuel back into the surge cylinder, why not use the stock return line that is a fixed part of the tank.

[BabyZ]

Definitely has merit - I've always thought external surge tanks to be a bit of an inelegant solution.

I'd stick with your idea of returning the fuel to the 'pot' - that's what the factory in-tanks do and it helps to ensure that it stays full when the tank is getting low.

I'd also consider making the whole thing screw-on to avoid welding on the tank entirely. The green disk on the top would simply need to be a bigger diameter with a second series of bolt holes on the part 'outside' the pot walls.

Edit: Holding the nuts captive inside the tank until you screw the pot down might prove a challenge, but there'll be a way

This would, of course, depend of whether you had a perfectly flat portion of fuel tank to bolt it to. But it would eliminate the need to have any leak-proof welding done at all, the pot could simply be tacked onto the bottom of the disk.

[brett_celicacoupe]

remember that with those purple holes in the cylinder, fuel will flow out just as fast as fuel is able to flow in. so you will be essentially just offseting the problem w.r.t time. this is also a similar problem people face with their sumps and oil delivery.


so going around long sweeping corner, the fuel will inevitibly flow out of the cylinder and once its all out, youll probly be waiting for quiet a while for the cylinder to fill again. this directly related to the size of the purple holes.


i would suggest just using a complete surge tank to drop into the hole in the top of the fuel tank. it doesnt have to be alloy, it can be steel so you can weld it to the tank.

i made my surge tank from 5" truck exhaust pipe (steel) i put my own fittings on it and built it to my specs. it doesnt look all that blingy but it literally cost me zero dollars to make.




if you made something similar and had the all the lines exiting the cylinder/surge tank via the top then it would essentially remove the space requirements but you hook it up as per any other surge tank. the compactness would be good and the engineer might like it better if its removing fuel hoses and fumes from the cabin.

BTW: i welded the surge tank with oxy-acetylene and brazed the metal hoses on. you dont neccessarily need TIG

[Hen]

I've thought about this a little recently. A couple of points to consider:

- When welding a fuel tank be very careful
- While it is best to ensure your welds seal fully, you can smear some 2 part epoxy over any suspect joins. This stuff is non fuel soluble so should be fine.
- On a similar note, if you need to paint any bare metal surfaces in contact with fuel use 2pack paint (which is also just an epoxy mix)

- The internal baffling I ripped out of an EFI Tarago tank recently looked like the diagram below.



Fuel wont slosh out due to cornering or accelerating, but will flow in (and out) due to gravity.

Hen

[DoctorDubb]

Witzl: Thanks, the idea was that since the fuel pump (WALBRO 255L/hour) i plan on using will flow way too much for my engine (1GGTE) then it will run though the regulator and back into the tank (and a lot of it). Since the bigger lines and reg etc will hold a fair bit of fuel, this hopefully will give me some 'antisurge' fuel from this return to inside the cylinder. If it exists on the outside of the cylinder then this fuel just joins the 'surging fuel' anyway.

BabyZ: Yeah good idea w.r.t. not actually having to weld anything to the tank. As you realised in your edit it would be hard to hold the nuts inside the tank . But you could do it if you used a two part lid. The first would just be a thickish ring with two sets of holes, the inner holes would have nuts welded to the bottom and the outside ones would bolt upto the tank. That way there is a hole in the ring to stick a tool/hand down to hold the nuts for tightening this ring to it. Then just use your idea to mount the top part of the lid as you now would have the welded nuts. This two part system has more tendency to leak (even if you used rubber) but does have merit for a lot of flater tanks. The RA28 tank isnt very flat at all, and i think it would be hard to find a flat top on the top of the tank and with that lining up with the lowest part of the tank.
Here i took some pics of the RA28 tank:




Brett: I can see what you mean about the flow in/flow out problem. Does your solution still only use one pump? As if i understand that using it as a 'normal surge tank' would require 2 pumps and i was trying to avoid this (cost and my alternator is probably not going to be able to handle 2 very well). But is a good idea for a compact and discrete surge tank and the use of 5" truck exhaust pipe . How i thought to rectify the problem with flow out is three ways i thought about this morning:
1) Have a 'low pressure' inside the cylinder and high pressure outside (in the tank) that way fuel would want to stay inside the cylinder (preferentially) and would run into the tank faster then runnin out. I thought possible with the sucking action of the pump this might happen, but would have to do more research.
2) I know ive drawn the holes so that they are all around the cylinder. If i could make a big enough hole(s) at say the front of the tank maybe (as this would only receive fuel running out mainly at braking), at this time im most likely going to be off wide open throttle (in a street environment) but may not work too well in a racing environment.
3) This idea would be a lot better if i knew wether it existed or not. In the holes place large-screw in one way valves (would have to make sure they flow enough fuel), that way fuel could get inside the cylinder but wouldnt be able to 'escape'. Problem is wether they flow enough and wether there are any that are non-soluble in fuel.

Hen: Nice work on the diagrams and the hints about welding and painting the tanks. This looks like a good idea you would just have to make sure it would flow enough out of one hole in order to 'refill' the tank.(The tarago wasnt exactly as sports car by any means ). Or theres other designs i have read about on the net in some WRX's with alternating conical 3/4 circles that get bigger and bigger, which is another idea too. Did you have any dimensions on the Tarago baffeling box ? Just to see how big?

[brett_celicacoupe]

would you really want to risk leaning out your engine at WOT just for the sake of one pump? the probability of this happening is probably slim but that all depends on your design. $50-100 spent nowcould be $500 saved later

a $150 lift pump is much cheaper then a $500 replacement engine.

i am and have been using a generic 'clacker' lift pump and its worked fine for the last 12months..... $30 repco item.


the idea about the 3/4 circles sounds alright.

[Hen]

Tarago box thing is about 200 long, 100 wide and 150 deep. I agree that the factory setup may not flow too much into the box, but another pipe could be added and their diameter could be increased.

Several 3/4 circular walls offset from each other is a good approach too.

Plus you could consider using an external lift pump sucking through the stock pickup to fill the internal surge tank. I agree with Brett above too, fuel systems, like brakes are something where overkill is a good idea.

Hen

[TA-022]

Efi pump $200
lift pump $60
surge tank $50
=
$310

not that bad when intanks are still $200 plus design and welding etc.

[Mr Ed]

My EFI pump cost me $120..........rated at 550hp!

Lift pump can be the existing pump, surge tank is farken cheap as (or free if you can weld).

Holes in the bottom of your surge tank completely defeats the purpose of having it. I can see no advantages in that design other than adapting the tank to work with EFI........basically the exact same thing as just plumbing a return line into your current tank and upgrading the pump.

Big fuel pumps can also require big fuel regulators though. If needed, that can add another $200 to the equation. I would also suggest an in-cabin fuel pressure gauge for any upgraded/custom fuel system.

[Mr Ed]

it would be hard to hold the nuts inside the tank . But you could do it if you used a two part lid. The first would just be a thickish ring with two sets of holes, the inner holes would have nuts welded to the bottom and the outside ones would bolt upto the tank. That way there is a hole in the ring to stick a tool/hand down to hold the nuts for tightening this ring to it. Then just use your idea to mount the top part of the lid as you now would have the welded nuts.

Pop-rivets FTW

[Hen]

The attraction of an intank setup is that everything is inside the tank. Do it properly once and never think about it again. Currently I have a big lift pump, low pressure filter, surge tank, EFI pump and lots of hose clamps shoehorned under the rearend of a Sprinter (I don't want it in the hatch).

If I could remove all that junk it'd be much neater and I suspect more reliable.

Hen

[Mr Ed]

The attraction of an intank setup is that everything is inside the tank.

I agree completely, but what he drew is not a surge tank.

[brett_celicacoupe]

to put my idea another way,
put the surge tank inside the main tank and hook it up as per normal

[TurboRA28]

Mmm yes I think that idea Brett has merit. Would you have the main feed line to the hiflow pump coming out the bottom of the tank?