Board logo

Fuel injection petrol tank design
Matthew_1 - 4/10/04 at 04:53 PM

Does anyone have a design I could see for a fuel tank for a car with fuel injection. As I understand it you're mean't to have a return "pot", but I'm unsure how/where this is placed in the tank. It's probably dead simple but all the pics on here show the outside of the tank! (Anyone got one in clear perspex!!)

Did try a search but it returned nothing - anyone know if the search just does titles?

Ta.


zetec - 4/10/04 at 05:41 PM

The feed from the tank normally goes to the fuel injection pump which pumps to the fuel rail supplying the injector(s).The presure in this section is set by the regulator and any excess fuel returns to the tank. As the fuel pump needs a constant fuel supply some people use a swirl pot of 1-1.5 litres which will hold enough fuel even if the fuel tank outlet is starved while cornering. There are a few different solutions, and even a baffled tank can work fine if you don't run it low on fuel all the time.

Try http://www.sbdev.co.uk for some good info.


Matthew_1 - 4/10/04 at 05:54 PM

Thanks for the info but that's more how to plumb the whole system in. I'm more interested in the design of the actual fuel tank.

Is the swirl pot just a simple, open-topped cylinder, attached to the bottom of the tank and the return feed feeds into the top and the injector feed takes from the bottom?

Presumably this cylinder then needs a few small holes near the bottom to let the remainder of the fuel in the tank top it up ?


JoelP - 4/10/04 at 06:30 PM

there are many ways a swirl pot can work - you know what is needed so just have a fiddle and see what comes to mind. You need small enough holes to prevent it all running out in a corner, but big enough to allow plenty of fuel in.


Mark Allanson - 4/10/04 at 06:45 PM

I made my tank to have the fuel outlet in the centre. It is baffled and I built in a chamber in the middle to act as a swirl pot. As fuel sloshes around in the tank, it must pass through this chamber. The chamber is also fed by the fuel return, so if the theory is bourne out, I should never get any fuel starvation


stephen_gusterson - 4/10/04 at 06:55 PM

ive explained what i did a few times

do a search on posts under my name with

swirl pot

as a search string.

think i put a crappy sketch up too


atb

steve

ps - didnt return into pot - would prob be a good idea


ady8077 - 4/10/04 at 07:51 PM

Hi

Some people prefer the return not to go direct to the pot as the fuel becomes heated and stale contiually looping around

Adrian


stephen_gusterson - 4/10/04 at 09:31 PM

i kinda thought that, as well as it may create bubbles and stir up sediment.

im loosing no sleep!

atb


steve


Spyderman - 4/10/04 at 10:12 PM

The fuel in the swirl pot is hardly going to become stale as you would be using at least 20% of it each pass. If not you're not driving hard enough!

Petrol doesn't foam like diesel so gassing up will not be a problem.

Where is this sediment that is going to be stirred up coming from? If you have sediment in the swirl pot then you should clean it out. Vigorous mixing in the swrl pot will ensure you don't get sediment!
The swirl pot should be just that "a swirl pot" so that the fresh fuel is mixed with the cycled fuel.

Having the cycled fuel return to the swirl pot will also enable a lower rate pump to be used from tank. This will only need to supply maximum flow for engine use at highest power. Whereas having the return go back to the tank means the primary pump has to be as efficient as the high pressure injection pump or you will get starvation problems.

Oops, just read original thread again!
This is all only relevant to a separate swirl pot external to the tank.
If you are just making a swirl pot in or under your fuel tank then where the return pipe goes back into the tank is irrelevant.
All you need to make sure is that the fuel can flow into the swirl pot from the tank faster than it could flow out due to cornering forces, so therefore a chamber below the main tank would be more efficient. Much like a reserve tank!

Terry


iwbunting - 4/10/04 at 10:23 PM

Swirrrrrrl pots.

I was talking to another locoster a while back who gave me the most obvious answer to the swirl pot debate.

All manufacturers of injection based cars have :

A - Swirl pot in petrol tank

OR

B - External swirl pot (not very common)


So why not use the pot out of the donors tank.This being the most simple answer as the pot was obviously designed to work with the original donor.

Take for instance a Ford sierra.The pot sits directly below the pump ,filter and return feed.It is a simple task to remove the pot by splitting the spot welds and tacking into the new tank.BTW this swirl pot is shaped like a snails shell with the filter etc sitting in the middle.


Hope this is of some help.


stephen_gusterson - 5/10/04 at 09:53 AM

spyder

your last para applies to my approach

iwbunting

how does an internal pot work?

surely fuel will slop out as fast as it pours in thro any holes cos it will be mounted level with tank floor.

if there is only 10mm fuel in tank, its not gonna fill pot.

an undertank pot (mines 150 tall) will be full of fuel from the 10mm layer above at any time. Same column in the tnak will only have 10mm in it, and any fuel that can be returned to the tank could top it up. But then the purpose of the pot is to supply a constant load of fuel, catch 22 methinks.

Using anything from the donor tank - it its integral - is risky. Do you want to cut a fuel tank open ?

kabooommmmm?

atb

steve


iwbunting - 6/10/04 at 08:10 PM

On the sierra tank the swirl (snail shell) sits directly below the return pipe and inlet filter.Dont ask me how it works but in my mind the fact that its fitted and obviously worked would be good enough reason to use it.

As regards cutting it out,i let the tank vent open to elements and then took a thin cutting disk and tin snips to it,honestly not a problem...i'm still hear to tell the tale.

Hope this helps.


JoelP - 6/10/04 at 09:28 PM

an internal swirl pot is more like a well thought out baffle. A true swirl pot, IMHO, will be fed by a low pressure pump. this has the advantage over a gravity fed pot that the fuel tank can be lower.


iwbunting - 7/10/04 at 07:16 PM

response u2u to stephen gusterson

I'm still here in one piece after cutting the donor tank to retrieve the swirl pot.If it worked in the donor why wont it work in a locost.

Be curious to here your comments.

Response from stephen gusterson -

i recon in order for a swirl pit to work properly, it has to be at a low point in the tank.

im guessing your donor tank was a typical weird road car shape, and not flat bottomed. Was it at a significanly low bit of the tank?

my tank is rectangular. The last gallon will only be about 12mm deep in the tank. How do you keep 12mm of fuel in a swirl pot, such that its more than 12mm deep in the tank?

Thats why I made mine as a 50mm tube 150mm long welded to the underside of the tank.



Correct me if i am wrong,but if it works in the donor why wont it work in anything else.

Any other thoughts ????


stephen_gusterson - 7/10/04 at 08:11 PM

why did you ask u2u if you wanted the response on the forum? And why u2u me the same thing three times?

You also didnt answer my question - is it at the lowest point of the donor tank?

perhaps the answer was in your other post

-------------
The pot sits directly BELOW the pump ,filter and return feed.
----------------

So, this means the lowest point of the fuel system?

why dont you post a pic so we can comment better?



atb

steve



[Edited on 7/10/04 by stephen_gusterson]


stephen_gusterson - 7/10/04 at 08:19 PM

this was a drg I posted of my system a few months back. Rescued attachment swirl.jpg
Rescued attachment swirl.jpg


iwbunting - 7/10/04 at 08:33 PM

Steve.

Thought i'd do u2u as i thought you where directing your questions directly to me,my appologies if you werent,i wont fall for that one again.

Christ,some people can be picky...Wasnt aware that message had been sent 3 times,not sure what happened.Whats up ?? thought i was trying on ??

Back to debate.

The swirl pot in the sierra tank as i said before sits directly below the inlet filter and return pipe,what i didnt say is that it sits on the bottom of the tank whicg is flat across its whole area.

I will be using this swirl pot as i still cant see why it shouldnt work,the other locoster that i reffered to in my earlier post is using the same and has had no noticeable problems.


stephen_gusterson - 7/10/04 at 08:37 PM

i think a basic rule is that if it works for ford, it should work for you! all things being equal

http://www.fueltanksdirect.co.uk/

lots of pics of fuel tanks!

there seems to be a difference in tanks for injection and non injection models on most fuel tank sites - must be the swirl pot........ and im assuming your tank is from an injection donor....... ?

atb

steve


JohnN - 9/10/04 at 03:57 PM

The drawing of the tank from Stephen Gusterson... That is just a sump, not a swirl pot, this will simply collect any debris in the tank - but will satisfy the need for a reliable source of fuel in a shallow tank.

A swirl pot makes use of the return flow from the injector rail to fill and provide a reservoir of fuel for the HP pump. This is normally arranged in the tank, so that the swirl pot will overflow back into the main tank.

The snail shell labrynth described in the Ford tank is simply to dissipate the energy and air present in the HP return.

So all in all I agree with IWBunting, use the one in the Sierra tank, in true Locost fashion.

There is a graphic on the following link which demonstrates the action of a fuel swirl pot

My two penneth! I'm using carbs anyway so just need a baffle or two - or even a sump followed by a good fuel filter

Swirl Pot

JohnN

[Edited on 9/10/04 by JohnN]


stephen_gusterson - 9/10/04 at 04:45 PM

my take off is a couple inches above the bottom of the tube for sediment reasons

Also, my return doesnt go into the pot as per the drg.

The idea is to keep a 'buffer' supply of fuel in the sump as you call it, and as such it complies.

The jag XJ-s has a similar - ish approach. It has a small (totally seperate and connected by a pipe) tank under the main tank. As far as I recall, the return does not go back into the lower tank, but into the top one.

in effect, is has no swirl pot but an under tank sump. Seems to work pretty well for a car that does 3mpg under high acelleration.

atb

steve

[Edited on 9/10/04 by stephen_gusterson]


stephen_gusterson - 9/10/04 at 05:02 PM

quote:
Originally posted by JohnN
T

The snail shell labrynth described in the Ford tank is simply to dissipate the energy and air present in the HP return.

[Edited on 9/10/04 by JohnN]



Thinking about this.... if the smail assembly is mounted in the level floor of the tank, it sounds more to me that its acting as a spiral trap for any fuel that might try to slosh out. If the fuel is taken from the centre of the spiral, and is returned to the centre, any fuel is going to find it harder to come out of the spiral on its return once its in there.

isnt the spiral to trap fuel rather than dissipate return energy?


sediment..... surely any sediment will make its way towards the exit..... ie migrate towards the spiral assembly........

atb

steve

[Edited on 9/10/04 by stephen_gusterson]


iwbunting - 9/10/04 at 07:12 PM

Stephen_gusterson

To save any more confusion view this pic. Rescued attachment AGF00011.JPG
Rescued attachment AGF00011.JPG


stephen_gusterson - 9/10/04 at 07:25 PM

great... now we can see what the ford bit looks like.

I imagine that the return and take off both dip into the box centre?

It must work, or ford would have had issues. But im obviously too daft to understand.


If there is 10mm of fuel in the tank, at rest, there must be 10mm of fuel in the pot, assuming its level at the flat bottom of the tank.

So, we start the car up. The pickup will hopefully draw fuel from the 10mm in the pot, shove it to the engine, and some will be returned.

new fuel has gotta come in thro the little entrace passageway around the sides of the box.

I wonder how easy this is when the tank is low on fuel and its all sloshing about.

when the level in the pot begins to rise above 10mm, it will tend to flow back out of the tank.


I dont see how this can work as well as a gravity fed lower sump UNDER the tank....

answers on a postcard please....

atb

steve


iwbunting - 9/10/04 at 07:38 PM

Cant we use the forum instead,might get a quicker response.

Postcards and royal mail dont seem to go together too well lately.

Sorry i couldnt post the pick earlier but without going into all the technicalities of the ford injection swirl pot,it must work or why fit it.

Ps.

The return and inlet filter both sit centrally in the pot


DEAN C. - 10/10/04 at 05:55 PM

Stephen,When the fuel gets to the bottom of the tank YOU STOP!
It doesn't matter if you got a swirl pot or not you cant run on fresh air..

My swirl pot is a rectangular chamber welded under the bottom of the tank sticking down about 40mm. The internal Rover fuel pump pick up and return sit in this.
Just imagine that it is like a small sump at the bottom.I'll post some pics in my archive tonight.And yes it works on the track under extreme forces,but there is a baffle in the tank as well.

I think the ford swirl idea would work fine but Stephen is right that with a sump idea youll get little or no surge untill you completely run out...
DEAN........


stephen_gusterson - 10/10/04 at 07:05 PM

yeah

think that happened to me once - in a hillman minx if i remember!

mounting it lower in the tank as you have is to my mind the best way.

thats why I have asked, and not got an answer, if the ford thingy is mounted at the lowest point in the tank. a lot of cars have whacky shapes that will tend to push the fuel to one end as it runs out. I suspect most locosts have flat bottomed tanks, hence my favoured 'sump' approach.

atb

steve