SeaBass
|
| posted on 25/5/04 at 12:38 PM |
|
|
Fuel Tank Swirl Pot
Anyone know if the PTM's plastic tanks have ever had any kind of swirl pot inside them. Also is a swirl pot absolutely necessary for fuel
injection?? Surely if you plumb in return on opposite side of tank the fuel feed will be bubble free? Baffles should stop surge etc.
Cheers
|
|
|
|
|
stephen_gusterson
|
| posted on 25/5/04 at 01:23 PM |
|
|
its even more importatn with efi to use a swirl pot. the pumps wear out fast if they run dry. a swirl pot, if attatched to tank bottom, is basically
filled with fuel via a hole in the tank. as the fuel sloshes around, there will be a minimum level in the pot. Without a pot, the fuel wont
continually cover the hole, and you will draw in air.
atb
steve
ps - i cant see how a swirl pot build in to the level floor of the tank would work...
|
|
|
zetec
|
| posted on 25/5/04 at 02:14 PM |
|
|
I fitted a PTM tank which he modded for me with a swirl pot. The swirl pot was about the size of a spary can lid fitted to the tank base, but
I'm not convinced it made much of a difference. He said the tank is baffled and it has been fine. You will need to take the pump feed from the
bottom of the tank, mine had a 10mm section which I drilled, taped and fitted with a 1/2" outlet. Might be worth try as is, and if you have
problems fit a swirl pot, remember to do it this way you should be using a low presure feed pump as well as the injection pump. I would have thought
an in tank unit would also work if you could find one which fitted.
[Edited on 25/5/04 by zetec]
|
|
|
mackie
|
| posted on 25/5/04 at 02:39 PM |
|
|
It's also super important on an EFI engine because as soon as it starts sucking air your engine cuts out, which on a fast bend = spin.
|
|
|
NS Dev
|
| posted on 25/5/04 at 03:07 PM |
|
|
Hit the nail on the head Mackie, had this happen on a tarmac rally car when the swirl pot wasn't big enough for the engine's fuel
consumption (even though the fuel return from the rail was direct to the swirl pot, and it had a Facet red Top feeding it) spun, hit bank and rolled
(went back together ok though!!)..........best to use at least 0.5 litre pot to do any good if you are venturing onto any tracks, corners can be long
and it's all very well saying you can run the tank full most of the time but on the track you won't cos you'll use it much faster
than you thought possible (5 mpg seems to be about the figure that most 2.0 16v cars achieve, regardless of the theory side of things!!)
|
|
|
mackie
|
| posted on 25/5/04 at 03:33 PM |
|
|
God knows what our 3.5 will be like :/
|
|
|
stephen_gusterson
|
| posted on 25/5/04 at 03:47 PM |
|
|
I know nothing about fuel dynamics, but this is the way I looked at it.........
Consider your tank is low and that fuel is sloshing about. Its not gonna cover the bottom exit hole a lot of the time. Consider the flow rate of the
fuel from the pump to the engine, and the size of the swirl pot.
Assuming (cos I dont know actual figures) that during max accelleration, you are doing about 5mpg, you can work out a flow rate. You can then work out
how much of a resevoir you need to cope with that fuel rate for say, a 5 second slosh of fuel.
I hope you can see what im getting at. And that is that I doubt that a swirl pot the size of a paint can top would hold much of a reservoir of fuel -
seeing as my granada pump has a 15mm inlet!
My jag V12 had a swirl pot of about 2 litres capacity - that did 3mpg under max accelleration.
How do you get a swirl pot INSIDE a tank?
If, say, your tank has 10mm fuel in the bottom, and its sloshing about, you can have at best 10mm in the pot. You need holes in the pot to allow the
fuel to slosh into it, and its surely gonna slosh out at the same rate - so net effect is it doesnt work.
Or have I got summat wrong here?
What I did was to make up a swirl pot (also prob too small) out of 150mm of 50mm tube. This is welded to approx 40mm hole in the bottom of my tank.
This holds about 0.17 of a litre (170ml) which at my (3 litre flow? - its a recirculating system - not all of the fuel is burnt) assumption is good
for a few secs fuel interruption.
At 60mph acellerating to give 5mpg instantaneosu fuel consumption would need approx a litre of fuel a minute. However, looking at the return flow from
the metro turbo I had, a fair amount is returned to the tank in order to maintain fuel rail pressure.
atb
steve
[Edited on 25/5/04 by stephen_gusterson]
|
|
|
ady8077
|
| posted on 25/5/04 at 04:23 PM |
|
|
Hi SeaBass
If you havn't bought your tank yet NF autos sell a plastic tank with a small surge tank underneath
http://www.nfauto.co.uk/fuel_tanks.htm
Adrian
[Edited on 25/5/04 by ady8077]
|
|
|
mackie
|
| posted on 25/5/04 at 04:49 PM |
|
|
It appears that place also have a Ferrari V12 engined kit car for sale too 
That tank looks really good though definately going in the bookmarks for when we need it.
|
|
|
SeaBass
|
| posted on 25/5/04 at 04:52 PM |
|
|
NS Dev, what stops swirl pot overfilling if return feed comes back to it and transfer pump filling it as well?
Cheers
|
|
|
zetec
|
| posted on 25/5/04 at 05:27 PM |
|
|
As said I don't think the swirl pot on mine does much, but I've never noticed fuel starvation yet. The thing about the MK tank is that it
is tall and thin where most production car tanks are flat. I would have thought a 1ltr swirl pot would be plenty, cost around £100 and tend to be tall
cylinder designs as per Steve's design. I suspect there are advantage and disadvantages to return to tank set ups, less flow rate but less
chance of bubble build up in the fuel. Some very good diagrams on http://www.sbdev.co.uk/
|
|
|
ady8077
|
| posted on 25/5/04 at 06:19 PM |
|
|
Hi Again
I've seen a few remote surge tanks for sale, they all seem to be about 1 or 1.5 litres, if you want a locost version early VW golf GTI's
use a combined surge tank high pressure pump. May be in other VW's too?
|
|
|
NS Dev
|
| posted on 26/5/04 at 07:15 AM |
|
|
Seabass, there is a return from the top face of the swirl pot back to tank, so the pot should be charged completely to the top outlet with fuel under
normal conditions. The ones I have made in the past were approx 200mm high and made from 4" exhaust tubing. I had a 1/2" outlet at the
bottom to feed the bosch injection pump, a 5/16" inlet at the top (in the side of the "cylinder" for the feed from the facet red top
pump, then a 5/16" port in the very top of the "cylinder" which split into a tee, one side was the return from the fuel rail, the
other returned to the tank.
|
|
|
SeaBass
|
| posted on 26/5/04 at 06:00 PM |
|
|
Many thanks guys for the advice. I'm going to fab up a swirl pot to the design I've produced below. Combination of your descriprion NS Dev
and others I've seen. That SBD site has very good plumbing diagrams for anyone else interested thanks zetec.
Cheers
|
|
|
stephen_gusterson
|
| posted on 26/5/04 at 06:34 PM |
|
|
dunno which way up that is, but have you ensured a nice big hole in the top to let the fuel from the main tank come in?
atb
steve
|
|
|
NS Dev
|
| posted on 27/5/04 at 08:00 AM |
|
|
if it's like mine it was pumped in by the facet redtop pump.
|
|
|
SeaBass
|
| posted on 27/5/04 at 11:47 AM |
|
|
Yep as per SBD diagrams standard 8mm hose in from and out to tank. 12mm hose from bottom to inlet to efi pump. I presumed return was 12mm from fuel
rail. I have no details yet of whether that is correct as I've purchased nothing yet. What diameter is high pressure efi hose usually?
Cheers
|
|
|
stephen_gusterson
|
| posted on 27/5/04 at 01:10 PM |
|
|
my car, using granda efi parts, has a 15mm feed from tank to pump, all else is 8mm pipe
atb
steve
|
|
|
mackie
|
| posted on 27/5/04 at 02:55 PM |
|
|
Same here I think although I'm not 100% that the pump inlet is 15mm (but it's definately bigger)
What are you doing for a pre-pump filter? ATM we have a VW 15mm (ish) out and 8mm in plastic thing, but I think we'll want something better
later.
|
|
|
stephen_gusterson
|
| posted on 27/5/04 at 03:47 PM |
|
|
to be exact, its prob about 16 - 17mm ish - I used some imperial petrol compatable pipe and it compresses ok with a jubilee clip.
If I remember, the granada (prob same as sierra) system has a direct feed to the pump, and a 8mm ourput that then goes to a filter. Seems that fuel to
the pump isnt filtered, but it is to the injectors....
atb
steve
quote:
Originally posted by mackie
Same here I think although I'm not 100% that the pump inlet is 15mm (but it's definately bigger)
What are you doing for a pre-pump filter? ATM we have a VW 15mm (ish) out and 8mm in plastic thing, but I think we'll want something better
later.
|
|
|
MustangSix
|
| posted on 27/5/04 at 06:09 PM |
|
|
Some first hand experience.
There's no swirl pot on my Mustang, but no problems either. Whether you need one or not will depend on other design factors.
The tank design plays a role. My tank has an internal "dam" that helps keep the pickup submerged. Likewise, a tank with a pickup sump
would not require a swirl pot either. It's analogous to an engine oil pan. Under most circumstances the pickup is submerged and the pan
baffles will help keep the oil there when cornering, accelerating, or braking.
The fuel rail design is also key. On mine the fuel rail is large diameter tube that holds a fair amount of fuel under pressure. The regulator is on
top of the rail feeding the return line. Even if air were to enter the fuel line, it would collect at the top of the rail and expelled thru the
return line, so the fuel pickup becoming momentarily uncovered really has little effect.
Under most circumstances, there is little chance of my rail running dry. If I were to do some constant high-G cornering with a low fuel tank, it
might be a concern, but in normal spirited driving it has never been a problem.
[Edited on 27/5/04 by MustangSix]
|
|
|
blueshift
|
| posted on 28/5/04 at 12:15 AM |
|
|
I wonder if we should build in an external swirl pot.. our pump + regulator system cycles fuel through very fast, like a litre in 5 seconds kind of
fast. This might be because the fuel pump doesn't have any balast resistor on it (it should really) but hmm. I can see a swirl pot emptying very
fast if the return doesn't feed into it, and I can't think of a satisfactory way of doing that in a swirl tank that's just a low
pocket of a normal fuel tank (think the fuel from the return would just gush and spray out of the swirl pot).
this is a rover v8 efi with a new bosch fuel pump.. correct type I think.
|
|
|
mackie
|
| posted on 28/5/04 at 09:02 AM |
|
|
An external one like what Jon Wolfe had on his race car would be nice. I think it's basically what seabass posted a drawing of. Would need a low
pressure pump from tank to top of pot then a takeoff at the bottom for the high pressure pump, a return from the fuel rail going in near the top and
an overflow back to the tank in the top (also letting any air escape). Shouldn't be too tricky to make at all, not sure how much they are to buy
but I'd hazard a guess at "not that cheap".
Not that those tyres that were on the sierra will give us any decent cornering g but hey.
Mackie goes to check if they do 14in CR500s. Oh they don't.
Actually just looking at the easytrack member list *everyone* seems to use avon tyres...
|
|
|
stephen_gusterson
|
| posted on 28/5/04 at 10:14 AM |
|
|
if you are overdriving your fuel pump it might burn out. Many pumps actually run in the fuel itself for cooling.... so id use a ballast resistor if it
was me!
atb
steve
|
|
|
JB
|
| posted on 28/5/04 at 11:57 AM |
|
|
Swirl Pots For Fuel
There has been a lot of talk on here about collector pots which is a small tank to ensure a good head of fuel under all conditions for the pump. A
swirl pot is designed to take air out of the system.
A collector is required for EFI and means you can run your talk low on fuel. I can run my tank down to 1 litre (but collector will have 2 litres of
fuel in it ) with no fuel starvation.
For pictures of my setup
http://www.beardmorebros.co.uk/website%20pages/update_10th_July_2003.htm
John
|
|
|
