Swirlpot for Pi

[john.r.davies]

I need a swirlpot on the Pi system I've fitted to my non-TR6 Triumph  The tank starts to deliver air on corners when less than half filled.

I can't find a suitable one commercially, so my local perfomance shop is making one for me.   The patron advises me that as well as fuel from the tank going to the 'pot via a lifter pump , I should have the excess fuel returning from the injectors go into the 'pot, with a vent back to the tank for spillover and any air.

 

I trust him, and can see the idea, to ensure that the 'pot is always as full as possible.  The arrangement is clear with a modern, low pressure electronic system, but what about the Lucas Pi?  It has TWO fuel returns, from PRV and Metering unit.  Which should be the auxilary supply to the swirlpot?

 

 That from the PRV will vary in flow from most at idle to least at full throttle.   Does the M/u return vary as much, or is it just there for a low, relatively constant, lubricating flow?   I notice that the PRV return is 6mm hose, while the M/u return is only 4mm.

 

I'll be grateful for information and views,

Thanks,

John

[Roger H]

John

 

I'm not sure if this will be appropriate for a "non-TR6 Triumph" but I used a Jag XJSV12 swirlpot for my ex-US TR6 (now PI). I need it as the US tanks were not baffled.

 

It is a perfect solution as it has all the required large diameter hose connections and is purpose made to do the job. I would recommend increasing the outlet from the main tank to 3/8" if you are using a US fuel tank. If you do this, you will also not need a facet pre-pump if you can locate the swirl tank below the main tank.

I re-routed the PRV return line into the swirl tank. If your metering unit is in good nick, there is not a lot of fuel that comes back from the engine.

 

There is a slight problem though if you are installing it in the wheel-well in the boot of a TR6 - it's fitted in the corner closest to the driver's seat and the spare wheel won't fit in!!!

This is not a problem for me though as I use that space for tool/kit etc rather than the spare! There is also a bonus though - it hold a couple of litres of extra fuel, which is a welcome addition to the piddly capacity of the US tanks.

[Peter Douglas Winn]

Roger

Can you explain "piddly US tanks"?

My 1971 US car has a tank with 11.45 gallons and I thought that was good! Perhaps I am wrong.

[john.r.davies]

Roger,

Thanks for that - I suspected that the return from the M/u was small, so the PRV return is clearly the one to use.

 

Thanks too for your other advice.   I should not have been so coy - mine is the Silverback Vitesse that measured the width of the Start/finish straight at Castle Combe recently (see http://www.putfile.com/media.php?n=Silverback !).    As it is based on an estate, the tank is very low and right across the back.  Triumph realised that this tank needed to be baffled and even provided a little gravity swirlpot, but that is so low that nothing could get under it and live.  And not adequate for the demands of racing.

 

John

[Guest andythompson]

The metering unit return should always vent to "atmosphere"... if it has any back pressure the metering unit gets most upset...

[GUY BENNETT]

In the rally world we are always having to improvise

Take a look at www.tas.uk.com or www.rallydesign.co.uk or www.demon-weeks.co.uk

[Richard Crawley]

If your metering unit is in good nick, there is not a lot of fuel that comes back from the engine.

The condition of the metering unit has no effect on the rate of fuel return from the metering unit to the tank. This is entirely dependant on the condition of the pump; fuel line pressure & sustainable flow rate!

[Roger H]

If your metering unit is in good nick, there is not a lot of fuel that comes back from the engine.

The condition of the metering unit has no effect on the rate of fuel return from the metering unit to the tank. This is entirely dependant on the condition of the pump; fuel line pressure & sustainable flow rate!

 

 

Hello Richard - I beg to differ on this point as my understanding is as follows:

The metering unit is the "end of the line" in the pressurised system and it's condition has everything to do with the amount of fuel which is returned to the tank. The pump (hopefully in excess of 110psi) pressurises the system to a point, that after line losses due to friction, the pressure at the MU is around 106psi. Any excess pressure in the system results in fuel being bypassed back to the tank by the PR valve (which is most active when idling and least required at full throttle).

 

The bypass fuel is used by the MU for lubrication. Early MU's had greater tolerances than later ones and bypassed more fuel. They had greater rotor/shuttle clearances and were more prone to wear. Later units had hardened faces, a 'hammerhead' design of the shuttle, hardened cam follower etc. Very subtle design changes were incorporated such as the piston being manufactured from aluminium so so that when the unit was warm, it expanded slightly and leaned off the mix. The later units also had far less tolerances and were fabricated in 'matched sets' with ratings of A,B & C (referring to tolerances). If there is too little tolerance, when the unit is hot, it will sieze. This is the problem now with rebuilding units as everything has to be 'matched' from a variety of used bits to get the best results. Having said all that, you are right that performance is dependant on the condition of the pump; fuel line pressure & sustainable flow rate, particularly if the MU has signs of excessive wear.

 

Peter - by 'piddly', I meant too small, particularly for thirsty PI's!!  The US models had 3 different sized tanks, 9.5gal, 10.25gal and 11.25gals, depending on the year - why they did this is anybody's guess!!  PI's had 11.25 or 10.75, depending on year and market (all quoted in UK gallons)

If not for the overdrive, I would have seriously considered making a larger ally tank for mine.

 

Regards,

Roger

[Richard Crawley]

Hi Roger

The bypass fuel is indeed used for to lubricate the metering unit. But as far as I can see it’s the working clearance between the rotor & the housing (or a knackered ‘O’ ring) that will dictate the fuel bypass rate, leakage past the fixed end stop should be minimal. Given that MU’s were subjected to different design tolerance, the components are machined & require fine tolerances to operate effectively & I think it would have to be a pretty well knackered unit for the bypass fuel return to increase dramatically. If the return rate on any give MU increases appreciably, I would initially be looking elsewhere for the problem, excessive fuel line pressure or possibly a blocked shuttle bleed hole – this will also blow the MU diaphragm off it’s seat.

[jean]

Roger concerning the tank size I have good reasons to believe that all TR6 tanks were of the same size.

For emission control reasons tanks on some models, the tank could not be filled up to max capacity. This was mainly on cars with canister ventilation, or to prevent fuel to spill via the vent tube.

To achieve the reduction the filler neck was positioned deeper into the tank and this made it very hard or impossible to further fill up the tank and to make full use of it's capacity.

Jean

[Peter Douglas Winn]

Roger

I think you should get about 270 to 300 miles on a full tank 11 1/4 gals when driving what I call "fast road style". If you get anything less then the PI system is not setup properly. I get on my old saloon with O/D, which is heavier and more powerfull than a TR a better consumption.

[Andy Moltu]

The full tank isn't entirely usable, particularly if you use a more powerful Bosch pump.

Even with baffles you can find air gets sucked in on left handers with a couple of gallons left in the tank.

This is where a swirl pot would come in handy - giving a litre or so of fuel to feed the pump when air might be sucked in. The pot should also be able to vent the air back to the tank.

The Lucas filter could have been made to do the job to an extent had the outlet been drawn from the bottom of the filter and the top had the capacity to vent back to the tank along with the return from the PRV being fed back to the filter.

You don't get such a problem with the carburettor cars because the pump doesn't shift as much fuel as the PI cars (where much of the fuel is returned by the PRV and the metering unit)

[jean]

Andy, this makes me thinking, if you can call that way...??

Originally on my 6 the PRV return pipe was connected to a plastic T piece with a larger opening on one side and a smaller opening on the other. The larger side was connected to the tank return and the smaller side connected to the filter. Comparing the diameters you could say that 2/3 went back to the tank and 1/3 was returned to the filter.

At one time I removed the T piece and connected the return direct to the filter in the hope to prevent starvation. The result was that the engine died after a few minutes.

Jean

[Roger H]

Interesting point about the emission control, Jean - it would have been impractical to manufacture all different size tanks depending on the model and market etc. I'm surprised that they did not just keep the baffles in the US tanks though - it was obviously a conscious decision to make them without them - or maybe it was the other way round?

 

Peter - economy is something I can't really come to terms with! It maybe the speedo I rebuilt badly (speed OK but odo??) - I always seem to be at the bowser. The MU and injectors are fully recond but set to deliver a touch more fuel than normal under power. Fuel delivery at 'cruise' is to spec.  I blame the engine!