Bias, fuel metering needle

[Richard Negus]

In a previous thread "Slide Needle Removal", the subject diverged into a discussion about bias of the needle, something that I hadn't given much thought to previously.
To confirm which way needles are biased in Norton carburettors, I stripped a NOS carb made by SU around 1989 to confirm.
The needle is clearly biased towards the engine side of the carb, this being accomplished by means of a little 'nib' on the top side of the needle mount and maintained in that position by a spring fitted to the top of the needle.

DSC_0293.JPG

DSC_0295.JPG

The fuel needle itself is stamped AEF

[Interpol2471]

Thanks Richard

[Richard Negus]

I'm hoping someone might give us the definitive reply on which allows the greater fuel flow - central or biased.
C'mon Clive, surely you must know more than a little about fluid dynamics?
R.

[redshiftrotary]

I'm still disappearing down the Internet wormhole on this one but, what I have found out:
The 'biased needle' was introduced by S.U. in the later 70's as it allowed more consistent wear, in an attempt to keep meeting tightening emissions controls in the U.S.A. Seemingly a 'biased needle' wears consistently (hence the rub marks we sometimes find down one side) and doesn't need the fiddle of 'centreing'; whereas with a 'fixed needle' you have to go through the aggravation of centring it and, as it's fixed, it wears inconsistently.
Also a 'crescent moon' shaped Venturi is more efficient than a 'doughnut' shape supposedly.
What I still haven't found out is why some S.U.'s have the needle biased to the engine and some have them biased to the intake!? Burlen can supply bushes with either bias, plus a bush to 'fix' the needle if you wish.
My slide needles are biased to the engine, Richards 'new old stock' are biased to the engine and yet my Police IP2 Workshop Manual says to "ensure that the needle guide gives the needle bias in the direction of the rear of the machine".
A mystery yet to be solved?
John.

[Clive603]

Combining Internet searches with old fashioned books and some hopefully correctly recalled conversations with various folk who I had reason to believe had real knowledge I think the thinking behind SU needle bias is as follows.

As John says it all started with trying to meet emissions requirements. Specifically American ones which, due to the test methods, put a premium on very lean running at low throttle, idling, coast down and start up on the test beds.

Airflow around around needle is, inevitably, turbulent so there are issues getting a consistent and accurately low fuel delivery through the small annular gap between needle and jet at very small throttle openings and power demand. Turbulent flow leads to variable fuel wetting of the manifolds too so actual instantaneous fuel delivery is uncontrolled due to variation between wetting rate and evaporation rate. Especially when cold. Centring the jet on old style SU carburettors isn't an exceptionally accurate process making things even more variable. Old style SU carbs tend to end up set to run close to the stoichiometric air fuel ratio. Good for power. Not so good for emissions but the extra fuel swamps the variations. Which objectively are small anyway.

Biasing the needle effectively enlarges the hole the fuel has to come out of. If nothing else the periphery is a lot shorter for any given flow area than the annulus surrounding a centred needle which must reduce drag and generally make things better behaved. As John says biasing concentrates wear on the touching side of jet and needle the actual metering orifice remaining unworn. Theoretically there is no jet and needle contact with the old style system providing they are properly centred so wear should never be an issue either way. Out in the real world wear is common.

As to direction of bias it appears that forward bias, away from the engine, was the design norm. Seems fairly obvious that having the fuel come out behind the needle gives it a clearer, better behaved, run into the engine than coming out in front of the needle and having to go round it. In some respects I'm a bit surprised that SU didn't go the whole hog by using a D shape needle like most flat slide carburettor designs. Digressing briefly the flat slide carburettor Ron Gardner built and set up for my DB32 Gold Star worked very well indeed. I believe (young'n stupid Clive didn't take a notebook or tape recorder) Ron said that the flat back acted like a Kamm tail on vehicle aerodynamics giving a clean, non turbulent airflow to pull the fuel out. That particular Goldie engine was in full race tune yet would trickle along in top gear at under 500 rpm and had a further 1,000 rpm of useful power at the top end over book. 80,000 miles of commuting convinced me that Ron knew what he was doing.

Forward bias clearly makes most sense as giving unobstructed flow but the bias spring strength is, of necessity, limited. The stronger the spring the greater the rate of wear on jet and needle so you'd want the weakest one that works. However a weak spring can, theoretically, be overcome by high manifold vacuum sucking the needle backwards. Most likely setting up some sort of oscillation which will be bad for both wear and fueling accuracy. I recall, but can't find, references to rearward bias towards the engine being advised for supercharged engines and engines with "high suck". Jaguar installations seem to have used rearward bias and, I think, (some) Rover V8 ones too. Odds are that the flow issues of fuel coming out in front of the needle and having to go round it are minimised if you have plenty of manifold vacuum.

Our rotaries suck pretty hard so rearward bias would seem to make sense. But I'd be unsurprised to discover that the heavily damped slide movement produced by running 40 grade oil in the dashpot changes things by giving a much more stable airflow so forward biased needles work just fine. Maybe just sucked back against the jet pretty much permanently so the spring does nothing!

Time to look at some high mileage jets and needles to see where they wear.

Realistically rotaries are pretty tolerant of air-fuel ratio variations so its probably academic as both can clearly be made to work well.

Clive

[Richard Negus]

My slide needles are biased to the engine, Richards 'new old stock' are biased to the engine and yet my Police IP2 Workshop Manual says to "ensure that the needle guide gives the needle bias in the direction of the rear of the machine".

John.

Thanks for that, John.
The Norton carburettor part numbers/drawings are, I believe, a bit light on detail, merely quoting the SU references FZX1333LH and FZX1334RH. My reference carbs carry an aluminium tag with those numbers stamped in and the date of manufacture, in this case WK 33 92.
Just a guess of course, but Fred's ride home went past the SU factory in Witton so any specification changes may have happened by word of mouth.
Perhaps Bob Rowley can offer some guidance?
R.

[redshiftrotary]

Thank you Clive for a comprehensive and understandable reply; it certainly provides a logic behind the forward or rearward bias for the needles.
Thank you Richard for the numbers quoted - they tally with my S.U. Book from Burlen as the Carb. Spec. Number for Norton Rotaries. The book provides several part numbers specific to the HIF 4 for Rotaries but doesn't specify the needle bush bias alas.
Anyway an interesting discussion and I think I'll leave the bias as just another one of those Norton anomalies!
John.