40 Dcoe Tuning Part 2

[tdafforn]

Hi,
Had very much the same "leaning out at high RPM" problem with a spitfire that I had fitted twin DCOE40s to. Tried all sorts of wacky jet/immulsion tube settings to get it right. Ended up solving it with a 3 step process.
1) bob wieghts on the dizzy had jammed (replaced).
2) Webbers were knackered (bought from friend) (replaced)
3) 200 quid on an afternoon on the dyno with an expert (great fun doing 90+ MPH in a shed on tethers and great to watch a pro set a car up)
Result 35 mpg and a whole lot of umph!
If I was doing it again (and may have to one day with the Dellortos on the +2 as it seemeds to bog a little at low revs, would go to the same guy again, well worth it.
Cheers
Tim

[type26owner]

Okay, so why does the air/fuel emulsion viscosity dominate the carbie? Simple. Because at the optimium viscosity the lowest pressure differential is required to efficiently maintain a steady state flow of the emulsion. What this does is extend the useful range of the venturi system towards the lower rpms. The fact that there is more emulsive air/fuel being being delivered at the point of the highest vacuum causes the vacuum to be extinguished at a faster rate because the vapor is filling the void at the mean vaporization value (maximum) of the bell curve.

This means when the carbie is tuned correctly they quiet down considerably. Because of this fact and the temperature effect from the latent heat behaving as thermometer I can walk up to any working carbie and diagnose in five seconds if it's out of tune and by touching it determine whether it's lean or fat if I've heard a proper tuned one running beforehand.

When I disclosed my theory to my son yesterday I could almost see the light bulb above his head come on brightly. He put two and two together immediately and stated this explains why the muscle cars folks can tune their carbies using a induction manifold vacuum gauge to get the best running engine by adjusting it till the smallest pressure drop is achieved. They been doing it like this for a long time but he has not run across anyone who knows why this actually works.
Keith

[type26owner]

Wow! I reserve the right to be wrong on this one because I'm reporting raw data without replicating it first. It appears the emulsion viscosity effect is much more dominant then even I envisioned. Apparently it's a bell curve with a FWHM of 0.1mm or less of the air corrector jet sizing. I was stepping right over it using 0.2mm increments on multiple tuning passes missing this effect each time because I believed the experts that said the air correctors had an affect primarily at the highest rpms and large steps were okay. THEY'RE ALL WRONG! Here's a link that generally explains what the FWHM is.
<a href='http://www.noao.edu/wiyn/images/fwhm.html' target='_blank'>http://www.noao.edu/wiyn/images/fwhm.html</a>

Got my mixture to remain within 1 lb/hr of air over the entire operating rpm range. Good enough to rival any fuel injection numbers. I won't have to fool around experimenting with any oversized main venturi sets afterall.

Now I can empathize with the ethical dilemma the Weber designer must have experienced. He or She most likely was ORDERED to keep this quiet. Makes me all warm and fuzzy to comprend it like the original designer obviously did.
Keith
p.s. I'm out of here soon so if you don't understand what I've covered ask it NOW. Bye.

[BillGavin]

Keith -

Do I understand correctly that you were seeing the same behavior on both sides of the correct point? Were you able to find the correct point with standard sizes, or were you going to smaller increments with your drills?

- Bill

[type26owner]

Bill,
I need to purchase all the sizes of air corrector jets around the sweet spot to confirm and define the curve dimensions with precision. Applying a leap of reasoning though this explains why the available step size increments is only 0.05mm. The emulsion viscosity bell curve is actually a sharp spike and with any bigger steps you'd miss it. Who'd have thunk it would turn out to be razor-edge adjustment! Have not drilled out any air corrector jets yet. Not expecting it to be required.

Possibly got it wrong were I said the tuning must be done from the bottom up. Have yet to explore the impact this has on the progressive hole tuning segment.

Speculating all of their commercial customer that bought these carbies and installed them knew this info and decided to screw their customers based on the added cost. Weber kept it quiet to not rock the boat. Got to love it when poor tuning practice makes for good business practice. Must have blown their minds to have designed the perfect carbie and not have it applied correctly intentionally. BTDT.
Regards,
Keith

[type26owner]

Hey Bill,
The solution set is a 3-axis graph. Main jets and air correctors makeup the X-Y axis and emulsion viscosity efficiency is the Z. When I said the curve is a spike that's not really correct. It's a X-Y diagonally oriented 3D cresting waveform which has a spiked profile cross-section in the Z axis. It extends across the matrix and comprises all the possible final mixtures at the maximum atomization efficiency due to the emulsion viscosity effect. Somewhere directly below that crest is a single X-Y point which is the desired mixture. Simple stuff.
Regards,
Keith

[steveww]

When you finished with this it would be great to write all this up in one article and get it posted on this site.

[type26owner]

Formula to derive the air/fuel mixture choices all the while maintaining the optimum emulsion viscosity for my twincam. This is close but may change slighty as I purchase all the air correctors in this area and refine the AFM measurement process. You'll get the general idea though.
Air corrector jet/Main jet = optimum emulsion viscosity ratio = 1.1
Main Jet X 1.1 = Air Corrector Jet
110 x 1.1 = 120 I happened to land on the crest here at 11:1 by dumb luck. Feels like I won the lottery!
105 x 1.1 = 115
100 x 1.1 = 110 Guessing this is about right for best power at 13:1.
95 x 1.1 = 105
90 x 1.1 = 100 Should be about right for best fuel mileage at 15:1.
Keith

[type26owner]

Steve,
My wife is presuading me real hard to stop 'fooling around' with the Elan and start reassembling my '55 Aston Martin DB2/4 mk1. That's my goal also and why I'm disappearing real soon. I've documented the Weber tuning process well enough if you just go back to square one and read everything I've posted here you'll be able to replicate the process.
Good Luck,
Keith

[type26owner]

Got this news from Doug at Pierce Manifolds today. Weber is in the Italian equivalent of bankrupcy right now. They are not supplying anymore assembled carbs at the moment and the cost of spare parts is skyrocketing. There's been two 25% cost increases in the last month alone.

Ouch, those air correctors were a tad bit expensive now. Last bits I need are the smaller jets for the cold start choke and I'm done. Going to pull mine apart this weekend and order those jets before they get prohibitively expensive.
Keith

[type26owner]

Hey, just reread my postings and have overlooked describing an extremely interesting effect I've never seen anyone mention before. This will give you a way to judge if you've achieved enough of an overlap of the progressive holes system and the main jet emulsion system from the venturis so there is seamless perfect performance. This should simplify the tuning by reducing the sensitivity (slope error) of the idle jets bigtime to cover an overly large rpm range from having the emulsion viscosity ratio wrong.

The procedure is simple. In fourth (top) gear at progressively decreasing rpms goto WOT instantly and observe if the engine continues running after the intial squirt of fuel from the accelerator pumps which gets consumed in about a second. If it continues to run then the only source of fuel that allows this to happen can only be from the main jet system. This is a direct measurement of the emulsion viscosity effect and it's minimum rpm threshold. At some rpm the airflow through the venturi and the resulting vacuum signal will be so low it can't suck the emulsion from the auxillary venturi anymore. A properly tuned carbie should run okay at 2k rpm or lower. How low I haven't found yet. It could as low as 1.5k rpm at an optimum state of tune. This reduces the scope of the demand on the progressive holes system required to maintain a steady state mixture. This is why I stated that maybe tuning from the bottom up is incorrect. Simple stuff!
Keith

[type26owner]

Another important point I need to touch on is the thermal environment of the engine bay which is required to operate the carbies at their maximum efficiency. Because of the latent heat of vaporization cooling effect of the gasoline enough heat must be supplied to not allow the induction system to be overcooled. If the induction manifold is allowed to cool down so a lower thermal equilibrium is maintained then the vaporization rate will be degraded so it will operate at less then it's optimum potential even though more then enough air/fuel emulsion is present. The main control of this thermal equilibrium is the choice of thermostat installed. That's why awhile back I installed a 195F thermostat prior to fiddling with the carbies. Not using the stock airbox will cause a problem by feed it air which is overly warmed and therefore possibly degrade the combustion process also. Best to supply it cool dense air and have the head portion of the induction manifold provide the necessary source of heating.

If you've got a race engine which has the carbs and induction manifold stuck out in the airstream then you've possibly screwing the pooch. This is also why I don't have any faith in any stationary dyno tuning getting it perfect. It ain't the real thermal environment.

Since Lotus installed way overly fat jets this also explains why the cold start choke is useless with the stock setup. Normally there's not enough heat to provide good vaporization so a choke provides an overly fat mixture which is burned inefficiently but keeps the engine running until the engine has a chance to warmup. With the Lotus stock jetting the affect was to have it burn inefficiently continuously.
Keith

[type26owner]

<a href='http://sel.me.wisc.edu/shedd/carb%20-%20model%20review%20-%2005-15-03.pdf' target='_blank'>http://sel.me.wisc.edu/shedd/carb%20-%20mo...%2005-15-03.pdf</a>
Check out the emulsion mixing possibilities shown on page 14. Obviously my emulsion viscosity theory doesn't match reality all that well. There's a whole branch of science devoted to this effect called 'two phase flow'. Suggest you do a web search to truly appreciate the amount of research done and the applicable scope of this subject.

Bottomline is the selection process of jetting is going to be most likely totally random. It's just a crapshoot. I'm finding out rapidly it's a fact. The implications of this are to find a suitable pairing of main and air corrector you must have all the available sizes of air correctors to swap and suck. So the question of when is close actually close enough becomes a determinant of plain luck or frustration. I wonder if the hard earned experience of a professional tuner really helps all that much? Kind of alarming to discover this general limitation of carburetion at this late point in the process. I'm really peeved none of the tuning books alluded to it at all. At least not that I can recall. Still, it's better to understand the situation then to be ignorant and perpetually baffled.
Keith

[rgh0]

Keith

I think what you are discovering is why the books are so vague on the topic.

The equations governing the air fuel mixturre are very complex as you are observing via your testing. No simple relationship between air flow, main jet and aircorrector jet sizes and resultant air fuel mixture. the 2 phase flow through the emulsion tube system is just to complex to easily model unfortunately add to it the pulsations in the air and fuel systems at changing freqency with engine revs and its almost impossible

The air fuel mixture through the jet system and emulsion tube has 2 main things affecting it, viscosity as you observe and density. As the air amount in the mixture increases the viscosity drops but so does the density. its compressibility also increases as it starts to flow more like a gas than a liquid as the amount of air goes up, all of these changes alter how much fuel gets to the engine and thus the A/F ratio in the cylinders. Finally the size of the air and fuel particles in the mix affects the outcome which is why the different emulsion tube designs which aim to change this parameter via the different hole pattern and sizes. On topic of all these chages you are starting and stopping the flow many times a second and you have all the accelaration dynamic affects of this.

The general guides in the books give what mechaincs with little knowledge of fluid dynamics have observed in practice over the years. I have spent 30 years in the fluid dynamics field and would not try to model it without six months of time and a very big computer with some sophistiated software to help. To try to get an individual engine right in practice requires an educated guess, lots of trial and error and no guarrantee that the outcome you settle on is optimum or a general soluton for all similar engines.

However it sounds like you have got closer than with yours than most people over the years with a twin cam.

Rohan

[type26owner]

Whoa! Calculated I need another $500 of air correctors just to fill in the gaps I've got between the range of 100 to 200. Can buy the drills, fixturing and the tiny number stamps to fabricate mine own for ~$150 and of course a full day's effort of machining. Could just go back and plug in the jetting I found which works perfectly at 11:1 and just live with that. I'm liking none of these options though. Investing the time and effort to install my TJ is looking better and better again. I don't have to overcome any CHAOS EFFECT to do it either.
Keith