E85 850 cfm Carburetor Calibration...

[shrinker]

Consideration of an Ethanol carburetor.
The current designs of petrol fixed venturi carburetors utilizing a booster use an established method of emulsion to correct the AFR delivery. Air Emulsion is commonly claimed as being necessary to correct the interaction of the velocity gradient of the main venturi and the positioning of the booster venturi. While this is a valid reason it’s not the sole reason. It is worth noting that not all fixed venturi carburetors have the booster centrally mounted.
The dynamic viscosity of petrol and methanol is low relative to the density of the fuel. Ethanol on the other hand has a high dynamic viscosity for nearly the same density as petrol or methanol.
When a liquid flows through a tube like a drilling in a carby body, the flow efficiency changes depending mainly upon the factors of viscosity, velocity and density. There is a special number called Reynolds Number that is a dimensionless number that is used to determine if the flow will be one of 3 types of flow. When the Reynolds number is calculated and the result is below 2300 the molecules of the fluid travel in an orderly fashion like an army marching on parade, the molecules close to the wall are nearly moving in line with the central ones. When the Reynolds number is between 2300 and 4000 the flow is called transitional, the army is a bit disorganized, but still recognizable as a parade. When the Reynolds number is above 4000 the flow will be turbulent, there is no organization of the parade its just a riot with people running everywhere going in circles and generally moving towards the destination. Because the Reynolds number is a dimensionless number it can be used to predict the onset of turbulent flow for any fluid, which means any gas or liquid. When the flow is turbulent the drag increases dramatically.
The density of most fuels is similar and those density variations don’t change the Reynolds number much. Reducing the density reduces the Reynolds number.
The main factors that affect the Reynolds number are the velocity of the flow and the viscosity of the fluid. Because Ethanol is similar density to other fuels the only thing that greatly alters flow velocity required is the stoichiometric ratio variance of the fuels. There are other things that influence the volume of fuel needed by an engine but generally the difference from ethanol to petrol results in a flow velocity increase of only around 30 to 39%. But it’s the viscosity difference that has the main influence on the Reynolds number. The Reynolds number goes down when the viscosity goes up and the relationship is one to one. So ethanol being sort of double petrol will have a Reynolds number of half petrol for the same velocity of flow. But ethanol needs to pass more volume in order to achieve the same Lambda so the difference in Reynolds number finishes up like this;
For a 4mm(.160”) tube (say an example single main well drilling) passing 100hp of ethanol with a BSFC of 0.669 it’s R=2498, clearly only just entering transitional flow.
That same tube with petrol to make the same power and BSFC of 0.41 it’s R=3866
From that we can see that Petrol tube is nearly turbulent and the ethanol one isn’t close.
If we look at a smaller diameter petrol tube say 0.120” then R=5100. To get the smaller petrol tube example down to R=4000 it has to be limited to 77hp of fuel, that’s just a 308hp engine for a 4 barrel carby. So the petrol small tube is turbulent above 77hp per barrel, the larger petrol main well is turbulent above nearly 100 hp per barrel and the ethanol one isn’t turbulent at all. In fact the ethanol one has to go to 160hp per barrel to start to become turbulent.
Now why we need to know this stuff is because once it becomes turbulent the force (read pressure differential between one end of the tube and the other) needed to move the fluid increases non linearly, in other words its starts to lean out as the CFM goes up.
So what do we do with this information, we use it to decide if we are going to build the carby to be in the laminar world or the turbulent world. If it’s a turbulent flow carby then it needs to have emulsion to reduce the density in the correct relationship to combat the non linear flow resistance. You can see now that a petrol carby is a turbulent flow carby on any mild hotrodders engine.
If it’s a laminar carby then we don’t use emulsion for that reason. Of course the vacuum generation of the booster and its relationship with main venturi may have a need to use emulsion correction, but that’s another reason.
If the carby crosses from laminar to turbulent flow at a point in the fuel demand of the engine then ideally you have to match the emulsion to the same point, that’s not easy to do because fuel level variations alter the pressure upon the main jet entrance, thus altering the flow rate into the main well. If the liquid inflow is reduced the emulsion system is able to supply more air, reducing the density and thus lowering the Reynolds number. So when you change a main jet size or a fuel pressure or float level you better think about a few other things sometimes. That’s why I say there is a range of successful operation for a main jet to go with a main well and booster etc. Also what are you going to do with the selection of a power valve opening point, is that opening at a point where the flow type is changing or not? Hotrodders don’t generally think about these issues but when presented with a special task these are just some of the things one has to consider.

[Slowpoke]

Shrinker,

Thanks for the info, it is very enlightening.

That's why a lot of Hotrodders use the "trial and error" method to establish their own data base for their combo and "their opinion on why it works". That's how I work, although I am collecting a lot more data and asking a lot more questions of people that are a lot smarter than I am, all in the interest of understanding "why".

Obviously Mark and Eric are very successful with their E85 carbs and if I was going to change to E85 and buy a carb, they would be at the top of my list. I'm sure this discussion will help all of us.

Mark, on a long post, when I am through, I highlight it and copy it before I submit it. Then if it times out, I can paste it in the new window.

Thanks again,

[Ken0069]

RE: Post time out.

If I'm going to write a long one, I usually compose it in word or word pad before I even open a reply window. That way I don't have to worry about the time out and since I copy and paste it into the reply, if something goes wrong then I've got a copy still in that open word document. If that post is successful then I just delete the document.

Spell checks? I use Opera for my browser. Others use FireFox. Both these browsers have "spell check" built into them so when you copy and paste into the reply window any miss spelled words will be underlined in red. MS Internet Explorer does NOT have a built in spell check.

Office Suite. If you don't have an Office Suite because of the cost, then go to Openoffice.org and download/install Open Office. It works just as well as Micro$oft Office and doesn't cost you a dime! It also has a real time spell check too along with a spread sheet and power point programs. Oh, and did I mention that it's FREE?

[Drag Chevette]

shrinker,
Excellent post.....


especially this point:

If it’s a laminar carby then we don’t use emulsion for that reason. Of course the vacuum generation of the booster and its relationship with main venturi may have a need to use emulsion correction, but that’s another reason. If the carby crosses from laminar to turbulent flow at a point in the fuel demand of the engine then ideally you have to match the emulsion to the same point, that’s not easy to do because fuel level variations alter the pressure upon the main jet entrance, thus altering the flow rate into the main well. If the liquid inflow is reduced the emulsion system is able to supply more air, reducing the density and thus lowering the Reynolds number. So when you change a main jet size or a fuel pressure or float level you better think about a few other things sometimes. That’s why I say there is a range of successful operation for a main jet to go with a main well and booster etc. Also what are you going to do with the selection of a power valve opening point, is that opening at a point where the flow type is changing or not? Hotrodders don’t generally think about these issues but when presented with a special task these are just some of the things one has to consider.

Eric and I have found a way to add some emulsion and keep the laminar flow consistant and the density/viscosity from going turbulant....
I believe this is working to our advantage when it comes to making a consistant running carburetor.

cant tell what it is, as this is just a small part of whats setting us apart from the other builders.

[Eric68]

Ok I gotta ask. What is the infatuation with E85 anyway?

I have never even seen or smelled the stuff but IMO fuel is such a small part of the racing expense why all the trouble?

Plus you mid western guys are the only ones that can even get your hands on it.

For me it's three things.

1) It works as good (or better) than 110 race gas and only costs $2.50 a gallon.

2) Its legal to run on the street -- unlike leaded race gas.

3) I'm obsessive compulsive by nature.

[Eric68]

I agree with Mark -- excellent post Shrinker.

Helps keep guys like me humble and wanting to learn more.

Also Mark -- I never said I couldn't get a stainless N&S to seal. I just happen to like the Viton, they seem to work well for me and hold up to the ethanol just fine. It's the big .150" stainless N&S that don't seem to seal worth a poo for me.

[shrinker]

There is good sense in running Ethanol, but E85 would probably be a bit different because when you mix ethanol with gasoline some other issues with the 2 fuels separating back into gasoline and ethanol come into it and that can create detonation of the gasoline component if the engine is optimized for the soluble state of the 2 fuels. But with care of storage, (keeping water out) and not operating in cold conditions thats all taken care of.
My preference is for straight Ethanol as the fuel curve should be easier to control because of the viscosity of Ethanol. However I think E85 is really a street fuel, its not always the same mix percentage and that would alter the viscosity unpredictably. Here in Australia the V8 supercar class runs E85 but its not the same fuel as sold at the pumps to the street. It is sold in drums and is specified for the class, it is consistent mix and the gasoline component isnt regular unleaded like the street version.

The viscosity problems caused by temperature with ethanol and mix percentages with E85, are probably the key issues to get around.
How it all reacts as a final production of power from the engine is not easy to predict or explain. Sometimes a combination can benefit from leaning out the high rpm only on hot days and you can get that effect if the fuel is hotter and goes into turbulent flow earlier because of that. Remember that the viscosity decreases with increasing temperature (for liquids that is, gases are opposite) so the Reynolds number goes up and you reach the turbulent flow point earlier in the power range of the engine. Carburetors can work in any flow state, its not important what it is its just important to stay within the one state. Obviously thats not going to happen for a turbulent flow carby like a petrol Holley because it has go through the different flow states to get to turbulent and once there, it will have stable conditions of pressure differential to fuel rate. Thats why the other design principles used by some other carburetors in the world are more efficient at fuel utilization.

[Eric68]

How most of us are getting around the variability in the E85 blend here is by testing it frequently. For the most part the blend doesn't change much down south where the weather stays warmer year round. Up north here (probably like you guys in Adelade) there is more gasoline content added for cold weather driveability. It doesn't seem to change much or vary often during race/cruise season, but we do have to watch it.

What the street strip guys do often up here is fill up a 55 gallon drum of summer blend E85 and keep that handy for racing. The serious race guys that are pushing the limits of the fuel get it by the drum in a "spec" version that is a little more optimized for racing.

The part many people forget about is that pump gasoline varies widely geographically and seasonally too. This applies to the viscosity as well. In the US pump gas can contain up to 10% ethanol, and it can contain all kinds of other "junk" (MTBE, ETBE, MMT, Butanol, Toulene, Benzene, etc.) the suppliers use to manipulate the emissions and vapor pressure of the fuel. Besides the "extras" gasoline itself consists of many different carbon chains each with their own density and viscosity. Now ethanol is ethanol -- no matter where it comes from or how it is processed, pure ethanol is C2H6O . . . always!

So my point is that other than the varying % of gasoline there is LESS variability with E85 than there is with gasoline. In fact, the most variable part of E85 is the gasoline part! So while E85 does vary in viscosity with temperature so does gasoline.

In practice, I find that pump E85 here in Michigan is consistent at the drag strip all season -- more so than gasoline. In fact, the gasoline guys often vary a tenth or two during an evening of racing (hot in the day, cools off a lot in the later rounds) and I vary less on E85. Part of that may be due to the cooling effect and being able to keep the temps consistent round to round I don't know.

OK, enough rambling for now!

[jmarkaudio]

Thats why the other design principles used by some other carburetors in the world are more efficient at fuel utilization.

How about some examples. And if that is the case, why have there not been more in the racing world, has Holley had that big of a lock on the racing community?

[Drag Chevette]

Also Mark -- I never said I couldn't get a stainless N&S to seal. I just happen to like the Viton, they seem to work well for me and hold up to the ethanol just fine. It's the big .150" stainless N&S that don't seem to seal worth a poo for me.

Oh, my bad...misunderstood... , it was only a two minute conversation....

The viscosity problems caused by temperature with ethanol and mix percentages with E85, are probably the key issues to get around.

I believe I found a way around this in most engines, some of the higher compression (14:1+)engines have to stay with the 85% blend.

but as Eric said the fuel must be tested.

Obviously Mark and Eric are very successful with their E85 carbs and if I was going to change to E85 and buy a carb, they would be at the top of my list.

Thank you very much, sometimes its hard to maintain your confedence in what your doing in this business.

a great carb builder told me just yesterday, "dont start second guessing yourself, you have something thats working pretty darn good." (or something to the effect)

its comments like yours and his that keep me motivated and moving forward, well that and the fact that your going to tell your friends about me..

[jmarkaudio]

You talk about the Reynolds number with a .160 passage being near turbulent at 100 HP. My Dominator has 4 -.160 booster pins and the engine makes almost 850 HP on racing gas. So you are saying that my Dominator is passing fuel in the turbulent range pretty much all the time at WOT? How big does it need to be to prevent that and how do you get fuel to flow at a much larger size? Or is that where emulsion plays into the mix? I can see where making sure passages make a smooth transition, although smoothing the mainwell to angle channel could be tough to work on. And if you make the mainwell to booster passage too large, the signal needed to start fuel flow gets too high and leave a hole in fuel delivery. And last, you talk about the range between a main jet and booster, any insight into this? .128's with a 1250 Dominator and same .160 angle channels and booster pins.

[shrinker]

Mark, dont be scared of turbulent flow, its not that the flow stops its just that the relationship of head to flow rate changes so that its not equal when its turbulent. When flow is laminar and you double the depression you double the flow but when the flow is turbulent you double the depression and the flow doesnt quite double. Because Laminar flow is difficult to achieve its not practical to deal with. With turbulent flow the ratio of depression to delivery is a curve not a straight line, remember that for a bit, now look at the booster to main venturi relationship. Because the booster venturi passes air and liquid so I dont know the actual viscosity of that but I assume it is in turbulent flow basically at all times. Its difficult to work out the actual answer for air as its compressible and it varies humidity etc. Let alone what it does in the wet booster, but basically the turbulent flow of the liquid is actually used to advantage to balance the turbulent flow of the venturi. Thats what sizing of the main well does.
In your example the fuel in the main well is turbulent and to get it non turbulent it requires a main well of 15mm (.590”) diameter to get the upper limit of laminar flow. If you tried that you would need that feed to the booster too, there’s no point having laminar on the main well and then turbulent in the booster feed and the main jet is going to be turbulent so WTF.
It’s the viscosity of gasoline that causes the design to be the way it is, but Ethanol has the advantage that its viscosity allows it to be laminar within the current designs of Holley style carbys. You simply cant do it with gasoline.
Beside those issues the main jet is going to be turbulent so therefore it will have a curved flow rate so you have to use emulsion to correct that and that’s why Mark only has a little bit of emulsion. He basically only has to correct for the main jet. The main jet is only short and little and it doesn’t take much correcting but to do that on gasoline is not practical. You could do it but the size of the main well to get laminar on gasoline is huge and the .080” main jet would take a long time to fill it up to respond to the depression of the booster as the load on the engine is increased from idle. No amount of emulsion is going to draw the fuel into the engine correctly with that setup. That’s what I meant with regard to the balance of mains jet, well size, booster feed etc.

[jmarkaudio]

After reading through this again, I saw something that stuck out a little more. I know Mark and Eric work the boosters for the carbs they set up, and I may be wrong but figure more of the carbs they do are 4150 versions. What stuck out was the booster being a point that may cause the fuel to go turbulent with E85. Would there possibly be an advantage using a straight leg booster over a downleg with the 4150, but even more so would a straight leg be more beneficial on a Dominator over an annular booster. A straight leg booster would not have the channel disruption in the booster that the fuel would have to navigate in an annular booster.

[shrinker]

What you are chasing is a flow change at a bend. But your chasing it after Emulsion. Emulsifying the liquid reduces the liquid viscosity. Reducing viscosity make a flow rate become turbulent earlier not later. Emulsifying the ethanol makes it closer to the viscosity trends of gasoline. The lower the viscosity the more turbulence going round a bend. So the answer is to do what Mark is already doing, use minimal emulsion.
It seems that other Ethanol carby producers emulsify the hell out of the main well, that's OK too provided you get it to turbulent flow at a point low enough in the horsepower supply that it can be consistent up the range.
There is probably much more knowledge that Mark and Eric have accumulated about boosters working with Ethanol than what I know, but that is their advantage. They have concentrated on that field to understand the interactions and good on them for doing so.
To measure the effect you are asking about would require quite an investment and for what? Ethanol is not turbulent until high enough power levels for a single carby motor. All that just stepping into turbulent flow does is it slightly alters the top end a little bit, it just leans it out. All one has to do is keep an eye on the AFR's and its easy, I suspect that the small change by going turbulent with ethanol would be difficult to detect unless you had a test engine on a stand. Differences from one run to the next would probably negate any such testing on the track. But Ive never ventured into Ethanol myself so I don't properly know those details.
But if your going to pursue thoughts about booster bends then you should include all the turns in the metering path.

I missed answering one of your questions about other designs that are better than a Holley at fuel utilization a few posts back-- If you look at the design of a Weber emulsion tube (its similar to Japanese motorcycle carbys etc) the air is introduced all around the circumference into the outer well. The air isnt stacked up on top of one another like a Holley well is designed, also the outer well diameter interface to the inner air tube results in close proximity and increased percentage of boundary layer flow. Boundary layer flow is reluctant to turbulence. Stromberg's OEM carbys have those principals, Carter thermo's and other carbys use drill sizing (with no emulsion basically) and double or triple booster stages to control the fuel rate. All these other designs are far more expensive to produce and are really only suitable for a mass produced engine. To use these other carbys when hotrodding you have to know a lot more about the physics of carburetion, Holleys are simple, change main jets and turn idle screws.

Years ago you went to the speed shop to buy a Holley for your hotrod and the guy hopefully knew enough to sell you the right model Holley, thats why they make something like 450 different models. Today they sell you a HP ultra and expect you to get it right yourself, Its this self-serve mentality we have nowadays. Trouble is the person buying the HP doesnt know much about how to tune all the tunable parts.

[jmarkaudio]

The flow change was my point with the boosters, the angled passage with a downleg versus a straight leg in a 4150, on the dominator it is the change where the fuel has to flow around the inner ring of the annular booster. This would be to maintain laminar mainwell flow for E85 until the point the fuel is pulled from the booster. And I do see the point on the entire path, using jets similar to BLP jets that minimize turbulent flow, smoothing all accessible passage bends, in the mainwell, etc...

In my phone conversations with Tuner, we had discussed the benefits of tube emulsion used in other carbs and in a few of the older Holleys over the emulsion well in current Holleys. Wouldn't this be easy enough to incorporate into a billet metering block for a Holley, as long as the mainwell was sized large enough to not be restricted by the emulsion tube?

And last, all of this still goes back to the issue of proper atomization to allow a more complete combustion on any given engine. I know in some cases it means sizing the carb down to maintain air velocity. I had talked about this to a friend that had a 550+" BBC with large heads, they found best power at the track using an alcohol 1050. Any bigger and it would slow down. Both the 447" and 461" SB2 engines we built have run faster every time we use a larger carb, both showed more power from below peak torque all the way past peak HP and did not fall off as fast with the larger carbs. Both engines do have a large stroke and very efficient heads, the ability for them to maintain higher velocities in the induction system I think is the reason the larger carbs are able to perform better. In the case of the BBC, I believe it had either the wrong cam that allowed air velocity to slow too much or the ports were too large for the RPM range they were trying to make power in or both. The larger carb in that case allowed fuel to fall out of suspension at lower RPM's, a critical area for a bracket car especially with an automatic to make it's ET.