Lotus Elan

40 Dcoe Tuning Part 2

PostPost by: type26owner » Sun May 23, 2004 3:21 pm

Steve,
Forgot about the rpm difference between our twincams and a high revving bike. Suspect a fourth tier of jetting is probably a necessity when considering a span of 15k+ rpm for a bike. Guessing no flow domain is that big that it would allow such a huge differential of velocities between the fuel and air and still remain within it's boundary. I'm struggling to find ones which work over only a 4k rpm range and have the right mixture! Actually the struggle part is over since I understand the game mother nature is playing. Now it's just work to be done to hopefully find the preferred jetting combinations or establish the fact that we're screwed Weberwise afterall.
Regards,
Keith
type26owner
Coveted Fifth Gear
Coveted Fifth Gear
 
Posts: 1616
Joined: 18 Sep 2003

PostPost by: type26owner » Sun May 23, 2004 4:17 pm

Hey Steve,
Must confess I'm kinda hoping there isn't any magic jetting combinations for the Weber. I've come up with a very simple way to force the emulsion to stay within the boundary of a flow domain. Already completed a patent search and there's nothing pertaining to this idea. It's an unproven concept only at this time but it's so simple, it's stupid and I'm liking my chances of being successful. This is why awhile back I commented I didn't think the Webers engineers understood two phase flow even today.
Regards,
Keith
type26owner
Coveted Fifth Gear
Coveted Fifth Gear
 
Posts: 1616
Joined: 18 Sep 2003

PostPost by: steveww » Mon May 24, 2004 9:24 am

Keith,

AFAIK bike carbs use constant vacuum as the main jet with an idle jet for smooth tick over and an accelerator pump for throttle response. Not too complex on the face of it.... However there are still a large number of variables. Needle profile, jet size, emulsion hole sizes, idle jet size, pump size/volume none of which have been worked out for the twinc. It would still take a long time on the dyno and a box full of jets/needles to come up with settings that were spot on.

It is always something that amazes me about engines. They will run with such a wide range of settings but the "sweet spot" is so small and difficult to find.

Keep us posted with progress on Weber jetting :)
User avatar
steveww
Coveted Fifth Gear
Coveted Fifth Gear
 
Posts: 1838
Joined: 18 Sep 2003
Location: Northamptonshire, England

PostPost by: type26owner » Mon May 24, 2004 2:38 pm

Have completed fabricating all the missing air correctors I needed from 1.0mm to 2.5mm by 0.05mm increments. Working on the missing the main jets from 1.3mm to 1.7mm and in about four hours will have them completed also. Not looking forward to all the swapping and sucking that's required. Really, really hoping I can screen out most of the the obviously poor performing jetting combos by the visual observation test.

After having driven Weber junk for years now it's fix-it time or move on. I can't stand driving a twincam that won't perform flawlessly and be somewhat greener while doing it anymore.
Keith
type26owner
Coveted Fifth Gear
Coveted Fifth Gear
 
Posts: 1616
Joined: 18 Sep 2003

PostPost by: type26owner » Tue May 25, 2004 4:28 pm

Coming up to the daunting task of evaluating several hundred jetting combinations and doing it in a time effective way because I'm really lazy. The visual criteria is cool but lacks consistency and precision. Been thinking about what attributes I'd like best in a jetting scenario and I remembered that it should respond to the smallest possible pressure gradient vacuum signal from the auxiliary venturi. Bingo, I can easily and accurately test for that in a few seconds with the car being stationary. Just need to temporarily disable the accelerator pumps by propping them up with a spacer off of their lever mechanisms on the throttle spindles. Next raise the idling to say 1500 rpm via the idle speed screw and instantly to go WOT. If it continues to run then it could be a two phase flow domain keeper. In which case that jetting combo is a candidate to be driven later and evaluated fully by the AFM. Should be able to whittle it down to about a dozen candidates in a weekend's bit of effort by this method. :)
Keith
type26owner
Coveted Fifth Gear
Coveted Fifth Gear
 
Posts: 1616
Joined: 18 Sep 2003

PostPost by: gjz30075 » Tue May 25, 2004 6:35 pm

Lazy? Keith, I think you're redefining 'lazy' with this project! :D
Greg Z
45/0243K Sprint
36/5727 pre airflow coupe
User avatar
gjz30075
Coveted Fifth Gear
Coveted Fifth Gear
 
Posts: 2835
Joined: 12 Sep 2003
Location: Roswell, Georgia, USA

PostPost by: type26owner » Sat Jun 05, 2004 3:01 pm

Seems from reading many of the postings that achieving excellent fuel mileage and having the Webers tuned to perfection is a real rarity. Many postings mention one or the other but almost none claim both. I'm seeing a pattern developing which indicates maybe the DCOE design isn't all that great afterall. Also many articles mention the importance of the vacuum signal strength from the venturis but none appear to have actually measured it. Going to do that today at least in a deadheaded state. Hoping it's substantial because I've thought of another possible solution if it is.

Where I'm going with all this is I think choosing an emulsion stack design that uses fixed sized apertures for the main and air jet was a mistake because of the dreaded two phase flow effect. If it were just for a static velocity (steady rpm) situation then it would be fine. What I suspect is happening in the DCOE emulsion stack is at the lower rpms it flows in one state then when at higher rpms it SUDDENLY transitions into another flow state. Suspect that by having a variable sized aperture reacting to the vacuum signal could delay this transition from even happening. However, I cannot find a single scientific study that supports my theory. Just got that feeling about it. If it will then finding the desired mixture and stabilizing it there could be easier to do then the swap and suck with fixed jets procedure we're stuck with now.

To visualize what two phase flow looks like have a look at this link. All the movies are just showing steady state conditions (no velocity changes like a carbie would exhibit) and the ratio of liquid to air is not close to the mostly air emulsion of the DCOE. For the DCOE case you have to imagine that as the rpms go up from say bubbly flow it transitions to slug and as a result the mixture usually leans out. Would have to actually modify a Weber carb body by adding a window to view the flow through the emulsion stack and passageway to know what type of flow it actually is. Who cares as long it provides the correct mixture at all times anyway!
<a href='http://microgravity.grc.nasa.gov/6712/2phase_flow/2phase.html' target='_blank'>http://microgravity.grc.nasa.gov/6712/2pha...low/2phase.html</a>
-Keith
p.s. Especially love the websites that state the Weber is poorly understood and then go on to give expert advice and to demonstate that they don't get it either. :D
type26owner
Coveted Fifth Gear
Coveted Fifth Gear
 
Posts: 1616
Joined: 18 Sep 2003

PostPost by: type26owner » Sat Jun 05, 2004 9:15 pm

Hey Bill,
That pipette idea for measuring the fuel level in the well worked out great. Measured the bottom edge of the passageway leading to the aux venturi to the deck up to where the jets screw in as 23mm. My rear carb fuel level measured in as 24mm and the front at 25mm. Going to raise the front one by 1mm to equal the rear. BTW, you can measure the fuel height with the engine idling.

Best ID for the pipette tubing is around 1-2mm.
-Keith
type26owner
Coveted Fifth Gear
Coveted Fifth Gear
 
Posts: 1616
Joined: 18 Sep 2003

PostPost by: type26owner » Sun Jun 06, 2004 12:27 am

Was able to measure the well aux venturi vacuum signal but with poor precision. Best guess it is in the neighborhood of only 1/4 to 1/2 on an inch of mercury at 6k rpms. My vacuum guage reads to 30 in/hg so it was to coarse for this purpose. I know now it's not enough for me to power a two phase flow compensator mechanism though. Okay, I've still got plan B. :)
-Keith
type26owner
Coveted Fifth Gear
Coveted Fifth Gear
 
Posts: 1616
Joined: 18 Sep 2003

PostPost by: type26owner » Mon Jun 07, 2004 2:51 pm

Bill,
Thank you! That pipette idea is by far the best way to be sure the fuel level is correct. Along with Rohan's simple solution for a quasi rear swaybar by means of an additional bumpstop cushion to reduce the body roll at high cornering loads. You two guys have provided the best two tips I've ever gotten off these lists. Thanks again, much appreciated. :)
Regards,
-Keith
type26owner
Coveted Fifth Gear
Coveted Fifth Gear
 
Posts: 1616
Joined: 18 Sep 2003

PostPost by: type26owner » Tue Jun 08, 2004 4:56 pm

Realized in a few hours of work I can replicate the well(s) of the Weber in a transparent acrylic plastic and still connect it to the aux venturi vacuum signal to run it in a stationary testing mode and probably see near real behavior. With that setup I could possibly see the type of flow for each jetting combo I try realtime. Now I would actually have a use for a rolling road.

It would valuable for testing and understanding the fuel level and behavior of the emulsion stack assembly also.
-Keith
type26owner
Coveted Fifth Gear
Coveted Fifth Gear
 
Posts: 1616
Joined: 18 Sep 2003

PostPost by: BillGavin » Fri Jun 11, 2004 1:03 pm

Hi Keith

I'm glad the pipette trick worked for you - it's simple but effective.

Have you been changing auxillary venturies? I don't have enough for my Elan, but I used to get some useful results swapping them on my racing Mini and my Street/track MG Midget.

I suspect that all the calibrated bits were designed to some standard, but not neccessarily to the most useful standard - e.g. I don't think we really care too much about jet diameters, we care more about area, and change in area. Diameter is very convenient for manufacture and for identifying and modifying in the field.

The glaring exception to this is the emulsion tube - the IDs appear to be completely arbitrary. It seems that there will be one or two ETs that will work for a given engine, largely independent of the state of tune. My guess is that this is where Weber accounts for the two phase flow. I seem to recall a discussion of the ET parameters in a manual, but cannot recall which one. The ID is important, as is the OD of the reduced diameter section, as well as the number, size and pattern of the holes, particularly the vertical distribution. I also guess that the ETs are, as their identification suggests, nearly completely arbitrary, and determined empirically. I have read of manufacturers and racing teams making trips to Italy to have Weber set up the carbs for their applications. I can see the test bed crew starting with a likely ET, then making changes based on dyno results and their experience. If they came up with a new tube that worked, they would give it a suitable ID and add it to their inventory. I have always gotten unexpected results when I've swapped ETs in my engines, apart from the recognized ETs for the engine.

If your acrylic well idea works out, and you can see what the fuel/emulsion level does while the engine is running, you may be able to solder up some holes, and drill others to adjust the mixture curve.
User avatar
BillGavin
Second Gear
Second Gear
 
Posts: 136
Joined: 19 Sep 2003

PostPost by: type26owner » Fri Jun 11, 2004 2:06 pm

Hey Bill,
Have you been changing auxillary venturies? I don't have enough for my Elan, but I used to get some useful results swapping them on my racing Mini and my Street/track MG Midget.

Nope, Mike Ostrov kindly loaned me a set of 33mm ones but I 've not gotten to the point where it makes sense to test them yet. They could give me better breathing at the higher rpms for the trackdays (BTW, I'll be at Thunderhill with my new rollbar installed where I can use all the performance I can get :) ). That comes later after I resolve the mixture issue.

If your acrylic well idea works out, and you can see what the fuel/emulsion level does while the engine is running, you may be able to solder up some holes, and drill others to adjust the mixture curve.

I'm of the opinion now that swapping and sucking of jets is an outmoded technique for tuning. I will do the jetting combo tests but only to see if the engineering solution that Weber used was ever valid. There looks to be alternatives that I'm pondering. Wouldn't it be nice if someone devised a retrofit upgrade kit for the DCOEs that would get the old carbies to nearly pass the current California smog regs. This is not really possible without a fully functioning cat and a ecu though but they can certainly be improved to be greener. The goal is to improve the mileage, performance, drivability and I'm finding the most important improvement would be to change the jetting solution so it's EASIER to tune them.

IIRC, the Weber Tuning Manual downplays the importance of the emulsion tubes saying their influence is most noticed at the lower rpm range of the main jet system. My onetime swapping of ETs seemed to confirm that to be the case. Seems likely this would be true if the the fuel level recedes down the well quickly to the main jet. This is why the transparent well could be the crucial observation I need to fully understand the flow dynamics that are real. If I can comprehend it correctly then the chances of me actually improving it go up exponentially. Fun stuff!
Regards,
-Keith
type26owner
Coveted Fifth Gear
Coveted Fifth Gear
 
Posts: 1616
Joined: 18 Sep 2003

PostPost by: BillGavin » Fri Jun 11, 2004 7:32 pm

I mean the auxilary venturies (aka 'booster'), not the chokes. Swapping them increses or decreases the signal to the jets. The main effect I noticed was to move the main circuit cut-in point up or down the rev range. This was a while ago, so I don't have any data - heck, it was all seat-of-the-pants tuning anyway - but it really helped smooth out the transition to the mains.

My engine has a Shankle head, Dave Bean BLL-14 cams, and a TTR big bore header and exhaust system. It's pretty soft in the lower ranges, but comes alive at 3500 and pulls strong from there. I have 40mm DCOE 18s with 34mm chokes, 4.5 boosters, 125 main, 175 air, 45F9 idle, F16 Emulsion tubes. I think it would like more ignition advance, but I'm at the edge of pinging now. I get 30-31 mpg in a mix of country roads and some highway. I really feel that it could be better, so I'm following your project with great interest.

You're getting like my terrier on a scent :-)

You're right, great fun!

- Bill
User avatar
BillGavin
Second Gear
Second Gear
 
Posts: 136
Joined: 19 Sep 2003

PostPost by: type26owner » Sat Jun 12, 2004 9:56 pm

Bill,
Had one of those adult moments and forgot you stated 'auxiliary venturi' and processed main venturi instead. Doh! In answer, no, I've not tested any other aux venturis. Don't have any and Pierce Manifolds only lists the 3.5mm as the only other size available.

Just completed some more vacuum signal tests of the auxiliary venturi and am presently baffled and confused. Made up a simple water manometer out of clear plastic tubing. Deadheading an emulsion stack gave a total vacuum reading of 1" of water. If the rpms were raised slowly then the signal only started rising at 3k+ rpms until it was 1" at ~6k. If the throttle is blipped then the signal rises immediately well below 3k rpms but still only reaches 1" at 6k. Clearly the system is exhibiting impedance with a strong coupling to rapidly opening of the throttles.

Perhaps there is a major contribution from savaging or regression of the combustion gases so a deadhead test might be misleading. Made up a small manifold so the vacuum signal could be measured downstream of the air corrector in a fully functioning emulsion stack so the cylinder was firing this time but I could measure the pressure drop across the air corrector and main jet realtime. Slowly raising the rpms resulted in NO observable signal no matter what the rpms. Blipping the throttle did register .1" to .2" rise momentarily but it fell back to zero after about a second. The only reasonable explanation which can allow this to work is the head pressure of the fuel level must dominate the fuel level in the well. The signal can't possibly lift the fuel more then a couple of millimeters to the passageway at it's highest value. The highest value being smaller then I can measure at all. It's so low the surface tension properties of the water in the manometer is the limit of the resolution. Now that I think about it maybe gasoline would have been a better choice to fill the manometer. Even factoring in the specific gravity of gasoline at .75 does not greatingly impact this situation.

I'm astounded that this stuff works at all at this point. I had no idea the potential energy was so low. The need for a transparent well to observe a functioning emulsion stack is now extreme. What in the heck is going on here? Reality is not obeying my best guess at the magnitude of the forces at play here. Hey, learning is good though and I'm keeping an open mind set at all times and not assuming anything or believing anybody. Extremely fun stuff! :D
Regards,
-Keith
p.s. Remembered I can but a load on the running drivetrain if I elevate the rear wheels off the floor and apply the brakes at WOT momentarily. That's the poor mans dyno! :(
type26owner
Coveted Fifth Gear
Coveted Fifth Gear
 
Posts: 1616
Joined: 18 Sep 2003
PreviousNext

Total Online:

Users browsing this forum: No registered users and 4 guests