Thursday, December 25, 2014
Trying to understand how vibrating engines can “shut off” fuel flow..
This has bugged me for years. Why would fuel, turning partially to foam in a vibrating carburetor float bowl, shut off the pressurized flow of fuel through the needle and seat? The carbs’ floats are buoyed by the weight of the liquid that they displace—so it seems natural that lightweight foam bubbles should allow the floats to drop—allowing pressurized fuel to enter the carbs easily.
Sled racers come to DynoTech Research sometimes to diagnose and try to correct erratic high speed engine operation often caused by fuel turning into foam in vibrating carb float bowls. Foaming carbs—especially on stiff mounted, high revving, slightly imbalanced engines—has been the bane of racers since I built this dyno testing facility 28 years ago.
The SuperFlow engine dyno measures fuel flowing from the tank to the pump, and if that flow is erratic it’s impossible to tune to max HP. Erratic fuel flow can result in diminishing top end HP and seizures. Racers look at dry pistons and white plugs and try to fix that with bigger jets and instead make things worse by running bowls dry even more quickly! But we often fix the erratic fuel flow by 1) softer motor mounts, 2) softer carb boots and weighted carbs, 3) larger needle and seats, 4) rebalancing the rotating assembly, and 5) higher fuel pressure—sometimes as much as 15psi is required to overcome the foam that seems to create an invisible barrier, preventing the stock fuel pumps’ 2-5psi of pressure from refilling the bowls.
Back in 1988 Tim Bender switched to solid motor mounts in his mod Yamaha Exciter 570 FIII oval racer a few days before leaving for Eagle River to keep the vibrating twin’s clutches perfectly aligned. But subsequent testing on Silver Lake (the ice fishermen hated seeing the Bender Racing trailer arrive) revealed erratic, surging top end power and dry pistons that couldn’t be cured by rejetting. So back to the dyno, and sure enough, the fuel flow was erratic or worse due to foaming fuel. Stock motor mounts cured the problem and the Exciter went on to win Eagle River three years in a row with stock motor mounts.
Lectron carbs are used by lots of racers, but seem very susceptible to foaming due to their light weight. An the clear plastic float bowls often allow us to see foam appear—coming and going at different revs on vibrating engines. Several years ago I watched then-Pro Stock Bike dragracer Antron Brown being interviewed on TV after losing a round at the Brainard MN NHRA race. He complained bitterly about his Suzuki (they all ran Lectrons) “surging” and “laying down” in high gear, causing the race loss. Vibes!
The next day I called Fast By Gast (supplier of Lectrons to all of the bike racers) owner Paul Gast and explained to him on a three-way conference call with his right hand man Kevin Gilham about vibration-induced foaming and the issues it causes the sled racers, and what we’ve done to try to rectify that. It had been hot in Brainard—close to 100F and several of the bike dragracers had the same problem in high gear.That discussion led to the design of the excellent high volume, high mass billet aluminum floatbowls for the Lectron carbs.Kevin Gilham, who’s a great guy, would eventually go on to become the owner of Lectron in Texas.
RVP testing and trying to relate to what’s happening in the vibrating float bowls…
Now we know that the hotter the fuel, the more front ends boil out of the fuel. But the RVP test is done with 20% fuel and 80% headspace—volume that must be filled to create pressure by the heated, agitated fuel. But a carburetor has a much different fuel/ headspace ratio—more fuel % and less headspace %. That ratio surely varies a bit from manufacturer to manufacturer. To examine the effect of reduced headspace on heated, agitated fuel, I “Carb Vapor Pressure” tested some Sunoco Maximal that was purchased from a bulk tank that was actually 2psi RVP tested properly. I retested it this time with 80% fuel and 20% headspace—maybe more like a carburetor might have. The result was 4psi—two psi higher than RVP or perhaps double the RVP. So that means that maybe real fresh 5 psi RVP fuel might test at 7 to 10 psi “CVP”! But the carbs have vents that should allow the boiling off front ends to escape—but are those vents adequate? And could the agitation from testicle-numbing vibes at 9000 RPM be more severe than the mild agitation of a RVP test? And could the violent high frequency vibrations be boiling off even more than just the front ends creating a rush of expanding gases that the vents can’t keep up with? If a sled’s fuel pump puts out, say, 3psi all it would take is a pressure rise of a couple of psi to slow down the fuel flow into the carbs. Then if it goes to 4 psi then voila fuel flow = zero lb/hr momentarily until the float bowls begin to run dry of front ends and base fuel, float bowl pressure drops and fuel flow commences once more (hopefully before seizure occurs!).
To examine that possibility, the next time we get a sled with erratic fuel flow and foaming float bowls I’ll connect one vent hose to a pressure transducer to see what might be happening there. If that’s the issue, a band-aid fix might be, simply, opening the vent passages or increasing their numbers. Then, all of those boiled-off molecules can be quickly replaced by the fuel pump, and the engine should be happy even if it’s ingesting primarily heavy-end fuel molecules.
There seems to be a difference in sensitivity to fuel delivery problems from manufacturer to manufacturer regardless of mass. Even some heavy weight aftermarket carbs that make good top end HP are plagued with vibe-sensitivity, even though they use standard Mikuni or Keihin floats and needles and seats. Perhaps insufficient venting?
More testing must be done. Next I’ll try the 80/20 CVP test with my 6psi combo of VPC16 and isopentane to see how high that pressure goes. This may be an incorrect theory, and maybe some engineer can tell me how bubbles can keep the floats floating, and needles and seats pinched shut. But for now anyone who’s plagued by inconsistent fuel delivery might just try adding a few morecarb vent fittings, or enlarge the existing fittings. And, if it cures/ helps the problem make sure you put a DynoTech Research decal on the hood!
Thursday, December 04, 2014
Today I think I got it figured out so we can fairly accurately determine RVP at home (HVP) as suggested in yesterday's entry.
Test 1--First, I refrigerated the complete test bomb for an hour, bringing its surface temperature down to 34 degrees F (measured with my $29 Harbor Freight infrared gun). Then I took the fuel sample chamber and filled it with cold water and immediately connected it to the cold air chamber. Then I submerged it in the circulating 100 F water bath and left it until the air pressure in the bomb quit rising. Result: expanding air pressure in the upper chamber EA 2.25 psi
Test 2--Next I removed the water-filled sample chamber, emptied and dried it, then put it back in the refigerator to chill it down to 34 F (this is to ensure that front ends are not lost from the fuel sample prior to testing) while the warm air chamber was left in the 100 F bath. Once it was chilled I filled I filled the sample chamber with a mixture of stale race gas and isopentane then immediately attached it to the 100 F and conducted an actual RVP test. Result: actual RVP 7.0 psi
Test 3--Finally, I removed the bomb from the bath, removed and emptied and dried the fuel sample chamber and once again chilled it and the upper air chamber down to 34 F, and did a test. A fresh sample of my volatile mixture was poured into that sample chamber and I immediately attached it to the air chamber. Then it was immersed in the 100 F bath and was agitated and heated to 100 F. So this time we had a combination of boiling off front ends and expanding air in the upper chamber. Result: test pressure 8.75 psi
I initially expected the HVP to be RVP + EA, but it was .50 psi lower. Mmm. We can surmise that the expanding air in the top air chamber was acting like nitrogen pressure on the sample--perhaps prevent some loss of front ends compared to an actual RVP test which has zero Expanding Air!
So it looks like the math for HVP testing is starting with a refrigerated homemade bomb 80% cold air and 20% fuel sample is as follows: fuel at as close to 100 F as possible and shaken to agitate = test pressure 8.75 psi - AE (air expansion pressure) of 1.75 psi = HVP 7.0 psi.
Now this is not perfect--for example if your fuel is completely dead, you should come up with a test pressure of 2.25psi so the AE will diminish some as actual RVP increases. But surely close enough to know whether your fuel is volatile or dead. And, if you build your own HVP test bomb, you should do the test once with 20% water to see what your AE pressure is on your gauge which is what counts. Then you can do the math based on your own instrument.
Wednesday, December 03, 2014
Today I had a sled racer from Utica with a sealed pail of C12 and it tested dandily--6-7psi. But last month Tripod Dan's sealed pail of Q16 was nearly dead at 2psi. Then Jesse Helwig bought 5 gallons of Sunoco Maximal from a bulk tank at a retailer--0 psi, totally dead.
It's great that we lucked into this RVP testing equipment so we can test fuel for our dyno tuners and local racers who bring their fuel samples to be tested ($40 charge). But what about the rest of the country? Is your race gas dead, or half dead, or dandily fresh and volatile? Mailing fuel samples here seems expensive and in some cases like USPS illegal. So why can't a person do their own RVP test in the kitchen sink?
OFFICIAL RVP TEST PROCEDURE:
The RVP test "bomb" (yes, that's what they call them!) is a two piece stainless steel chamber with an accurate bourdon tube 0-15psi pressure gauge at the top. The bottom chamber is where the fuel sample goes--then its attached to the top air chamber by a threaded Oring sealed connection. The top air chamber is 4x the volume of the fuel sample chamber.
The water bath that heats the bomb and the air and fuel inside to 100.0 F constantly circulates to provide turbulence that helps heat the bomb chambers and their contents quickly. The procedure is to immerse the top chamber--open at the bottom--and bring that up to 100.0 F. That expands the air inside, causing the excess air to bubble out the open bottom. Meanwhile, the fuel chamber is in the refrigerator being cooled to 32 F before the gas sample is poured into it. This prevents the loss of front ends that would occur if poured into a warm chamber. So with the top chamber at 100.0 F, the fuel sample is poured into the chilled fuel chamber and immediately attached and sealed to the top air chamber. Then the assembled bomb in inverted--allowing the chilled fuel to drizzle into the preheated air chamber, and front ends begin to boil off and pressure rises (as seen in the Jim Czekala DrumPreserve.com RVP YouTube video) some. Then the bomb is immersed in the water bath to bring the whole bomb and sample up to 100.0 F. While it's warming up, the bomb is removed and agitated and immersed again every five minutes or so until the pressure in the bomb quits rising.
If we didn't preheat the top air chamber, and assembled it with a filled fuel sample chamber, the cool air in the top chamber will expand and add to the pressure exerted by the boiling off of the front ends!
RVP TESTING AT HOME:
My idea is if we understand what the RVP test is all about, why cant we take a sealable aluminum bottle of some sort and make a homemade bomb and test your fuel in your kitchen sink? If we have, say, a 16 oz beer or Monster Energy drink can we can punch two holes into the cap (or even the body of the can) and JB weld (1) a 0-200 F meat thermometer with a probe long enough to reach within 1/2" or so of the bottom of the can and (2) a 1/8" hose barb. Then a short piece of 1/8" ID hose can be attached to a 0-15psi boost gauge.
UNOFFICIAL APPROXIMATE RVP TEST:
Let's call this a HVP test (home vapor pressure). Chill the aluminum home-made bomb in the refrigerator until the thermometer says 35 or 40F. Then to determine the effect of expanding air, remove the bomb and immediately pour in 3.2 oz of tap water and seal the bomb (or 20% or 1/5th of the volume). Now hold it totally submerged in a sink full of running hot water until the thermometer registers 100 F, and note the pressure buildup on the gauge. Empty and dry the bomb, and chill it in the refrigerator again. Now, pour in 3.2 oz of your fuel, reseal the bomb and bring it up to 100 F in your sink of hot water, and then shake it to agitate the fuel. I surmise (I'll need to do this same test for real, to be sure) that you should subtract the WVP from the unofficial RVP that you just got.
It might not be a perfect test--the grade 3A precision gauge on our bomb is accurate to .25% or with fuel that tests at 5 psi, it would be + or - .0125 psi. A boost gauge is surely more than that, but close enough to get and idea of the relative volatility of your fuel. And a thermometer for checking roast beef is probably not spot-on, but once again close enough!
And what if your $100 pail of 6psi fuel is really low on HVP? We are experimenting with Isopentane that is available from chemical supply companies. It has a RVP of 20.5psi and you can do the math to figure out how much isopentane is required to bring your dead fuel up to a reasonable level. Some people are currently running isopentane in their fuel and it is working as expected.
I plan to do a HVP test here, and then a real RVP test on the same fuel to see if my math is correct.