what issue does that cause because I'm not seeing it at 700+awhp

what trouble? the cars come stock with adequate lines for most people

 

You get a 5-6psi pressure wave that oscillates through out your fuel system. When the wave matches the frequency of the injectors firing (similar to Helmholtz resonance on exhaust pulse tuning) You can have injectors firing with up to 6psi less fuel pressure than the other injectors.

The solution to fix this is dampening the fuel either by using a dampner or by increasing the volume of the rail/lines. The issue with the increased diameter fuel lines if you get more heat rejection to the fuel... same reason why you see OEM's going to returnless fuel systems to reduce the amount of evaporation within the fuel tank from the increased fuel temperature generated by the return system.

Like I said... just because you can't see it doesn't mean it's not happening. I doubt any of you are running sensors that can sample fast enough to see the phenomenon. Mind you... you guys probably aren't running individual EGT probes...and you're getting your AFR's from your downpipe after each cylinder has mixed into a common tube. You're getting an average from the 4 cylinders...

You may see an oscillation in your power from your standing fuel pressure waves. If you inject at a lower pressure your atomization will suffer and so will your power. It can explain why your dyno graph looks like you need more ignition timing, yet you're already at MBT...

Dave B... I KNOW I've seen your dyno charts do this...in fact I asked you in a thread about why your dyno graph was waving on E85/E98, remember?

 

As far as I know Buschur is the only company that modifies the Walboro in-tank to be able to run more fuel line pressure.
Do you need this modification? How much fuel pressure will you run, as the relief valves open in the 85psi range at the pump which lowers your flow rate on the stock Walboro.
-James

 

It's actually a 53% increase in flow area (proportional to flow capacity), which is the important parameter.

The pressure oscillation is interesting. Kind of funny because just about every modern fuel system has a fuel damper. The EVO fuel rail is designed to be the damper, as it has some flexibility to it.

I've always wondered why none of these aftermarket FPR companies don't incorporate some kind of damper into their regulators or design an aftermarket add on damper.

At the very least, companies making fuel rails could incorperate a Honda damper or something.

 

R/T, not to argue but I think you are over thinking this a bit. I don't need a formula or scale to see if the stock fuel lines have corrosion and such in them. The factory rubber lines are soft, pliable and are not falling apart. I've seen rubber that is effected by E85, it swells up horrible and tears apart.

I have a 5 bar map sensor in my fuel rail, I log fuel pressure, I can tell you the fuel pressure is very stable. I don't think there is any disputing 600+ on the stock lines seeing as even on our dyno we've made 730+ on the single stock line.

Not being "worth the trouble to run small diameter lines" doesn't make sense, there is no "trouble" involved, they are on the car and have proven dead reliable for 700+ whp.

 

R/T whatever it is you've seen in my dyno charts have not been from a fuel delivery problem. Find the dyno sheet and post it or give me a link.

 

I'm not one to argue, but my company spent "lots of monies" to look at this particular issue... Kistler's piezoelectric pressure transducers are like the Ferrari of sensors and you have a Kia.

http://www.kistler.com/pressure-transducer_en

Granted we inject at much higher pressures (2500 bar), but the physics still apply. You're not logging fuel pressure at a fast enough rate. What are you logging at 100Hz? What you're observing with your sensor is an aliased signal.

On the Q16 issue...
You cannot squeeze a rubber hose and determine if it's been affected by the fuel you've run in your tank. That is my point. It's like a chemical engineer revving an engine and then telling you...LOOK...it makes power. I took some plastics and composite engineering classes while at WWU and one of the way to determine if the rubber was affected by the chemical it was submersed in is to weigh it before and after and see if it has gained weight or lost weight. That's really not getting that scientific... but its an honest laugh to see you squeeze it and say "look its good." lol. Not hating, it's just funny.

More on the Q16 note...it was Paul's tank that I saw the corrosion with. I've also been privy to see the corrosion on my engine parts due to running Trick 114L It's rusted in the inside of my ss o2 housing and downpipe. Only while running trick. C16 doesn't do this, pump gas doesn't do this, 100 octane doesn't do this, 94 cenex doesn't do this... only the Trick 114. It has a LOT of MTBE. Methyl Tert-Butyl Ether. It's an oxygenate additive used in Q16... and just happens to be a great solvent.

03whiteEvo,
Aftermarket fuel pressure regulator companies DO make dampers.
http://www.injector.com/cart/pc/view...H&idCategory=4

 

One distinct advantage of a larger fuel line is it decreases pressure losses due to turbulence inside the line. Maxing out a set of 1000cc injectors on E85 creates a pressure loss of 5psi or so with the factory lines. Maxing a set of 1450s creates a loss of over 9psi. This means the pumps are working 9psi harder than whatever the rail gauge reads, and that becomes significant when the power levels go up. Change the line to an AN-8 (1/2" ID), and the losses drop to virtually nothing. A quicik, easy way to estimate those losses can be found with this simple calculator:

Fuel Line Sizing Calculator

As far as pressure dampers go, I see them being more important to engines running batch fired injectors. In those setups, all injectors fire at the same time,with each engine revolution, and that creates larger pressure waves within the system. Off the top of my head, I can recall every OE batch injected car I've owned had a damper in the line, but not the sequentially injected ones. YMMV

 

Laminar Flow in a tube... Reynolds number... hey! I did a physics project on that in college.

--------------------------------------------------------------------------------

"C23 is good stuff. A little expensive and probably overkill for 85% of the Honda guys, but extra insurance if you want it. I run it on my car for a little extra piece of mind. 50+psi on gas, a little extra octane never hurts. People are quick to forget about C16, it's an awesome fuel and was 'the' fuel to have for years. As i've stated before, i've run C16 and Q16 back to back on my car at the track and saw no difference in ET or MPH once the a/f ratio was dialed in for each fuel. Q16 is highly corrosive and hard on engine parts because of it. VP now recommends that it's drained from the fuel system when not in use. Q16 is a good fuel, don't get me wrong, but it's 'high' maintenance and there are plenty of other options that make the same power without the downsides."

-Tony Palo
from...
http://www.honda-tech.com/showthread.php?t=2723106

 

Ah, very nice piece R/T Ernie, particularly since it has a Schrader valve for fuel checks at events.

David, sorry to go further off topic, but I'm wondering if maybe modifying the stock fuel rail might actually be a better solution than replacing it. From what I recall, you found the stock fuel rail to be an issue around the 600HP range? The stock rail has a rather small inlet and outlet, I imagine this is to induce a pressure differential in an effort to control pulse amplitude.

Out of curiosity with respect to pulse damping, I wanted to do a quick comparison of fuel rail flex. The stock fuel rail is designed to flex to dampen out fuel pulses. I compared roughly a rectangle like shape found on the stock rail and a thin wall round and square tube. Interestingly, the factory shape provides about 1000 times more pulse damping flexibility than a 0.02" thick 0.75" OD round tube and 50 times more flexibility than square tube. I'm sure it's several more orders of magnitude "softer" than the thick wall aluminum fuel rails typically used.

I suppose my question is, would opening up the inlets and outlets provide the needed fuel flow capacity but still offer better damping characteristics than the billet rails? Also, the factory rail has radius inlets to the injectors, which while a minor detail, would provide gains in injector flow in theory.

Does the EVO run batch or sequential at WOT? It's probably a minor issue anyway as IDC is usually high enough that there is constantly a draw on the fuel and they are likely causing significant pressure pulses.

Anyway, sorry, back to your regularly scheduled program.

 

David,
thanks for the great videos, very good stuff.

I've got a few questions about running ethanol that i think you'd be the man to answer.

1) Pump electric connectors,
afaik. gas does not conduct electricity but ethanol does. Have you protected the pumps electrical connectors somehow or just running the stock ones?

2) Valve seats and other bits not from alu or stainless,
Ethanol is tougher stuff on some materials, for example brass. What are the evo valve seats made off, and have you checked the seats or other possible non-stainless components in the chain for corrosion?

I've been thinking about switching to E if it ever gets available in this goverment-gas-monopolized arctic ****hole of a country.

 

You college boys talk way over my head, there's no hiding it. I'll make you laugh harder though R/T when I do my next video and show you what a fuel line that won't hold up looks like when put in E85 or Import, how about that? I've run the "wrong" fuel line and within 24 hours the hose starts to detiorate and swell. The hose is not going to "gain weight" if it doesn't swell as it has to take on the extra volume of the fuel somewhere or so my un-college educated guess would say so. So in the case of me squezing my hose for all the world to see was to show that the hose is not swollen, not falling apart and not having problems.

I did see Pauls' pictures. That tin can probably had water in the tank, it was a cobbled mess. His pictures were on my mind when I've inspected mine over the last few years, always to find nothing wrong with my tank/lines.

Alright, R/T, I took the time to pull up my logs of the fuel pressure. I've sent them in with a question to AEM to get the sample rate information just for the hell of it. I'm open minded enough to atleast discuss something as there's always something to be learned.

I wanted to compare the boost signal to the fuel pressure signal from some logs I have but then just as I was going to agree with you I realized that the boost signal channel has a smoothing feature on it used to get rid of some of the peaks/valleys in the signal. So I can't compare the two signals.

Looking at the fuel pressure though, the pressure doesn't vary more than 2.5 psi in the peaks and valleys and that is on the high side of what the log shows. I can live with that.

 

Just bear in mind that anytime you're going to make gobs of power (e.g. 600+) with E85, a larger fuel line gets you more pump capacity, more consistent pressure, and the pumps work less hard.

As far as oxidation due to Q16, VP Import, etc., it seems these fuels may be killing some ID2000 injectors, which are claimed to be the only ones in the series with some internal rubber parts. I can't verify this, but it's something worth researching. If you're going to upgrade your fuel system and want to run these fuels, Teflon fuel lines would be the best option (which I have in my EVO now). You can pump sulfuric acid through those without worry.

Since some report corrosion in the tank and others not, it may be that some are using parts in the fuel pump assembly that are susceptible to corrosion with these fuels.

AFAIK it is full-time sequential, but if you know something different, I'd like to know.

 

I think the issue R/T is pointing out with signal aliasing is that you might not be picking up the important feature of the signal. If you are sampling at 100Hz, the fastest signal you can pick up is 50Hz, which is about 3000 RPM. Anything above 3000 RPM, you are not getting an accurate sample for each revolution. Then when you figure 2 injector pulses per revolution, you really can only get a rough idea of the pressure waves below 1500 RPM. Anything above 1500 RPM and you aren’t catching each injector firing at all.

If you wanted to get a more realistic picture of the dynamic pressure at 8000 RPM, you'd need to be sampling around 1000 Hz. Then you need to have a sensor that has a response time fast enough to even react at that frequency. You also have to have a high enough ADC resolution and speed to pick it up too. ADCs are usually MUCH faster than that, but the code controlling the look ADC lookup has to accommodate the high speed request.

Fast signal processing is much different then what you typically use in automotive use. The fact of the matter is, when people say the signal is "noisy" it's usually not. It's just they are picking up different portions of the wave based on their sample rate. You see this in MAP sensors all the time. If you ever see a MAP sensor graph done at high speed with respect to crank angle, the pressure varies GREATLY in an intake plenum (several PSI) yet unless you are seeing compressor surge, your MAP sensor rarely varies more than 1 psi, if that. If you are seeing 2.5 psi, it could be substaintially more then that. Or it may not and 2.5 psi may be a true max. Hard to say.

Nope, I don't know. I was curious.
I know it does batch fire on TPS based async accel enrichment. Not sure about WOT though. It would be nice to have the injection angle tables for the stock ECU though if the main fueling is always sequential.

 

Now we're talking! Injection Phasing FTW! We saw a HUGE pick up in power on the Engine dyno on several different setups. The DTA Fast only had a single angle of injection, but the AEM IIRC had a curve that varied with engine speed. Good ish Maynard.

I stole this hotlink from someone that shows what an Aliased signal is. Your sample rate is too slow to pick up what's actually happening in the rail. What you're seeing is an aliased singal that could be also be filtered.