Why are MFI injector fuel lines equal length?
 

I guess the obvious answer is related to the timing of fuel pulses at each cylinder.

BUT, if the lines are full of gasoline, a liquid, therefore non-compressible, the actual length if each pipe should not affect pulse timing, within reason of course.

Is it related to fuel volume in each pulse, or something more complex that I am missing completely??

I am asking because a non-Porsche friend, very knowledgeable in precise plumbing type matters (former machinist mate on a boomer), asked me, and I was not able to satisfy even myself with my reply.

Help!

There must be hysterisis is all I can think of--there must not always be 220 PSI injection pressure in the lines at all times, so that the length makes a difference. Either that or it's Germans being German.

I'm not sure Henry Schmidt's flex lines are the same length. Maybe he will chime in here. Also, someone look up a pic of the 917 12 cyl. Are they the same length?

Are brake lines or CIS lines the same length? If the fluid line contains no air, the contents move as a solid column.

Sherwood

maybe just cheaper to make from the same stock..... :rolleyes:

Henry's lines are not equal length from some of the photos I've seen, and I don't think that (all) 917's had equal length lines. Using the Darcy-Weisbach equation we can see that there is a linear relationship between the length of the pipe and pressure drop.

So, it appears that there will be a small difference in flow, but not enough to significantly impact fuel ratios. So John is likely right - just Germans being Germans.

I know you say this in jest or maybe it's sarcasm but it's probably the reason. The lines have a nipple on the end so it's not just a matter of cutting to length. They would have to manufacture six different versions to have them different lengths. They each need to be bent differently but at least making them all the same length simplifies one step of the production and keeps cost down.

Bobby, I really meant it that way. If there is no benefit either way they making them the same length takes one step out of the complexity of making all these parts.

Got it. :rolleyes: = sarcasm. :)

I don't think it has to do with production costs. If I were to hazard a guess, it's for when you're starting - the pump works by having a little piston push a certain amount of fuel through the lines. If the lines are full of gas, then no problem if they are not equal lengths. But if fuel drips out of the lines through the injectors, then they will not be full. If they are not full, then having unequal lengths will mean that some cylinders will get fuel before the others upon starting.

That would be my guess.

Less likely would be the friction and/or the compressible effects of the gas in the lines. MFI is pretty high pressure, and different length lines will have different pressures at the end injector, if the gas doesn't act like a 100% perfect hydraulic fluid. I hesitate to say that if Porsche was building the cars today they would have different length lines, but back then, they seemed to do things "the right way" for the sake of doing them. There are a bunch of other 911 details I can point to that fit into this category as well.

-Wayne

I'm liking Wayne's thinking here, I always wondered how the system gets filled at starting, esp. when the car has been sitting a long period. Surely at that point the lines are no longer filled with fuel.

I am also thinking about the "head pressure" effect on the in-line pump at speed. I can not find any work done in support of this idea, but it seems to me that a rapidly spinning (up to 4000 rpm or thereabouts) 220 psi six piston pump may have some harmonic or other difficulties if the loading on each piston is "unmatched" so to speak.

Any takers on that idea??:confused::confused:

Yeah, it's a German thing!
 

Probably the same reason that for the first 40 years the 911 had a fuel gauge that had widely spaced increments toward the full end and tiny lilttle increments at the empty end...where you might actually need to know more precisely how much gas is left in the tank!!

Go figure.

Just not a German thing. Also a US thing, many a Caterpillar diesel engine with mechanical fuel pump/distributor have equal length injector lines, to the extent of lenghts of pipe with U turns etc. to maintain the equal lengths. Only the other month the topic of compression of fluids was raised with some hydraulic experts, and was advised that there is minor, minor compression of liquids. The concept of equal length (and equal everything else) in the fuel lines is to eliminate any variences.

Paul

its all physics
 

Poiseuille's equation
states that flow is inversely proportion to length
and proportional to radius to the 4th power
so length is a relatively non critical factor but does matter.
a kink in your line is a disaster.
i don't know why i remember this.
you can wikipedia for the actual formula
bob

The lines are not full of gas, MFI is a metered injection system unlike CIS which is continuous. Therefore the lines need to be equal length or it will affect the fuel timing.

I think they are not the same length. If they were it would be for the simplicity of production or some other mundane reason. Of course the lines are full of fuel. The injectors open at something like 200 PSI. If there were air in there no fuel would flow. Unfortunatly I sold my MFI engine and installed a 3.0 or I'd go into the garage and check right now.

-Andy

We have built lines, both even and uneven lengths according to the customer's wishes. You will see no difference between even and uneven lines (verified with an LM-1). You also have fuel in the lines at all times if the pump is running properly. The check valves in the pump guarantee this. If air were to be in the system it would diminish both timing and volume demending on the amount of volume of air that is present in the line.

At 220 psi the lines would have to be drastically different in length for it to make a difference. It would take several hundred feet for it to make a difference if even then. The weight of the fuel in these short distances makes matching lengths inconsequential at that kind of pressure.

If you have a leaky injector then having equal length lines will not help. The pressure will be down at the time of injection.

The lines are only full of fuel during an injection event. The fuel is metered to each cylinder by the injection pump, not the injectors as in EFI. Read Dempsey's book.

FWIW,..the fuel lines on our old MFI RSR and 935 were not equal length; I used to replace them annually. :) :)

I never observed any power balance or EGT issues between cylinders that were attributed to different length injector lines.

We don't have a street MFI system here at the moment to confirm but I recollect they are not all the exact same lengths.

Wrong. Fuel is in the line at a minimum pressure of approx 24 psi.

Has anyone considered the inertia that a column of fuel presents. During an injection event, the fuel must be accelerated from rest then brought back to a complete stop in as little as 1min/4000 or .015 seconds. Inertia could have an effect on this process. (Think of "water hammer") If the lengths are all equal, it would eliminate a possible source of mixture variation as the mass of each column would be equal.

Matt

The volume of fuel that moves with each event is very small (anyone have a range?). I would therefor expect inertial and head impact to be insignificant.

I'm with Ingo on the manufacturing economies.

If so, this "residual pressure" is 10% of the injection pressure...but you miss the point. If a 220-240 psi fuel impulse is released from the pump, there will be a propagation delay (travel time)through the injection hose as the fuel impulse travels to the cylinder. If all the hoses are different lengths, the arrival times of the fuel impulses will be different. It may not matter much in the real world, but I would make my injector lines the same length.

Your "impulse" (if there was one) would travel at the speed of sound in fuel (I'm guessing 1800 FPM). Not much time required to go through a 2 foot line. The fact is there is no impulse. The injection pump uses a cam to push a piston that squirts the fuel. The cam accelerates the piston at a reasonable rate just like the cams in your valve train. You wouldn't call a valve opening an impulse event.

The injectors are just spring loaded valves, they are not controlled in anyway. As the pressure in the line rises above 200 PSI or so the spring compresses allowing the fuel to flow until the pressure drops back below 200 PSI. The pressure is supplied by the piston moving in a cylinder full of fuel inside the injector pump. You could think of these springs in the injectors as the equivalent of the valve springs in your cylinder heads.

-Andy

The valve opening (inlet/exhaust) is not a good analogy. Tuned induction and exhaust systems are bases on using the "impulse(s)".....

Lets expand on the effects of different length fuel lines. If you have two lenghts of hose, one is 1 foot long and the other is 1 mile long. With an identical supply source (pressure and pulse) the resulting timing and flow at 1 mile is different to 1 foot. This example is a major exaggeration, to make the point that there is a difference. Now back to the MFI, what diifference does it make if there are unequal lenght fuel lines? not much, but there is a difference. On diesel engines that use a common mechanical pump with individual fuel lines to each injector this difference is more important as it also effects the firing time for cylinders, hence the fuel lines are equal length as in the below photo.

http://forums.pelicanparts.com/uploa...1195797835.jpg

If all converted to EFI then this thread would fade away :)

Paul

Sorry, that's not true - the lines are full of fuel at all times when the engine is running. If they weren't full of fuel, then you'd have a compressible situation (with air in the lines), and you wouldn't get the pressures that you needed out of the lines. It works like a hydraulic line at that point. You can have fuel in the lines but not under pressure at a certain moment.

-Wayne

Mercedes used unequal length fuel lines on their MFI cars. Fuel is in the lines at all times, and relatively incompressible. The MFI pump is a positive displacement pump. A tuned intake relies on a compressible fluid.
I vote that it was for production efficiency...

Think of it as a pipe with a row of ball bearings inside. When you push a ball bearing in one end of the pipe a ball bearing will fall out the other end of the pipe at the same exact time of the one being pushed in.

Now if you have two pipes of unequal length and you push a ball bearing in both at the exact same time a ball bearing will fall out the other end of each at the exact same time. What is different between them, however, is the amount of pressure required to push the ball bearing out the other end. The longer pipe requires more pressure.

Timing is not an issue in different lengths pressure is the issue.

The weight difference of the fuel between different length lines in an MFI setup is of no consequence at a 220 psi pressure. There is more than enough pressure to push both easily.

Thanks Wayne, you are of course correct. What I was thinking, and should of said, is the lines are not pressurized all the time like EFI.

This has been a fascinating discussion. My personal opinion is that the lines are equal length because they are subject to the conservation of mass law of physics.

B-b-loo's ball bearing analogy is a good one...

Fuel impulse=Pressure impulse
Absolutely I would call a valve opening an impluse event. In fact, the ideal cam "profile" would be a rectangle...fast open to max height, dwell, then fast close. Similar to the improvment attained by roller cams over flat tappets.

Actually more like 1200 MPS...(meters per second)..the denser the fluid, the faster the speed. I think that @ 220psi the expansion and contraction of the steel line is more of a factor than compression (yes, it does compress) or weight of the fuel...equal length lines will give you the same rate of fuel with linear movement and pressure drop...all of which, combined, will affect timing to one extent or another...why have any variation when you can just make them all the same and be done with it. Theoretically, spark plug wires should be of equal length also .

I agree that a rectangular cam profile would result in an impulse event :)

-Andy

I am gonna bump this in hopes that some more of our experts will join the discussion. As far as I can interpret, we still don't know what the real reasons are here.

I suspect the answer is the expansion of the LINE during the injection stroke. With an increase in line pressure, the line itself will expand in diameter, and therefore the mass of the fuel in the line will increase. Albeit by a small amount in steel lines...but more so in the flexible RSR/Racing lines.

I'm sure somebody here has the inclination to do the calcs for steel vs nylon lines at the working pressure...

I believe the Bosch test procedure calls for equal length flexible test lines 6x2x600....test fluid at 45°C (can anybody confirm this ?? and or post a copy of the procedure??)

The change in the mass of fuel in the line with each injection stroke will be in direct proportion to the length of the line.

With respect to presure drop due to flow rate (and line internal roughness) - the flow rate is too low to introduce an appreciable pressure differential

In any event, equal length lines with minimise variances due to all effects - (in)compressibility of fuel, line expansion both diameteral & linear, and flow rate pressure drops)

Now if you could only guarantee that all of the line were at the same temperature........

Anyway, thats my 2 cents worth...

I think we're overanalyzing this knowing that other manufacturers used different lengths.

As Bobby and Ischmitz suggest at the beginning, it's just a production step. I tend to agree. PAG can order x thousand straight lines from their vendor, then specify 6 bend patterns for each. It's easier on inventory. Instead of ordering 6 different part numbers and keeping track of how each line gets bent, all lines are bent 6 different ways.

Sherwood