Webinar 092 — Fuel System Design

Forgive me if I posted this question incorrectly; newbie here.

Just watched webinar 092 on fuel system design and I have a question regarding the return line. Andre mentioned he typically uses the next AN size down on the return vs. the feed. He also mentioned that the smallest ID of the system is going to be the determining factor on flow rate and pressure. While the return orifice on my Aeromotove FPR is -6 the orifice inside is only 0.21” (see attached photo).

I’m confused as to how any return line size greater in size than this orifice size will make any difference in return flow capacity or pressure? Admittedly I have no formal training on fluid dynamics so excuse my ignorance.Thanks so much and you guys are amazing in your knowledge, passion for the subject matter and presentation skills. I hope you’re making good money and will be around for a long time!

I'm no doctor of physics, but I'll try to explain, until someone more experienced pitches in.

The orifice on your pressure regulator (in combination with a spring loaded valve) regulates the amount of pressure on the inlet line. That's where your injectors are, and by connected vessel fluid dynamics (apologise for the poor translation), injectors are under this pressure too. After the orifice, the line is opened to atmosphere (indirectly via fuel tank), so there's atmosphere pressure there, or no gauge pressure.

Ifyour return line was the same size as the orifice on the regulator, then the fluid passing through would have to flow at higher speeds, directly dependent on the pressure and amount of fluid that is flowing. More fluid/pressure, higher flow, less fluid lesser. But fluid is sticky, and likes to grab onto surfaces. So in your same diameter return line scenario, you'd have a lot of drag. Very little fluid would flow, because most of it would catch the inside wall of the line. This in turn would increase the pressure in the return line, and thus affect your regulated pressure as well. So to counteract this, a return line diameter is increased. This allows for more space for the fluid to flow, but also, because the same amount of fluid having to fill in more space, fluid slows down, in turn reducing drag.

Now, if you consider that most fuel is returned at idle (least fuel injected), then you need to engineer the line to be large enough in diameter, to prevent pressure building up in the return line. One size down in comparison to supply line is a good rule of thumb, for added comfort.

Hope this makes sense... (*scratches the head*)

:)

Rikko,

That’s for taking the time to explain your answer; makes a lot of sense.

So essentially, if the resistance to fluid flow of the return line is greater than the resistance of the regulator, the return line, in essences becomes the regulator, with no means of adjusting regulation.

Yes, I'd agree.

Anothersimple analogy is your water tap, tap being fuel regulator. There's no real pressure of the water flow after the tap, until you try to stop it by hand (this being analogue to reducing return line diameter).

Cheers

Thanks again!

~Rich

Other affects to consider is drag = friction thus adding temperature to the returned fuel.

Splashing in the fuel tank when being pushed back into the tank also aerates the fuel which is also an undesirable affect.

For these two reasons I have a braided line to and from my engine bay from the previous owner which I am changing to a hard line to add some indirect cooling as the lines will have air passing over them all the time which should cool the lines marginally. Not a huge gain but a consideration.

Brian,

Thanks so much for the additional insights to good fuel source design. I really haven’t seen any comprehensive write-ups on the effects of fuel temp and how/if to attempt to address.

Thank you!

~Rich Laws

Akron, OH

Paul Yaw (the guy behind Injector Dynamics) wrote this back in 2013. I had to use the Internet Wayback Machine to find this from his Feb 2013 newsletter:

Paul’s TechnoRant

In last month’s TechnoRant, I discussed the difference between injector volume flow, and mass flow, and described how mass flow is dependent on fuel density. I also hinted at the fact that fuel density is dependent on temperature. This month, I will detail that dependence, describe how it affects your tune, and give you a tool to compensate for it.

Let’s start with what really matters, which is how it affects your tune. Have you noticed that even a well-tuned engine runs leaner when the fuel is hot?

There’s a reason for that…

The reason is that the density of the fuel changes with temperature, which changes the injector mass flow rate. Predictably, your air fuel ratio changes because your fuel injector is now effectively smaller. (Remember, air fuel ratio is mass based.)

This change in density can be predicted with a parameter called the “Coefficient of Thermal Expansion” (CTE) which is a description of how much the fuel expands with temperature.

It sounds complicated, but like many things related to performance, most of us already have an intuitive feel for it. For instance, we all know that our pistons expand when they get hot, which is why we specify a cold clearance. And we all know that the fluid in the radiator expands when it gets hot, which is why we need an overflow tank. It should come as no surprise then that our fuel also expands when it gets hot, and in fact it expands at a greater rate than either water or aluminum.

So let’s consider how this expansion changes the fuel density.

Let’s start by filling a cylinder to the very top with ethanol. Ignoring evaporation, if we heat the ethanol, it will expand and overflow just like the fluid in our radiator. The cylinder contains the same volume of ethanol, but because a portion has spilled out, the mass of ethanol in the cylinder has changed.

Referring to last month’s TechnoRant, we know that density is defined as mass divided by volume. As we divide our unchanged cylinder volume into a lesser mass, we get a lesser density, which we already know will reduce our injector mass flow rate.

The situation is the same on a running engine. The fuel injector delivers a specific volume of fuel to the cylinder, but as the temperature of that fuel changes, so does the mass, and so does our mass based air fuel ratio.

The hotter the fuel gets, the smaller the fuel mass delivered to the cylinder, and the leaner the air fuel ratio becomes.

Let’s move beyond generalizations, and be more specific.

Gasoline, ethanol, methanol, toluene, etc all have a similar CTE value of approximately .001 per degree Celsius. This means that for every change in temperature of 1 degree C, our volume will change by a factor of .001, which you math geeks will recognize as 1/10 of one percent per degree Celsius.

Let’s apply this to a sample 1000 cc/min injector, and calculate the mass flow rate at a few different temperatures.

We start by looking up the density of ethanol, and we find that the density of .789 is stated at a temperature of 20 degrees C.

Multiplying our 1000cc/min by ethanol’s specific gravity of .789, we get a mass flow rate of 789 grams per minute at a fuel temperature of 20 degrees C. Now let’s consider what happens after a half hour of driving in stop and go traffic when the fuel temperature climbs to 70C, using the following formula:

(((Reference Temp – New Temp) * CTE) + 1) * Reference Density = New Density

Using the values from our example:

(((20 – 70) * .0011) + 1) * 0.789 = 0.746

As a result of the density changing from 0.789, to 0.746, our injector mass flow rate has now gone from 789 g/min (13.2 g/sec, or 104.4 lbs/hr) to 746 g/min (12.4 g/sec, or 98.7 lbs/hr)

Putting that to numbers that we can more easily relate to, our air fuel ratio of 11.5 to 1 has just become 12.2 to 1 and would need a trim of approximately 6% to bring it back in line. Not the end of the world as long as our tune is conservative, but potentially ugly if we are cutting it close.

And I think most would agree that a 6% trim is unnecessary if we can compensate for it in the tune right?

While half of you are nodding your heads in agreement, and planning to perform the calculations to account for this density change, the rest of you are shaking your heads violently, thinking “There is NFW I am going to struggle with a bunch of math just to fix a 6% trim!”

For those of you suffering from a complete lack of mathleticism, I have provided a handy little tool in the form of an Excel spreadsheet that you can use to generate a fuel temperature compensation table to plug into your favorite ECU.

It is hopefully self-explanatory, and useful. Feel free to use and distribute it, but don’t even think of removing the logo and trademark information, or you will be the subject of my next rant!

Moving on to the subject of rants…I am left without the energy or motivation to let it rip. Let me end by saying that we put a man on the moon in 1969. We can sure as hell account for a simple temperature change in the year 2013.

I sincerely hope that you found this information helpful.

Paul Yaw

Injector Dynamics

David,

Thanks SO much for your efforts in finding that article--extremely informative!

I came away with a couple new questions that maybe you have some knowledge of. Seems like every time my eyes are opened up to an answer to one question I come up with 2 more to take its place! Grrr.

In the example of the fuel temp extremes we can reasonably expect to see in the course of a day's driving, Paul mentions going from 20C leaving the garage to 70C after some hot day driving...here in the US, that's 68F to 158F :-) My questions is this, I have a conceptual understanding of how ambient air temp obviously effects fuel tank fuel temp, but I really have no idea to what degree (pardon the pun) the recirculation of the fuel (on a recirculating type system, such as my Supra) adds to fuel temp? I get that the warm pump itself and the friction of the gas moving through the lines and circulating through a hot injector rail will add temp to the fuel, but I have no idea how to quantify the amount of heat added due to line resistance or more importantly, what the total temp. effect would be--is Paul's example of a 50F temp change a realistic number? If so, once you factor in substantially larger injectors than the 1000cc/min sized injectors that Paul uses and bump those tot he 1725cc/min I'm using on my build, that only exacerbates the amount of fuel mass change, correct?

In any event, I'm getting off my original topic of why we need a -8AN return line for an orifice hole of 0.021" on the aeromotive fuel regulator, but it's all fascinating stuff! :-)

I need to find the spreadsheet Paul talks about that he created to account for fuel density variations that can be input into my Haltech 2500 as a compensation table to adjust injector dwell time for changes in fuel temp...good stuff! Forgive me if I just described how the AF would be adjusted for fuel temp density changes; I'm new to tuning and am trying to learn from HP Academy...I've yet touch an aftermarket ECU or take my Haltech Elite 2500 out of it's box...going to have to soon though!

Thanks again!

Rich Laws

Akron, OH -- USA

Paul Lives in AZ, so it's totally possible he see's fuel temperature rises like that. Looking at my own race car data, I think the highest fuel temp I've seen is something like 45c (113f). My fuel return goes to a small 1qt makeup tank and the overflow from that tank returns to the fuel cell. So a small amount of my fuel will heat up just re-circulating through the pump.

Of course, I can't find any files right now where I was logging Fuel Temperature -- I do have a sensor, and the MoTeC M1 will compensate for fuel temperature if a sensor is fitted.

Also remember heat sources can be ambient engine bay temps too. So if your not passing enough air over a hot engine which adds heat to the likes of the intake manifold which depending on how your fuel rail is mounted will absorb some of those temperatures as well. If outright performance in all environments is a major factor a sensor as David has mentioned above could smooth it all out for you.

PS, attached is the spreadsheet Paul referred to in the article.

David and Brian, thanks so much! Really appreciate the input and the spreadsheet.

I’m pretty sure the Haltech Flex Fuel Sensor I’m installing does fuel temperature sensing as well…need to verify.

~Rich Laws | Akron, OH

A very while back, just to add to the subject, I was reading a blog about nitrous injection systems. Amongst other things, concern was a supply line, and it's thermal mass depending on material it was made from. Goal was to minimize the illeffect of heating the nitrous while in liquid form, to maintain accuracy of injection quantity. As a result, a PTFE tube was author's best choice, as it offered minimum heat soak.

Same could be translated to fuel supply line, if heat was a concern.