posted on December 02, 2013 13:31
Project [Rotary] FD RX-7: Part 3 - Fuel System (Chassis Side)
Fuel systems upgrades vary drastically and sometimes only a simple pump or injector change will suffice. Other times, like in this project, it seems like every component in the entire system needs to be changed. Recently, in Part 2 of Project RX-7, we covered the engine side of the fuel system, where the injectors, rail, and pressure regulator were upgraded. Now it's time to discuss the remaining fuel system upgrades that we are so eloquently dubbing the "chassis side."
In Part 1 of this project, we talked about how to prevent oil starvation during cornering and acceleration by adding baffling to the oil pan. Similarly this can happen in the fuel system as well, just on a significantly larger scale. The difference in scale and the fact that gasoline is much less viscous than engine oil means that simply adding some trap door baffling around the fuel pump isn't a viable solution. Adding an external fuel surge tank however, should do the trick.
Radium Engineering specializes in fuel surge tanks that are designed for both internal and external fuel pumps. Many of the tanks designed for internal fuel pumps can accommodate two pumps. When configured with a single pump our surge tank has a capacity of 1.5 liters. Two pumps would decrease that capacity to 1.38 liters.
A fuel surge tank (commonly referred to as a "FST") acts as a secondary fuel reservoir that is usually tall and narrow to decrease the effects of fuel sloshing generated by cornering and acceleration g's. To get a better understanding of how the Radium FST works and how they are plumbed, check out this video produced by Radium Engineering.
The top of the Radium fuel surge tank houses three black fittings that are used for surge tank feed, engine return, and primary tank return. The sole green fitting is a bulkhead fitting attached to the fuel pump inside and is reserved for engine feed. The Radium FST also features a 4-pin mil-spec sealed electrical connector for providing power to the fuel pump(s) contained inside.
Here are the guts of the Radium FST. The green bracket on the bottom cap holds one or two fuel pumps to feed the engine. The green bulkhead fitting attaches to a piece of 5/16" ID fuel submersible hose that attaches to the outlet of the fuel pump contained inside. The top and bottom caps of the fuel surge tank, as well as each mounting bolt, are o-ringed to prevent leaks.
We turned to Aeromotive for the next step this build, choosing a fuel pump. Since the surge tank requires an internal pump and because we need to upgrade the fuel pump inside the RX-7's factory fuel tank as well, we needed look no further than a set of Aeromotive's 340 Stealth fuel pumps.
The Aeromotive 340 Stealth fuel pump is a drop-in pump that delivers an impressive 340 L/hr of fuel flow at 13.5V and 40 psi base fuel pressure. At 60 psi (20 psi boost pressure) it still flows 280 L/hr, which would be enough fuel for over 500 horsepower to the rear wheels of a rotary powered RX-7.
For this build we are using two Aeromotive 340 Stealth fuel pumps. One is for the low pressure in-tank pump that feeds the Radium FST (from here on out we will refer to it as the "lift pump") and the other will reside inside the surge tank to feed the engine (the "feed pump"). The benefit of using the same pump for both locations is that it guarantees that fuel will enter the surge tank faster than it exits. This is because fuel pumps have a higher flow rate at lower operating pressures, so the low-pressure lift pump will always be outflowing the high-pressure feed pump, keeping the surge tank full.
The two Aeromotive 340 fuel pumps used for this project.
We are also using an Aeromotive 10-micron fuel filter and billet aluminum filter bracket. This will be placed inline between the feed pump in the FST and the fuel rails on the engine.
Tuesday, December 03, 2013 8:55 AM
Wow, beautiful work. Do you have an overall diagram of which tubes connect to what? I assume this is not a returnless system, so does the return from the engine bay go to the FST, the fuel tank, or both?
Tuesday, December 03, 2013 9:56 AM
Excellent work. I went through a similar procedure running new fuel lines for my fuel injected Datsun 510. Tube bending is an art and not for the faint of heart (and it rhymes) My only concerns are running your surge tank inside the car and perhaps some vibration from the hard mount from the surge tank's fuel pump. I'm sure it will be minimal to the rumble of the Wankle, but worth considering.
Tuesday, December 03, 2013 10:35 AM
Great write up, I have been researching hard lines for a while now and this is by far the best write up I have found. Thank you.
Tuesday, December 03, 2013 10:53 AM
Tuesday, December 03, 2013 10:54 AM
I only see one mention of the AN fitting size, are they all -6 with the 3/8 tubing?
Tuesday, December 03, 2013 11:54 AM
Very cool fuel system. Since the lines all free floating in the air with no bracing is there any chance that harsh vibrations could damage the lines? Or is the tubing stiff enough with the bends in it.
Tuesday, December 03, 2013 12:22 PM
Dusty Duster: I do not have an overall diagram, but you are correct about it not being a returnless system (neither is the factory system). Basically it goes main tank lift pump -> surge tank -> feed pump inside surge tank -> inline fuel filter -> fuel rail on engine -> fuel pressure regulator -> surge tank -> main tank return.
jeffball610: Thanks. The surge tank inside the car will probably have a sheet metal enclosure surrounding it before the end of the build.
mrk1: Thanks and yes you are correct...all fuel hose for the entire build is -6AN and all tubing is 3/8" OD, 0.028" wall 304 stainless with the exception of the super short piece of 5/16" OD coming off the lift pump.
Protodad: You would be surprised how stiff the stainless tubing is after bending and mounting it. Also, stainless is much less susceptible to damage from vibration (greater fatigue life) than aluminum.
Tuesday, December 03, 2013 12:47 PM
So the return line goes to the surge tank and then to the main tank. Got it. I was trying to wrap my head around the way you would install one of those in a "generic" fuel system.
Tuesday, December 03, 2013 9:17 PM
Are you finding any problems with fuel heating? I'm trying to wrap my head around the rate of heated fuel return from the regulator/rail into the surge tank versus the dilution rate of fresh fuel from the main tank, along with the rate of return to the main tank. That's an interesting mixture rate calculation.
Protodad: You want the tubing to be stiff enough that it's properly self-supporting under vibration, but flexible enough to not transfer stress from relative motion between the hardpoints like the fittings on the truck floor and the surge tank. The tubing diameter and thin/light wall provides the stiffness to resist vibration and excitation (by raising the natural frequency), and the large radius bends provide the compliance to decouple stress between the attachment points.
Tuesday, December 03, 2013 11:27 PM
I've heard that those surge tanks when using 3 pumps do get fairly hot. Adding a series resistor to the pumps under low load as most OEMs do wouldn't hurt. You could always add a thermocouple into the surge tank and use a fuel density correction in the standalone too.
Great write up. The amount of work required to get good looking hard lines is significant and should not be under estimated. I've done stainless, coated steel and aluminum lines and they all have their own difficulties. Starting from coiled tubing like you get from summit is a PITA. If possible, find annealed straight tubing to start from.
Wednesday, December 04, 2013 1:07 AM
I've done tube bending back in my HVAC days... definitely requires some experience to get the length just right when bending!
Wednesday, December 04, 2013 8:33 AM
Fabrik8: The car is not running yet, so I can't comment on fuel heating. I did think a bit about the mixing of "cool fuel" from the tank and "warm fuel" from the engine though. Since all the returns are located at the top of the surge tank and the feed is taken from the bottom, the fuel in the surge tank will never really be stagnant. Couple that to the fact that the fresh fuel flowing from the main tank into the surge tank will be at a greater flow rate than that returning from the engine (~370-380lph for the lift pump compared to ~330lph at 43.5psi or ~270lph at 63.5psi for the feed pump).
The calculation of mixing becomes very complicated once you take into account injector flow-rate which decreases the surge tank return flow rate, but even at the worst case scenario (at idle) the lift pump will always be supplying a minimum of 10% fresh fuel. The fresh fuel content in the surge tank only improves as the injector duty cycle increases.
Wednesday, December 04, 2013 8:43 AM
mike156: I'm assuming you are talking about the thread in the Supra forums about running three of these pumps as parallel feeds. I have to imagine that does get quite hot. Luckily I only have one feed pump that is operating at higher current and one lift pump operating at lower current, so I don't expect to see high fuel temps like seen in that Supra setup. I have thought about running a thermocouple as well, but think I would put it just before the fuel rail to capture any effect of heat lost through the stainless tubing running under the car and any heat gained from the engine.
Also, all this tubing was from coil and you are correct that it is a pain to straighten prior to bending. I did look for straight tubing, but the price difference between the coil and the straight was more than I figured was worth the convenience.
Wednesday, December 04, 2013 9:13 AM
I'd imagine the lift pump inflow/outflow should be pretty large without much flow restriction to raise the pump pressure and a relatively small amount of outflow to the fuel rail compared to that, so there should be a significant amount of turnover in the surge tank from the main tank. I guess the lift pump also really acts as a pressure pump cooler.
Thursday, December 05, 2013 9:38 AM
Awesome write up and stunning work.
that fuel pump is when `function > form` = `function and form`
you should add little stickers like they do on pipes in factories indicating what is in the pipes and which direction its flowing....
Thursday, December 05, 2013 11:12 AM
I'm thinking a laser etch or silkscreen for maximum impact.
Tuesday, December 10, 2013 2:40 AM
Beautiful! Great work guys.
Thursday, December 12, 2013 4:34 PM
How are you guys straightening (sp?) the coil of stainless tube before bending it?
Thursday, December 12, 2013 4:47 PM
Tube straightener. Not joking, there's tools meant specifically for that. It's usually 5 or so grooved rollers two rows, and one row has an adjustment to set the distance between the row. Run the tubing through the roller, and presto...
There are also convex grooved pliers that are cheaper and don't work as well, they work like line bending pliers in reverse.
Thursday, December 12, 2013 6:33 PM
Fabrik8 stated the "correct" way to do it if you are doing a lot of it.
Everything in this build was straightened the old fashion way...by hand.
Monday, December 16, 2013 10:50 AM
Fellas, I'm seeing some SS tubing options on summit but am having a pretty tough time finding 1/2" SS (8an) tubing :( I've seen a few options on amazon but they were from a brewing supply company and I'm not entirely sure it would work for an automotive application. Any suggestion as to where I might find some -8an SS tubing?
Monday, December 16, 2013 11:01 AM
Monday, December 16, 2013 12:19 PM
Agreed, inlinetube.com is the place many people get their SS tubing from.
Wednesday, December 18, 2013 9:57 AM
Thanks guys, looks like they have what I'm looking for!
Sunday, January 05, 2014 10:02 AM
From a DIY perspective, why not use the hard line tubing adapters? It eliminates the need for flaring tools. They also work well for converting an OEM fuel pump hanger to AN, you just cut the barb off the end and put the hard line adapter on.
Friday, March 21, 2014 10:18 AM
I would love to see a project update.
Friday, March 21, 2014 10:45 AM
Hoss: You are in luck. The work for Part 4 is complete and the article will be done very soon.