Fuel tank v2.0

Issues, solutions, helpful hints, shortcuts, tools, etc
Post Reply
Midlana1
Site Admin
Posts: 2917
Joined: 19 Dec 2008, 16:44
Anti-spam question: 4
Are you a spammer: No
Location: Southern California

Fuel tank v2.0

Post by Midlana1 » 14 Jan 2015, 21:16

For several reasons I'm gathering parts to make another fuel tank, call it tank v2.0, and this thread will cover its design and fabrication.

1. The goal is to be able to run ethanol, which requires a compatible pump, and ones with sufficient flow are in-tank units. I'm currently running an external pump that's neither E85 compliant nor enough flow.
2. I used aircraft fuel tank sealant during construction of the first tank and it's unclear if it's ethanol resistant.
3. The tank weeps slightly - it never drips outright, but I can still smell it and that bugs me.
4. Due to the design of the filler, it constantly trips the auto-off on gas station filler valves. The only way to fill it is to throttle it to as slow a flow as possible. This bugs me too.
5. The old tank distorted terribly due to excess welding heat. No one else can see it but I know it's there.

Are these enough reasons to make a new tank? If I have to ask, the answer's probably no, but I'm going to do it anyway because I enjoy building stuff.

This one will be slightly simpler than v1.0, with only three baffles, though they'll be larger. The in-tank pump has a sock filter at the foot of the pump so the sock must be kept away from the one-way fuel doors.

Someone asked if I'm going to use Holley's new "HydraMat" (https://www.youtube.com/watch?v=4rG97UuVgmQ) that'll be on the market next month. After thinking it over, I surprised myself a little by answering "no":
• The mat is retained with magnets on the floor of the tank so that it doesn't slide around or bunch up, but that doesn't work in a plastic, stainless, or aluminum tank. Yes, it can be trapped with straps, but it's still going to move around during cornering and there's no information on the effects of abrasion on the membrane, but it's probably not good.

• How much flow can it handle - not specified

• How much flow is available as a function of fuel coverage - not specified

• I need a mat larger than the single size offered. No word on future sizes or release dates, if any.

• No prices listed. The rumored price was very high.

• E85 compatibility - not specified

• Physical baffles are still a good idea because they limit having 50 pounds of “ballast” slamming left and right through turns.

Tank material:
I really wanted to go with aluminum this time because it's so much more pleasant and easy to work with. Chemical property tables show that alcohol in an aluminum tank is okay, but there's a big catch. That is, no matter what, there'll always be some water in the mix and this is the killer. What muddles the data further is the presence of 15% gasoline. I'd have thought that the oils in the gas would protect the aluminum from the water, but apparently this isn't so. I understand that aluminum can be hard anodized which will protect it, but the plating costs doubles the overall expense, so it'll be stainless.

Pump:
I received the Walbro 400 E85 fuel pump (PN F90000267) today and was shocked how small it is, roughly half the volume of the Bosch 044 pump. Maybe it's because it runs submerged in fuel that it stays cool. Due to its size, there certainly won't be a problem getting it to fit, which was a real concern since the top of my tank is only 4" wide.

Fuel level sender:
I currently use a capacitive sensor with no moving parts. It's an awesome unit but surprise, having an aluminum tube and with ethanol affinity for water, it won't operate properly. I went old school and picked up an Autometer mechanical unit, one with an arm and float. I asked Autometer if it was E85 compliant of course it wasn't, them saying the steel would corrode. That was an interesting comment because several chemical compatibility charts I've seen show bare steel as being okay with ethanol. I think what Autometer was talking about was the effects of water in the fuel, it's not the alcohol that's destructive. I'm going to use it anyway because I have seen any alternatives, and the sensor's fairly easy to replace.

This thread will be added to as the project progresses.

dpkilty
Posts: 223
Joined: 27 Jan 2014, 20:13
Anti-spam question: 4
Are you a spammer: No
Location: Indiana, USA

Re: Fuel tank v2.0

Post by dpkilty » 15 Jan 2015, 11:22

Not that this will sway your decision to use the hydramat, but I though it might help some others.

I talked to the guys at Holley at the PRI show about the mat.

It is indeed E85 safe.

They did not say a set GPM it can flow, but they did say it was getting tested in alcohol dragsters. So I would assume that would be a high enough flow rate for anybody building a street car.

Pricing was to start at around $100 for a small sized mat. They did not specify what a small sized mat was though... Going up to about $1500 for a large 2 ft x 3 ft (approximate) mat.

As long as its touching fuel you don't starve the pump. They do recommend that you pipe the fuel return line so its spraying directly onto the mat. That way it limits your chances even more of suffering fuel starvation.


The baffle plates are still a good idea if only to keep fuel from sloshing around as much. Personally I am going to try a combination of the mat and baffles in my Lotus tank. If I can get a mat that at least fits most of the section where the pump is; as long as that section has fuel in it I'm good. But I am also still thinking about utilizing a fuel surge tank on top of all that just to gain a little extra fuel storage capacity so its a moot point anyway. Using the Lotus tank means I have a smaller tank than the book design is.

As for mounting, since the mat uses magnets to hold it in place, is it possible you could weld short threaded studs onto the bottom of the tank (on the outside of the tank) and use them to bolt on a small piece of steel (or even another magnet). This might allow the magnets to grip tight enough, but I am not sure how many pounds of pull their magnets are.


All that being said I am still looking forward to seeing this thread progress.

Midlana1
Site Admin
Posts: 2917
Joined: 19 Dec 2008, 16:44
Anti-spam question: 4
Are you a spammer: No
Location: Southern California

Re: Fuel tank v2.0

Post by Midlana1 » 15 Jan 2015, 12:53

I wonder what happens after the mat has slipped around against the bottom of the tank for a while. Because they said it operates via surface tension, it would seem critical to not have any wear. I wonder if the active surface is only on the top of the mat... could be, but then it would be giving up a small amount of usable fuel. Who knows.

Also, are we to assume that it serves double duty as both a liquid-gatherer and filter sock? I assume so.

steve
Posts: 37
Joined: 03 Apr 2014, 13:40
Anti-spam question: 4
Are you a spammer: No
Location: Durham UK

Re: Fuel tank v2.0

Post by steve » 15 Jan 2015, 12:59

I have used a plastic boat tank in my jeep and mounted a Ford Fiesta fuel sender, its a biggish bit of plastic with a swirl pot for returning fuel, a pump built in which can be replaced with whatever rating required and has a fuel level sensor on the side.

Got the idea from Sylva Autokits here in blighty cause they use the setup in a bespoke Ali tank on the mid engine cars they build.

Midlana1
Site Admin
Posts: 2917
Joined: 19 Dec 2008, 16:44
Anti-spam question: 4
Are you a spammer: No
Location: Southern California

Re: Fuel tank v2.0

Post by Midlana1 » 15 Jan 2015, 13:24

I'd do that in a second but the lack of online specifications ruins some of the fun; there's no way to know whether they'll fit and especially how they mount.

Some have a bayonet-type attachement which could be rather involved to make a fuel-tight mate for. Some have multiple bolts, which is more universal and easier to fit, but without dimensions, it could get costly guessing from tiny online pictures what "might" work.

There's also no way to know the diameter of the mounting flange, other than guessing. There's also guessing how tall the units are; yes, they can be modified, but the more modifying that's needed is just another step toward doing it yourself from the start.

Lastly, unless specifically buying a unit out of a flex-fuel car, who knows whether the materials will hold up.

PS: I think the term "Bespoke" is really cool :)

dpkilty
Posts: 223
Joined: 27 Jan 2014, 20:13
Anti-spam question: 4
Are you a spammer: No
Location: Indiana, USA

Re: Fuel tank v2.0

Post by dpkilty » 16 Jan 2015, 09:03

Midlana1 wrote: Also, are we to assume that it serves double duty as both a liquid-gatherer and filter sock? I assume so.
Yes this doubles as a filter sock. Which is why I plan on using at least a small one as I will need a filter sock anyway...

Midlana1
Site Admin
Posts: 2917
Joined: 19 Dec 2008, 16:44
Anti-spam question: 4
Are you a spammer: No
Location: Southern California

Re: Fuel tank v2.0

Post by Midlana1 » 19 Jan 2015, 08:18

Never have I gotten myself so confused about hole placement! The problem was that the sheet had a protective film on one side, so of course that would be the outside surface. To protect it, I did all the marking, cutting, and drilling on the unprotected side, but kept getting confused by the multiple levels of "inversion." That is, I had the sheet on the floor with the inside up, with the tank "unwrapped." The trouble was figuring out if the filler went on the left or right end (it's on the right side of the car) but depending where I was standing and how the sheet was oriented, it kept looking wrong, then right, then wrong, wrong again, then right again. I must have erased and moved the cut marks four times before thowing my hands up, placing the sheet on the floor next to the car, and pointing the forward side forward. Only then did it become clear what needed to go where, but I came really close to ruining a $100+ sheet of stainless. Even worse would have been to have the sheet metal shop bend it, increasing the financial hit even more. Close call, and I'm revising the rule "Measure twice, cut once" to "Measure five times, consider what it costs, then cut once."

The stainless sheet is cut, trimmed, and drilled, now it just has to be bent.

schwac2
Posts: 14
Joined: 11 Feb 2009, 14:47
Anti-spam question: 4
Are you a spammer: No
Location: Providence RI

Re: Fuel tank v2.0

Post by schwac2 » 02 Feb 2015, 10:53

Hey Kurt -

One of the things we do for leak checking at my work (biotech) is vacuum checking, not just pressure testing. The advantage is a little better than soap/water - during a vacuum test we apply a foam to any suspect interface and when vacuum is applied, the foam pits exactly where the leak is (even tiny ones). This may help? Maybe not?


Casey

Midlana1
Site Admin
Posts: 2917
Joined: 19 Dec 2008, 16:44
Anti-spam question: 4
Are you a spammer: No
Location: Southern California

Re: Fuel tank v2.0

Post by Midlana1 » 06 Feb 2015, 07:02

Another suggestion was to use Magniflux Spot Check, a two-bottle test kit that identifies cranks and holes in metal. You apply Agent 1 to the surface, wipe it off, then apply the Agent 2 developer. While not exactly cheap (~$50 for both), if the testing isn't done and the tank leaks, it'll suddenly seem like a great idea!

I was told to spray the stuff inside the tank and develop the outer surface, but it may not work well due to the baffling and how the stuff comes out the spray nozzle. I may just apply it to the outside surface of the welds as it's intended to be used. We'll see

Midlana1
Site Admin
Posts: 2917
Joined: 19 Dec 2008, 16:44
Anti-spam question: 4
Are you a spammer: No
Location: Southern California

Re: Fuel tank v2.0

Post by Midlana1 » 06 Feb 2015, 07:32

There were two issues with this tank:

1. As noted in the Bug section, the fuel tank drawing has an incorrect dimension such that an additional strip of material had to be welded in.

2. The sump should go in the center of the tank to maximize fuel scavanging. However, the coolant lines also pass through this area and the fuel supply and return fittings may block where the coolant lines. The coolant lines can move somewhat but the fuel return fitting may need to be offset to maximize space for the coolant lines.

Post Reply