RobzWorkz #8b: Improving the 328 Fuel System and Peripheral Hoses (fuel tanks/pump)

Welcome to the second installment of my 328 fuel system refurbishment effort!

This thread will address the cleaning/improvement of the tanks while installed, the replacement of fuel tank hoses and the replacement of the fuel pump/accumulator/filter. The hoses replaced are those for the fuel tank and it's vent control only; thus this does not include the evaporative emissions, which, of course, attaches to the right tank, but is deferred to its own thread.

My tanks required some TLC since they are old. The tanks were cleaned out and all nipples reinforced with QuickSteel. Also, the nipple for the CIS fuel return was weak and failed during hose removal and thus I repaired it.

The hoses include SRI hoses and Aeroquip as stated before (RobzWorkz #8a). Reproduced below is the pic showing the hoses I used with measurements made.

The replacement fuel pump is of the common (non-Ferrari licensed) type, as well as the Bosch accumulator; both of these items should present a reasonable pressure (~5bar) so the CIS can controll pressure adequately. The pump assembly is supported by rubber mounts, which I replaced with new from TRutlands. The fuel filter is part of a maintenance kit from Eurospares (its diameter was smaller than previous, so I need a smaller mount ring...done later as part of the hose rebuild effort).

1. CLEAN/IMPROVE THE TANKS:

Both of my tanks, even at 26yo were in good shape. I saw no leaks nor holes.

FYI, it is an established problem that ethanol-blend fuels, when static and separated, generate galvanic corrosion in aluminum tanks which can cause leak pits in the tanks. If your tanks are leaking through the walls, it's time to remove and do a TIG-weld repair, or raplace with something new and non-aluminum (as an example; using common racing fuel cells relocated to suit the car); let a pro do these repairs.

My only action for the inside of the tanks was to vacuum them out to minimize any particle accumulation. IT IS NOT A GOOD IDEA TO VACUUM OUT THE TANKS UNTIL YOU ARE SURE THERE ARE ABSOLUTELY NO FUMES SINCE A VACUUM CLEANER WILL HAVE AN ARC SOURCE IN ITS MOTOR!!! So, to do this, well after I removed all the gas and removed the hoses (RobzWorkz #8a), placed the gas in a proper container, and moved it away from the work area, I allowed my open tank system to ventillate for over a week. Afterwards, ABSOLTELY NO GAS SMELL WAS PRESENT! I then set up my shop vac to BLOW AIR and sent air into the tanks first...this was to ensure that airflow ventillation was accommodated; then, I DID NOTICE A GAS SMELL!!! Thus I continued to flow air through the tanks (each individually) until no evidence of gas smell was present. Then, and only the, did I feel ready to vacuum out each tank.

For both the blowing and vacuuming processes, I duct-taped a heater hose to my shop vac hose so I could get deep into the tanks via the filler hole (for the left tank only) and the crossover lines (both tanks). See pics 2 and 3 for this.

After I was satisfied that I vacuumed the tanks enough, I then repainted both tanks and the crossover pipe. I used Rust-Oleum Silver paint. Let cure for at least a day before installing hoses, etc. At this time, I did not loosen the tank mount straps; I painted arount them. See pics.

After the repaint, I reinforced all nipples on both tanks with QuickSteel. I believed this to be essential since the tanks were old enough and the aluminum joints weakend by age and stress. It is well-known that aluminum has a memory of strees, so reinforcement is advised. After the QuickSteel cured, I touched up the paint. BTW, the best way I found to finish up the QuickSteel after applying to a surface, was to use rubbing alcohol so smooth out the joints. I used 91% rubbing alcohol.

Also, as a reminder, I had removed my AC system, including the compressor. This can be seen in the pics.
Image Unavailable, Please Login
Image Unavailable, Please Login
Image Unavailable, Please Login
Image Unavailable, Please Login
Image Unavailable, Please Login
Image Unavailable, Please Login
Image Unavailable, Please Login

 

As mentioned before, the fuel return tube/nipple had broken off as I removed the hose. Thus I needed to repair and reinforce it. I used available brass tubing and QuickSteel to do the repair (no heat...if you decide to use heat, like with TIG welding, remove the tanks altogether). I tapped in the repair tube into the original tube. After installing the repair, I reinforced with more QuickSteel as with the other tubes. This is seen in the last pic; sorry for the blurry pic, but it's well-reinforced.

Please note, the pics for this effort show me using a hose which I chose not to use in final; I used the SS braided hose identified in the hose measurement diagram as the final replacement. I was leary of using a lower pressure hose for a return line which could have significant pressure in it albeit at a transitional rate (testing will be done later).

To prepare the hole in the tank, I had to use a tap (thread drill) to optimize the opening. I also used a needle file to finalize the hole in the tank.

Also, well after my repair, I researched the propensity for the brass aluminim joint to cause galvanic corrosion. I found that the potential difference for brass to aluminum was at the allowable limit of 0.5VDC. Steel tubing would have been a better tube choice. The good thing is that I'm aware of this and can often check this line at the tank for corrosion and repair again if needed. Even with the potential, galvanic corrosion in a sealed system with fuel vapor is minimal (the tanks are vapor sealed when the cap is tight and the evap lines are in equilibrum).

After making the repair, I vacuumed out the tanks again to remove any debris. Please review the concerns in the earlier post. I also touched up the silver paint.

It's very strong now, but I will be watching it.
Image Unavailable, Please Login
Image Unavailable, Please Login
Image Unavailable, Please Login
Image Unavailable, Please Login
Image Unavailable, Please Login
Image Unavailable, Please Login
Image Unavailable, Please Login
Image Unavailable, Please Login
Image Unavailable, Please Login

 

When I removed and inspected my upper crossover vent pipe (the aluminum one; not the fuel pressure pipe or the braided line), I noticed several pits, some even corroded all the way through. I thus chose to repair and reinforce. I did this using QuickSteel and repair aluminum panels cannibalized from other aluminum tubes. I uses a hammer to tap into shape prior to fusing. I held in place with duct tape until fusing was complete. After making the repair, I reinforced the y-tube section with QuickSteel and painted the entire pipe (the final paint was silver and is not shown...sorry). Installation with my hose of chouce was easily done from under the fender wells. I did very little hose tightening from above.

Also, since removing the line clamps from the firewall was such a chore as to be characterized as an impediment to further repairs, etc., I chose to install an aluminum U-channel to the firewall which I would attach hoses and lines with heavy-duty plastic straps. The OEM attachments become embedded into the flimsy firewall metal, which caused the bolts in my 328 to become unreachable.

The last 5 pics show the aluminum U-channel installed, the last three of which are after other hoses are installed.
Image Unavailable, Please Login
Image Unavailable, Please Login
Image Unavailable, Please Login
Image Unavailable, Please Login
Image Unavailable, Please Login
Image Unavailable, Please Login
Image Unavailable, Please Login
Image Unavailable, Please Login
Image Unavailable, Please Login
Image Unavailable, Please Login
Image Unavailable, Please Login
Image Unavailable, Please Login
Image Unavailable, Please Login
Image Unavailable, Please Login
Image Unavailable, Please Login
Image Unavailable, Please Login
Image Unavailable, Please Login
Image Unavailable, Please Login
Image Unavailable, Please Login

 

Now to install the SRI fuel tank hoses. These went in very easily. The only trimming I had to do was for the fuel filler, which I removed about 1/2" from one end. All the SRI clamps are excellent.

After cleaning, I reinstalled the filler neck metal piece first. Spend enough time to properly install the big rubber grommet before bolting into place.

Then, I loosened all the tank straps on both right and left tanks. This allowed for flex in the installation process. I installed all hoses and refrained from tightening the clamps until all hoses were in place. For me, the best installation lube was spit (yep...that stuff from your mouth...it works well for installing hoses). The hoses were free to flex and loosen as I re-tightened the tank straps and the lower crossover mount. After I was satisfied that I could no longer adjust the tank straps and crossover clamp, I clamped the SRI hoses into place. Notice that there is some "stretch" look in the SRI lower crossover lines. These hoses are not stretched as they appear to be (there is some press-flex there), only that the crossover line is not exactly in-line with the tank attachments...there is a displacement which is apparent in the look of the hoses. I welcome any feedback on aligning the crossover pipe and the tank pipes. BTW, when tightening the fuel tank straps, be sure that there is room between the other pipes and the tanks and straps; a good example is the coolant line on the right side in front of the timing system.

Also, as part of this effort, installing the metal crossover vent line was done. There are two hose lines on the left side, one for the tank and one on the filler neck. The right tank only has one hose attachment for the vent line. I was able to gain access to the clamps from under the fenders.

I also reinstalled the heat shields removed from the lower left side, after cleaning, of course. I can't identify the rough-looking stuff...my guess is that something with flux-chemistry got spattered onto the shield.

Also, as part of this, I inastalled the filler overflow line from the body tube next to the filler tube having the gas cap.

The pics don't show the process as detailed as I would like, but the idea is there. Spend ample time and don't take anything for granted.
Image Unavailable, Please Login
Image Unavailable, Please Login
Image Unavailable, Please Login
Image Unavailable, Please Login
Image Unavailable, Please Login
Image Unavailable, Please Login
Image Unavailable, Please Login
Image Unavailable, Please Login
Image Unavailable, Please Login
Image Unavailable, Please Login
Image Unavailable, Please Login
Image Unavailable, Please Login
Image Unavailable, Please Login
Image Unavailable, Please Login
Image Unavailable, Please Login
Image Unavailable, Please Login
Image Unavailable, Please Login
Image Unavailable, Please Login
Image Unavailable, Please Login
Image Unavailable, Please Login
Image Unavailable, Please Login
Image Unavailable, Please Login

 

Some will have noticed in earlier posts, that the new fuel pump was installed in the hose replacement pics. This is because I replaced the fuel pump and accumulator first, before installing tank hoses. However, my instruction to you would be to be certain the tanks are without gas fumes since the pump is electrical and could generate sparks. I did fully vent and clean the tanks first; however, for better process description in this thread, I felt that continuing with the tank effort to completion was more pertinent.

In this effort, I replaced the fuel pump and accumulator altogether. The new fuel pump included a check valve as part of its assembly and was rated for the pressure required for operating the CIS. The first 3 pics show the old pump adn accumulator as well as the removed (assumed) OEM check valve.

As an aside, I have seen "much ado" about the fuel pump check valve and its outrageous pricing. Recall that the CIS distributor is designed to regulate pressure based on having a macroscopic fuel pressure and flow rate presented to it. It is my opinion, that, if the CIS pressure testing is within specs, including the rest pressure, that the provided check valve is proper enough. I, therefore, defer to the CIS pressure test in the final thread of this series to declare if the OEM check valve is as desireable as the zeitgeist has revealed. If no apparent concern is realized, I will certainly offer up my used OEM check valve for the very-competitive price of $100US.

The fuel pump and accumulator are mounted as a removable assembly. Thus improvement necessitated disassembly and cleaning; then pump and accumulator installation and remounting into the car. It's fairly straight forward to do and quite rewarding to see all cleaned up. As stated, I also replaced the rubber mounts with new.

As is recommended, it is a good idea to replace all crush washers in the rebuild. This I did with one annoying disappointment. I went to NAPA to gather crush washer preplacements; they had the correct I.D. and O.D. but were too thin for the fuel pump exit banjo fitting to fit with pressure. I thus had to use three of these stacked together. To ensure good sealing, I used INDIAN HEAD varnish to stick together prior to tightening. Also, for any of the fitting threads and surfaces, I applied varnish to aid in sealing the system and prevent future loosening. When all was done, I was satisfied with my assembly. I will keep watch on this as I begin my testing...I do not forecast any problems as yet.

Electrically speaking, the FP has 2 terminals, one which is a ground line attached to the frame at a mount stud, and the other which is attached to the wiring harness. My FP included nuts, washers and boots for these terminals. I fabricated a 14gage ground line as can be seen prior to installation.

You can see that, after installation, I installed the accumulator-to-filter hose. This is only temporary since I will be rebuilding this hose. Thus it's "just for looks." Also seen is my new fuel filter. As previously stated, the new filter was of a smaller diameted and will need a smaller mount. Also, I had to use the fittings in a different order to properly affix...thus is the fun of the aftermarket. Of course, the upper pressure line was attached to the top of the filter, as expected. The filter and pressure line will be properly fitted later as part of the CIS/hose rebuild effort...since they all need to line up.
Image Unavailable, Please Login
Image Unavailable, Please Login
Image Unavailable, Please Login
Image Unavailable, Please Login
Image Unavailable, Please Login
Image Unavailable, Please Login
Image Unavailable, Please Login
Image Unavailable, Please Login
Image Unavailable, Please Login
Image Unavailable, Please Login
Image Unavailable, Please Login
Image Unavailable, Please Login
Image Unavailable, Please Login
Image Unavailable, Please Login

 

As said in RobzWorkz #8a, some of my replacement hoses can be considered as "over-kill." This is certainly true for the (usually dry) accumulator vent line to the left tank. All that's really needed for this line is a good quality fuel injection hose and some good clamps. However, to have more fun (and since I ordered the hose used as part of my larger order), I replaced with a 1/4" AQP elasomer multi-braid hose. This is excellent hose, but the bend stress applied to the accumulator vent was great enough for me to split the hose and install a 1/4' metal pipe, pre-bent to allow for no real bend stress. Also, as with the tank nipples, I reinforced the accumulator vent line with QuickSteel (I did this on the tank and the accumulator itself). I used new clamps for this as well.

FYI, cutting the multi-braided hose (this had SS inner weave and external cloth weave) is easy to do with a Dremel cutoff wheel. However, I improved the method by first coating with Ultra Black RTV at the cut line and allowing it to cure hard (at first, I used large heat shrink as well; although it looked better, it is not advised since it can come off...the RTV is plently strong and minimizes weave mess). Thus the pics show a failrly nice cutting of hose. After cutting, blow through the hose with pressurized ait and clean before installation.

The same method was used for the braided hose of the CIS fuel return line to the right tank.

Thus completes RobzWorkz #8b. Now onto the evap system.
Image Unavailable, Please Login
Image Unavailable, Please Login
Image Unavailable, Please Login
Image Unavailable, Please Login
Image Unavailable, Please Login
Image Unavailable, Please Login
Image Unavailable, Please Login
Image Unavailable, Please Login
Image Unavailable, Please Login
Image Unavailable, Please Login

 

My car had a bad gas smell since weeks. After checking underneath, suspect areas of the right tank could be seen.

We disassebled the right wheel, and disc (otherwise the gas tank cannot be removed). We found the a water pipe was touching the tank, which lead to wear of the tank aluminium. We think this hose was touching the tank because the clip was assembled in the wrong way. In order to give equivalent clearance around the water pipe, we modified the angle of the clip (from 90° to ~45°).

The leak could be confirmed by putting some air pressure in the tank : small bubbles appeared in the water we put in the suspected area. We put some aluminum epoxy in this area to try to make it tight. I hope this will not be damaged with the gas. We changed some gas hoses. Majority had not bad looking, but as it was disassembled, better to put new ones.
I will fill up next week.

Going to my friend for repairing, I discover another issue : right window does not lift up anymore : it can go down but not up at all (it is not very slow, it don't lift up at all).
- Using for right and left buttons : same issue.
- Putting directly 12 V at the motors wires (from external battery) : motor goes down and up very quickly.
- checking fuse and relays : all are OK.
- measuring the voltage at the button : +12 V when down, 0 V when up...

So, we continue to investigate the thing...
Image Unavailable, Please Login
Image Unavailable, Please Login

 

Great Catch, Denis!

This is definitely a place to check for abrasive tank wear.

So, as a note, it seems that, for the right tank, removing the wheel, wheel well and air boxes, and brake disk will allow tank removal.

Can't say for left, but I expect more suspension components will need to be removed than just the wheel and brake disk alone.

When cleaning and inspecting my tanks, I loosened the straps and positioned to minimize hose flex. When tightening, I positioned to minimize chafing, especially in the coolant pipe area.

 

What torque specs are you using to 'seal' the copper crush washers in your various fuel pipes?

 

No specs have been seen, however, hand/wrist tight is fine...definitely tight. If you do by hand with a nominal box wrench (of course, you will be using the open-end), you will be fine and not over-torque and strip threads; also, use Indian Head varnish to help seat and keep tight on all surfaces. Be sure the crush washers are flush with the hose/pipe surfaces when tightening.

BTW, I re-did the crush washers on the fuel pump. I do not think stacking them was appropriate, so I replaced with new, thicker ones. I still used the varnish to stick together. Definitely seek proper thickness washers.

The pic shows inappropriate washer stacking at the top joint of the fitting.
Image Unavailable, Please Login

 

I'm curious how you managed to 'need' additional washers when, in an earlier picture, you didn't....something changed?

BTW, there are torque specs for these fittings - I just can't decipher which spec to use for which fitting (I'm re-building the same assembly currently):
Image Unavailable, Please Login

 

The washers I purchased from NAPA were too thin for the fittings to compress the washers, so I stacked them. Thus, later, when I found more correct washers, I replaced altogether with the thicker ones.

Thanks for finding the torque specs. I forgot that I actually had seen them in the 328 Tech manual before; when reviewed, I assessed these values as commensurate with expected torques applied by hand using common wrench lengths and tight application (tight to seating the fittings at the stop...no give=crushed washers). The wrench size will basically dictate the added force. Thus the higher values will corresponds to using longer wrenches (bigger fittings). Just be sure it's tight and has varnish, and no leaks present. This should help with deciphering the fittings as listed. Also, don't forget to use two wrenches where two hexes are available at the fitting to minimize bending.

 

Although that's an interesting "one size catches all statement", I find this advice pretty dangerous. There are many well-intentioned folks who don't have any idea proper torque figures are supposed to be. Or, perhaps more importantly, don't have the experience to know, or feel when a fastener is indeed tightened correctly. In that we're working with the fuel system, I'd rather know than guess "based on the wrench size."

Using your thought process, help me understand how with two fasteners of the same thread pitch and size (8mm) have two different torque values, per the chart provided. A typical 8mm DIN-spec fastener will have a 13mm hex size. Using SAE, ASTM or DIN standardized torque tables, a torque value of 16-22 lb-ft. would be specified. Yet, Ferrari/Bosch is specifying less!

There are many variables which determine correct torque; not limited to fastener size. Making broad-based statements that wrench size/length is the determining factor for proper torque is dangerous particularly considering that your audience may perceive your advice as factual. It is not. In your defense, I will agree that finding the actual torque figure for many Ferrari fasteners can be daunting. However, the lack of immediate (technical data) availability is no excuse to guess.

 

David, in general, I agree with you...however, this is way too analytical. Apart from the top hex bolts on the distributor (not for the line to the WUR, though), the WUR and the cold start injector (all less than 10ft-lb), all torquing hexes will need special access (crow's foot or hard to access with a socket). This will add significant error for most DIY'ers. Basically, the DIY'er will torque fairly tight and be ok. If leakage (or smell) shows up, tighten some more. The use of crush washers allows for error on the tighter side.

The list Michael provided from the WSM (post 11) shows that only two of the torques are "special": 1) the union with the filter on the distributor (higher torque for bolt size), and; the thermo time switch (low torque for bolt size). All the rest are nominal. The general idea for all the fittings with crush washers is to seat the washers. For the hoses, it is essential to torque fairly hard to prevent leakage; thus with new hoses (new fittings), the torque specs should suffice, but when reusing, torque a little more to stop leakage...this would be a judgement call on the mechanic.

BTW, for anyone using a torque wrench with a crow's foot on the in-line fittings (flex hoses, filter, etc), be sure to adjust the torque setting down to accommodate the added length (current torque = greater torque with longer lever; therefore, lower the torque setting).

Also, BTW, I havent posted my efforts on the CIS and WUR stuff. I will do so after pressure testing the hoses and doing the CIS cold pressure test. I will share the torquing effort there.




Also, as a reminder to all DIY'ers, an implied, and often necessarily needed to be presented in writing, would be the following GENERAL CAVEAT:




****************************************CAVEAT FOR DIY MAINTENANCE***************************************

ALL DIY MAINTENANCE IS TO BE CONDUCTED AT THE RISK AND AGREEMENT OF THE ONE CONDUCTING THE WORK AND THE OWNER OF
THE VEHICLE. ANY EXTERNAL ADVICE NOT PRESENTED IN FACTORY-AUTHORIZED PROCEDURES AND BULLETINS SHOULD BE TAKEN AS
EXTERNAL AND SHOULD THUS BE SCRUTINIZED AS SUCH.

****************************************CAVEAT FOR DIY MAINTENANCE***************************************



This applies to anything posted by myself as well. I may have to make this my signature and/or get FChat admin to post in plain sight of all.

 

Hey Rob,

In some of your pics (from all the hoses threads) you show making some small tubes to take care of some sharp bends the more modern hose simply can't make. I get that. My question is about the the tube preparation which isn't shown. Did you bead the ends to prevent the hoses from slipping off, and if so what tool did you use? I've been looking for a good beading tool which can go down to 1/4" ID for a different non-automotive project, but haven't found one that doesn't cost big $$$.

Concern: If you didn't bead the ends I assume you're going to go back and double-clamp or use some proper hose fittings to make those bends? Prototyping in-place is one thing, but now that you've got the lay of the bends you need to make what are your plans here? Just concerned for your safety.

 

hmm.. I always wondered, could you, as the tubes are small diameter, put them in the brake flaring tool and do just a little bit of a flare to get a passable bead? Larger diameter it's a problem but sizes up to 5/16' should be no biggie.

Lester

 

I considered that, but the edges would be pretty sharp. There is a hand-beading tool sold by aircraft tools suppliers, but again it's big bucks for the one-time project I'm working on.

 

The only systems I used the tubes were for the evap and the accumulator overflow, both non-pressurized. So, no flaring/beading here. Also, the AQP inner tube was hard enough to make difficult with raised material (flared lip) during installation with the small hose diameters. Larger hoses were fine. I simply clamped each hose end. This was really no different from the OEM setup anyway (see tube edges on fuel tank and vapor separator, etc). The finished arrangement is quite strong.

Again, for these systems, my hoses could be considered "overkill." Using the alternatives (high quality fuel injection hose) would suffice and require no tubes...just clamps.

Thanks for your input.

 

Flaring would make a "sharper" edge. Installation of exact diameter hoses (mine would be more difficult) as well. Double flaring would be smoother. Recall that flaring is used for fitting placement, not hose connections.

When I cut my (pre-bent) tubing to length, I used a cutoff wheel and sanded the ends before installing. I used steel line commonly available at auto parts stores.

 

I've added this thread to the tech/DIY sticky.

 

What I was thinking but didn't express well is that the first part of a single flare causes the bubble flare. Do that gingerly and you end up with with a raised gentle ridge without the need for new tools. Not suggesting that you need them just that it should work..

Lester






When I cut my (pre-bent) tubing to length, I used a cutoff wheel and sanded the ends before installing. I used steel line commonly available at auto parts stores.[/QUOTE]

 

I didn't think of that, Lester...perhaps that is a good idea. Would you care to share with pics? (Flaring 101)

 

As a professional tester I'm skeptical by nature, so please take this as additional feedback with no disrespect intended: The accumulator connection has a bead on it as does the one on the fuel tank at the other end - at least on my car it does. One might guess that's there for a reason?

 

This is the best that I can do on short notice.. I use the Fedhill flaring tool and this is the first thing one does to begin a double flare. It's the DIN/ISO bubble flare pic.

http://www.fedhillusa.com/webnuts/common%20flares6.pdf

Lester

 

Jeff, this is excellent feedback. BTW, in my opinion, these DIY posts require the form of "peer review" that is present here. Without it, we would not bring each others' skill sets to reasonable quality. We also need to feel free to express our concerns openly and teke and dish out criticism. If you read enough of my posts, you will see that I receive valuable inputs and have made corrections as well.

In general, I feel that the FChat community respects and enjoys my efforts; this is why I have enjoyed sharing my work here. The community really enjoys all the pics! Just please realize that my priorities are safety, excellence and affordability, in that order.

I bought my Ferrari to challenge my abilities and use viable alternatives to save cost; otherwise, I would just keep on wrenching on the Japanese and American cars I've worked on already.

FYI, my accumulator did have a lip and posed extra difficulty installing the AQP hose; I did get installed though. I don't recall the tank tube having a lip. My recollection is that all tank tubes for the small hoses were straight on my 328. Tube lips existed for the crossover lines and the fuel return hose.