Pressure testing fuel system with no fuel

[rossman]

I'm getting close to my first start up with my new fuel system. It has never had fuel in it. I'm planning to pressure test the system with compressed air prior to adding fuel. Here is what I plan to do.

 

1) Remove the NPT fittings from the tank

2) Cap the return line

3) Add an adapter fitting for my air compressor to the feed line

4) Attach the compressor to the feed line adapter

5) Verify that the compressor regulator is turned all the way out (no pressure on the outlet)

6) Turn on the compressor and let the tank pressure build until it shuts off

7) Slowly turn the regulator knob until fuel line pressure gets to 60 PSI at the fuel rail gauge

8) Check for air leaks

9) Depressurize the system by opening the air compressor regulator knob

10) Disassemble the test setup

11) Reconnect the NPT adapters to the tank

 

Does this sound reasonable? Suggestions welcome.

 

Cheers,

Ross

Edited January 11, 2012 by rossman

[Tony D]

Attachment of a ball valve on the inlet, and a high resolution gauge on the end of the return line will allow you to hold air in the system and check for slow bleed down.

 

Minuscule leaks can be found using either "Snoop" leak detector fluid, or a spray bottle of water with dish soap dissolved in it.

 

Leaving it overnight and not seeing any pressure drop would give me absolute confidence in the fuel system integrity.

 

This sam method is used to test hydrocarbon transport containment piping before putting it into service.

[rossman]

Thanks for the input Tony. That's what I will do.

[EvilC]

You are better than me, I just double check all connections and fill her up - prime and listen for any leaks and smell of gas!

 

Good luck!

[rossman]

Better? Not really. I'm just not willing to accept as much risk

 

I'm paranoid that I'm going to have leaks since this was my first time bending/flaring stainless fuel line and working with AN fittings and hose ends.

[neotech84]

"Snoop" is what we use on aircraft hydraulic and fuel lines and it works great, but it is pretty much soap and water as TonyD stated...... So I guess this is a ditto post.

 

Any way hope you are leak free!

[cygnusx1]

 Just don't look for fuel leaks with a match.  I mean, it works, just works too good.  Clive, your fuel leaks are diagnosed by the EPA.    

[EvilC]

Yeah, my last build I had a nicked o ring on the FR. You know what that looks like at 60 psi! lol

 

 

Ross have someone there with you when you do the first true start up.

[Tony D]

Actually, Snoop is FAR "bubblier" than any dish soap mixture I have ever tried. I don't know what they put in it, but it shows leaks where Dawn/Joy/Palmolive don't bubble at all. I can't explain it, that is just the way it is...

[rossman]

I pressure tested the fuel system this morning. As I feared, I have leaks at the stainless tube flare to male AN fittings. I think the tubing is too hard for my tooling. I can still see extrusion marks in the flare sealing surface. I'm going to try aluminum conical seals like these instead of redoing all the flares, which will probably yield the same results. Anybody have experience using these? I assume aluminum is acceptable for fuel with 10% ethanol.

Edited January 22, 2012 by rossman

[Tony D]

You need a flat sealing face. The conicals will help as they "embed" and turn it into a gasketted joint, and not a metal seal.

 

On hydraulics running 6kpsi I would sometimes lap the tubing to smooth it if I could see divots or imperfections.

 

I did this with #600 or 400 wet-or-dry cut in an arc to fit on the face of a spare fitting, and then supergluing it in place. I would dab a little oil on the face, tighten the fitting to the tube with the flare nut lightly with my fingers to align it, then carefully "sand" my way around the face checking occasionally until I got a "clean" tubing flare face.

 

It worked really quickly on hardened aluminium aircraft tubing, it took longer on steel and SS line flares. Just move around evenly and you may find you don't need the seals at all. A metal-to-metal fit is preferable to a gasket being in there.

[rossman]

Thanks Tony, that seems to have worked! I lapped half of the tube segments and those fittings passed the soapy water test and overall the leak rate slowed considerably. I kinda feel stupid though. I know the surface must be smooth but I ignored it for some reason. Oh well, another lesson learned.

 

Oh, and I switched to 320 grit sand paper when became apparent that it was going to take all day with 400 grit and finished off with 600 grit. I have several spare fittings and made one with 320 grit and one with 600 grit.

Edited January 23, 2012 by rossman

[rossman]

Here is a picture of the tube, as flared. Those grooves are big enough to catch my fingernail. Big surprise it doesn't seal eh?

 

 

Here is the lapping tool I made based off of Tony's description. This one has 320 grit bonded to the sealing surface. I worked on one flare this morning for ~20 minutes and I can still see an feel the extrusion marks. Starting with 220 grit then moving up to 600 grit might make this process a little faster.

 

Edited January 27, 2012 by rossman

[pete]

Here is a picture of the tube, as flared. Those grooves are big enough to catch my fingernail. Big surprise it doesn't seal eh?

 

 

Here is the lapping tool I made based off of Tony's description. This one has 320 grit bonded to the sealing surface. I worked on one flare this morning for ~20 minutes and I can still see an feel the extrusion marks. Starting with 220 grit then moving up to 600 grit might make this process a little faster.

 

[pete]

Hey Ross,

I can give you a hand with that AN tubing if you want. I have access to a nice flaring machine. Looking at the photo of your flare I suspect you have some issues with the extrusion marks on the tubing ID transfering to the flared surface.This is more of a problem with some imported tubing than domestic produced. At the smaller diameters (less than 3/4")it is workable. What wall thickness do you have.035"? and are you using 316 or 308ss?

 

Sorry I was unavailable last week,I will be back at work Monday. Pete

[Tony D]

I'll agree-that tubing interior looks terrible! Like their drawing tool had burrs on it. The flaring process should swage that mostly smooth. Are you using a flaring tool with rollers on the cone? The cheaper tools produce a "upset flare" like the more common SAE 45 flare. The better air raft tools have rollers on the cone to roll the flare larger which usually smooths out this kind of stuff. But some stuff you just won't fix. Then you're stuck gluing sandpaper on old fittings and laying on your back 20 minutes at a stretch...

 

I'd caution against using 220. Stick with 320 and slow and steady. It will get the job done. You don't want scratches going all the way around, that can cause leaks as well. You end up takin off maybe too much.

 

I found if I left a gap in the paper, the removed material migrated there in the oil, and didn't load the paper up. Once it loads, it just slips over the tubing surface.

 

Couple of turns, back off, squirt some lube, snug and turn a couple more turns, repeat.

 

On thing is also possible: Loctite 545 (I think) they do make a good hydraulic joint sealant. Like Loctite Green. I'm NOT a fan of ANY compound on what should be a 100% metal joint. That metal joint kept tight will never leak. A gasket or compound can always deteriorate and leak.

 

Just keep at it with your "surface conditioner" with maybe some gaps for removed material and see if that speeds it up some for you.

 

Glad to see you got some positive results. Beats buying something you have to keep using over and over. Those metal cones will go forever now!

[rossman]

Hey Ross,

I can give you a hand with that AN tubing if you want. I have access to a nice flaring machine. Looking at the photo of your flare I suspect you have some issues with the extrusion marks on the tubing ID transfering to the flared surface.This is more of a problem with some imported tubing than domestic produced. At the smaller diameters (less than 3/4")it is workable. What wall thickness do you have.035"? and are you using 316 or 308ss?

 

Sorry I was unavailable last week,I will be back at work Monday. Pete

 

I believe you are correct. Written on the side of the tube is "ASTM/ASME A/SA 269/213 AW TP316/316L 5/8"X 0.035"X20' HEAT TREAT NO: YT090276" and "MADE IN CHINA." I pulled the spec and it allows 10% variation in wall thickness or .004" for .035 wall. The extrusion marks are considerably deeper than that although they are difficult to measure or even see without flaring the tube. The spec doesn't specify a surface finish and actually says very little about the ID of the tube.

 

I'll be working on the car early Saturday morning if you want to swing by.

 

I'll agree-that tubing interior looks terrible! Like their drawing tool had burrs on it. The flaring process should swage that mostly smooth. Are you using a flaring tool with rollers on the cone? The cheaper tools produce a "upset flare" like the more common SAE 45 flare. The better air raft tools have rollers on the cone to roll the flare larger which usually smooths out this kind of stuff. But some stuff you just won't fix. Then you're stuck gluing sandpaper on old fittings and laying on your back 20 minutes at a stretch...

 

I'd caution against using 220. Stick with 320 and slow and steady. It will get the job done. You don't want scratches going all the way around, that can cause leaks as well. You end up takin off maybe too much.

 

I found if I left a gap in the paper, the removed material migrated there in the oil, and didn't load the paper up. Once it loads, it just slips over the tubing surface.

 

Couple of turns, back off, squirt some lube, snug and turn a couple more turns, repeat.

 

On thing is also possible: Loctite 545 (I think) they do make a good hydraulic joint sealant. Like Loctite Green. I'm NOT a fan of ANY compound on what should be a 100% metal joint. That metal joint kept tight will never leak. A gasket or compound can always deteriorate and leak.

 

Just keep at it with your "surface conditioner" with maybe some gaps for removed material and see if that speeds it up some for you.

 

Glad to see you got some positive results. Beats buying something you have to keep using over and over. Those metal cones will go forever now!

 

I used a Rigid hand flaring tool that has a hardened steel flaring cone, eccentrically mounted in needle bearings. I had to put a cheater wrench on it to the the cone to bottom out on the flare. The tool is probably not the best choice for this tube size/material.

 

Thanks for the caution but I already went at it with the 220 grit paper. There doesn't appear to be any scratches under a 20X magnifying glass. I'll try putting a small gap were then ends come together and finish it up with 600 grit for good measure.

 

Agreed, no gasket/sealant is preferred. Those options are plan B if this doesn't work.

 

Thanks for your help guys. This has definitely been a learning experience for me.

Edited January 27, 2012 by rossman

[rossman]

So far, two of the flares still didn't seal after using the "surface conditioner." I think the fitting was bottoming on the inner non-lapped surface causing it to leak. Working the inner flared area by hand fixed the problem.

 

On a positive note, the 1/2" return line is much smoother. It took just a couple of minutes to get one of the flares smooth.

 

Four more flares and I will hopefully be done. Of course I saved the hardest, access wise for last. I may have to drop the differential to get to them.

Edited January 29, 2012 by rossman

[Tony D]

Of course, you make such good headway, get excited and start grabbing fittings...

And then when you think you're done...you realize there is ONE left...the one you never see, the one tucked up there so high you forgot all about it. The one you tightened FIRST because it was the biggest beyotch to get to and after that it was all downhill!

 

In this case, 'rectification is not the reverse order of assembly'!!!

[rossman]

Well, the saga continues...

 

All the flares are smoothed out to what looks like a smooth shiny surface under 20x magnification. I pressurized the system (10 flares, 10 unions, 6 injector o-rings, 6 ORB fittings, 4 NPT to AN fittings, 12 hose end fittings) this morning to 60 PSI. There is still a slow leak (compared to what I originally had). The pressure degrades approximately 1 lb in 4 minutes. Soapy water doesn't seem to detect any leaks at the fittings. I ordered a bottle of Snoop to see if that helps with detection.

Edited February 2, 2012 by rossman