pressure testing rubber bungs

[Piero]

Hi Guys

Having completed a petrol tank extention on my classic car,

its time to test it,

I have made air tight all the smaller outlet holes, and for now, I pressed a rubber mallet into the petrol tank filler pipe.

I pumped up the tank with air, via a cycle pump, and tested for air leaks with a damp cloth.

I must say, at the recommended 3psi, some pin holes didnt leak enough to show up with a wet/damp cloth.

I think I have found all of the pin holes in my tig brazing (three)

BUT before if paint said tank, and fill with petrol,

I wanted to step up the game a do a proper pressure test.

I bought a 0 to 1 bar testing gauge, and will be fitting that to the set up.

I intend to leave it pressurised for a good few days, and take readings.

Here is my problem, I dont trust the rubber mallet in the filler outlet pipe.

I dont want a bodged plug either.

The best I have found is a expanding plug as used in testing pools (swimming/relaxing) and there pumps for pressure.

a tad on the rich side at £15 pluss the dreaded for a one off rubber plug.

was hopping some one might now of a better idea????

 

[addjunkie]

Taking it up to 1 bar 14.7 psi, may do more harm than good, it doesnt take much pressure to deform a vessel not designed to hold pressure. I know this as over the years in my day job i have been involved in more pressure tests than i care to remember up to 900 bar. I did once see an atmospheric tank that wasnt atmospheric deform at 1/2 a bar. I was part of the investigation team to understand the failure.

If it were me, id stick at the hand pump 3 to 4 psi and cover everything in soapy water, check regularly over a day or so.

If its ok, then wash out, and coat the inside with por 15 petrol tank sealer. Follow all the instructions dont cut any corners.

 

[Piero]

thanks for that.

although I didnt intend to pressure the tank to one bar,

I bought a 0 to 1 bar gauge as it was the lowest bar reading I could get, its a 4 inch with lots of incriments,

 

[rtbcomp]

Get a rubber bung with a hole through the centre (Ebay?) and use a nut & setscrew with a couple of penny washers to squeeze it so it expands into the pipe.

I'd fill the tank with water, pressurise and cover the welds & seams with french chalk, if it leaks you'll see the damp patches.

 

[Piero]

its a very clean steel petrol tank, no rust on the inside in 40 odd years,

wouldnt be happy to add water

 

[gordon stephenson]

Pessure test as per rtbcomp, then soapy water test with a nice solution of washing up liquid, even the smallest leak will be self evident if left pressurised for a short while, Done lots of testing in industry using this method.Just like checking for leaks on a tyre where it meets the rim,

 

[Piero]

hoping to get advice in were to get a proper pressure testing plug

 

[pedrobedro]

I made a tank and to test it I blocked the filler hole, put a pipe on the outlet and blew in it with my mouth. I tried to blow more air in and couldn't so just assumed none was coming out It was on the car for years and never leaked.

 

[Piero]

thats one way of testing it lol

 

[hotrodder]

Probably won't exist/be easy to find in typical filler neck sizes but https://www.google.co.uk/search? silicone hose caps (typically 4 - 32mm) are a simple/cheap way of blocking off inlet/outlet tubes for leak testing stuff at lowish pressures. For larger dia stuff and/or hydro testing at higher pressures i usually make my own... cut a short piece of tube, weld a cap on one end and roll a hose retention bead on the other. A short length of hose and a couple of jubilee (or better still mikalor) clamps and you're good to go. A convenient chunk of bar stock is another way, can either drill and tap the cap (if it's thick enough) or the bar for a BSP schraeder valve where necessary

For fuel tanks i fit a 0 - 15psi gauge (0.25psi resolution) into the plate i use to blank the sender hole

and put a couple of psi in it through one of the capped fittings. If it hasn't lost pressure after a tea break then get on with something else for a few hours, if it has and you're the glass half full type then the soapy water gets brushed over all the capped fittings before you start looking for pinholes in welds etc
A few days is overkill and possibly more likely to lead to paranoia due to the permativity of rubber hoses, air very slowly seeping past capped fittings etc. Especially if the filler neck is a one piece deal as the bayonet fitting for the cap on older stuff often isn't designed to seal 100% as the cap either doubles as a tank vent or pressure relief valve

 

[Pete.]

I use leak detector spray. It shows up the very very tiniest of leaks at 5psi. 1 bar will blow it out of shape, I did that to an aluminium motorbike tank this year.

 

[Piero]

Hotrodder
Sorry to have to say.

But I zoomed right (100 × zoom) in on the welding on the photo of the tank (probably one you done)

I took a micrometer out. And did found one puddle of weld a mega mega fraction different to the rest

 

[Bill Edwards]

Washing up liquid reveals the majority of leaks within seconds. An hour will show any tiny ones, then call it done.

I'm surprised that nobody has mentioned dye penetrant for testing for pinholes.

 

[hareng]

Even better if you can, submerse the lot in water, slightest bubble will show instantly.

 

[brightspark]

Even better if you can, submerse the lot in water, slightest bubble will show instantly.

ever tried pushing a large tank full of air down in water

 

[hareng]

Yes best way to do it, no areas go unchecked.