Hi, welding on flat sheet- warping question...

[topcat]

Hi all, first post here. I am pretty much self tought with TIG welding, but I feel my level is pretty good, the only thing that lets me down is not being quite perfectly neat but I guess thats where more practice comes in! Here's a quick phone pic so you can see where I am level-wise!

I'm making a square flat fuel tank at the moment, 600x600x130, which needs to have a filler neck attached near one of the corners, and a fuel sender in the centre. To get over too much welding in the middle for the sender I have turned a flat ring on the lathe which I intend to fix onto the inside of the tank (maybe rivet) which I can bolt the sender to, sandwiching the tank skin along with a gasket so that is not a problem.

The problem I have is how to fix the filler neck. It will be a short piece of 1.1/2" tube about 3" in from the corner, I'm considering welding a flange onto the tube and riveting it to the tank because I'm worried that if I weld it, it will distort the tank top so badly it will be scrap, does that seem like a viable option to anyone or is there an easy way round it? I have swaged 3 lines in the top and bottom of the tank for rigidity, but nowhere near enough to stop warping I'd have thought.

I also have a tube of stixall (some type of mastic stuff that my local fixings place supplied) which amongst other uses says it is completely petrol resistant, but I don't trust glue that much!

All thoughts and advice welcome!

 

[the snooper]

you could weld it in stages allowing it to cool down between welds so there wont be too much heat at any 1 time

 

[^neo^]

It will distort if you weld it to some degree, but as above, weld it in 3 or 4 hits letting it cool and just tap it back into shape. Ideally you should have put the sockets on first then you would have a chance of less distortion.
Did you stitch weld it on the inside too, or just the outside?

 

[Engineer]

Whats the thickness of the material used? As long as its not overly thick you will easily be able to weld in filler neck (these are available on Ebay complete with cap), and a flange for fuel sender. Dont try to bodge these in with rivets or glue, as you will find that they will leak almost certainly. Fittings need to about 3mm thick for filler neck, and 5mm for sender unit, and should be welded in using a 1.6mm tungsten and 4043 filler.

 

[topcat]

Great thanks for all the replies already, I am using a 1.6 tungsten, can't remember the filler

It is 1.5mm sheet, I have folded up the base and 4 sides and welded both inside and outside of those corners, I have made slot together baffles (like a noughts and crosses board) which I will tack in place, and small sumps for underneath for fuel pick up and to stop surges even more.

The lid will be welded all the way round, and I will drill about 6 7mmish holes where I can weld through to tie into the baffles (I folded a 1/2" lip on top and bottom edges).

So I have not welded the lid on yet- because I know I should weld the filler neck in first, and I will also weld the sender unit's mounting plate on the inside with about 3 short welds (it's only there for some meat to screw into with M5 fixings (already drilled and tapped)).

I'll roll a bead on my filler neck then (got a piece of fuel hose to go on not directly a cap) and try welding it on in small stages then! Am I best off using low heat which will take a while to warm up and keeping the weld small or should I crank the power up and just do say 4 tacks at a time in 90° intervals slowly working my way around?

edit (sorry for long post)- if I position the filler tube (about 1.6mm wall) about 2.5mm through the lid could I get away with using no filler and using the couple of mm protruding instead?

 

[hotrodder]

You won't get away with autogenous (no filler) welds, there's a few applications were autogenous welds can work with ally but only certain alloys and joint types- the tube will be a 6xxx series alloy and welds will typically crack unless diluted with some filler wire

For sender mounts i typically plug weld them in place...

The bolt holes for the sender take care of fixturing, simply bolt the sender ring in place and plug weld between each of the bolts.

I often weld baffles and the like to the back of stiffening beads- as long as you're welding directly on the bead distortion isn't a problem. Plug welding through the skin as you plan is another good way

Filler neck... passing it through the skin and welding from the inside is the way many radiator necks are attatched. On the plus side need less heat to make the weld on the downside you have to make the weld before fully assembling things i.e. there's less geometry to resist distortion. The later isn't much a problem with rad end/header tanks due to their size and the order of construction

I would weld that in with a fillet weld after the tank is assembled and at least tacked up. Personally i would use three tacks (red dots) like this...



and make the first weld between 1 and 2, then 3 - 1 followed by 2 - 3

 

[topcat]

Thanks, lovely looking tank, do you think it would help if I swaged the filler hole upwards so it is more of a butt weld around the end of the tube where it joins rather than a fillet which would need more heat and time? The shape may reduce the distortion as well..?

 

[hotrodder]

Swaging the hole should further reduce distortion but it shouldn't be a huge problem with a regular fillet weld.
Have a look at this thread... http://www.mig-welding.co.uk/forum/showthread.php?t=18630&highlight=fuel Not the best view to gauge straightness but the filler neck is 3mm wall welded into 1.5mm tank (no baffles connected to the skin). I put more distortion into that tank with the bead roller than i did when welding it up. Similar story with the tank in my last post- bead rolling the top put a twist in it, pressing the relief for the sender put a slight ripple in it too.

 

[topcat]

Thanks again for the advice, I finished the tank and it is not as neat as I was hoping when I started out, but I decided I could have designed it a lot better so just got on and welded it up, I did get a fair bit of distortion, mainly from the small welds through the top skin onto the baffles, which I may not bother with next time.

As said, I would change the design of the construction next time, and have learnt a lot more about the fabrication (for one how many clamps/screws etc I need to put into certain parts to keep it from pulling apart when the heat gets in).

Here's a few pics anyway, it's nowhere near perfect but it is only for my own car so will do the job, when I've finished the car I may take it out and build a new one which should be a lot neater.

Underside:

To be honest the amount of fustration it caused when making it has made me want to do another one having learnt from the bits that went wrong! I'm going to pressure test it with just a couple of psi and some fairy/water solution, as at 600mmsq it's a bit big to put in a water butt!

Thanks again,
TC

 

[Ducati]

Nice work

Out of curiosity what engine / fuel system are you using, carb or injected ?

The only reason I ask is that on some fuel injection systems a swill pot is used / required in the tank to prevent hesitation or cutting out on such as long sweeping bends when the fuel level in the tank is low.

 

[topcat]

It's a single carb system on a rally mini, so I've put the 2 fuel pots (in effect) underneath the tank as in the 3rd picture, they are connected to the main tank by just 2 9mm holes at the rear, so they should be full all the time, and even if one empties by means of a very long corner (unlikely) the other will be full. There is a small facet pump going infront of each one and feeding a Weber DCOE 45.

I tried to plan it so there will always be at least 1 surge tank full with either heavy breaking, accelerating and long corners.

 

[Engineer]

Will the tank hold enough fuel for serious competition use?

 

[Ducati]

It's a single carb system on a rally mini, so I've put the 2 fuel pots (in effect) underneath the tank as in the 3rd picture, they are connected to the main tank by just 2 9mm holes at the rear, so they should be full all the time, and even if one empties by means of a very long corner (unlikely) the other will be full. There is a small facet pump going infront of each one and feeding a Weber DCOE 45.

I tried to plan it so there will always be at least 1 surge tank full with either heavy breaking, accelerating and long corners.

Beat me to it

The only reason I asked was that my oldest son used to fettle MK1, MK2 Escorts, many of his pals used to come round to my workshop on the borrow, many still do and now their children

One came with a MK3 Escort that occasionally hesitated on large roundabouts, sometimes would start then immediately stop when parked on a hill and had an occasional unusual noise that he couldn't put his finger on.

The swill pot had come loose

Being a skint apprentice, as they all were at the time, I got him to do the hard work, dropping and draining the tank, then spot welded the swill pot back in place.

RE. Engineers post, he may be thinking the same as I am in that have you thought of tank foam as a safety feature ?

Next, seeing as you are using electric pumps, have you thought about what form of safety cut out to use in the event of a shunt, ie inertia switch ( difficult to position correctly and can trip with rough race / rally type driving) tachometric relay ?

 

[topcat]

It will hold about 42litres, the twin tanks found in original "S" are 50litres and the big single tanks as in standard minis hold around 33litres, so it will do for most stuff! I won't be doing anything too serious anyway for now.

It has baffles fitted that divide it into 6 sections (plus the little tanks underneath at 0.32l each) and I had considered fitting an intertia switch but to be honest I can't see it being reliable in a rally situation where the car will be getting hammered by the rough terrain and harsh driving all the time, so I will have to rely on the batter master switch, though I will have a dash board switch for each pump as well (can disconnect one or the other if have any problems and run on one only), of course this will require a non return valve inline with each pump to stop them pushing fuel past the dead one back into the tank!