While I’m probably not the first to have problems with M3W fuel tanks, when I got my second chassis change at MMC quite a few years ago they replaced the right hand tank (and possibly the left too – I’m not sure) without any explanation beyond “it was necessary” so technically I may actually have been one of the first but what it really means is that my proper entry to the tank problem club was delayed to a few months ago when I started getting the dreaded strong fuel smell in the lock-up. Initial searching didn’t locate anything definite apart from a badly degraded gasket on the level sender. This became obvious after a slightly over-zealous fill up followed by a 2-mile drive to my lock up where parking overnight revealed a small pool of petrol on the floor under the tank and signs of liquid on top of the tank. A new gasket cut from Viton sheet sorted out that leak and there was a distinct lessening in that petrol aroma, but only a lessening not an elimination.
After some searching with the tank back in situ the next step was to take the right-hand tank out again, plug the balance pipe outlet, hang it from the ceiling and fill it with petrol to see what would happen. Nothing obvious was the initial answer but after leaving it over night there were signs of some petrol on the underside of the tank which isn’t a good place for petrol. With some chalk rubbed over the weld seams where these traces were and squeezing the sides of the tank there was definitely some leakage from a bottom seam but it wasn’t clear where on the weld it was leaking or over what sort of length. A visit to the inspection department at work saw me attacking both tanks with a set of dye penetrant aerosols and sure enough there was a lovely 50mm crack on the lower toe of the inboard floor to sidewall weld of the right-hand tank, about halfway from the underside mounting point to the back of the tank. The mounting brackets of both tanks were also in poor shape, in particular the underside brackets had tears in them heading through the bracket parent material into the welds towards the tank shell.

Red line = crack!


In the process of taking both tanks in and out I can’t honestly say that I was overly impressed with the mounting arrangement; two very thin aluminium brackets welded to the shell of each tank with minimal effort made to spread the loads over a larger area of the tank shell and a length of seat belt with a simple tensioning system that on my car was only possible to undo on the left tank; the strap for the right hand tank was slightly shorter and the only way short of taking the entire car body off to remove it was to cut a section of the threaded tensioner off and lever it out of its bracket, getting it back into the bracket was a total swine of a job involving far more brute force than seems reasonable, and yes, I did try various combinations of wrapping it at an angle over the tank to get it to slide down into place but to no avail. The way that a considerable part of each tank is cantilevered beyond the back mounting bracket would also suggest that the loads from the contents slamming down would cause excessive flexing of the tank floor and the side walls perhaps leading to fatigue in the seam welds, say along the bottom toe where the petrol was getting out…
I could simply have got nice new tanks from MMC but that would just postpone the problem for another few years or just find somebody to weld up the crack and re-fit the tank but that too struck me as just another delaying tactic and I wanted to at least try to sort the problem with a whole new mounting arrangement and this is where the sledge hammer to crack a nut reference comes in; a sledge hammer of course only being the appropriate tool to crack a nut if a hydraulic press isn’t available.
First thing after some rolling about under the car with verniers and tape measures getting some dimensions for clearances and tube diameters was to remove the bottom brackets from the RHS tank leaving a flat base. A support tray of aluminium 10mm sheet was then cut to the shape of the tank, lengths of aluminium angle (1”x3/4”x1/8”) attached to the underside of the tray giving a 15mm flange around most of the tray though some of this would be trimmed back later to ensure clearance for shock absorbers, some chassis bits and the wheel spindle. Strips of 1mm nitrile rubber sheet were used to line the tray and the side flanges temporarily for fit-up purposes and M5x12 button head screws for holding the flanges on…quite a lot of screws in fact, so a lot of drilling and tapping. For the front mount a couple of blocks of aluminium were drilled and tapped to hold them together, drilled for four hidden screws to attach them to the underside of the support tray, then drilled and bored to give a closed, tight clamp on to the horizontal tubular that runs across behind the seats once lined with the 1mm nitrile rubber.
More rolling about under the car ensued with the tank on the tray and the back end of it supported on a jack until I had just the right thickness on the front clamp to hold the tank and tray parallel to, and tight to the woodwork on top of the tank. Holes were then marked, drilled and tapped for the front clamp. More aluminium angle was then used to make a bracket between the chassis tank mounts and the tray. This bracket was bolted to the chassis tank mounts with hard rubber washers to give some insulation between the steel and aluminium and to allow some flex. It was fitted with slotted holes for its attachment to the tray to give some adjustability. The front tank mounting was re-used but with better rubber washers to allow it to at least be held firmly in place with the bracket in an un-bent, neutral position.
Once all this was assembled in place and looking right, the tank went off to a fabricator for the weld repair and I both started on the same arrangement for the LHS tank and also cut sections of the tray out to reduce weight and give less of a water trap. The trays both had their one-piece 1mm Viton linings bonded on at this stage and trimmed to size.

Freshly cut support tray on bench with assorted required bits.


In the car.


From below with tank in place.


Front mount on to chassis tube.


Tank in place.


Both tanks fitted.


The fabricators weren’t sure of the grade of aluminium the tank is made from so ran their PMI machine over it; it turns out to be a high silicon alloy. Apparently this is fairly normal in automotive aluminium applications but pretty much everyone outside automotive uses the magnesium based alloys. It’s a bit important to use the right material for the welding so the PMI was definitely a good move.

Repaired tank without bottom bracket.


Next was the webbing strap arrangement and what to do with it. Having not come up with a simple, elegant replacement for the strap I gave in and just went for a minor modification to allow it to be easy to fit and remove without removing the body. I just cut a small piece of steel flat bar to give as short an extension of the hidden attachment bolt as possible which allowed the tensioner rod on the other end to easily fit into its bracket once wrapped round the tank. This if course meant that the strap could no longer be fully tightened but all it needed was a piece of aluminium flat bar inserted between the webbing and the upright box section that it wraps around, and full tension can now be reached.

Webbing strap extension.


Aluminium spacer on web strap.


The first support tray assembly took more than a little time and effort to do, the second one while considerably quicker still took a fair old amount of effort to do. I have the tools to do these things but not to do any sort of production work. In the end, very little of the whole assembly was actually measured, most of it was done by matching sizes then trimming to suit to get as good a fit as possible. The tanks were actually a very good mirror match side to side but I’m not so sure about the chassis to woodwork measurements being such a good match so while a CNC machine could churn out the components, I’m not sure they would fit anyone else’s cars.
Is this the perfect solution? Of course not.
Will it keep petrol inside the tanks for longer than the original arrangement? Ask me in a few years…I certainly bloody hope so!
As an aside, while I was doing all this I also knew that the Centa rollers were deep into their end of life stage so I’ve also just fitted Mr Bleazey’s latest own brand cush drive and clutch plate which took a reasonable amount of effort to do, being an engine out job so I’ve only had one little run about in the wee car but nothing has dropped off yet!

Nice one Calum, thanks for that.

Great work and a very useful set of photos...another mid-life crisis to look forward to sorting....

That is quite a sledgehammer! Hopefully the extra bonded on load spreading mounts that some of us have done, along with regular checks of the belt tension, will prevent further cracks from the brackets and the seams will hold up. As you say, doing nothing isn't a good option as they will crack eventually....and again if replaced with the same standard of tank.

Impressive job! Congratulations for your skills. I really hope it does last longer than the original design!

Impressive stuff (as ever) Calum. I wonder if the fuel tanks are fabricated by the same company who produced the early oil tanks (the ones that popped their end plates off when given the opportunity)?

That is a magnificent job on engineering a better support system for those tanks. I always thought they were suspended in a very pendulous manner and certainly nothing that would withstand a good military inspection. Kudos for such a fine job, you have raised the bar once again!

Some very well designed and executed engineering Calum

Me too the tank cracked underneath, I had been smelling gasoline for a week without seeing any leaks and Sunday morning there was gasoline on the ground, the only chance I have is that I just rebuilt two circuit engines to a person who has a business building and repairing fuel tanks and builds aluminum stainless steel and metal.

Those bottom brackets have certainly seen better days!
Even with the straps done up tight there is quite a bit of movement on the tanks, partly because of the rubber mounts on the brackets but mainly its just that the steel tubes that are cantilevered out from the chassis cross rails flex a remarkable amount even with just by hand so throw in a fuel tank of petrol with a lot of the weight out beyond the bottom bracket and there's a lot of flex going on.
Incidentally, the man that sorted out my split tank reckoned that the original welding was perfectly okay which kind of means that the problem is with the mounting arrangement, the thickness of the aluminium the lack of stiffening of the floor of the tanks or any combination of the three!