Top Fuel race car oil tank ..boom

[bentwings1]

I build a lot of race car fuel and oil tanks. Most are made from .060 or .050 3003 aluminum. This one was .060. I use 4043 rod unless they are anodized. Even some of those get anodized even though there is a darker shade where they are welded. I use a square wave TIG welder. The tanks have rounded edges all around produced by using a form roller called a tank roll. 3/8" radius. The joint is a butt joint with a carefully fitted "0" gap. I spend a lot of time and effort to get very good fit up as most of the tanks will warp is there is any gap at all. I get really deep penetration but as with most TIG welded tanks there is still a seam visible on the inside. Gas welding is nice but a lost art. I'm going to a 4 day seminar in a couple weeks featuring gas welding aluminum. Anyway the weld is smooth and about 3/16 wide, consistant all around, no stuck electrodes, no holes melted thru, no other uglies. Blowing my horn I haven't had a tank leak after welding for a long time.

Now finally what happened. The tank in question is/was 12 x 12 x 9 inches with a 2 3/4" id vent tube. This is fixed at an angle and not capped. This tank is to capture oil and vent fumes from the motor...crankcase vent. The valve covers vent into the lower chassis tubes then are vented into this tank which vents fumes and captures oil. This type of tank is mandantory on all top fuel cars.

So the car is brand new, run only one warmup a week ago. No problems sonuds very good. Last night it was warmed up in prep for a demo at a big local cr show. During the warmup while running on alcohol the tank exploded. I mean big time. One side was blown completely out tearing the mount fastener right thru the welded tab. The other side was blown flat...ripped 3 sides totally out right down the weld. The band or center of the tank is stretched bigger than a bowling ball. Of course both sides are stretched in the same manner. This is the firts time anyone has seen or heard of this happening but I would have to guess statistics would say it has happend before.

I tried to take pictures but they didn't come out worth posting. The seams split right down the center of the welds except for about 1 inch on the side blown out which ripped near the mount tab. I looked at the welds with a fairly high power magnifying glass and they appear to be stretched and ripped rather than a chrystalized cracked area. They are thinned out and there are stress marks into the sheet skin. The bead itself is flattend some as though rolled or planished.

Now why did this happen?? Some information. The motor gets started on gasoline for about 30 sec then the alcohol is turned on. It runs several minutes at fast idle then the nitro is turned on about 1/2 to continue the warm up. Quite a bit of fuel mix gets into the oil pan during the warmup. Sometimes you can see the vapor come out of the catch tank. Since this was brand new there was no oil in the tank at all. Once oil gets into it, it is very hard to get it all cleaned out. This tank was bone dry after the explosion. The exact duplicate of the car (and tank) lives in the next stall at the shop. It has had numerous oil venting due to burned pistons and motor catstrophic distructions and the tank is still there in one piece and same shape.

Obviously there was a fume explosion but what set it off???? Maybe metal chips inthe chassis tube caused a spark??? Static electricity fromt the rotating tires??
Something else???
Pat let's hear from you.
Thanks.

99 Dodge CTD dually.

 

[unclesyd]

You have a least three scenarios to get and explosive mixture in your fuel tank. As you know air is common to the three mixtures, Hydrogen, gasoline, and methanol. Again you have to have an ignition source through the autoignition temperatures are quite low for these materials. Hydrogen would be generated as corrosion product from the Methanol and Aluminum. I don't think you have this as the tank was new. Gasoline and Methanol vapors will creep over surfaces and it hits an ignition source there can be flash back to where the concentrations are in the explosive range. Your event was at relatively low energy as it had the time to stretch the Al at the welds. Under your conditions as you posted a little Oxygen sucked or diffused into the fuel tank will go a long way. Try to remember the conditions prior to the event. Also remember that Methanol burning can be invisible so you may not have seen the flashback if ethanol was the cause.
Your welds failed in the correct manner, down the middle, under tonsil.

One pont thank you may want to try in welding the thin materials is but them up with a square edge. Don't use any method to braak the edgess. It takes a little praciice but a full pentration weld can be made.

 

[bentwings1]

I got a little more information on this today from the team. The 'event' happened about 5 minutes into the alcohol part of the warmup. The motor runs about 2800 rpm. It 'popped' or miss fired then the tank blew up. They continued the warmup and switched to the nitro for another 10 minutes and shut the motor off.

At the show they ran another complete warmup without the tank.
Here the #8 cylinder was blowing much more smoke than the other cylinders. These motors smoke a little as they start and warm up then they dry up as we say. So we have the first clue something is amiss.

So tonight back at the shop we started to look things over. the crew just wanted to write it off but I being the dad of 3 generations of crewmembers on car insisted they get to the root cause of this as there really isn't anything that could cause a spark to set the fuel vapors off under normal conditions. I've only been in top fuel 40 years so I felt I had a point.

I told them to take the valve covers off and run the valve adjustment which they should have already done but had not.

5 minutes later crewchief son says "hey we have a damaged pushrod on #8 intake". Looking at the pushrod it was blue/black on the top end and the adjuster on the rocker arm is missing the ball and the end of the adjuster smashed badly as well as the rocker arm.

Damaging an intake valve or any part of it can easily cause a blower explosion roughly equavalent to a grenade with many more pounds of shrapnel. Very bad situation.

Well the pushrods are made of high quality hardened tool steel. D-2 I think as is the adjuster and it's ball end. Tool steel sparks very nicely when ground, or 2 sharp edges struck hard together or abrasive rubbed against it. All 3 likely happened very quickly here.

So when the adjuster broke it caused a number of sparks and lit off the fuel vapors inside the motor. I think the flame traveled down the chassis tube to the tank where it encountered more vapors and lit them off too, thus blowing the tank.

I looked duch more closlely at the tank welds tonight too. Except for the first 2 inches which are 100 percent or very close but no weld bead, the rest were full pen with a bead. I always purge these tanks with argon too. An old aircraft maint guy told me to preheat to about 150 deg F at the start too so the weld is not on cold metal and it tends to drive off moisture. I do this and measure using my infrared heat gun. It does work quite well if you measure at very close range.

So I think we have a reasonable cause.

99 Dodge CTD dually.

 

[EdStainless]

My hunch is that you are right, and that you don't see this because once there is a layer of oil in the tank the initial fuel fumes will be soluble in the oil. Once they dissolve in the oil it reduces the amount of them free to burn.
It sure doesn't sound like there was a problem with the tank.

= = = = = = = = = = = = = = = = = = = =
Plymouth Tube

 

[Tmoose]

A screen is often an important part of many effective spark arrestor/flame trap. I used to think that metal screen was required to conductively cool the fast moving fireball, but Volvo used a plastic screen in their PCV system to separate the crankcase full of several liters of potentially mixture from intake puffs for decades.

http://www.carpartswholesale.com/images/parts_100/beck_arnley/BA-0460075-1.jpg

 

[bentwings1]

Edstainless....that's exactly what I'm telling these guys. The tanks always have a little oil in them after a run that is hard to get out so it is just left there. I believe the residual oil raises the flash point just as you say.

So the new tank will get a quart of fresh 70 wt sloshed thru it before I give it to them.



99 Dodge CTD dually.

 

[kingnero]

A little sidenote about using an IR-gun to measure shiny metals like Aluminum and stainless:
Make sure your measurements are correct: IRguns are standard set on measuring dull finish surfaces like standard steel. Rusted steel gives different results because of the more dispersed radiation. shiny objects do mess with the reaadings as well.
Maybe you already knew, maybe you didn't (?) but try to doublecheck your results once in a while to make sure your readings are (still) trustworthy.

 

[bentwings1]

King..Glad you posted this. I preach this all the time to the crews. Even demos of this fail to convince the guys at the shop.

Nobody will take the time to look at Omega's site.

99 Dodge CTD dually.

 

[Mr168]

X3 on kingnero's comment. Even a dull/unpolished surface finish on stainless steels can have a significant impact. I have seen differences in excess of 300°F+ between a calibrated infrared pyrometer and a tempilstik just measuring interpass temperatures on stainless and duplex steels.

 

[Tmoose]

A piece of masking tape is a decent uniform emissivity surface, if time and temperatures permit.

Plenty of info suggests even the black paint = high emissivity / white paint = low emissivity chestnut can be deeply flawed.