Originally Posted by
cwarner
I wondered about this, particularly after my second OEM tank failed in 12 months. So after I replaced it yet again, I decided to do a failure analysis on the failed tank, and was shocked at how bad a few of the mistakes were that lead to failure.
So here is the tank. I had already cut the top off and cut out the side port, so this is just to show the whole tank:
Next, looking down into the tank from the top. Note that I have removed a bunch of internal parts already, but one can quickly see that the side that bulges out is not supported to the center for most of the length. Not good.
To make things more clear, I then cut off one corner of the tank to expose the central rib that goes down the middle of the side which bulges out.
Now that the inside of that portion of the tank is exposed, the problem is remarkably clear. The mistakes here are really bad. They clearly blew it on multiple accounts including the decreasing thickness of the stiffening rib and the insufficient cross section of material connecting the side of the tank to the middle. The loading is clearly more than the fiber reinforcement to the plastic can take and it fails in shear (indicated by the 45 degree orientation of the cracks. Then the whole tank self destructs due to excessive deflections. It is not to supprising that these fail all the time.
Clearly if the side of the tank was connected to the middle at some point above the side port, the stresses would be significantly lower. More to the point, if the stiffening rib did not thin out as it does the deflections in the side of the tank would also not contribute to the problem. I also noted that one corner of the tank did not crack (the corner in the lower right corner in picture 2 above). This corner had a much higher radius than all the others. again stress concentrations.
This is very sloppy engineering. It is very much on par with the abismal design work done on the rear subframe mount. In both cases the engineer who did the design work shows a complete lack of comprehension of stress concentrations, how stress is distrubuted in a structure, and the design limits of the materials they are using.
I sincerely hope the new tank (part number 17 11 7 573 781) has structural improvements inside versus this one (part # 17 13 7 787 039)!