On Friday the 1st of March the Hybrid team performed a test of their recently produced tank at Advanced Lightweight Engineering in Delft. The aluminium tank was a shortened version of the one that will be produced for the eventual engine. However, although this test tank is only approximately 1.5m in length compared to the 4m of the eventual tank, it is still a good test case. This is because all deformations in the tank and also the eventual failure occur only near the bulkheads at either end of the tank. The tank is with 1.5m still of sufficient length to make sure that all transient effects that occur at the ends do not effect the middle section of the tank.

The tank was tested at ALE where it is possible to put the tank in a sealed container and then pressurize it with water. The test was conduced over a number of steps. First the tank was pressurized to 10 bar to test if it was leak tight. The same was done for 30 bar as well. In both cases the tank held the water without a problem. So far so good. Then the pressure was pushed up to 60 bar, the intended working pressure of the tank. At this point the tank showed some small deformations at the top bulkhead (the side that is intended to fail first). This was however no harmful deformation but showed that the bolts settled into position to take the loads best as they could. In the picture you can see this deformation at 60 bar in comparison to the unpressurized case.

Tank deformation Dawn tank test 01-03-2013

After this the tank pressure was raised for a final time before it would be taken up to failure. At 90 bars the tank showed mayor deformations on both bulkheads. The bolts slowly started shearing out of the wall. Soon after this, when the pressure was at 91.8 bar, the wall finally gave way and the bulkhead sheared out as intended. This gives the tank a safety factor of 1.5. Usually this is considered perfect in the aerospace industry. However for our project we would like to increase this factor to 2, to account for any possible uncertainties in production and manufacturing and to make sure that it is perfectly safe to work with the tank.

Because of this the tank was originally designed for a burst pressure of 120 bar; Two times the operating pressure. Why did it fail earlier then? This was due to some small production errors and modifications in the design. There were two main differences . First, the bolts used needed to be countersunk about a millimetre further then was originally planned, this weakened the aluminium wall. Second, the holes for the bolts were positioned about half a millimetre closer to the wall edge then was intended, also lowering the tank burst pressure.

Although the team is pretty confident that, when solving these issues, the tank will be strong enough, they will need to test the tank again.