Icebergs, World War 2, aircraft carriers, and glaciology: the study of pykrete and the bergship

Artists rendition of a Bergship

War brings out the best in people. Okay, so that needs some explanation. What I mean here is that solutions that would normally be labeled ridiculuous or insane are considered plausible and explored with fervor.

Which brings me to World War 2, icebergs, aircraft, glacial flow, and an awesome paper from 1948.

Note: The following information is taken from Perutz 1948.

By the Fall of 1942 a major disadvantage of the Allies was a lack of air support. Any invasion of a far off land would be held back by the lack of air support until the Allies could establish airfields.

Therefore the cheap construction of gigantic aircraft carriers was considered. Mr. Geoffrey Pyke submitted a plan in October of 1942 where he proposed that an iceberg should be hollowed out to shelter aircraft and leveled to provide a runway. This craft would travel at a few knots.

Pykrete ship design

Why make a boat out of ice? Well for one it would unsinkable. Ice is difficult to break with explosives (shown by resistance of icebergs to shellfire) and melted very slowly when insulated. Thus, the Allies began researching the possibility of constructing a "bergship". An ice thickness of 15m was considered essential to operating aircraft from the deck of a carrier and the minimum runway size for bombers was 600m with 60m the most desirable width.

To design such a ship the mechanical strengths of ice needed to be known. However it was found that some icebeams would fail at stress as low as 4.9 kg/cm^2. To compare, pine wood has a rupture modulus of 800 kg/cm^2.

It was discovered that the inclusion of wood pulp in the ice increased its strength considerably. This substance is known as pykrete.

A fantastic amount of research was done on pykrete to learn about its behavior under stress. A cylinder of pykrete was loaded with a stress and the resulting deformation measured. Differing percentages of wood pulp was tested with 14% wood pulp to ice performing best for the materials. Pyrkete's resistance to projectiles was also tested as seen in Table III from Perutz 1948:

Table III from Perutz 1946.

There's plenty of videos online demonstrating the strength of pykrete versus ice:



Based on the results of underwater explosives tests the effect of a torpedo hit would produce a crater 60 cm deep and 4.5 m in diameter. Regardless, a wall thickness of 9.0 m was suggested to keep the bergship safe from torpedo attacks.

The result of the stress tests on the pykrete showed that the strength of pykrete was dependent on the rate of loading. The load at which pykrete failed would be greater the slower the rate of compression. An added difficulty became apparent. With a very slow compression rate the pykrete would not fracture at all. Instead the cylinder would undergo plastic deformation until it was the shape of a flat disk.

Thus the possibility was considered that a ship of this size would deform under its own weight. The creep properties of pykrete had to be tested. Samples of pykrete were loaded with differing stresses:

Figure 2 from Perutz 1948.

What we see in the above figure is total compression in inches is plotted against time for contours of differing initial loads (kg/cm^2). What we see that loads have a rapid initial compression and then slow until a constant rate of compression is achieved.

Based on the creep curves, it was decided that a bergship would undergo plastic deformation for a period of several weeks until deformation ceased.

Thus the design of a pykrete ship was undertaken. The hull would be surrounded by a waterproof, artificial skin and the walls would be maintained at a constant temperature of -15 degrees C through artificial refrigeration. Each ship would have 16 refrigerating plants necessary to cool the ship. The berg ship would be propelled by 20+ electric motors. However difficulties in design and also the shear scale of resources  necessary (1,700,000 tons of pykrete per ship) to construct a bergship led to the project being abandoned.

Although the project to construct a carrier out of pykrete failed, much insight was gain into the behavior of ice and glaciers. The tests on pykrete showed that the creep mechanism was "quasi-viscous" flow and suggests the same for glaciers. Quasi-viscous flow is achieved through rearranging atoms in the crystal lattice either within individual crystals or across grain boundaries. This creep mechanism differs from melting and refreezing creep which was originally thought to be the creep mechanism for glaciers.

One cool feature of Perutz 1946 is that after the paper finishes, there is a discussion attached to the end. It is concluded that pykrete could be used as an effective runway material in arctic countries.

References:

Perutz, M. F., A description of the iceberg aircraft carrier and the bearing of the mechanical properties of frozen wood pulp upon some problems of glacier flow. Journal of Glaciology, Vol. 1, Issue 3, pp.95-104, 03/1948