Efforts to estimate a factor to relate the ultimate overturning moments found from destructive pulling tests to those which results in trees being windthrown have been ongoing.

The suggested mechanism of failure is as follows:

• A strong wind causes a build-up in pore water pressure which could result in hydraulic fracturing and liquefaction
• Subsequent to this even the tree could be rocke with large displacements by winds of lesser magnitude and these large displacements could lead to increased bending moments that could eventually overturn the tree.

In addition to the above it is known that the shear strength of a soil is greatly reduced when subjected to a cyclic loading. The results of cyclic tests conducted at UCG indicate that the hear strength in cyclic loading was between 31% and 49% of the shear strength in monotonic loading.

It has also been noted that when subjected to cyclic loading the stiffness of the tree system reduces and thus a smaller force is required to cause the same deflection than was originally the case. Tests in the field have shown that the stiffness of the tree system does decrease when subjected to wind loading.

The above points suggest that a tree may be windthrown when subjected to wind loading.

The above points suggest that a tree may be windthrown when subjected to wind levels which may be considerably less than those predicted from destructive pulling tests.

Previously we suggested that a useful factor could be one which related the ultimate overturning moment as found from destructive pulling tests to the overturning moment causing substantial buildups in pore water pressures under the rootplates of dynamically loaded trees.

The factor suggested is as follows:

Factor = O.M.Pull/O.M.Dyn

Where:
O.M.Pull = the overturning moment recorded from pulling tests, and

O.M.Dyn = the overturning moment initiating pore water pressure buildups

An application of this formula to ninetrees resulted in the following estimated average factor:

Factor = 0.39

It is difficult to analyse the relationship between a monotonic pulling test and a natural cyclic loading as both regimes are significantly different. Therefore this simple relation should be considered only as a starting point for developing a more significant relationship.

The difficulty arises from the number of factors which influence a trees ability to resist wind loading, factors such as the following:

• The length and duration of any given storm event
• The possibility of accumulated damage and subsequent recovery
• The nature of the soil at the particular location
• The nature of the soil moisture at the particular location
• etc.

Therefore it is suggested that a comprehensive series of tests in addition to a reanalysis of existing data would be required before any reasonable suggestion could be made.