"The largest single windmill project in the last 150 years"

Ian Pritchett of IJP Building Conservation Ltd was the speaker at the Annual General Meeting of The Friends of the Windmill at Windmill Hill on Tuesday 16th March 2004.

With the aid of computer graphics Mr Pritchett gave the audience a flavour of the largest single windmill project in the last 150 years.

Following a detailed study of the mill in Summer 2000, a schedule of repair was drawn up, based upon the timbers "as seen".

In view of the poor condition of the structure, they decided to dismantle the mill body and to then transport it in sections to their workshops. This decision was based upon partly on financial considerations and partly as its was felt to be impractical to repair the mill from a scaffolding platform on site.

In November and December 2003 the mill body was dismantled and taken to IJP's workshops. The upper sidegirts and one front corner post were found to be in a worse state than had been predicted in the 2000 study.

This Grade II* listed building presents the millwrights with the challenge of balancing both safety and engineering issues. A CAD model of the mill was produced, based upon Ron Martin's drawings. To Mr Pritchett's knowledge, computer modelling of this kind has never before been applied to a windmill.

From the model is was clear that when the mill was built in 1814, the stresses on the timbers were on the limit.

The mill body is "headsick" in windless conditions so that as wind speeds increase the mill body gradually levels itself, becoming fully upright in a 55mph headwind.

Based upon the publication "The Stability of Post Mills" by Paul Jarvis and a two-dimensional model of the mill, there was evidence of instability in the trestle as wind speeds reached 60mph.

The next step was to produce a three-dimensional model of the mill's main components. This revealed that the cross trees and quarterbars make the mill more stable than had been indicated by the two-dimensional modelling.

It is proposed that, after restoration, the mill's trestle will be monitored over time to see how the stresses and bending in "real life" compare with the projected stresses. The knowledge gained from the monitoring could then be applied to help other post mills in the future, for example Chillenden Mill, which collapsed on 26th November 2003.

During the course of examining the timbers of Windmill Hill Mill, many carpenters' marks were found, which has enabled the millwrights to piece the timbers back together in order to determine the original lengths of the sidegirts and some of the other major timbers.

It has been found that the front corner post is flared; it varies in section from 12 inches square at the top to 8 inches square at the bottom. The upper sidegirt slopes downwards by 4 inches from front to back.

The amount of weatherboarding that can be salvaged was less than expected. Some rot and woodworm has been found [5 inches in the sheers and 3/4 inch in the post]. The south-west quarter bar will be replaced as it has a major "shake" and has been attacked by deathwatch beetle.

Computer generated design alternatives [by K F Hume Design & Engineering] were then presented to the audience showing what may happen during the restoration project. 3D computer modelling will be used to determine the assembly of all primary timbers, whilst the secondary timbers will be assembled in 2 dimensions.

The proposed sequence of works is to repair the roundhouse before the buck, then to work upwards on the post and crosstrees. Next will be the ladder and tailpole followed by the walls of the buck. The windshaft, tailwheel and brakewheel will be the next components to be fitted, followed by the roof and weatherboarding of the buck.

The risk of a mill falling over is greatly increased if it is not facing the wind. Various options have been considered, including the installation of either a discrete electric motor or a rear fantail to turn the mill body.

There was a suggestion that the mill body could be anchored down in bad weather, however this would significantly increase the strain on the trestle.

The final part of the project will be the sweeps. It is planned that the mill body will be assembled by Autumn 2004, the sweeps will be up in Spring 2005 and that the sweeps will be capable of turning by May 2005.

No definite plans or decisions have yet been made. It is considered that the problem of stability does not arise until the fitting of the sweeps, so the mill body can be assembled in Autumn 2004 without having to resolve the method by which it will turn to face the wind.

  [Report compiled by Simon Potter]     Windmill Hill Windmill website

Reinstatement commences on site