I was interested to read Steve Turner's reference to the problem of 'ballast memory' rising from the inability of the track laying machine to remove sufficient depth of ballast to eliminate the problem (NCE 7 June).
I experienced a similar problem when working in the 1980s on the upgrading of the South African Railways coal line between the Broodsnyersplaas in the Transvaal coal fields and the Indian Ocean port of Richard Bays. South Africa Transport Systems, as it was then known, transported coal in 2.5km long trains carrying around 18,000t with axle loading of 28t, at speeds up to 80km/hr.
The inland half (about 320km) of the upgrading works of the sea-bound line was undertaken by a contractor using Donelli gantries. I was involved in this contract to replace the existing 54kg/m rail with UI60kg/m 1% chrome/ manganese steel rails on concrete sleepers.
Difficulty was experienced preparing the old ballast to receive the new track which was being laid with closer, 620mm, sleeper centres. An improvised plough, used to break down the series of stubborn crib voids left after the sleepers had been lifted, proved not to be particularly effective.
The solution was to place a ballast cleaner ahead of the gantry operation. This had the effect of destroying the 'ballast memory' and, at the same time, leaving a neat flat ballast top under the old sleepers. This eventually allowed the contract target of 576m per four hour possession to be achieved and maintained.
Tom Bryer (F), 2 Crosskeys, Bearstead, Maidstone, Kent, ME14 4HR