On-farm service loading on a rotary cultivator

Abstract

Service loading data are required by agricultural implement manufacturers to design against failure due to overload and to ensure an adequate fatigue life. Long term on-farm data are the most realistic but also the most difficult to collect, and more often data are obtained in short duration field tests in severe conditions using an arbitrary range of machine settings. In this investigation, long term on-farm data for a p.t.o. (power take off) driven rotary cultivator were collected and compared with data from short tests to establish whether a laboratory fatigue endurance test based on short duration test results alone would produce the optimum cultivator design for farm usage. On-farm recordings were made of p.t.o. torque and the structural strain at a potential fatigue failure location on the cultivator. The records were taken during transport and field manoeuvring as well as cultivation on nine farms, and covered periods ranging from 20 min to 12.5 h. The short duration field test data were obtained in a separate and earlier study using the same cultivator and included the same data measurements. These short tests lasted 2 or 3 min only and did not include transport and field manoeuvring. Torque records from the p.t.o. were analysed to give histograms of the proportion of time spent at different torque levels, and the strain records were analysed in accordance with BS 5400 to calculate fatigue damage and fatigue life. The productive working time from the on-farm tests, which was taken to be the time spent above the zero torque level, was in most cases 60–70% of the total but could be as low as 40%. Maximum p.t.o. torque was similar in both short and long term tests. The severity of loading between short and long term tests was compared using the mean fatigue life calculated for the two groups of tests. On this basis on-farm usage was one-sixth of the severity of the short duration tests. A laboratory endurance test based on short duration test data alone, without regard to their relationship with on-farm data, would lead to an unnecessarily heavy and costly machine.