ABSTRACT
The most recent product safety standards for machine tools such as ISO 16090 for the safety of milling machines follow the three-step-method of risk reduction, as it is explained in ISO 12100. Step 1 is always the design of inherent safety of the machine. Step 2 follows with the design of additional safety measures, and step 3 focusses on instruction for use (including warning signs at the machine) and training. A very effective risk reduction can be achieved, if the work area of the machine is fully enclosed (this belongs to step 2). In doing so, the automatic machining processes can take place without exposing the operator to the hazard. However, operator access to the work zone is sometimes necessary for manual intervention, such as setting actions inside the work area or workpiece change. Then careful instructions are necessary, which shall enable the operator to fully control the situation and protect himself by awareness of the respective hazards.
The many different kinds of operator activities, which vary with the selected machining process, can be allocated to specific modes of operation. In ISO 16090, a modes of operation concept comprises five selectable options: 0: manual mode/1: automatic machining/2: setting mode/3: special mode with limited manual intervention/and separately for selected operators only: service mode. The purpose of such a concept is to reduce the relative exposure of the operator to occurring hazards as far as for the intended use possible.
The combination of full enclosure and modes of operation concept brings about a significant risk reduction, however the effect is being discussed only qualitatively on an intuitive basis. Unfortunately, a scalable model is still missing, which could quantify the risk reduction effects, e.g. for the purpose of parameter optimization.
In order to improve the engineering abilities, a very first simplified Markov model is presented here, which is founded on a probabilistic concept for the description of the operator activities at a machine. As a result, the a.m. risk reduction effects can be scaled probabilistically. This is of advantage, if it comes to the quantitative reliability requirements of safety functions according to ISO 13849-1 such as e.g. a safe standstill of gravity loaded vertical axes. Because of the potentially severe hazards of those axes in setting mode or during manual intervention in the work area, the required reliability is high, e.g. performance level PL = d according to ISO 13849-1. This paper explains for typical manual interventions, how this requirement and supplementary safety provisions can be justified probabilistically.
