In a process chain, workpiece quality features may be affected by several sub-processes. Hence, an optimisation of quality features throughout the process chain requires a process-wide approach, which also adapts the flexibility concerning the arrangement of the sub-processes. This paper presents a holonic control strategy for a process-wide optimisation of selected workpiece quality features. In particular, the modelling and optimisation procedure are designed such that a rearrangement or extension of the process chain involves no intervention in the control strategy. The approach is demonstrated along the production of bearing rings, regarding especially geometric quality features affected by distortion phenomena.
In the SFB570, several sub‐processes of the standard process chain for manufacturing bearing rings have been modified in order to compensate distortion. This paper presents control strategies that enable a precise deployment of the compensation potential, on a local (sub‐process) level as well as along the process chain. Two different control strategies for local compensation are demonstrated at the example of the cutting and the gas quenching processes. Furthermore, a concept for a process‐wide compensation of distortion is presented.
During the turning process of a ring, the clamping force of the chuck causes circumferential deviations in the ring wall thickness and diameter. In order to minimise the wall thickness deviations, an experimental set-up for an in-process control of the cutting depth during turning is being developed. The set-up comprises a fast tool unit, an ultrasonic measurement system and a global control unit. In preliminary tests, the ultrasonic measurement proved suitable to acquire the changes in wall thickness. First turning experiments show that the developed fast tool unit and the applied control strategy fulfil the requirements for an in-process compensation of wall thickness deviations during standard turning conditions.
Kurzfassung Um das Kompensationspotenzial der Gasabschreckung für Verzüge optimal anzuwenden, wurde eine aus zwei kaskadierten Rückführungen bestehende Regelungsstrategie entwickelt. Diese Strategie wurde bei der Abschreckung von Ringen umgesetzt. Für eine automatische Kompensation von Rundheitsabweichungen konnte ein analytischer Ansatz hergeleitet werden, der als Basis für ein Radial-Basis-Functions-Netz dient. Simulationen zeigen, dass sich Rundheitsabweichungen um mehr als 50 Prozent reduzieren lassen, wenn das Kompensationspotenzial des Abschreckprozesses entsprechend angewandt wird.
For an optimal application of the compensation potential of the gas quenching process, a quality control strategy, consisting of cascaded in-and post-process control loops, has been developed. The implementation of this strategy in regard of controller hardware and software is demonstrated for the quenching of a ring. For an automated compensation of roundness deviations, an analytical approach has been set-up in order to develop a Radial Basis Functions (RBF) network. Simulations of this approach indicate that the roundness deviations can be more then halved if the compensation potential of the gas quenching process is applied adequately.
The turning process requires a clamping of the workpiece. If the fixing of a thin-walled workpiece (i.e. a ring) is realised with a three-jaw-chuck, distortion of the outer diameter of the workpiece occurs, leading to a varying circumferential wall thickness on unclamping. This paper presents an ultrasound system for the in-process-measurement of these workpiece distortions. The measurement principle is based on the determination of phase differences between echoes of an ultrasound burst. The measured in-process values can be used to control the turning process with a fast tool servo.
This paper presents strategies for an automated control of distortion at the operational level of a production process chain, as well as first steps towards a combined implementation in heat-treatment processes. In order to minimise the loss of production due to quality defects, the first quality control loop will be implemented on an in-process level. The second control loop within the reach of operational control includes the feedback of post-process measurements of quality features. As such, the post-process control loop optimises the set-values of the in-process control loop.
To compensate the geometric distortion of steel parts, a strategy for an in-process quality control during the heat treatment was developed and tested. The control concept consists of a cascade, in which a state space controller is combined with PI-controllers. A first application of the control strategy demonstrated the suitability regarding the required dynamics, stability and accuracy.
Tactile distance measurement systems cannot be used in hot environments with temperatures above 800degreesC. A green laser-triangulation system with appropriate optical filters can meet these requirements. A triangulation system was build up and tested in two different heat treatment furnaces at temperatures from room temperature up to 900degreesC. Good linearity of the system and acceptable correspondence between measurements and published reference values could be observed. The repeatability of measurements of different workpieces is also good. These results show the suitability of the system for distortion measurements in heat treatment furnaces.
A new combination of four organic solvents is proposed for the optimization of TLC separations of basic drugs and alkaloids. The solvents are diethylamine (DEA), methanol (MeOH), chloroform (CHCl3) and ethylacetate (EtAc). They were selected from a collection of ten solvents used in Normal Phase TLC mobile phases recommended for the separation of alkaloids and basic drugs in the literature. The selection was based on the classification of solvents according to selectivity and solubility parameters. Excluded were apolar and weak solvents that show no selective (polar) properties and are used only for the adjustment of the solvent strength. Polar solvents from different selectivity groups were selected to combine as many as possible selective effects in one solvent system. The final choice was made considering the displacement theory for Liquid Solid Chromatography.The four solvents have been intended for application in an optimization procedure that uses mixture designs and response surface modelling. The factor space is a tetrahedron of which only in a part suitable experiments can be performed, i.e. the design space. Experimentation in the design space should allow the simultaneous optimization of solvent strength and solvent selectivity. The quality of TL chromatograms was characterized by two criteria: the separation of the worst separated pair of spots (R(s)min) and the R(f) value of the slowest moving spot (kmax). In this way it is possible to influence not only the separation of the spots but also their place on the TL plate. The new four solvent system has been tested successfully in the separation of the parent alkaloids of four dry plant materials: Ipecacuanha root, Cinchona bark, Belladonna leaf and Opium. The predicted mobile phases gave better or equally good separations than the mobile phases prescribed by the European or Netherlands Pharmacopoeia for these separations. Moreover a substantial reduction of the number of solvents necessary for the composition of the mobile phase was achieved as the official procedures need eight different solvents.