Reglering av avgasningen i en dialysmaskin.

Abstract: The current control strategy in the degassing system in a Gambro AK200S dialysis machine leads to a rather noisy flow upstream in the machine. This is where different concentrates are added to the dialysis fluid and as the control of this adding is rather slow the unsteady flow leads to an unsteady conductivity of the dialysis fluid, which makes it difficult to evaluate the effect of the dialysis treatment. The main components in the degassing system are an adjustable three-way valve, a constrictor valve, a pump and a degassing chamber. The pump is used to create the necessary underpressure needed to degas the fluid and the three-way valve is used to recirculate fluid from the degassing chamber, as the flow needed to create the underpressure is greater than the overall dialysis fluid flow. The pump and the three-way valve are available actuators and the underpressure and the fluid level in the degassing chamber are available measurement signals. The system is investigated thoroughly and different control strategies are tested. It is found that the desired underpressure in combination with the constrictor valve leads to what probably is cavitation. This makes the flow approximately constant for big underpressures and this makes it hard to implement a control strategy where the pump is used to control the degassing level and the three-way valve to control the pressure, this strategy was beforehand thought to be the solution to the problem. The three-way valve is controlled by a stepper motor, and the best method tested to achieve a steadier flow has shown to be to increase the resolution of the motor, using microstepping control. Strategies where the three-way valve is changed as seldom as possible are also tried and the result is a piecewise steady flow. The best flow that can be achieved is limited by disturbances, especially from air bubbles entering the degassing system, but also disturbances from other parts in the machine. Although the flow is stabilized the conductivity is not improved, showing that more has to be done to achieve a steady conductivity. Probably the conductivity control loop has to be redesigned, but this is outside the scope of this master thesis.