Piston pump in simulink/simscape

Short Discription

For my project/assignment I have been asked to design and build a controller for a piston pump (syringe pump). While working out the plant model for possible simulation and testing I stuck an issue where it seems Matlab/simulink/simscape does not have a block for these types of pumps.

.

working description

1.A to be defined controller will control a motor and this generates rotational mechanical movement.

2.The rotational mechanical movement will be converted to linear by using an spindel (gear and rack in simulink)

3.This connects to a piston that moves up and down pumping in and out a fluid (using hydraulic library components).

4.The piston has 1 connection that gets switched by a 3 way valve with the "A" connector connected to the pump, "S" Connected to the plant, and the "T" connected to the tank.

5. The valve will switch to tank when the piston is retracting (creating an vacuum and so sucking fluid in). And switch to the process when the piston is compressing (creating pressure pushing the fluid out)

.

Problem

The main issue is that I can not find a correct way to make the body of the pump. I have been using A piston and connect the shaft to the linear part(gear and rack) and the body to the reference point but it seems the piston is unable to generate a vacuum and drain the fluid out the tank and then in the process.

.

Asked

How to create a proper piston/syringe pump using simscape components that can both suck and compress liquid.

[Nicolas Schmit]

Please attach your Simscape model to your question.

[Maximilian Partl]

Dear Michael,

I have the very same problem. If you found a solution to it, can you please share it? I tried it now with a ideal translational velocity source and a single-acting-hydraulic cylinder, but it still doesn't work properly for me.

Greetings Max

Answer 1

[Michael Lammers]

I attached a file I tried using for testing. (after uploading i found out that 1 did not add a 2 pos valve between pumpsensor1 and the hydraulic reference. this would be needed to generate pressure as it would push to a closed area.)

As i am not an expert within simscape the operational way might be odd.

    The position block before the sum block contains a slider that sends the value to the motor and gets the pump moving
    The valveAcommand1 is to switch the valve between its 3 settings

to explain what i wish to simulate and how i tried doing it i typed out the ways

(Pump extracted means pump volume is 0 (no fluid inside))
(Pump retracted means pump volume is 100 (fluid inside)).
.
1.the situation starts with the pump extracted and the valve in T/A position
2.the pump goes to retracted position filling the chamber of the pump)
3.the valve switches to A/P pushing out generating pressure and flow (2 way valve is closed) and pump ends in extracted position.
4 the valve switches back to T/A and the pump retracts this drops some of the pressure but not between the 3 and 2 way valve..
5. the valve switches to A/P and the pump extracts.
.
or in other words
take fluid from tank and compress it using piston.
add compressed fluid (5 bar) to process and isolate it and maintain pressure.
take new fluid from tank and compress it using piston.
add compressed fluid (5 bar) to process resulting in 10 bar and maintain this pressure. 
take new fluid from tank and compress it using piston 
add compressed fluid (5 bar) to process resulting in 15 bar and maintain this
repeat

the main issue is that even if i seal the pump right after its output it leaks and causes the drop the pressure in a mere few seconds. This means that if i fill the pump push to 50 % and generate 20 bar the 20 bar is not maintained and drops fast. for what i can find in the tonnes of documentation i can not set the leak to 0 thus making my simulation not possible.

as my main assignment to design an motor speed/position control based based on the output pressure i stopped working on the simulation/plant model and focused only on the controlling part.

If an explanation/example could be provided in where with a piston-motion generates pressure (so not any type of constant pump block) and can increase the pressure each stroke. i could work this out into the possible solution.

[Landon Sugar]

I'm looking for a solution to a similar problem. Maybe you could try to simulate a hydraulic accumulator? Hydraulic accumulators (essentially) are used to store intermittencies of power, and release them at a constant rate.

For example, think of a bike pump. when you lift up the handle of the bike pump, nothing happens. When you compress it, a quick surge of air pressure comes out of the nozzle (but we want constant air pressure). When the bike pump is attached to a bike tube, the tube stores energy. It stores 10-15 strokes, in a singular place and holds it. Then when the pressure is ready to be released, you press on the stem valve, and the air rushes out at a (relatively) constant rate. The bike tube, in this example, acts as the accumulator.