# System Design III: A Cheap Scheme With More Possibilities

According to the disadvantages of Design II, we make the following improvements:

1. Put the transducers outsides: the transducers won't interfere the flow anymore and it is easier to adjust their angle!
2. Use a tube as the main part: the radius is no bigger than 10mm, so that the powder won't get deposited easily. What's more, it is much much cheaper!

As what is shown in the image above, with a rotatable structure, it is easy to change the transducer angle. Additionally, with a slider, we can flexibly adjust the position of transducer on the tube. The sliders are mounted on the tube tightly, but there is still air filling the space between the transducer's piezoelectric chip and the tube's surface, which make most of ultrasound reflect on the tube-air boundary. In order to make ultrasound transmit more into the tube, we use a kind of soft rubber, which is actually a kind of cheap keyboard cleaning mud that we can easily buy, to fill those empty space and build up a route for ultrasound to transmit from piezoelectric chip to the tube outer surface and go inside more smoothly. With these efforts, we can gain a stronger signal of reflected wave without influencing the flow, and a wider flow speed range as a bonus! Although oxidized aluminum particles still settle down a little at the bottom of the tube, the situation is improved a lot.

However, there does be a replacement for these $Al_2O_3$ powder. By introducing some air into the pump system purposely, we can mix adjustable amount of bubbles into the flow. These bubbles act somehow like a real particle.