1. Updates Summary
This week I mainly explored two topics:
First, I studied the wave breaching simulation tool and attempted to recreate it in C++.
Second, I experimented in designing a new waterproof shell that includes a membrane mechanism to help with pressure measurement.
This weekly update mainly focuses on the second exploration. After talking with Mark, I decided to concentrate my effort on the hardware platform, as it will be the major content for my upcoming paper.
2. Detail Updates
From the previous experiments, we realize that it is difficult to get an accurate measurement of ambient pressure with on-board barometer, mainly due to two reasons:
a. The shell adds an attenuation effect to pressure reading. As a result, the pressure change of the air within the shell does not directly reflect the surrounding water pressure change.
b. Electronic parts such as ESCs and crazyflies would heat up during the flight. Hence, the air pressure within the shell will be influenced by temperature change.
A flexible membrane design is proposed to offset the side effects mentioned above. We will wrap the electronic chamber with a layer of flexible membrane material. With the presence of membrane, the air pressure inside the shell will always balance against the ambient water pressure. Thus, we will be able to directly measure ambient water pressure by measure the air pressure sealed by the membrane.
Three iterations of CAD designs were attempted:
Iteration 1 features a side-aperture with a circular membrane holder attached to the shell with six screws. The disadvantage of the design is that the side-aperture makes the design asymmetric and the window is too small for the shell to get enough volume change under pressure. Also, the fastener penetrates the shell, inducing potential problems in terms of water proof.
Iteration 2 features a top window design. Extra extrusion is made in the shell to waterproof fastener connections and a larger window is cut out to ensure sufficient volume change. However, this iteration of design requires to drill into the membrane, which causes stress concentration and may potentially tear the membrane apart during tests.
Iteration 3 travels from the realm of over-design back to simplicity. It chops off the whole top surface as the membrane window and it has a lip and grove that helps rubber band tie down the plethora of the membrane.
3. Planned work for near future
The CAD design is sent to Jacobs’ Maker Space and is being printed. Once it is manufactured, I will test it, assemble the drone and adjust the scaling parameter of the depth estimator. Afterwards, I will carry out another underwater hover test to evaluate the performance of the system after modification. Meanwhile, I will start composing the first draft of the paper of UAUV and documenting down design decisions related to UAUV.