Table of Contents
Stepper motor assembly V22.1
Render from CAD model September 2022.
Introduction
Stepper motor assembly is a rotating actuator for underwater applications. It is used in this project to rotate the wings of the towed ROV. The base design is made by the 2021 ROV group. 2022 group has done some modifications described in this document. I will in this document describe the main design decisions, discuss the problems, and explain ideas and improvements for future designes.
Oil system
This motor assembly has a oil system to keep the water out. Oil will help equalize the pressure between inside and outside of the wall. Water will not flow from a lower to a higher pressure. Keeping the pressure inside the tube higher than the outside, will ensure no water can enter. Filling the oil is done through a hole made with a syringe case. A vent is there to let the air out. The hole is pressurized and sealed by pressing down a syringe. A cover/cap is screwed to keep the syringe in place. Sunflower oil is selected because it is cheap, has better expiration date, and has slightly lower viscosity than alternative oils. All food oils will eventually turn rancid and smell bad. Keeping the motor house airtight will slow the expiration, because it is reaction with oxygen that turns the oil rancid. Other non-food oils were considered, but was too expensive, too thick, or a danger to the environment. No oil is good in the environment, but food oil is degradable, and therefore better than petroleum based oils. Silicone liquid/oil/grease can potentially be utilized, but none were found available in stores.
It is possible for oil to seep inside the cable and into the electronics tube. The cable penetrator on the tube side was filled with epoxy to reduce the issue.
Printing material
The stepper motor house is printed with PETG. PETG is stronger and has better water resistance than PLA. (layer adhesion is better). Part cooling fan speed was set to 0 to reduce warping of the part. Infill was set to 100% to increase water tightness and compression strength.
PETG has some downsides. The main problem is excess material sticking to outside of the printing nozzle. The material will slowly build a clump and then drop it on the print. The clump can make a gap in the surface. Oil was leaking from the parts with a clump in the surface. Low viscosity super glue was used to fill the gaps. This stopped the oil leak.
Magnets and sensor
The design uses a magnet sensor and two magnets for calibration. Magnets are placed with a 90 degree offset inside a magnet holder. The magnet holder is mounted to the rotating shaft. A magnet sensor is placed stationary next to the stepper motor. The wing mounted to the shaft, will be level when the sensor is in the middle between the two magnets. When calibrating, the program will find the location of both magnets, and then rotate to the center. More info on the magnet sensor is in the electronics part documentation. Detailed description of the calibration sequence can be found in the StepperWing library ReadME. TODO: link
Note: Magnet holder must be in correct position (calibrated) when assembling and disassembling the stepper assembly. The magnet holder can break the syringe if the position is wrong. Port side syringe is broken because of this.
Bearings and shaft spacer
The bearings will hold the shaft sturdy, while still enabling smooth rotation. Using two bearings helps with reducing the load on the stepper motor. The shaft spacer is there to keep the bearings apart. The spacer has holes for enabling oil to flow. One side has a tab to ensure none of the holes are blocked by the wall. When inserting the spacer, it is important to align one of the holes in the direction of the syringe. This reduces the chance of air bubbles getting trapped.
Problems with this design
Gears are inaccurate. The wing is not rigid because of small gaps between the gears. It is therefore possible to freely rotate the wing a few degrees. This could make controlling it harder.
One of the shaft seal is leaking oil. The 3D printed surface is uneven. Contact between surface and seal is not watertight. Sanding the surface will help, but it is hard and time consuming. A new design should have the seal on the outside of the house. Sanding would then be easier.
The shaft seal only worked on one of the assemblies. A new seal was glued to the outside of the stepper house. This did not leak.
Using a rubber endcap is a bad design decision. The cap will bend outwards when filled with oil, defeating the whole purpose of using syringe to make an overpressure. It is aslo hard to find and buy.
Ideas for future design
The gear wobble problem has to be researched first before making a new design. This might be a big problem for stability and control. Other solutions must be considered if it turns out to be a big problems.
There are many parts of the design that can be improved. As mentioned, the shaft seal can be mounted externally. It can be sandwiched between the main body and a plate. The plate can be tightened with screws, where the nuts are "printed in place" inside the main body. The main body can be slimmer, where the inner diameter of the tube is the same as the bearings. Magnet holder can be compacted to have the same outer diameter as a bearing. Main body can then be expanded to fit the stepper and magnet sensor. (Important: Test first if magnet sensor can operate this close to the stepper motor) The shape can be oval, with a flat side facing backwards relative to the ROV. The flat side of the oval can fit the cable penetrator and connections. The other side can fit the sensor and a syringe house. Vent and syringe can be mounted on the top cover facing the middle of the ROV. The cover can be screwed to the main body with a rubber seal in between. Main mounting point should be align with the outer bearing. This is were the force from the wing is highest. Another mounting point can be close to the servo, and screwed to the center aluminum extrusion.
Disclaimer
Currently the two built prototypes are bad and is not ment for reproduction or fixing. Some parts are glued/forced together and disassembling will be difficult. Port side has a broken syring part loose inside the house and can jam the system. One shaft was damaged after an attempt to disassemble the stepper and fix the syringe.
Motor shaft
The shaft connects stepper and wing. It is made from a 1m anodized aluminum shaft.
There was an attempt to make the shaft with hardened steel. The metal was very hard to cut and drill with the available tools.
The result was poor. Therefore the shaft was made with aluminum instead. Aluminum is weaker and has less accurate dimensions.
Build method
- Shaft was cut to size with a Femi metal saw and lubricant oil.
- Center hole was drilled with a mini lathe.
The hole was first drilled with a center drill. Then it was drilled with a 4 and 7,5mm twist drill bit.
Make sure the shaft is not wobbling before drilling. The shaft needed a lot of adjustment before it was stable.
It is also important to use lubricant oil when drilling. - Chamfer on the sides of the shaft was also made on the mini lathe.
A Form tool was carefully used at an angle. Moving the tool too fast resulted in the shaft rotating off axis. - Holes on side was drilled with a drill bench.
A straight line was first drawn to use as a reference. Hole on top and bottom must be aligned.
Hole for the wing was marked by placing the wing on the shaft, aligned with the line.
The first hole for the magnet holder, was measured from the edge.
The second hole was marked by aligning the magnet holder on the first mark.
All holes was first pre-drilled with a smaller twist drill.
Using a center punch is recommended to make sure drill bit is not drifting.
Drilling speed was selected from the table at the wall. (fast with small bits, slow with thick bits) - The two holes for the stepper and magnet holder was tapped to fit a machine screw.
- All hole edges was smoothed by hand with a larger drill.
Reference
CAD files is located on github: towed-rov-fusion
Electrical schematics is under the Electrical Schematics page.
Parts are found under the Materials page.
More info on this design in the Report - Summer 2021.docx