10/5/22 Altering the design to be cut by local laser cutting shop

 Colin Szeto

Lengthened the drone top plate so holes are not so close to the edge

Motor mount exporting in inches

Side wall exporting in inches

front plate exporting in inches

The below plates were laid out by hand 

12 motor plates

2 side walls

1 front wall

1 mid plate

1 top plate

1 bottom plate

8 leg mounting plates

Auto Layout Arrange (Fusion 360 Feature)

Utilizing this tutorial to auto layout the parts with the correct spacing and edges

Basic steps

Preferences > Preview features > manufacturing workspace > 3D arrange

Enable Nesting trial

1. Import components into file

2. Create manufacturing workspace (blue means within manufacturing workspace)

3. Right click create manufacturing model 

4. Right click edit manufacturing model to enter into the workspace

5. Create new component with the workspace as the parent

1. Under each component create a sketch 

6. Arrange, select all parts and select the sketch

7. For the parts left over select and arrange on second sketch

8. Create sketch, project all needed geometries into sketch (different that the intal component sketch)

9. Export sketch

9/15/22 Drone Legs and Mag Gripper Edits

Colin Szeto

Finished standing down clean the 19cm legs

Current leg options

For each of the leg options the magnetic gripper is a bit too high off the ground.

Adding in 10 mm spacer to lower the claw lower to the ground

Before

After

9/13/22 Gimbal CAD

 

Colin Szeto

Inspiration for the Gimbal here

Edited the forward facing camera to be gimbaled

Edited the downward facing camera mount to interface with the low light camera

Side view of the FOV of the downward facing camera

Back view of the downward facing Camera FOV

9/7/22 Attaching the Payload

Colin Szeto

Here with the payload plate mounted onto the rail inserts the pins were not able to insert through the holes

By adding spacers between the t rail and the payload plate this eliminated the force the payload plate was pushing on the rail blocks and allowed the pin to slide through the holes easily

Interference was found between the magnetic claw interface plate and the mourning rail. This was due to the socket head m3 screws not modeled into the drone CAD

The drone causes quite a bit of down draft, in testing environments we want to eliminate debri jumping from the ground and cracking the oak-d lite glass lense. 

We repurposed plastic sheets used for face shield as the camera cover

The plastic was cut to size to be glued onto the oak-d lite camera holder

The glue up was transferred to the vice to apply even constant pressure along the mated faces

This is the final result with the plastic adhered onto the front of the oak-d lite case

Here is both the HSI camera and the bottom facing oak-d lite mounted onto the payload rail mounted on the drone

9/6/22 Work Holding For Rails

Colin Szeto

Printed out in ABS a 3 part work holding Jig with the holes printed into the top and sides. Enables meeting the desired tolerances with available tools in the lab.

Attached this fancy clamp to the drill press for locating the rough location of the holes. There is some backlash between the x and y screws so holding the jig in place by hand is still necessary.

One problem has been chip clearing. This has been caused by keeping the part within the holders for all 3 drilled holes. This can be remedied through extracting the part out of the clamps and clearing the chips between the drilling.

New jig will take 2 hours.

Later on this jig was not required as the the cut rails were cut slightly larger than one another. This was unintentional but aided in work holding.

First to fourth

1.5085

1.5105

1.5235

1.5430

The material was inserted into the jig from smallest to largest. This meant that the jig would be slightly enlarged after each set of hole drilling. Yet due to the rail increase in length the jig would still snug around the part.

9/4/22 Designing the Downward facing OAK-D camera Mount

Bottom ISO view

Bottom view

Side view

In the side view we see that the oak-d camera field of view is not blocked by the magnetic gripper.

9/3/22 3D printing

 

Printing out the servo housing in ABS at UCSD printers, note the mounting holes will utilize screws threading the plastic rather than brass heated inserts

Magnetic latch in PLA at UCSD printers, note the mounting holes will utilize screws threading the plastic rather than brass heated inserts

9/3/22 Notes on the assembly of the claw

4.233 mm

Will have to cut down m3 16mm screw to 12 mm to have enough clearance 

Alternative: 10 mm screws or revert back to using heated inserts and screws.

When knob rotates there is still clearance

9/1/22 What About Galvanic Corrosion

 

Colin Szeto

Galvanic corrosion when contact with the ocean

Damon McMillan here speaks about galvanic corrosion when stainless steel is submerged with carbon fiber for a long period of time in the ocean

Video overviewing prevention methods of galvanic corrosion: here

About Galvanic Corrosion here

Ways of avoiding Galvanic Corrosion here

8/31/22 Fabricating and Mounting Minimized Drone landing plates

 

Colin Szeto

Fabricating the smaller landing plates

Configuration used for cutting the drone mourning rails. The rails were taped together to get a consistent cut across both pieces. A Backer was set up behind the chop saw to catch all the metal particles

Cut and drilled holes for the quick mounting blocks which integrate with the bottom drone plate

Drill bits used for fabricating the rail blocks:

3/16 for the pin hole

3 mm for the mounting holes

Able to mount the smaller profile landing plates

8/30/22 Designing Minimized Drone Landing Plates

Colin Szeto

Deemed that one clamp was enough on the motor arms. Changed the clamp location directly above the legs

8/30/22 Fabricating Drone Landing Plates and Assembly

 

Colin Szeto

Utilized desktop cnc, did not use air compressor and water mixture, the carbon fiber dust got everywhere

Edges of the parts before wet sanding

Edges after wet sanding

Front view of the drone landing plates connected to the legs. Utilizing 2 clamps instead of the intended 4 clamps. This decision was made as only had 8 clamps on hand. Demonstrated that landing plates only required two clamps rather than 4 as there was more than enough clamping force