Big Project | Project | Project Manager |
|---|---|---|
Comp 2025 | Payload |
Task Description
In order to deliver water, we’ll need somewhere to store it, this task is to design what we use to do so. We’re open to ideas for how we do this, but it needs to be fully sealed other than the pump inlet and outlet.
The tank can be made by us somehow, or can be a modified existing water tank of some kind, but you’ll need to figure out how this mounts to the drone.
If you have ideas to be able to “dump” the water from this (ie don’t need an exit pump) that will not leak, please write them out and we can discuss, but the initial scope of this will be just a tank solution.
Constraints
Constraints | Written By | Append Date |
|---|---|---|
holds ~2.5L of water |
| |
total weight (no water, no pumps), <=800g |
|
Tasks
Research
Tank design
Component sourcing
Component testing
Manufacturing / mechanical installation
Integration testing
Task Progression/Updates
Author: Alison Thompson Date:
Initial Ideas
We can experiment with how water-tight various 3D print materials are at different infills, then pick the lightest.
Ideally light weight PLA would work for this if we’re 3D printing, but off the shelf tanks should be investigated as well.
...
Research
Really liking the idea of an OTS container because it’ll be more reliable than 3D printing things, from my first look. Amazon link that I’ve been searching: https://www.amazon.ca/s?k=plastic+water+storage+tank+2.5L&crid=2XY0SQTAAPRKO&sprefix=plastic+water+storage+tank+2+5l%2Caps%2C149&ref=nb_sb_noss. Uline also a good place to look.
Tank requirements
watertight
able to install a pump/valve
lightweight
Tank nice-to-haves
baffles in the interior to prevent sloshing
Some COTS options:
https://www.uline.ca/Product/Detail/S-22889/Jugs/Jerrican-1-1-4-Gallon-Natural (4.731765 litre, 410 g)
must buy 4x at a time, 13$ each. 52$ base
https://www.uline.ca/Product/Detail/S-22920/Jugs/Square-Utility-Jugs-1-Gallon-Natural (3.78541 litre, 440 g)
must buy 12x at a time, 3.80$ each. 46.60$ base
Could install baffles in a COTS jug by cutting a slit into the side, pushing a flat piece of plastic in (about 50-75% of the way), and sealing the base of this piece into the cut exterior with epoxy.
Re: Oct 3 mech meeting, will be trying out 3D printed tank before buying COTS tanks.
Calculations for tank volume in PLA: https://docs.google.com/spreadsheets/d/13ZkewgjBpSqwBHvnumQ3UP_YPBGpoKL36r2nGgX6v9E/edit?usp=sharing
Assumptions for this tank:
Cylindrical
Fillet top and bottom 1 cm
full area filled-in top and bottom
thickness 4 mm
Tank option | Pros | Cons |
|---|---|---|
PLA cylinder | cheap easy to prototype and change can set to specific max capacity | has to be water treated may be fragile |
Lightweight PLA cylinder | ||
COTS jerrican | designed for carrying water (watertight) | expensive |
COTS jug | designed for carrying water (watertight) | expensive |
1/4” tube activity
tube <> aquarium pump (1/2” female BSP to 1/4” OD push) https://www.amazon.ca/uxcell-Straight-Connect-Purifiers-Connectors/dp/B07WGRLN3D
...
Status update oct 10
...
Tank Design
Objective is a tank that holds 2.5 L of water securely and can infeed/outfeed water at request.
Nice-to-haves include vaguely aerodynamic design, easy to manufacture, assembles and disassembles quickly. Uses parts we already have/ are easy to get.
Captured in 2024-10-10 Mechanical Meeting Minutes
Idealized connection architecture shown above.
CAD shown below.
This screenshot does not have an o-ring groove. The printed one in ABS shown below includes that update. Bambulabs 0.20 mm white ABS
Component sourcing happening concurrently with design process.
Bunch of issues with parts arriving wrong/ not arriving.
With reference to purchase threads https://discord.com/channels/776618956638388305/1299101413342843011 and https://discord.com/channels/776618956638388305/1301328298659418112
I was looking at options with @ravoik, we can replace both fittings that did not arrive / arrived wrong with some tomfoolery (epoxy) we can throw the BSP 3/8"<>push-to-connect 1/4"OD directly into the tank. this connector:
https://www.amazon.ca/uxcell-Straight-Purifiers-Fittings-Connector/dp/B07WCS3WTXaction plan would just be to return the 1" (wrong size) o-ring connectors, Sarah goes ahead and gets a refund on the ones that didn't arrive, and we order another 15$ connector (and maybe buy some epoxy).
CAD Review
Notes from 2024-11-06 Pegasus Payload & Landing Gear + Eclipse CAD review
Water tank solution:
2.7 L internal volume > 2.5 L
Change bolts attaching the top plate and the tank to M3 instead of M5s
Polycarbonate for the top plate
look for gaskets instead of a funny o-ring
Float sensor to measure water level for payload
Chamfer the tube piece
flip the counter-sink screws
Maybe countersink the top plate, maybe not. Easier to manufacture if no countersink. ALso need to add space for the socket heads at the drone center block.
Standalone assembly (TANK_ASSY)
Unsuppressed on main assembly
Getting floats to check water levels - https://www.amazon.ca/Micro-Traders-Plastic-Dispensers-Humidifiers/dp/B0CXSH7YDF/ref=sr_1_13
Debate between this or current sense. issue with current sense is that tank must be full to generate a current feature. Requires a current sense shunt resistor on the water control board.
Floats | Current sense |
|---|---|
Easy to read (binary) | Lighter |
More infra required | Don’t have time to implement atp |
this table is kinda throw tbh but see https://discord.com/channels/776618956638388305/1112817387734057020/1307368097547812875 https://uwarg-docs.atlassian.net/wiki/spaces/Meetings/pages/2838495233/2024-11-14+Comp+Arch+Sync#Mechanical
Worksession 2024-11-16
Testing again. somehow failed to realize that the pump is NOT N/C (normally closed), and there needs to be a resolution to close the opening so that the water won’t drain out during inoperation.