79 lines
3.4 KiB
Markdown
79 lines
3.4 KiB
Markdown
# Agenda
|
|
|
|
- [ ] Workshop
|
|
|
|
- [ ] Control System
|
|
|
|
- [ ] Electronics
|
|
|
|
- [ ] Prototype
|
|
|
|
- [ ] Summary von USA
|
|
|
|
# Workshop
|
|
|
|
- Research summarized in [sheet](https://docs.google.com/spreadsheets/d/1YP47rCptCled3IzDZ0RmcIJ_H5GzLnqsTZDGfUqnUEQ/edit#gid=0)
|
|
- Most workshops work in a offer based way. We need to send them an email with the enquiry and they will answer with the lead time and cost.
|
|
- Do we want a workshop that helps with the design? #question
|
|
|
|
# Control System
|
|
|
|
# Prototype
|
|
|
|
- What is the status? #question
|
|
- Is it Inigo's or my job? #question
|
|
- Does it make sense to keep trying with old design? #question
|
|
|
|
|
|
|
|
## My thoughts and strategy
|
|
|
|
- We're going down a rabbit hole. We do not understand the system properly and we're burning our time and money with this.
|
|
- We need a PX4 expert --> either one of us learns it and becomes a good pilot or we hire someone who knows it.
|
|
|
|
- We need to understand _exactly_ how motor commands are issued. #important
|
|
- Wouldn't it make more sense to redesign it properly before putting all our efforts in this?
|
|
|
|
- Within the design we can include the possibility to change certain aspects to make up for uncertainties (e.g. threaded holes to attach the blade holders at several different azimuth angles and not only one fixed one)
|
|
- Include learnings from current swashplate failure
|
|
|
|
- Blade holders are not strong enough --> one solid piece
|
|
- Blade flapping (bending) is considerable --> how is the distance between the two rotors calculated? #question
|
|
- We do not know how the system works --> get to understand it properly
|
|
- Swashplates always introduce vibrations (why?), how can we minimize those? #question
|
|
- Use MatLab and books to create a model and a simulation (See Nasa papers for simulators)
|
|
- Make a prototype of the prototype, where both swashplates are directly actuated by servos. This removes coupling between the two rotors and allows to measure the effect of each swashplate individually.
|
|
- We should get our hands onto a force-torque sensor --> System identification becomes easy
|
|
- Can we get in touch with experts that have designed other helicopter systems? [Swiss helicopter startup](https://koptergroup.com/)? Martin Stucki (Founder) #question
|
|
- I think we should invest in a well thought out test bench. It allows to thoroughly analyze prototypes in a safe environment where we can measure important aspects and learn from them. It allows to create a more scientific approach to the design. Later it will help with the control system design.
|
|
|
|
- Important components are:
|
|
|
|
- Slowmo camera
|
|
- 6 axis force/torque sensor
|
|
- Time synchronisation
|
|
- Desktop computer to control and monitor everything
|
|
- Temperature and humidity sensors
|
|
- Can it be combined with a 6 dof robot arm to simulate different inertias and therefore different drone bodys? This might be a very fast iterative approach to control system design. #question
|
|
- Is peek a viable option to print the swashplate? To iterate faster through prototypes?
|
|
- Are those tight tolerances really needed? It makes it very expensive to manufacture.
|
|
|
|
# Electronics
|
|
|
|
- Hire or not?
|
|
- Outsourcing or not?
|
|
|
|
- Outsource small scale --> upwork
|
|
- Outsource large scale --> Thundercomm
|
|
- Power System:
|
|
|
|
-
|
|
|
|
|
|
|
|
|
|
# Summary USA
|
|
|
|
- Does OneSec Inc. Exist? #question
|
|
- How many discussions did you have?
|
|
- Verdict of evaluation and interest? |