Week 8 Design Review
Team Vision for Week 8 Design ReviewOur vision for week 8 was to:
- Continue along the process flow chart on designs
- Gain design approval from Ed
- Gain manufacturability from ME Shop
- Create drawing package
- Build dancer subsystem
- Create build and test plans
Our vision for this review:
- Information sharing
- dancer progress update
- design progress update
- project outlook
BOM Update and Design Completion PercentageFor the majority of our components, we are at the RIT approval point, seeking guidance from Hanzlik and the machine shop. After meeting with the guys in the machine shop, designs have been updated to meet manufacturability and commercial grade standards. However, there are still a couple major design areas that require review. The entire traversing pulley system still needs to be designed. The frame is almost complete, but integration with full assembly as well as fixture assembly need to be considered. The flexible coupling for the spools and the control cabinet still require specification.
Live Revision of the BOM can be found here.
Dancer System Assembly
- Design is finished
- Machining is finished
- Mounting system is assembled
- At point for proof of concept and knowledge gain
- Dancer Pulley Mounting System, complete
- Bracket #1 hardware acquired
- Bracket #1; complete - 
- Bracket #2 machined
- Bracket #2 hardware acquired
- Bracket #2; complete - 
- Dancer Pulley System is now able to be tested
- Gather proper pneumatic connections
- Connect pneumatic system to supply air
- Test Dancer Pulley System
Supply Spool Purchasing
ME Shop Conversation and Changes Update
Fixture Frame Design
- Smaller: greatly reduce weight and cost
- Gussets: plan to add gussets for support
- Material: A36 ground hot rolled plate with corrosion resistant coding as opposed to expensive stainless steel
Spool Subsystem Design
- 3 part separable design: increases machinability, decreases material cost
- Locating features: ensure concentricity to decrease vibrations due to imbalance
- Separable design: increases machinability, decreases material cost, parts can be replaced without total remake
The design for the guide pulley system has remained the same as seen in the week 5 design review. All components are designed for reuse on the final machine. Integration with full assembly has been considered. Final engineering vibration and structural tests are being looked at before seeking approval from the machine shop.
Traversing PulleyThe speed for the traversing pulley is a function of the radius, wire-laying angle and wire speed. In simple terms, it can be thought of as the distance per wrap over the time of one revolution. However, as the spool radius decreases, the angular velocity increases; as such, the time per revolution will decrease.
- V = wire velocity [m/s]
- R = spool radius [m]
- t = revolution time [s]
By looking at the geometry of the spool, we can determine that the distance per wrap, x, is related to the spool diameter and wire-laying angle.
- x = distance traveled per wrap [m]
- D = diameter
- Theta = wire-laying angle (relative to surface of spool)
With this information, we can solve for the traversing speed, T. We can vary the wire-laying angle or the distance traveled per wrap.
We decided to ideally bound the wire laying angle to be between 85 and 89.98 degrees (laying wire right next to itself at max radius for takeup spool). This results in the following extreme cases for the takeup spool.
Takeup SpoolMax Velocity = 1.0025 m/s
- R = 0.06 m
- V = 18 m/s
- Theta = 85 degrees
Min Velocity = 0.00253 m/s
- R = 0.12 m
- V = 10 m/s
- x = 175 um
Dancer Pulley and Spool MotorsThe dancer system can be thought of like a flow issue - you have a certain amount of wire coming into the system, and a certain amount coming out. The dancer will accommodate for the velocity mismatch by changing the path length. If we define vm to be the wire velocity from the motor, vp to be the wire velocity on the guide pulley (and presumably the guide rollers), we can define the change in the path length to be the difference between these two multiplied by the change in time. By convention, extension will be defined as positive displacement. Assuming we can decouple from the traversing pulley system and look only at the dancer system, we can see that the extra path length would be split half and half in the 180 degree wrap, such that the change in path length is equal to twice the dancer extension, X.
- L = Path Length
- vm = Motor wire velocity [m/s]
vp = Pulley wire velocity [m/s]
- NOTE: This is a measurement taken from the guide pulley on the other side of the dancer system, and should reflect the guide roller wire velocity.
- X = Dancer displacement
Taking the derivative with respect to time, we find:
As such, we can find out the exact tangential velocity mismatch between the rollers and the spools if we can measure the dancer velocity. Using a position sensor with a known polling rate, we should be able to accomplish this. Thus, we have the following:
- No slip conditions on pulley
- Sampling/polling rate of dancer position sensor is known
- Dancer system can be decoupled from effects of traversing pulley
- Position sensor
- X = Dancer Position
- t = Refresh rate (ideally known/set)
- Guide pulley with shaft encoder
- Rp = Pulley radius (constant)
- wp = Pulley angular velocity
- Motor shaft encoder
- wm = Motor angular velocity
- Position sensor
Using the relationship defined above, we can plug in radius*angular velocity for the wire velocities, and solve for the instantaneous radius of the spool, R.
This gives us several possible control options:
Possible Control Options
- Change the instantaneous value of the reference signal (angular velocity) to reflect the instantaneous radius
- Pros: Better-defined relationship between motor and reference
- Cons: Processing overhead due to re-calculating reference
- Calculate the reference signal magnitude at runtime relative to pulley - act as if the motor is driving the pulley rather than spool
- Use dancer feedback to accommodate for the mismatch between the pulley and spool
- Pros: Only need to calculate reference signal once per run (or upon input parameter change)
- Cons: Feedback to motor is less direct
- Change the spool motor velocity based solely on the dancer extension. Speed up motor if dancer is going inwards, slow down if dancer extends.
- Pros: Simple to implement
- Cons: Not as precise, possibility of continuous speed fluctuation
Remaining DesignsThe design for the pulley has been reviewed by Hanzlik and will look very similar to the design shown below.
Shared Team Vision for Week 11 Review
- Leave enough detail to pick back up right where we left off after break
- Prepared to present ~90% of our designs and drawing package to GF for review at week 11 meeting
- Present controls in algorithm form
- Present traversing pulley design progress
- Present dancer subsystem air cylinder tests and resulting knowledge gain