Embedded Files
Automated Pressure Gauge Filling Station
GAUGE HOLDER
The previous iteration of the gauge holder stand from the Winter 156B team had issues that needed to be addressed. The pillow block bearing was not strong enough as a 3D printed part and too complex to be CNC. Thus, a redesign was required. Ultimately, the pillow block bearing was replaced with a simple aluminum bearing block that increased both manufacturability and durability. The vertical shafts were wrapped in PTFE tubing to function as sleeve bearings, allowing for tolerance in the alignment of the two vertical shafts without risk of binding. Horizontal rods were press fit into the bearing block to provide support for the gauges. Gauge holder with PM25, PR35, and PT45 gauges (from right to left).
CONVEYOR LOOP
The conveyor loop system was chosen due to its perceived safety and minimal footprint. Integrating the loading and unloading areas into a single loop reduces the amount of material needed for each section. While the other two systems required an incline to introduce the gauge holders to the conveyor belt, every surface that the gauge holder rests on in the loop system is flat. This reduces the probability of a gauge tipping or spilling fluid. While the rollers were a more expensive solution than low friction material, early risk reduction tests comparing the two overwhelmingly proved the efficacy of a roller track over a low friction track. Even with the roller track to significantly reduce friction, concerns about friction between the guide rails and the gauge holders remained. To ensure that the conveyor belt could push all gauges through the curved roller track, a free body diagram was constructed. A quasi static analysis was justified because of the low conveyor speeds.
Based on the analysis conducted, it was determined that the higher friction SBR rubber conveyor belt was required to produce the coefficient of friction necessary to avoid slip between the gauge holder and the belt material. In order to choose between an off the shelf conveyor belt and a DIY conveyor, a DIY system was designed and the prices were compared. While the DIY conveyor was less expensive than the off the shelf conveyor, it was determined that the potential risks of constructing a conveyor system, including belt tracking and reliability, were too high to justify the reduced cost.
PLC information flow
PLC FILLING STATION
Previously, the team designed the system to send pulses of digital square waves that have high and low signals of 0.01s to activate the function of stepper motors. However, because the PLC refresh rate of 50 Hz was too slow, the team decided to increase the filling speed with a faster controller first through an Arduino and later the Controllino. Because the Arduino/Controllino has a communication rate of 490 Hz, sending discrete signals would be significantly faster. Therefore, the PLC system was left intact with the Arduino responsible for only faster controls of the stepper motors. Because filling times were shortened, the PLC was simplified to a time tracker. Once a proof of concept of the Arduino showed its potential to significantly increase filling speed, the team decided to proceed with the Arduino integration into the existing PLC system.
Moreover, the team decided on heating the glycerin to reduce the viscosity and increase the fluid flow in the funnel. This design decision was necessary because even with a faster peristaltic pump, the fill rate inside the gauges would still be constrained by the glycerin build-up in the funnel. The team decided to proceed with a heating tape to wrap around the copper tube that reached a maximum of 43 degrees Celsius. After doing analysis of the necessary wattage given gauge volume, heating time required, and temperature parameters, the team found the optimum option with a 1” x 48”, 288W, 120V tape. A Reotemp temperature probe was fitted to monitor the temperature of the glycerin.
PLC Main Front Box (Left) and Back Box (Right).
Heating system. The peristaltic pump draws glycerin from the bottle, and the heating tape heats the glycerin.
Next, the PLC read the 4 combinations of inputs. If the sensor detected a new color, the system would correspond the filling time to that of the gauge size. If the sensor detected the gauge holder's aluminum color, the system would preserve the previous colored gauge's parameters. The color sensor continued this cycle until it detected a purple color reading. Then the conveyor belt movement would stop, hence indicating all gauges have been filled.
CAD of temperature probe fittings (left) and temperature probe and readout mounting position (right).
AUTOMATED GAUGE SIZE DETECTION
The gauge holder detection system included a thru beam sensor and a color sensor. The thru beam switch was reliable for detecting the presence of the gauge holder. For size detection, a color sensor was chosen because the gauge holder orientation had no effect on the accuracy of the size detection, which significantly reduced the chance of error, and thus, the chance of overfilling a gauge.The CM1000 sensor was connected to the Color IO board. The Color IO board allowed the programming of the sensor using an application that ran on Windows software. It also connected the sensor with the digital inputs of the PLC. There were four digital outputs from the sensor, each which output a 12V signal. Depending on which color the sensor detected, a certain combination of digital outputs were activated. With this setup, the sensor was capable of differentiating up to 15 different colors.
Position of thru beam and color sensors.
Color sensor concept. Gauges of the same size are placed next to each other.
The PLC updates the fill time if it detects a colored baseplate.
Performance Results
Conveyor Loop
The conveyor loop system successfully pushed 4 gauge holders and 43 gauge holder baseplates without stalling. The conveyors were also able to stop a gauge holder under the filling station accurately.
PLC Filling Speed
The filling speed was dependent on the communication speed of the Arduino. A shorter delay corresponded with a faster filling time. Below are the calibration curves for 3 delays (10, 15, and 20 milliseconds) the team experimented with to obtain the optimal balance between filling speed and resolution.
Ultimately, the 15ms delay was chosen. This filling speed resulted in the filling times listed in the table below. The upgraded PLC was capable of filling the largest gauge (PR60) in under 5 minutes, which was an improvement of 15 minutes from the PLC without the Arduino.
Filling times for 6 different gauges.
Note: PT45 required multiple pauses in between filling. Its actual fill time is around 260s.
Page updated
Report abuse