Embedded Files
Project Deliverables
Design and assembly of both static and dynamic testbeds
Evaluation of rolling-diaphragm actuator performance in terms of position transparency, force transparency, response time, system stiffness
Suggestions of improvements on rolling-diaphragm actuator system
Rolling Diaphragm Actuator Functional Requirements:
Force tracking errors of < 15% of the reference force over 0-10 N
Position tracking errors < 2 mm on average and < 5 mm at maximum over 0-10 cm
Can detect membrane puncture during needle insertion into phantom tissue with 90% accuracy
Can detect relative changes in stiffness of 20%
At least 2000 Hz recording frequency for DAQ
One DOF (degrees of freedom) >300 Hz bandwidth and three DOFs >40 Hz bandwidth
Test Setup:
Our ultimate test setup connected a pair of rolling diaphragm actuators via individual water and air hoses, and placed sensors throughout the system to evaluate force and displacement transparency as well as system stiffness. Each actuator includes a pair of DM3-50-50 rolling diaphragms from IER Fujikura, 3D printed pistons, an air cavity and a water cavity. The experimental testbed includes two force and position sensors, an air hose, a water hose, a motor, linear ball screw, and a linear slide.
Sensors are used to determine and quantify the:
impact of friction on motor dynamics (derived from motor speed vs. time)
backlash for data calibration (derived from encoder position vs. time)
stability of input power supply (derived from motor input current vs. time)
sensor noise and assist in controller design (derived from load cell force vs. time)
positional transparency of the system (derived from output displacement vs. input displacement)
force transparency of the system (derived from output force vs. input force)
response time (derived from Bode plots of the system’s frequency responses)
system stiffness (derived from both force and positional data)
Key Components:
Position Sensor:
The purpose of the position sensor is to detect the displacement of an MRI needle at the output compared to the doctor’s hand motions from the input side as he/she is inserting a needle. When the rolling diaphragm transmission activates needle insertion, the position sensor will detect and record the motion so that the system can be characterized by analyzing the recorded data.
For this project, the position sensors must detect linear motion over a range of 0 – 10 cm, position tracking errors < 2 mm, high resolution of at least 12-bit (4096 PPR), high accuracy, small size and light weight.
Sensors from US Digital, AMS, and CUI Devices were considered but the EM2 linear encoder from US Digital was chosen due to its high resolution, availability and lead time, size, weight, and two channel quadrature with index feature.
Force Sensor:
The force sensor is used to track the force input of the rolling diaphragm hydrostatic transmission and the force output through the rolling diaphragm hydrostatic transmission.
The main requirement for the force sensor is that it can have sufficient measurement range and accuracy to track and monitor the range of force generated by the needle piercing the skin and in human tissue and movement in the body. Therefore the sensor is at least within the range of 0-6N. However, it is better for the sensor to exceed twice the target range to obtain a more accurate measurement. The force tracking error is less than 15% of the reference force. And real-time feedback of measurement data (low delay).
The system uses an S-type load cell DYLY-103 as the sponsor provided this sensor.
Actuator Mount:
The rolling diaphragm mounts are employed to fasten the rolling diaphragm cylinders on the optical table. Each mount supports the weight of a cylinder and restricts it from undesired motion when horizontal forces are applied on it.
When a maximum horizontal force of 20 N is acted upon the rolling diaphragm cylinder, it needs to prevent it from undesired motion and keep it parallel with the table. In this case, this mount should provide enough support without any translation, deformation, or fracture. It needs not only to support the cylinder but also to set it at a desired height so that each cylinder is aligned with other components to minimize the error due to misalignment.
This component was designed by our team.
DAQ:
One of the driving factors behind selecting an appropriate data acquisition system were enough signal channels and portable-type design. The DAQ needs to contain enough input and output channels to deal with both digital and analog signal streams from and to the rolling diaphragm system.
Another driving factor is a user-friendly interface and high connectivity. The interface needs to be easy to operate so that users are able to process experimental data with simple commands rather than a heavy load of programming
Final Test Results
Part 1 - Air Hose Configuration Optimization
The objective of this testing was to determine the optimal air hose configuration to minimize the resistance due to the variation in air pressure. Four different types of hose connection were tested to obtain the resistant forces respectively. For the initial hose configuration, the air chambers of the two rolling-diaphragm actuators were connected to an air compressor individually and supplied by two air flows with the same pressure. For later configurations, the system was tested with unpressurized chambers, and then chambers interconnected in two configurations. Ultimately it was determined that the optimal hose configuration would be the T-junction one, which generates only about one third of the resistant force of the original design.
Image - various air hose configurations at 60 psia.
2 - Air Pressure Optimization
The purpose of testing was to determine the best operating air pressure to optimize the position and force transparency between input and output. It was found no matter which water hose length was used, the pressure of 60 psi provided a slightly higher force gain than other values at a certain moment, although the output forces in all of the tests finally achieved around 80% of the input forces in terms of the magnitude. Therefore, it can be concluded from this set of data that 60 psi is the most suitable pressure for the force transparency. Similar tests were carried out to determine the effects of pressure on the position transfer relationship, but it was found the magnitude of air pressure exerted has a little effect on the position transfer effectiveness.
Image - position performance at various air pressures
3 - Hose Length
The purpose of testing was to determine the best hose length to optimize the position and force transparency between input and output. Two different hose lengths were tested (30cm and 69 cm) under the 60 PSI and T-junction configuration. Data indicates, the transfer coefficients of force and position with a longer tube are both higher than those with the shorter tube length. This means a longer tube performs best.
Image - various air hose configurations at 60 psia.
Page updated
Report abuse