College - Author 1
College of Engineering
Department - Author 1
Biomedical Engineering Department
Degree Name - Author 1
BS in Biomedical Engineering
College - Author 2
College of Engineering
Department - Author 2
Biomedical Engineering Department
Degree - Author 2
BS in Biomedical Engineering
College - Author 3
College of Engineering
Department - Author 3
Biomedical Engineering Department
Degree - Author 3
BS in Biomedical Engineering
Date
3-2022
Primary Advisor
Christopher Heylman, College of Engineering, Biomedical Engineering Department
Abstract/Summary
Our team was asked to design a syringe pump that would deliver fluid at a controlled flow rate to cells in a microfluidic device. The design process of our syringe pump proved to be a very dynamic one. The beginning research of both microfluidic devices and existing syringe pumps helped our team get an idea of ways we could implement existing aspects that work into our design. There were many existing devices that resembled the one that we were asked to make closely; however, due to our resources as students, we had to be a bit more creative in figuring out how to afford and assemble each component to the best of our abilities. Developing customer requirements was a huge step in the process of understanding what exactly you as our customer wanted to see delivered in our syringe pump. The main requirements of our pump were that it was able to deliver accurate shear stress values so that they could mimic those found in true physiology, that it was able to deliver an accurate flow rate to the device, that it was easily usable, and that it was compact to both fit in a desired location and have ease of mobility when needed to be moved to or from that location. Next, it was our job as the engineers to turn those requirements into quantitative engineering specifications that our device needed to meet via testing of the device once the prototype was finished. Once we determined what numbers needed to be hit to quantify the requirements set by you, we were able to create a network diagram of tasks in order to organize the design, manufacturing, and testing processes that we had ahead.
Our design process then became a series of brainstorming via tools like a conjoint analysis, morphology, and Pugh matrices. We did these exercises in order to compile a multitude of ideas for each component of the pump to determine which combination of these ideas would produce the optimal pump that is attractive to the user and does the best job at meeting the customer specifications. We determined the main functions of our pump were inputting flow rate parameters on the interface, having a power source for the pushing mechanism, the physical pushing mechanism, and lastly the mechanism through which the fluid would be delivered into the tube. Ultimately, through the many exercises as well as iterations due to a multitude of realizations down the road, we settled upon using a stepper motor linear actuator for the pushing mechanism and a screen with buttons for the input from the user, powered by a 24 V DC Power Supply and connected by a needle attachment to the syringe. Next came acquiring the materials and aspects of the pump that were to be purchased from a manufacturer as well as designing the aspects that we were going to manufacture ourselves. The primary component of our design that we purchased was the FUYU stepper motor linear actuator, to which we programmed electrically and designed adapters to fit onto. Our electrical programming revolved around the Arduino UNO and the Sketch coding software. The chassis was our last component to design, and its main purpose was to keep the user safe from any potential harm from the pump and protect the pump from any water or other wear. When we had performed the Hazard Safety Assessment, we determined a lot of the risk involved the user having their hands in the pinch points as well as having the device fall on the user, both of which were mitigated by having a chassis that covered the pinch point and made the device more compact and mobile. Once we had those components designed, we determined how we would both manufacture and assemble the final prototype. These plans were surely dynamic as we changed materials and found new ways to better manufacture each piece. Critical changes included changing the chassis material from acrylic to polycarbonate, and thus changing the manufacturing process from laser jetting to water jetting to using a variety of saws to cut the pieces. Another critical change came after having manufactured the pusher block adapter, as we were sent back to the design process when the adapter did not perform the way we wanted it to. Additionally, the electrical side of our design manufacturing had to be iterated multiple times as we determined what was feasible and still effective for inputting the parameters. Our design changed from a 4 x 4 keypad to two buttons, one increasing the flow rate value and one decreasing the value. Once the prototype had been built, it was time to verify that we had made a device that met the customer specifications. We created protocols for how we would test these specifications and executed each of the four, the most time-consuming ones being the flow rate and shear stress tests.
Our testing plans for shear stress included both an analytical COMSOL simulation through the solid model of the microfluidic device as well as physical testing of the velocity of the particles moving via the LabSmith Micro Particle Image Velocimetry microscope. The physical testing was to verify that our analytical model accurately displayed what velocity and thus shear stresses the cells in our microfluidic devices would be experiencing. Next, we tested flow rate via running water through our pump at specified flow rates for a given period of time, measuring the mass acquired on a sensitive scale to back-calculate what flow rate was actually being delivered. Additionally, we used a gauge to measure the displacement of our pusher block over a specified time to first ensure that the correct speed was being programmed to the motor. In terms of surface area testing, we simply used a ruler to measure the dimensions of the bottom of our chassis to verify it would fit in the desired location in the lab. Lastly, our ease of use testing included simply numbering the steps in the operations manual.
Ultimately, our data showed that we did in fact create a pump that received an input and delivered a controllable flow rate and shear stress to the cells in the microfluidic devices, all while being compact and easily usable. After inputting a flow rate of 28.8 ml/hr, we measured the delivered flow rate to be 25.5 ml/hr, which was within our target percent error range of 15%. For shear stress, when entering a flow rate of 75.8 uL/hr, our physical testing showed a particle velocity of 295.6 um/s and our COMSOL velocity showed one of 358.91 um/s, putting these within range of our 20% error goal. We measured the bottom surface area of our pump to be 431.85 cm^2, which was well within our specification of 695 cm^2. Lastly, we measured 5 steps to program our device, which was our target specification. There were surely limitations to our data, as when flow rate decreased to smaller and smaller values it was increasingly harder to acquire data, and then additionally extremely difficult to have that data be accurate. Thus, at the flow rate of 0.76 uL/hr, which is the flow rate at which the pump will typically be used at, both the shear stress and flow rate specifications were not met via our testing. There are a multitude of reasons why our data may have been skewed, and we have plans for future testing to discover where errors might be introduced in our pump. Overall, our team learned much about the design process and grew as engineers while designing this syringe pump.
URL: https://digitalcommons.calpoly.edu/bmedsp/150