Team Vision for System-Level Design Phase
The team goals for the System-Level design phase are to identify potential design changes to the original heart pump and circulatory system design that would make the system functional and user friendly. Brainstorm concepts that could implement the desired design changes. Then, using concept selection techniques, decide on a design that will satisfy customer and engineering requirements from the Problem Definition phase. Also, to meet with subject matter experts to ensure that the chosen design concept is feasible. Basic prototypes are designed in this phase.
To review the customer and engineering requirements see Problem Definition.
Action Items From Problem Definition
All action items from the problem definition phase were completed during this time phase.
- Customer interview two conducted with Dr.Bailey and Dr.Stiehl to brainstorm possible lab procedures. COMPLETED
- Engineering requirements and House of Quality reviewed to ensure all costumer requirements are satisfied by testable conditions. COMPLETED
- Interview scheduled with previous member of the P16081 COMPLETED
- Dr.Day contacted to request a tour of his laboratory to research his circulatory and heart pump systems for bench-marking. COMPLETED
Draft Laboratory ProcedureExercise one: The observation of pressure-volume (PV) loops.
In this portion of the lab, the students would be required to operate the loop under an array of pre-determined conditions while observing a series of real time plots presented in the labview program.
Goal: The goal of this lab would be to predict and observe how the changes between the health of the “patient” alters the shape and position of a P-v loop. Loops could be calculated once per heart cycle and could be displayed in a first in first out method with three graphs overlapping.
Exercise Two: The creation of a Starling curve In this portion of the lab students would be required to collect discrete data by adjusting the model to achieve certain conditions of cardiac output, pressure, and stroke volume. During the lab, data points would be collected using the LabVIEW program and exported once all points were achieved. In the analysis portion of the lab, students would use software such as MATLAB or Excel to plot the data and make observations based on the output.
Goal: The goal of this portion of the lab would be for students to take control of individual factors of the human body. This will help to enforce an understanding of the subsystems of the body by observing how individual components such as flow, resistance, and compliance affect the graphs.A draft laboratory procedure was created and the active document is seen below:
Benchmarking will be used to assist with concept development. The overall system, resistance model, and tubing were benchmarked.
The overall system needed adaptation because it is not currently functional. Other similar models on campus were compared against the current model. Compared Models Overall System
The current resistance model is a gear assembly that has multiple moving parts and would unexpectedly disassemble. Mechanical and electrical pinch valves were researched to find a user friendly way to model the pressure drop due to the change in tube area. Compared Models Resistance
The current thick walled tubing does not accurately demonstrate compliance. Compliance is a term used to describe how the blood vessel expands as blood passes through the system. Using thinner wall tubing or alternative material would change the demonstration of compliance.
 p.3-5; Encyclopedic Dictionary of Polymers, Volume 1 By Jan W. Gooch
 Tubing Glossary; http://www.freelin-wade.com/glossary.php
Subject Matter Expert Information and Feasibility
Subject Matter Expert Information and Conclusions
LabVIEW: Professor John Wellin
Overall System: Dr. Steven Day
Silicone Ventricle Production: Mike Buffalin
- The previously generated LabVIEW program was not a complex program. However, given the current condition of the code a redesign is recommended.
- Our current model does not accurately simulate pre-load adjustments and the ventricle system would be more physiologically accurate.
- It is possible and simple to create the desired silicone ventricle
The idea for the ventricle design started in an SME interview with Dr. Day. The current model was based on a piston and raised some concerns to the biological accuracy of such a model. He presented his heart model loops and the ventricle design appeared to be the most compatible with the current system. A graduate student was interviewed for a shape and volume of the ventricle. The shape is unique, but only relative shape accuracy is desired. Using CAD a draft of a ventricle design was created and brought to Mike Buffalin at the construct. Mike had made the previous molds for Dr. Day’s and Dr. Mix’s loops. The mold is currently in the process of being 3D printed. The ventricle will be made by pouring silicone into the mold.
The ventricle has a volume of 120ml. The silicone mix is estimated to cost $30 per pint. Which would provide enough for 3 molds with almost enough for an additional mold remaining. The calculations show the ventricle is approximately $10 per mold, where the diaphragm was $50 per unit. One extra plate is needed to hold the ventricle in place, at a cost of $10. There are very few risks and enough reward to justify a prototype for further testing
Piston Heart Pump Feasibility
The heart pump currently is a piston design with a spring loaded plate in a diaphragm. Updating this system would only require minimal changes in design with a few updates to current parts. The resulting cost is therefore minimal and the system would perform more reliably.
The current design was not physiologically accurate. It was close, but there were some issues with the collected data. After speaking with Dr.Day, it was concluded that this type of system does not accurately model the pre-load phase. This would slightly through off the data collected. The cause of this is from the spring system which is creating a negative pressure, similar to a suction, thus pulling the fluid into the chamber.
The current solenoid has a maximum switching frequency and the customer requirements state that the pump must operate in a range with a maximum of 2.5Hz. The solenoid was tested by firing relay with an Arduino controller. The speed was slowly increased until the solenoid was unable to return to the closed position. The upper limit of the solenoid was calculated by counting the number of fires in a ten second interval. The upper frequency of the solenoid was approximately 2.8Hz. The solenoid fails at approximately 3Hz. The conclusion is that the current solenoid is feasible to use to operate the pump from efficiently for the desired range.
Benchmarking and research concludes the material choice for system connectivity, Latex tubing demonstrates the best choice of material due to its capability of allowing for pulsatile flows to model physiological vessels. Latex is also economically friendly costing approximately $16.60 per 50ft at 0.4 inch diameter. In order to better model a physiological system, the tubing going back into the heart pump must be much more compliant opposed to the tubing leaving the heart pump. The tubing leaving the heart pump must be able to withstand much higher pulsatile pressures where the latex material tubing wall would need to be thicker. Latex is also non toxic and is environmental friendly.
An important factor in the successful implementation of the design is being able to control all of the sensors and devices through the provided NI DAQ’s. Furthermore, we are interested in achieving a more cost efficient and compact design. Therefore, we are also interested in exploring the feasibility of operating all of our sensors and devices on a single device.
To begin, a list was gathered of all of the devices in the proposed design, and the required data connections need for each one. Combined, the design includes a regulator, a relay, two pressure sensors and two flow sensors. The necessary channels needed tow run these include 2 analog-out ports, and 4 analog-in ports. A benchmarking was performed using the NI myDAQ and the NI USB-6211 device. Through this, it was determined that the number of analog in/out channels provided in the USB-6211 would make it the only choice for running the system on a single DAQ.
This conclusion means that our final system will not only be more compact and economical but will require less effort when designing a LabVIEW control system.
Model 1: Ventricle Heart Pump and Mechanical Pinch Valve Resistance with Pressure Reading Location Changed
Model 2: Multi-Spring Heart Pump and Mechanical Pinch Valve Resistance with Pressure Reading Location Changed
Model 1: VentricleThe Ventricle Heart Pump uses two air lines. One line goes to the lab air to Pressurize the heart pump and be the ejecting force for the heart. The other line is attached to the lab vacuum to assist with the pump inflating. This model is expected to demonstrate the physiological condition of pre-load more accurately than the pump models 2 and 3. Pre-load is the volume of blood in the ventricle at the end of diastole. Diastole is the phase of the cardiac cycle when the heart relaxes between contractions. This model also uses one way valves tat are not directly attached to the pump so that they do not cause issues with the deflation of the ventricle. This also allows for the a pressure reading to be taken and sent to the LabVIEW program at a location outside of the pump, between the valves and the pump. This reading is the heart pressure and is one of the required outputs for LabVIEW. The volume of the fluid in the ventricle will be calculated with a known integration method instead of the previous system with a distance sensor.
The resistance chosen for model one is the mechanical pinch valve. Through benchmarking it was determined that the electrical pinch valve was not within the budget and the current model was not user-friendly.
The last change is the pressure reading location on the circulatory system. The locations originally were in a T-shape, assumed reason was for higher accuracy. The tubing diameter would also be changed because the current method is 1/4in tubing for the pressure sensor attached to 1/2in tubing for the system. That large difference in diameter is not desired for accurate results.
Model 2: Improved Heart Pump with Multiple SpringsThis model also uses the mechanical pinch valve and the different pressure sensor locations for the same reason as model 1. These changes were determined necessary by the team to implement the customer requirement of the system being user friendly. For the adaption to the accuracy of the physiological condition most of the problems were identified with the heart pump. This is expected because it is the driver.
The proposed adapted heart pump has multiple springs with a softer spring constant than the original springs. The reasoning for this is due to the discoveries during preliminary observation of the current heart pump. The current heart pump uses one spring which has a stiff spring constant. The large spring constant causes the diaphragm to jolt back into place during each cycle, as well as, created a suction that brings new water into the pump. This does not model pre-load as previously discussed. The diaphragm also tilts in the pump. The multiple springs would limit the amount of possible tilt in the diaphragm. Benefits to the piston design is that the infrared sensor can be easily used to calculate volume instead of using integrals. This model also only uses one line to the laboratory air instead of two, one to pressure and one to vacuum.
Model 3: Updated Current Model
The updated current model has all of the current components however either the fixture, material, or connections have been updated and replaced. This model would not use 3D printed parts for the heart pump because those are porous and allow for leakage to atmosphere. It would also adapt the gear system so that it could not come apart and would have to be slid on and off the tubing if disassembled. This would limit the ability to replace the gear but would increase the ease of use for the students during laboratory time.
Model 1:The Ventricle Heart Pump System was selected because it had more benefits than the Multiple Spring Heart Pump System. One benefit is that the ventricle is a more accurate example of the physiological conditions of the heart. The ventricle heart pump demonstrates pre-lead more accurately because it is made from a silicone shell that can inflate and deflate instead of the piston model that is a ridged system. The rigid system uses suction to pull water in instead of allowing the fluid to pour into the heart pump. Also, the ventricle system has less moving parts than both of the piston models. This decreases the amount of friction in the system and increases the expected part life.
The location of pressure measurements was changed for both generated concepts because the current design was susceptible to user error. The two-line pressure readings from the current design was controlled using a valve that could be easily turned to the wrong location. It is an easy mistake to make and would generate incorrect data for the students. Eliminating the two-line pressure reading and converting it to one line is done in anticipation of human error. This change creates a more user-friendly system.
The mechanical pinch valve is also used in both generated concepts. The current resistance model uses 3D printed gears and 2 square plates. The plates were used to make the measurement of the pinched tube height simple. However, all of the moving parts over-complicated the resistor and allow for more opportunities for the part to fail. Using a simple pinch valve is simpler for the students to assemble and still provides the ease of measurement.
These three changes are the main adaptations to the system. Other aspects require change such as;
- Sealing of the tanks and tubing connections
- Replacing broken tubing and connections
- Creating a user-friendly LabVIEW code
The only additional risk after the systems design phase is if the ventricle design does not interface with the current system as desired. In anticipation of failure, the current heart pump is also being adapted to be functional. The housing of the adapted current heart pump will be used for the ventricle heart pump as well, thus removing any concerns of time or money waste.
The ventricle heart pump design is for a silicone shell that would inflate and deflate to simulate the heart. This model is expected to be able to demonstrate pre-load in the heart more accurately because the water will enter in more of a pouring action instead of a suction filling. The ventricle heart pump will be created using a 3D-printed mold which will create a flange around the outer edge. This flange will be used to create a seal between the ventricle and heart housing. The ventricle will use the same housing as the improved piston heart pump.
The current design used a 3D printed cap for the Heart pump housing. This cap was located at the top of the pump and connected the pump to the air supply and the emergency release valve. The 3D printed material was porous and leaked air to the atmosphere causing the pressure readings of the heat to be inaccurate. This was discovered when soapy water was put onto the cap to clean it. The soap was still on the cap when the pump ran. Bubbles began to form on the cap surface. The solution was to make the cap out of aluminum so that it would no longer be porous and supply and an accurate pressure reading.
Design Review Materials
- Create and assemble the ventricle prototype
- Test the pump against the Engineering Requirements
- Decide if the ventricle pump should continue as an option.
- Identify all material in the circulatory system that needs to be replaced.
- Test various tubing for compliance and identify the desired tubing
- Generate preliminary LabVIEW program