Repositorio Institucional de Documentos

Optimization of the in-vitro model equipment for future heart-valve studies BACHELOR’S PROJECT - APPENDIX Author: Ana Ochotorena Portales Student number: 201301955 Spring semester 2014 Mechanical Engineering Department Ana Ochotorena Portales 1 Ana Ochotorena Portales 2 Index 1Structure of the report and briefly explanation of each section …………………………….… 3 2Requirement for the new in vitro model ………………………………………………………………….. 8 3Images from Solid-Works (Compliance chamber and aortic root) ……………………………. 13 4Real images (Compliance chamber and aortic root) ………………………………………………… 24 5Drawings (Compliance chamber and aortic root) …………………………………………………….. 27 6Images and drawings of the mould for the silicone rings …………………………………………. 62 7Equipment ………………………………………………………………………………………………………………. 63 8Fluid mechanics theory to achieve Bernoulli equation …………………………………………….. 65 9Compliance chamber pressure calculation ………………………………………………………………. 67 10Compliance chamber’s wall force calculations ………………………………………………………... 68 11Compliance chamber screws force and stress theoretical calculations ………………….... 71 12Compliance chamber screw’s stress simulations ……………………………………………………... 79 13Pressure drops in the entire model ……………………………………………………………………….… 89 14Fluid passing through the new model ……………………………………………………………………... 99 15Silicone ………………………………………………………………………………………………………………..… 100 16List of materials ….……………………………………………………………………………………………….... 102 17Transonic flow sensor specifications ……………………………………………………………………… 105 18Glue options for the wall’s connection …………………………………………………………….……. 107 19Tests …………………………………………………………………………………………………………………………….. 19.1. Signal processing ……………………………………………………………………………………….…. 109 19.2. Comparison of results – Analysis methods ………………………………………………….… 110 19.3. Tests results ……………………………………………………………………………………………….…. 115 20Project management planning’s tools. FMEA and Fishbone analysis ……………………………. 20.1. FMEA ………………………………………………………………………………………………………….… 128 20.2. Ishikawa diagram or Fishbone analysis …………………………………………………………. 134 21List of figures, tables and equations ……………………………….…………………………………….. 135 Ana Ochotorena Portales 3 1Structure of the report and briefly explanation of each section 1Introduction 1.1. General description A general view of the model is presented in this section 1.2. Project specification The abstract of the project is written in this section, not only the part done MIBAC4 group but also the part done by M7BACH. 2Background In this section are presented different some theoretical concepts around the heart and the prosthesis valves, due to the purpose of the designed model is to analyse them 2.1. Heart 2.2. Blood 2.3. Cardiac cycle 2.4. Diseases 2.5. Overall about the kind of valves 3Previous in-vitro model – Description 3.1. List of the parts of the current model A list of the different parts in the previous model is presented in this section, and its explanation in the next one. 3.2. Description of each parts of the current model 3.3. Aramis Some concepts around Aramis which may affect the design are explained in this section. 3.4. Discussion of the current in-vitro model Here are described the problems seen by the previous groups working on the model. There is also a list of requirements added in the appendix. 4Updates in the in-vitro model In this section is presented a view of the changes done in each part of the model, not only the ones done by myself but also, briefly, the ones made by the other group 4.1. Compliance chamber 4.1.1. Requirements In this section are explained briefly the requirements to be fulfilled. Ana Ochotorena Portales 4 4.1.2. Versions of the new compliance chamber Here are described the final chambers designed – square and trapezoidal. There are also going to be described the previous options, explaining briefly how the process was to achieve the final result. Also are going to be shown the Solid-works graphs and some photos (In the appendix are the drawings attached) 4.1.3. Pressure calculations inside the chamber Explained the results achieved for pressure calculations inside the chamber 4.1.4. Calculation of forces and stress in the screws A force analysis and stress analysis is done theoretically in order to compare the results with the one done by simulations. By this calculation is possible to know if the model is an improvement or not of the previous one. This calculation has been done for the square and trapezoidal chambers 4.1.5. Forces inside the compliance chamber Calculations to achieve the forces that the walls and therefore the connections inside the compliance chamber 4.2. Aortic root 4.2.1. Requirements In this section are explained briefly the requirements to be fulfilled. 4.2.2 Versions of the fittings Here is described the final fittings for the aortic roots designed. There are also going to be described the previous options, explaining briefly how the process was to achieve the final result. Also are going to be shown the Solid-works graphs and some photos (In the appendix are the drawings attached) 4.2.3. Flow meter Here is going to be described the flow meter chosen and the requirements to be fulfilled for its usage. 4.3. Ventricle chamber and atrium chamber Here is described the final ventricle and atrium chamber designed by the other group. 4.3.1. Ventricle chamber 4.3.2. Ventricle module 4.3.3. Atrium chamber 4.4. Analysis of the new model Some analysis done to the new model designed 4.4.1. Pressure drops It has been calculated different kinds of pressure drops in the model, due to different agents. Ana Ochotorena Portales 5 4.4.2. Fluid passing through the entire model It has been calculated the flow passing through the model, this value should be reduced in comparison with the old model 5Connection test Due to problems seen in the bolt connections, some test were done in order to find an improvement of them. Reducing the time to break or avoiding it, by the usage of quickserts or helicoils 5.1. Expected results – Theoretically The expected results are described in this section. 5.2. Test procedure The procedure explanation and things to take into account are explained in this section. 5.3. Diverse connections – Results of the test and explanation In this part all the test done are explained and the results achieved. 5.3.1. Connection 1 – Test 1 - Characteristic of this kind of connection - Present the test notes and the results - Discussion 5.3.2. Connection 2 – Test 2 - Characteristic of this kind of connection - Present the test notes and the results - Discussion 5.3.3. Connection 3 – Test 3 - Characteristic of this kind of connection - Present the test notes and the results - Discussion 5.4. Other relevant tests Some other tests done are explained in this section, such as the water test and its non-success. 5.5. Discussion of the results In this part is discussed which option is best one and also the cracking problem seen in the previous compliance chamber. Ana Ochotorena Portales 6 6Mould for the silicone rings A silicone ring was desired in order to connect some parts in the model minimizing the leakage between them. In this section it is going to be explained the procedure to achieve it. 6.1. Theoretical part In this section is explained the chemistry and formulas given to explain the reason of the initial problems seen on the silicone bag construction, which are relevant also for the ring design, due to same material is used. 6.2. Mould design – SolidWorks In this part is the description of the mould and Solid-works graphs (In the appendix are the drawings attached) 6.3. Silicone ring obtained Explanation of the silicone ring obtained and how it works in the model 7Materials As some problems are seen in the previous material, was considered a change of it. 7.1. General description In this section are explained the different materials evaluated, the previous and the new ones. PMMA, POM and PC are described. 7.2. Analysis of characteristics (in order to explain why the material was changed) In this section many properties are analysed in order to choose the best material. The properties analysed are the ones connected with the fracture of the material, so as to the bolts are the critical point. Mechanical properties, optical properties, chemical properties, physical properties, thermal properties and price have been analysed. 7.3. Glue As the walls are connected each other by glue, in this section is explained the chosen one and the other options evaluated. 8Evaluation of the new model 8.1. Test procedure In this part is explained how the analysis sequence was and the ideas to do the different test in the new in-vitro model 8.2. Standards that must be followed In this section are explained briefly the standards that must be followed Ana Ochotorena Portales 7 8.3. Signal processing How the signal obtained while recording have been treated and analysed 8.4. Comparison of results – Analysis In this section is explained the way to get and analyse the data 8.5. Different tests done in the new model In this section are defined and explained the different tests achieved. The test results are added in the appendix 8.5.1. Square compliance chamber and ventricle chamber without the bag 8.5.2. Square compliance chamber and ventricle chamber with the bag 8.5.3. Trapezoidal compliance chamber and ventricle chamber without the bag 8.5.4. Trapezoidal compliance chamber and ventricle chamber with the bag 8.5.5. Discussion of the waveforms 8.5.6. Comparison between the new model without the bag and the old model 8.5.7. Comparison between the new model with the bag and the old model 8.5.8. Addition of compliance in the ventricle chamber 8.5.9. Additional tests 8.6. Interpretation of the results 9Conclusion 10List of figures, tables and graphs Ana Ochotorena Portales 8 2Requirements for the new in-vitro model (cave) 1 Some requirements were provided by the people who have been working with this model before. This document was provided at the beginning of the semester in Danish, and here is attached the translation into English. General design requirements • The components must as far as possible be designed in order longevity. That is, the following parameters should be considered when working with: - Metal Parts - corrosion - Plastic items - cracks and internal stresses • For the items should be used acrylic sheets with a thickness of 20 [mm]. • In collections density should be carefully assessed. Good collection (see Annex 2): - Small recess (immersion) of 1-2 [mm] for controlling the second party - O-ring or gasket at all joints - All surfaces must be close to each other must be ground level • Plastic Thread is very fragile and damaged over time. Possibly alternative assemble principles can be incorporated Atrium chamber Atrium chamber acts as a large open reservoir before the ventricle chamber. Filling the chamber during systole, the mitral valve which is opened during diastole and ventricle chamber filled. Design Requirements a. Instrumentation: - No b. Construction: - Atrium Comrade shall work with the mechanical mitral valve from the laboratory. - Mitral valve must be positioned vertically i.e. with the entrance on the side of ventricle chamber. - The volume of the chamber must be at least 6 [L], to ensure a more stable flow. - Atrium chamber must be open to the atmosphere. - Atrium chamber coupled with compliance buddy via 14mm [mm] silicone tube. - Atrium chamber must be designed so that it is able to provide a water column in ventricle chamber (at the aortic valve) during diastole between 10 and 20 [mmHg]. - No requirements for transparency. 1 Provided by the Cave team Ana Ochotorena Portales 15 Figure 4 - Top view of the square compliance chamber (A.O.P.) Before achieving the square chamber final design, some changed were done. The first design was bigger than the final and the connection between it and the aortic root was by using an Oring. Then this connection was changed for a silicone ring built by using a mould. Afterwards, another big change was made, the dimensions of the chamber. Initially was followed one of the requirements which was having 4[L] above the aortic hole, but after talking with the supervisors there was no need for it. The last change made on the chamber, was cutting one of the corners, it was needed to do it due to space problems with the atrium chambers in the assembly. Following is possible to see some pictures of the previous chambers. Figure 5 - Initial square compliance chamber (A.O.P.) Ana Ochotorena Portales 16 Figure 6 - O-ring connection (left) and silicone ring connection (right) (A.O.P.) Figure 7 - Almost final square compliance chamber design – without corner cut (A.O.P.) Ana Ochotorena Portales 17 The second compliance chamber designed was a trapezoidal one. The entire model is also going to be used to analyse biological aortic root, which length is shorter than the acrylic ones used to test the aorta valves. So that, the chambers have to be closer, and by using a square one was no enough room for it. For that reason a trapezoidal one, with less front dimension was designed. Following is possible to see some images of the chamber. Figure 8 - Isometric view – trapezoidal compliance chamber (A.O.P.) Figure 9 - Left and right side of the trapezoidal compliance chamber (A.O.P.) Ana Ochotorena Portales 18 Figure 10 - Front view of the square compliance chamber (A.O.P.) Figure 11 - Top view of the square compliance chamber (A.O.P.) Ana Ochotorena Portales 19 Aortic Root For the aortic root, the main problem to focus in was the fittings in order to connect avoiding the leakage between it and the chambers. Some fittings have been developed to end in two designs. The first one is used for all the analysis which the model will be used for but for flow analysis. For this last analysis a flow meter is required. Due to the big diameter of the aortic root a big clamp flow meter is therefore required. For that reason a lot of room is needed for it. For this usage, new fittings have been designed. The first fitting was designed to be used with an O-ring connection between them and the chambers, the first one with a circular shape and the second with a rectangular shape. But due to leakage problems seen in the previous models, was decided to attach them by a silicone ring connection designed by using a mould. In all the models was needed to add some room to insert the Millar catheters. The first idea was to insert the catheters in the sides of the fittings but later on, the holes were moved to the top area of the fittings because it was easier and more convenient for the user. The original shaped was circular and then it was changed into rectangular, with Millar catheters in the sides of the fittings. Below are presented some images of the initial models. Figure 12 - Circular shaped fitting connection (A.O.P.) Figure 13 - Rectangular shaped fitting connection (A.O.P.) Ana Ochotorena Portales 20 Due to leakage problems, the connection between the fittings and the chamber was improved into a silicone ring connection. Below is possible to see the both different connections, the Oring connection in the left and the silicone ring connection in the right. Figure 14 - O-ring connection and silicone ring connection (A.O.P.) Afterwards the Millar catheters were moved into another possition. From the side to the top. Below are the images for both positions. Figure 15 - Millar catheters' positions (A.O.P.) The final assembly is possible to see below. Figure 16 - Arotic root assembly (A.O.P.) Ana Ochotorena Portales 21 So as to be able to insert the flow meter was needed to relocate some of the rods that connect both fittings, in order to increase the required room to do it. Therefore, a new design was desired because there was not enough space in the designed fittings. This new design not only changes the rods position, but also a connector in acrylic has to be added. This acrylic connector is used to connect the aortic acrylic root with the silicone tube which will be used to attach the flow meter. Pressure drops are expected due to the length, so that it is recommend the usage of this second model only for flow calculations, more about this topic in other sections. Before achieving this solution, other ones were tough. Figure 17 - Fittings – rods changed in position (A.O.P.) Below is shown the hand-drawing and the solid works image for the connector design and the way to join it with the acrylic sinus of Valsalva. The connector is made in acrylic as well and has a silicone tube attached. Figure 18 - Connector sinus of Valsalva and silicone tube – Hand drawing (A.O.P.) Figure 19 - Connector sinus of Valsalva and silicone tube – Solid Works (A.O.P.) Ana Ochotorena Portales 22 Before this design was though another kind of connection which can be seen below. It was neglected for two reasons. The first one is because is easier and more secure to attach the silicone tube into the chosen connector, due to its conical shape. Another important reason is a construction one. Some similar pieces have been manufactured before; it will be easier and faster therefore for the workshop to produce it. In the graph below is presented a hand drawing of the previous thought connector. Figure 20 - Previous design for the connector - Hand drawing (A.O.P.) It was also needed to design a connection between the silicone tube and the fitting that is connected to the compliance chamber since the tube was not big enough to be attached directly to the fitting. The design is possible to be seen below. Finally, there was no need to manufacture it since was possible to achieve a 31 [mm] diameter tube. Note 31 [mm] is the dimension which fits perfectly in the fitting connection. Figure 21 - Connector silicone tube and fitting – Hand drawing (A.O.P.) Figure 22 - Connector silicone tube and fitting – Solid Works (A.O.P.) Ana Ochotorena Portales 23 The final assembly of the fittings, the sinus of Valsalva, the silicone tube and the O rings are possible to be seen below. Note the rods are not drawn in the graphs. Figure 23 - Aortic root assembly for flow calculations – front view (A.O.P.) Figure 24 - Aortic root assembly for flow calculations (A.O.P.) Ana Ochotorena Portales 24 4Real images (Compliance chamber and aortic root) Square compliance chamber Figure 25 – Square compliance chamber assembly (A.O.P.) Figure 26 - Parts of the square compliance chamber (A.O.P.) Nº Part Material Quantity 1 Top plate Steel 1 2 Top plate Polycarbonate 1 3 Bottom plate Steel 1 4 Bottom plate Polycarbonate 1 5 Side plate for Millar Cth. Polycarbonate 1 6 Side plate Polycarbonate 1 7 M8 Bars – 335 [mm] length Stainless stell 5 8 M8 Nuts - 5 9 Pressure gauge - 1 10 Compressed air inlet - 1 Table 1 - List of parts in the square compliance chamber assembly (A.O.P.) Ana Ochotorena Portales 31 Ana Ochotorena Portales 32 Ana Ochotorena Portales 33 Ana Ochotorena Portales 34 Ana Ochotorena Portales 35 Ana Ochotorena Portales 36 Ana Ochotorena Portales 37 Ana Ochotorena Portales 38 Ana Ochotorena Portales 39 Ana Ochotorena Portales 40 Ana Ochotorena Portales 47 Ana Ochotorena Portales 48 Ana Ochotorena Portales 49 Ana Ochotorena Portales 50 Ana Ochotorena Portales 51 Ana Ochotorena Portales 52 Ana Ochotorena Portales 53 Ana Ochotorena Portales 54 Ana Ochotorena Portales 55 Ana Ochotorena Portales 56 Ana Ochotorena Portales 63 6Images and drawings of the mould for the silicone rings In order to improve the connections between the aortic root and the chambers, minimizing the leakage or trying to avoid it totally, a silicone ring was decided to be used instead of the Oring used in previous models. To achieve this silicone ring, a mould in aluminium was designed. Below is possible to see a solid works image of the final silicone ring. Figure 31 - Silicone ring (A.O.P.) The mould was designed in aluminium because it was an easy material to achieve in the workshop and due to chemical properties explained in other sections. Below are the images for the mould assembly, the male part and the female part. There are also attached two bolts in order to be possible to have the male and female part centred. Figure 32 - Mould assembly for the silicone ring (A.O.P.) Figure 33 - Female part (right) and male part (left) (A.O.P.) Ana Ochotorena Portales 64 7Equipment Following it described the equipment used to do the connection tests analysis. - Torque wrench: it is a tool which is used to apply the precisely torque to a fastener, such a bolt or a nut. Figure 34 - Torque wrench (A.O.P.) - Bench vice: is a mechanical device used to secure an object to allow work to be performed on it. It has two parallel jaws, one fixed and the other mobile, threaded in and out by a screw and lever. Figure 35 - Bench Vice (A.O.P.) - Milling machine: it is a machine which works using rotary cutters to remove material from a work piece, advancing in a direction at an angle with the axis of the tool. - Manual thread milling machine: it is a tool which works using rotary cutters. By using this tool, it is possible to thread a hole manually. - Fastener elements: bolts, nuts, washers are elements used to do the analysis. - Specimens: acrylic specimens were built in the workshop in order to connect and do the experiments with them. Figure 36 - Specimens (A.O.P.) Ana Ochotorena Portales 65 Below is shown the final assembly to carry out with the connection tests. Figure 37 - General assembly (A.O.P.) Ana Ochotorena Portales 66 8Fluid mechanics theory to achieve Bernoulli equation 2 In order to calculate the pressure inside the compliance chamber, it is necessary to explain briefly the theory that supports the calculations. To calculate the pressure, it is going to be assumed fluid statics. Therefore, the fluid is going to be considered at rest owing to be evaluated. If a fluid is in rest, it does not experience any net force anywhere. If a small part of fluid, all forces, external and internal must be balanced. By force balance calculations in ‘x’,’ y’ and ‘z’ direction it is inferred that the fluid cannot depend on the ‘y’ direction and neither on the ‘x’ direction. However, it is dependent on the ‘z’ direction due to the gravity force. ( ) ( ) ( ) ( ) Equation 1 - Force balance in ‘y’ direction ( ) ( ) ( ) ( ) Equation 2 - Force balance in ‘x’ direction ( ) ( ) ( ) ( ) ( ) ( ) Equation 3 - Force balance in ‘z’ direction Consequently, it can be inferred hydrostatic equilibrium for a fluid at rest. Hence, fluid at rest experiences that the pressure is constant in any horizontal plane and a decreasing pressure with height at the ratio of g . 2 Basics of fluid mechanics, Notes from Uffe Vestergaard Poulsen (Assistant professor – Department of Engineering – Sustainable Energy Systems) Ana Ochotorena Portales 67 Fluids obey energy conservation; to explain the mechanical energy conservation in particles it is used the mechanical energy equation as the starting point and then do derivative time. Finally it is achieved that the mechanical energy equation derivate is equal to zero, which means that mechanical energy is preserved between two situations. Equation 4 - Mechanical energy equation Equation 5 - Derivate of the mechanical energy equation The same procedure is used to fluid particles, achieving Bernoulli equation. There is need to take into account the hydrostatic equilibrium, suppose steady flow and constant density. Therefore, the Bernoulli equation for inviscid, steady and incompressible flow is the one bellow. Equation 6 - Bernoulli equation Ana Ochotorena Portales 68 9Compliance chamber pressure calculation The maximum pressure in the chamber will be in the bottom plate surface, according to the properties explained in the theoretical part – section 8. In order to calculate pressure, the limit water height used (z) is the minimum possible one, just above the aortic root inlet. Figure 38 - Simplified compliance chamber (A.O.P.) The height value, z, for the new compliance chamber designed is 101.96 mm ≈ 102 mm Equation 7 - Bernoulli equation applied in this problem 3 [ ] [ ] [ ] [ ] [ ] [ ] [ ] Equation 8 - Solution to the Bernoulli equation applied in this problem (A.O.P.) So the pressure at the bottom plate is 121 [KPa]. This value is significantly close to the compressed air pressure, 120 [KPa] therefore its contribution can be ignored. The water contribution in the new model is consequently possible to be neglected. Same hypothesis has been used for the old model. For screws forces’ calculation a value of pressure inside the chamber of 120 [KPa] has therefore been used. 3 Basics of fluid mechanics, Notes from Uffe Vestergaard Poulsen (Assistant professor – Department of Engineering – Sustainable Energy Systems) Ana Ochotorena Portales 69 10Compliance chamber’s wall force calculations In this section is the explanation of force calculation in the compliance chamber. For this calculation is going to be supposed the air is since the beginning with a value of 1.2 [bar] constant, as there is no pressure gauge to ensure more accurate pressure values. It is going to be calculated the force due to the water and due to the air in the back plate, isolating the deposit in order to simplify the problem. Moreover, this chamber is the only desired to be calculated. However, in the reality there are more agents that must be taken into account, such as the pump forcing the water direct to the chamber according the value of the cardiac output and the peripheral resistance (clamp). In order to calculate the force due to the air, below are going to be explained the steps needed to be followed. Pressure can be calculated by using the following equation; [ ] [ ] [ ] Equation 9 - Pressure The pressure is supposed constant and the area is a known value, therefore the force value is known too. To calculate the force in the water section is needed to know the pressure due to the water column by Bernoulli equation as explained in previous sections – appendix section 8. [ ] [ ] [ ] [ ] Equation 10 – Bernoulli equation In the following table are shown the values of force owing to the water column and due to the air. It has been calculated since the chamber is completely filled of water without air until it is filled only with air. And below is the graph that represents these forces. It is seen that the force created by the air is much higher than the water column influence. Ana Ochotorena Portales 70 Table 4 - Forces calculation GrapH in the compliance chamber's back wall because of air and water (A.O.P.) Figure 39 - Graph of forces in the compliance chamber's back wall (A.O.P.) air [%] water [%] Area air [mm²] Area water [mm²] Height water [mm] Force air [N] Pressure water [Pa] Force water [N] Sum of forces [N] 0 100 0 34060 262 0 2570.22 43.77 43.77 10 90 3406 30654 235.8 408.72 2313.20 35.45 444.17 20 80 6812 27248 209.6 817.44 2056.18 28.01 845.45 30 70 10218 23842 183.4 1226.16 1799.15 21.45 1247.61 40 60 13624 20436 157.2 1634.88 1542.13 15.76 1650.64 50 50 17030 17030 131 2043.6 1285.11 10.94 2054.54 60 40 20436 13624 104.8 2452.32 1028.09 7.00 2459.32 70 30 23842 10218 78.6 2861.04 771.07 3.94 2864.98 80 20 27248 6812 52.4 3269.76 514.04 1.75 3271.51 90 10 30654 3406 26.2 3678.48 257.02 0.44 3678.92 100 0 34060 0 0 4087.2 0.00 0.00 4087.20 0 500 1000 1500 2000 2500 3000 3500 4000 4500 water [%] 100 90 80 70 60 50 40 30 20 10 air [%] 0 10 20 30 40 50 60 70 80 90 Forces in the compliance chamber's back wall Force air [N] Force water [N] Sum of forces [N] Ana Ochotorena Portales 71 When the sum of forces produces a shear stress higher than the one that the connection can support, then the connection breaks. The connections in the walls are suffering a shear stress which can be calculated by the following equation 4 5 ; the shear stress in glue connections depend on the material and its thickness, the glue and the glued area. √ √ Equation 11 - Volkersen’s Equation Equation 12 - Middle shear stress value Note that is the maximum stress the connection can support, G is the shear modulus, E is the Young Modulus, F is the force, A is the shear area (with b the width and the overlapping length), s is the plate thickness and d is the glue layer thickness. Figure 40 - Glue connection of two plates with dimensions 6 Figure 41 - Shear stress distribution in a glue connection of two plates 7 4 Notes from Thomas Greve (Aarhus University school of Engineering – Materials) 5 http://biblioises.com.ar/Contenido/500/550/Tecnologia%20Adhesivos.pdf (pages 83 – 86) 6 Notes from Thomas Greve (Aarhus University school of Engineering – Materials) 7 Notes from Thomas Greve (Aarhus University school of Engineering – Materials) Ana Ochotorena Portales 72 11Compliance chamber screws force and stress theoretical calculations In order to calculate the forces in the screws it is going to be used the following equation which relates force and pressure. [ ] [ ] [ ] Equation 13 - Relation between force and pressure As it was explained in the previous section, the pressure that is going to be used is 120 [kPa], supposed uniform in the whole chamber. The pressure is always normal to the surface in which is applied. The volume chamber that is going to be taken into account is the internal one, neglecting the thickness of the walls. The walls are going to be analysed alike very thin layers, i.e. simplifying them to a 2D problem since one of the dimensions (the thickness of the wall) is much smaller than the other two dimensions (height and width of the walls). Furthermore; it is not going to take into consideration the glue between the walls neither the force caused by the other kind of connections, such as the recess in between some walls (top plate and bottom plate with the side plates). Therefore, the only force analysis which is going to be analysed is the screw connection. The chamber assembly is symmetrical and that simplifies the problem. However, there are only screws in the top plate. The bottom plate as is in direct contact with the table thus there was no need for those screws. Focusing in the side walls and in the back and front plate, in order to make them symmetrical for the force calculation, the glue part in the bottom area is going to be approximated to the same number of screws as there is the upper part. Figure 42 - Pressure distribution in the compliance chamber (A.O.P.) Ana Ochotorena Portales 79 To calculate de stress in the screw, first is calculated the cross section area perpendicular to the force; [ ] [ ] To sum up, in the following graphs are going to be resumed the forces and the stresses in the screws. Forces: Figure 53 - Forces in the screws (A.O.P.) Stresses: Figure 54 - Stresses in the screws (A.O.P.) Ana Ochotorena Portales 80 12Compliance chamber screw’s stress simulations 8 All the theoretical calculations done in the previous section were also done with Solid-Works simulations, achieving enough closed values to be compared. It was simulated not only the old model but also the new one, showing the new one will present less cracking problems due to the change of material and the new configuration. Note that, even though the results are fairly closed to the theoretical calculations, many assumptions were done. In this section are going to be explained the simulations and the assumptions done. Also it is detailed a comparison between the theoretical and Solid Works calculations. First, it is explained the old model and the three walls analysed: top, side and front plate. 1Top plate: Two simulations were done, the first one with threaded holes and the second one without them. Once the first simulation was done, the threaded results achieved were not good enough and it was recommended to analyse it with clearance holes. Therefore, the first assumption is using clearance holes, that would not provide the exactly stress that the screws are suffering but it would be able to compare the old model connection with the new model ones. This assumption will be used for all the walls. For simulate the behaviour of the top plate, there was supposed roller fixture in the sides of the plate and fixed geometry in the 16 holes. The pressure is distributed along the whole plate. Below are seen the results of the simulations with threaded holes: Figure 55 - Pressure distribution and fixtures (A.O.P.) 8 More accurate simulations were done by M7BACH, they can be found in “Otimering af In Vitro model til fremtidige hjerteklap studier”, 2014 report – section X Ana Ochotorena Portales 81 Figure 56 – Stress/Top plate old model (A.O.P.) Figure 57 - Displacements/Top plate old model (A.O.P.) Figure 58 - Strain/Top plate old model (A.O.P.) Ana Ochotorena Portales 82 Below are seen the results of the simulations with clearance holes: Figure 59 - Pressure distribution and fixtures/Top plate old model (A.O.P.) Figure 60 - Stress/Top plate old model (A.O.P.) Figure 61 - Displacements/Top plate old model (A.O.P.) Figure 62 - Strain/Top plate old model (A.O.P.) Ana Ochotorena Portales 83 Below is a table that sums up all the results and compares them with the theoretical calculations. Maximum Stress [MPa] Maximum Displacement [mm] Maximum strain Threaded hole 18.9110 0.05008 0.006173 Clearance hole 6.7945 0.05008 0.001381 Theoretical calculations 7.7922 - - Table 5 - Simulation's results/Top plate old model (A.O.P.) Simulating the threaded holes does not drive into good results, which was also a recommendation before. The displacement achieved is very small and it may not produce any problem in the model. Also the maximum stress does not reach the yield point of the material which is 45 [MPa] due to it is PMMA. But, the stress caused by the pressure is therefore an agent to increase the probabilities of breakage in the screws, adding other agents such us moisture absorption and the stress problems due to a bolt connection. 2Side plate: As the previous results with the threaded hole were not successful, this simulation was avoided in the side plate. The assumption of using clearance holes is therefore applied. For simulate the behaviour of the side plate, there was supposed fixed geometry in the three holes, in the sides and also in the bottom of the plate. The pressure is distributed along the whole plate. Figure 63 - Pressure distribution and fixtures/Side plate old model (A.O.P.) Ana Ochotorena Portales 84 Figure 64 - Stress/Side plate old model (A.O.P.) Figure 65 - Displacements/Side plate old model (A.O.P.) Figure 66 - Strain/Side plate old model (A.O.P.) Ana Ochotorena Portales 85 Below is a table that summarises all the results and compares them with the theoretical calculations. Maximum Stress [MPa] Maximum Displacement [mm] Maximum strain Clearance hole 3.9864 0.02684 0.001109 Theoretical calculations 3.4312 - Table 6 - Simulation's results/Side plate old model (A.O.P.) The displacement achieved is again negligible and it might not produce any deformation in the model. On the other hand, maximum stress does not reach the yield point of the PMMA material, which is 45 [MPa]. This value is small, but may be one of the reasons for the cracking problem. 3Front plate: The assumption of using clearance holes is used. For simulate the behaviour of the front plate, there was supposed fixed geometry in the five holes, in the sides and also in the bottom of the plate. The pressure is distributed along the whole plate. Figure 67 - Pressure distribution and fixtures/Top plate old model (A.O.P.) Ana Ochotorena Portales 86 Figure 68 - Stress/Front plate old model (A.O.P.) Figure 69 - Displacements/Front plate old model (A.O.P.) Figure 70 - Strain/Front plate old model (A.O.P.) Ana Ochotorena Portales 87 Below is a table that summarises all the results and compares them with the theoretical calculations. Maximum Stress [MPa] Maximum Displacement [mm] Maximum strain Clearance hole 10.481 0.1192 0.003061 Theoretical calculations 3.888 - Table 7 Simulation's results/Front plate old model (A.O.P.) Equally to the side wall and top wall, the displacement value achieved negligible. There is a notable difference between the stress value reached by theoretical calculations and SolidWorks simulation. Even though the maximum stress is bigger than the previous ones, it does not reach the yield point of the PMMA material, 45 [MPa]. This value is still small compared to the tensile strength of the material, when the material starts cracking, this value is 76 [MPa] but may be one of the reasons for the cracking problem. 9 More accurate simulations for this wall using advanced fixtures in Solid Works have been done, taking into account the glue connection between the walls. However, threaded holes were substituted by clearance holes for the simulation since not possible realistic results can be achieved in Solid Works. Even though it is known that in each thread of a bolted joint there is a stress concentration. This study has been done for the back plate of the old model. Achieving the maximum stress value in the middle hole with a value of 17.2 [MPa] and 0.158 [mm] with 1.6 [bar] as pressure value. Figure 71 - Stress distribution in the holes of the old model’s back plate 10 9 Done by Tobias (M7BACH) – appendix section number X of M7BACH group’s report 10 Photo from M7BACH group’s report Ana Ochotorena Portales 88 Besides, the calculations of the new model to compare it with the old one are expressed below. Note that there was only needed to simulate the top plate since the connections are outside the pressure area, so it only would affect the top plate screw’s connections. 4Top plate: The assumption of using clearance holes is again used. For simulate the behaviour of the front plate, there was supposed fixed geometry in the five holes of the steel plate, not the PC ones since the holes are some millimetres bigger than the threaded bars. Furthermore, the sides of the plate are also supposed as fixed geometry due to the new configuration designed for it. The pressure is distributed along the whole plate. Figure 72 - Pressure distribution and fixtures/Top plate new model (A.O.P.) Figure 73 - Stress/Top plate new model (A.O.P.) Ana Ochotorena Portales 95 Pressure drop [mbar] Cardiac Output 5 [L/min] 11 [L/min] Aortic Root Short 0.01 0.05 Long 0.03 0.13 Table 11 - Summary of the pressure drops in the aortic root due to length (A.O.P.) Can be concluded that there is an increasing pressure drop with length but its value is not larger enough to change significantly the results. However, as a recommendation to optimize the model is the use of the shorter aortic root for pressure calculations and to use the longer aortic root for flow measurements when the flow meter is needed to be inserted in the system. The second calculation is the analysis of pressure drops due to cross section changes in all the connections of the model for a CO of 5 [L/min] and 11 [L/min]. This calculation has been done by Bernoulli equation and neglecting the fiction factor because is not relevant. The height factor will be also neglected because it is the same. So, the equation use is the one seen below. Equation 18 - Bernoulli equation (velocity) This calculation has been done for all the connections seen in the model, achieving the following pressure drops for a CO of 5 [L/min]: - Aorta to compliance chamber: achieving a pressure drop of about -10.47 [Pa] - Compliance chamber to the tube that connects it to the atrium chamber: achieving a pressure drop of about 271.08 [Pa] - Tube connected from the compliance chamber to the atrium: achieving a pressure drop of -271.09 [Pa] - Atrium chamber to the mitral valve: achieving a pressure drop of 14.25 [Pa] - Mitral valve to ventricle chamber: achieving a pressure drop of -14.07 [Pa] - Ventricle chamber to aorta: achieving a pressure drop of 10.31 [Pa] - Aorta to aorta: achieving a pressure drop of 0 [Pa] CO = 5 [L/min] Connections Aorta Compliance Tube Atrium Mitral Ventricle Aorta S [mm²] r² = (27/2)² = 572.55 130x111.32 = 14471.6 r² = (12/2)² = 36 220x271.81 = 59798.2 r² = (25/2)² = 490.87 r² = (74/2)² = 4300.84 r² = (27/2)² = 572.55 V [m/s] 0.145 5.75e-3 0.737 1.393e-3 0.169 0.019 0.145 ∆P [Pa] -10.47 -271.09 -14.07 ∆P [Pa] 0 271.08 14.25 10.31 Table 12 - Pressure drops due to cross section changes for CO = 5L/min (A.O.P.) Ana Ochotorena Portales 96 For a CO of 11 [L/min] was achieved the following pressure drops: - Aorta to compliance chamber: achieving a pressure drop of about -51.03 [Pa] - Compliance chamber to the tube that connects it to the atrium chamber: achieving a pressure drop of about 1309.76 [Pa] - Tube connected from the compliance chamber to the atrium: achieving a pressure drop of -1309.84 [Pa] - Atrium chamber to the mitral valve: achieving a pressure drop of 69.43 [Pa] - Mitral valve to ventricle chamber: achieving a pressure drop of -68.56 [Pa] - Ventricle chamber to aorta: achieving a pressure drop of 50.23 [Pa] - Aorta to aorta: achieving a pressure drop of 0 [Pa] CO = 11 [L/min] Connections Aorta Compliance Tube Atrium Mitral Ventricle Aorta S [mm²] r² = (27/2)² = 572.55 130x111.32 = 14471.6 r² = (12/2)² = 36 220x271.81 = 59798.2 r² = (25/2)² = 490.87 r² = (74/2)² = 4300.84 r² = (27/2)² = 572.55 V [m/s] 0.320 0.0126 1.62 3.06e-3 0.373 0.042 0.320 ∆P [Pa] -51.03 -1309.84 -68.56 ∆P [Pa] 0 1309.76 69.43 50.23 Table 13 - Pressure drops due to cross section changes for CO = 11L/min (A.O.P.) Note: S = surface, v = velocity, ∆P = pressure drop The bigger pressure drops are seen in the tube that connects the compliance chamber and the compliance chamber. Then the ones that connects the chambers and the valves (aorta and mitral). The third calculation of pressure drop is done in the bent of the tube that connects the compliance chamber and the atrium chamber. This one has been done by using the following equation and graph. Equation 19 - Bend loose equation 18 Note: = Moody’s fiction factor, = density, = mean flow velocity, = bend radius, D = diameter of the tube, = bend angle and = bend loss coefficient. 18 http://thermopedia.com/content/577/?tid=104&sn=1422 Ana Ochotorena Portales 97 Figure 82 - Bend loss coefficient calculation graph 19 The calculations and the explanation for 5 [L/min] are seen below: The friction factor has been calculated by Moody’s diagram. First is needed to be known the Reynolds number for the tube and the velocity. With this number and the factor where is the roughness and d is the diameter is possible to enter in the diagram ti obtain the friction factor. [ ] Equation 20 - Velocity calculation with a CO = 5 [L/min] in the tube (A.O.P.) Equation 21 - Reynolds number for the tube with a CO = 5 [L/min] (A.O.P.) Equation 22 - Relation between the roughness and the diamter of the pipe with a CO = 5[L/min] (A.O.P.) Equation 23 - Friction factor for the tube with a CO = 5 [L/min] (A.O.P.) 19 http://thermopedia.com/content/577/?tid=104&sn=1422 Ana Ochotorena Portales 98 In order to calculate the bend loss coefficient is needed to use the graph seen in the figure 68, achiving a value of: Equation 24 - Bend loss coefficient with a CO = 5 [L/min] (A.O.P.) Putting all these values together is achieved the pressure drop: [ ] [ ] Equation 25 - Pressure drop calculation with a CO = 5 [L/min] due to bend (A.O.P.) This value was also done by the same web-page used before, ahieving a value of 4.97 [mbar]. The same calculations are followed to achieve the pressure drop due to bend in the tube with a CO of 11 [L/min]: [ ] [ ] Equation 26 - Pressure drop calculation with a CO = 11 [L/min] due to bend (A.O.P.) This value was also calculated by the same web-page used before, achieving a value of 19.43 [mbar]. Finally, the fourth calculation is to analyse the pressure drop in the ventricle chamber due to bends. The flow does not go straight inside the ventricle, as it is seen in the photo below; therefore two bents have been assumed to simplify the calculations. These calculations have been done in the same way as the third one by using the Bend loose equation and then the results have been compared with the same web-page. Figure 83 - Ventricle chamber with the two valves' connection (A.O.P.) Ana Ochotorena Portales 99 The first bent hypothesis is between the mitral valve and the aortic root. The flow will go from the mitral to the aorta once is closed. The angle is supposed around 136.33º and the radius of the bent is 26.934 [mm]. Figure 84 - Ventricle chamber bent one (A.O.P.) The second bent hypothesis is between the pump valve and the aortic root. The flow will be drove by the pump into the aorta. The imaginary bent tube’s angle is supposed around 37.22º and the radius is 111.681 [mm]. Figure 85 - Ventricle chamber bent two (A.O.P.) In the table below are going to be summarized the results in the bents supposed in the ventricle achieved by the web page and hand calculations (same procedure as explained in the third one). Pressure drop Bent Mitral to Aorta Bent Pump to Aorta Web-page calculations Handcalculations Web-page calculations Handcalculations 5 [L/min] 0.12 [mbar] 0.084 [mbar] 0.03 [mbar] 0.017 [mbar] 11 [L/min] 0.45 [mbar] 0.36 [mbar] 0.13 [mbar] 0.072 [mbar] Table 14 - Pressure drops due to bend in the ventricle chamber (A.O.P.) Ana Ochotorena Portales 100 14Fluid passing through the new model In order to know if the new model achieved has a more laminar flow, one of the changes done is increasing the size of the atrium chamber. The way to calculate the time that the volume of water needs to complete the entire cycle is the calculation done below. [ ] [ ] [ ] [ ] [ ] [ ] [ ] [ ] Equation 27 - Fluid passing through the new model calculations 20 Comparing this value with the previous model, which has an exchange of water of 1.6[ ], it is possible to say that the volume of water going through the new model per minute has a lower value, meaning that the system goes slower than the previous one, so the flow may behave more laminar. 20 Comparison with the “The effect of sinus of Valsalva on TAVI valves” results Ana Ochotorena Portales 101 15Silicone 21 22 Some chemical reactions happened between the mould and the silicone, and that is why some troubles were seen along the design mould period. In this section it is going to be explained that chemical reaction mainly. The silicones provided by the supplier are RTV-ZA12 A+B. The final silicone used was a half – half in weight proportion mixture in order to improve its properties, such as hardness. Furthermore it was provided some SH 350 oil. In order to avoid some of the bubbles that appear during the mixture procedure was also though about using just one kind of silicone. The mould was done in several materials, in order to try which one was the better option. Firstly it was built using a 3D printer in ABS material. However, as some problems associated with this plastic material were seen, it was decided to make it in aluminium. To make the silicone elements was initially tried only by using the ABS mould and filled it with the silicone. Nevertheless, the solution was not accurate since the silicone element was sticky and soft, not achieving the desired properties. Later on, some of the oil provided and wood polish was used in order to cover the surface expecting better results. Finally the mould was decided to be made in aluminium, and it was covered with some solutions in order to avoid the chemical reaction that happens between it and the silicone, due to the OH groups formed. The reason for the inaccurate properties achieved in the silicone was because of the OH groups that the silicone and the aluminium have. So they react creating the HO₂ bubbles. The reaction of the silicone and the aluminium can be seen below. Figure 86 - Chemical reaction of the silicone (A.O.P.) 21 Finn Monrad Rasmussen (Associate professor – Engineering College of Aarhus – Materials) 22 Stevens, Malcolm P. Polymer chemistry: an introduction. 3rd ed. New York: Oxford University Press, 1999. Print. Ana Ochotorena Portales 102 The aluminium has covalent bonds and its formula is the following one: Al₂O₃ Al₂O₂H Thereby there are also OH groups that react with the OH groups of the silicone, creating those bubbles. The material will react adversely with the silicone. For that reason some additional materials, to cover the mould’s surface, are needed in order to avoid that reaction between hydroxides. There were some ideas suggested such us polish like nail polish, hairspray, a mixture made of Styrofoam and acetone and a mixture made of Styrofoam and toluene. The Styrene has no OH groups; hence it cannot react with the silicone so it does create a coat between the aluminium and the silicone. Figure 87 - Chemical reaction of the Styrene (A.O.P.) Ana Ochotorena Portales 103 16List of materials 23 Following are the data sheet for the materials used given by the supplier company, Vink. 23 Information taken from the supplier http://vink.dk/ Ana Ochotorena Portales 104 Ana Ochotorena Portales 111 19.2. Comparison of results - Analysis methods The goal and aim of the project is to reduce the amount of recorded noise. In order to determine whether improvements have been made, based on the amount of noise present, it will be necessary to somehow quantify the entity. A method could be simply to calculate the area under cardiac cycle curve, and compare the size of the areas. However, that could create too many uncertainties since the measured pressure sometimes exceeds or subceeds the targeted pressure of 120 [mmHg] when recording. This method is therefore rejected. The character of the fluctuation and the frequency size do however appear to be unaffected by small pressure differences. The solution may therefore be to isolate the fluctuations and calculate the area bounded by peaks and valleys that defines the noise. Figure 91 shows the noise in start systole and start diastole as being the area of interest. Figure 91 - Noise in the cardiac output to be eliminated Labview can be programmed to detect peaks and valleys. The amplitude line is the trend line between the amplitudes. To evaluate the amplitude line trend lines are created for the peaks and the valleys. The amplitude line is the difference between these two and defines the zero line see figure X. The area can now be calculated by integrating for the absolute values of the curve. Figure 92 - Isolated fluctuations in start systole Noise Ana Ochotorena Portales 112 Method for isolating the area of interest For each dataset the area can be calculated simply by evaluating the number of peaks that represent the area of interest. By integrating pressure with respect to time, the unit of the noise is [Pa*s] which is also the unit for absolute/dynamic viscosity. When deciding the number of peaks to be included for calculation of the area beneath the curve should include as many peaks as possible as long as the period remains constant, the rest of the curve can be considered to be random and should not be included. In figure 93, the circled area represents the range of the calculation. The calculated frequency is grossly misrepresented; therefore another method is used to calculate frequency. Figure 93 - Fluctuations during start diatole isolated with Labview Method for isolating frequency of interest The number of peaks included in the calculations will have to be evaluated manually for each curve, to make sure the calculations are comparable. The frequency is calculated by measuring the distance between the peaks and valleys. Peaks are represented with red points and valleys with green points. In figure 94 it is seen that the point do not mark the distance between what can be considered to be a period. By reducing the number of peaks, the program will choose only the extreme peaks and valleys and thereby obtaining a correct frequency value. Figure 94 - Noise curve 23 peaks (left) and 3 peaks (right) Ana Ochotorena Portales 113 The program does not discriminate between peaks in the constant systemic noise and random noise, therefore this also needs to be evaluated manually. In figure 95 the program has chosen accurate peaks for the systemic noise, but it has also chosen peaks in the area of random fluctuations. In this case it would be necessary to reduce the number of peaks from 5 to 3. Figure 95 - Fluctuations in start systole isolated in labview Spectral analysis 28 Fourier transform is a mathematical tool for transforming signals in the time domain to the spectral domain. The cardiac cycle is the sum of sinus of different frequencies. If our curve only consisted of a single sinus the spectral image of it would be a straight line in the frequency domain. Figure 96 - Sprectral analysis for one frequency 28 http://classes.yale.edu/fractals/CA/OneOverF/PowerSpectrum/PowerSpectrum.html Ana Ochotorena Portales 114 The cardiac cycle is the sum of sinus with different frequencies, if the cycle consisted of the green and red line the blue curve in figure 97 would be the cardiac cycle. The spectral analysis would then show that the red has a lower frequency and higher amplitude, than the green sinus. Figure 97 - Spectral analysis for more than one frequency The components of sinus displayed in figure 98 (left), shows that the cardiac cycle consists of every single frequency of sine from 0-500 [Hz], with various amplitudes. However many of the frequencies are due to noise, by considering the spectral analysis of the mean cardiac cycle, some of the frequencies are eliminated figure 98 (right). There are peaks around 200,300 and 400 [Hz] which suggests that there is something in the in vitro system that vibrates with these frequencies. These frequencies are very high and will affect the shape of the cardiac cycle. The cardiac cycle from old and the new model will be compared to determine if these signals have been eliminated, and to make sure a different set of frequencies does not occur in the new model. However the sources of the frequencies are unidentified and no active step has been made in the development of the new model, and should therefore be investigated further. Some suggested methods for locating source of high frequencies are Solid works frequency study or by the usage of accelerometers. Figure 98 - Spectral domain – cardiac cycle Ana Ochotorena Portales 115 As the noise seen has a similar shape compared to a step response, a method for recreating the step response has been proposed. The idea is show in the figure 99. The way to recreate it is by closing a chamber with a glove (or a similar object) and add compliance. Once is fulfilled, break it with a needle or a lighter in order to change the state as faster as possible obtaining the graph seen below. Figure 99 - System to recreate the step response Figure 100 - Step response Ana Ochotorena Portales 116 19.3. Test description and results Equipment used The equipment used for the test is named following: - New In Vitro model (Atrium, compliance chamber and ventricle chamber) - Silicone bag with 3 layers of silicone (if required) - Pump - Millar Catheters (pressure sensors) - Flow meter - Peripheral resistance - TAVI valve no. 3 - Terminal for adding compressed air to the compliance chamber. - Equipment for collecting the data Data set for the different tests carried out Discussion of the waveforms This analysis has been done with the new ventricle chamber without bag and the square compliance chamber. In this test an analysis of all the waveforms has been done in order to determine if waveforms have a significant influence on the way the results are being evaluated. It has been seen that the results will vary depending on the chosen waveform but nothing conclusive, see figure 101. For this reason all the tests will be performed with all the waveforms. Waveform B seems to behave better for diastole noise while waveform D behaves better for systole. However, irregular tendency with cardiac output increase has been seen for these waveforms as well. Hence, further tests will be performed with all the waveforms to conclude if the tendency is normal or is a matter of unsuitable measurements. The frequency is not dependant on the waveform as similar values have been achieved for the four waveforms. See figure 102. Ana Ochotorena Portales 117 Noise New model – VC without bag and square CC Systole noise A B C D CO 1-2 5.27305 7.32741 3.2661 6.9317 CO 2-3 6.16017 5.80488 4.5494 4.13645 CO 3-4 8.28195 10.9551 6.9317 4.935 CO 4-5 11.0607 10.5316 6.9317 4.59364 Average 7.6939675 8.6547475 5.419725 5.1491975 Diastole noise A B C D CO 1-2 2.16871 1.60103 2.55628 5.16281 CO 2-3 2.52883 1.76866 2.84054 5.24442 CO 3-4 2.88751 3.21079 5.16281 5.6498 CO 4-5 4.84071 4.37133 5.16281 6.23477 Average 3.10644 2.7379525 3.93061 5.57295 Figure 101 - Systole and diastole noise. Waveforms A, B, C and D Frequency New model – VC without bag and square CC Systole frequency A B C D CO 1-2 38.2627 36.4651 37.324 39.713 CO 2-3 39.7486 37.0199 38.2597 38.0869 CO 3-4 39.1001 36.7049 38.873 38.6228 CO 4-5 38.2527 38.5784 38.873 38.5825 Average 38.841025 37.192075 38.332425 38.7513 Diastole frequency A B C D CO 1-2 40.8022 44.8462 40.9105 41.999 CO 2-3 41.8146 40.6696 41.9404 42.166 CO 3-4 41.4954 41.2609 41.4841 43.1013 CO 4-5 42.8827 43.9037 41.8441 42.9482 Average 41.748725 42.6701 41.544775 42.553625 Figure 102 - Systole and diastole frequency. Waveforms A, B, C and D The waveform from the following test with the bag was also analysed in order to determine if the tendency seen before remained. This test was performed with ventricle chamber with the bag and squared compliance chamber. The results vary with the waveforms; again it is observed that waveform D behaves best for systole noise while waveform B is best for diastole. The normal tendency is for the noise to increase with the cardiac output but with some anomalies. These irregularities seem to reoccur for waveforms B and D, same behaviour as seen in the test without the bag. Again is seen that the frequency is fairly constant. Ana Ochotorena Portales 118 Noise New model – VC with bag and square CC Systole noise A B C D CO 1-2 9.50493 7.74308 7.54304 4.53959 CO 2-3 11.7904 9.41204 9.17174 3.83601 CO 3-4 12.8533 12.2057 9.09903 3.16448 CO 4-5 12.4849 9.78726 10.1701 6.08805 CO 5+ 17.5863 15.9889 9.97306 8.28016 Average 12.843966 11.027396 9.191394 5.181658 Diastole noise A B C D CO 1-2 3.99203 1.75161 3.71356 3.52167 CO 2-3 5.77231 2.66389 4.99715 5.11579 CO 3-4 4.75373 3.37348 4.77296 6.46826 CO 4-5 5.16877 3.66013 5.55667 7.33703 CO 5+ 5.22591 4.71436 6.61562 8.15047 Average 4.98255 3.232694 5.131192 6.118644 Figure 103 - Systole and diastole noise. Waveforms A, B, C and D Frequency New model – VC with bag and square CC Systole frequency A B C D CO 1-2 21.3583 27.0612 27.1455 33.6153 CO 2-3 21.6096 27.1298 27.5494 33.7924 CO 3-4 25.6599 26.6736 29.0966 33.2312 CO 4-5 26.1756 26.1099 29.9896 33.0847 CO 5+ 33.9228 27.3679 31.5101 33.4117 Average 23.70085 26.743625 28.445275 33.4309 Diastole frequency A B C D CO 1-2 29.5407 35.236 36.3596 42.1129 CO 2-3 29.9689 37.1337 36.6945 43.3881 CO 3-4 34.9756 35.9982 39.093 42.9314 CO 4-5 35.5012 36.768 38.9684 43.2502 CO 5+ 43.7516 37.1853 41.7736 44.8123 Average 32.4966 36.283975 37.778875 42.92065 Figure 104 - Systole and diastole frequency. Waveforms A, B, C and D To conclude what can be said for the waveform analysis is that the waveform D is better for systole and waveform B is better for diastole. Further tests with the trapezoidal chamber with and without bag indicated that the waveform analysis was consistent with what was previously observed. Ana Ochotorena Portales 119 Comparison between the old model and the new model (the ventricle chamber is without bag and the two available new compliance chambers) This test was performed to determine if any of the new chambers behaves better in the new model without bag in comparison to the old model the following was found: - The model with trapezoidal chamber has a smaller area of noise than the model with square chamber in systole and diastole. - The old model has less area of noise compared to the new model regardless of the compliance chamber in diastole. - Frequency for the new model was generally higher compared to the old in systole - In diastole the model with square chamber has the same frequencies as the old. - Because of the irregular frequencies recorded for the trapezoidal chamber in diastole the data is considered invalid. Noise Old model New model – without bag Systole noise Old chamber Trapezoidal chamber Square chamber CO 1-2 2.53 3.28482 6.9317 CO 2-3 5.35 4.56418 4.13645 CO 3-4 6.033524 6.21045 4.935 CO 4-5 8.33947 4.82717 4.59364 Average 5.5632485 4.721655 5.1491975 Diastole noise Old chamber Trapezoidal chamber Square chamber CO 1-2 1.688 3.04032 5.16281 CO 2-3 3.4758 4.01024 5.24442 CO 3-4 2.56 4.09329 5.6498 CO 4-5 3.56965 4.9726 6.23477 Average 2.8233625 4.0291125 5.57295 Figure 105 - Systole and diastole noise comparison old model and new model without bag Frequency Old model New model – without bag Systole frequency Old chamber Trapezoidal chamber Square chamber CO 1-2 32.927 44.1247 39.713 CO 2-3 32.2672 40.3695 38.0869 CO 3-4 33.6683 38.2834 38.6228 CO 4-5 32.0948 38.7858 38.5825 Average 32.739325 40.39085 38.7513 Diastole frequency Old chamber Trapezoidal chamber Square chamber CO 1-2 44.8663 47.461 41.999 CO 2-3 45.4804 121.727 42.166 CO 3-4 45.6438 43.9603 43.1013 CO 4-5 44.2663 45.386 42.9482 Average 45.0642 64.633575 42.553625 Figure 106 - Systole and diastole frequency comparison old model and new model without bag Ana Ochotorena Portales 120 Comparison between the old model and the new model (the ventricle chamber is with bag and the two available new compliance chambers). This test was performed to determine if the model with bag has an effect in the noise. Comparing it with the old model the following was found: - The square chamber has less systole noise than the trapezoidal new chamber and the old model. - The old chamber has less diastole noise than the new model - The trapezoidal chamber has a lower frequency in comparison with the other models for systole and diastole. Noise Old model New model – with bag Systole noise Old chamber Trapezoidal chamber Square chamber CO 1-2 2.53 5.79515 4.53959 CO 2-3 5.35 7.01218 3.83601 CO 3-4 6.033524 7.37454 3.16448 CO 4-5 8.33947 5.23202 6.08805 Average 5.5632485 6.3534725 4.4070325 Diastole noise Old chamber Trapezoidal chamber Square chamber CO 1-2 1.688 3.49664 3.52167 CO 2-3 3.4758 4.75574 5.11579 CO 3-4 2.56 6.20574 6.46826 CO 4-5 3.56965 3.05462 7.33703 Average 2.8233625 4.378185 5.6106875 Figure 107 - Systole and diastole noise comparison old model and new model with bag Frequency Old model New model – with bag Systole frequency Old chamber Trapezoidal chamber Square chamber CO 1-2 32.927 23.0417 33.6153 CO 2-3 32.2672 26.8021 33.7924 CO 3-4 33.6683 27.0973 33.2312 CO 4-5 32.0948 26.8662 33.0847 Average 32.739325 25.951825 33.4309 Diastole frequency Old chamber Trapezoidal chamber Square chamber CO 1-2 44.8663 35.5702 42.1129 CO 2-3 45.4804 35.5116 43.3881 CO 3-4 45.6438 35.7547 42.9314 CO 4-5 44.2663 36.4439 43.2502 Average 45.0642 35.8201 42.92065 Figure 108 - Systole and diastole frequency comparison old model and new model without bag Ana Ochotorena Portales 127 Figure 122 - New model running with high CO -0.4 -0.2 0 0.2 0.4 0.6 0.8 1 1.2 1.4 1 42 83 124 165 206 247 288 329 370 411 452 493 534 575 616 657 698 739 780 New model Ana Ochotorena Portales 128 20Project management planning’s tools. FMEA and Fishbone analysis 29 Quality planning is part of the project quality management. In this stage is where the relevant standards are going to be clarified and the decision of how are going to be fulfilled will be made. In order to prepare a good quality management plan, some tools have to be carried on, for instance: brainstorming, FMEA or Fishbone analysis. In this section is going to be described the two tools that have been developed at the begging of the project in order to clarify, organize and decide which problems should be resolved and how. 20.1. FMEA FMEA explanation 30 FMEA is the acronym for Failure Mode and Effect Analysis. It is a method used to analyse potential failures in a system determined by gravity or effect of failures. With this method all the possible stages of failure are analysed. Applying the equations (explained after) is possible to achieve which of them has the biggest risk and therefore be able to focus on it. FMEA worksheet Following are presented the different parts in the model which are going to be taken into account in the analysis. There are also listed and rated all the potential failures. The result of that rating will show in which component are the highest risks and how it is possible to solve the problem. - Overall model o Connections between parts and chambers o Drainage - Compliance chamber o Maintain pressure o ARAMIS - Ventricular chamber o Silicone bag o Connections module between Mitral and Aortic valve - Atrium chamber o Pressure o Dimensions/Volume - Aortic root o Measurements o Flow (Smaller root, silicone root) o Pressure o Ultrasonic transducer o Longer root 29 Kousholt, Bjarne (2012). Project Management. Theory and Practice, 2. ed. Nyt Teknisk Forlag 30 This analysis is done towards the M7BACH14 group Ana Ochotorena Portales 129 - Tests o Equipment failure o Pump failure o Lack of knowledge o Problems procedure o Problems with the model o Verifying the model Risk level calculation The rating of the components will be made by using the following rating schedules. Therefore, it is possible to find which of the components has the highest risks. The risk level is found by multiplying the probability (P) with the severity (S). - Risk level: (P x S) and (D): It is a combination of the End Effect, Probability and Severity. Probability/Severity I II III IV V VI A Low Low Low Low Moderate High B Low Low Low Moderate High Unacceptable C Low Low Moderate Moderate High Unacceptable D Low Moderate Moderate High Unacceptable Unacceptable E Moderate Moderate High Unacceptable Unacceptable Unacceptable Table 15 - Risk level (P x S) and (D) - Probability (P): It is necessary to look at the cause of a failure mode and the likelihood of occurrence. Rating Meaning 1 Extremely Unlikely (Virtually impossible or No known occurrences on similar products or processes, with many running hours. 2 Remote (relatively few failures) 3 Occasional (occasional failures) 4 Reasonably Possible (repeated failures) 5 Frequent (failure is almost inevitable) Table 16 - Probability (P) Ana Ochotorena Portales 130 - Severity (S): Determine the severity for the worst case scenario adverse the effect. Rating Meaning I No relevant effect on reliability or safety II Very minor, no damage, no injuries, only results in a maintenance action (only noticed by discriminating customers) III Minor, low damage, light injuries IV Moderate, moderate damage, injuries possible. V Critical VI Catastrophe Table 17 - Severity (S) - Detection (D): Determine the likelihood for detect a failure mode. Rating Meaning 1 Certain – fault will be caught on test 2 Almost certain 3 High 4 Moderate 5 Low 6 Fault is undetected by Operators or Maintainers Table 18 - Detection (D) Conclusions Watching at the FMEA table is possible to know that the construction failures are the biggest predictable problems. The problems with higher risk level are the ones related with the connections’ section, especially problems due to leakage and rust in the connections in the design and assembly phases. There is also a big issue related with the silicone bag design in the ventricle chamber. Therefore, construction failures can be mitigated by doing exhaustive work in the design phase; however there is a deadline for finishing the models. So it will be prioritized finishing the model over having the best possible one. Ana Ochotorena Portales 131 FMEA table Below is the table of the FMEA model. Ana Ochotorena Portales 132 Ana Ochotorena Portales 133 Ana Ochotorena Portales 134 20.2. Ishikawa diagram or Fishbone analysis Diagram explanation The Ishikawa diagram is also known as the fishbone analysis and it is also another planning tool. This diagram resembles the skeleton of a fish, in its head is written the problem that a solution is required for and on its bone is written the causes of it. Normally this diagram is also connected with brainstorming. The problem used to solve by Fishbone analysis was the entire design of the new in-vitro model taking into account all the requirements given and the possible problems associated. Diagram 31 Below is represented the fishbone analysis done. Figure 123 - Fishbone analysis 31 This analysis is done towards the M7BACH14 group Ana Ochotorena Portales 135 21List of figures, tables and equations Figures Figure 1 - Isometric view – square compliance chamber (A.O.P.) .............................................. 13 Figure 2 - Left and right view of the square compliance chamber (A.O.P.) ................................ 14 Figure 3 - Front view of the square compliance chamber (A.O.P.) ............................................. 14 Figure 4 - Top view of the square compliance chamber (A.O.P.) ................................................ 15 Figure 5 - Initial square compliance chamber (A.O.P.) ................................................................ 15 Figure 6 - O-ring connection (left) and silicone ring connection (right) (A.O.P.) ........................ 16 Figure 7 - Almost final square compliance chamber design – without corner cut (A.O.P.) ........ 16 Figure 8 - Isometric view – trapezoidal compliance chamber (A.O.P.) ....................................... 17 Figure 9 - Left and right side of the trapezoidal compliance chamber (A.O.P.) .......................... 17 Figure 10 - Front view of the square compliance chamber (A.O.P.) ........................................... 18 Figure 11 - Top view of the square compliance chamber (A.O.P.).............................................. 18 Figure 12 - Circular shaped fitting connection (A.O.P.) ............................................................... 19 Figure 13 - Rectangular shaped fitting connection (A.O.P.) ........................................................ 19 Figure 14 - O-ring connection and silicone ring connection (A.O.P.) .......................................... 20 Figure 15 - Millar catheters' positions (A.O.P.) ........................................................................... 20 Figure 16 - Arotic root assembly (A.O.P.) .................................................................................... 20 Figure 17 - Fittings – rods changed in position (A.O.P.) .............................................................. 21 Figure 18 - Connector sinus of Valsalva and silicone tube – Hand drawing (A.O.P.) .................. 21 Figure 19 - Connector sinus of Valsalva and silicone tube – Solid Works (A.O.P.)...................... 21 Figure 20 - Previous design for the connector - Hand drawing (A.O.P.) ..................................... 22 Figure 21 - Connector silicone tube and fitting – Hand drawing (A.O.P.) ................................... 22 Figure 22 - Connector silicone tube and fitting – Solid Works (A.O.P.) ...................................... 22 Figure 23 - Aortic root assembly for flow calculations – front view (A.O.P.) .............................. 23 Figure 24 - Aortic root assembly for flow calculations (A.O.P.) .................................................. 23 Figure 25 – Square compliance chamber assembly (A.O.P.) ....................................................... 24 Figure 26 - Parts of the square compliance chamber (A.O.P.) .................................................... 24 Figure 27 - Trapezoidal compliance chamber assembly (A.O.P.) ................................................ 25 Figure 28 - Parts of the trapezoidal compliance chamber (A.O.P.) ............................................. 25 Figure 29 - Aortic root assembly (A.O.P.) .................................................................................... 26 Figure 30 - Parts of the aortic root assembly (A.O.P.)................................................................. 26 Figure 31 - Silicone ring (A.O.P.) .................................................................................................. 63 Figure 32 - Mould assembly for the silicone ring (A.O.P.) ........................................................... 63 Figure 33 - Female part (right) and male part (left) (A.O.P.) ....................................................... 63 Figure 34 - Torque wrench (A.O.P.) ............................................................................................. 64 Figure 35 - Bench Vice (A.O.P.) .................................................................................................... 64 Figure 36 - Specimens (A.O.P.) .................................................................................................... 64 Figure 37 - General assembly (A.O.P.) ......................................................................................... 65 Figure 38 - Simplified compliance chamber (A.O.P.)................................................................... 68 Figure 39 - Graph of forces in the compliance chamber's back wall (A.O.P.) ............................. 70 Figure 40 - Glue connection of two plates with dimensions ...................................................... 71 Ana Ochotorena Portales 136 Figure 41 - Shear stress distribution in a glue connection of two plates ................................... 71 Figure 42 - Pressure distribution in the compliance chamber (A.O.P.) ....................................... 72 Figure 43 - Top plate dimensions (A.O.P.) ................................................................................... 73 Figure 44 - Pressure distribution in the top plate (A.O.P.) .......................................................... 73 Figure 45 - Side plate dimensions (A.O.P.) .................................................................................. 74 Figure 46 - Pressure distribution in the side plate (A.O.P.) ......................................................... 74 Figure 47 - Side plate dimensions (A.O.P.) .................................................................................. 75 Figure 48 - Pressure distribution in the side plate (A.O.P.) ......................................................... 76 Figure 49 - Forces in the screws/Old chamber (A.O.P.) .............................................................. 77 Figure 50 - Stresses in the screws/Old chamber (A.O.P.) ........................................................... 77 Figure 51 - Top plate dimensions (A.O.P.) ................................................................................... 78 Figure 52 - Pressure distribution in the top plate (A.O.P.) .......................................................... 78 Figure 53 - Forces in the screws (A.O.P.) ..................................................................................... 79 Figure 54 - Stresses in the screws (A.O.P.) .................................................................................. 79 Figure 55 - Pressure distribution and fixtures (A.O.P.) ............................................................... 80 Figure 56 – Stress/Top plate old model (A.O.P.) ......................................................................... 81 Figure 57 - Displacements/Top plate old model (A.O.P.) ............................................................ 81 Figure 58 - Strain/Top plate old model (A.O.P.) .......................................................................... 81 Figure 59 - Pressure distribution and fixtures/Top plate old model (A.O.P.) ............................. 82 Figure 60 - Stress/Top plate old model (A.O.P.) .......................................................................... 82 Figure 61 - Displacements/Top plate old model (A.O.P.) ............................................................ 82 Figure 62 - Strain/Top plate old model (A.O.P.) .......................................................................... 82 Figure 63 - Pressure distribution and fixtures/Side plate old model (A.O.P.) ............................. 83 Figure 64 - Stress/Side plate old model (A.O.P.) ......................................................................... 84 Figure 65 - Displacements/Side plate old model (A.O.P.) ........................................................... 84 Figure 66 - Strain/Side plate old model (A.O.P.) ......................................................................... 84 Figure 67 - Pressure distribution and fixtures/Top plate old model (A.O.P.) ............................. 85 Figure 68 - Stress/Front plate old model (A.O.P.) ....................................................................... 86 Figure 69 - Displacements/Front plate old model (A.O.P.) ......................................................... 86 Figure 70 - Strain/Front plate old model (A.O.P.) ....................................................................... 86 Figure 71 - Stress distribution in the holes of the old model’s back plate ................................. 87 Figure 72 - Pressure distribution and fixtures/Top plate new model (A.O.P.) ............................ 88 Figure 73 - Stress/Top plate new model (A.O.P.) ........................................................................ 88 Figure 74 - Displacements (A.O.P.) .............................................................................................. 89 Figure 75 - Strain/Top plate new model (A.O.P.) ........................................................................ 89 Figure 76 - Short aortic root relevant dimensions (A.O.P.) ......................................................... 92 Figure 77 - Long aortic root relevant dimensions (A.O.P.) .......................................................... 92 Figure 78 - Table result for the short aortic root with CO of 5 [L/min] ...................................... 93 Figure 79 - Table result for the short aortic root with CO of 11 [L/min] .................................... 93 Figure 80 - Table result for the long aortic root with CO of 5 [L/min] ....................................... 94 Figure 81 - Table result for the long aortic root with CO of 11 [L/min] ..................................... 94 Figure 82 - Bend loss coefficient calculation graph .................................................................... 97 Figure 83 - Ventricle chamber with the two valves' connection (A.O.P.) ................................... 98 Figure 84 - Ventricle chamber bent one (A.O.P.) ........................................................................ 99