Senior Design Project P07110 METEOR Vertical Test Stand Rochester Institute of Technology, Rochester, New York

Transcription

1 Senior Design Project P07110 METEOR Vertical Test Stand Rochester Institute of Technology, Rochester, New York Michael Saitta (Team Leader)- Mechanical Engineering Brian Frost- Mechanical Engineering Amberly Prill- Electrical Engineering Kevin Saboda- Mechanical Engineering Daniel Uerz- Electrical Engineering Ethan Willistein- Mechanical Engineering Team Guide: Dr. Jeffrey Kozak- Mechanical Engineering May 16, 2007 Contents I Introduction and Design Generation 3 1 Introduction 3 2 Customer Needs Mechanical Needs Data Acquisition Needs Mechanical Concept Generation Initial Concepts Rails on Rails Ethan Orthogonal Frame Ball-Bearing Tube Axis Pivot Single Cantilever Double Cantilever Gimbal Initial Concept Evaluation Additional Concept Evaluation Final Concept Chosen Gimbal with Tube Concept 8 5 Initial Design Constants 8 II Mechanical Analysis 8 6 Tapered Roller Bearing Assembly Tapered Roller Bearings Bolts Holding Bearing Case Together Summary of Standard Bolt Preloads Bolt Sizing Shaft in Bearing Case Initial Hand Calculations FEA Analysis of Shaft Universal Joint FEA Analysis on Modified Universal Joint Shearing In Bolts Load Cell Connection to Rocket Assembly Tension and Compression in Bolt Shear of Threads One Thread Carries Load Threads Carry Load Frame Stress in Frame Uprights Weld at Base of Frame Uprights.. 19 III Data Acquisition Analysis Load Cell Amplifier First Design Final Design

2 11 DAQ Housing Box Connector Choice Power Supply Analog to Digital Converters Thermocouple Inputs Load Cell Amplifier IV Force Transducer Selection and Calibration Transducer Selection Vertical Thrust Force Lateral Force Roll Moment Transducer Calibration Vertical Thrust Force Lateral Force Roll Moment- Torsional Force Sensor 26 V Risk Assessment Risk Management Philosophy Standards for the Ranking of Hazardous Conditions Standards for the Probability of a Hazardous Condition Occurring Hazards in the Design Frequent Hazards Probable Hazards Occasional Hazards Remote Hazards Improbable Hazards Risk Management Summary 31 List of Figures 1 Rails on Rails idea Ethan 2 design Orthogonal Frame Ball-Bearing concept Tube concept idea Axis Pivot Single Cantilever concept Double Cantilever Gimbal Diagram of tapered roller bearing assembly Dimension schematic for hand calculation of bearing case shaft stress calculations FEA model of bearing shaft in tension FEA model of bearing shaft in compression FEA model of bearing shaft under 100 in-lb f torque Bolts through universal joint Free body diagram for bolt through the universal joint FEA analysis for compression case FEA analysis for tension case Bolt connecting load cell to rocket assembly Force applied to beam for stress computation Detail of plate welded to frame upright CAD model of test stand design with frame highlighted First design of amplifier circuit Final load cell amplifier circuit DAQ housing box Vertical thrust force sensor Lateral force sensor Roll moment load cell Steel tube surrounding the rocket (concrete bunker not shown for clarity) Hole in the top of the concrete bunker. 28 List of Tables 1 Initial concept selection scoring matrix Initial sizing and force/moment specifications Final ranking of concepts Summary of developed bolt preloads Strength properties of SAE Grade 8 bolts Hazard ranking categories Hazard probability scale Frequent hazards Occasional hazards Remote hazards Improbable hazards

3 Part I Introduction and Design Generation 1 Introduction The Microsystems Engineering and Technology for the Exploration of Outer Space Regions (ME- TEOR) project at Rochester Institute of Technology is an undergraduate student effort to launch small payloads into outer space. This design report details the work completed by four mechanical and two electrical engineering students in the process of fulfilling their senior project requirement for an undergraduate degree in engineering. The responsibility of the P Vertical Test Stand Team was to create a design for a vertical rocket test stand that meets the requirements necessary of other teams in the METEOR project. The main purpose of the vertical test stand is to provide accurate test data to the P Guidance and P Flying Rocket teams so that they can design and optimize the rocket that METEOR intends to produce for launch into orbit. Our team has heavily built upon the work done by the senior design team that created a horizontal orientation test stand last year. We have used the horizontal test stand as a learning experience and have adapted relevant concepts to the vertical test stand. The data acquisition system of the horizontal test stand is also very similar to that of the vertical test stand. However, the collection of more dynamic forces of the rocket is required with the new vertical rocket stand design. By placing the rocket in the vertical orientation, the effect of orientation of the rocket can be studied and the final rocket design can be verified. 2 Customer Needs Our principal customers are other METEOR senior design teams. Data collected from the vertical test stand will help in the development and verification of the designs of teams charged with supplying the rocket itself, rocket guidance system, and the hybrid rocket engine. In light of this fact considerable cooperation between teams within the METEOR program has been necessary. 2.1 Mechanical Needs Two forces and one moment are of particular interest in the development of the guidance system for the rocket. The thrust force of the rocket along with a measure of how much the thrust force may deviate from the longitudinal axis of the rocket need to be measure. This deviation from the axis of the rocket gives rise to a lateral force component to the rocket. There may also be some amount of roll moment around the longitudinal axis of the rocket. The team designing the hybrid rocket engine is most interested in the thrust measurements and the guidance team is more interested in the lateral force and roll moment measurements. Another secondary area where experimental data is desired is to quantify any vibrations or acceleration of the rocket while it is in the test stand. This data is of interest to the guidance team in the short term and for future development of the METEOR program. Vibration data may be of particular interest to future teams designing the systems of the actual rocket stages that will be launched into orbit. To achieve safety during testing and to capture the test data required the rocket will need to be securely restrained during the test. Any hazards created by a malfunction of the rocket during testing, up to and including explosion of the rocket, will also have to be contained and mitigated. The very nature of the type of data desired, however, necessitates some freedom of motion of the rocket. Therefore, it is essential to have a test stand design that will keep the rocket constrained safely. Also, the flame plume of the rocket will need to be handled in a way that ensures accurate test data and mitigates the risk of fire. 2.2 Data Acquisition Needs Data will need to be collected from a variety of different types of sensors. Measurements of force, pressure, temperature, and acceleration need to be made during testing. As a result of previous work done for the METEOR project, there is already a supply of some commercially produced sensors that our data acquisition system needs to interact with. Also, commercially produced data acquisition cards that interface with laptop computers are already possessed by the project. Some of the currently owned transducers do not have amplified outputs, so it is also necessary to provide amplification for these signals. A high gain amplifier circuit and its necessary power supply 3

4 Figure 1: Rails on Rails idea. Figure 2: Ethan 2 design. source need to be provided by our team. Additionally, it is desired to organize all of the signal inputs, data acquisition cards, and amplifier circuits into a modular system that allows for quick set-up during testing. 3 Mechanical Concept Generation A concept generation process was followed to develop ideas for the structure and mechanics of the vertical test stand. Design ideas were evaluated for their ability to capture a thrust magnitude and direction for the rocket, the ability to measure the roll moment created by the rocket, and ability to restrain the rocket safely. 3.1 Initial Concepts Rails on Rails The Rails on Rails idea shown in 1 uses two orthogonal sets of rails with linear bearings to provide freedom of movement for the rocket. Torsional force is not easily measurable with this design. linear bearings to provide provision for thrust measurement. Lateral forces could then be captured at the point at which the frame is held Orthogonal Frame The Orthogonal Frame concept proposed to measure the magnitude and direction of the rocket thrust force by measuring the reaction forces at the base of the three orthogonally arranged frame supports. This is displayed in figure 3 on the following page Ball-Bearing In the Ball-Bearing concept, the rocket is allowed freedom of movement by securing the rocket in a large ball joint. Figure 4 on the next page shows how the rocket would be allowed limited movement using the ball joint Ethan 2 Figure 2 shows the Ethan 2 design. This design proposed to contain the rocket in a frame with a pair of Tube The major feature of the Tube design, shown in figure 5 on the following page, is that the rocket is 4

5 Figure 5: Tube concept idea. Figure 3: Orthogonal Frame. contained in a metal tube for safety measures and restrained by the lateral and thrust force transducers. Figure 4: Ball-Bearing concept Axis Pivot Two nested rings pivoted at ninety degrees to each other give lateral freedom in this design. The Axis Pivot shown in figure 6 on the next page is similar in principle to the Ball-Bearing concept idea Single Cantilever As shown in figure 7 on the following page, the Single Cantilever idea holds the rocket by rigidly attaching it to a cantilevered beam. By measuring resultant stresses in the beam and frame of the structure while the rocket is being fired, the forces created by the rocket could be determined Double Cantilever The Double Cantilever design is very similar in design to the Single Cantilever concept; this can be seen in figure 8 on the next page. The fact that the rocket is held on two sides using two cantilevered beam is the differentiating factor between these two ideas. 5

6 Figure 8: Double Cantilever. Figure 6: Axis Pivot Gimbal Figure 7: Single Cantilever concept. A universal joint provides freedom of motion in the Gimbal idea. The top of the rocket is attached to the universal joint and force sensors contacting the rocket near its center of mass constrain the rocket and measure lateral force. These details are visually shown in figure 9 on page Initial Concept Evaluation To establish a quantitative method for evaluating each of the concepts they were scored as better, worse, or equal to a benchmark idea with regards to the customer needs. The benchmark idea was selected to be the Rails on Rails concept. Table 1 on the next page shows the scoring matrix for the initial concept evaluation. From this it was decided to proceed on with the number one and two ranked concepts. It should be noted that there was a for way tie for the ranking of number two. The ideas proceeding onto the next step are: Ethan 2, Orthogonal Frame, Tube, Gimbal, and Dual Cantilever. 3.3 Additional Concept Evaluation At this point, some slight improvements were made to the concepts that were advancing in the design process. The evaluation process in this stage served 6

7 Table 1: Initial concept selection scoring matrix. 7

8 Figure 9: Gimbal. Table 3: Initial sizing and force/moment specifications. Specification Value Maximum Rocket Diameter 12 in Maximum Rocket Length 50 in Estimated Plume Length 3 ft Maximum Rocket Thrust 200 lb f Maximum Rocket Weight 50 lb f (including fuel) Maximum Lateral Force Expected 25 lb f Maximum Roll Moment Expected 100 in-lb f to further differentiate and more finely rank the ideas that we were considering using in the final design. Weighing factors were then attached to each of the customer needs based on their relative importance to the design. How well each concept satisfied each of the customer needs was then rated on a zero to five scale, with five denoting the highest possible fulfillment of the need. The sum of the scores times the weighing factors effectively assigns a suitability score to each of the design ideas. Table 2 on the following page contains the detailed scoring results for each of our designs. of the data required and provide safety during testing. Appropriate freedom of movement is allowed to the rocket, while still securing the rocket safely within the test stand and the concrete bunker. Freedom of movement is provided via a linear bearing system and an universal joint. The rocket will be attached to a universal joint that allows it to rotate and swing laterally. This whole assembly will then be able to move up and down on a linear bearing system. A force transducer at the top of the linear bearing system will capture the thrust force of the rocket. Lateral forces will be recorded by force transducers attached to a ring that surrounds the rocket. This ring also functions as a safety component of the design because it will prevent gross movement of the rocket. Roll moment is measured by using a torque bar connecting the universal joint to a force transducer. Vibration and/or acceleration data is found by direct attachment of accelerometers to the rocket body. 3.4 Final Concept Chosen The final concept chosen to proceed with in the design process was the Gimbal with Tube concept. This design had the highest ranking score in the concept evaluation process and was a combination of the best attributes of two different concepts from the initial concept evaluation. 5 Initial Design Constants Table 3 below summarizes the initial a priori design specifications that the vertical test stand is being tailored to. These constraints and guidelines were formed for two reasons. They provide a starting point for our team s calculations and serve to define the relationship that our project has with other projects in the METEOR program. 4 Gimbal with Tube Concept The Gimbal with Tube concept fulfills the customer needs very satisfactorily by being able capturing all 8

9 Table 2: Final ranking of concepts. 9

10 Figure 10: Diagram of tapered roller bearing assembly. Part II Mechanical Analysis 6 Tapered Roller Bearing Assembly 6.1 Tapered Roller Bearings Two tapered roller bearings mounted in series provide freedom of roll movement; this is shown in figure 10. Tapered bearings were selected because the shaft coming from the universal joint will see a torque and an axial force. Since the axial force will change in direction depending on whether the rocket is hanging from the stand or firing with an upward thrust, two bearings were used. In light of the fact that the bore of the universal joint is 1 in., the bore of the tapered bearings were selected to be 1 in. Elaborate engineering analysis was not done on the bearings because they will not support a constantly rotating shaft. The supplier of the bearings gives dynamic load capacities of 1620 lb f in the radial direction and 1040 lb f in the axial direction. Both of these values are well within any loading expected to be placed on the bearings. 6.2 Bolts Holding Bearing Case Together A bolt pattern, around the outer edge of the tapered roller bearing case, was decided as the method for holding the top and bottom of the bearing case together. For engineering simplicity, it was also decided to use a fastener going through the bearing case and terminated with a nut and washer. Figure 10 shows how the bolts go though the bearing case. The bolted joint is designed so that enough preload is developed in the bolts to prevent the three sections of the bearing case from separating in any loading case apt to occur. At this point in the design, the only concrete loading data is that the maximum expected thrust of the rocket is 200 lb f and the maximum weight of the fueled rocket will be 50 lb f. An unknown piece of loading data is the weight of the testing apparatus connecting the rocket to the load cell. The total load placed on the rocket will be the net sum of the three forces described above. Additionally, it can be said that the rocket thrust will act in the opposite direction of both the weight of the rocket and the testing apparatus. Therefore, if the combined weight of the rocket and testing apparatus are less than the rocket thrust, the net force never exceed the magnitude of the thrust force. We feel that using the maximum thrust force as the design load on our test apparatus because we feel it is very reasonable to assume that the weight of the rocket and testing apparatus will not be greater in magnitude than the rocket thrust Summary of Standard Bolt Preloads The preloads (F i ) obtainable for a standard reusable bolted joint containing are given in table 4 on the following page for fasteners ranging in size from #6 to 1/4 in. The decision to use stainless steel fasteners is made up front because of the corrosion resistance that stainless steel provides. With this in mind, the proof strength (S p ) of the stainless steel fasteners used is 34 ksi. This value was obtained by using the relationship that the proof strength is generally equal to 85% of the tensile yield strength of the fastener material. F i =.75A t S p A t standard tensile stress area 10

11 Table 4: Summary of developed bolt preloads. Size A t (in 2 ) F i (lb f ) Bolt Sizing # # # # # / / From table 4 it can be safely said that many of the bolt sizes given will satisfy the necessary requirements. So that a symmetric clamping force is applied to the end plates of the bearing case, a bolt pattern of 4 bolts was selected. The size of the bolts selected are #10-24 because they provide a balance between overall strength of the bolt and the bearing case. Using the preload results obtained before, the final total preload force holding the bearing case together is computed to be about 1785 lb f. If the net force created by the rocket during testing is transferred as a thrust into the bearings, then the tendency of this force will be to try to pull the bolted connection between the three components of the bearing case apart. Our design case uses a design load, F, equal to the maximum thrust of the rocket. This leads to a factor of safety, against separation of the components of the bearing case, of 8.9. n = ΣF i F n = 1785 lb f 200 lb f = Shaft in Bearing Case On the shaft inside the bearing case there is a shoulder that supports the weight of the rocket and any thrust forces during the rocket firing. Stress concentrations occur at changes in diameters of shafts and the bolt holes that connect the shaft to the universal joint. Initial hand calculations and then FEA models were run on the shaft because of the stress concentrations Initial Hand Calculations The hand calculations were done assuming a 200 lb f tensional load applied to the shaft. Only the tension case was analyzed because common stress concentration factor charts only provide data for loading in tension. In addition the tensional loading cases generally result in higher stress states. Figure 11 on the following page shows the dimensions required for the shaft to fit in the bearing housing and the variables that those dimensions correspond to in the stress concentration factor equations. The stress concentration in the shoulder region of the shaft is dependent on the ratio of the two shaft diameters and the fillet radius between those diameters. These diameters and the fillet radius were determined from the confines of the bearings. Using figure A-15-7 in Mechanical Engineering Design the value of the stress concentration factor was visually determined. D in = = d 1 1 in r.050 in = d 1 1 in =.050 K t,shoulder = 2.2 This stress concentration factor is then applied to the basic formula for pure tensional stress in the region of the minor diameter of the shaft. σ = K t,shoulder F A, A = π(d 1 2 )2 σ = lb f π( 1 in 2 )2 = 560 psi Since there may be play in the drilled bolt holes at the end of the shaft, it is possible for only one hole to carry the entire load placed on the shaft. Accordingly the calculation assumes this fact. The stress concentration factor was again visually determined using figure A-15-1 in Mechanical Engineering Design. In the formula below A min refers to the minimum cross sectional area of the shaft at the hole and this value was calculated to be approximately equal to.319 in 2. d 2.5 in = d 1 1 in =.5 K t,hole = 2.2 σ = K t,hole F A min σ = lb f = 1379 psi.319 in2 11

12 Figure 11: Dimension schematic for hand calculation of bearing case shaft stress calculations FEA Analysis of Shaft Models were run for a 200 lb f force in compression and tension applied at the bolt connection closest to the end of the shaft. The shoulder of the shaft was fixed. The material was chosen as AISI 304 stainless steel shaft with a yield strength of 32 ksi. Figure 12 shows the results from the tension case. The maximum von Mises stress in tension was 2756 psi, resulting in a factor of safety of 11. Figure 12: FEA model of bearing shaft in tension. Figure 13 on the following page shows the results from the compression case. The maximum von Mises stress in compression was 1509 psi, resulting in a factor of safety of 21. A third case, where a 100 in-lb f moment was applied to the bottom of the shaft, was also run to simulate seizure of the taper bearings. Figure 14shows this and the maximum von Mises stress in this case was 5438 psi. In this case the resulting factor of safety was 6. 7 Universal Joint A Curtis universal joint was selected to provide freedom of lateral movement for the rocket. The 2 in. universal joint comes with a 1 in. bore for mounting the shaft from the tapered roller bearing case to the universal joint. The manufacturer of the universal joint gives maximum loads that the joint can withstand. Additional analysis was also done on the universal joint because the design requires modification of the stock universal joint. 12

13 Figure 13: FEA model of bearing shaft in compression. Figure 14: FEA model of bearing shaft under 100 in-lb f torque. 13

14 The manufacturer of the universal joint gives a static maximum torque rating of in-lb f and a maximum axial force value of lb f. Both of these values are so large that they are not limiting factors to the design. It is desired though to drill two holes into the universal joint to attach the shafts however, so an FEA analysis was done on the modified universal joint. 7.1 FEA Analysis on Modified Universal Joint Two FEA models were run on the modified universal joint yoke. One assumed a 200 lb f compression loading and another assumed a 200 lb f load in tension. Both models used the constraint that the center pin of the universal joint remained fixed. It is entirely possible for only one of the bolts between the bearing case shaft and the modified universal joint yoke to feel the load, so the models were run with the 200 lb f load applied to the hole closest to the opposite end of the universal joint. The force was placed at this location because it is of importance to gain insight to the stress flow around the other hole in the shaft. The yoke was assumed to be made out of alloy steel with a yield strength of 90 ksi. Results for the compression case are shown in figure 15 on the next page. In compression the maximum von Mises stress in the part was 778 psi. This yields a factor of safety of 115. Figure 16 on page 16 shows results for the FEA model involving tension. The maximum von Mises stress for this case is 2341 psi, resulting in a factor of safety of 38. Figure 17: Bolts through universal joint. Figure 18: Free body diagram for bolt through the universal joint. free body diagram for the bolt as analyzed. Since the bolt is in double shear, the total shear force being applied across each load bearing section of the bolt is 100 lb f. The minor area of the bolt that the shear force is applied over is A s =.1486 in 2. Using these conditions the shear stress in the bolt is calculated below. 7.2 Shearing In Bolts As explained in the previous section, the worst case scenario is for one of the bolts shown in figure 17to be carrying the entire 200 lb f load. Another worst case scenario assumption is that the loading was applied as a pure shear load over the minor area in the threaded portion of the bolt. SAE Grade 8 1/2-20 bolts were selected for this joint. A fine thread series and Grade 8 rating were selected for maximum strength. Figure 18 shows a τ s = τ s = F s A s 100 lb f = 673 psi.1486 in2 The resultant shear force of 673 psi is very small. For calculating the resultant factor of safety Distortion-Energy Theory was used to predict the shear yield strength based on the known tensile yield strength. That relationship is that the shear yield strength is approximately equal to.577 times the tensile yield strength. For a Grade 8 bolt the minimum yield strength in tension is 130 ksi. Now 14

15 Figure 15: FEA analysis for compression case. 15

16 Figure 16: FEA analysis for tension case. 16

17 Table 5: Strength properties of SAE Grade 8 bolts. Property Minimum Allowable Value Proof Strength (S p ) Yield Strength (S y ) Tensile Strength (S t ) 120 ksi 130 ksi 150 ksi the resultant factor of safety is calculated for the bolt to be 111. S y,s.577s y,t S y,s = ksi = 75.0 ksi n = n = S y,s τ s 75.0 ksi 673 psi = Load Cell Connection to Rocket Assembly that is less than the proof strength of the bolt. The load factor of an externally loaded bolt is generally calculated as the ratio of the proof strength of the bolt to the stress state created in the bolt as the result of an external load. In a bolted connection the net stress state in the bolt is made up of several different factors. Those factors are the external load applied to the bolted connection (P), the fraction of the external load carried by the bolt (C), and the preload force developed in the bolted connection (F i ). Our postulation that the entire load could be carried in tension on the bolt area is a possible situation because this could happen if the bolt was not tightened with a preload. In this case F i =0, by definition, and C=1 because the entire external load would be carried by the bolt. C can only be less than one in cases where and elastic reaction force is developed in the members as a result of the clamping reaction of the bolt. Taking into account the parameters of our design case, the general equation for the load factor of a bolt can be simplified. The connection between the rocket assembly and the vertical load cell needs to be analyzed because only the load cell only accepts one 1/2 bolt (location shown in figure 19 on the following page). Therefore there is only one point of connection ultimately holding the rocket up. In light of this fact the bolt and threads of this connection were heavily analyzed. The load cell is threaded to accept a 1/2-20 bolt as the member that actually connects the load to be measured to the load cell. Therefore, the size of the bolt is a driven quantity. One design choice that we could make was to specify a SAE Grade 8 bolt for this connection. This will provide maximum possible strength. Table 5 provides a neat summary of the minimum strength values, for Grade 8 bolts, used in our calculations. 8.1 Tension and Compression in Bolt For failure analysis purposes the bolt can be analyzed as it is carrying the net load entirely in tension or compression on the standard tensile stress area (A t ) of the bolt. From a stress analysis point of view, one of the design targets in a bolted connection is to have maximum stress state in the bolt n = S pa t F i CP n = S pa t P Our design case assumes an external load (P) of 200 lb f and the standard tensile stress area (A t ) of a 1/2-20 bolt is.1599 in 2. The final calculation for the load factor equals 95. n = n = S pa t P 120 ksi.1599 in2 200 lb f = Shear of Threads The other possible area of failure is if the internal threads of the hole that the bolt going through the load cell shear off (strip). The bolt itself was selected to be SAE Grade 8, and the shaft material that the bolt is going to thread into is going to be annealed AISI 304 stainless steel. Of the two thread materials, the stainless steel has the lower tensile yield strength. The tensile yield strength (S y,s ) of AISI 304 stainless steel is 40.0 ksi. Again using the Distortion-Energy Theory, the shear yield strength (S s,y ) of AISI 304 stainless steel was approximately calculated to be 23.1 ksi. S y,s.577s y,t S y,s = ksi = 23.1 ksi The other piece information needed for developing an analytical expression for the shear on the 17

18 Figure 19: Bolt connecting load cell to rocket assembly. thread(s) is the area that the shear force acts upon. This area is a function of the thread pitch (p), thread root diameter (d r ), and percentage of engagement between the internal and external threads (w i ). For a UNF 1/2 thread the thread pitch is.05 in, the root diameter is.435 in, and the percentage of engagement is 80%. Using an expression for the circumference of the thread and the known values previously stated, the area that the shear force acts on for a single thread (A s ) is calculated to be.055 in 2. A s = πd r w i p A s = π.435 in in =.055 in 2 If the maximum permissible shear stress is set equal to the shear yield strength and the shear stress equation is resolved for the external load (F) an expression for the maximum force (F max ) that the thread(s) will support is derived. n t is the number of engaged threads in the joint. τ max = F A s F max = S y,s n t A s Two cases were explored in our analysis. The first case assumes that the entire load is carried by a single thread and and the second case assumes that 1 in of tapped hole or 20 threads carry the load One Thread Carries Load F max = S y,s n t A s F max = 23.1 ksi in 2 = 1270 lb f Using the design load (F) and the maximum load (F max ) an expression and value for the factor of safety in this case is found to be 6.3. n = F max F n = 1270 lb f 200 lb f = Threads Carry Load The calculated factor of safety in this case is 127. F max = S y,s n t A s F max = 23.1 ksi in 2 = lb f n = F max F n = lb f 200 lb f =

19 9 Frame The vertical test stand design utilizes a frame to hang the rocket from and attach the sensors to. A preliminary CAD model of the vertical was designed in order to show the practicality of the concepts. All analysis done on the frame assumes that a very weak structural steel was used in the construction of the frame. Using this approach allows the use of stronger materials for increased factors of safety if necessary later on. The steel type for the calculations is AISI 1040 hot rolled steel. The specified minimum yield strength of this steel is 42 ksi. Figure 21: Force applied to beam for stress computation. 9.1 Stress in Frame Uprights Three 6 X 25 lb f /ft wide flange I beams were selected to be the uprights for the frame. The design proposal assumes that the two tallest uprights are roughly 9 feet high. With regards to analysis, the main area of concern with the frame is that people or equipment could lean up against the frame. Forces applied perpendicular to the web of the upright will result in higher bending stresses in the beam than forces applied along the web of the beam. Analysis will focus on a force applied perpendicular to the web because the stress will be higher. For the design case, a 400 lb f force was applied, as shown in figure 21, to the top of the tall upright. The maximum bending stress was then found at the base of the beam. M is the resultant bending moment for a force, F, applied at a distance, d, from the plane of analysis. Since the beam selected is a standard shape values for c max and I were obtained from tables. c max =3.040 in and I=17.1 in 4. M = F d M = 200 lb f 20 in = 4000 lb f in σ max = Mc max I σ max = 4000 lb f in in 17.1 in 4 =.711 ksi A factor of safety against yielding can now be computed based on the yield strength of the steel and maximum stress just calculated. n = S y σ max Figure 22: Detail of plate welded to frame upright. n = 42 ksi.711 ksi = Weld at Base of Frame Uprights To secure the frame uprights to the ground, it is desired to weld square plates to the bottom of the frame uprights (see figure 22). Since this weld is in a comparatively high stress area, compared to the rest of the frame, it is desired to use a butt weld with complete joint penetration. This will take more effort to construct, but will yield a weld with strength comparable to that of the base metals. According to AISC (American Institute of Steel Construction) guidelines, weld joints subjected to bending stresses should conservatively have a maximum stress at the weld that is no greater than 19

20 Figure 20: CAD model of test stand design with frame highlighted. 60% of the minimum yield strength of the metal(s) in the weld joint. If materials of several different yield strengths are present in the weld then the lowest yield strength among the individual materials should be used. Most common filler metals for steel have higher yield strengths than than the minimum yield strength of the steel used in the frame design, so the frame material is the limiting factor. S y,weld =.60 S y,min S y,weld = ksi = 25.2 ksi Since complete joint penetration is being used in the welds the area of the weld area is at least as large as the cross section of the beam. Since there is no area reduction, the maximum stress calculated in section 9.1 on the preceding page is representative of the stress in the weld. With this a factor of safety can be calculated for the weld. n = n = S y,weld σ max 25.2 ksi.711 ksi = 35 Part III Data Acquisition Analysis 10 Load Cell Amplifier Upon taking over the data acquisition system for this project, the given task was to redesign the original amplifier circuit for the load cell to make it more effective and accurate and interchangeable between the horizontal and vertical test stand setups. The original design of the amplifier circuit consisted of a single TLV2783 amplifier chip that contained two separate operational amplifiers. The circuit was designed for a gain of 100. Since the load cell outputs readings of approximately 3mV/V or a maximum voltage of 30mV based on a 10V range, a gain of 100 would only show a maximum of 3V at the output of the A/D converter. Using this information, it was determined that a gain of approximately 333 was needed to fully utilize the 0 to 10V range that the A/D converters are capable of. Gain = Gain = V out V in 10 V 0.03 V =

21 10.1 First Design The first redesign of the amplifier included a TLV2783 chip and two LM741 operational amplifier chips set up in a non-inverting op-amp configuration as shown in figure (1) 23 on the next page. The total gain of 333 was split across 4 separate amplifier stages. The gain equation for a non-inverting op-amp, was used in designing each stage of the circuit in figure 23 on the following page. Gain = V out V in = 1 + R f R in To power the TLV2783 chips, two 1.5V batteries were connected in series for a total of 3V. A total of four 9V batteries connected in series to produce a positive 18V source and a negative 18V source, which were then sent through a negative regulator, LM320T-15, and a positive regulator, LM340T-15, to produce + 15V, were used to power the LM741 chips. This design was not successful, partially due to extreme weather conditions that were below the operating temperature of the chips. Another problem that was encountered was the consistent failure of the TLV2783 chips during test firing of the rockets. Therefore, this design was discarded in favor of a more reliable amplifier circuit and research was performed to find amplifier chips and regulators that could withstand extreme weather conditions Final Design The final design of the load cell amplifier, shown in figure 24 on page 23, consisted of a single precision instrumentation amplifier, INA114, which contains 3 internal op-amps. A single resistor, R G, was the only component value that needed to be calculated during the design of the amplifier circuit. To calculate the resistance, R G, needed to produce the desired gain the following equation was used: Gain = V out V in = kΩ R G Two battery packs, comprised of two 9V batteries each, were made to supply +18V to a positive 12V regulator, AN78L12, and -18V to a negative 12V regulator, AN79L12, which powered the INA114 chip with +12V. A PCB layout was designed to accommodate a total of eight INA114 chips on a single board for use with both the horizontal and vertical test stands, since all seven load cells output readings of 3mV/V or a maximum of 30mV. The PCB utilizes terminal blocks for the inputs to the INA114 chips from the load cells and from the outputs of the INA114 chips to the A/D converters to reduce the possibility of error due to poor wire connections. This design has been successfully used during test firings of the rocket and has been incorporated into the final design of the data acquisition system. 11 DAQ Housing Box The main purpose of the box was to reduce the amount time needed to setup the data acquisition system and create a centralized area to collect and transmit data from the sensors to the computers. A concept drawing of the box is shown below in figure Connector Choice The first step in designing the physical data acquisition (DAQ) box was to decide what type of connectors to use when interfacing the various sensors with the circuitry contained within the box. The Mate-N-Lok? plugs and caps appeared to be the most practical, since the connectors were easy to plug in and did not have a tendency to fall out. Either three circuit or four circuit connectors were used depending on the sensor s signal and power configuration. The pressure sensors, in which there are three total, required the three circuit connector; (+) signal and (+/-) power. The six load cells and three infrared temperature sensors required the four circuit connector; (+/-) signal and (+/-) power. The plugs were fitted and secured into.25" milled slots on the faceplate of the box arranged in five columns with three plugs in each column and the caps were attached to the ends of each sensor Power Supply The next step of the design process was to determine how to power all of the sensors and amplifiers. The original setup included multiple sets of two 9V batteries, which were connected in series to produce 18V. This method was determined to be inefficient for a few reasons. First, the load cells could not handle an 18 V excitation so the voltage 21

22 Figure 23: First design of amplifier circuit. 22

23 Figure 24: Final load cell amplifier circuit. 23

24 Figure 25: DAQ housing box. 24

25 would have had to been regulated down to 12V, leading to more circuitry and less space. Second, the number of 9V batteries that would have been required to power all the sensors would have been very high due to their low power output. Third, since 9V batteries do not last very long they would have had to been replaced frequently which means spending more money and risking them going dead in the middle of testing. As a result of the fore mentioned reasons two sets of two 6V batteries connected in series, producing 12V, were designed to power all the sensors Analog to Digital Converters The third design step involved improving the way data was transferred from the sensors to the computers. Since the sensors output analog signals and the computers are only able read digital signals multiple analog to digital (A/D) converters were used to covert the data from all the sensors. These A/D converters were used in past testing however this time they were placed in a box and not lying on the ground where they could have easily been damaged. Three RS-232 connector shaped holes were punched through the back of the box and the converters were bolted down to allow easy connection Thermocouple Inputs Per request of one of the other M.E.T.E.O.R. teams a thermocouple data acquisition system was added to the main data acquisition box. The system was purchased from DATAQ R and allows for four thermocouple inputs. The decision to purchase the system, as opposed to building one, is that this all in one system provides a higher degree of accuracy, resolution, and durability than anything that could have been built. The thermocouple system was secured in the box by two custom built brackets & spacers, and a slot was milled in the faceplate so that the inputs could be connected easily. The output of the thermocouple couple system exits via RS-422 and then is converted into RS-232. The converter was placed on the back of the box for simplicity and the lack of space inside the box Load Cell Amplifier The before mentioned load cell amplifier was placed inside a small box with two openings for the inputs of the load cells and outputs of the amplifiers. Figure 26: Vertical thrust force sensor. Part IV Force Transducer Selection and Calibration 12 Transducer Selection 12.1 Vertical Thrust Force For the vertical thrust measurements a tension and compression 250lbf S-Type load cell was used. The load cell shaft is guided into the load cell, shown in figure 26, using linear bearings so that only axial force is measured. The reason for using a tension compression load cell is that it can measure the dead weight of the rocket when it is in tension then it will read compression values when the rocket is fired. A certified bolt holds the S type load cell in place and secures it to the channel frame Lateral Force For the lateral force measurements mini-beam compression load cells were selected. As shown in figure 27 on the next page, they were placed in a cruciform pattern at 90 degrees to one another perpendicular to the thrust axis. The lateral force is transferred from the rocket to the individual minibeam load cells via threaded rods with tapped brass 25

26 Figure 27: Lateral force sensor. 13 Transducer Calibration It is essential to calibrate all six load cells before testing to verify the amplification to obtain proper force readings Vertical Thrust Force balls on the end. The brass balls allow the rocket to spin and move vertically freely while still transferring the lateral force when impacted. The minibeam load cells have a maximum output of 25lbf which will allow minute lateral thrust values to be read. In addition the lateral force support frame is attached to the channel uprights with bolted fasteners which allow the frame to be moved up or down for variable length rockets. The threaded brass ball connections can be threaded in or out depending on the diameter of the rocket Roll Moment A load cell was installed with an attached torque rod to obtain the torque readings if the rocket were to roll about its thrust axis during firing. This assembly can be seen in figure 28 on the following page. The load cell utilized is a 25lbf tension and compression S type load cell. The load cell is attached to the universal joint via threaded rod and two ball joints. The ball joints allow the connection to deflect under vertical force but still transfer the torque of the roll. As the rocket rotates the torque can be obtained by simply multiplying the length of the lever arm times the force read by the tensioncompression load cell. The load cell is attached to the upper yoke of the universal joint which only rotates and moves vertically. For the vertical force sensor the calibration is simply done through weights and a connection cable. A threaded eye bolt connection can be connected to the bolted connection for the nose cone coupling of which a steel cable can be attached via spring connecting rings. An aluminum platform is connected to the opposite end of the cable which calibration weights can be placed on to calibrate the load cell. The weights can be loaded on in 50lb increments to determine the amplification and linearity of the load cell output Lateral Force The calibration of the lateral force sensors is done is a similar manner to that of the vertical force sensor. A cable is threaded through the load cell and attached pulley of which there is a weight support on the end. The end of the cable threaded through the load cell is then clamped with nuts to offset the clamp from the load cell. The weight support can then be loaded with 1lb calibration weights to obtain the amplification and linearity of the load cell output Roll Moment- Torsional Force Sensor A special torsional force sensor calibration frame will have to be attached to the torsion cell anchor. The universal joint will also have to be removed to allow for clearance of the torsion cell calibration frame. The torsion cell calibration frame contains a cable and pulley that can be threaded into the load call and support the weights for calibration. 26

27 Figure 28: Roll moment load cell. Part V Risk Assessment 14 Risk Management Philosophy Current and future rockets constructed by the ME- TEOR project are all experimental devices. Sound judgment and scientific knowledge is applied during the design of the rockets, however, there is still an uncertainty in knowing what exactly is going to happen when a rocket engine fires. In light of this fact, our team has come up with a plan that lists the risks present and then describes the measures being taken to mitigate the risks. It should be noted that the safety analysis done by the vertical test stand team builds on previous safety analysis done for the METEOR project. These shared safety measures provide proven strategies for dealing with the hazards of rocket testing; which in turn brings on less new risk. The basic mentality that our team has taken in developing the plan to manage the uncertainty in rocket testing, is to have engineering and physical barriers in place to help contain the hazards. From an engineering standpoint safety was built into the design and excess strength exists in the structural design. Physical barriers are also present to help contain any uncontrolled occurrences that may occur during a test. The first of those is a steel tube that surrounds the rocket, shown in figure 29 on the next page, to help minimize property damage inside the concrete building that testing is done within. The concrete building itself is another safety measure. This structure was put in place by a previous METEOR team to help contain the rocket during testing. Our test stand holds the rocket in the vertical orientation with the thrust forcing the rocket upward. As a safety measure there is only a 2 inch diameter hole in the top of the concrete bunker. As shown in figure 30 on the following page, this hole allows the 1 inch shaft that supports the rocket to connect to the load cell, but will not allow the rocket to come out the top of the concrete bunker. 27

28 Figure 29: Steel tube surrounding the rocket (concrete bunker not shown for clarity). 15 Standards for the Ranking of Hazardous Conditions Hazardous conditions are categorized based on the severity of the outcome experience for a particular condition. The framework shown in table 6 on the next page gives the standardized categories, and the consequence criteria for each category. 16 Standards for the Probability of a Hazardous Condition Occurring The probability of each hazardous condition occurring is also rated according to a standardized scale. That scale is described in table 7 on the following page. 17 Hazards in the Design Figure 30: Hole in the top of the concrete bunker. The hazards for this project were collected and are explored in detail in the following subsections. To help with the ranking of the hazards, they were separated into groups based on their probability of occurrence. Each subsection contains a different probability of occurrence Frequent Hazards See table 8 on the next page Probable Hazards None 17.3 Occasional Hazards See table 9 on the following page Remote Hazards See table 10 on page