See presentation notes in Planning & Execution
Risk AssessmentFor a full list of all the risk assessment items, see the Risk Assessment document.
The high risk items are highlighted in the image below.
Prototyping, Engineering Analysis, Simulation
Glider Dynamics Analysis
Based on the flight orientation from the image below, a 3-D simulation was created in MATLAB using a simple force balance approach.
- Side slip negligible
- Glider is a point mass
- Aerodynamic forces calculated based on 3-D glider
- Tether is rigid and mass-less
- Constant flight conditions
- Glider is orthogonal to the tether line
- Wind is always horizontal (-j and -k vectors)
- No Tether Drag
Force BalanceDRAG: Drag is always in the same direction as the apparent velocity.
LIFT: Lift perpendicular to the drag and wing span
WEIGHT: Weight is pulling the glider down (-j and k vectors)
TENSION: Tension equation is determined from a force balance with Drag, Lift, Weight, and the summed forces (mass times acceleration).
The model is set up to allow user input of the following critical variables:
- Wind Speed
- Glider size
- Glider configuration
It outputs the magnitude of the tension and position information of the glider. The magnitude of the tension and instantaneous velocity of the kite as it goes through its flight are plotted. These gave us data that was easily visualized and critically analyzed.
The following are plots of the glider flying at varying wind speeds. The flight configuration of these simulations are as follows:
- Beta angle = 92 [deg]
- Tether length = 30 [m]
- Flight path radius = 12.5 [m]
A major constraint in our system is the load cell. We can at most afford a 500kg load cell. Because of this, we cannot exceed 5000 N tension force without damaging the load cell. As shown in these plots, a wind speed of 10 m/s is right around this limit. Thus, we can conclude that we have to conduct the experiments in winds less than 10 m/s.
Another key point to note is the instantaneous velocity of the kite, plotted on the right side of each picture. These speeds are very high so we will have to develop a detailed and complete flying procedure and safety measures.
Tether Drag Calculations
In order to help verify our model, we need to determine if our assumption of no tether drag is accurate. This was done by using a model created by Rajani et al. which calculates tether drag, tension, and tether angle numerically at various intervals on the tether. Using our flight conditions and 100 steps, the total tether drag is 27.063 Newtons. As this is significantly lower than the other forces acting on the glider, the tether drag is assumed to be negligible. This implies that our assumption is correct.
The tether angle and tether tension was also plotted against the tether length. The figures below show the results. Tether tension does not change significantly over the length of the tether so the readings obtained by the load cell will be assumed to be accurate. However, the change in angle in the tether is significant. Because the tether length is fairly long, a small change in angle can cause large inaccuracies in the glider position measurements. The data acquisition code will be required to account for the change in tether angle.
Tether Elongation Calculations
Because there is such a large load on the tether, the tether may extend a significant amount. This can be calculated very simply, using standard tensile stress and stress-strain equations. Using the properties in the table below and those two equation, the strain was found to be 0.01558. The tether will extend 0.4674 meters for a 30 meters long tether.
|Modulus of Elasticity||113.5||GPa|
The wind speed, beta angle, flight path radius, and tether length have been identified as having the largest affect on the tether tension (aside from the glider size which has already been set to the Bixler's parameters). A model sensitivity analysis was conducted and yielded the following results:
High Level Plan Develop a quantitative analysis of the sensitivity of the factors in relation to the response. Using Minitab a regression equation was formulated in order to determine the sensitivity of each factor. This equation takes into consideration the sensitivity of each factor in relation to tension as a response.
Force = 55.4 + 2.89 WindSpeed - 1.86 Radius - 0.169 Beta + 1.11 TethLen
A regression equation was developed and reiterated to remove unusual observations and close the gap of the prediction interval. The issue with this regression was that there is no constant mean of zero (Figure 2); hence the r-squared value is extremely low. However, the significance of each factor can be checked when P stat is compared at an alpha of 0.05. This comparison leads to Wind speed, Flight Path Radius and Tether Length being highly significant. We concluded to important outcomes, which were that these factors are very sensitive to tension and would still consider the Beta angle sensitive because it is a factor when flying and testing the glider. The following equation displays the sensitivity of each factor in relation to response
Tether and Bridle Design
The tether needs to be attached to the glider in a way that won't break the plane apart and and allows for an adjustable beta angle. The EPO foam from which the Bixler glider is made has a maximum allowable stress of 30 MPa.
We created displacement, shear, and stress diagrams in MATLAB to identify the forces acting on the glider in flight. Using that we cannot exceed 5000 N of force, we applied a factor of safety of 2.5 and input that the maximum force the glider will feel is 12500 N. This factor of safety accounts for the possibility of non-uniformity of the EPO foam, wind gusts, and other jarring encounters.
Two Point Bridle Analysis
This configuration consisted of a bridle point attached on the fuselage between both wings and the second bridle attachment point was on the rear of the fuselage. The displacement, shear, and stress on the wing are plotted below.
This showed that, with a bridle point attached on the fuselage between both wings, the maximum stress is 180 MPa, 6 times more than the allowable stress and the glider would break apart from tension at the wing roots. A two point bridle design will not work.
Three Point Bridle Analysis
The three point bridle design consisted of one connection point on the rear of the fuselage and one connection point on each wing. We varied the location of the attachment point on the wing from the wing root to the wing tip to find the best location for the connection point. We also varied the angle at which the tether makes with the span of the wing as it angles towards the fuselage. This analysis produced the following contour plots:
From these plots we needed to find the location of the minimum stress. The minimums identify the ideal location for the bridle attachments on the wing as 0.51 meters from the wing root on each wing at an angle of 54 degrees.
The following plots are the shear, stress, bending, and displacement plots for the wing using the ideal bridle attachment configuration:
The maximum stress from this is 15 MPa, which is half of the maximum allowable stress for EPO foam. This proves the wings will not break apart from the tether forces.
Fuselage Attachment Analysis
The analysis for the fuselage was conducted in the same manner as the Two Point Bridle Analysis section. The displacement, shear, and stress on the wing are plotted below.
The maximum stress from this is under 20 MPa, thus less than two thirds of the maximum allowable stress for EPO foam. This proves fuselage will not break apart from the tether forces.
Base Station Analysis
After applying Pugh analysis to our initial concepts, we selected the following type of system for our base station. The figure below depicts a rough sketch of the idea.
Initially, a single detailed design concept was proposed, and this concept can be seen in the CAD model shown below.
It was determined that the load path for the lower portion of the base station in this design was overcomplicated, so a second iteration of the design process was conducted. After reviewing the functions that the lower portion needed to complete (transmit torque and absorb thrust), several additional concepts were brainstormed. A summary of each is given below.
Concept 2: One thrust bearing to absorb thrust, two pillow block radial ball bearings to absorb any radial load, and support shaft. Design is similar to a propeller shaft assembly. Top portion same as initial concept. A picture of the concept is shown below.
Concept 3: Single thrust ball bearing for the vertical shaft. Top plate uses a rail guide to allow the top plate to spin. A drawing of the concept is shown below.
Concept 4: Two mounted thrust ball bearings for the vertical shaft. Top portion same as initial concept. A picture of the concept is shown below.
Concept 5: Two tapered roller bearings mounted which withstand both thrust and radial loads. Two custom mounted bearings on the top plate to allow the horizontal shaft to rotate freely. Shaft collar with threaded rod welded to it which connects load cell to the horizontal shaft, without drilling a through hole. A picture of the concept is shown below.
After brainstorming the additional concepts, they were compared against each other in order to select our final one. Pugh analysis was utilized as the tool to do this. The following shows the first iteration of Pugh analysis, where the datum was our initial concept.
After completing the analysis, not much was gained as 3 of the new concepts seemed to tie, at least when compared against the initial concept and the selected criteria. That being said, a second iteration of Pugh analysis was completed, this time using Concept 4 as the datum. The second iteration can be seen in the following image.
Final Concept Selection
After going through the second round of Pugh analysis, it was determined that Concept 5, shown below, was the best design due to it's simplicity, and this was the design that we decided to move forward with.
Again, this design used two tapered roller bearings mounted vertically to absorb any radial and axial loading, and two radial ball bearings mounted horizontally on the top portion. Both the vertical and horizontal shaft were coupled via flexible couplings to potentiometers in order to track the vertical and horizontal sweep of the glider. A load cell is connected to the horizontal shaft and at the tip of this load cell would be the connection point for the tether. Some feasibility calculations as well as some calculations involved in component selection are summarized in the following sections.
One of our engineering requirements states that the angular position resolution must be minimally 0.5 degree. As a result, we wanted to see what kind of tension values would be necessary to get the system moving, while still mantaining this resolution.
The following depicts the free body diagram of the forces acting on the load cell starting to rotate about the z axis (out of the page)
In order to calculate the minimum angle that the tether can deviate from the line along the load cell, we used our resolution requirement. The following was the basis for these calculations
After applying a force and moment balance, the tension was solved for as a function of the ground angle, and this is plotted below. This is the necessary tension to get the system rotating vertically while still preserving the resolution of 0.5 degrees.
The following depicts the free body diagram of the forces acting on the load cell starting to rotate about the y axis (into the page)
The resolution considerations were conducted in the same manner as the previous vertical case. The following is the Matlab plot for the tension required to get the system rotating horizontally (parallel to the ground).
Main Design Goals
- Wanted to minimize flexing of the top shaft, in order to prevent seizure of pillow block bearings
- Wanted to ensure tapered roller bearings were sized to absorb necessary radial and thrust loading
Main Component Selection
As previously mentioned, a shaft that flexes minimally, yet has a small moment of inertia about the center axis is desirable for our application. After resolving the forces on the shaft and doing a shaft deflection and bending stress analysis for various lengths and diameters, a 1/2" x 4" inch shaft was chosen. The free body diagram can be seen below
For the vertical shaft, after analysis, a 3/4" diameter shaft was chosen.
As mentioned previously, two tapered roller bearings were mounted vertically, to allow the base station to spin freely while absorbing radial and thrust loading. The following calculation of the forces on the tapered roller bearings under worst case scenario loading provided the basis for our choice for these bearings. Following the calculation is a Figure showing the choice of bearings, as shown on McMaster-Carr.
Radial Ball Bearings
Two radial ball bearings were employed to allow the horizontal shaft to rotate freely. As these bearings only needed to withstand radial loading, two ball bearings were chosen. The reaction forces from the shaft calculation (shown above) were used to size these bearings. The following Figure shows a picture of the bearings, as shown on McMaster-Carr.
The pillow blocks to mount both the radial and tapered roller bearings will be constructed out of aluminum. The following figure depicts the CAD model of the pillow blocks for the horizontal and vertical shaft bearings respectively.
Drawings, Schematics, Flow Charts, Simulations
Tether and Bridle Design
Based on the analysis from section 1.2 above, the following is the tether and bridle configuration.
Base Station - Final Design
The final design that we expect to build is as follows:
Base Station Custom Part Drawings
|Angled Bar Anchors||
|Dead man Anchor||
The external weights method was discarded because it is impractical to carry large weights around every time we want to fly. Also, the weights would have to be positioned directly on the base station which would take up a lot of space and could possibly get in the way of the tether's path.
While the dead man anchor would supply sufficient support for the base station, it is a permanent structure that we could not move. This would greatly limit us as we would need to chose a single spot to fly at. This could cause issues if the wind is flowing in a bad direction. Also, it requires a six foot deep hole to be dug which would be hard when the ground is frozen in the winter.
The two remaining ideas are an auger and a duckbill earth anchor, which is a combination of the angled bars and the dead man anchor. Both of these methods are semi-permanent and if the correct item was purchased, could support the required load. The auger was chosen over the duckbill earth anchor, because the duckbill earth anchor is more expensive and also requires a drive steel tool to install.
The auger chosen is a 4 inch auger that has a 30 inch shaft. It is usually used in high tensile applications such as fences that keeps large animals contained. It is rated at 2000 pounds vertical pull out. One of these is sufficient to keep our base station anchored to the ground as long as the site chosen has normal soil.
The site that was chosen to fly our glider at is RIT's material storage field, located on Miller Road, just off Andrews Memorial Drive at RIT. There is a large isolated field that would be perfect to fly at because there is a low risk of the glider being flown near any pedestrians. Also, there is a mound roughly 20 feet high, which could be used to anchor our base station. This would allow the glider to fly lower than the base station, reducing the chance of collision with the ground. The soil was tested by using a thin metal rod which was pressed into the ground to determine how densely packed the earth was. The soil appeared to be alfisol, which is a common type of soil that is known to exist near Rochester. Because the field is composed of alfisol and alfisol is a common type of soil, we believe that the 2000 pound rating of our chosen anchor is accurate and would be sufficient to anchor our base station.
Below is the connectivity from the load cell and potentiometers to the DAQ devices (Bridge Amplifier and NI USB-6210) to the laptop for LabVIEW processing and saving.
The following sections display the individual components along with the technical specs. The first DAQ system called for the S Type Load Cell (100-500kg), based on the theoretical model results. After initial testing began and budget constraints, the Micro Load Cell (0-50kg) was chosen. The original amplifier was the Phidget Bridge, however this was changed for the Logos Electromechanical Bridge Amplifier v2. This change was made because of the complexity in programing for the Phidget Bridge. Initial testing of the DAQ system was done using the NI USB-6008 obtained from Professor John Wellin. This was suggested by our customer in case, while learning, something was setup that fry the DAQ device , the NI USB-6008 model is the less expensive one to have to replace.
Bill of Material (BOM)
Total budget: $500
For the "Detailed" version of the BOM see the Bill of Materials Spreadsheet.
Note: BOM only includes shipping costs for the gliders.
The test procedure will require refinement especially concerning getting the glider in and out of the flight path. The current version shows detailed setup and disassemble procedures as well as our first iteration in brainstorming a procedure for entering and exiting the flight path.
Testing the engineering requirements will be executed over various stages throughout MSD II. See the preliminary outline and timetable below.
|Component/System Tested||Specification Tested||Responsibility||Completion Date|
|Experimental Proof of Theoretical Model||Tension||Team||02/28/2014|
|Varied Tether Length||Model Sensitivity||Team||03/14/2014|
|Varied Wind Speed||Model Sensitivity||Team||03/28/2014|
|Varied Beta Angle||Model Sensitivity||Team||04/04/2014|
|Varied Flight Path Radius||Model Sensitivity||Team||04/11/2014|
|Varied Glider Mass||Tension||Team||04/18/2014|
- The sequence of tests is subject to change based on environmental conditions. Varied wind speed testing is the test in question as the outside wind speed cannot be controlled. This test would be conducted on an opportunistic basis.
- The varied glider mass would be an additional test if time allows. This would expand the versatility of the theoretical model.
- Total testing of the theoretical model would require different gliders of different flight dimensions and flight characteristics. This is beyond the scope and resources of this project.
Last Updated - 18 November 2013