hanskeh-blog

[hanskeh-blog]

Week 10

This week was the last week of AERO 307! Alex, Sam, and I presented our findings on the project Mobius Solar Plane winglets. As was said last week the data obtained from this experiment did prove that the addition of a rounded wingtip would increase the lift to drag ratio of the wing by a significant amount. However, the addition of the hoerner wingtip and the endplate greatly reduced the efficiency of the wing. This is unlike what is expected for both winglet types and the error in these results may be attributed to the manufacturing technique used to make the winglets. Yet, even with this error, the rounded wingtips performed far better than the other wingtip designs. From this data it was determined that it would be beneficial to use a rounded wingtip on the wing because it would increase the lift to drag ratio of the wing by reducing the drag caused by the wingtip vortex and result in a more efficient wing. The wingtips will be made from foam due to its light weight and durability. All other manufacturing methods are either too expensive or result in a part that is too heavy.

Now that my aerodynamics courses are over, I am happy to say that it is hard to pinpoint exactly one thing I feel I understand most about aerodynamics. This being because one aerodynamic thing can depend on so many other aerodynamic things. If you don’t understand one of those other things very well, it is hard to say you understand the one thing a lot. With that being said, after 307 and 302 I feel like there isn’t really much in aerodynamics which confuses me. There are, without a doubt, some aerodynamic phenomena which I think are fascinating and strange such as the Von Karman vortex street (so cool) and the boundary layer. But lift is by far one of the strangest and most AWESOME phenomena which has so many different theories. The unknown about aerodynamics is one of the many things which makes it so interesting.

AERO 307 has been a pretty awesome experience this quarter. It was great to be able to come up with our own tests which we could preform in the wind tunnel. So much of the class time was spent either in the wind tunnel preforming a test, or somewhere else preparing for the test. This much hands on time and time in the wind tunnel is super rare for an undergrad and I found it to be such a great learning opportunity.

I thoroughly enjoyed AERO 307 this quarter and it has been great learning more about aerodynamics and testing. The only “lowlight” I can think of is when testing just didn’t go as planned. However, this is bound to happen, and it is most likely one of the best learning experiences you can have. When something doesn’t go as planned you have the opportunity to help solve the problem or come up with another solution which will enable you to continue with testing. It keeps you on your feet.

I am the type of engineer that wants to be physically working with something. During my two previous internships I worked as a test engineer. I enjoyed test engineering because I was able to work directly with the product; however, the only thing that was missing was that it wasn’t an airplane I was working on. I’m interested in avionics, but I am also interested in general aircraft maintenance and I hope to have my private pilot’s license within the next few years when it becomes affordable. I have been trying to do research on what kind of positions offer this kind of engineering experience, but it has been difficult finding any information on this kind of work online. I have found that the best way to learn more about this kind of work is by speaking with certified mechanics and other people working in the aircraft maintenance field. If I could find a career where I am working with and on aircraft every day this would be the dream career I have for the future. AERO 307 has provided me with so much valuable hands on experience which will be useful when I begin me career.

Thank you so much Tynan and Dr Doig for such an awesome quarter.

Hans

[hanskeh-blog]

Week 9

This week in lab Alex, Sam, and I finished forming the winglets which were to be used for testing. The winglets were made from blue foam which was carefully sanded to the desired shape. We also mounted the wing to the mounting plate we had cut the week prior. The wing was attached to the plate by making two holes in the plate and wing, filling these holes with an expanding foam glue, and then placing a screw in the holes. This resulted in a very strong bond between the plate and the wing. However, after doing this we realized that we would need to run a test with the plate only to subtract this drag out for all the other tests with the wing. In order to get this set of data, we decided that after completing the testing with the wing and winglets, we would carefully cut the wing off of the plate and test with only the mounting plate.

Unfortunately, we were unable to vary the angle of attack during testing due to the way our wing was being mounted. However, this mounting method proved to result in acceptable drag data. Because we could not vary the angle of attack, we decided to instead vary the speed that the wing was tested at.  This would result in a different lift and drag force but would result in the same L/D, coefficient of lift, and coefficient of drag. The wing was tested in the wind tunnel with each winglet at 5m/s, 6m/s, 7m/s, 8m/s, 9m/s, and 10m/s. A video of the wingtip vortex was taken for each winglet at 7m/s using smoke directed onto the tip of the wing and a remotely controlled camera. During testing we also noticed that, unlike the previous winglet test for flow visualization, we needed to maintain the same reference area for each wing/winglet setup. Each winglet was made such that each had the same span; however, to maintain the reference area while performing a no winglet test, we needed to increase the span of the wing. Because the wing was attached to the plate and could not be removed we attached a small wing section which was the same span as the winglets.

When analyzing the data, the -x-direction (port side) was the lift and the y-direction (with the flow) was the drag. The test with only the plate produced a lifting force which had an average near zero and a large standard deviation, so this force was not subtracted out of the data. The drag force from the plate was subtracted out of the data. After the analysis, it was found that the endplate and the hoerner wingtip produced very similar L/D results. However, the endplate did produce the most lift out of all the winglets, but it also produced a large amount of drag. The rounded wing tip produced a very minimal amount of drag but produced a lifting force identical to that of the hoerner wing tip and had a very large L/D. The negative drag seen in the rounded winglet plot may be because most of the drag is coming from the plate and not actually the wing, so when the plate drag was subtracted out, it brought the drag near zero. Below are some of the plots from this analysis.

Next week we will meet to discuss the data and prepare for the presentation of our findings on Wednesday.

Thank you for reading!

[hanskeh-blog]

Week 8

This week our group for the Aspect Ratio lab worked on completing the lab report. I finished preforming the data analysis for the report early last week. The data from the analysis was very interesting. In the plot below, it is clear that the high aspect ratio wing had a positive impact on the coefficient of lift as would be expected. It can also be seen that the high aspect ratio wing stall at about 12 degrees while the low aspect ratio wing staled at 14 degrees. The decrease in the stall angle was another expected result of the high aspect ratio wing.

The drag coefficients are greater than a theoretical drag coefficient by nearly an order of magnitude. So, it was hard to make any definitive conclusions about the drag results obtained from the experiment. It is currently unclear where this additional error in the drag is coming from. The error could be due to the flow around the wing interacting with the flow around the force balance hub. This could be potentially increasing the force that the device reads. This poor data also effected the validity of the experimental lift to drag ratio.

This week in lab we also worked a lot on our mini projects. Ours of which is the winglet design for the mobius solar plane. The original plan we had for testing was to create a 1/5 scale model of the wing and mount it to the force balance in the wind tunnel. Using this wing, we were going to mount various winglet designs to both sides of the wing and determine which winglet design produced the greatest increase in the lift to drag ratio. However, after using the CNC foam cutter to cut the wing(photo below), Sam and I realized that the 1/5 scale model wing would be far to weak to mount in the wind tunnel even after reinforcement with a spar. This mainly being due to the large aspect ratio of the wing and the small chord length of the 1/5 scale model. So, rather than testing a 1/5 scale model in the wind tunnel, we have decided to test a 1:1 scale wing section. This section is roughly 27cm long and will be mounted to a plate attached to the force balance on the top of the wind tunnel in the test section area. This will be much like our previous experiment, but we should be able to obtain more definitive empirical data from our results about the performance of the wing with winglets. The plate will be oriented such that it ingests the boundary layer. In the previous lab the aerodynamic interaction between the wing mounted on the sting and the force balance produced inconsistent and inaccurate drag data. This experiment heavily relies on accurate drag data and we are hoping that this mounting method will improve the results.

To visualize the flow around the wingtip, we will also be mounting a camera on the side of the wind tunnel so that the camera is out of the wake of the wing. Smoke and a laser will be directed onto the tip of the wing in order to view the vortex. Currently, the apparatus we have built to perform this testing does not have the ability to change the angle of attack. This is something we plan to investigate more on today. We also plan to complete the manufacturing of the winglets by Tuesday. The winglets were originally going to be 3-D printed, but because the shapes of the winglets are so basic we have decided to manufacture them out of foam. This will save time and filament. We are planning to perform our winglet testing Wednesday during our normal lab time.

Thank you for reading!

[hanskeh-blog]

Week 7

This week we performed testing on two wings with the same airfoil and two different aspect ratios in the wind tunnel. During this testing we ran the tunnel at speeds which allowed us to test at the same Reynolds number for each wing. We were also able to analyze our results with a program I wrote to calculate the coefficient of lift at different angles of attack. However, we found during testing that the best way to ensure the quality of our results was to verify that the standard deviation of our data was not too large. When a large standard deviation was seen in the data, it normally signified that the wing was not tightly secured to the force balance. This insecurity caused the wing to vibrate which resulted in inconstant force data. However, the large standard deviation was not always the result of how secure the wing was mounted. It was also seen that the standard deviation of the data increased when the wing stalled. This was because the flow over the top of the wing was no longer attached and was creating vortices causing the wing to vibrate. Below are images of tuffs on the wing which allow for the visualization of the state of the flow over the wing. (attached, detached)

While testing we also noticed that we had taken a set of data which we did not need. We found that we did not need to measure the force values of the wing mount without the wing at various angles of attack and without wind. This set of data was unnecessary because this force is already taken into account when doing the same test for wind on without the wing.

These incidents while testing caused us to only obtain one set of data for each wing which was half of what we had planned to complete. But, the data we have recorded will still give us a good idea of how the aspect ratio affects the efficiency of the wing.

While analyzing the data we also found that the drag component of the force balance is off by nearly an order of magnitude. It is still unclear what is causing this interference, but it could be caused by the interacting flow around the force balance hub which increases force that the device reads. There are also other methods which can be used to measure the drag using the same force balance; Alex, Sam, and I may need to look these methods for our winglet experiment for the solar plane.

Alex, Sam, and I plan to manufacture the wing and winglets for our experiment this week. Because the wing will be a scale model of the full-size glider wing, it will have a large aspect ratio and will need to be sturdy. The wing must withstand the high speeds we need to run the wind tunnel at in order to match the Reynolds number which the glider flies at. To ensure that the model wing is strong enough, we will determine the lift the wing will generate at the operating tunnel speed and using this we will mount weights to the tips. This small pre-testing experiment will help us make sure our wing will not deflect too much and alter our results and it will also ensure that the wing will not break. To give the wing it structural support the wing will be cut out of solid foam and a 1/4” balsa rod will be used as a spar. The wing will then be monocoated and a thin layer of fiberglass will be adhered to the center of the wing where the wing will mount to the force balance.

When trying to determine which winglet will produce the best results we will be analyzing the L/D of the wing configurations. However, with drag results that are an order of magnitude to high, we may need to use the difference in the drag rather than the actual drag to perform analysis on each winglet. Because the wing is a scale model we will also be able to calculate an appropriate coefficient of lift for the glider wing. During testing we would also like to try to determine the stall speed and stall angle of the wing using smoke and the force balance. This would give us some helpful in site on how we can maneuver the aircraft in various flight conditions.

Thank you for reading! I will have some interesting video and photos to share next week!

[hanskeh-blog]

Week 6

This week Alex and I gave a presentation on our findings for the winglet flow visualization lab. There is some more information on this experiment in the previous post. Alex and I had some very interesting findings, we were astonished to see how much the addition of a simple rounded wingtip decreases the size of the wingtip vortex. This visualization really helped give us a good idea about which winglets were most effective. However, without any concrete data it was very difficult to make any definitive conclusions about which wingtip design was the best. Below is a photo from testing while using the rounded wingtip.

To investigate this further, Samuel Chung and I have decided to conduct another experiment to help us determine which winglets will use on the solar glider to reduce the induced drag caused by wingtip vortices (one on each side... plural). The full-size wing of the glider has a span of 3.82m and a 34cm chord. A 1/6 scale model of the wing will be made and mounted to the force balance in the wind tunnel. The wind tunnel will be set to operate at the same Reynolds number as the glider which flies at an average wind speed of 25m/s. A trip strip will also be installed where the solar panel slot begins on the full-scale wing. Various winglet designs will be mounted to the wing and from this we will be able to determine which winglet results in the greatest increase in L/D.

On Wednesday I worked with another lab group to begin preparing for the lab which will take place next week. We will be investigating how much the aspect ratio affects the efficiency of a wing. For this experiment we will be using two wings with the same airfoil; a 4412 wing with an 18.8125” span and a 6.1875” chord, and a wing with a 26.5” span and 4.625” chord. Relatively these two wings have the aspect ratios of 3.04 and 5.73.

To help prepare for testing I wrote a MATLAB code which will allow us to analyze the lift and drag during the testing, this will allow us to ensure that we are seeing the trends we expect in the data. We expect to see the L/D ratio to increase for the wing with the high aspect ratio. This is because the drag is inversely related to the AR. However, because the effective angle of attack at the tip increases for higher aspect ratio wings, we would expect to see that the wing stalls at a smaller angle of attack. The photo below helps describe both of these flow phenomena.

Next week we hope to get some great data and hopefully even some flow visualization using smoke during the experiment. Sam and I should also begin doing some preliminary planning to go over exactly how we will conduct the testing of the winglets. This upcoming week is going to be a busy one, as we hope to complete the manufacturing of the wing and fuselage of the solar plane as well. Below is photo of the wing design for the solar plane.

Thank you for reading!

[hanskeh-blog]

Week 5

In the beginning of this week, Alex and I completed the manufacturing of our winglets and practiced setting up the wing fixture in the wind tunnel.

This allowed us to prepare for testing which was to take place on Wednesday. However, upon arriving to the wind tunnel on the day of testing we discovered that the smoke machine was overheated and that the laser which was to be used to assist in flow visualization was no longer working. Thankfully the smoke machine only needed to cool down, which it could do while we were preparing for the experiment. The wing was mounted to the top of the wind tunnel test section as shown below.

The wind tunnel was set to 20 m/s for each winglet and the smoke was directed such that it flowed over the tip of the wing; this allowed us to visualize the vortex forming off the end of the wing. Two cameras were set up in the tunnel; one camera was mounted at the rear of the tunnel on the floor and another was mounted on the ceiling.

Each winglet was mounted to the wing, and a series of photos and videos were taken for each wing configuration. The camera at the rear of the wind tunnel had zooming capabilities which, in some cases, allowed us to see the vortex better. The camera on the ceiling was never moved which allowed us to take measurements of the size of the vortex to perform comparison. The vortex size was originally going to be analyzed by measuring the diameter of the vortex (seen using the laser) at the same point for each wing configuration. Because we did not have access to the laser, we have noticed that the size of the cone shaped vortex coming off the wingtip changes and have decided to measure the angle of this cone.

We are currently in the process of analyzing this data, but it appears that the rounded and endplate wingtips performed best. It can be seen in the pictures below that both the rounded wingtip and the endplate had a much smaller wingtip vortex than the bare wing.

The hoerner wingtip slightly decreased the size of the wingtip vortex and also forced the vortex outwards resulting in a greater aerodynamic span. This wingtip may not have reduced the wingtip vortex as much as expected because of its design. If a more in-depth approach was taken when designing the wingtip it may have proved to be the best candidate for reducing wingtip vortices. With this being said, the current best candidate is the rounded wingtip. The rounded wingtip reduced the size of the vortex and is also very versatile and easy to manufacture. The endplate also greatly reduced the size of the vortex; however, an endplate is not efficient in normal flight conditions with crosswinds and other flow inconsistencies common at low altitude flight.

Next week we plan to investigate the effect the aspect ratio has on the efficiency of a wing. We will do this using a force balance and various aspect ratio wing sections. This will be very interesting because it is extremely applicable to the glider we are developing. But, more on this next week!

Thank you for reading!

[hanskeh-blog]

Week 4

This week in lab we mainly worked on the NACA 4412 Experiment report and came up with a testing method for the Flow Visualization Lab. Our group had the rough draft of the report completed by Wednesday. Each of us then reviewed the report and the last to review (me) was to complete the final formatting and submit the report for grading. The final draft of the report can be viewed here.

Alex and I also made some great progress for the flow visualization lab. We completed our winglet designs and compared them to determine who’s design we should use. Using the CNC hot-wire we were also able to cut a 10” chord NACA 4412 airfoil out of a 36”x4” piece of foam. From this, we cut out a 16” section which could be mounted through a slot in the wind tunnel test section. This wing section needed a fair amount of sanding due to ridges caused by the hot wire. If these ridges were not sanded down they would have acted like turbulence generators and may have negatively affected our results. Images will be posted in the next blog, after testing.

We also had the opportunity to 3D print one of our hoerner winglets. However, after 3D printing a hoerner wingtip out of PLA, we have decided that it would be better to form our winglets out of the extra foam we have. The part made with the 3D printer had a very poor and rough surface finish which would alter the flow off of the wing. The winglets will be attached to the wing in the wind tunnel using foil ducting tape. We are not worried about these winglets fall coming detached; however, we are worried that the wing may begin to deflect at high speeds which would alter the flow off the wing.

Here is a good example of a gliders wings deflecting due to the large lifting force being generated during this banking turn.

Next week we plan to test on Wednesday during our normal lab time. We are using a camera at the back of the wind tunnel to measure size of the vortex coming off of the wing. Because of the nature of our experiment, we do not need the room to be dark for the laser. Monday we will finish manufacturing our winglets and we will also do some experimental photography with the smoke machine to determine which settings we should have the camera at.

MORE PICTURES AND VIDEOS TO COME!

Thank you for reading!

[hanskeh-blog]

Week 3

This week in 307 our lab group finalized the summarized data from our NACA 4412 Stall Strip Experiment done in the wind tunnel. This data was put into plots (which can be seen in the previous posts) to make the data easier to visually understand. We also divided up the work for the report and plan to have the rough draft completed before this upcoming Wednesday.

In addition to the report for the previous lab, I also began working with Alex Meraz on our flow visualization lab. We have decided to investigate the effects that various winglets have on aircraft flying in low Reynolds Number flight conditions such as gliders. It is uncommon for these types of aircraft to be equipped with winglets due to the very minor decrease in drag that they produce. For large aircraft, there is a large amount of lift being preduced by the wings which results in a large wingtip vortex. The addition of winglets to large aircraft such as this slightly reduces the wingtip vortex by reducing the interaction that the top and bottom surface pressures have at the wingtip. However, this decrease in the vortex size has a large impact on the planes fuel efficiency because of how large the vortex, plane, and distance the plane will travel are. For low weight, low speed aircraft the wingtip vortex is very small and a minor decrease in the size of this vortex will decrease the drag. However, this decrease in drag does not always outweigh the extra weight, and parasitic drag that the winglet introduces.

With this in mind, Alex and I have decided to investigate the effects of adding what is called a hoerner wingtip to the wing of a NACA 4412 wing with a chord of 10”. This wingtip seems to be specially designed for glider-like aircraft because of it aerodynamic characteristics. Unlike many winglets, the hoerner wingtip acts as a small continuation of the wing with a concave contour on the bottom which speeds up the flow on the bottom surface of the wing such that it has a speed closer to that of the speed on the top surface of the wing. This greatly reduces the size of the wingtip vortex while also directing it outward. It also increases the effective wingspan without increasing the wing area.

To investigate this, we will be mounting the wing at a 5° AOA to the top of the wind tunnel test section. The wing’s AOA will not be altered throughout the testing. We will design and 3-D print three different wingtips: a hoerner wingtip, an endplate wingtip, and a rounded wingtip. Each of these wingtips will be attached to the end of the wing using foil tape. We will also conduct a test without a wingtip as a baseline. The wind tunnel test section velocity will be set to 20m/s which will correspond to a Reynolds number of approximately 5x10^5. A stream of smoke will be directed to the tip of the wing which will allow us to visualize the flow and specifically the wingtip vortex. To capture this, a camera will be placed at the back of the wind tunnel which will be focused onto the wing tip and will take a series of photos for each test. These photos will enable us to measure the diameter of the wingtip vortex which will help us determine which type of wingtip will be more effective at reducing wingtip vortices.

[hanskeh-blog]

Week 2

Next week we will work on completing the report for this research. We will also begin devising a test plan for a (currently undetermined) flow visualization study. It has been discussed that the flow visualization may involve the study of the addition of various types of winglets to a wing to determine which model is most effective in reducing wingtip vortices for small scale aircraft such as RC gliders. This would be very applicable to a solar plane project I am currently working on and will give me some good insight on which type of winglet will be most suitable for our plane. This week our team was able to take a deeper dive into the data that we recorded last week, and we also began more rigorous testing on the effects the stall strip had on the wing’s viscous drag. The data taken during last week’s testing was very useful in determining the effects that the stall strip had on offsetting the angle of attack at which the wing stalled at. However, the data was not so useful in determining how much the stall strip affected the wing’s drag. After some more analysis of last week’s data, it was found that the CL was slightly decreased, the CD was relatively unaffected, and that the wings stall angle was increased by 1 degree after the application of the stall strip. This data implied that the application of a stall strip could increase the maximum angle of attack without having any drastic effect on the drag. But, the drag analyzed only included the self-induced drag. This led to the need to do more analysis on the change in parasitic drag through wake rake testing. The wake rake testing conducted this week provided an abundance of useful data. From this data it was found that the stall strip had a major effect on drag and actually caused the parasitic drag to nearly double.

The wake rake testing was conducted by panning a pressure probe through the wing’s wake at a distance of 2.5 inches behind the trailing edge. The data was taken at 2mm increments until the probe was within the wake. Once the probe was within the wake the step size was reduced to 1mm. During the testing I analyzed the data in real-time in order to ensure that there were no poor data points and to determine when the sensor had left and entered the wake. The outcome of this kind of analysis produced an outstanding set of data showing the effects of the stall strip on the wings parasitic drag.

Next week we will work on completing the report for this research. We will also begin devising a test plan for a (currently undetermined) flow visualization study. It has been discussed that the flow visualization may involve the study of the addition of various types of winglets to a wing to determine which model is most effective in reducing wingtip vortices for small scale aircraft such as RC gliders. This would be very applicable to a solar plane project I am currently working on and will give me some good insight on which type of winglet will be most suitable for our plane.

Thank you for reading!

[hanskeh-blog]

Week 1

Thank you for reading!