mattwambach-blog

[mattwambach-blog]

Aero 307 Reflection

What did you find to be the most memorable "moment" of 307?

The most memorable moment for me has been the past couple of weeks spent working on the mini projects. Designing a fairing for wind tunnel traverse turned out to be a fun and challenging project. Also, the winglet challenge was a pretty valuable experience for me. I enjoy doing CAD work, and both of these projects provided the opportunity to get better at it. What did you think was the most challenging aspect from all 10 weeks? (other than time... yes.... time... always need more time....)

The most challenging part of the past quarter was the entire 4412 campaign. We collected way too much data for some of the experiments. This made the analysis tedious and cumbersome. Do you feel better prepared for the no-single-answer, sometimes-don't-even-know-what-the-question-is stage of senior year and the workplace? And/or has the course helped you better get to know yourself and how to manage your approach to such situations even if that's not the environment in which you feel you thrive?

I definitely feel more comfortable working in a testing environment than I did at the beginning of the quarter. By the end of Aero 304 last quarter, I wasn’t yet entirely comfortable conducting experiments in the wind tunnel. After 307, I feel fairly confident performing any role in lab, whether it be running Labview to setting up equipment. In order to get to this point, I basically just asked a ton of questions until things started to make sense. In 304 and 307, you've worked in several different groups in the lab setting - what are your main takeaways for how to manage group situations and teamwork and team dynamics going into senior design or other group projects next year?

Communication is important. Being active in the group chat (or any other form of communication) and having each person complete specific tasks definitely helps in the development of an effective, cohesive group. The experiences working with different groups will certainly be beneficial in design next year. Do you feel like the course made you more interested in aerodynamics and testing, or did you "get your fill" from this course and will strike out in another direction?

This course did make me more interested in aerodynamics and testing, but I still think I’m more interested in advanced composites/structures. I definitely appreciated how hands-on this course was. Much more fun than a typical lecture-only class. It was a lot of work, but it was one of the best classes I’ve taken so far at Cal Poly.

[mattwambach-blog]

Traverse Fairing Prototype

We 3d printed a prototype for our traverse fairing this week and as expected, we ran into some problems. The first problem being we ran out of material half way through the print, which is why it’s two different colors. Additionally, the part warped and split quite a bit on the print bed. This may have to do with the size of the part and the manner in which the layers cooled. Sometimes 3d printed parts will experience warping and splitting caused by differences in temperature and cooling rates in between the layers. It also took about 14 hours to print, which is about what we were expecting.

We may have to find better print settings for these parts so that they don’t warp or split. We could also try printing them in smaller sections, which may help prevent this problem.

[mattwambach-blog]

Traverse CAD Design

The CAD work for the traverse fairing is mostly complete. We were originally considering building the fairing out of sheet metal, which likely would have resulted in a finished product that looked quite different from any CAD model we would have come up with. No one on our team had much experience with sheet metal fabrication and we came to the conclusion that it probably wouldn’t be the best option for this project anyway. The fairing needed to be easily removable, which would have been difficult if it was made with sheet metal. Because of this, we made our design into something that could be 3d printed and put together in pieces. 

Our working design consists of 10 separate pieces (eight body pieces and two rounded end caps).  Although 3d printing is probably our best option, we have considered several factors that may be problematic with this method of manufacturing. Printing these pieces would require A LOT of material. Moreover, printing them would take a very long time. But assuming there is enough material available and if it’s printed during the summer when less people need to use the printer, then it should be feasible. 

[mattwambach-blog]

We designed and 3d printed a test piece for our traverse fairing using Solidworks and the water tunnel printer. It is 8 mm thick and about 350 mm long. The hollowed out part in the middle was designed based off of measurements we took of the traverse. The longer flat side is to be attached to the side of the traverse that faces the control room. We plan on attaching it via the screw holes that are already drilled into the traverse. We had to make a few adjustments to the dimensions of the hollowed out portion to accommodate some bolts that were sticking out on the side. Other than that, the test piece was a good template for the actual fairing.

[mattwambach-blog]

Traverse Update

One of the biggest challenges we’ve been facing with the traverse is accommodating its complex geometry. Designing an aerodynamic fairing that fits around it while making room for all the bolts and probes that are sticking out of it has required a lot of time taking measurements and getting a feel for the shape. We decided to design the fairing based off of the use of a smaller chain. The current chain (seen in picture 2) is needlessly large. As a result, the current plan is to accommodate a chain that is half the thickness of the current chain. 

[mattwambach-blog]

Mini Project: Streamlining the Traverse

Our final project for the quarter involves streamlining the traverse. The z-axis strut on the new traverse currently without a fairing. This is causing a lot of problems such as vibration, strain on the belt, and noise. 

We are tasked with designing a fairing that covers the z-axis strut that does not interfere with the movement of the pressure probe that protrudes out in front. 

We may base our design off of the ME senior’s design that was never manufactured. It will probably be a challenge to get all of the parts to fit around the fairing without interfering with its function. Additionally, manufacturing the fairing will present its own challenges. We may possibly 3d print the fairing, or portions of it at the very least.

[mattwambach-blog]

These diagrams show the effects of the fuselage on the lift and drag of the wings. Both the lift coefficient and the drag coefficient increase due to the presence of the fuselage. Also, the effective angle of attack increases because of the fuselage. This is known as alpha flow. This would cause the portion of the wing nearest the fuselage to stall before any other portion.

We essentially observed the opposite of these effects during our flow viz experiment. The flow remained more or less attached near the fuselage and separated near the mid span. 

[mattwambach-blog]

Here is a diagram depicting the flow over a swept wing. For a swept wing, the flow meets the wing at an angle. As a result, the flow gets broken up into normal and spanwise components. The spanwise component flows outward and is relatively constant along the span while the normal component can vary quite greatly. This variation in the normal velocity component affects the lift on the wing, as the lift only depends on the normal component.

For our test at AOA = 10 degrees, the outward facing tufts may be the result of the spanwise flow component.

[mattwambach-blog]

Flow Viz Update

The above photo shows the wing at an angle of attack of 10 degrees and a speed of 15 m/s. One unexpected thing we observed was that the wing stalled at around the mid span while the root remained unstalled. Due the washout of the wing, we expected that it would stall near the root before any other portion of the wing. 

The unstalled portion near the root is somewhat triangular shaped. This could possibly be due to the buildup of a boundary layer near the fuselage/wing junction.  

Additionally, the tufts seem to be pointing outward near the wingtip. This may be the result of spanwise flow. 

[mattwambach-blog]

Flow Viz Challenge

We decided to use Matt Paul’s 3d printed Common Research Model for the flow viz challenge. Our objective was to observe the flow conditions around the wing and determine the influence of the fuselage on the wing near the root.

Our setup included four rows of tufts taped to the wing in an offset fashion near the root and one row of tufts for the remaining span of the wing. The actual setup of the wing in the test section took a little bit longer than anticipated because the pins that hold the wing in place weren’t fitting in properly at first. Once the setup was complete, everything else went by quite easily.

We tested at angles of attack of 5, 10, 12, and 15 degrees at a speed of 15 m/s. Video was shot using the mini Go-Pro that is clipped onto the top of the test section. 

[mattwambach-blog]

4412 Campaign Reflection

The past six weeks have been very busy to say the least. Every week we were in the wind tunnel gathering data and testing one of the three 4412 wing sections. Being in the wind tunnel so often definitely made me feel more comfortable performing all of the different available duties, from Labview to Matlab to setting up equipment inside the test section. The most challenging aspect of the whole campaign for our group as a whole was the analysis portion of each experiment. We found it was very easy to become overwhelmed with the amount of data we were collecting. Additionally, we found it challenging many times to get our CL vs alpha and Cp plots to look the way that they should. Oftentimes, the our plots would follow the correct trend, but the values wouldn’t make sense.

For me, the most exciting aspect of the past four experiments was the winglet challenge. It was pretty neat to have the opportunity to design and manufacture our own designs from scratch. We got to 3d print them and they actually turned out pretty well after some sanding. 

Overall, the 4412 campaign was a lot of work and I gained a lot of practical experience as well as confidence conducting experiments in the wind tunnel. I’m looking forward to flow viz this week and next!

[mattwambach-blog]

Week 6

The amount of data that we wound up collecting over the course of the 4412 campaign was probably too much, at least for the first two experiments. For the finite wings, we tested at seven different angles of attack and three speeds. We gathered even more data for the full span wing, with ten different angles of attack at four different speeds. In addition, we collected data twice at each flow condition for both the finite and the full span wings. Having this much data was pretty overwhelming when it came time for analysis. If we had the option to do everything all over again, we agreed that we probably would cut down on the amount of data we collect as well as come up with a better file naming convention.

[mattwambach-blog]

Week 5

For week 5, we tested the full span wing with a twenty port pressure rake placed in the wake region. The purpose of this experiment was to observe how the pressure behind the wing changes as angle of attack changes. We chose to test at angles of attack of 5 and 9 degrees and Reynold’s numbers of 300k and 750k. Angles higher than 9 degrees could have potentially led to flow separation which would have rendered our pressure data essentially useless. The mechanism pictured below moved the pressure rake in increments of 50 mm. We tested at several different rake positions for each condition. 

[mattwambach-blog]

Here is our final winglet design. We decided to go with a slightly more traditional looking winglet as opposed to the wingtip fence design we were considering earlier in the quarter. This design features an upward swept, trapezoidal shaped 4412 wing section. If it works as intended, there will be an angled lift force that points up and inwards, and a component of that lift force will contribute to the total lift of the entire wing section.

The winglets were printed on the Lulzbot Taz 6 3D printer in the wind tunnel lab.

[mattwambach-blog]

Testing of the full span wing went well. We took data at angles of attack of 0, 5, 7, 9, 11, 13, 15, 17, 20, and 23 degrees and Reynold’s numbers of 227k, 400k, 600k, and 750k. Data was collected twice at per angle of attack per and Reynold’s number. 

We did not encounter any problems with any of the hardware and the data seemed to be good. We’re now working on the Cl vs alpha plots and other analyses.

[mattwambach-blog]

Testing on the red and blue finite wings is now complete and we are in the process of analyzing our data. All of the data was a bit overwhelming at times and we spent a lot of time just loading all of the data files into Matlab. 

For both wings, we collected data at 0, 5, 7, 9, 11, 13, 15, and 17 degrees angle of attack. At each angle, we ran at three different Reynold’s numbers: 150,000, 175,000, and 200,000. We were particularly interested in the stall regions of the wings.

We ran into some problems with the linear drive while blue wing was in the tunnel. Sometimes when we would change pitch, the linear drive wouldn’t move at all, and sometimes it would move only a little bit. Once Riess helped us fix that problem, we started having problems with the pitching mechanism. The pitching mechanism was an easy fix, however. All we really had to do was change its rotation speed. It took a lot of back and forth communication between our team to work through these problems. I was in the test section manually measuring the height and pitch of the wing in order to verify its true position while Colton, Aaron, Alex, and Riess were in the control room troubleshooting what could be wrong. 

These minor problems delayed our testing somewhat, but we were still able to acquire all of the data we needed. 

[mattwambach-blog]