scott-kaufman-blog

[scott-kaufman-blog]

Reflection

Over the course of 307 I feel like I learned a lot of things that I will take with me in the future. 

One thing I really enjoyed about 307 was getting to work with all the equipment in the tunnel and learn how to use it. Even though it didn’t all work all the time, I think I learned a lot about trouble shooting equipment and figuring out ways around broken equipment. 

In terms of  being “prepared for the no-single-answer, sometimes-don't-even-know-what-the-question-is” scenarios, I don’t know if I’m better prepared, but I think I know how I would go about them now. I think kind of setting your own goals and reasoning for why you’re doing what you’re doing is important, and as long as you can stand behind that, I think it’s a good way to approach those kind of scenarios. 

In 307 we worked in the same group the whole time and I think it went well for the most part from my perspective. In terms of working with a group it was one of my better experiences.

I really enjoyed doing the aerodynamic testing, but in a way I kind of missed actually learning new concepts. I don’t mean I didn’t learn anything, but I felt like there were a lot of times we would collect our data and then use our past knowledge to analyze it, instead of learning new methods to analyze data or learning about phenomena in the flow that we haven’t covered before. The experiments were really great though for seeing the things we’ve learned in class in real life, which I enjoyed.

[scott-kaufman-blog]

Final Group Meeting

Today was our last official day of lab, and I think we made a good use of time working on the memo/report. After reading the goals of the project again and looking back on what we did, I feel content with he amount of work we accomplished. We were able to figure out a new mounting method to obtain higher angles of attack. We confirmed the pressure ports work well, and we were able to get accurate calculations using them. We found the wake regions affected by wing, and which regions data should be recorded in to capture the whole wake and obtain a good velocity profile. And finally, Andrew was able to get the handheld hotwire anemometer working and got some results from that. 

Over the last couple days since last lab though I have been working on doing a better analysis of the pitot tube data, and I have been able to generate 2 plots that I find interesting. The first one is the velocity profile, which shows the effect that the Active Flow Control has, as well as where each wake region is located. 

The second plot is the Coefficient of drag. I still haven’t been able to figure out how the distance that the probe is place behind the wing affects the coefficient of drag calculations, so the magnitude of my calculated values are pretty far off, but the trends seem to confirm the predictions we had going into the lab, as well as confirm previous calculations done with the pressure ports.

[scott-kaufman-blog]

Hotwire

We have been experimenting with the handheld hotwire anemometer a fair amount lately in addition to doing wind tunnel testing lately. We wanted to find out how the frequency and voltage inputted into the actuators affects the velocity of the jets through the slits on the wing. To do this, the hotwire was held  directly above the slits, recording the velocity. This was done for each slit at each frequency and voltage. The data recorded is plotted below.

After looking at the plot, it’s easy to see that at the frequency and voltage we were running at (500 Hz, 60V), the velocities coming out of each port vary a fair amount but generally have an average around 3 m/s. 

Another thing that I find interesting and nonintuitive is that the jet speed decreases when the frequency is changed to 100 Hz and the voltage is unchanged. Based on my understanding, the voltage should determine the amplitude of the displacement of the disk, so the velocity would not change much because the displacement of the disk would be the same. However, this is clearly not right, so I’ll probably do some more research to figure out why this is.

[scott-kaufman-blog]

Final Testing Day

Today we ran our Pitot Probe test with the wing mounted to the floor in the tunnel and then used the pitot probe to record total pressures in the wake behind the wing. We defined the position directly behind the wing as our 0 location. Due to the way the traverse and wing were set up, + indicated the region beneath the wing and - indicated a position above. We ran at a single Reynolds numbers and 3 different angles of attack (0,10, and 25 degrees). For each angle of attack we recording data with the Active Flow Control (AFC) on and off.  At each test condition we traversed the pressure probe across the wing recording pressures at every 5 millimeters. The range that we traversed though varied with each test condition. This was determined by viewing (almost) live data in matlab to get a sense of what the pressures in the wake were looking like. We started each position at 0 and increased in each direction until the probe appeared to be reading the same as the stagnation pressure of the tunnel, indicating the wing had little to no effect on that location. When looking at the data we were hoping to see a “U” shape in the total pressures behind the wing. Points of low stagnation pressure should indication a point experiencing a large effect of the wing. When we tested at 20 degrees with the AFC on and off we recorded the pressures in the figure below.

In this figure, the blue circles indicate when the AFC is on, and the red circles correspond to when it’s off. At this angle of attack the flow was fully separated when the AFC was off. I believe this may be why the wake region appears to be much wider than when the flow is attached in the scenario when AFC is on. Also, the total pressure drops should relate to the amount of drag experienced by the wing, so I am thinking the drag should be much lower with the AFC on. 

One case I haven’t analyzed yet that I think should be interesting though is at an angle of attack around 10 degrees where neither of the flows are separated. I’m curious whether the AFC control effects the drag and the location of the wake when the flows are attached, of if it is useless in this region. Should be interesting to see.

[scott-kaufman-blog]

Making Progress!

This week has been one of the most productive weeks we’ve had in 307 so far, and I think our group has been working together and dividing up tasks very efficiently.

Starting with Tuesday... We used the wooden spacers we made to mount the wing further into the test section. Our hopes were that this would allow us to pitch the wing at unlimited angles of attack and not be limited by the walls of the tunnel. Additionally, with this set up, we would be able to mount the Scanivalve next to the wing using the tracks on the outside portion of the test section. As always, there were a few flaws in our plans, that we found when we began mounting the wing. First, the spacers worked fantastically. They moved the jig closer to the test section, freeing us from the limitations of the traverse. However, we ran into another issue. The bolts on the clamps, that keep the wing set at a specific angle of attack, extend and would collide with the walls of the tunnel before the wing is fully placed in the tunnel. This is shown here:

After taking further measurements, even if the bolts were shortened, the clamps themselves would collide with the walls as well. Another issue we encountered (which we somewhat predicted) was the pressure lines on the wing were very short, making it difficult to mount the Scanivalve in a location where the jig would not collide with it. We determined that mounting directly below the jig was the best option after testing multiple locations.

At this point, we accepted that we would be unable to pitch the wing to higher angles of attack using this setup, so we carried on with our experiment anyways. We recorded pressure data on the upper and lower surfaces of the wing at a range of alphas up to our max available alpha, which was around 20 degrees. 

Over the last couple of days, I have spent some time analyzing the code and came up with a few plots:

This Coefficient of Pressure plot looked promising to me, because it demonstrated that the Active Flow Control resulted in great levels of “suction” on the upper surface, which is what we were hoping for, but it did not seem to make as much of a difference when looking at the Coefficient of Lift shown below.

I calculated the Coefficient of Lift using a panel code, and according to this plot, the Active Flow Control did not make a significant difference, which I found interesting, so I may check my calculations once more; however, the results are similar to other wind tunnel tests, giving me confidence in the calculations. Additionally, this plot says that we never reached a stall point, which is what we kind of wanted to focus on, so it may be beneficial to take further pressure data if we have time, but there are other tests we would still like to run.

When looking at the Coefficient of Drag though, the Active Flow Control made a clear difference, which I find really interesting, and am still unsure how to explain why the drag changes significantly but the lift does not. 

Finally, one other thing I wanted to visually see through the analysis was the pressure distribution over the wing, which I thought was best done through a 3D plot shown below.

I think the analysis is going well so far, but I definitely still need to sit down and think about what everything means.

Today, we switched gears a little bit. We did not initially have plans to test in the wind tunnel, but ended up doing a little bit at the end since everything went better than expected. Initially, we wanted to spend the day figuring out another way to mount the wing such that we would be able to pitch to higher angles of attack, which was the main failure of the previous experiment. We decided to do this by bolting the jig to the floor of the test section. We accepted that by doing this, we would be accumulate additional limitations and problems. These include not being able to adjust the angle of attack with the wind tunnel on (which we were unable to do previously as well) and we cannot record pressure data using the Scanivalve unless we mount it in the tunnel next to the wing, which I am not sure if that solution is acceptable. One problem this created was connecting the wing to the linear amplifier. Initially, the chord was too short to reach the linear amplifier outside the tunnel, so today we used additional wire to create and extension which lengthened the chord, resolving the issue. 

Here, Adam and I are connecting the extension wire to the wing.

To mount the jig to the floor of the test section we predrilled holes and used 2 5-6″ lag screws to screw the jig into the floor. We made sure that these screws went into the crossmember beneath the floor for extra security. These bolts can be seen here:

We also swapped out the clamp on the jig for a more secure one shown above to minimize the chance of the clamp flying off. Once this was complete, we still had an hour or so left in lab so we figured we’d use the time to collect some flow viz and see where the wing stalls at with the Active Flow Control on.

Below, the wing is pitched to 22 degrees, which ended up being the max angle of attack where the Active Flow Control would still reattach the flow.  We also, tested 23 degrees, and the Active Flow Control still appeared to make a difference, but was not quite able to reattach the flow.

Finally, we tested additional deep stall angles of attack to see where the Active Flow Control made absolutely no visible difference, and found that this occurred around 30 degrees, which we found to be an interesting piece of information.

[scott-kaufman-blog]

Presentations/Experiment Prep

The presentations went well today and I feel like I learned a few things that will help out in the future, especially in design next year. After the presentations, I think our group used really good use of the time we had left to get ready to do some testing on the active flow control wing. 

We finished building our jig that will act as a spacer, allowing us to pitch the wing at higher angles of attack. The jig we made is pictured here:

Once we finished building the jig, we hooked up the pressure ports to the scanivalve to verify that they work, and in doing so, we got some experience hooking them up. The pressure ports do indeed work; however, the pressure tubes are slightly small for the scanivalve, and they are too short to hook into the scanivalve once the wing is mounted in the tunnel. So, we may need to find some adapters for this.

While part of our group was working on the stuff above, others were working on learning how to use the hotwire anemometer, and hopefully get that working in time.

[scott-kaufman-blog]

Final Project-Active Flow Control

After working with the active flow control wing for the flow viz project, I am excited to do a continuation of it for the final project. After talking as a group we decided we would really like to learn more about the airflow through the slits using the hotwire anemometer (and learn how to use the hotwire anemometer). In addition, we would really like to utilize the pressure ports on the wing and use them to see how the active flow control changes the lift and drag on the wing. During the flow viz experiment, we were limited to a max angle of attack of 20 degrees, which wasn't steep enough to cause a stall with the active flow control on. So, we have already begun building a jig that will allow the wing to extend far enough into the tunnel that the walls of the tunnel will not limit the AOA. We have cut plywood squares that we will place between the mounting plate on the traverse and the back of the jig that the wing mounts in.

Finally, we are curious if the vibration alone has an effect on the flow. I was reading the thesis papers and they had experimented with this already, and according to their results, the vibration didn’t seem to have much of an effect, but it would be interesting to see if we can replicate those results. The last thing we may experiment with is the frequencies and voltages created by the waveform generator and the linear amplifier to further understand the effects those have. 

[scott-kaufman-blog]

Active Control Flow Viz

Today we ran our flow viz on  the NACA 0015 with Active Flow Control, and everything went really well! I felt like I learned more today in lab than I have in the rest of quarter. I learned a lot about swapping out test equipment and setting up the Active flow control wing. I enjoyed actually getting to do it and not just watching someone else set it up.

When it came to actually running the flow viz, everything worked just as expected, which was awesome! Alex had shown us videos of when he tested the Active flow control, but I still wasn’t expecting it to work as well as it did. I’m really curious now though as to what angle of attack we could pitch this airfoil and still have the active flow control effectively reattach the flow back to the wing. I’m really hoping we can turn this project into a project for the last few weeks of the quarter so I can learn even more about how the active flow control works and possibly ways that it could be improved. 

Here are some of the photos from lab today!

[scott-kaufman-blog]

Flow Viz

We had some trouble thinking of what we wanted to do our flow visualization on. Initially, we wanted to do a frisbee, but making a frisbee spin seemed too difficult for a 1 week project. Instead, we found an even cooler project. We are going to do flow visualization on the active flow control wing. The wing uses electric frequencies to drive diaphragms to oscillate airflow through slots on the upper surface of the wing, around 10% chord. After watching some videos of it and learning how to set it up this morning, I’m really excited about it!

[scott-kaufman-blog]

Finishing the Report

Today, we finished sat down as a group and read through the report and finalized the document. Overall, I’m pleased with the way the report turned out and the way our group worked together. 

Throughout each experiment, we have had some kind of issue come up, but after putting all of the results in the report, the tests were definitely worth the time. I think a lot of the issues with the data encouraged us to think more about why we were getting the results that we were getting, and it made us think more about what was going on that we couldn’t see. 

I think the equipment struggles weren’t all for nothing either. I think being the first group to run each week allowed us to learn a lot about the equipment in the wind tunnel and how to set up and operate different equipment. 

I think some skills that I’ll take away from this experiment are primarily hard skills when it comes to operating the wind tunnel, but one soft skill I will take away is working with a group. As with any group project, you have to be able to work together, and I think this project gave me extra practice in that.

[scott-kaufman-blog]

Finishing up the Analysis

I had been struggling trying to correctly calculate all the forces and coefficients correctly on the infinite wing up until yesterday. The panel code method that I was attempting to use had a few mistakes in it that I kept missing, and it would result in drag forces that would drop significantly when separation occurred, and the lift forces were much greater than the theoretical. Now that it is (hopefully) correct, the coefficients of lift and drag are very close to the theoretical values. The l/d that we produced from these however, is slightly larger than the theoretical, but I think it just has to do with the fact that we are dividing by such small numbers that small errors can result in a large deviation from the theoretical. From here, I would like to plot the theoretical lines on the same plots as a comparison in the report, and then the goal for this weekend is to finish up the report. 

Here are some of the plots we have now, which I think are looking pretty good. 

[scott-kaufman-blog]

Winglet Test

The winglets that we 3D printed turned out really well, and Justin sprayed some cool gold paint on them that we were hoping would give us a smoother finish than the 3D printed surface. When mounting the winglets to the airfoil, everything ended up fitting well, but we had to put tape around end of the dowels on the side that went into the winglet , because when we designed the winglet, the holes were sized much larger than we needed so there would be no risk of the dowels not fitting in the slots on the winglet. 

Once seeing the winglet on the wing, it looked like tapering the rear of the winglet could have significantly decreased the drag, but the reason the design was somewhat blocky is because we were unsure of the limitations of the 3D printer, and we didn’t was to attempt to print a part that was too thin and would warp. 

After looking at the data though, we were achieving a L/D of around 5-6. For an infinite wing this should theoretically be around 60-70. There is a way to convert this to a specific finite wing, but I need to do more research to figure it out. The best thing to do next is to compare these results to the previous tests to see what improvements were made by the winglet. 

Pictured below is the winglet during the test.

[scott-kaufman-blog]

Pressure Rake Testing

Today in lab we began testing the pressure rake. We  spent the first 40 minutes of lab setting up the pressure rake and connecting the pressure. We were testing 3 different Reynolds numbers at 3 different angles of attack, 0, 5, and 10. We started with the pressure rake near one end of the tunnel and tested each condition at that location. Then, we moved the rake to the next position which was 50mm over, and then we tested the conditions all over again. This would have been a good approach if we would have had more time to get the rake to the other side of the tunnel. When analyzing the data briefly during the lab, the pressures were decreasing as the pressure rake entered the region behind the wing, which is what we were expecting because the static pressures should be the same at each point, but the velocities should be lower directly behind the wing due to the loss of momentum. Since the pressure rake is recording total pressure, the pressures should decrease. This was great; however, we never got to the point where the pressures started increasing on the opposite side of the wing, so we may need to go back in and test more data points. 

[scott-kaufman-blog]

Winglet

Yesterday throughout the day we 3d printed our winglets. We split each winglet into 3 parts in order to give us flat surfaces to print on and to avoid having to print any scaffolding for certain geometries on the part. Originallly the printer estimated about 40 hours total to print both winglets. But after splitting it up the way we did and adjusting the layer height to 0.2 mm instead of 0.1mm, we were able to print both winglets in just one day. 

Once each piece was done, they were all epoxied together to form a single winglet. We were considering adding aluminum tape to the whole winglet to improve the strength of the winglet and give us a smoother surface finish since the 3d printed surface looks nice but is still somewhat rough. 

Here’s some pictures.

[scott-kaufman-blog]

Infinite Wing Testing

Today we tested the infinite wing in the wind tunnel. Compared to previous labs,  this testing went really well. All the pressure ports were giving consistent and reasonable data except for port 8, but I corrected this port when doing the data analysis using the ports around it. The only surprise we really had today was there was no pressure rake. The code I had written ahead of time was written to include the pressure rake data, so I just had to make some changes in the beginning of the lab. 

When we were running we had a similar strategy as before. We tested in increments of 5 degrees with multiple Reynolds numbers. Once we had this data we were able to narrow in on where the stall point occurred, using a Cl vs alpha plot. We tested a combination of 1 degree and 0.5 degree increments around the region. 

I’ve also been spending quite a bit of time today analyzing this data. I have been able to plot the Cl vs alpha and get reasonable results. I did this using a panel method; however, using this same method for drag, I got negative results, so I need to look into that. I also plotted the pressure coefficient and tried to use that data to calculate Cl again to see if the results were the same. They aren’t. So I’ll try to fix that too.

One cool thing I was able to do with the Coefficient of pressure though was plot the pressure distribution over a 3d airfoil using a surface plot. When looking at it at different angles of attack, it’s easy to see how the pressure differential on the upper and lower surfaces changes with angle of attack. Also, as the angle of attack increases in pitch you can see the separation point move forward on the upper surface, and by the time the angle of attack is 20 degrees, you can see it’s fully separated on the upper surface. 

[scott-kaufman-blog]

Infinite Wing Prep

Today we worked in the conference room focusing on improving our code to analyze the data from the blue wing, and working on putting some code together for the infinite wing lab next week. I primarily focused on the code for the infinite wing. The code I made runs through all the angles of attack and Reynolds numbers tested up to that point in the lab and separates the pressure data from the wing and the rake. Using momentum and the rake pressures the code calculates the total drag. I made the assumption that the static pressure is constant behind the wing, which I read was true but am not sure why. Also, I am unsure how the distance between the wake rake and the wing affects the results and how we would account for this. Hopefully this will produce some good velocity profiles behind the wing

I have also been working on calculating the pressure drag on the airfoil using the static pressure points. I am making progress but am just having trouble figuring out the incidence angle between pressure and the x-direction. Once I figure this out it should be pretty simple to calculate the viscous and pressure drag.

[scott-kaufman-blog]

Blue Wing Test Day

Yesterday, we continued our testing with the blue NACA 4412. Overall, the data we collected was better than the red wing; however it still was not nearly as good as we were hoping for. During the testing, I was checking the results using a code of Brandon’s that he helped me adjust to work for our data; allowing us to graph Cl and Cd plots against alpha. He showed me some pretty cool tricks you can do with Matlab that I had never seen before, so I thought that was pretty cool.

The first few data points we collected seemed reasonable based on what we predicted; however, when we jumped from 10 to 15 degrees the Clplot began to  decrease, so we assumed our stall point occurred somewhere in between. We ran at 20 degrees anyways since that was our plan in the beginning, and we wanted to get some deep stall data. When we ran this angle, the Cl increased again which doesn’t seem right to me, so I would like to look into why that happened. 

After that, we went back to 12.5 degrees to narrow in on the stall point. When we did this and calculated Cl from the data, Cl became negative. This could be a calculation issue or it could be related to bad data, but I’m not sure at this point so I plan on looking into that as well. At this point we were already beyond our lab time, so we didn’t get nearly as much data around the stall point as we had liked, so hopefully we’ll be able to make some up.

We got some cool flow vis though.