bitter-307-blog

[bitter-307-blog]

Week 10: What a ride.

Wow, what a journey we went on from start to finish. So here we are. Monday of this week, we finally got some testing properly done in the water tunnel. Flipping things over and doing it upside down made things work. I’d show you visuals from it, but literally all of them were put into our presentation. You have that link, I hope a few clicks won’t kill you.

So time to reflect.

One thing I feel like I understand better about aerodynamics after the class is the realities of testing. Everything you can think of matters when you are testing, and the 100 things you didn’t think of probably matter too. In the force balance lab for example, the drag from the sting was just wrong for everyone. Every group found that, but having us all know about it and do nothing was interesting.

One thing that confuses me still about aerodynamics is boundary layers. I never am sure when they matter and when I can ignore them. It makes me concerned about all measure of testing in a relatively narrow wind tunnel or water tunnel. I know I theoretically should know more about it from taking 302... but I forgot lots of it.

307 highlight: Baseball aerodynamics. It was just a cool story from start to finish. We started with an idea of something we wanted to test. Then, we struggled to figure out how to make it work, but we got ourselves in the water tunnel with a drill and did some science. AND the results we got were so inconclusive on the thing we were trying to study! However, we learned something that we weren’t even looking for. Sure, we couldn’t see what difference seam configurations made, but spinning vs non-spinning was awesome to see the difference.

Lowlight.... I’ll give it to the Panel Gap.

In hindsight, it was kind of funny and a problem that we just eventually had to work around instead of solving. As frustrating as that is. That’s engineering, and more importantly that’s life. I still hated it, so it’s a lowlight. But I see the value in the struggle.

The future of me in aerospace. Man...

I’m not that good at aerospace, but I like it a lot. I’m also more comfortable with pressure on me and when I have to be quick on my feet (hence the improv.) So a way I see that going is being something like an imagineer for Disney. Using a general knowledge and understanding of aerospace and engineering to make things that might not be real, but aren’t so far fetched that people question it could be. Then, I might be able to inspire others to greatness to make what I pretended was a real something actually real.

Thanks for reading.

[bitter-307-blog]

Week 9: Everything is Fine

Okay, this week was miserable. You would think that it would be great because we didn't have class on Monday, but it was somehow the worst.

So Wednesday before class we got an email from Daniel with Doig pointing out all these flaws with our model and testing plan. We needed to get a model with a larger cavity, and the front ramp leading up to the model needed to be less steep so the flow wouldn't separate.

So we printed a model with a bigger panel gap and a larger cavity. Check.

Now, we have to make a ramp. We've done this before, so at this point I think "second verse same as the first." I was wrong as hell.

Well that was a pain in the ass. We struggled with it enough to eventually get something that is good enough. We sanded those and glued the ramp to the first model, and then we set a plan to come test on Friday. That's right, we were coming in on our day off to try and do some science.

Friday roles around. And the struggle bus keeps on rolling.

First of all shout out to Tynan and Daniel for letting us be there NDAs aside.

Second of all blue foam is the most buoyant nonsense I have ever had the disappointment of dealing with. We put something close to 16 pounds of weight and a steel plate on to the bottom of our model, and it still floats.

Model floating in the tunnel pictured above. So I sat there and I took a hot wire to the foam to start cutting it out and lower it's lifting force.

And it didn't work. The thing still floated. Then Doig came along and threw out this idea. Test with the model upside-down so it can float.

That's a brilliant idea, and it's the only thing I can think of that will work, so that is our Monday plan. If it doesn't work... Then we might just passively be SOL.

[bitter-307-blog]

Week 8: Things float.

Okay. This week was all about that mini project. Monday, our model had not started printing yet, so we jumped ahead with building a ramp so the flow would be less disturbed going over the panel gap. We dug around the wind tunnel looking for scrap, and we ended up settling on foam.

Cutting foam into proper shapes was something none of us had ever done before, but we knew where the cnc hotwire was.

We had to figure out how to create G code for the wire, but it wasn't that complicated once we got a little guidance from people who did know what they were doing.

Two triangles like that and we were in business.

So we put our model and the ramps in the tunnel flush to the sides and gave it a shot. Unfortunately, everything floats. Foam, plastic, even when it was jammed into the tunnel it floated up so we couldn't test.

Plan B: Glue the ramp and test section to a piece of metal and put that in the tunnel. However, after sanding it and finishing the glueing we ran out of time with the tunnel. So we will test at the next available opportunity.

[bitter-307-blog]

Week 7: Really though

Our new project is to investigate the panel gap on a Tesla and the aerodynamics associated with it. Having a project like this is excellent. It is real engineering. We were given a problem, and told “We don’t really know what is happening here. Figure it out and give us some good flow viz that shows it.”

That is a gift. We get to design our own test and make it work. Our test will consist of a 3-D printed version of the panel gap. The gap that we are interested is only 4 mm wide, so there is not scaling down that can be done. Kendrick learned in his flow viz testing that 3-D printed stuff floats, so we are going to glue it to a metal plate that can go on the bottom of the tunnel. The plastic will be sanded down to a very smooth finish, and it will be covered with clear coat so that it matches the finish of a Tesla.

Testing in the water tunnel, we will not be able to match Reynolds number. We were not asked to. We are being told to create some visuals, so that is what we are going to do. Monday, we will make sure the model construction is underway, and we will address any other hiccups that could emerge along the way.

Hope you liked the sonnet.

[bitter-307-blog]

Week 7: How We Spend Our Time

I walk into three hundred twenty-two.

The tensions high because of 405

But now we stop to do 4412

Or else the Doig will skin us all alive.

The difference in the aspect ratio

Creates a sooner stall on the “red wing”

The stall emerges from detachéd flow.

The force balance, she is a fickle thing.

Yeah I forgot to zero out the weight.

It could have ruined all of our data.

By Wednesday our report will be so great.

We’ll have it in at midnight no latah.

“To read an angsty post just read Bennett.”

So I respond with a bitter sonnet.

[bitter-307-blog]

Week 6: Finite Wing and Presentation

Not a ton to talk about this week

Starting of the week was the lovely day of flow viz presentations. Our presentation went well because we were well prepared. Simple as that. Sure, we didn’t win the flow viz challenge, but that is because we didn’t 3-D print an F-16. If I knew that was how to win the presentations, I still wouldn’t have done it because I don’t have that kind of CAD skill or a 3-D printer.

(Okay, the F-16 was cool, sue me.)

That being said, we had some lovely visuals of a baseball with dye flowing around it all pretty-like. For any reference, just check my previous tumblr posts. All my best pictures are there and I am not a reposter.

Wednesday comes rolling around, and we got to spend the first hour and a half planning on how we wanted to do our testing. Not going to lie, our group was crippled. The wind had really picked up, and there is no drug on the planet that can prepare a man for SLO Spring allergies. Couple that with the wind tunnel itself really picking up dust and moving air around even more, and we were toast. Before our group did any testing, we were building our test matrix and some code to run the data. Ben Kenworthy was the king of the code, but we all chipped in with commands and indexing advice here and there. We came up with some angles of attack that would hopefully paint a strong Cl vs alpha plot for each aspect ratio wing. I did the angle of attack adjusting for the first wing and did a great job watching it for the second one.

There was a little mishap (my bad) with forgetting to zero the force balance when we added the wing on, but we got that data afterwards, so it should be all fine.

Anyhow, Yev and I also both took pictures of the wing at the same time.

[bitter-307-blog]

Week 5: Writing a lot

Like a lot

Well, this week I didn't get to do any testing in the wind tunnel which blows (Pun intended.) That being said, I just had to write a bunch of stuff for my flow-viz presentation and for the NACA 4412 report.

So, Monday was all about doing that NACA 4412 report. In my personal opinion, we are being too long winded with how we handle it, but I’m not going to say something about it and come off as lazy. I’ll write my sections as requested and keep on keeping on with that. Getting exact data to measure our experimental data against is tricky and is mostly interpolating from airfoiltools data. A lot of it is a non-exact science, but we are still able to get a good comparison to our data. But if you (Tynan/Doig) have any more questions about that, then you can check out my 4412 report that you have by now.

Wednesday:

Pixel counting is the best.

In all seriousness, this is how we got a decent amount of our quantitative data for our flow viz.

We counted (measured) the pixels across the ball to get the percentage of the surface of the ball where the flow stayed attached. Pretty fun huh? There is a significant difference in separation points when the ball is spinning vs when it isn’t (About 11 percent) That difference Unfortunately, on our scale and with our viz, we were unable to draw any conclusions on differences in the flow with different seam configurations. From a baseball perspective, I can tell you that there is a difference. From science though, I have no evidence to offer.

Another thing I worked on with our visualization is getting the frequency of the vortex shedding. This took a little bit of guess work and we did used expected Strouhal Numbers to help us figure out what we were looking for, but our data was better than expected. We got a Strouhal Number of .2143 compared to published data on a sphere that is about 0.19. For there being a stick holding it in place and a bunch of seams that theoretically matter to the spinning ball, this is well within the acceptable margin.

I wish we could have gotten to better and more exciting conclusions, but working with dye at a Reynolds number that isn’t even close to the real life situation, this is it.

All of that being said

Man it looked cool.

[bitter-307-blog]

Week 4: Baseball

Well, we worked on the water tunnel today, and we have it down to a psuedoscience. I know how much I need to pump the dye to get a clean flow in the tunnel. (Side note, Tynan just ripped on someone for having bad grammar. Watch your ass Bennett.) We also learned that the tunnel takes forever to drain. Using just the hose, it's about an hour and a half. We can probably push that back to ~30 minutes using buckets to drain as well.

After working in the tunnel, I used remaining time to prepare our baseballs for their water tunnel testing debut. 

Tada! Before I knew it, I had two baseballs on sticks. Each baseball was set up to spin in either a "two-seam" or a "four-seam" configuration. The one pictured here is the four-seam.

So, come Wednesday it was time to run the tunnel for real. I "figured out" how to get a decent stream of dye when we were messing around on Monday, so I pumped the dye again. We got some really cool results from the dye. Josh made a grid to put behind the tunnel, and it makes our footage look much more official.

When the ball was spinning, the flow stayed attached significantly longer regardless of the seam configuration. I will have to study the footage more to back out the RPM of the ball as well as any qualitative difference on the seam configuration.

Outside of the data we gathered it was worth noting a few things about the water tunnel.

BE CAREFUL WHEN YOU ARE FILLING AND DRAINING!

We got water everywhere. I went to my car for a towel midway through testing to continue cleaning. I did everything I could to clean it up, but I know I did an imperfect job. The group after me got chewed out for leaving the place a mess, but part of that is on me.

Testing took up the majority of Wednesday, but afterward I checked in with my infinite wing lab group. Our report is due Monday 4/30 at midnight. Everyone has been just pecking away at the tech memo this week while I did flow viz. I feel a little guilty about that, but I’ll make it up to them while they do more of their testing on Monday.

[bitter-307-blog]

Week 3: See you next wake

Not sure if Tynan was serious or not when he said I could vlog. We’ll see about that later.

Also, sorry for the horrible title.

Monday: 

Back in the tunnel with the classic 4412 airfoil. We were using a pressure port behind the infinite wing to measure the wake generated. This, was a trial of patience. We decided to move the pressure port two mm at a time until we got close to the trailing edge, and then one mm at a time after we crossed the trailing edge. This involved a lot of very small movements of the pressure port and many many files.

We tested at 15 m/s and 30 m/s for an angle of attack of 7 degrees. Seven degrees is supposed to be the angle of attack that yields the greatest Cl/Cd. We get the total drag from the wake, so we will see how our measured Cl/Cd is to the ideal once we analyze our data a little more. Below, is our wake measurment for 15 m/s and 7 degrees aoa. (Forgive me father for I have no graph axes) The x axis is mm from the edge of the test section that the pressure port was located.

For this run, the wake was relatively clean. However, when we increased the angle of attack to 14 degrees, our wake became more turbulent and less consistent. Still at 15 m/s and 14 degrees

So less consistent pressure changes in the wake of the wing, but this can be expected as 14 degrees is on the edge of stall. After testing some members of our group worked on implementing our new data into Cl/Cd code, and some worked on beefing up our test plan. I beefed up our introduction, abstract, and testing procedure. I also kept Louis from going insane working on Cd code.

Wednesday:

I walked in three minutes late and Tynan stared me down. It seemed all in good fun, but he can expect the same treatment if he is ever late.

Matthew Walters and Joshua Matthews are my noble and proud group members. We have decided to test the flow characteristics around spinning baseballs in the water tunnel. We will test two different spin configurations at two different RPMs. We have a little bit of work ahead of us as far as buying baseballs and drilling into them, but we’ll get there when we get there. We checked out the water tunnel, and we are pretty sure we can make it work. We have to spin from the top, which is unfortunately our primary filming angle as well, so how we handle our visualization will have small amounts of trial and error.

[bitter-307-blog]

Week 2

You know what happens when you make reliable code and go to the wind tunnel with a strong test matrix and a plan of action? Something goes wrong. When the wing was reinstalled over the weekend, the pitot-static tube was hooked up incorrectly. Initially, the lab view had our wind at a speed of about 15 meters per second. However, the hot wire showed a reading closer to 25 meters per second. With a little bit of troubleshooting and help from Tynan, we started collecting data.

During our first run through at varying speeds and angles of attack, I was the guy on the computer looking at our data in Matlab. Comparing our data with what we were hoping would be there based on airfoiltools.com as well as Xfoil. I noticed some small flaws in our code that were giving us weird plots, but after investigation of the values we were getting from the ports some more ports had to be averaged out.

For our next cycle through testing, I was on the other side of the tunnel adjusting the angle of attack of the wing. Not going to lie, I messed up the first two because I didn't see the little paper arrow pictured above. We got data for those values at the end, but it was a blunder that cost us a little time.

Wednesday came around and we were coding away. Louis and I worked on the panel code trying to find Cl/alpha plots that were reasonable based on our testing Reynolds number. At points, we were getting a maximum Class as high as 2.6. However with careful observation of what our code was actually doing, we found something that worked.

We are still struggling to get good drag numbers, but we plan on using data from the wake portion of the lab to supplement the data from our first tests. Using the wake data for drag, we should be able to get decent Cl/Cd data.

[bitter-307-blog]

Week 1

Welcome to Aero 307. A two unit class that somehow justifies 6 hours of class time per week. I never understood how a 3 hour lab counts towards 1 unit of class, but a 3 hours of lecture goes to 3 units. But I digress.

Lab 1 involves testing using a NACA 4412 airfoil in the Cal Poly low speed wind tunnel. The experiment that our lab group got to run was up to us, so I suggested that we find the best Cl and Cl/Cd of the infinite wing at a few speeds.

We’ve calculated lift and drag from the wind tunnel before, so it didn’t seem like it would be an issue. However, when we did it in the past during the Aero 302 winglet lab, our math came from the load cell. The data that we have available to us is pressure from 20 ports unevenly spaced around the wing. 11 on the top and 8 on the bottom plus one more right on the leading edge. We understand that the Cd that we can get from the pressure data will only include the pressure drag and not any of the viscous drag, but it will have to do for now.

My first individual task was to make Matlab code for the Cp plot.

Nailed it.

The real struggle comes from trying to get the Cl and Cd from the pressure points. To try and find the direction and area associated with each pressure port, we have to develop some rudimentary panel code. Making that all work with the angles, panel lengths, sines, and cosines is, believe it or not, a huge pain in the ass.

But, with a little elbow grease and some matrix dimensions that may or may not agree, we will have code that spits out Cl/Cd by the time that we test on Monday.