calpolyaero307tester-blog

[calpolyaero307tester-blog]

307 wrap-up

307 is coming to an end and I have to say the most memorable parts of the class were being able to design and conduct our own experiments. I especially liked working for ‘clients’ because it gives a small taste of what a real world testing job would include.

I think this class has prepared me for the uncertainty of senior year design and testing. We had to deal with a lot of problems in testing the NACA airfoils and had to find ways around them, which taught me a lot. Bring on senior design!

These two tunnel classes have gotten me quite used to working in different groups and I found it quite helpful to be able to interact well with them. I think the key to this is splitting up the work and maintaining good communication so everyone knows where the project is at. I’ve enjoyed all the group work I have done and it is nice to have other perspectives that sometimes one person can miss.

I am much more interested in aerospace testing after taking this class. This along with working at ES Aero’s testing facility have put me in hands on situations which I really enjoy. I’d be happy to pursue a career in this type of stuff once I graduate.

Thanks for a good quarter Dr. Doig, was a fun one!

[calpolyaero307tester-blog]

These are the quantitative filter results we obtained from the periodic spectrometer tests. Obviously there is some great inaccuracy/strange behavior in the middle but overall the trend-line has a negative slope. This gives some hope for the filter being effective. In this test we ran the pump at 3.5 inches per second, enough to pass the water forcefully through the filter, and we ran it extensively to see what the long term effects would be. After ten hours the color barely changed to the naked eye and was likely still too polluted to use again. 

The final verdict from the filter experiment is that some more research must be done on an effective way to filter out dye particles, because both filters, the old and new, were largely ineffective. For the long amounts of run time required to make little change is not worth even using the filter after the experiment is over. Changing the water at this point would be more effective.

[calpolyaero307tester-blog]

The last thing we were tasked to do was to utilize a UV laser to make a laser sheet in order to analyze the dye. The previous set up was just two UV  lights hanging in a hood which provided little to no illumination. Our hope was to create a much more intense UV reaction with a laser. We attached this to an acrylic rod which dispersed the laser into a sheet. We got good results from the build but because of the dye we had in stock were not able to create an acceptable UV illumination level.

The suggested fix for this is to mess around with different types of dyes that will illuminate well and have the correct density/solubility. More research will have to be conducted by the next investigative team on dyes as well as a mounting device for the laser.

[calpolyaero307tester-blog]

Our group was also tasked with finding out why the current UV dye mixture was not the best for use in our water channel. Our first discovery came when we added the dye straight to water and it congealed up and sank. Not only is this dye completely insoluble, it is much denser than water. These are not good characteristics of a dye to add to a water channel.  

Furthermore this is UV dye designed to be used to find leaks in car radiators and hoses. A much better suited dye could be found to use, but for this lab, we tried to improve the characteristics of the dye.

We did this bye adding various chemicals. Dish soap claimed to have soluble properties so we mixed it with the dye and some water and it actually managed to keep the dye mixed in. The soap was blue, which turned the mixture slightly green. This was the best combination we found, though the UV properties were not strong.

We also tried adding mineral spirits to the dye, to make it less dense. I then added water to those two ingredients which turned the mixture opaque, giving me good hopes for a strong UV mixture. This did not turn out as the mineral spirits made it less dense than water and the opaque clouds were barely lit up by the UV laser.

The conclusion Yoav and I made was that appropriate water soluble dye should be purchased for the best flow viz potential.

[calpolyaero307tester-blog]

Our first task in improving the water channel was to replace the old filter that hasn’t been changed in many years. Before doing this we recognized the dye filtering qualities of the channel were quite sub-par. This is the old filter we replaced with a new charcoal style 3M replacement, hoping for some improvement.

I checked out a spectrometer, beakers, and a LabQuest tablet from the chemistry department so we could quantitatively analyze the quailty of the water over time. We took samples after short periods and over much longer ones to get a good scope of the filters effectiveness. These samples were the absorbency of light at wavelengths corresponding the max value of absorbency. We then monitored this same region on the wavelength graph to see how the max absorbency changed over time. Results will be posted shortly. 

[calpolyaero307tester-blog]

Water Channel Improvement!

My next task in 307 is to improve the fluorescent dye characteristics of our previously used water channel. We are tasked with finding a new type of UV illumination, the best recipe of flouro dye, and a study on the dye removal qualities of the current filter installed. Hopefully we will come up with some solid solutions to these tasks. 

To improve the UV qualities we are investigating the use of a UV laser and an acrylic rod that will split the beam into a sheet.

To find the best flouro dye, we will be doing a good amount of research and testing with different diluting agents for the concentrated dye we already have.

To see how well the filter is filtering....we are going to take test viles at certain times after a specified amount is circulated through.

Will update as research progresses.

[calpolyaero307tester-blog]

In this video our “roof racks” are placed on the car and it is run at the same speed. This video is a lot more interesting because we can see the large effect of the roof rack bars. Then dye is seen hitting the bars and splitting into reciprocating Von Karmon vortices. This creates a much larger disturbance profile behind the car than before, so the low pressure zone will be bigger. 

It is also interesting to note the location of flow separation in this video compared to the last. Immediately under the first roof rack is where the flow separates from the car in this setup rather than further back at the rear window. This will contribute to a significant growth in drag. We also concluded that the vortices at a high speed would be quite loud in the cabin and possibly cause some vibrations.

In real life this effect will be mitigated somewhat because roof racks are usually designed more aerodynamically than the cylinders we used to mimic them. Regardless this test is a good one to visualize the flow around a very nice model car.

[calpolyaero307tester-blog]

Here is the first video we took of the Tesla Model S in the water channel run at 2.75 inches per second. Interestingly enough this produces a Reynold’s # of about 40,000. A real Tesla Model S at 40 mph produces a Reynold’s # of about 6 million, so this test is not accurate on that scale. It is however accurate for visualizing the characteristics of flow over a car and can provide some clues as to what a higher speed flow would be doing.

In the video it is easy to see the dye flowing over the cabin and separating near the back window, at which point it begins to circulate behind the car creating a stagnant, low pressure zone. This lower pressure behind the car compared to the higher pressure in front is what creates pressure drag.

[calpolyaero307tester-blog]

This last week we were tasked with creating our own experiment to do some good Flow Viz Analysis on. We decided to take a look at the Tesla Model S. Cal Poly was given this expensive 1/18th scale model of the Tesla Model S and its the perfect size to put in our water channel. 

At first, we wanted to visualize how the car would perform drafting behind a large object, like a semi trailer. We quickly realized there was not enough room in the test section of the water channel to do a meaningful test .

Our next idea was to analyze the flow consequences of adding roof rack bars to the top of the Tesla. I was able to fashion these makeshift replicas out of round, rigid wire. We ran the car in the tunnel twice.  Once with the bars and once without. We were able to get some pretty interesting video I will upload in the near future.

[calpolyaero307tester-blog]

The load cell data for the Blue and Red finite wings came out a lot nicer, especially the Red wing data which we were finally able to test after the wind tunnel sting broke the week we were supposed to test it. Our Blue wing data was off by a factor of about 1 on each side of the Cl axis and this was most likely due to an electrical gain from a wind tunnel traverse arm that was left on. This was fixed with the red wing which is why the coefficients are much more reasonable. Also we can see there is a much greater difference between runs of the red wing versus that of the Blue wing, we deem some of this to be because of the Red wing’s higher aspect ratio causing some small perceivable flutter at the speeds we tested.

[calpolyaero307tester-blog]

Our NACA 4412 report was due last Thursday and I finally found time to finish the panel method code analyzing the black full span wing. While the curves initially seem reasonable, looking at the Cl vs. Alpha curve, it never reaches a positive value like is expected. This is most likely due to an error in the calculation of forces around the airfoil. My initial attempt was done by calculating the pressure on a panel multiplied by the area to find the force. Had I had more time to complete the assignment I would have gone back and calculated the forces by using the Coefficient of Pressure, but the pressing deadline didn’t give me the chance to optimize it.

[calpolyaero307tester-blog]

During our winglet experiment, threw some smoke over the winglet area to see what kind of effects the winglet was having on the wingtip vortices. These arent the best pictures but you can see some of the swirling effect of the vortices. This means our wingtips were having some effect on the flow around the edge of the wing. For a definitive answer on performance improvement we will have to wait for the winglet load cell data.

[calpolyaero307tester-blog]

In between working at ES Aero all week I have been working on the panel method code to analyze lift and drag for the large machined NACA 4412 we had in the tunnel vertically. This picture shows how i am getting the slope at each pressure port location. Two closely placed dots on either side of the port location to get the most accurate slope. This was done using the NACA 4 digit airfoil equation. Yoav took our raw pressure and Cp data and analyzed it during the test, finding that one of the leading edge ports was blocked. We were able to interpolate around this to mitigate the error. I will be finishing up the panel code results this weekend.

[calpolyaero307tester-blog]

Here you can see Yoav and Cooper setting up the blue NACA 4412 for the winglet test competition. Besides matching colors, the sleek simple design looks good and should have a noticeable effect on the L/D ratio. The data will tell!

[calpolyaero307tester-blog]

Last weekend our group was able to finally 3-D print our winglets for our NACA 4412 wing section. They came out great! These took about 11 hours total to print and get ready for the competition.

[calpolyaero307tester-blog]

Last week our team tested the finite wing at the same conditions as the week before, but this time we swept a pressure rake across the flow coming off the back of the airfoil. We did this to determine the momentum loss caused by the wing interrupting the airflow. This is another analytical way to determine the drag caused by the NACA 4412 airfoil. Yoav is doing the momentum loss and Cp distribution graphs while I am determining the CL vs. Alpha curve with a panel code. Results will be up later this week!

Also Cooper has 3D printed our winglets, they match up perfect on the first try! We just need to sand them to get them smooth and then we will be ready for the winglet challenge this week!

[calpolyaero307tester-blog]

In the meantime we continued our testing of NACA 4412 airfoils. This time with the ‘Infinite Wing’ made of machined metal with pressure tap ports. This means we do not have to deal with the error of the load cell, which significantly compromised our data. We are analyzing the Coefficient of Pressure distribution over the wing to get CL vs. Alpha curves on this wing. We also did some flow visualization to see where the flow separates in increasing negative angle of attack.