Baby's First Wind Tunnel

Of Collars, Latches, and Ebay Items

2009-02-22T11:36:00.001-08:00

Three hitherto unsolved problems have been solved since the last post:
* Collar Mechanism
* Airspeed Measurement
* Screen Attachment

Collar Mechanism:
It's pretty easy to find relatively cheap latches with ~1/4 in. of travel (despite the calculations all being done in metric, much of what I work with is in English units), which can be used for compression-purposes. Nails and drilled holes do a pretty decent job as dowels and guides, and surgical tubing provides both the seal, and the resistance to the tension provided by the latches. The end-result is a simple-to-manufacture, interchangeable, quick-release collar. The
specific mechanism will be described in more detail on the website.

Airspeed Measurement:
Finding a pitot-tube anemometer is pretty simple, but finding one that fits in my budget is not so simple. Thankfully, the design airspeeds for the tunnel are known, and good Pitot Tubes are reasonably priced (~$175 for a pitot+static tube with pressure outlets), and ~$150 for a decent differential pressure gauge. From there, using a local machine-shop (or one that a friend of mine has access to), and suddenly you have a differential pressure gauge that shows the dynamic pressure. Combined with absolute pressure (from a barometer) and absolute temperature (from a thermometer), and dynamic pressure can be turned into airspeed.

Screen Attachment:
As Barlow et. al. put it, screens have an uncanny ability to collect dust, and will need to be cleaned, frequently. Given that stipulation, it seemed a bit daft to place the screens in positions that would preclude easy access. It was suggested that I make the screens removable, like a lint-trap in a dryer, but even that seemed like a bit too much work, because then the screens must be securely mounted on something that can be readily removed and reinserted, and seal properly -- even then, the number and position of the screens is variable enough at this stage to warrant a different approach. The collar requires ~3cm per section to operate properly -- I've been using 5cm for good measure (and round numbers). With Sections that small, I've simply decided to make each screen its own section.

Unrelated Comment:
Working with a battery-operated circular-saw on a deck can be quite obnoxious, I'm considering upgrading to a chop saw and a table saw...In the purchase queue they go...

Of Wide-Angle Diffusers and Kitchen Equipment

2008-11-23T14:53:00.001-08:00

The first component that will be designed and built is likely to be the wide-angle diffuser. Now that I am resolve to building a blower tunnel (with the option of attaching a smoke unit). I'm ready to start construction etc. While reading up on wide-angle diffusers, I noticed that we have (as of the publishing of the 1977 paper I'm reading) absolutely no idea how they work -- all our techniques are emperical.

Accepting the fact that I'm going to have to use emperical methods, I've taken a look at some historical data and decided that a value of 25-degrees seems pretty reasonable for 2-theta (Mehta, 69). From there, I'll have to use emperical methods.

On an unrelated note: I'll soon be purchasing a centrifugal blower from lowes or something for testing and development, and an airspeed meter.

REFERENCES:
1. R.D. Mehta, THE AERODYNAMIC DESIGN OF BLOWER TUNNELS WITH WIDE-ANGLE DIFFUSERS; Progressive Aerospace Science, 1977, Vol. 18, pp. 59-120.

Welcome Back

2008-10-20T19:49:00.000-07:00

The old is gone and the new is in. Happily settled into the new job, I am now back to working on the wind tunnel.

After much though, I've decided to drop the ambition-level of the project from a closed-circuit wind tunnel to an open-circuit tunnel. Doing so will, eventually, afford me the flexibility to turn it into a smoke tunnel, and, in the mean time, it allows me to make the test-section larger.

Back to the proverbial drawing board.

Compartmentalized

2008-08-13T01:37:00.000-07:00

The debate of whether to make an open or closed-circuit wind tunnel has come to a close. The winner is: both? In the interests of convenience, I'm going to make the various components of the tunnel highly separable. I have a collar mechanism in mind, and will try to put the information up on the website, sometime tomorrow. For the sake of making the engineering easy, I'm going to try to have two primary sizes: test-section, pre-constriction (also known as small and large).

If you model the wind tunnel as a test section, followed by a really long diffuser, then a contraction, then the axial fan would, usually, go somewhere in the middle of the diffuser. In my case, diffusers are complicated to build, and so I'm going to try to make the diffuser fit between the test-section and the second corner. Doing so would allow me to hide all of the intermediate sizes in three sections (first diffuser section, first corner, second diffuser section). All other collars would be either small or large.

I'll start designing tomorrow, and start stubbing out a python file to perform the pressure-drop estimates, etc., and we'll see how feasible this plan is.

Design Order

2008-08-05T17:56:00.000-07:00

I have, now, had a chance to read a decent amount in the wind-tunnel design book (~50 pages). The equations and methods used in the wind-tunnel book are starting to click with those employed in the introduction-to-flight book, cementing my general aerodynamic understanding. My parents and brother will be arriving in just over three weeks, and I am hoping to have a component designed and ready for construction by the time they all arrive -- thus the design-order question.

The test-section is seeming to be the most logical thing to start with, as, for the initial version, I won't require the ability to take precise measurements (provided I have, at the very least, a good idea of how to amend that shortcoming later). The next component that I would foresee us working on would be the diffuser, immediately proceeding the test-section, as doing so does not require knowledge of the target size.

Partners In Crime

2008-08-03T19:47:00.000-07:00

Two of my swing-dancing buddies might be helping out, once construction begins. Autumn has done a great deal of wood-working in the past, and Nick is a fabricator by trade.

Moving On

2008-08-02T02:08:00.000-07:00

It does not appear as though construction of a high-reynolds-number wind-tunnel is going to happen. The wind-tunnel testing book states pretty clearly that small tunnels like this one will not reach reynolds numbers that even approach reality. This project is a bit of a hail-mary as-is, so I shouldn't be trying to push the edges of too many envelopes at once.

Brick Walls - Theoretical Limit

2008-07-31T12:51:00.000-07:00

So, apparently, most of the data of interest that can be collected from a wind-tunnel is going to be governed by the Reynolds number, not the freestream speed. Consequently, I need to design my wind tunnel to maximize the Reynolds number. Not only does the tunnel need to be designed to maximize the Reynolds number, the model does, as well. The Reynolds number (Re) is proportional to the length along the aerodynamic body. If your aerodynamic body is 1/20th the size of the 'real thing', then you have to up the other components of the Re by a factor of 20 to compensate...That puts the theoretical max for my toy truck's simulated speed at about 15 mi/hr...

Let the Posting Begin

2008-07-30T10:00:00.001-07:00

I ran my first 'experiment' last night with my first 'wind-tunnel'. The 'test-section' was the size of the room, and the axial fan used, maybe, 1/8 hp. It was very 'scientific'.