How can I design an affordable test rig for a wind tunnel?

[h2chohan]

Hey,

I have to design a test rig for my university's wind tunnel. The test rig is required to measure lift, drag, pitch moment and rolling moment. The professor has asked me to keep the budget around 2k. I was looking online about what is available and came across 2 types of balances. Internal and External. Internal balance has a 6 component balance system which usually cost 43k with the sensor and you need to design a sting for that. It is way over the budget.

The previous design could only measure lift and drag. The picture is attached to the forum. Wind tunnel is 2' by 2'

Can anyone help me with how can I design a new test rig which can measure forces and moments too. I was told to try and keep 1 support because more the supports it will effect the airflow.

Harry

 

[h2chohan]

The weight of the model (plane) which will be tested is 200gms

 

[Cyrus]

I would get yourself a copy of Wind Tunnel Testing by Pope, Rae and Barlow. Look in Chapter 7, references. Doing this thing under 2k seems formidable, because I can see your load cell alone eating up close to 500 bucks, and a DAQ board another 500-1000, which leaves very little for materials and machine shop time.

 

[h2chohan]

Thanks a lot for your reply ... Some1 recommend me to use multi component Load cell as it can measure forces and moments
http://www.ati-ia.com/products/ft/sensors.aspx
I am not sure how much it will cost. If i don't find any solution within price range ... I can go upto max 5k dollars.

With the multi-component load cell ... Right now I was working on how i am going to connect plane/model to the cell.

Any help is appreciated .. Thanksss

 

[KalamMekhar]

What will the rig be made out of? Aluminum would be my first choice, PVC second. It wouldn't be that hard to make with the diagram provided, and Alum isn't very expensive at all. Something like 18 or 16 gauge would be just fine for your purposes (assuming the wind tunnel doesn't have too high of speeds). You could whip one of those up in an hour. Aluminum is going around 1$/lb ingot, so let's say around 100 bucks for everything, including nuts and bolts (depending on the size of the rig, I am thinking of something around 12" high, 14" long). The sensors are going to be where most of the money will be going to.

 

[h2chohan]

Thanks Kalam, I am thinking of aluminium too ... But If i use a multi component load cell to measure force, drag, rolling and pitch ... I would require only 1 cell .. can you help me up with the design. How should it look like?

 

[KalamMekhar]

Well, The design I drew up real quick like on pad and paper had the sensor on the base plate, with a vertical shaft that connected to the object. It would have a low friction ball socket support beam with a u-shaped teflon bushing to hold the vertical support beam in place. It shouldn't be super hard to design, but a 2 sensor (one is roll, pitch, lift. One is force, drag, etc.) would be much easier to implement, and a lot less head scratching.

 

[Cyrus]

If you want pitch and rolling moment, I would use four load cells, and back out the moments based on the geometry. You can increase the sensitivity of the measurement this way.

 

[h2chohan]

Yea test rig needs to measure 4 essential measurements ... Lift force, Drag force, Rolling moment and Pitch moment. Is it possible to have 1 support to the model and use 4 load cells to measure 4 components?

If i use 3 vertical beams to support the plane .. one at the tail .. the other two along each wing ... It will effect the airflow and the values. My professor is like try and keep single support.

 

[h2chohan]

Kalam,

I didnt understood "It would have a low friction ball socket support beam with a u-shaped teflon bushing to hold the vertical support beam in place". I can't picture that in mind.

 

[Cyrus]

No, you have one exposed support strut and windshield in the flow. Outside and below the test section is where you have the four load cells. I am not saying that each load cell measures Lift, drag, etc. I am saying two of them will measure a moment, which is roll. All four will add up to measure lift, etc. Its not a one-to-one correspondence. Get the book I referenced.

 

[Phrak]

Yea test rig needs to measure 4 essential measurements ... Lift force, Drag force, Rolling moment and Pitch moment. Is it possible to have 1 support to the model and use 4 load cells to measure 4 components?

If i use 3 vertical beams to support the plane .. one at the tail .. the other two along each wing ... It will effect the airflow and the values. My professor is like try and keep single support.

I agree. You need four measurements, no matter what these measure directly. I assume there will be the capability to post-process sensor data.

Without transducers internal to the model or the use of telemetry, at least 4 structural elements will have to couple to the model. Say, you have a single vertical support beam. This is one structural element, and sufficient to measure lift. The obvious choice is to use rods, internal to the support beam. The obvious is not always the best, but I can't see any other method right off hand.

 

[sunshine4all]

Hi,

I have to design the same test rig. Anyone who can help me. I am going through the book ''Low Speed Wind Tunnel Testing''. h2chohan, did you get it work?

I have to measure lift, drag, pitching moment and roll moment.


Thanks in advance

 

[Wah]

hi,
im making a device that will measure lift and drag on different aerodynamic shapes in a wind tunnel. i have been researching on sensors, Dual-Range Force Sensors, but however as this is a project for my studies, i can't simply just buy the sensors and software to show it digitally. i think i have to make it from scratch. i don't know how else i can go about this to do it.
can someone please help me!

 

[FAlonso]

How about measuring lift by calibrating the force developed in a spring that gets compressed due to lift and you could calibrate the deflection for the force it has actually been acted by

 

[sunshine4all]

I think the response time is the issue. Specially for the dynamic case. Even the flexures are not resolving the problem of response time.

 

[DylanB]

Not to mention deflection issues. You need the force balance setup to be fairly rigid so that the flow is not affected, or deflects out of alignment so that you may be measure lift as drag, and so on and so forth.