Wind Tunnel Load Cell

[wonderboy]

Yes, this was the last thing I needed: another project. But I just couldn't stand seeing it get thrown away...

Long story: our local high school robotics team used to occupy one of the old school buildings in our district. The building was no longer needed for students due to population decline so all the extra curricular groups (including First Robotics) were able to use the building to house their activities. The robotics team maintained an e-bay store for fundraising and sold all kinds of donated items. Well, the district decided to tear down the building, displacing all the teams. The robotics team packed up all the essential stuff and left most of the "junk" behind, including a lot of the e-bay items that were just too hard to move. The wind tunnel was one of those items. I didn't want to see it thrown away, and I figure that I may be able to incorporate use of it at some point when doing rocket building activities with Scouts, science groups, etc (I've been involved with quite a few different groups over the years).

The wind tunnel is a LabVolt Model 40576. It is very basic, and really just meant for basic science class room lessons, but I figured heck, it's pretty cool.


So on to my question. The setup includes a balance which measures both lift and drag and displays the values on two separate LCD displays on the front of the unit. Only the lift is working. The drag force is not working. I opened up the electronics, and observed basically two identical sets of circuits one to read the lift load cell, and the other to read the drag loadcell. I swapped the amplifier IC (BurrBrown INA120) to see if it followed the IC or the loadcell, and it follows the loadcell.

I'm figuring the load cell was abused in it's former life. I'd like to find a way to replace it. The problem is I've never seen a load cell like this one and I can't find anything with google searching.

Here is the loadcell (removed from the wind tunnel):


The rod is fixed in the tunnel where I'm holding it. I "THINK" that the rod is the first load cell and measures the lift force, and the square piece at the end is the drag load cell. This one is fixed into the aluminum frame with two screws on the bottom. The test sample is mounted on a rod that inserts exactly where I've drawn the drag arrow (the sample extends to the left in the pic).

Here is a closeup crop of what I think is the drag (only) loadcell:


Anyway, anybody know what type of loadcell this is, or what type of load cell I'd look for if trying to replace it? I think I need one that would measure shear forces, as the loadcell is fixed at the bottom and measures force along the same axis. I've written to the manufacturer, and am waiting to hear back from them. But I'm not optimistic so I'd like to see if I can come up with a solution.

Thanks for reading!

 

[bjphoenix]

Does the load cell have strain gages on it? Those work by changing resistance so it's possible there could be a wiring issue.

 

[wonderboy]

Yes, I think that's exactly how it works. I agree to the wiring issue. However, typically the strain gauges are a very tiny fragile component as is the wiring that connects to it. My assumption is that this is damaged. I think a repair of the strain gauge is outside of my skillset (at least without having a microscope).

On the bright side, the company that makes it did e-mail me back and are in the process of generating a quote for me to get a new "balance" as they call it. I hope this doesn't mean it'll be cost prohibitive.
Fingers crossed!

 

[prfesser]

Load cells are doggone cheap these days---the $10 kitchen scale I have contains one, as well as all the associated circuitry. If the company can provide some specs for the load cell, it might be worth buying one that approximates those specs and seeing whether it will work. Hopefully the dimensions of the cell aren't critical as cells vary in size. Beam-type cells seem to be the cheapest.

 

[Kelly]

We used to make our own load cells - it's fairly simply, just involves gluing a strain gauge down to an appropriately shaped piece of aluminum. Here, you already have the aluminum piece, so it might be repairable just by scraping off the old gauge, and gluing on a new one. Something to consider if the quote from the manufacturer comes back unreasonable.

 

[wonderboy]

We used to make our own load cells - it's fairly simply, just involves gluing a strain gauge down to an appropriately shaped piece of aluminum. Here, you already have the aluminum piece, so it might be repairable just by scraping off the old gauge, and gluing on a new one. Something to consider if the quote from the manufacturer comes back unreasonable.

That's good to know! I was curious about it. I know at work we have instrumentation labs that will attach strain gauges to shafts (for example). I wasn't sure how involved it was.

I noticed after inspecting the existing balance more closely that my initial suspicion was correct. The long bar measures the lift, and the aluminum square measures the drag. I saw on the bar that there are 4 strain gauges located along the length. Two on top, two on bottom. One set close to the square, once close to the mounting end.

I know I can google it but since we're talking about it here: is this a common arrangement for strain gauges? Is this a "wheatstone bridge"?

I don't want to dissect the rest of this yet, and the design of the balance obscures my view of the drag load cell, but I'm assuming that there is also an arrangement of 4 strain gauges on this as well.

Thanks for the info!

 

[Kelly]

Is this a "wheatstone bridge"?

Yes, exactly.

 

[shockie]

Go to their website abd and contact them. I'm sure they can give you the specific specs for the parts you need.

 

[prfesser]

We used to make our own load cells - it's fairly simply, just involves gluing a strain gauge down to an appropriately shaped piece of aluminum. Here, you already have the aluminum piece, so it might be repairable just by scraping off the old gauge, and gluing on a new one. Something to consider if the quote from the manufacturer comes back unreasonable.

This brought back a memory. Back when load cells were pretty expensive, say $500 for a decent 100 lb cell (more like $20-30 today) I thought about making my own from strain gages, which were much cheaper. When I read the directions that the gage company had for dealing with the substrate and attaching them I decided no-thanks.

Yes, you can just glue strain gages to a piece of metal or whatever...but for good linearity, reproducibility, etc. there seemed to be a lot more involved. The substrate should be strain-relieved, the surface was to be carefully prepared (I don't recall the details) and for a wheatstone bridge arrangement, four identical gages were needed, carefully located. Notice that most load cells aren't just blocks of metal but have holes milled in a particular pattern to produce maximum response from the gages.

That doesn't mean you can't make your own load cell. And since you already have the substrate it'd probably be a fair bit easier than starting from scratch. Still, for optimum performance I'd personally try a $20 commercial unit.

Disclaimer: as I said, I've not used strain gages, simply because the instructions for a good setup were more complex than I'd expected.

 

[Cape Byron]

The wind tunnel is a LabVolt Model 40576. It is very basic, and really just meant for basic science class room lessons, but I figured heck, it's pretty cool.

Some of us get excited about free lawn furniture on the curb.

You get a freakin' wind tunnel.

 

[prfesser]

I know I can google it but since we're talking about it here: is this a common arrangement for strain gauges? Is this a "wheatstone bridge"?

A Wheatstone bridge isn't specific to load cells; it's a circuit with four resistors, balanced and arranged as shown below (Wikipedia). In most load cells four gages are used for the resistors, for greater sensitivity. The gages' resistances change when flexed. It takes very little change in resistance to give a significant and very accurate change in I(g) or V(g).

(Useless memory: Some decades ago I was emphatically told by a supposed knowledgeable person that "It's a strain GAGE! G-A-G-E!" Wikipedia says either gage or gauge. So I learned something new today.)

 

[rocket_troy]

I used a strain gauge for pressure sensing for my deployment device - SG mounted to the side of the tube to measure the material strain as the internal pressure changed. The SG was the actual Wheatstone bridge as it was a full bridge (4 way) SG. However, these wouldn't be suitable for load cell applications where you actually want your reverse SGs on the compressive side of your bending beam.
For some applications like mine, it's actually pretty straight forward to implement a full bridge SG but for something like a load cell where you want good linearity and temperature compensation, then it's much wiser to simply purchase a ready made one - and don't trust the really cheap Chinese offerings if accurate linearity is important to you (well, that's my experience from the past).

Anyway, here is a brief description of my application with a spreadsheet I whipped up to simulate.

www.propulsionlabs.com.au/Pyroless_Release/page9.html

TP

 

[bjphoenix]

When I read the directions that the gage company had for dealing with the substrate and attaching them I decided no-thanks.

When I worked in the research lab in college I glued on a lot of them. One of the projects I was working on involved testing large anchor bolts such as for highway signs. I would file a flat spot on the side of the bolt, glue on the gage using a form of superglue with accelerant, solder on some wires and secure them to the bolt with tape and/or wire ties. I would lay the little gage on a smooth surface, put a piece of clear tape over it and pick it up, put the tape on the bolt with the gage carefully positioned, then peel it back and put on the 2 glue components, then push the tape back down with my thumb. I would try not to glue my thumb down. Generally this worked well but the little gages weren't 100% reliable.

 

[Tech 68]

I would need a refresher course on strain gage/wheatstone bridge instrumentation to remind me of everything I forgot working flight test for the Army in the late '70's.
Doing a quick look on the internet I see you should be able to find replacement gages of assorted configurations you can mount. I recall cleaning/polishing the mounting surface and then using a mild chemical etchant on the surface. I assume you're sure the gage(s) are bad, and not the electronics? Lastly, if you do replace the gage(s), you'll have to calibrate the load cell with the support electronics.
I don't think the gages themselves are that expensive, and you can probably find some cheap surplus ones on eBay you can practice with on a scrap of aluminum.
I suspect you won't have that much trouble with the project... after all, you build rockets.

EDIT: I reread your first post and see you already tested the electronics as good... only other easy possibility is that there is a connection problem. A genius electronic tech I worked with was adamant that 70% of electronic failures were mechanical in nature.

 

[rocket_troy]

A genius electronic tech I worked with was adamant that 70% of electronic failures were mechanical in nature.

I wouldn't argue with that - at least for 70s-80s electronics. I remember my work experience time at Data General back in the days when it was feasible to repair stuff. 1st thing we did with monitor repairs and circuit boards of similar nature was to spray the PCB with freeze spray to see if that helped - which it often did; which I think (was a long time ago) suggested a mechanical issue with a solder joint or cracked something.

TP

 

[wonderboy]

Sigh... good grief!

After having my e-mail bounced back and forth within the corporate bureaucracies since I made my original post here, on about the 10 try, I FINALLY got a quote for the new load cell.

You guys sitting down for this??? Guess how much they wanted?

$1600!

I'm afraid that may be the end of the line for the "free windtunnel". I did read up on the installation of strain gauges and am very slightly considering buying the chemicals and glues necessary to prep and bond a new set to this loadcell. High quality strain gauges are a bit pricey, but amazon has quite a few that are affordable (in a convenient 4 pack). I'm just not sure this is all worth it just for a toy to screw around with. Did I mention I already have too many projects?? Yeah, yeah I know: first world problems...

EDIT: I think as has been mentioned, it's time to dig a bit deeper and perhaps I've just got a broken lead wire somewhere in the system. Wouldn't that be nice?!

 

[Tech 68]

Did you get the manuals for the system?
Even if you have to install your own strain gages to make the load cell work, as long as it's clear what the force axis is and the full scale load, you should be able to make it work and calibrate it.

 

[wonderboy]

I got a very basic user manual, but nothing with too much technical information. They do indicate that full scale drag measurement is 2 kg.

I need to type up a much more detailed response when I get some time. I've noticed a very interesting arrangement of the strain gauges (4 for drag, 4 for lift). It seems that there is some intermingling of the strain gauges between the drag load cell and the lift loadcell, perhaps to decouple the readings (in order for lift to not register on the drag loadcell). I'll post more details and pictures once I find some time. It's gonna take some diagrams to explain what's going on I think.

Oh, and to cut to the chase: I think I've found the culprit:


Repairing / replacing it is a whole-other can of worms though. Stay tuned...

 

[user 41914]

Good afternoon, I came across your post while I should have been doing more important things. I have exactly the same tunnel and had exactly the same problem. It turned out that the potentiometer was dirty so I just turned it all the way from one stop to the other and that wiped it well enough that it's working again. This might work for you as well. (Of course I tore the whole thing apart before I decided to try the simple thing). If it doesn't work, one of your posters was correct that you could gut a kitchen balance for the load cell and display without having to do anything to the electronics.
We're in the process of setting ours up as an educational display for an aircraft museum. It would be really cool if we could tap the lift and drag signals to plot them interactively, but that's beyond my experience as an aerodynamicist.

 

[user 41914]

I'm unsubscribing from the list but if anyone wants to contact me about this post I can be reached by email at [email protected]
Good night.
Nick