Vertical Test Stand

[Richard Dierking]

I'm creating a vertical test stand to eventually test research hybrid motors. I want to test solids first to work the bugs out, and I have a larger version I would like to build.

Following are some photos. I ordered a 50 Kg button cell from Robotshop and hasn't arrived yet. Because the stand is vertical configuration, I installed a pneumatic cylinder to balance the mass of the slide (80/20 linear bearings) and whatever I'm testing on the stand. So, the cylinder is pressurized to bring the slide up to the load cell. Then, I run the tare program on the Arduino (actually an Adafruit Metro) to zero the read before conducting the test. The analog to digital converter/amplifier is from Oscar Davidson in the UK. He has a video showing his ADC on Youtube and reached out to him and eventually purchased one. It's capable of a high sampling rate.

And, because the stand is vertical, I can use it as a launch pad with 20 mm mini-rail. Also included a photo of that.

The question I have is regarding load cells. The S-type I originally got for this was not working out. Thought it would be great because it can be secured top and bottom. But, any movement side to side would cause readings to vary up to 200 grams. It was an inexpensive load cell. And, the load cell I ordered from Robotshop is cheap too frankly. The Omega and Honeywell load cells are about 6 times the cost (about $400). Are they really that much better? Should I "invest" in a better load cell?

 

[o1d_dude]

Rocket science!

 

[Steve Shannon]

Prove the concept with the cheap load cells and then see whether you need the industrial grade ones.

 

[user 30299]

You can find some good load cells on ebay. The picture is a new 1500# load cell I picked up last year for $75. But it was 2-months of keeping an eye out for this size cell. If you're not in a rush, you'll find quality load cells at very reasonable prices.

 

[Richard Dierking]

Wow, you got a great deal on that load cell!

I'm going to do what Steve suggested. Try the cheap buttons cell out with a static fire of a CTI G-33 and see what the data looks like.
With vertical static testing (pointing down) there are some considerations for a reduction in mass as the motor burns, but I'm going to try and work that out.
Also, the first test will be in a 29 mm TVC mount. Servos on but not actuating and straight down. I want to see if the TVC holds OK and be able to compare to thrustcurve.org data.

 

[user 30299]

Here's a few more of the parts & pieces, and more spread around my garage. We were working on a motor test stand to be used in the physics labs in our local high school. But the virus put a stop to that project - for now.


You mentioned getting the A-to-D Converter from someone in the UK? That's the one item I have not been able to track down at a "reasonable" price. And the price seems to go up as the sampling rate goes up. Can you share some info on the unit you purchased - or send me a link?

 

[user 30299]

Wow, you got a great deal on that load cell!

I'm going to do what Steve suggested. Try the cheap buttons cell out with a static fire of a CTI G-33 and see what the data looks like.
With vertical static testing (pointing down) there are some considerations for a reduction in mass as the motor burns, but I'm going to try and work that out.
Also, the first test will be in a 29 mm TVC mount. Servos on but not actuating and straight down. I want to see if the TVC holds OK and be able to compare to thrustcurve.org data.

Check out NRI Industrial. They carry new and used load cells, and components to go with test equipment.

 

[beeblebrox]

I have been using the micro load cells you can get from robotshop. They are actually Phidgets, which you can also get from their site. I went there because the one I wanted was not in stock at the time. They are great, inexpensive and reliable. I got the Information for them from Scott Sympruch, President of MDRA a couple years ago. He likes the button ones, but I prefer the type shown because you can mount it with a bolt, although I don't know what thrust range you are looking to test. I have 5,20 & 50kg ones. Scott tests motors up thru "O" size. He makes the motors Tom Cohen flies in his gi-normous 600lb rockets. I do know that he does a lot of tests using a pressure transducer rather than a load cell for thrust. BUT, your one flaw in your design...Test Horizontally not vertical. The mass of the motor changing will alter your results. Also use 1010 Rail so it is compatible for your rockets at most all lunches in the US. Standard rail buttons, like Apogee, or Railbuttons.com fit that, not the 20mm square stuff that you see on 3d printers.

Note: load cell platform has four 1/2" holes at the corners, they are for 11 inch nails driven into the ground. They are strong enough to keep the unit from moving even with an N motor...

 

[Richard Dierking]

I'm using the 20 mm rail for a special project. Here's what will be tested first in the test stand and what's in the rocket.



First flight off the rail will be straight. For the second flight, I'm going to bend the rail toward the top and see how the rocket recovers using the TVC. It's model rocket size and mass with a 3 grain G motor. Anyway, that's why the 20 mm rail.

I would like to get the vertical test stand working because the hybrid will use LOX and have cells (for liquid) inside the 3D printed fuel grain. I know that some people place the grain horizontally for static test with the tank vertical and connected using oxidizer lines. Fingers crossed that the vertical stand will work without too many issues.

For the ADC, here's a link to the Youtube video:


Oscar describes the specs on his converter-amplifier in the video.

Please don't quote me on the price but it was about $70 and shipping. Maybe he will offer it through eBay in the future if there's demand for it. Here's a photo of the ADC:



I'm planning on collecting the data using a Adafruit data logging shield. I was thinking that log cables (>50') for data transmission might have problems. I would suggest waiting for me to collect some data before contacting Oscar. I will post my results.

 

[beeblebrox]

Cool, I'm using a 100ft ethernet cable for my set up, have no issues with noise using the Dataq unit with their software, usb to the pc. Amp and dataq units less than $75 total

 

[user 30299]

Thanks for the info. I'm a big Adafruit fan. Use their components on a number of small science projects. The picture shows a setup I'm working on with an aerospace student at the local university. Data collection add-on at the Base station for one of the rocket tracking systems. We're using the Feather modules.



Your concern about the 50+ ft. transmission is well founded. The load cell is putting out miniscule voltages. I have the same issue because of safety concerns with high school students. I plan to use the item in the picture to boost the voltage. I have not tested it yet. Attached is the spec. sheet on this Cooper unit.

 

[user 30299]

Cool, I'm using a 100ft ethernet cable for my set up, have no issues with noise using the Dataq unit with their software, usb to the pc. Amp and dataq units less than $75 total

Awesome. You just saved me some work !

 

[Richard Dierking]

Yes, that range is great! Always good to be in touch with data collection during testing, and at an appropriate distance.

 

[pbahorich]

I suspect that having the blast deflector close to the bottom of the rocket will change the measured thrust significantly. If testing vertical you might try putting the nozzle up with no blast deflector.

 

[Sooner Boomer]

You know, you can *make* your own load cells. Omega has great technical info on CD (or maybe DVD now), and the strain gauges are fairly inexpensive. Yeah, it's rocket science. But then, all of this is...

 

[Richard Dierking]

I suspect that having the blast deflector close to the bottom of the rocket will change the measured thrust significantly. If testing vertical you might try putting the nozzle up with no blast deflector.

This is an action-reaction question. Glad you are considering things like this.
The thrust out of the motor hitting the pad should not affect the reading. Well, unless it was really-really close and the blast was coming up and hitting the base of the rocket/test article. But, the deflector is at about 30 degrees. If it did, then the reaction would decrease as your rocket gets farther from the pad, and it doesn't. Same reason thrust hitting the atmosphere doesn't affect the reaction.

However, the thrust (mass of propellent hitting the pad at high velocity) could affect the movement of the pad because the blast deflector is attached to the base of the pad. So, I'm going to use 3 sand bags (about 180 lbs of sand) to hold the stand down.

It's very important to consider thrust hitting a blast deflector if the deflector is attached to the pad. Movement of the deflector can transmit all the way to the top of the rail.

 

[Richard Dierking]

You know, you can *make* your own load cells. Omega has great technical info on CD (or maybe DVD now), and the strain gauges are fairly inexpensive. Yeah, it's rocket science. But then, all of this is...

LOL, you know I'm already over my head about 90% of the time.
I am learning some electronics, which could easily be a hobby in itself.
So, are you a sparky? ;-)

 

[user 30299]

This is an action-reaction question. Glad you are considering things like this.
The thrust out of the motor hitting the pad should not affect the reading. Well, unless it was really-really close and the blast was coming up and hitting the base of the rocket/test article. But, the deflector is at about 30 degrees. If it did, then the reaction would decrease as your rocket gets farther from the pad, and it doesn't. Same reason thrust hitting the atmosphere doesn't affect the reaction.

However, the thrust (mass of propellent hitting the pad at high velocity) could affect the movement of the pad because the blast deflector is attached to the base of the pad. So, I'm going to use 3 sand bags (about 180 lbs of sand) to hold the stand down.

It's very important to consider thrust hitting a blast deflector if the deflector is attached to the pad. Movement of the deflector can transmit all the way to the top of the rail.

This is why some people point the nozzle straight up to the sky. The motor is pushing against the Earth - and you can't move the Earth. This also puts your motor casing soundly against the load cell thanks to Gravity. There are so many scenarios to setting up these test stands. I have quite a few pencil sketches!

 

[Steve Shannon]

This is why some people point the nozzle straight up to the sky. The motor is pushing against the Earth - and you can't move the Earth. This also puts your motor casing soundly against the load cell thanks to Gravity. There are so many scenarios to setting up these test stands. I have quite a few pencil sketches!

Inverted test stands are useful for solid motors, but I believe Richard wants to be able to test hybrids.

 

[user 30299]

Inverted test stands are useful for solid motors, but I believe Richard wants to be able to test hybrids.

Yup, I forgot about that. Thanks for pointing that out.

 

[pbahorich]

This is an action-reaction question. Glad you are considering things like this.
The thrust out of the motor hitting the pad should not affect the reading. Well, unless it was really-really close and the blast was coming up and hitting the base of the rocket/test article. But, the deflector is at about 30 degrees. If it did, then the reaction would decrease as your rocket gets farther from the pad, and it doesn't. Same reason thrust hitting the atmosphere doesn't affect the reaction.

However, the thrust (mass of propellent hitting the pad at high velocity) could affect the movement of the pad because the blast deflector is attached to the base of the pad. So, I'm going to use 3 sand bags (about 180 lbs of sand) to hold the stand down.

It's very important to consider thrust hitting a blast deflector if the deflector is attached to the pad. Movement of the deflector can transmit all the way to the top of the rail.

I am aware that rockets do not generate thrust by pushing against air or other things.
The concern I have would be the pressure field in the proximity of the blast deflector could change the way the converging/diverging nozzle converts the pressure into momentum (and the momentum produces the thrust). A high pressure area at the base of the rocket might reduce the momentum involved, and the resultant thrust. Rocket engines produce more thrust at higher altitude due to the lower ambient pressure. The lower pressure allows more expansion or conversion of pressure to momentum.

 

[Richard Dierking]

I am aware that rockets do not generate thrust by pushing against air or other things.
The concern I have would be the pressure field in the proximity of the blast deflector could change the way the converging/diverging nozzle converts the pressure into momentum (and the momentum produces the thrust). A high pressure area at the base of the rocket might reduce the momentum involved, and the resultant thrust. Rocket engines produce more thrust at higher altitude due to the lower ambient pressure. The lower pressure allows more expansion or conversion of pressure to momentum

That's interesting. Yes, that could happen. Btw, I could not tell you knew that only from your original post. I hope you realize that. I need to think that through and I guess I could even raise the set-up more. I'll measure the vertical distance of the motor nozzle to the blast deflector and post that too.

Inverted test stands are useful for solid motors, but I believe Richard wants to be able to test hybrids.

Yes, I would like to work out potential problems before hybrid testing. And, the hybrids I'll be testing are more like BiPro engines using Nitrogen as the pressuring gas and LOX as the oxidizer.

So, here's one thing already I found out. The slide is 934 grams (unattached to the stand). When I checked it again after installation, it was 930 grams. Probably not too bad, but obviously the mass didn't go anywhere so that's friction of the 80/20 linear bearings.
If you test inverted, retention is not a big issue with solid motors. But, there are construction considerations if you are also testing delay grains. For larger solid motors I think that horizontal configuration has advantages over inverted.

For vertical up or even horizontal config you have to assure your static test stays static. ;-) It must have freedom of linear motion and also be restrained that means undesirable friction in the system. I might be adding too much restraint, but here's a case of being cautious.
I already know the force of the thrust is not going through the center axis of the bearings, so that could be a significant problem. Nice thing with having the pneumatic is that I can simulate force of thrust by varying the pressure in the cylinder. I need to get a more accurate pressure gauge. The bore of the cylinder is 7/8" and Parker rounds it to 0.88". The cylinder spec is 150 psi max so I could simulate a lot of thrust.

 

[Sooner Boomer]

LOL, you know I'm already over my head about 90% of the time.
I am learning some electronics, which could easily be a hobby in itself.
So, are you a sparky? ;-)

I've spent the last thirty years or so working with load cells and associated "stuff" for measurement and control. Most of that time was spent in education or research programs at the University of Oklahoma, in the School of Industrial Engineering or the School of Chemical Engineering. I also worked at the National Severe Storms Laboratory, designing and buiding mobile radar systems and mobile mesonets systems (storm chase vehicles).

PM me a mailing address. I found an old CD from Omega with very good reference materials on it, I'll send it to you. It explains how the different sort of load cells work, puls it gets into the math of *why* they work the way they do. Omega used to send the same thi gs out as hardback books. A CD is a lot cheaper, I think.

 

[StanO]

I'm creating a vertical test stand to eventually test research hybrid motors. I want to test solids first to work the bugs out, and I have a larger version I would like to build.

Following are some photos. I ordered a 50 Kg button cell from Robotshop and hasn't arrived yet. Because the stand is vertical configuration, I installed a pneumatic cylinder to balance the mass of the slide (80/20 linear bearings) and whatever I'm testing on the stand. So, the cylinder is pressurized to bring the slide up to the load cell. Then, I run the tare program on the Arduino (actually an Adafruit Metro) to zero the read before conducting the test. The analog to digital converter/amplifier is from Oscar Davidson in the UK. He has a video showing his ADC on Youtube and reached out to him and eventually purchased one. It's capable of a high sampling rate.

And, because the stand is vertical, I can use it as a launch pad with 20 mm mini-rail. Also included a photo of that.

The question I have is regarding load cells. The S-type I originally got for this was not working out. Thought it would be great because it can be secured top and bottom. But, any movement side to side would cause readings to vary up to 200 grams. It was an inexpensive load cell. And, the load cell I ordered from Robotshop is cheap too frankly. The Omega and Honeywell load cells are about 6 times the cost (about $400). Are they really that much better? Should I "invest" in a better load cell?

View attachment 430596View attachment 430597View attachment 430598View attachment 430599View attachment 430600

Just curious. What does it weigh without the blast deflector?
It looks really good, I'm jealous. Someday I'll make one.

 

[Richard Dierking]

Thank you. 80/20 structural framing is a good material for stuff like this. However, cutting aluminum to size (I use a chop saw with aluminum cutting blade) makes a mess. And, the brackets are expensive. I would tell how much I spent, but that's Top Secret.

So, the stand on the sheet of 3/4" plywood is 28 lbs, and the blast deflector would add 15 lbs. I thought about staking it down, but bringing sand bags sounded better. It's high desert, so lots of sand available at the site. I'll replace the plywood with a sheet of aluminum when I find the right scrap at the metal recycler. Also, going to add a bit of ballistic protection to the uprights.

When I get done with the this test stand probably within 6 months, I'll give it to whoever wants to come to FAR and get it. I already offered it to someone but they never got back to me. So, up for grabs when I'm done with it and moving on to larger testing.

 

[Richard Dierking]

I have been using the micro load cells you can get from robotshop. They are actually Phidgets, which you can also get from their site. I went there because the one I wanted was not in stock at the time. They are great, inexpensive and reliable.

Thank you!
I still haven't received the order from Robotshop. Phidgets has good information on their site including detailed specifications.
There's an uninsulated wire included in the cabling For my application, where the load cell is mounted on aluminum framing, should I connect this wire to the frame? The videos I've seen where people use these types of load cells usually say to disregard this wire.

 

[StanO]

Thank you. 80/20 structural framing is a good material for stuff like this. However, cutting aluminum to size (I use a chop saw with aluminum cutting blade) makes a mess. And, the brackets are expensive. I would tell how much I spent, but that's Top Secret.

So, the stand on the sheet of 3/4" plywood is 28 lbs, and the blast deflector would add 15 lbs. I thought about staking it down, but bringing sand bags sounded better. It's high desert, so lots of sand available at the site. I'll replace the plywood with a sheet of aluminum when I find the right scrap at the metal recycler. Also, going to add a bit of ballistic protection to the uprights.

When I get done with the this test stand probably within 6 months, I'll give it to whoever wants to come to FAR and get it. I already offered it to someone but they never got back to me. So, up for grabs when I'm done with it and moving on to larger testing.

Thanks for the feedback. What do you anticipate as the maximum thrust for your testing of a hybrid motor?

 

[Richard Dierking]

Thanks for the feedback. What do you anticipate as the maximum thrust for your testing of a hybrid motor?

That's a good question. The 80/20 linear bearings are rated 90 lbs static. I think I'm just going to see how it goes with smaller 38 mm solids, then go from there.

 

[Richard Dierking]

I received the 50 Kg button load cell and installed it on the test stand. I'm very encouraged so far because it's working much better than the S-type load cell.

To calibrate it, I actually inverted the stand. I used the slide which I already knew weighed 934 grams. After determining the initial Kg to bit factor, I placed known masses on the stand slide to refine it. Max variation was about +/-20 grams to known mass which is about +/-0.2 Newtons, so that seemed good to me.



Back to normal orientation:
With the pneumatic at 11.5 psi (it has a 0.875 bore) the Arduino serial monitor again shows a variation of approx 20 grams and drift is about 20 grams an hour. Creep in the specs is 20 grams/hour.



So, what sample frequency should I use for the actual static test?
The G33 burns for 4.4 seconds. Max thrust is about 65 N.

 

[Dan Griffing]

I'm creating a vertical test stand to eventually test research hybrid motors. I want to test solids first to work the bugs out, and I have a larger version I would like to build.

Following are some photos. I ordered a 50 Kg button cell from Robotshop and hasn't arrived yet. Because the stand is vertical configuration, I installed a pneumatic cylinder to balance the mass of the slide (80/20 linear bearings) and whatever I'm testing on the stand. So, the cylinder is pressurized to bring the slide up to the load cell. Then, I run the tare program on the Arduino (actually an Adafruit Metro) to zero the read before conducting the test. The analog to digital converter/amplifier is from Oscar Davidson in the UK. He has a video showing his ADC on Youtube and reached out to him and eventually purchased one. It's capable of a high sampling rate.

And, because the stand is vertical, I can use it as a launch pad with 20 mm mini-rail. Also included a photo of that.

The question I have is regarding load cells. The S-type I originally got for this was not working out. Thought it would be great because it can be secured top and bottom. But, any movement side to side would cause readings to vary up to 200 grams. It was an inexpensive load cell. And, the load cell I ordered from Robotshop is cheap too frankly. The Omega and Honeywell load cells are about 6 times the cost (about $400). Are they really that much better? Should I "invest" in a better load cell?

View attachment 430596View attachment 430597View attachment 430598View attachment 430599View attachment 430600

Before retiring I spent about 15 years working professionally with load cells and strain gage technology.

There are many subtle aspects that are important, including mechanical linkage, electronics, and required data rate, precision, and data bandwidth.

High speed strain gage data collection of a rapidly changing signal requires much different electronic circuitry than the standard bridge-balancing that a stationary weight measuring uses.

As you will surely discover, an upward thrust test stand for measuring an entire thrust profile that might last less than a second — is far different from putting something on a static load cell scale and recording the weight.