Load Cell Resources?

[Johnnie]

I have been looking for an S-beam type load cell to go on to a new test stand, and I was wondering if anyone has a bead on one? I have checked with ebay, and I have had little to no success.

I do have a line on a local source, but the pricing seems a little high at $175 new ( 1000 lb. 3mV ).

 

[Loki]

That's what they cost, you aren't going to find new cells for less. They show up used on eBay occasionally, but there's no way to know if they are any good.

 

Nakka has a page on his website about how to build a strain-gage load cell.

Take a look - perhaps that will work for you.

 

[Johnnie]

Originally posted by Loki
That's what they cost, you aren't going to find new cells for less. They show up used on eBay occasionally, but there's no way to know if they are any good.

Luckily it is a local source, and our company has done business in the past. Thanks Jeff.

 

[Johnnie]

Originally posted by dosco
Nakka has a page on his website about how to build a strain-gage load cell.

Take a look - perhaps that will work for you.

I have looked over the strain gauge set-up as well as the brake cylinder set-up that mr. Nakka has used. These are great alternatives, and worthy of investigating further.

 

[Tom Rouse]

Hi J-
I spent a lot of time working on load cells ,AD converters and then the software trying to get just a simple system together to use for testing motors.
I tried three things.
First, a few guys from Aero-Pac who are engineers (Steve Preston etc) got together and we made a test stand and then used a lab jack interface and a program called something like Data Q ??? and spend months of spare time dorking with it.
The software was cumbersome and required a lot of attention. Its one of those programs that is hard to use unless you use it everyday. I ended up spending more time learning about something that I really no interest in. I just wanted to make motors and not get a BS in EE.
Then I bought the set-up from Oliver Schubert, a member of Les Derkovitz's gang who has a nice motor test system. I had a load cell and sent it to him for calibration to the software. The problem was that we couldn't the Windows software to inferface with it because Oliver wrote code on Win 98 and we had Win 2000. Some freaking windows clash yet again. (I think its worked out now.)

Finally, I bought the system from RCS. WOW, that is a **** hot system. Talk about plug and play. We go out to test motors all the time and get perfect results with NO problems, and, on a Mac to boot!
The person who wrote the program for RCS is a genius.
I can now make rocket motors and not have to worry about getting into another hobby of electrical engineering and computer programming just to get results of a motor burn.

Its easy. Just go buy a load cell and sent it to RCS. The system will be calibrated to that load cell and you are DONE. PERIOD.

Tom

 

[jwagner61]

Hi Johnnie,

I got my membership to the SugPro listserv - thanks for your help!
The interface I have for is the DataQ (www.dataq.com) DI-194 starter kit which includes both interface and software. While it may not be the best out there, its more than functional for logging thrust curves and at $25 complete its an incredible value.
A bonus is the ease in calibrating the load cell- just connect the interface to the loadcell/stand and set the high an low level using software. I set the signal to zero with the stand empty, and at 980 N with a 10 kg mass loaded. Calibration constants are stored in the software and no need to worry about recalibration. The data is displayed and saved as N, lbs or any other unit specified. Its really a snap.
The manual is a little dense as it covers many options not needed for simply logging data. If you opt for the DI-194 interface, I'd be happy to help with any configuration problems. Downsides are its a little clumsy to open and close files used to save data and a 2nd program has to be used to view the data and convert to CSV or other format needed for Excel. But for $25, it aint bad.
The load cells I've used are the compact button form, not the larger S-shaped cells you're looking at. I bought two off Ebay - both unused and sealed in their original packaging with calibration certificates. They were 100 and 200 lb capacity and I got them for under $50 plus shipping. I you keep an eye on Ebay, this and other types frequently are listed.
I've attached a picture of my current test stand. Cables/wires can quickly get out of control and making setup a real pain, so I've tried to simplifiy them as much as possible. A relay launcher and two NiCd 12 V power supply are in the picture as well.
The RCS system Tom mentioned is exceptional well designed, but if you don't mind playing with the electronics/wiring yourself, its not that difficult to make your own.

Just opinions based on my limited experience,
John Wagner

John Wagner

 

[Johnnie]

Originally posted by jwagner61
Hi Johnnie,

I got my membership to the SugPro listserv - thanks for your help!
The interface I have for is the DataQ (www.dataq.com) DI-194 starter kit which includes both interface and software. While it may not be the best out there, its more than functional for logging thrust curves and at $25 complete its an incredible value.
A bonus is the ease in calibrating the load cell- just connect the interface to the loadcell/stand and set the high an low level using software. I set the signal to zero with the stand empty, and at 980 N with a 10 kg mass loaded. Calibration constants are stored in the software and no need to worry about recalibration. The data is displayed and saved as N, lbs or any other unit specified. Its really a snap.
The manual is a little dense as it covers many options not needed for simply logging data. If you opt for the DI-194 interface, I'd be happy to help with any configuration problems. Downsides are its a little clumsy to open and close files used to save data and a 2nd program has to be used to view the data and convert to CSV or other format needed for Excel. But for $25, it aint bad.
The load cells I've used are the compact button form, not the larger S-shaped cells you're looking at. I bought two off Ebay - both unused and sealed in their original packaging with calibration certificates. They were 100 and 200 lb capacity and I got them for under $50 plus shipping. I you keep an eye on Ebay, this and other types frequently are listed.
I've attached a picture of my current test stand. Cables/wires can quickly get out of control and making setup a real pain, so I've tried to simplifiy them as much as possible. A relay launcher and two NiCd 12 V power supply are in the picture as well.
The RCS system Tom mentioned is exceptional well designed, but if you don't mind playing with the electronics/wiring yourself, its not that difficult to make your own.

Just opinions based on my limited experience,
John Wagner

John Wagner

Nice set-up John, do you have any curves that you have produced from your tests?

It looks like we have lucked into a 2,000 lb s-beam load cell, Data Q and amplifier...I have not seen any of it, but we may take posession sometime this week or next weekend.

 

[edwardw]

I've been looking for something in the 500 lb range - anyone have a good source on them. Also, what kind of AMP is better? I know the PSAS site has a great schematic for an AMP and was wondering if that was overkill.

Link:

https://psas.pdx.edu/LoadCellAmp



Edward

 

[jwagner61]

Hi Johnnie and Ed,

If you're looking for a large load cell, you might be interested in this:

METTLER TOLEDO MODEL 0743 3,000 LBS CAPACITY LOAD CELL
ebay Item number: 7538237735.

By the looks, it never been used and the starting bid price is $70 with a buy-it-now price of $90.
I'll try and post some data from my test stand later this morning.

John Wagner

 

[dr wogz]

We sell them.

www.scaletron.com

 

[jwagner61]

Attached is a thrust curve I took this morning on my load stand using the following:

Transducer Technique SLB-200-KE (200 lb load cell)
Futek JM-2 load cell amp
DataQ DI-194 10-bit interface.

The data was logged at a rate of 120 readings/sec.

It was a 38 mm sugar motor with grains cast 2 weeks ago that I hadn't got around to firing. The peaked thrust curve isn't the test stand's fault but the motor. The motor didn't ignite as fast as others I've tried and I'm still trying to get a neutral burn on KNSU motors. I'm considering coating the end of the grains with smokeless gun powder.

Anyway, at least the stand seems to work.

John Wagner

 

[Johnnie]

Originally posted by jwagner61
Attached is a thrust curve I took this morning on my load stand using the following:

Transducer Technique SLB-200-KE (200 lb load cell)
Futek JM-2 load cell amp
DataQ DI-194 10-bit interface.

The data was logged at a rate of 120 readings/sec.

It was a 38 mm sugar motor with grains cast 2 weeks ago that I hadn't got around to firing. The peaked thrust curve isn't the test stand's fault but the motor. The motor didn't ignite as fast as others I've tried and I'm still trying to get a neutral burn on KNSU motors. I'm considering coating the end of the grains with smokeless gun powder.

Anyway, at least the stand seems to work.

John Wagner

Wicked peak!

I have a program that a gentleman in our club wrote that will give you all the info you need including Isp. The data was easily converted to Rocsim, so you could virtualy fly your creation. I'll see if I can get permission to send it too you. It poulls all of the data from the dataq, it is really a nifty program.

 

Originally posted by jwagner61
Attached is a thrust curve I took this morning on my load stand using the following:

It was a 38 mm sugar motor with grains cast 2 weeks ago that I hadn't got around to firing. The peaked thrust curve isn't the test stand's fault but the motor. The motor didn't ignite as fast as others I've tried and I'm still trying to get a neutral burn on KNSU motors. I'm considering coating the end of the grains with smokeless gun powder.

Anyway, at least the stand seems to work.

John Wagner

1. Is the time hack started when you energize the ignitor? If so, your ignition system has some sort of problem.

2. I think you have some sort of other problem as well, the peak is......well.....very peaky. What is the grain geometry? Can you hand-sketch a "representative cross section?"

 

[jwagner61]

Hi Johnnie and dosco,

1) The time axis in my previous thrust curve is arbitrary because I start logging the data before igniting the motor. Ignition is typically faster than this morning's data but all my curves look spikey.
2) Attached is a picture of the 3 grains used in the sugar motor - they are supposed to be typical bates geometry grains with 1.25"OD and 0.5" ID. My coring skills leave much to be desired so perhaps they are moon-burners. I put them inside a poster-board liner as per Jimmy Yawn's excellent web site, and ignite them with a homemade ignitor made with nitrocellulose and blackpowder.

I've tried 3/8" and 1/2" cores with a 5/16" nozzle, but the spike remains. I'm either doing something wrong with making the KNSU propellant or the homemade ignitors are not up to the task. I will try to get some real ignitors and see if that makes a difference.
Thanks for the advice,
John Wagner

 

Originally posted by jwagner61
1) The time axis in my previous thrust curve is arbitrary because I start logging the data before igniting the motor. Ignition is typically faster than this morning's data but all my curves look spikey.

OK. I can handle that, although it may be helpful to start the hack when the ignitor is "lit" so you can tell how long it's taking the chamber to come up to pressure.

2) Attached is a picture of the 3 grains used in the sugar motor - they are supposed to be typical bates geometry grains with 1.25"OD and 0.5" ID.

What are you using to hold the grains in the case? The reason I ask is that there needs to be space between the grain ends....if you just stack them/jam them in there the ends may not be exposed sufficiently which could be contributing to the spike. There needs to be a gap between them that allows the flamefront to ignite the grain ends while also igniting the grain core.

Nakka has also used 'pyrogen' applied at the grain ends to aid in proper ignition.

and ignite them with a homemade ignitor made with nitrocellulose and blackpowder.

Have you tried Nakka's ignitors?

Additionally, perhaps the ignitor is too small or incorrectly placed.....

 

[edwardw]

For sugar motors I used with great success a pyrodex pellet. You can find them at your local gunstore/outdoor store. About 3/8" OD with about an 1/8" core through them. I put my ignitor in them and the motors lit fast.

Edward

 

[Johnnie]

Originally posted by dosco
OK.

What are you using to hold the grains in the case? The reason I ask is that there needs to be space between the grain ends....if you just stack them/jam them in there the ends may not be exposed sufficiently which could be contributing to the spike. There needs to be a gap between them that allows the flamefront to ignite the grain ends while also igniting the grain core.

I have never "gapped" my grains in order to get the ends to burn. The space between the grains will ignite with or without space, much like a flame front will travel thru the crack of a defective grain. If there is nothing there to inhibit the ends from burning, they will burn very well. I have enclosed a pic of one of my liners that clearly shows that the grains ignite near instantly and burn throughout the course of motor run.

The ends of the grains could be painted with a pyrogen mixture to insure instant ignition, but with that I would be hesitant due to possible over pressurization due to the extra pyrogen burning inside of the motor.

The pic was the bi-product of a 6-grain RNX variant, which ignites a little slower than KNSU...

 

[shockwaveriderz]

do you think this pressure sensor would work:

https://www.imagesco.com/catalog/sensors/pressure.html

or perhaps this flex sensor:

https://www.imagesco.com/catalog/flex/FlexSensors.html

or perhaps these piezo sensors:


https://www.imagesco.com/catalog/sensors/piezo.html


https://www.imagesco.com/articles/piezo/piezo00.html

 

Originally posted by Johnnierkt
I have never "gapped" my grains in order to get the ends to burn. The space between the grains will ignite with or without space, much like a flame front will travel thru the crack of a defective grain. If there is nothing there to inhibit the ends from burning, they will burn very well. I have enclosed a pic of one of my liners that clearly shows that the grains ignite near instantly and burn throughout the course of motor run.

Hey, no problemo. And FWIW I agree.....however, got an explanation for the "spike" in the chart? Seems odd to me that a Bates-grain would have that problem.

And if end-ignition wasn't a big problem, why did Nakka perform experiments with Bates-grains by putting pyrogen between the ends?

Of course, this is all assuming Mr. Wagner didn't make the propellant incorrectly, or botch his K-sub-N.....

The pic was the bi-product of a 6-grain RNX variant, which ignites a little slower than KNSU...

Could that be the source of the difference in the sugar and you motor?

 

Originally posted by Johnnierkt
I have never "gapped" my grains in order to get the ends to burn. The space between the grains will ignite with or without space, much like a flame front will travel thru the crack of a defective grain.

A quick review of Nakka's website....looking at the "Kappa" engine, there are spacer rings that are designed to hold the Bates grain ends .25" (that's a quarter inch) apart (!!).

And he uses a pyrogen ignitor at the top of the motor that "spits a flame for 2 feet for a duration of 2 seconds....the effects in the motor are different due to confinement."

I'd still lay odds that Mr. Wanger isn't getting good end burning of the Bates grains, combined with poor ignition.

 

[Johnnie]

Originally posted by dosco
Hey, no problemo. And FWIW I agree.....however, got an explanation for the "spike" in the chart? Seems odd to me that a Bates-grain would have that problem.

And if end-ignition wasn't a big problem, why did Nakka perform experiments with Bates-grains by putting pyrogen between the ends?

Of course, this is all assuming Mr. Wagner didn't make the propellant incorrectly, or botch his K-sub-N.....



Could that be the source of the difference in the sugar and you motor?

Sorbitol has the personality of being a "spike" propellant, and I would assume Sucrose is too. Nakka gets a neutral burn on his sucrose based on his thrust curves, but he really needs to turn the resolution up to catch it in a span of .4 seonds...

As for the RNX possibly being a source of the difference in the two propellants....maybe so, but the pic I offered up was showing that using a bates configuration works in theory to try to produce a neutral burn. My RNX tends to burn a little longer, so the flame front stays for longer period of time as noted by the hot spots on the used liner...I have not observed this on any of my "candy" motors as the propellant is consumed entirely too fast to burn for any length of time to cause a hot spot in the liner.

 

[jwagner61]

Shockwaveriderz had a link for interesting, low cost force sensors that could work in test stands however they lack the precision and accuracy of standard load cells. Comparison of performance characteristics of typical load cell vs Flexiforce:
Linearity: 0.1 vs 5%
Repeatability: .02 vs 2.5%
Hysteresis: .06 vs 4.5%

On the plus side, the electronics appear to be much simpler since the signal/load coefficient seems to be a million times greater (I didn't read the specs that carefully).

The value for my last motor was in the 180-200 range. What values do you typically use? Failure to ignite the grain ends would result in a progressive burn. Is that what happens in the first part of my thrust curve? The peak is assymetrical.

John Wagner

 

Originally posted by Johnnierkt
Sorbitol has the personality of being a "spike" propellant, and I would assume Sucrose is too. Nakka gets a neutral burn on his sucrose based on his thrust curves, but he really needs to turn the resolution up to catch it in a span of .4 seonds...

Wagner's motor burned for what - 1 second?

Nakka's "Kappa" motor burns for 1.8 seconds.

Nakka's "B" engine (an H-impulse category) burns for .4 seconds.

And fwiw, Nakka's resolution appears to be pretty good, based on his graphs. So I don't buy your "resolution" argument, except that maybe Wagner's resolution is too low (although he's claiming a sample rate of like 100+ samples per second....).

As I said before, I think Wagner's problem is either the end gaps or ignition. Or both.

My RNX tends to burn a little longer, so the flame front stays for longer period of time as noted by the hot spots on the used liner...I have not observed this on any of my "candy" motors as the propellant is consumed entirely too fast to burn for any length of time to cause a hot spot in the liner.

Yeah, that's not too suprising. What kind of Kn are you using in your RNX motors?

 

[Johnnie]

Originally posted by jwagner61
Shockwaveriderz had a link for interesting, low cost force sensors that could work in test stands however they lack the precision and accuracy of standard load cells. Comparison of performance characteristics of typical load cell vs Flexiforce:
Linearity: 0.1 vs 5%
Repeatability: .02 vs 2.5%
Hysteresis: .06 vs 4.5%

On the plus side, the electronics appear to be much simpler since the signal/load coefficient seems to be a million times greater (I didn't read the specs that carefully).

The value for my last motor was in the 180-200 range. What values do you typically use? Failure to ignite the grain ends would result in a progressive burn. Is that what happens in the first part of my thrust curve? The peak is assymetrical.

John Wagner

John,

from what I am able to find out, Richard Nakka was able to obtain a neutral burning KNSU by using a free standing grain. He may have used bates at one time, but I did not see any notes to that effect right off. Try and design a free standing grain for the same case and see what the net results are, could be quite interesting. Would be great workout for your test stand.

 

[bobkrech]

Originally posted by shockwaveriderz
do you think this pressure sensor would work:

https://www.imagesco.com/catalog/sensors/pressure.html

or perhaps this flex sensor:

https://www.imagesco.com/catalog/flex/FlexSensors.html

or perhaps these piezo sensors:


https://www.imagesco.com/catalog/sensors/piezo.html


https://www.imagesco.com/articles/piezo/piezo00.html

The first two sensors are interesting, and I was not aware of their existence. They are capacitive pressure sensors made by Tekscan. https://www.tekscan.com/flexiforce/flexiforce.html You can read about them at the USPTO https://patft.uspto.gov/netacgi/nph...,272,936.WKU.&OS=PN/6,272,936&RS=PN/6,272,936 A pretty clever design for an otherwise expensive type of pressure sensor.

Good things:

They are really cheap. 4 for $55 from the manufacturer. They have ranges from 1 lb, 10 lb and 100 lb and at least the 100 lb one can be overanged by 10 to 1000 lb by an electronics modification. Another good feature is that they have a fast response time (5 or 20 microseconds listed but I a bit skeptical.).

Bad Things: Low temperature rating for standard product. Sensitive to temperature. Poor repeatibility. Poor hysterisis. Poor linearity (At least with simple electronic circuits shown on their website.).

It could be used for test stand applications if 5%-10% accuracy is acceptable, however since the load cell is a relatively small cost in a test stand, a load cell is probably a better all around choice.

The last two sensors are PVDF polymer based piezoelectric sensors. They are not suitable for test stands as they have no DC response. They require calibration which will be time dependent and require special amplifers if you are doing something quantative.

Bob Krech

 

Originally posted by Johnnierkt
John,

from what I am able to find out, Richard Nakka was able to obtain a neutral burning KNSU by using a free standing grain. He may have used bates at one time, but I did not see any notes to that effect right off. Try and design a free standing grain for the same case and see what the net results are, could be quite interesting. Would be great workout for your test stand.

Johnnierkt:
The "Kappa" motor is a Bates grain, for KNDX and KNSB. 1/4 inch spacing between grain ends, ignitor built into the upper enclosure, etc etc.

Here is a link: https://members.aol.com/kappadx/kappa.html

EDIT: (Nakka does like the free-standing and rod-and-tube grains, though - I'll give you that)

 

[Johnnie]

Originally posted by dosco
Johnnierkt:
The "Kappa" motor is a Bates grain, for KNDX and KNSB. 1/4 inch spacing between grain ends, ignitor built into the upper enclosure, etc etc.

Here is a link: https://members.aol.com/kappadx/kappa.html

EDIT: (Nakka does like the free-standing and rod-and-tube grains, though - I'll give you that)

Dosco -

Jwagner is having trouble with Sucrose not Dex or Sorbitol...Nakka uses Sucrose in a free standing grain only config. YES! The KNDX and the KNSB propellants work as a bates I do not dispute that (except for maybe KNSB) but sucrose is the delima on the test stand here...

 

[Johnnie]

On topic -

It is possible that a new loadcell will be on the new test stand by Friday, not an s-beam like I would have liked, but a canister should still work fine. I will have more details in a few more days.

Thanks to all who posted...still watching ebay though.

 

[jwagner61]

Hi Johnnie, dosco, Edwardw:

Thanks for all the discussion on my thrust curve problems with sucrose motors. I really like Johnnie's suggestion for trying an uninhibited, single free standing grain. This should minimize grain-end ignition problems and hopefully provide some answers for this debate.

Examining the first 0.5 s of my thrust curve more carefully, the increasing thrust falls into two regions ~0.1 and 0.4 s long. Wild speculation that the first 0.1s corresponds to ignition of the inner surface followed by a 0.4s progressive burn linear from 60 N to 190 N is close to what one would expect from the grain geometry. Without considering grain ends, the final to initial burn area ratio is 2.8. Dosco's concerns about end ignition may well be right. As for botched Kn, incorrect time measurements etc..., I like the words of the great Charles Barkley: "I may be wrong - but I doubt it."

I'm out of town for a few days, but I will fire a few test motors when I return.

Thanks again for all the discussion,
John Wagner