Choosing A Piezo Material And Conditioning the Signal

[Ron_W]

Good afternoon,

I am the team lead for the NASA SPACE Flight Design Challenge at Blue Ridge Community & Technical College. I am trying to figure out the best type of Piezoelectric ceramic to purchase for measuring the harmonics of a sub-orbital rocket launch (RockSat-X initiative). Last year I decided upon the SMD15T02F140412S made of the material SM412 from StemInc (

http://steminc.com/).

We were able to record high amplitude, low frequency (80-100 hertz at sea level) and then low amplitude, higher frequency (300-400 hertz above 10 miles) on the x, y, and z axes within a Terrier-improved Orion Sounding Rocket. This year we will be launching inside of a Terrier-improved Malemute Sounding Rocket. The technical info is as follows:

-Altitude (km) ~ 160 km

-Spin Rate (Hz at Burn-Out) ~4.8 Hz at Malamute burn-out ~ 0 Hz at apogee

-Maximum Ascent G-Load 25 G (Sustained) 50 G+ Impulses Possible

-Rocket Sequence (Burn Timing) 5.2 second Terrier burn (Max sustained could be 26g), 12.2 second coast, 11.7 second
Malamute burn (Max sustained could be 32g)

a series of vibration testing of up to 2,000 hertz will be performed at wallops before launch. Any piezo must survive this testing environment.

Last year's experiment piezo information:

Piezo Ceramic Disc
SMD15T02F140412S


Part Number: SMD15T02F140412S

Piezoelectric Ceramic disc. Silver electrodes being one on each side (S configuration). Radial mode vibration

Piezo Material: SM412
Dimensions: 15mm dia. x 0.2mm thickness
Resonant frequency fr: 140KHz ± 3%
Electromechanical coupling coefficient Kp:≥58%
Resonant impedance Zm: ≤4.0 Ω
Static capacitance Cs: 13500pF±20%@ 60Hz/1V
Test Condition: 23±3 °C 40~70% R.H.
fr, Zm, Kp => Radial mode vibration
Cs => LCR meter at 1KHz 1Vrms

Applications: Piezo transducer vibration, matter dispersion, sonar transducer, Ultrasonic Sensor, wall thickness sensor, material stress sensor, pressure sensor,piezo electricity harvesting, fish finder transducer, compression sensor, piezo expansion sensor, biomedical probe and others.


I have put about 80 hours of research into Piezoelectric theory and components, however we do not have any electronics engineers or mechanical engineers to walk me through choosing the proper piezo for the job and how to condition the signal for this specific application. We were successful last year, but this year I want to maximize the resolution, calculate the frequency and amplitude of the vibrations, and plug our analysis into a VR simulation to show future teams what the harmonics of the rocket looks like during each second of the launch. Some teams have had their experiments ruined by the harmonics from the testing, or from the launch because they did not have the understanding of how to mitigate vibrations. I am seeking to rectify this. In the handbook we are instructed to be able to withstand 2,000 hertz, but that doesn't cover the strength of the forces involved.

My apologies for the length of the post, but I figured I would color in the history and experiment idea a bit to prevent misunderstandings. This is not for a grade, nor am I going to profit from this. This is something I am looking to share freely after completion of the launch with the general public and the other RockSat teams. We are already working on the 3D environment, have the designs for the PCB almost ready for fabrication and are working on telemetry.

My main requests are as follows:

Would anyone be willing to point me to a resource, or explain the type/size piezo that would best fit this application? To reiterate, I choose a 15mm diameter, .2mm thick rigid ceramic disc piezo last year.

Would anyone be willing to point me to a resource, or explain the best way to condition the signal. We are using the Raspberry Pi HAT - 8 Channel ADC - MCP3208 - SPI to convert for the Raspberry Pi 3B.

Would anyone be willing to point me to a resource, or explain the best way to analyze the data from a scratch made program (including calculations if possible) My younger brother is the software lead, but we are self taught in all of this, so we may need someone to slap a ruler to our coding fingers. (this specific request may be for a different forum section on this website)

I would like to cite my sources, so please feel free to request to remain anonymous if you help us out and do not wish to be added to my final report that I submit to NASA in August of next year. (I will only be using first names anyway, but I don't want to be rude)

I will be making a full-in depth YouTube video breaking down how we built our experiment after we finish it and launch it in August, so if anyone wants to see it upon completion, I will be happy to notify you when I upload it.


Thank you for your time and consideration,
Ronald Willis
(SPACE Flight Design Challenge Lead and local NASA SPACE Club President)

 

[GregLocock]

That's a pretty neat summary, but I seem to have missed what you are trying to measure. Acceleration? sound? what?

Cheers

Greg Locock


New here? Try reading these, they might help FAQ731-376

http://eng-tips.com/market.cfm?

 

[btrueblood]

...and, if measuring a physical parameter, why you wouldn't buy a ready-made sound/pressure/acceleration transducer that is already scaled and calibrated to the limits you are trying to measure?

 

[Strong]

Ronald,

Following on comments by others, what sensor characteristics do you need that are not already present in a commercial off the shelf (COTS) product. Look at Analog Devices MEMS accelerometers or piezo accelerometers from PCB Piezotronics, Endevco, and others. Don't you have enough work to do for the signal measurements and analysis? Get a standard sensor product, since they have already been in space and came back!

Walt

 

[Ron_W]

Hello everyone,

we are looking for Physical hertz data (Just frequency and Amplitude of vibration). We do not need anything more than that with the piezos. Taken from the summary above:

"but this year I want to maximize the resolution, calculate the frequency and amplitude of the vibrations, and plug our analysis into a VR simulation to show future teams what the harmonics of the rocket looks like during each second of the launch."

I already have a few degrees (AAS) from this college, so I could easily buy a calibrated sensor, but if I am trying to be an engineer, why would I skip the process of learning how to build a system from scratch? My long term goal is to work with neural interfacing and the R&D that accompanies it. An intimate understanding of why something works is much more pragmatic in R&D than knowing how to plug in numbers. I could easily grab libraries (Which we are for the ADIS gyro we are using as a redundant system) and use another persons work with a fairly in-depth understanding of the topic, but that won't show future students here and in the RockSat initiative why the waves move as they do, why signal conditioning is important, why which math is used and how to apply it. I am 2 classes from achieving an AAS in mechanical engineering, and a few after that before a bachelors, but I haven't been shown how to build something myself that's of this nature. It's high level and it's fun.

I am assuming most here who use a CNC mill or lathe know that it's important to learn manual mill or lathe first to get an understanding behind the work involved in performing the tasks required and understanding why the cutting tool has to be positioned a certain way. Why would I skip the manual set up for a physical hertz transducer? I am not looking for professional quality per sey, but if our best is close to that standard, I would be ecstatic about it.

The SPACE Flight Design Challenge is an initiate from the Colorado Space Consortium and NASA to teach future engineers practical skills and let them build complex experiments to test themselves. Not building anything custom would go against the very nature of the initiative. And a little boring.

TLR - We want the challenge and we want to know what we are doing. This is a trial by fire to test our abilities. We have degrees and certifications in CAD, Instrumentation, and Robotics between us, but we want to learn the higher level processes required to put ourselves at the front of the up coming AI/Automation revolution. You can't simply plug and play with brand new technology.

I've already reached out to companies and was unsuccessful finding any engineers with any experience in our area. I have been trying off and on for about a year in. So, this isn't some lazy attempt to have someone do my work. I legitimately want to learn this stuff and need assistance from... well, anyone that can point us in the right direction. This is akin to a hail mary play at this point.

I hope I was able to accurately clear up anything confusing. Thanks guys!

 

[Ron_W]

I came here because I am a technician already, do not have any staff on campus that can assist with this (Piezo), do not have a campus library with books on this subject and have not received help from any other source I have tried to use thus far. Even a link to a few resources that show applied use and not simply theory is all I really need.

 

[IRstuff]

"but if I am trying to be an engineer, why would I skip the process of learning how to build a system from scratch"

This is precisely what an engineer DOES NOT DO on a time critical project. You cannot stand on the shoulders of giants if you keep reinventing the wheel bearing. If you are doing this solely for your own amusement, that's a whole separate thing. For actual real-world engineering, the MAKE/BUY decision has to be done with consideration of benefiting the program as a whole.

TTFN (ta ta for now)
I can do absolutely anything. I'm an expert!

https://www.youtube.com/watch?v=BKorP55Aqvg

faq731-376 forum1529 Entire Forum list

http://www.eng-tips.com/forumlist.cfm

 

[GregLocock]

It sounds as though you are using a disc as a bender (plate in bending). That is how cheap hydrophones and microphones work, it is not the typical construction for an instrumental piezo accelerometer, which mount a mass to the piezo and crush it in compression.

https://en.wikipedia.org/wiki/Piezoelectric_accelerometer

. That's not to say that it won't measure vibration, microphones in general are very sensitive to vibration. I'd have thought there was some danger of it responding to the sound at launch, which can be 170 dB, enough to cause structural problems.

The reason you won't get much traction here is that we've all just gone and bought accelerometers, not tried to reinvent the wheel. As it happens I have worked in a plant that manufactured benders, but guess what, somebody else designed them, I (and 3 others) just designed the thing that used them.

I'd take the one from last year, that works, and a commercial one that has the required spec, and test them on a shaker. If your solution has a reasonable performance then persevere with improving it. If not, gracefully bow out.

Cheers

Greg Locock


New here? Try reading these, they might help FAQ731-376

http://eng-tips.com/market.cfm?

 

[btrueblood]

Can't disagree with Greg. But his cautionary statement is valid too. You are likely measuring a compounded signal that includes both physical accelerations of the object the sensor is attached to, as well as acoustic noise from the engine and aero-shell of the rocket. Good microphones are designed to (as much as possible) ignore physical vibrations; good accelerometers are designed to ignore pressure signals. Testing and calibrating the device would include determining the sensitivity of the measurement both to the physical parameter being measured as well as sensitivity to possible confounding parameters and uncontrollable enviornmental variables (temperature, humidity, barometric pressure, etc.) Without that calibration and testing, what you are measuring is just noise.

 

[Strong]

I have been working in the fields of sound and vibrations for 50 years. I have worked with a wide of variety of microphones, dynamic pressure sensors, accelerometers, seismic velocity sensors, proximity and LVDT displacement sensors. I neve felt compelled or the need to design a custom sensor, but I can appreciate someone who wants to do it. I only hope that you would honor the space program and give flight to the best and most reliable sensors and not send junk (sorry -- unproven design) into space with taxpayer dollars. I would venture to say that most engineers who when asked for the time of day would look at their watch, smart phone or computer or clock on the wall and not start designing a time piece. I think everyone is trying to make the point of encouraging you to use COTS components unless you have an engineering evaluation that shows the need for a custom sensor or component. If you want more help, then at least provide detailed specifications so that recommendations can be made.

Walt

 

[Ron_W]

IRstuff,

We have a redundant system set up to compare with our Piezo data (ADIS16460 Gyro and Accelerometer). We are working on this project to create a low cost, decently high resolution solution for future teams and the layman to be able to create at home. It's not about reinventing the wheel. It's about creating a straight forward guide for fabricating a homemade sensor for a multitude of purposes, rather than having to purchase multiple sensors for non-critical applications. We are not on a time crunch. The experiment period just began and I am getting in front of the timeline in order to work out all of the details and start fabricating before the final design review is even do. Our team is almost ready to design the PCB and fabricate and we haven't even reached the preliminary design review stage in the program as a whole, so I am quite aware of the fact that time critical jobs require off-the-shelf solutions. We have accounted for the time necessary to complete this task and have off-ramps just in case we need to scrap certain features. As you are probably aware, there are Augmentative and alternative communication devices that use non-typical piezo configurations. These needed to be reinvented to accomodate the application. With automation coming swiftly, it is important to become self sufficient in the ability to design new systems on the fly for testing. There is utility in this.

Edit: My father is an Electrical and Optical engineer with over 30 years of experience. When he read your reply, he wanted me to remind you that engineers in certain positions are required to reinvent the wheel. He, for example, had to completely redesign lenses for a certain application when lead was no longer allowed to be impregnated within the structure of the lens. He had to work out the math, the physics, and then look for the lens that would suit his purposes. He then had to redesign the inards of the device he was working on to account for the subtle differences in the lens that couldn't be accounted for in the lens design alone. That is a real world example of a time he was required to reinvent the wheel to "match his car," if you will. Also, any company that makes piezo sensors constantly keeps a look out for new piezo materials and tries new and more sophisticated designs for sensors.

btrueblood,

Last year we insulated the piezos from noise and used Epoxy CL-30 to adhere the piezos in place. we amplified the signals from the X, Y, and Z axis piezos by a factor of 10 and were able to see very clean voltage spikes, with only low level noise. I also designed a shielded cable that attached to ground to protect against radiation that could cause noise. We succeeded in monitoring the physical vibrations within a canister mounted inside the rocket. We are running the same experiment this year, looking to fine tune the piezo material, shape, and size to maximize the resolution. I have accounted for many factors, but just want some experts to post their thoughts. Again, WE ALREADY SUCCEEDED last year. I am just curious as to what the experts say, so I can match what I already know against them and improve my skills, which will allow me to improve my team's skills.

This is our fourth year, I am already a graduated technician, and I thought this would be a good place to come and check with others. I definitely appreciate the cautioning from both of you, but be aware this experiment is more about learning something specific that is not usually taught in many universities. That is the purpose behind this forum, yes? Specialized information to help others expand their knowledge base? I don't want anyone to waste their time. I just genuinely want to match my education to someone who is more seasoned.

Strong,

We are a pretty low budget initiate. Last year we bought some of our own parts. The initiative is designed to give future engineers and scientists the ability to design and fabricate experiments. Here is a link for you:

https://spacegrant.colorado.edu/national-programs/rs-x-home

(Home Page)

https://spacegrant.colorado.edu/images/RockSatX/2018/RockSat-X_Users_Guide_Rev_9.pdf

(User Guide; deck layout begins on page 9 of the PDF)

The program focuses on community colleges and others that don't have the ability to participate in these initiatives on their own. To save money over the past 3 years, I suggested using more cost effective housing and provided free applications, like Slack and Trello, to help with project management. Our school fund-raises money to go to wallops every year. I have personally been involved with improving the program to eliminate waste, including recycling components from the previous year. I take it very seriously. Saying that, this is one of the programs that is well worth the funding. this isn't a wasteful endeavor. Schools that don't pass the design reviews, or look to be wasting everyone's time are cut from the initiative. There is some bureaucratic oversight, but all in all, its a pretty tightly kept ship. The RockSat program has increased the competence of many new engineers and freely offers findings from all experiments.

The application of the piezos in this experiment are rather unique. I might be able to find a sensor that could do what we are trying to do (I looked, but wasn't able to find what we were looking for; could be lack of knowing what exactly is needed), but the expense might be higher than we are willing to spend (the major reason we don't want to buy COTS). I haven't been able to find a component that fits perfectly yet for this application, but I am speaking with StemInc right now about piezos for this application as well as trying this forum out.

As for the details, they are up on my original post. The rocket, the forces, 3 piezo ceramic discs (One attached to each axis of the canister). We are looking for the physical vibrations felt by the canister in the rocket to describe what teams will need to account for in the way of physical vibrations. The canister we are in is on the website I provided. The piezos and controllers will be housed within an enclosure, most likely made of aluminum (We are waiting for details on that as we speak. WVU is designing the enclosure now. We had some teams that experienced resonance in their experiments, ruining calibration on some and destroying others at certain points in the flight. This experiment will provide vibration data and how we designed the experiment so other brand new teams can do something similar, or simply use the data to protect their experiment. If we find (or you know of) cheap-ish COTS sensors that can be used for this application, I will add them to the experiment as an isolated system and still design our own and see how close the results are.

I appreciate all of the answers/replies everyone has given. Thank you for your time!

Ron

 

[Strong]

The OP does not mention how the piezo disk is/was mounted. It can respond to acoustic pressures and vibrations. It is usually designed into a microphone or an accelerometer (sub-audio, audio, or ultrasonic frequency range). An accelerometer design places the piezo material between the enclosure base (mounting surface) and a seismic mass. The piezo material produces an electrical charge proportional to the dynamic tension-compression force on the piezo material (F=M*A). A charge amplifier is needed to convert charge (high impedance input) into voltage (low impedance output). Common accelerometers have the amplifier built into the case and are called ICP or IEPE, and a DC voltage is needed to supply constant DC current to the accelerometer.
There are plenty of engineering challenges for this project including high temperature (500-F), sensor mounting and resonant amplification, frequency analysis (onboard or from telemetry data). Note that analog vibration data cannot be sent via the 0-5 volts telemetry system, because time domain data is +/- 5 or 10 volts. It could accept RMS overall vibration levels, but no frequency analysis would be possible. Do you have another scheme to do onboard frequency analysis and/or store the dynamic data for analysis after rocket recovery?
Here are some wandering links of resources:

Endevco Accelerometer Provides 100pC/g Charge Sensitivity At Up To 500 Degrees Farenheit; Low Cost Product Designed for General Vibration Measurements.

From <

https://www.thefreelibrary.com/Ende.../g+Charge+Sensitivity+At+Up+To...-a020818816>

High Temp Accelerometers ≥ +490 to +1200 °F

From <

http://www.pcb.com/testmeasurement/accelerometers/hitemp>

High Temperature ICP® Accelerometers (≤ +356 ºF/+180 ºC)

From <

http://www.pcb.com/testmeasurement/accelerometers/hitemp_icp>

High temperature charge amplifier for geothermal applications

From <

https://www.osti.gov/biblio/1230012>

Search: how to make a piezoelectric accelerometer

https://www.bing.com/search?q=how t...-41&sk=&cvid=81221261397E410C847B5115A1D28216

Accelerometer Images

https://www.bing.com/images/search?...C3E19BB819DB3154F7FBD5C6DE4F5366B&FORM=IARRTH

Make Your Own Precise Vibration Sensors

From <

https://makezine.com/projects/make-40/vibration-sensors/>

Search: analog devices accelerometer

https://www.bing.com/search?q=analo...x8JkhYQb!Di!6hMQPwRG3NtwKondsowKnZYv8BTWSjD0D

Accelerometer Design

http://shodhganga.inflibnet.ac.in/bitstream/10603/2272/8/08_chapter 2.pdf

Accelerometer Transduction Types

From <

http://www.modalshop.com/calibration.asp?ID=190>

Search: pvdf accelerometer

https://www.bing.com/search?q=pvdf ...-18&sk=&cvid=438C7E492BBC473CA25F96C7923599B1

https://www.masterbond.com/

for adhesives and circuit protection

https://www.analog.com/en/education/education-library/scientist_engineers_guide.html

Accelerometers

From <

https://www.analog.com/en/products/sensors-mems/accelerometers.html>

What You Need to Know About MEMS Accelerometers for Condition Monitoring

From <

https://www.analog.com/en/technical-articles/mems-accelerometers-for-condition-monitoring.html>

Condition-Based Monitoring (CBM)

From <

https://www.analog.com/en/applicati...avilion-home/condition-based-monitoring.html>

Analog Devices Engineering Zone

https://ez.analog.com/

A high temperature 100 mV/g triaxial accelerometer

https://endevco.com/news/archivednews/2010/2010_04/TP329.pdf

AC STANDARD – HIGH TEMPERATURE ACCELEROMETERS
500-F rating
From <

https://www.hansfordsensors.com/pro...tandard-high-temperature-accelerometers-top/>

Signal Conditioning Basics for ICP® & Charge Output Sensors

From <

https://www.pcb.com/Resources/Technical-Information/tech_signal>

Signal Conditioning Piezoelectric Sensors

http://www.ti.com/lit/an/sloa033a/sloa033a.pdf

Walt

 

[btrueblood]

So, did you compare the accelerometer data to your sensor, and confirm that you can correlate them?

Can you test your sensor on a shaker table, to both verify and calibrate the device?

Can you build multiples of the sensor, so that you can show what effects variations in construction have on the calibration/range/sensitivity of the sensors?

What you have now is a sensor that gives an unknown output in response to a somewhat known input. Useful as a demonstration, perhaps, but of limited use to an engineer. It's a bit like me making a series of marks an a stick, and telling you it's a ruler. Sure, my ruler can measure something, but what is it measuring relative to known quantities? Can you see how a "store bought" ruler is a bit better than my stick carving? Can you see how at least comparing it to a known good ruler would improve it rule-iness?

 

[Ron_W]

Btrueblood,

Yes, we used a few different devices, including a back massager (we didn't have a shaker table available), and were able to match the voltage spikes to the frequency of the massager's oscillations. I wasn't sure how to calibrate for amplitude though, seeing as piezos exhibit hysteresis and other complex effects that are difficult to account for. We are adding amplitude this year. We were able to verify that the vibrations after the burn ended, remained in the same frequency pattern, with the exception of the amplitude dropping rapidly. The high altitude engine burn also showed the same results. The frequency at higher elevations was more stable without the friction from the atmosphere acting heavily upon the rocket's skin. After the high elevation burn the vibrations exhibited the same pattern of behavior, where the votage spikes didn't change in frequency much (About a 20-40 hert change over a few seconds) as the high DB sound source ceased. It was an effective, albeit crude success. That's why I called it a low level experiment. Where I could be wrong in my calculations, the hertz ranges during the flight match what we were expecting to see, i.e. low freequency/high amplitude at in the lower atmosphere and the opposite in the upper atmosphere as the rocket material has more physical freedom to naturally exhibit vibrations without outside fluid friction and engine forces. We are matching the vibrations to an accelerometer this year during testing. We have the ADIS gyro that has a 3 axis accelerometer. We made sure to purchase a fairly safe calibration source.

Strong,

Thank you for the links. I have seen, read a couple of these already, but I will still verify that I have and read them all. We are using a single core of the Raspberry Pi 3B to record analog vibration data being converted by an 8 channel 340 megahert ADC, another core to run an analysis, another core to sift and send the calculated data to the telemetry unit as digital data and the last core to store the analog data in onboard memory. There is a small chance the rocket may not be recoverable, so we are going to send the frequency analysis back to wallops via the telemetry lines and store the original data to on the Raspberry Pi to hopefully collect later. This way, we will at least secure the analysis, even if the rocket in unrecoverable. We will be sending ADIS gyro/accelerometer data back as well to use as our working standard. The piezos will be attached to the protected enclosure being built inside of the canister. Each piezo will be placed inside of a 3D printed PLA or ABS casing that is then screwed to the inside of the enclosure. We may use Epoxy CL-30 if required. The white sheets on the American Piezo website notated Epoxy CL-30 as being an appropriate adhesive. We will run testing for all configurations to assess the difference (delta) in readings from a single source.

Thank you both for your replies!

btrueblood - That pun made me chuckle a bit. I almost want to say I am disappointed... almost.

 

[btrueblood]

Dad jokes are no extra charge.

 

[Strong]

"8 channel 340 megahert ADC"
Really fast ADC for the low frequency data you are expecting! What about anti-alias filter?

"Each piezo will be placed inside of a 3D printed PLA or ABS casing that is then screwed to the inside of the enclosure."
Will you attach a seismic mass to the piezo to make it into a real accelerometer?
Are you following this guide: Piezo Accelerometers

https://www.americanpiezo.com/piezo-theory/piezo-accelerometer.html

"Epoxy CL-30"
Loctite E30-CL Temperature rating for contact with surface is 250-F, so what about protection from reentry temperature of 500-F?

Walt