The speed of sound in materials

Hi there

In principal I know what to do. Send a sound pulse into a material and measure at the other end when the pulse arrive. But if I would like to measure the speed of sound in ... lets say custard... How would one attack that.

I have some Ultrasound distance devises for my arduino (ala: http://www.dx.com/p/hc-sr04-ultrasonic-sensor-distance-measuring-module-133696#.VNpYWixRbkY) so...

1) Is it possible to modify these and cast them into the custard and let the sound bounce off on the other side of the custard (like in normal air distance measurement)? (I am thinking to determine the time from Ping out to echo back, then by known custard thickness - bar-da-bing Speed of sound in custard)

2) Is it possible to use one unit as a sound giving device and the other as a sound receiving device?

3) (related to 2)) could I just desolder the microphone add wires and then make it work from A to B (without the bouncing).

All comments are very welcome - please remember that I am not an electronic engineer - just a plain Geek having fun. As you guessed - next step is elastic modulus of custard from these measurements....

Yours Mogens

It is certainly possible to measure the speed of sound in materials with an Arduino, but you will probably have to build a custom apparatus. Inexpensive ultrasonic ping devices won't work well as the transducers are designed for use in air.

A piezoelectric disk as the emitter, a microphone and some circuitry should do the trick.

That could work.

But, I don't see any detailed specs or application notes...

The "bounce" method would work if you can make good contact with the media (you might have to bury the transmitter & receiver in the custard) and if you have a solid surface to bounce off of.

As far as I can tell that senor doesn't directly give you distance, but if it does, you could compare the reported distance to the known distance, and knowing the speed of sound in air, calculate the speed in the unknown media.

Otherwise, it's just a matter of sending a pulse and timing how long it takes to return (or the reach the sensor on the other side).

The biggest limitation will probably be the speed of the processor. Every step the processor takes, takes a certain amount of time. So, if the distance is too short and the pulse comes back before the processor can get around to reading it, you'll miss reading it, or you'll get an error. (If you program in assembly/machine language you an look-up the time for each processor instruction, but in C/C++ it more uncertain.)

There is a micros() function that measures time in microseconds. I shouldn't do this... But, I think sound travels about 1/100th of an inch in 1 microsecond (in air). That's just a quick-and-dirty rounded-estimated calculation and I could be way off. And there could be more than one microsecond of "overhead" in your program, but I have no feel for that.

The speed will be faster in solids, liquids, or gels, than in air, so the minimum distance will be greater. The specs say that particular device can go down to 2cm (in air) so minimum distance might not be a problem. But , It doesn't actually say it can measure distance at 2cm (maybe it just detects an object at 2cm). That's the bounce distance, so you'd have to double it for a one-way measurement.

I think you'd have problems with ultrasonic in custard for a couple of reasons.

Firstly, ultrasonic waves are greatly attenuated in liquids. So echos would become much less distinct.

Secondly, sound travels much faster through liquids than solids, so you'll have issues with the speed of the arduino.

Thirdly, custard is a non newtonian fluid so it'll likely stop ultrasonic waves even more rapidly than other fluids. So I don't think it's going to work (for custard at least).

Unless you have a very large container of custard, you are going to get false echos off the sides of the container.

Paul

I think the biggest problem is going to be coupling.

The piezo element should not have an air gap between it and the custard.

Something like a fish finder head .

Ever tried running on custard :slight_smile:

Thank you Riva, beat me to it.

Dear all

Thank you very much for your input - highly appreciated.

In the walking on custard vid's... They are not walking on custard! Custard is boiled corn starch, flavored with vanilla (normally). After cooking you cool it down and it sets into the wonderful dessert/cake filling (worked 10 years at professional Baker - I have used my share of custard). What the lad's at walking on is a water starch mixture - close to a saturated starch concentration. Now when you try to deform this mixture you will experience shear thickening (like walking on wet sand) due to the repulsive forces between the starch PARTICLES under your foot that will try to move other particles and other and other - at the end they are trying to move the swimming pool. When standing still the "solution" have time to relax and the particles have time to move around your foot - thus you will sink into the solution until your boyency balance the gravitational forces.

If you try to walk on custard - you will sink in right away (but you will be tasty right after). As on cooking the starch you will remove the particles and have the "single" stranded starch molecules (polymer) partly hydrolyzed and solvate it in the water phase at concentrations above its gel point.

But this was not my problem - I just do not know how to measure the speed of sound in custard. And now when it seems to be hard I would like to do it even more.

ok recap: 1) Speed of sound in custard is faster than in air (check) 2) The freq of the arduino is 16 MHz - which might be to slow (check) 3) I need a lot of custard (Ok - it is cheep and i have a bath top somewhere - check) 4) The speakers and microphone have to have a close contact to the object (check) 5) The item (that I have 5 spear off) I have might not work (bummer) 6) I am now obliged to try (check)

So the experiments first at hand would be: a) dip the electronics into custard - did it survive that? b) Rip-apart one of them and send sound Across (thus no bounce) c) combine a and b

So code examples for FAST (from the sound of it (ho ho)) really fast sampling. One write "machine code" - hmmm have not done that since commodore 64....

Any comments and ideas are very welcome.

yours Mogens

Do you think the speed of sound in custard is likely to be much different from the speed of sound in 40% cornstarch suspension?

The speed of sound is found to range from 1483 ± 10 m/s in pure brine to 1765 ± 9 m/s in the 40% cornstarch suspension. The bulk modulus of a granule of cornstarch is inferred to be 1.2(± 0.1) × 10^10 Pa.

http://scitation.aip.org/content/asa/journal/jasa/133/3/10.1121/1.4789926

hinge: ... next step is elastic modulus of custard from these measurements....

If you're going to be slightly scientific about this, you'd better define the term Custard first. :)

Next time I'm in the lab at work (and remember to take a custard sample) I'll take a few velocity measurements so we can compare notes.

I think this is what we'll find:

jremington: ... the speed of sound in 40% cornstarch suspension ... 1765 ± 9 m/s

Another few issues. I think the speed of sound through custard will also vary with temperature. The speed of sound generally increases with density. As custard warms it will become less dense and therefore the speed of sound through it will slow. This will be FAR from a straight line graph though and would probably make for interesting research.

Another issue is that the container could provide an alternative ( and quicker ) path for the sound. Thereby invalidating your readings. Using a container made of, something like, foam rubber may prevent this.

Martin-X you are right - I did not define my system very well - I will estimate polymer concentration in custard ASAP and post it.

On the speed of sound issue I am sure that it will make a difference running the sound in custard (heat treated corn starch) and in a particular suspension - I will assume that all the particles will give an echo - where as the custard I assume that it will behave as one homogenous material. But that is all guess work by now.

Comparing data - I love it!

I will return as soon as I have tried out some experiments.

Yours Mogens

I am planing:

150 g corn starch (Mejsmel 98%, Coorp, DK) in 1L of milk at 5 degrees C (Minimælk, 0.5% fat, Arla, DK), Mixed cold in a 4 L pot and heated to a boil during virgours steering with a whisk. When at boil keep boiling for 4 min still whisking hard. The now thickened pail yellow solution (not colloid suspension any more) is poured into a container (Hjemmeis, DK) and directly placed into the fridge (5 degrees C, Gram, DK). When cold the custard is applied in further experiments.

So compered with the article who apply 40% by mass starch particles in a colloidal suspension I am expecting to use approx 15% by mass in a true solution (I have not corrected for the lower density of milk, and temperature (5 degrees C when measured), hence assumed that 1L of milk is 1000g - which is wrong - I know).

Quick calc (thank you jremington for the reference): Seed of sound in 40% colloidal suspension is 1770m/s = 177cm/ms =1.7mm/us Thus making a custard that is 100 mm long would give about 56.5 E-6 sec for measurement. Assuming that I am late by 2 us then I will get 1724m/s. Then my error would be (1770-1724)/1770 = 2.5%. I do not think that I can reproduce my custard with that precession. So it is possible with an arduino. I do not know how to interpret the 16MHz clock of the arduino in relation to this calculation.

I am still blank when looking int the programming of my arduino for very fast sampling. And suggestions/experience on sensors would be nice to.

Yours Mogens

Sorry Riva - I did not answer your question - Yes

I have. We made the same experiment in a kids inflaterbel pool and with potato starch at an sciences exhibition once - grate fun - but only the first few days - the fermentation set in.

Lets focus on the eletronics and dataaquaiaion here forth.

Yours Mogens

hinge: Quick calc (thank you jremington for the reference): Seed of sound in 40% colloidal suspension is 1770m/s = 177cm/ms =1.7mm/us Thus making a custard that is 100 mm long would give about 56.5 E-6 sec for measurement. Assuming that I am late by 2 us then I will get 1724m/s. Then my error would be (1770-1724)/1770 = 2.5%. I do not think that I can reproduce my custard with that precession. So it is possible with an arduino. I do not know how to interpret the 16MHz clock of the arduino in relation to this calculation.

I am still blank when looking int the programming of my arduino for very fast sampling. And suggestions/experience on sensors would be nice to.

Using an external timing chip (GP22) connected to an arduino I can measure (*specific) cable length quite accurately so something as slow as the speed of sound should be no problem for timing. The big problem is how and what sensor will be suitable and I have no idea on that front. *specific meaning I need to know the Velocity ratio of the cable being measured. Normally about 0.8C

Non-scientific results from my quick experiment…

20% by volume custard powder in water, boiled in a microwave. Sonatest Sitescantm UFD, 5.0 MHz 10mm diameter twin crystal probe, pulser voltage 200V:

Hot fluid state: 1603 m/s
Cold jelly state: 1623 m/s

Martin - X Super cool. Thank you very much.

Yours Mogens

Dear all

Sorry for the long reply time - but my day job....

Ok I did some experiments.

1) You can de-solder the ultra sound transmitter and receiver. Then by adding wire one can send directly across a giving distance (above 4 cm). This works as a charm and after a series of measurements at different distances I got a nice estimate of the speed of sound i air.

2) I "mounted" a slap of custard in between the transmitter and receiver and wupti I got a much higher speed - almost the speed of sound in water - I was excited until I tried on a shorter slap (and a longer slap) and got the same time (not speed) out... It turned out that if i touch (wet fingers) the trans and receiver - then I get the same value. Bommer! It showed that when electrical contact between trans and receiver = no measurement (a number... yes but not a measurement).

Back to the drawing board and now i finally got a waterproof trans/receiver for my project so soon i will try again - ASAP.

End of update.

Martin - X you are not forgotten and I appreciate your readings.

Yours Mogens