Hi, I think you will struggle here. I don't know what a hele-shaw cell is. For ten minutes you will need a near perfect match on the density. I used non-expanded polystyrene in water for LDA work in pumps. I think the error was about 5% OK for a strong convection situation. Your case is stagnant and long time base so perfection is needed. You need to match density maybe you could introduce a soluble salt into the water to change its density progressively until it matches the particles? Who knows! Another angle you might try and exploit is some other force on the particles. Maybe magnetism or electrostatics, designed to counteract the buoyancy issues. This way you might have suspended particles at time zero and switch off the additional body force before the experiment starts.

Dear Martin, thank you for your answer. Unfortunately the lightest plastic commercially available still has a mass density above 1050 kg/mc; adding salt to water in order to reach that density modifies too much the properties of the system of my experiments. Also electrostatic forces cannot be used since the windows of my Hele-Shaw cell are in Lexan, a quite active material in presence of charges.

However it seems that pine pollen (raw) is a good candidate. It is quoted as having a mass density of 1000 kg/mc. I will test it.

Sincerely

Sandro

Hello Sandro,

This is a so difficult and unsolved problem ! i had always problems of non 1000 kg/m3 densities. Sure Pine pollen would be good choice, but i thought that pine pollen particle diameter was greater than 30 microns, very often near 50 micron. . May be now companies propose now smaller pine pollen particles ? The lighest polyamide particle are now 1030 kg/m3: which is 3% too high (not 5%) .  see:  http://www.dantecdynamics.com/seeding-materials    . But i guess that 3 % is also too much: you need no more that 1/1000.

cheers

hubert

Ciao Hubert,

I thought the same thing. However we are working with a macro lens and the field of view is 40x40 mm2 at 1024x1024 pixels. The overall resolution is nearly 40 micron/pixel. Hopefully the pollen particles will lump in order to occupy at least one pixel. I will let you know, thank you.

Sandro

Did you try milk! I think low concentration would be sufficient. Problem: as organic material, with time, can deteriorate and produces bad smell, so you'll need to change the samples often.

There is a relatively easy solution, but it may get expensive if you need a large volume of water: using D2O (heavy water) instead of H2O.

The average density of D2O is 1110 kg/m3,  which will leave you with a very large choice of adapted plastic materials. At the same time, thermal expansion coefficients, fusion and boiling temperatures, and even surface tension with air of D2O are only marginally different from H2O.

I cannot give you the average price of D2O right off the tip of my mind, but if you compare it with the overhead salaries associated with long and difficult manipulations with a less handy solution, chances are that it may even become cost-effective.

If you ever try it, let me know, I would be interested by some feedback.

Good luck !

Pierre

Dear Antonio,

I will try also milk but I suppose that it will mix with water as soon as the movement starts or as a consequence of Brownian diffusion. I also think that milk does not organize itself in droplets but in other larger coherent structures that may help PIV analysis but probably without the needed resolution. Thank you for your answer, I will let you know.

Dear Pierre,

I really need a small quantity of fluid, so I could try following your suggestion. However I checked that the main difficulty is in selecting particles having exactly the same density of the fluid, not in increasing the density of the ambient fluid. Surely increasing the density of the fluid may help since more plastics are available in the range of the D20. Thank you for your interest and suggestion.

Meanwhile I have tried pine pollen (raw) which is quite promising since putting a small quantity of it in a glass of water, you can observe a cloud of suspended pollen. Part of the pollen slowly plunges or moves upward, but a decent quantity of it is still floating. I will select the neutrally buoyant pollen particles to be used in my Hele-Shaw cell. I also think that you need to wait one day or more since pollen particles absorb water. It helps the plunging of some of them which otherwise are floating on the free surface.

Some preliminary tests confirm what Hubert Branger wrote about the size: the pollen particles are very tiny and surely are perfect in LDV but with PIV you need a macro lens to observe them moving.May be it is possible to 'glue' some particles in clusters but it requires several attempts in order to detect the proper glue, and a proper technology.

Thank you all

Sandro

I did PIV in water sloshing. At the incipit of the motion, the fluid is at rest and particles are quasi-still. The perfect density matching is difficult to be obtained but if your transient motion is strong enough to respect Stokes number for PIV, you can neglect the residual motion of the particles due to buoyancy.

Are you using water as test fluid or something else?

I used fluorescent particles of 45um with a density of 0.995g/cc in water: rest I applied a sinusoidal motion with A=1.2mm and F=2Hz.

I'm sure that those particles are suitable for your experiment if the fluid is water.

Dear Alessia,

thanks for the information. The problem is that the set up of the PIV, made with the fluid at rest, requires the presence of the particles in the flow field. In addition, the thermal induced convection is extremely small and the particles start to move only when it has increased enough. Since we want to detect the critical Rayleigh number, we need tracers with a quick response, otherwise a systematic error is introduced in the evaluation of the critical Rayleigh number.

Could you please specify the commercial name of the product, and where is it possible to buy it?

Thank you

Sandro

Dear All,

the particles with a given specific weight and diameter can be purchased online at

https://www.cospheric.com

The stuff sold by Cospheric is not cheap, but presumably could solve some specific problem.

Thank you all for your valuable suggestion

Sandro

Dear Sandro,

Sorry for the late replay, somehow the notification of Researchgate didn't inform me about your following up to the message!

Indeed, I purchased my particles by Cospheric! I have to say that, particles of 0.995g/cc where quasi static inside water at 20deg at rest. I tested buoyancy of this condition (cause you never match perfectly the density) and it resulted in a peak motion of 0.002m/s. In my case this was acceptable respect to my fluid motion.

Happy that you finally found Cospheric.

Dear Alessia, thanks, it happens to me as well and indeed I did not receive the notification of the last message of yours.

May be you meant 0.002 mm/s, it seems to me that the peak value of 2 mm/s you referred to in your message is too high. Am I right?

No, I meant m/s, but indeed it depends on the applications and on the largest  separation time one need to use for his particular applications.In my case it was an harmonic motion, so in the moment of zero velocity in order to appreciate a displacement of 8 pixel I set the separation time at 0.099s

If I consider this separation time respect to the buoyancy flow which has a velocity of 0.002m/s with a magnification factor of 0.105mm/px, I calculate a displacement caused by the buoyancy terms of (0.002*0.099/0.105)=0.002pixel.

Normally the subpixel resolution of the cross-correlation function is of the order of 0.05pixel.... I would say that I can not appreciate this mouvement. And I consider it as negligible respect to my main flow.

Maybe I didn't specified that to be able to have the 8pixel displacement for only the buoyancy flow, without mainflow I had to set a separation time between images of 1s which gave the 0.002m/s.

I really think that it will depend on your typical flow velocities and resolution that you use. For example, which is your initial velocity from rest? Could you have an estimation on that?

Dear Alessia,

there is an horizontal imposed flow of less than 1 mm/s. Then thermal convection should induce a vertical velocity with initially value equal to zero (fluid domain), up to a fraction of mm/s in stationary regime. However, I am using power-law fluids with high apparent viscosity (two order of magnitude larger than in water), hence the vertical speed of the particles due to gravity should be much less than in water.

Sincerely

Sandro

What is the fluid? can you find iso-density particles (neutrally buoyant) or is it impossible due to temperature changes? would a "micro-bubble" work? sometimes we observe those in gels and they're not moving due to viscosity (although not perfectly non-buoyant) 

The fluid is a power-law shear thinning. It is possible to find controlled density particles, but the variation of temperature generates a variation of density of the fluid. It is true that constant density particles (even microbubbles) have a very slow motion in the fluid. However, since we need to measure very tiny velocities to detect the incipient instability (up to 10 microns per second), they are still too fast.

what about microporous particles? These would absorb the fluid and hence their total density would adapt to the fluid. Looking through articles they are used in medical applications for drug trnsport and also called "microsponges". Here an article:

http://www.ncbi.nlm.nih.gov/pmc/articles/PMC3255415/

It seems a good idea. However, microsponges must be prepared in a chem lab, since I could not find any commercial seller on the web. I have downloaded some articles which explain the preparation of microsponges, it is not easy if you have only an hydraulics lab