Let's talk about centrifuges

[Sillychris]

I need to spin the hell out of a glass vessel with fluid inside. So far I have gotten an erlenmeyer flask up to 2400 rpm with water inside before it shatters. It seems to pick up some kind of vibration which shatters the glass. The setup seems fairly balanced up to that point and I suspect the fluid inside has a self balancing effect on the entire apparatus. ultimately, I'd like to push the vessel up to 13000 rpm but I need to kill this vibration in order to do it.

So who knows stuff about spinning things?

I have since removed the driveshaft bushing. That did reduce the vibrations but not enough. I am also starting to suspect that a vaccuum is not necessary.

Edited March 22, 2014 by Sillychris

[Stargate525]

I'm shooting from the dark here, but it may have to do with vibrations in your engine/shaft assembly? Could you get the desired effect by increasing the arm length? Less rpm, but the same rotational speed at the edge.

[Sillychris]

I'm shooting from the dark here, but it may have to do with vibrations in your engine/shaft assembly? Could you get the desired effect by increasing the arm length? Less rpm, but the same rotational speed at the edge.

I like the shooting in the dark because it may just introduce something we haven't thought of. I'm really hoping it isn't vibrations from the motor because then we need a new motor. But it probably is. Maybe I should post a figure...

[KASASpace]

How are you rotating the glass?

On a stick like a shishkabob? (lord how do you spell that!)

Or completely?

Because if on a kabob it will eventually start gyrating due to the fact that not all the glass is getting accelerated at once.

That's what I know.

[Sillychris]

How are you rotating the glass?

On a stick like a shishkabob? (lord how do you spell that!)

Or completely?

Because if on a kabob it will eventually start gyrating due to the fact that not all the glass is getting accelerated at once.

That's what I know.

I posted a figure of the apparatus. I have since removed the driveshaft bushing since it was catching on fire. GOing to give up on the vacuum for now and see if it is actually needed.

[KASASpace]

I posted a figure of the apparatus. I have since removed the driveshaft bushing since it was catching on fire. GOing to give up on the vacuum for now and see if it is actually needed.

The shaft is very thin compared to the table and the glass.

This will eventually cause problems with vibrations as you can never perfectly have the mass centered on the plate, some of it is leaning towards a certain side, and less of it the other, this difference causes the glass to move, and since it is rotating rapidly, it vibrates.

If you got a smaller plate and glass, or instead of using a drive shaft used a gearing system, that might help.

[Stargate525]

...Why do you need to spin a flask like that? I'm just curious, never seen it needed like that before.

And I'd imagine at those speeds you might be getting negative impact from imperfections in the glass throwing off your balance. Dunno how you would account for that, though. Would adding a flywheel dampen vibration?

[Camacha]

...Why do you need to spin a flask like that? I'm just curious, never seen it needed like that before.

I ask myself the same thing, as I think useful applications typically involve an asymmetric center of gravity (like a blood centrifuge) or an agitator (something inside the flask to stir the liquid). Now you just spin the glas quickly, but allow the liquid to stay relatively stationary inside.

Edited March 22, 2014 by Camacha

[Nuke]

i bet your vibrations are hitting the glass's resonant frequency. the frequency of the vibrations from any imbalence will probibly be proportional to the motor's current rpm. so if you change up the design so that the motor spends as little time as possible at an rpm which will produce the resonant frequency it should stay together. once you get passed that you wont need to worry about it until you spindown.

[jwenting]

also remember that the forces on the glass from the spinning liquid inside (which spins at a different rate from the glass itself) are considerable. This may well be enough after a while to shatter the container.

[Sillychris]

The goal is to create a particle detector. Recoil momentum from a weird particle needs to be sufficient to fracture the liquid while under high centrifugal forces.

All the ideas so far are ****ing excellent. My favorite is make the driveshaft thicker.

[Tommygun]

If the vibrations are coming from the motor, then could you set it up to run on a drive belt to help isolate the vibrating motor from the glass.

Does the container have to be glass?

Maybe a heavier base plate could act as a flywheel too? If it has more mass it may take more energy to create significance vibrations.

This might be worth looking at, as it lists sources of torsional vibration and methods for reducing them:

http://en.wikipedia.org/wiki/Torsional_vibration

Edited March 22, 2014 by Tommygun

[Ravenchant]

Also, it may not hurt to block the casters, maybe with hard rubber chocks or similar...just in case there is any vibration coming from the pump and being transferred to the table.

[Nuke]

yea do a drive belt if you can, and be sure you mount the motor on rubber bushings.

[Creature]

Finally decided to register on the forums just to answer you

So first thing, I'd strongly recommend not using an erlenmeyer as they're built mostly just to be chemically inert and if it's pyrex then to withstand heating and not so much to withstand heavy mechanical forces. In a typical lab centrifuge the g-forces involved are in the range of several thousands so most likely just the pressure acting on the container will shatter it even if you could eliminate all vibrations. Unless it's absolutely critical to use glass, go for plastic which is also a lot safer.

If the liquid moves fast in relation to the container wall, it might be possible that the friction between the liquid and container wall gets too high and causes turbulence inside, but this is just a speculation as I really have no idea how the liquid behaves when spun that fast. Also the acceleration time can be a factor, if you push it too fast it might again cause turbulence. Then there's always the chance as someone mentioned already that the erlenmeyer has an uneven weight distribution as it's not a factor in their manyfacturing really.

I don't know how what's your background, but the importance of balancing can't be overstated. When we use in a lab a big washing machine-sized centrifuge to spin bottles that weight around a hundred grams or so, using a rotor that weighs several kilograms, the opposing bottles still need to be balanced so that there's less than one gram of weight difference between them. So even tiny differences are important.

I didn't really understand what you're doing with it, but is it necessary to spin it around it's axis? If you can use two containers, then using a swing might be a solution.

[diomedea]

Given how fast you want to spin your flask (13000 rpm), I would use a technology similar to how jet turbines are spun, that is, using compressed gas on a turbine rotor. There may still be vibrations, if any of the blades on the rotor is not symmetric, but it requires a very high speed for such vibrations to be relevant; instead you will certainly eliminate any from the electric motor and gear you are currently using.

[shynung]

Given how fast you want to spin your flask (13000 rpm), I would use a technology similar to how jet turbines are spun, that is, using compressed gas on a turbine rotor. There may still be vibrations, if any of the blades on the rotor is not symmetric, but it requires a very high speed for such vibrations to be relevant; instead you will certainly eliminate any from the electric motor and gear you are currently using.

For this idea, an automobile turbocharger's turbine could do the job.

[Prime flux]

Given how fast you want to spin your flask (13000 rpm), I would use a technology similar to how jet turbines are spun, that is, using compressed gas on a turbine rotor. There may still be vibrations, if any of the blades on the rotor is not symmetric, but it requires a very high speed for such vibrations to be relevant; instead you will certainly eliminate any from the electric motor and gear you are currently using.

Or a air-tool motor

[K^2]

Liquid is only going to provide self-balancing effect if certain conditions are met. If you model your system as a damped harmonic oscillator, there is going to be a phase angle between direction of centrifugal force and deflection. If the natural frequency is higher than your 2,400 RPM, the angle is small, and the effect is destabilizing. If natural frequency is low compared to the angular velocity at which system is spinning, the angle tends to 180°, and it becomes self-balancing.

So it's entirely possible that your system is too stiff, preventing self-balancing.

The other question is purely on the ability of glass to withstand this kind of abuse. At 5cm and 2,400 RPM, you'd pick up something like 300G. Are you sure the glass you are using is designed to handle this?

[Alchemist]

First, the conical Erlenmeyer flask is not meant to sustain high pressure differences or have high structural integrity (both because of the form and the thin walls), if the liquid is going to produce significant centrifugal force take spherical (or flat-bottom) flask instead. Then again, breaking at some specific frequency might be a sign of resonance effects (don't forget that any shift from platform's center already gives vibrations), that can destroy even thicker glass. Try maybe passing this value as fast as possible, also a bigger flask should have lower resonance frequencies that can be safer from this perspective.

One more (probably stupid) option - take Bunsen flask and probably cut the asymmetric top part off. At least it's thicker glass (meant to sustain 1 atm pressure from outside)

[Sillychris]

I can't thank you guys enough for all the suggestions. You've all given me a pile of ideas to work with. This is for a senior research project I am scheduled to complete in a few weeks so I obviously will only be able to implement a couple suggestions by the deadline, but I plan to incorporate all the suggestions into my paper.

I suppose I owe an explanation of what I am trying to accomplish.

The centrifugal forces will generate negative pressure in the central axis of the fluid. If the cohesion of the fluid is sufficiently high that we can keep it on the threshold of pulling apart, then an interaction with a particle could deliver enough recoil to fracture the liquid and create a tiny bubble. I am then hoping that the bubble continues to pull apart under the centrifugal forces and grows to something that can be optically detected.

For what purpose? I would like to establish a proof of concept with a neutron source but ultimately the goal is to use this apparatus as a detector for WIMPS (Weakly interacting massive particles) ie dark matter.

[Alchemist]

Also, don't forget that water at reduced pressures tends to boil at lower temperatures (or are you trying to make low temperature version of bubble chamber?). If you want something that has high cohesion and won't boil when you turn on the vacuum, consider using ionic liquids

[magnemoe]

13000 rpm is fast, very fast, one idea is to use an metal flask. this also let you increase the width of the flask.

[Sillychris]

(or are you trying to make low temperature version of bubble chamber?)

Pretty much. I'm just using water for getting the thing working, the liquid used in the actual experiment will have much different properties.

As much as I would love to machine a flask out of metal, the chamber needs to be transparent. Plastic, on the other hand is quite tempting...

[K^2]

Water will already boil at 2.3kPa and 20°C. So if you drop pressure in the flask to just above that, it will dramatically reduce the centrifugal effect you need for this to work. In fact, if this is going to work at all, that sounds like you'd have to drop the pressure.