Science Buddies: "Ask an Expert"

Currently, I'm developing an idea about viscous couplers, a type of automobile differential. In a nutshell, a viscous coupler transfers torque from rotating plates with different rotation rates via a shear-thickening fluid medium. When the plates in the viscous coupler rotate at different speeds, it generates shear pressure, resulting in a rise in viscosity of the fluid to a near-solid state.

I would like to replace the standard shear-thickening fluid of colloidal silica with a kaolin clay suspension, as kaolin clay is far cheaper, being a common, naturally occuring mineral. Kaolin clay has been demonstrated to exhibit significant shear-thickening properties as a reinforcement for Kevlar in the following experiment:
http://www.ccm.udel.edu/STF/PubLinks2/M ... EMay07.pdf

Before I can conduct any sort of experiment, or even be sure that this is feasible, there are many things I need to know about viscous couplers that I don't know. I'm no mechanical engineer, sadly.
Would it even be possible to open up the unit and replace the fluid? Would there be a way to accurately test the response of the shear-thickening fluid at certain speeds of rotation? How would I go about attaining these speeds accurately? Also, I would like to learn some more about the in-depth mechanics of how the viscous coupler functions.

Any help would be greatly appreciated!
Thanks for your time,
- Richard

Richard,

I don't know much about Viscous Differentials, but a little bit of research suggests that many, if not all, are pressurized devices and may be difficult to open up and replace the fluids. Also, it's my experience that running experiments with the real specifically engineered devices without some speciallized equipment tends to be difficult. I would suggest designing a simple analog test article like what I describe below.

Your suggested research is very interesting, and could be a study of the torque transfering phenomina known as viscous coupling which is of course leveraged in a viscous differential.

You could fairly easily build a test article that could be used to identify materials that provide the best effect and the optimum surface areas needed with each liquid to provide the desired effect.

My idea of a test article is a free hanging Cylinder inside of a bucket. When the bucket is filled with a liquid and rotated, how long does it take for the cylinder to start rotating. The bucket could easily be filled with various "liquids" and the torque transfer measured. Also, different size cylinders could be used to determine the effect of change in surface area ratios. To relate to your project idea, if you were able to obtain the two materials you mention, these could be tested along with the others.

I'm sure some refinement of this idea would be required, but I hope this helps.

Cheers,

I'm very grateful for your response. However, I have been thinking about this idea for awhile, and I am having some trouble devising a procedure which would produce quality quantitative data. I am unaware of any method to maintain a consistent rotational speed of this bucket, or or any way to accurately measure the rotational speed/force of the cylinder at any given time without direct contact, which would alter the results. I am also having some difficulties in devising a way to design the experiment so that it could accurately represent a viscous coupler - what sort of speeds would I need to test for it to be applicable to a VC unit, for instance.

The idea has been a great help, though.
Thanks again,
- Richard

I don't know much about viscous couplers except the basic mechanism of operation, though I do know a fair amount about torque converters, which are fluid couplings that use the momentum of the fluid as a coupler rather than the viscosity.

It might make an interesting experiment to compare a viscous coupler to a fluid coupler. In particular, a viscous differential and a torque converter from an automatic transmission (both can probably be found at automotive junkyards). There are several different areas for comparison, such as heat loss, efficiency at various speeds, etc., etc.

Ray,

I've researched quite a bit about fluid couplers after reading your suggestion. However, how would I be able to control any variables of an experiment like that? I could not control the type of couplers I found at the junkyard, and I am unaware of any way to apply the same torque to each coupler, for instance, but I will look into this further. If you have any knowledge of how I could create a more detailed procedure, that would be amazing.

Thank you so much,
- Richard

While waiting for responses, I have been thinking of a procedure design, of sorts, following theborg's guidelines. If anyone could help me revise it, or give me any advice on whether it is practical, that would be very helpful.

The system consists of a hollow, cylindrical container of sorts has a free hanging cylinder inside, suspended in shear-thickening fluids of varying types: a standard silicone-based coupling fluid, and multiple kaolin clay suspensions in glycerol of varying wt%s. The surfaces of the cylinder and of the container will have protrusions imitating that of the plates of a viscous coupling. The system will be rotated at a certain speed, and the rotational speeds of the cylinder will be measured at certain times with a tachometer. And from thereon out, the rest should be just calculations.

That still leaves me wondering how I will consistently rotate the system at a certain speed, and how I should go about creating protrusions on the cylinder and container. If anyone has thoughts on how I might possibly do either of those, I would very much like to hear them. It doesn't seem that using a centrifuge of any sorts is plausible.

Also, would this be able to accurately determine the potential of a kaolin-based STF as a coupling fluid? Would the conditions of such an experiment be able to reliably model the performance of this fluid in a VC unit? If an expert could give me their opinion on what I have thus far, I'd be much obliged.

Thanks so much,
- Richard

It looks like you are really attempting to work out a good test proceedure. I'm really impressed.

It's important to remember that you are testing the properties of the different "liquids" and thier ability to transfer momentum from one shaft to another. This is an important first step before determining suitability as an operational material in an actual VC unit. Remember, the VC on a vehicle must perform in extream environments...significant velocities (interstate speeds) while propelling a 2,000 lb or heavier vehicle. The slightest miss-alignment or underperformance of the fluid, could cause loss of control of the vehicle. To test this would require some very speciallized safety and measurement equipment. So let's just start with what we have.

I think that such a test device as you describe, would start pointing to the potential of your different fluids. I think a key measurement would need to be "reaction time". This I would define as the amount of time it takes a change in angular momentum of your drive shaft to be transfered to your test shaft. What happens if it's gradual, sudden, low to high velocity, high to low velocity, etc? ...many, many posibilities for testing. As for the protrusions you discuss, i'm not sure you need them. I'm not 100% sure, but I imagine they are there to increase the surface area and create a bit more "drag" one each opposing plate in a VC. However, if you feel you need them, a simple lag bolt in the sides of each cylinder, sealed with calk or something to prevent leaking should work fine. The thing to remember is to keep an eye on balance! If you're unballanced at all, the whole contraption could rip itself apart at higher speeds.

A thought on controling rotation speed. I suggest hooking your drive shaft to a motor and using a variable resistor to control to speed of the motor and therefor the rotation of the drive shaft. A little experimentation will help you calibrate turns of the potentiometer knob with speed of the motor.

I've uploaded a .pptx file as a quick sketch of my idea for reference. Note that the drive shaft and test shaft should be as in line as possible, and steady in the X and Y axis, while alowed to rotate about the Z axis. See how it matches with yours ideas.

theborg

After looking into the design formed above, I have formed a lot of questions. My chief concern would be this: since the axes are rotating at such high RPM's, and correct alignment of the shafts is so crucial, what sort of measures would be necessary to consider the shafts "properly aligned"? There would have to be some mechanism to allow the upper shaft to free-hang and remain aligned when such torque is applied.

The diagram seems to represent my own design quite well, filling the hole of the rotation mechanism as well. However, as I seek to refine the details of my procedure even further, I wonder: what type of motor would I be supplying the torque with?

Just a few of many, many questions that I feel I will have difficulties finding the answer to on my own. If you have any ideas, I would love to hear them.

P.S.: If anyone, especially theborg, is willing to let me get in contact via e-mail or other methods that are more convenient/faster, an e-mail address or something of the sort would be amazing! Thank you for your time!

Once again, thank you very much!
- Richard

P.S.: If anyone, especially theborg, is willing to let me get in contact via e-mail or other methods that are more convenient/faster, an e-mail address or something of the sort would be amazing! Thank you for your time!

Thanks yet again,
- Richard

I'm sorry, Science Buddies does not allow contact outside of the forums. Any posting with email addresses or other contact information will be removed.

Regarding your question about alignment, I think theborg was trying to say that couplers actually mounted in cars have serious environmental and safety requirements that make taking measurements directly from them while mounted on the vehicle difficult, dangerous, and hard to control.

An experimental setup like the one he suggested won't be nearly as difficult to deal with. Any variable speed electric motor would work fine in such an experiment. The alignment could be done with a variety of tools, ranging from things as simple as a plumb-bob (a weight on a string to measure that both shafts are perfectly vertical, combined with using the same plumb bob to take 3 measurements ensuring that the two shafts have the same center point (essentially by triangulation).

Actually, a very interesting independent variable you could consider varying would be the alignment: how does offset or angle variation affect the efficiency of power transfer?

BTW, I will say that it's very encouraging to see such original thought being put into a project.

Thanks Ray, excellent comments.

I highly suggest performing low speed tests to understand the physics involved and to be able to identify all the variables and effectively isolate them for testing. For now, you wouldn't come anywhere near the environment conditions of a VC. Again, the key is to understand the fluid and how it reacts in an analog to the real system.

Remember, the nature of a true scientist and researcher is "patients".

As for contact, my SciBud account is best, so keep posting here and I will continue to check back.

I think this is really original! Keep up the good work.

By the way, that would be "patience" vs "patients"

It's good to know that I'm somewhat on track with my ideas. Thanks so much for your replies!

After giving this a lot of thought, I've started to design a detailed procedure, based around the quick sketch above. The biggest concern of mine is, how would I be able to keep the upper shaft properly aligned as the rotation applied? Would having the same center point and correct alignment at the beginning be sufficient to guarantee that it does not change in the middle of the experiment? I would prefer not to add another variable and make things more complicated. Also, I can't seem to think of a solution for the actual suspension of the upper shaft. I wouldn't think that hanging it from a ceiling of sorts with a string would work well, since there would be a point where the string itself would be twisted to a point where it would counteract a portion of the rotational force, although the error in both experiments could cancel each other out, I'm not quite sure.

This may sound silly, but I can't think of a way to keep the whole model upright without affecting the spin, either. If I were to use a flat-bottomed shaft, with the shape of an upside-down funnel, of sorts, would the friction between the ground and the shaft produce an error in torque transferred from the motor to the shaft?

Also, I'm under the impression that VC's are sealed units. Would using an unsealed bucket or something of the sort affect the results?

I'm having trouble determining a specific motor to be using here, too. (I'm not the most knowledgeable on variable speed electric motors, please bear with me )

And finally, how would I decide whether my STF is a suitable replacement? Since the conditions I'm applying won't be anywhere near the conditions of a VC, I can't judge whether it will function as a proper replacement, only what the general properties are in relation to the standard STF, such as the speed at which it transfers torque.

I know these questions are specific and tough to answer, thanks so much in advance!
Sorry for the delay, and many thanks,
- Richard

There seem to be too many models of electric motors out there - in particular, I've been looking at DC motors with variable potentiometers, none of which are too cheap. However, I don't know which motor would best suit my needs, and the size of my drive/output shafts and many of my other materials rely on the type of motor I use, so I would like to make a good choice.

As for my problems with the previous configuration, would something like the schematic below be practical? There aren't many resources on the web about these mounting hubs, so I'm curious as to if they will be a proper substitution. After looking on the web for a drive cable and coming up empty-handed, I thought that the drive cable in the schematic from earlier may just rotate around the drive shaft without actually applying torque to the drive shaft itself. Am I right, or am I completely off target?

At the moment, I'm considering simply conducting the experiment to find the torque transferring capabilities of the kaolin clay suspension alone, as well, since obtaining a sample of VC fluid (a polyorganosiloxane suspension, upwards of $100 in the smallest sales units) is quite expensive. Whether the kaolin clay suspension demonstrates a significant amount of torque transfer in a reasonable amount of time could be a relatable result, perhaps? I'm having trouble finding a way to relate the properties of the kaolin clay suspension alone with no comparison to its applicability in a VC.

If anyone could help answer any of my questions, I'd be very grateful. As always, thanks so much for your help!
Many thanks,
- Richard

I guess my posts have been a bit too specific... To broaden the questions I have a bit, maybe I could reword them like this?

Would it be possible to estimate how well a fluid transfers torque at a high speed without actually exposing it to that speed?
Will the surface area/volume ratio of the model affect the torque transferring of the fluid? (so I can determine dimensions of the model I'll make)
How can I connect a motor to the drive shaft?

As always, thanks so much!
Happy holidays,
~ Richard