View Full Version : Scale model - viscosity

09/02/2017, 04:13 PM
I'm building a scale model of my future tank. It's 1/3 so the 12' tank becomes 3', etc...

The intent is to look at flow along the long path (12' down, 8' side, 12' back = 32' travel)...

I'm using the concept of similitude from fluid dynamics so my scale model has an equivalent dynamic viscosity to my water tank:


But that means that I need to find a liquid that has 3 x the dynamic viscosity of water. Milk is close, but I'd need 150 gallons of milk... $$$

So - I'm thinking of mixing 50:1 splashless bleach to raise the viscosity of the scale model water instead...



Vinny Kreyling
09/02/2017, 04:26 PM
Keep any and all clothes far away.

09/02/2017, 05:49 PM
Thanks for helping me on the Scrubber Forum, hope this gives you some ideas.

How about diluting some of this or similar product. (Bath Gel, Bath Slime), if it can thicken a BathTub of water, then close to your tank size and costs less than milk. Won't rot.

Not clear since colored, but will smell better than Bleach.

Here is another Link to Water Gel. Skip the Color part and it might be clear enough.

The Polymer powder they mention may not do the trick since beady, and possibly way thicker than what you need.

But the slime formula might work if diluted right, to thicken water to your requirements.

Not sure if any of these will work for you.
Just some out of the box thinking for you to consider.
Good luck with your Model.

09/02/2017, 08:01 PM
thanks! glad for any ideas that could make this scale model more representative of the real system.

I'm sure there's a thickening agent that clorox uses. If I could get just that, it could do the trick.

I don't have a way to measure viscosity, so I'd need to know the properties and get to it by design/mixing.

09/02/2017, 09:38 PM
I'm sure there's a thickening agent that clorox uses. If I could get just that, it could do the trick.

Good luck. Over my head.

09/02/2017, 10:46 PM
nice searching!
It's over my head too - but this is the chemistry forum... someone might be able to tease it out.

The basic question is "how much of what chemical do I need to mix in to triple the dynamic viscosity of water at room temperature?"

09/03/2017, 05:20 PM
Compounds like methyl cellulose increase viscosity and might fit well within your budget. Googling methyl cellulose might lead you to other cheap thickening agents. Food and pharmaceutical preparations often require thickening agents. Food examples include flour and corn starch.

09/03/2017, 05:22 PM
Is there an off the shelf product of methyl cellulose?

Found it


Also citrucel is apparently based on the same


09/03/2017, 05:25 PM
Is there an easy way to tell when a mixture is 3x the dynamic viscosity of water?

It's = absolute viscosity / density.

I can get density by using a weight scale and graduated container.

That leaves viscosity to be empirically determined.


I love google. So it's a little complicated, but doable.

Anything easier?

09/03/2017, 05:50 PM
What about polyethylene glycol?


Or plain old food starch?


09/04/2017, 02:52 PM
I must admit that I don't know enough chemistry to help much here. The polyethylene glycol will be safe to handle, but I don't know how much to add.

09/04/2017, 03:02 PM
I'm getting all three - polyethylene glycol, food starch and methyl cellulose

I'll add the same % and measure dynamic viscosity.

09/05/2017, 06:10 AM
I'd suggest trying to measure the viscosity of water itself several times with your wiki method prior to anything else. If you're not able to get repeatable results, than the method will do you no good. The reason I bring this up is the viscosity of water is very small, and the measuring utensils are cheap and have a fairly large variation, especially for measuring density. This will likely result in large variations in your results. This method would work better for thicker fluids like hydraulic oil, but with substances as thick as water or even a few times thicker, you may find it's not repeatable enough to trust.

With all of that said, I don't know how close you need your viscosity to be in order to meet your needs.

Also, do you plan on doing a mock up of your rockwork for the full size tank to check the flow on the scale model? The rockwork in the tank will play a very significant role in how the fluid flows.

You mentioned your intent is to look at flow along the long path, I'm guessing doing a "U" shaped path around the perimeter. What is the purpose of checking that flow? are you trying to size your pumps/power heads? determine if the flow will be too aggressive in one area vs another? determine power head placement prior to the main tank in case you want them built in? Depending on what your need is, you might be getting carried away with the model where your efforts could be on adjusting the real thing.

09/05/2017, 06:44 AM
I'm running a large surge and counter surge system that hasn't been done before.

The tank is 1600 gals with a subterranean section that's 500gal. So the reef is 1100 gals and each surge is 180 gals. So - one event will release the surge on one end and then a counter surge of equal size on the other end ~ 28' away around the U.

A counter surge is when the DT acts as tube surge container and releases the volume into another.

This entire event will take 5 seconds since the openings are 4' x 1'. The intent is for this to repeat every 15 seconds. This is peak mode.

The intent in creating this flow is to create a current loop without substantial loss / either to the overflow, floor or in wasted wave energy. I wanted it to be as directionally uniform as possible. The target flow rate everywhere is to hit a peak of 4in/s and then alternate directions.

My fear is that the positive and negative water waves will take too long to equilibrate and a lot of water energy will be wasted or lost as surface waves instead of a circulating current..

I can adjusting timing/flow.., or make changes to he model to fix this.

09/05/2017, 06:53 AM
There's also a second surge under the reef that creates a flow up from the far reef edge.
With 4 channels of flow, there are 16 permutation of flow dynamics.

I will also have powerheads for backup to create loop flow against the back end.

If this was a straight 8 or 10 ft tank, I'd just build it. But this is effectively a 28' x 4' tank with a U channel. Water energy moves too slowly in that scale and I can't predict it's behavior without getting much more capable at fluid dynamics -or- running a finite element simulation -or- building a scale model and trying things out.

I tried the first two options with little success. Before spending a lot of $$, I'm going to experiment.

09/05/2017, 06:55 AM
Here's my build thread (one of them).


The design has changed a lot since page 1 but it's getting closer.

09/05/2017, 06:55 AM
<a href="http://s1062.photobucket.com/user/karimwassef/media/Designs/5_zpsdcvkcxti.jpg.html" target="_blank"><img src="http://i1062.photobucket.com/albums/t496/karimwassef/Designs/5_zpsdcvkcxti.jpg" border="0" alt=" photo 5_zpsdcvkcxti.jpg"></a>

09/05/2017, 06:56 AM
<a href="http://s1062.photobucket.com/user/karimwassef/media/Designs/1_zpsybteiv92.jpg.html" target="_blank"><img src="http://i1062.photobucket.com/albums/t496/karimwassef/Designs/1_zpsybteiv92.jpg" border="0" alt=" photo 1_zpsybteiv92.jpg"></a>

09/05/2017, 01:02 PM
Corn starch + Wolfram Alpha for the calculations...

Sent from my iPhone using Tapatalk

09/05/2017, 01:04 PM
Did I mention free or cheap simulation software?

Can wolfram do 3D fluid dynamics?

09/05/2017, 01:08 PM
I was worried that corn starch varied too much over temp. I don't expect massive swings but still.

I guess part of my question in this thread was whether a small additive can effectively change the viscosity of water by 3x?

09/09/2017, 09:08 PM
I could not get an accurate measure of water viscosity using the method above. Looks like this range of viscosities needs to be measured using capillary action.

I found a viscometer, but it's $141... for a single use, it's a little pricey for me.



looks like I'll have to call some local universities and see if I can use their equipment

09/14/2017, 06:59 AM
I'm going to experiment with a DIY capillary measure of viscosity ... will update once I get some pics and data.

If you want to see the full scale plan, I added a video to my main thread http://www.reefcentral.com/forums/showthread.php?t=2589632&page=25

09/14/2017, 08:39 PM
Glycerol is used to change viscosity of aqueous solutions. we use it in the lab when a series of different viscosities is needed for experiments. It is widely used for this purpose.

I did a quick google search and found this paper.


09/14/2017, 08:54 PM
Great. Thanks. That's perfect.

It points me to about 35%-40% aqueous glycerol by weight.

Unfortunately, it shows me the scale of the problem. I need 60 gallons of water ~ 500lbs so I would need 130lb of glycerol!!

I think that's ~$1000 if I buy it retail. I need a wholesale source of glycerol.

09/14/2017, 08:57 PM
I found this.. I would need about 15 bottles


so that's ~$400.. better, but that costs as much as my entire scale model.

09/14/2017, 08:59 PM

This is ~ $150... so getting closer.

09/14/2017, 09:12 PM
That bulk apothecary site had 50lbs for $53 in bulk (basically a 5 gal pail). Just google glycerol for sale and it popped up.

09/14/2017, 09:15 PM
Sorry I just saw you needed 130 lbs... yeah 3 pails... $150.

09/14/2017, 09:28 PM
I can't find anything cheaper... unless there's a cheaper chemical thickener out there.

09/14/2017, 10:40 PM
That's fairly pricing for modeling. I don't know that I'd invest that much cash in glycerin.

09/14/2017, 10:42 PM
I know... that's why I'm looking for a cheaper thickening chemical.

The full scale model would be very expensive or impossible to fix, so I need a scale model or a way to run a numerical model.

<iframe width="560" height="315" src="https://www.youtube.com/embed/BIjj3Qcmbf4" frameborder="0" allowfullscreen></iframe>

09/14/2017, 11:00 PM
cornstarch? $60 assuming the wt% is comparable to glycerin.


09/14/2017, 11:32 PM
or make my own glycerol...

<iframe width="560" height="315" src="https://www.youtube.com/embed/HaWe0Af_0hg" frameborder="0" allowfullscreen></iframe>


09/23/2017, 05:04 PM
Ooze baff?

Slime baff


09/24/2017, 12:10 AM
It might be a bit foamy, but at least it'd be cheap to try.

09/30/2017, 04:27 PM
To measure the viscosity of paint the sell cups that hold about 2 cups and you just time how long it takes to run out a hole in the bottom. For your use I would just use a 5 gallon bucket with a 3/16-1/4" hole in a bottom corner. Set it at a slight angle with the hole at the lowest point. It should take 3 times as long for your test fluid to drain as plain water. Of course you need to fill the bucket to the exact same level each time.

09/30/2017, 05:40 PM
Yes. I saw the cups. But they each have a range of viscosities that they measure in.

None were at water viscosity or 3x.

10/01/2017, 01:56 AM
Yes. I saw the cups. But they each have a range of viscosities that they measure in.

None were at water viscosity or 3x.

That is why I sugested making one your self out of a gallon bucket.

10/01/2017, 05:40 AM
Should that be 1:4 not 1:3?

What is the purpose of the model? Are you wanting to visualise mixing by injecting dye so that you can change the geometry if there are dead spots or something?

Vinny Kreyling
10/01/2017, 07:57 AM
Hi Karim,
Would something like this help?, time the water & adjust your liquid mix appropriately.

10/01/2017, 10:12 AM
Hi Karim,
Would something like this help?, time the water & adjust your liquid mix appropriately.

That's like the cups Laverda is talking about.

10/01/2017, 10:14 AM
That is why I sugested making one your self out of a gallon bucket.

I think it's more complicated than that. I started looking at how those things work. Got something along the same direction.

10/01/2017, 10:18 AM
Should that be 1:4 not 1:3?

What is the purpose of the model? Are you wanting to visualise mixing by injecting dye so that you can change the geometry if there are dead spots or something?

The model is 1:3 because that's about the size I can build using similar materials to the real thing and small enough so I can make changes easily.

Why? I have several complex flows in my system and it's very large. I'm trying to establish a circulating current loop flow with several surges. I'm concerned that it won't work and I want to figure out what will work.

I talk through it earlier in this thread. Posts#14-18

10/01/2017, 10:26 AM
There are plenty of places to purchase glycerol that are much cheaper than $1,000 per 130 lbs. The price of glycerol varies greatly depending on the purity. Spectrophotometric grade is very pricey, but food grade glycerol is comparatively cheap. I spent no effort on this, just one search, and came up with this (http://www.bulkapothecary.com/raw-ingredients/other-ingredients-and-chemicals/glycerin/?gclid=EAIaIQobChMIhLzN0-fP1gIVi2SGCh3TDwPPEAQYASABEgL-k_D_BwE).

A couple of suggestions/comments to reach your goal. While real-world empirical experiments are easy to understand and provide a "hands-on" experience, most engineering firms will at least start with a numerical model. The necessary theoretical underpinnings of finite element analysis models have been around for 50+ years, but the gigantic drop in computational cost in the modern era has made them very practical.

I'm not familiar with Wolfram's system - if it allows a drop-in CAD shape and flow field, that might be a viable solution. But one typically does these sorts of things in ProEngineer. It's a very expensive program, so not really suitable for hobbyist use. However, ProEngineer is almost a de rigueur course in most US university engineering departments, particularly in mechanical engineering departments.

And in many cases, university programs are looking for "real world examples" for their design classes. If you've a local university with a strong engineering college, and can make friends with one of the department professors that teaches either a dedicated ProEngineer class or that teaches a Senior Design class, you might be able to convince them to use your proposed system as a case study. One thing to offer to pique their interest is the scale model - they do the computational model, you build a prototype to test their results and present it to the class.

That might put you under the gun to get the prototype done on a schedule so that the students that do the modeling get the feedback from the scale model testing, but the results might actually be publishable in a magazine like Coral, if that interests you.

By the way, regarding using glycerol for your scale model: in my opinion, it's not worth scaling the viscosity. You can get very good approximations from mathematical transformations of measured velocity fields in a scale model to estimate a full-scale system. This is one reason why hydrology/fluid dynamic laboratories that do nautical engineering use unadulterated water to do testing of scale-model ships. But if you choose to use the glycerol, recognize that glycerol will support bacterial growth. So you will want to add a bacterial growth inhibitor to your solutions if you want to keep it around for a few months to do repeated experimentation.

10/01/2017, 12:02 PM
Thanks! The bulkapocatherapy site is the one I ended up with also.
I'm very familiar with numerical modeling and FEA but I don't have access to the tools. Interesting idea to reach out to the local universities to see if I can get them interested.

The scale of the system is such that the viscosity will play a significant role. Specifically, if my concerns are true, the positive and negative waves will cause water to overflow on one side and expose the corals on the other line with every cycle. The viscosity will change the wave shape, height and the propagation speed.

To put it in context, I'm looking at a wave traveling 32 feet in the real system modeled with 11 feet in the model.