Large volume laminar flow machine

[karimwassef]

The bottom plate is 1/8" thick. Each fin is 2 7/8" high (total is 3") and 1/16" thick.

The fins won't be carrying any weight. There are 1/2" pillars in the corners that will be doing that. The fins just have to stand up to the incoming flow and channel it in the direction they're pointing.

The construction is like this:

1/8" sheet of acrylic
2 7/8" ABS fins (3D)
1/8" ABS base (3D)
1/8" sheet of acrylic

And between each sheet of acrylic and the next, there are 1/2" thick acrylic sections.

 

[karimwassef]

<a href="http://s1062.photobucket.com/user/karimwassef/media/0_zps9z4mrm37.png.html" target="_blank"><img src="http://i1062.photobucket.com/albums/t496/karimwassef/0_zps9z4mrm37.png" border="0" alt=" photo 0_zps9z4mrm37.png"/></a>

 

[thejuggernaut]

Ok, it isn't as tall as it looked in the drawing. I wouldn't think that would be to difficult to produce. Probably very fragile until it is all assembled though. I'm very interested to see the final result! Let me know if you want me to reach out for some quotes!

 

[karimwassef]

The amount of printing capacity out there is much larger than I thought.

I converted to an STL file and asked for a quote of 20pcs of the center section (that's 2 per section x 5 section x 2 sides).

The price variability is insane. The best I could find was $1800 for middle resolution (3 edges to make a 1/32") to $2300 for high resolution (12 edges to make a 1/32").

So on average it's $110 a piece. Still expensive compared to what I was thinking ($10 a piece).

 

[karimwassef]

Material also counts - PLA seems more rigid and cheaper than ABS, but melts at a lower temperature. no idea of the implications in a reef (anyone know?)

 

[karimwassef]

I loaded all the parts (total 40pcs) and the best quote I got in ABS is $1,900. The price can vary up to $2,400 (for the full model).

This is a decent quote site www.3dhubs.com

That's not terrible, but I wasn't expecting the baffle to cost more than the pumps ($800)...

So it's feasible, but expensive.

 

[OllieNZ]

The wax and epoxy can work but heat can be a VERY BIG ISSUE!

Just to add to this when I did my composites training we did a little test by pouring various resins into a 4x4x1" tray and used a thermocouple to measure how hot they got during cure and the epoxy hit 450 deg C and the tray was sat on a large mahine vice as a heat sink. You can get specific casting resins that will cure more slowly but they still get warm. We used to use latex for complex shapes but I'm not sure that would be dimensionally stable enough for what you're trying to achieve. A negative mould milled out of aluminium perhaps?

 

[karimwassef]

yikes. milled aluminum... sounds even more expensive.

The worst case cost to have it all made in 3D is ~$2000 so the solution to get around that needs to be less.

Here's how I think I can use a couple of these boxes in my current tank - retrofit

<a href="http://s1062.photobucket.com/user/karimwassef/media/0_zpskxndxktm.png.html" target="_blank"><img src="http://i1062.photobucket.com/albums/t496/karimwassef/0_zpskxndxktm.png" border="0" alt=" photo 0_zpskxndxktm.png"/></a>

 

[OllieNZ]

yikes. milled aluminum... sounds even more expensive.

Possibly not you've done the cad so that would be put into a programme for a cnc mill that would add the cutting paths. Set up the mill and press the go button. You'd be able to pull many components off a single mould and you can easily strengthen the resin by adding aerosil/micro fibre blend or milled fibers.

 

[theatrus]

A "soft mold" (Al) can be several grand, and depending on your tolerances could last for hundreds of shots. Worth finding a plastics injection tool person.

 

[karimwassef]

I'll need 5 gal of epoxy resin ~ $300 + 5 milled aluminum molds and release.

I have to modify my current CAD because it doesn't take curing shrink into account. I don't know how to do that yet. I'm assuming I just need to scale the mould up x%.

So how much would 5 aluminum moulds be?

 

[karimwassef]

I think five injection moulding moulds would cost more than $2000.

The material is probably around the $300 range of the epoxy.

Looks like 3D is still most viable unless I turn this into a business.

The biggest difficulty in using these uniform-flow-spreaders is their size.

Even in my current 380g DT, it's a lot of space to give up, just for a single 64 slot box retro fitted in.

 

[thejuggernaut]

The smaller ones could be great for closed loops! Spread flow. Although I doubt they would be for everyone, but if you were installing a large custom tank, they could be very nice!

 

[herring_fish]

Plastic injection molds are way way expensive (add another zero at least) but the unit cost, aside from that, would be low.

 

[herring_fish]

Going back to 3-D printing. "¦.PERHAPS you could look at building a printer of suitable size and output quality for less. Then you could graduate to the inevitable versions B, C and so on. With all the advice in the world, you never hit it out of the park the first time.

If you did make it a business, the printer might hold you long enough to financially justify going to the next level.

 

[karimwassef]

Going back to 3-D printing. "¦.PERHAPS you could look at building a printer of suitable size and output quality for less. Then you could graduate to the inevitable versions B, C and so on. With all the advice in the world, you never hit it out of the park the first time.

If you did make it a business, the printer might hold you long enough to financially justify going to the next level.

I'm not making it into a business. Just saying that I would need to to justify making this DIY

I have a business to run already. This is just for pleasure.

We ran out of Lego blocks... It's amazing how many pieces go into a model of this size, or just a piece of it. I've spent $40 on Legos and couldn't get half a shape out of it. If I were to make it all out of Legos, it would cost $1600...

3D or aluminum/epoxy look to be the best options other than just contracting out the build with acrylic to the local plastics shop.

 

[herring_fish]

I'm not making it into a business. Just saying that I would need to to justify making this DIY

At the price of some 3-D printer kits, it's looking like you could even save money for just one complete flow system for your tank.

We ran out of Lego blocks... It's amazing how many pieces go into a model of this size, or just a piece of it. I've spent $40 on Legos and couldn't get half a shape out of it. If I were to make it all out of Legos, it would cost $1600...

Couldn’t you just make a test run using a sheet of acrylic, and PVC? You could prove a lot from just making one half of one side. …say the front half of the right side …at full scale, only shorter. Then you could use regular water.

You could get a sheet of acrylic split into strips at Lowe’s or HD. They could be anything 2”, 4”, 6” or more, wide. That would be the height of the test fixture. You could cut the PVC into the same length pieces as the strips and again lengthwise, into quarters for the rounded corner arounds and splitters.

Then cut the strips into the lengths that you want, in almost any way that you want. Jig saw, chop saw, karate chop, anything. Build up the “T”’s and glue them in place with silicon and add in the PVC for corners and splitters, again using silicon. Smooth out the edges of the PVC as they approach the acrylic using your finger. This would be a short term test run.

With the PVC added, the acrylic should stand up to the pressures that you are using. The “T”’s will be strong going left, right, forward and back. The curved PCV will be strong in the other directions. If you want, you could squeeze out enough silicon into the gaps, behind the PCV it add some confidence.

Just move a paddle by hand or use a left over pump to generate the flow. You only really need to test the flow in one direction for laminarization.

With that, you could test the resistance of the system, how laminar the water comes out of the fixture, how fast any turbulence decays after leaving the fixture, returning to laminar flow. You could pull off the last straight pieces and cut them in half and reinstall them. Test if laminaration stays high or drops off a lot. Reversing the flow would show you a lot about resistance in the pull phase.

 

[karimwassef]

Thanks. I'll finish the Lego model of 1/20th of the flow. (1/4 of a section).

Then I'll test it in water.

 

[JamesHolt]

This is made up of several layers correct?? (thats what post 502 looks like anyway)
You are stacking them to get the desired height?

 

[karimwassef]

Yes. The cost is a lot more than I expected so I'm looking at alternatives in design.

For $2500, I can buy a 3D printer myself and all the material I need.

I have a friend who can run CFD analysis to confirm the back pressure and flow calculations and improve the design.

The LEGO concept failed. I used cheap Lego knockoff parts and they don't hold very well. I can try the real Legos but I'm going to wait for the fluid flow analysis instead.