Distribution plates/reservoirs

LuckyBenski

LuckyBenski

LuckyBenski

Soldato
Joined
28 Dec 2017
Posts
8,922
Location
Beds
Right, as a chronic tinkerer and DIYer I have a new conundrum and I need YOUR help to scratch the itch.

I really think I need to develop and make a custom distribution plate. I love the idea of putting my fittings exactly where they're needed. It opens up options like fill and drain ports, parallel routes (if wanted), and can just tidy up the entire case.

I have access to a laser cutter. I'm familiar with working with acrylic, mechanical assembly etc. Here are my expected challenges:

- I need to make sure it's watertight. This is easy if I glue the parts together, but will be visible if using clear acrylic. I'd love to find a good adhesive process that can look OK. If not...

- I'll need to make rubber gasket seals. This will involve cutting a channel in the acrylic and using gasket strip. I'll experiment with this in the near future to see if a decent channel can be made using the tools I have.

- It needs to be useful! No point making a plain old reservoir unless it looks awesome, so I want to design something that serves a need in a particular case or project.

I'm open to input from everyone as to what I need to consider and how to go about it. If you have a build coming up and you'd like to collaborate, and test my prototypes, even better. I won't be happy until I've given this a go and I intend to learn a lot through the process. I'm ordering clear perspex in 6mm and 10mm thicknesses, let me know if you think I ought to order black or some other solid/transparent colours!

So what do you think? Am I barking up the wrong tree, or are you excited to see the results?

N.B. Mods, if this would be better suited to the Project Logs forum please go ahead and move it :)
 
LuckyBenski said:
Right, as a chronic tinkerer and DIYer I have a new conundrum and I need YOUR help to scratch the itch.

I really think I need to develop and make a custom distribution plate. I love the idea of putting my fittings exactly where they're needed. It opens up options like fill and drain ports, parallel routes (if wanted), and can just tidy up the entire case.

I have access to a laser cutter. I'm familiar with working with acrylic, mechanical assembly etc. Here are my expected challenges:

- I need to make sure it's watertight. This is easy if I glue the parts together, but will be visible if using clear acrylic. I'd love to find a good adhesive process that can look OK. If not...

- I'll need to make rubber gasket seals. This will involve cutting a channel in the acrylic and using gasket strip. I'll experiment with this in the near future to see if a decent channel can be made using the tools I have.

- It needs to be useful! No point making a plain old reservoir unless it looks awesome, so I want to design something that serves a need in a particular case or project.

I'm open to input from everyone as to what I need to consider and how to go about it. If you have a build coming up and you'd like to collaborate, and test my prototypes, even better. I won't be happy until I've given this a go and I intend to learn a lot through the process. I'm ordering clear perspex in 6mm and 10mm thicknesses, let me know if you think I ought to order black or some other solid/transparent colours!

So what do you think? Am I barking up the wrong tree, or are you excited to see the results?

N.B. Mods, if this would be better suited to the Project Logs forum please go ahead and move it :)
Click to expand...

Check out DerBauer’s YouTube channel, he’s recently posted some videos involving making custom blocks. Also Bykski make a lot of custom water boards already for many cases.

Also the other option is to use a plate to pass through the back connectors to another part of the case, check out GGF Events on YouTube as they do some great work.
 
I was going to suggest looking at Der8auer's youtube as well. He had the same issue with glue being visible on the distro plate, hopefully you'll be able to pick something up from what he's done and perhaps avoid his mistakes. :confused:

There's a thread in the project logs section about a thermaltake competition build which also has a decent amount of distro plate related photos and trial and error.

Channels and gasket seems to be the way to go if you have the tools and patience for it.
 
I have made my own reservoirs with acrylic back in the day. Machining is not that hard once you get feeds right. But you need the joints accurate, as they're not glued but solvent welded. Trick is assemble two parts by jig. Arrange it so gravity helps feed capillary action from a small squeeze bottle with needle applicator. The solvent is thinner than water and wicks into the joint. There's no bead like glue as there's no filling action. Drill and tap is fine, takes a thread well. Just regrind the drill tip angle to stop it cracking, way down from 120 degree jobber. Like 70 , 0 rake, 15 degree back clearance. Slot drills or end mills for gasket rebates. Oh and don't ever wipe it with IPA, disaster time.
 
Last edited:
Thanks fornowagain :)

I've used IPA on acrylic loads, always seemed fine. Acetone on the other hand is evil and for emergencies only. That said, I don't recall using IPA on clear acrylic yet...
 
I’m gonna get mine CNC machined so if you buy one it’s good for both of us ;)

Not sure how easy a distro will be to make manually particularly small channels for gaskets. Tolerances are not huge but I guess you will have to rig something up to keep the depth and width really consistent or you can assume it will not seal well..
 
LuckyBenski said:
Thanks fornowagain :)

I've used IPA on acrylic loads, always seemed fine. Acetone on the other hand is evil and for emergencies only. That said, I don't recall using IPA on clear acrylic yet...
Click to expand...
I used to think that, found out the hard way. Like you, avoided Acetone, with it melting thermoplastics an all. But depending on the polymer used, IPA can react spectacularly. Here's an old image from one of my projects. Swiftech MCRES-Micro reservoir vs Cast Acrylic sheet after cleaning with IPA. Reacted overnight.

le60hguh.jpg

Just looking over a radbox project, 10 years ago.
Even with larger joints you can get a clean seal.

DlBbUJCh.jpg
 
Last edited:
Alex Banks (MakiRole) recently glued a distro plate rather than screwing but I never saw the end result. I think he just used some mental adhesive because his top half wasn't transparent so he could get away with it. It might be on bit-tech somewhere, or I could ping him on FaceAche and ask for an update.

I'm looking at doing something similar in making a DDC pump top from layers and then laminating them together with Tensol 12. I can get away with it (in theory) because I have very little surface area to cover so Tensol 12 will do the trick, even though it's intended to weld ends, not faces. Notes from my own research here.

You're at odds with your tools and materials. EXTRUDED acrylic works better on laser cutters (I've had a few problems with cutting cast recently) but only CAST acrylic can be welded without visible joins. So if you're really thinking about gluing the distro together then you have to use cast really.

If you're gluing then why do you need gaskets? The gaskets are there to seal between layers as you screw it together, but if you're intending to weld the layers together then you'll get the seal that way. But you'll need something beefy like Tensol 70 because you're laminating sheet faces together, not sheet ends, and that's a different thing.

But I wouldn't weld anyway. Cast acrylic has wild manufacturing tolerances and given the way it's produced you can see variations in thickness across a sheet. As a result, you may not have perfectly flat faces to mate together. Hence the gaskets and screws.

If you do end up with o-ring gaskets, then you really need round channels for the cord to sit in, rather than flat channels you'll get by engraving with the laser. Also, unless you're a super human expert, it's incredibly difficult to engrave a channel with a laser to a precise and repeatable depth, and to a standard that doesn't require a boat load of sanding down to make perfectly smooth for the gasket cord to sit in.

tl;dr I don't think you can make a distro plate with a laser cutter, these things really are something to mill out of cast acrylic.


On a related topic, you fancy cutting some acrylic for me? :p I don't have the time or money to get to my local maker space and just mess things up again, so I'd rather pay somebody more competent than I to get it right.
 
LePhuronn said:
If you do end up with o-ring gaskets, then you really need round channels for the cord to sit in, rather than flat channels you'll get by engraving with the laser. Also, unless you're a super human expert, it's incredibly difficult to engrave a channel with a laser to a precise and repeatable depth, and to a standard that doesn't require a boat load of sanding down to make perfectly smooth for the gasket cord to sit in.
Click to expand...

Strictly speaking you don’t need the channel to be round as you use a smaller channel to oring, so the ‘shallow’ channel gets filled by the gasket as it’s compressed. Agree it’s unlikely to be done by laser, you could rig a fixed height router for a simple shape but you need to be careful not to kill the acrylic.

I think drawing comparison with Maki is futile because he’s insane ;). It’s a bit like gluing an aquarium really, there is always some visible residue if you look hard enough if gluing a flat to flat plate but if you get uniform coverage of glue it will all look fine.
 
Thanks chaps, tons of good advice there especially about different types of acrylic @LePhuronn. I'm by no means an expert with the laser yet (it's largely for my girlfriend's jewellery business) but happy to help get your stuff cut, I've done a few projects for mates including one fan shroud on here.

Regarding gluing, it would be that OR gaskets rather than both. I have some ideas about roughing out a curved channel using multiple laser passes but it's true, it might be too rough. That might allow slow ingress of water through capillary action, so not good. I'll have to do some experiments! Also trying different glue - solvent cement, a more goopy one like E6000, and proper capillary solvent stuff are all on the menu.

I also remembered a friend/colleague has a small CNC and wants to use my laser, so hopefully we'll do some skill exchange soon!

For now, perspex arrived! Just need to settle on a design which has a purpose - not much point just going "tank with inlet and outlet". Unless getting the most basic concept working and adding features seems like the safest bet?

Sg65L1Kl.jpg

zeTTum0l.jpg

9XSZWahl.jpg
 
LuckyBenski said:
Regarding gluing, it would be that OR gaskets rather than both.
Click to expand...

From what I understand, when you weld 2 bits of cast acrylic together the fibres are chemically unstitched from the individual pieces and then mesh back together to reform a single piece. So in theory by welding a flat piece atop your carved channels piece, you've sealed any and all gaps around the channels by chemically fusing the pieces together. And therefore no need for gaskets to act as a barrier against microscopic gaps between layers.

Now that's the theory. In practice your pieces may not be totally flat because of thickness variations in cast acrylic, welding faces together ("laminating") is a different kettle of fish to welding edges together, and goodness knows what else.


As for my own stuff, I'm waiting for somebody local to get back to me so I can save on postage costs, but I'll send you a trust shortly so we can talk turkey. Cheers!
 
Right, cut some test pieces to compare gluing methods. Peeled the backing off and I'd gone and used half-frosted Perspex :mad: Cut a few spares out of proper clear for the second stage of testing, in the second picture. Excuse potato photos, my phone died on 26% battery as I sat down at the laser.

l5Y9j43l.jpg
NkyGHt9l.jpg

I tried using plastic weld solvent stuff, didn't seem to work that well I assume because the laser deforms the surface of the plastic slightly. Along the cut tends to be a slight lip or raised edge. So I tried sanding the sheet flat just to see how that improved the joint - much better penetration but still not all the way through. Then I tried cyanoacrylate (super glue) as it's something I'm very familiar with. This worked really nicely as it filled the gap neatly by just having a small continuous bead and not squeezing too hard. The picture shows the three test pieces in this order, left to right. Any other suggestions for glue?

AbZBiAbl.jpg

I also had some fun trying to figure out how to cut through 10mm Perspex - experimented with increasing laser power, and slowing down the feed speed. Out of 8 test cuts only the one that was both slowest and highest power even reached the back of the plastic. So I added some moar powah and halved the speed and cut some test squares! Definitely a little more learning to be done here.

My friend has offered to lend me his simple CNC - I will try and make a few files and if his cuts don't work out well, I'll take him up on the offer!
 
Ah yes, I forgot to mention that!

I've found I need to overpower the laser to get perfectly straight edges. Because the laser cuts by evaporating the acrylic, you almost need to "shoot" the laser too far to ensure you get even evaporation down the full thickness of the material, otherwise you get that lip effect. Much sanding is needed to flatten it out.

Also, laser cut edges are almost like a flame polishing so in order for any adhesive to work you need to give the edges a good buffing up again with sandpaper. Nothing too coarse but it's gotta lose that fused glass edge.

Extruded acrylic is much more consistent to laser cut than cast. I've found it takes more power to cut the same thickness of cast than it does extruded, and extruded exhibits the above problems to a lessened degree, although I still have to overshoot my laser.
 
You must log in or register to reply here.
Back
Top Bottom