Project Leviathan - Fully Immersed PC Build Project Log

[cavemanoc]

Super accurate and dead straight?

Out of interest, how are you planning on configuring this? Are you going to such the fluid out of the tank/case and run it through the rads? How are you going to make sure you get a good through flow and cool all of the fluid? Do you have a mock up of the final design (struggling to visualise it)?

 

[Mcnumpty2323]

^ good questions
Am also curious
Assuming it's not some sort of submerged pump
I would probably go through the panel
Into the other chamber using bulkhead fittings
At the bottom of the chamber
And have pump,res,radiators in the other chamber
So mavity/weight of fluid pushes the coolant through to the pump
In the other chamber
How the heat/all the coolant circulates
May depend whereabouts the main heat is produced
Since hot coolant will rise/displace colder coolant
Or they might just run tubing up and over into the other chamber

 

[TrojanWhore]

Or they might just run tubing up and over into the other chamber

That's how I'd do it - hose from tank to rads just near the top, the other all the way to the bottom. Not sure if there's anything you could do to keep it mixing around rather than having some fluid stuck around the bottom of the tank, I suppose with enough flow that wouldn't be much of an issue.

 

[Mcnumpty2323]

That's how I'd do it - hose from tank to rads just near the top, the other all the way to the bottom. Not sure if there's anything you could do to keep it mixing around rather than having some fluid stuck around the bottom of the tank, I suppose with enough flow that wouldn't be much of an issue.

I guess would depend where the heat sources are
Since rising hot coolant would help "mix" it
So there wasn't one particular bit of coolant didn't circulate
Not sure it even matters if there's a "dead zone"
Where coolant just sits there in the tank
It will still heat up through conduction with the other coolant
And eventually rise

Since its non conductive
You could even have the pump sitting in
the bottom of the coolant i guess
This isn't mineral oil
Pretty sure it's lower viscosity so shouldn't be an issue to
Pump it

It's all interesting
Does make you wonder if the op has
A definitive plan
Or if some of it will be try this see what happens
And adapt from different results with different placements

 

[kindai]

The largest and most complex panel is all ready for machining now.

Hopefully it will be ready next week

 

[Threepwood]

I'm really looking forward to the finished result!

 

[moppy89]

Super accurate and dead straight?

Figured that was the case but wasn't sure if I was missing anything else, thanks

 

[Cenedd]

Figured that was the case but wasn't sure if I was missing anything else, thanks

That and a couple of other things: how clean the cut is - the powder coat hasn't chipped - how thin the kerf is at 0.5mm (maybe slightly less). The wire doesn't have to take any cutting force (like say a hacksaw would) so it doesn't need to be wide enough not to snap. How accurate? 10 to 20 microns should be quite easy, maybe single digit microns.

 

[kindai]

So the front panel prototyping has come back all good, and sent off for machining of the acrylic panel. That should be here in a couple of weeks.

In the meantime, the fan brackets for the radiators simply had too much metal on them. While great for 120mm fans, a huge section was simply blocked up for 140mm ones.

So popped them onto the cutter and cleared off that excess material.

Before:

After:

 

[Gepetto]

That's an eye for detail.

I wouldn't have thought the slots for the 120mm fans would have created much resistance to the airflow, and would have ignored it.

I imagine having the right tools for the job does help. Nicely done.

 

[moppy89]

I feel like I need some red circles here!

 

[Gepetto]

Are you making any progress with this @kindai?

 

[kindai]

Are you making any progress with this @kindai?

Hi!

Yes so the front pannel is all machined tested and ready to go. And the side and back ones are currently with the CNC company.

I am also trying to find some "space" to document the build a bit more nicely now stuff is coming together, which is why I havent put anything up at the moment.

 

[Gepetto]

Hi!

Yes so the front pannel is all machined tested and ready to go. And the side and back ones are currently with the CNC company.

I am also trying to find some "space" to document the build a bit more nicely now stuff is coming together, which is why I havent put anything up at the moment.

Brilliant! Thanks. I'm really interested in what you are doing with this!

 

[Tefal]

^ good questions
Am also curious
Assuming it's not some sort of submerged pump
I would probably go through the panel
Into the other chamber using bulkhead fittings
At the bottom of the chamber
And have pump,res,radiators in the other chamber
So mavity/weight of fluid pushes the coolant through to the pump
In the other chamber
How the heat/all the coolant circulates
May depend whereabouts the main heat is produced
Since hot coolant will rise/displace colder coolant
Or they might just run tubing up and over into the other chamber

Why risk bulkheads just have a long pipe to the bottom as a cold return short pipe at the top as a hot take off both gping up and over like the wires

Edit:I should bother reading the thread fully

 

[kindai]

So its been a while since an update.

How the day started:

All the Acrylic plates are now done and ready:

A quick.... cat to help....

Test fit, spot on....

The first 4 plates in and bonded.

While waiting for the glue to set, mounted up the radiators onto the cut down mounts:

And with that done, re-assemble the metal work.


The back can be bonded in later, but for now using that as my way in to take measurements etc more easily. Also have some polishing to do.

Yes that is blood on the dust sheet... acrylic can be sharp

 

[Mcnumpty2323]

Very Nice been waiting for updates to this
As its something totally different to the usual stuff
The acrylic i assume is permanently bonded in there?
If one was to crack or break
No way to replace individual panels easily?

 

[kindai]

If one was to crack or break
No way to replace individual panels easily?

Correct, that is why ive gone overkill with 15mm thick sheets, in theory could get away with as little as 6mm.

 

[Cenedd]

Yay overkill! T'is the only way
I have to admit, I do picture the finished product with some Neon Tetras swimming about!

 

[Mcnumpty2323]

Yeah 15mm should be way over specced
Shouldn't be any issues with expansion and contraction

Sadly it's dielectric fluid not water (I assume)
Won't be any real neon tetras in there