Fluorinert eats plastic
 

I saw an article on fluorinert @ houseofhelp and OCforum; posted my 2p worth there so I thought I'd share it with you cos you've all been very helpful with my cooling efforts.
Would that I had come across the posts at houseofhelp.com a year ago :( I have been using fluorinert for about a year now and I have just realised it does indeed react with certain plastics. I had it in both my water cooled systems one with a swiftech micro reservoir and and the other with a coolplex-25; both with O-CuK 1/2" ID Steel Reinforced PVC tubing. The first coolplex developed a crack and I put it down to general manhandling and just bought another. The second one then developed increasing cracks throughout the tube ( looked like scratches ) but did not leak until yesterday.. I will post pics tomorrow. Between these two events I came home one evening to find the swiftech reservoir almost disintegrated the glues lines and around the nozzles were all brittle like superglue when hardened. I was more concerned about the loss of the expensive FC-77 that had cost me £150 for 5 litres. One pump an Aquaextreme 50Z works well although choking intermittently after being down for a few days. I will take it apart and look for erosion. the Laing D5 seems to be okay.
The chemical cools just like water and does nothing when inadvertently spilt on electrical components-definitely non conductive. I haven't noticed any irritation when in contact with the skin and I tried my very very best not to put in my eye or inhale it ;P It does evaporate quicker than my monthly pay cheque lol. I am now considering fluid-XP; it's where I got the idea in the 1st place but wasn't readily available in the UK at the time.
hope this helps someone

tubing-http://store.over-clock.com/Tubing.html

Re: Fluorinert eats plastic
 

Hmm... and to think I was about to submerge a motherboard and it's overclocked components in an Acrylic Plexiglass tank of quickly circulating 3M Fluorinert. :(

Re: Fluorinert eats plastic
 

Looking forward to pics, thank you!

Use an aquarium (glass, silicone seals).

Re: Fluorinert eats plastic
 

I have been running Fluorinert (PF-5080) for at least two years now, and I can't say that I have ever noticed any issues with plastic (in my loop that is the pump and the flow sensor). Tubing is Tygon --again no issues here.

Fluorinert does expand much more than water does when it gets warm, hence you want to leave a bit of headroom in the tank.

I note that you use steel reinforced PVC tubing, which has very little give in it in terms of expansion. Depending on how full you fill your reservoirs, you may find that inadequate expansion room will, through repeated cycles of heat expansion/contraction by the Fluorinert, put stress on the reservoir (particularly at the glue joints) in much the same way that metal fatigue does its work. Many people have reported some reliability problems with glued reservoirs even with conventional water-based cooling liquids.

Re: Fluorinert eats plastic
 

more than likily that some solids are soluable in flourinert due chemical stuff.

FC-77 has a lot less cooling performance the water. Water as a base cooling fluid is about the best you can get.

Re: Fluorinert eats plastic
 

Not sure about that. Fluorinert's big application point is that it is inert. That means that it does not chemically react with anything, at all, ever.

There are Perfluorcarbons out there that are not so friendly, and will eat plastics faster than a depressed bulimic hits the chocolate cake.

FC-77 and its brethren has marginally less cooling performance than water. I have used both in the same loop, under the same conditions, and can confirm other people's findings that there is no more than a degree or so in it.

Re: Fluorinert eats plastic
 

I thought chemically inert and soluble were different. Either way the effect on the glue appears slow. I notice from wikipedia that Flourinert is highly soluble for oxygen for example so it could still be possible.

(if memory serves and memory is always unreliable)
The Prandlt number (http://en.wikipedia.org/wiki/Prandtl_number)of FC-77 is half that of water therefore heat transfer performance of the fluid (as opposed to the system) is significantly lower as it is harder for the fluid to transfer heat to it self by convection, so you have to rely more on conductive heat transfer (which is worse but might be the dominant form in something like confined jet impingement; storm style blocks). Although since the response is non linear its effect on overall performance is small (think is proportionally to Prandlt^1/3 or 1/4)

1 degree in water cooling can thermodynamically speaking be pretty huge for changing one component.