Radiator Equivilency: 2x120 = 240?

[LyL18]

Please forgive my algebraic title

Generally speaking, is having two 120mm radiators just as effective as having 1 radiator with 2 120mm fans?

Hopefully my question is clear enough.

 

[ashmedai]

Roughly. I believe a single one should be slightly more effective with all other things being equal, but it would be easy to have other factors overshadow this.

 

[LyL18]

This was my assumption. However i somewhat dissuaded myself when thinking about complex loops where radiators (if using two 120mm) are between blocks.

EX: ... Rad> CPU > Rad > GPU >NB ...

My logic on in this brings me two opposing views:

1- Cooling the water all at once with a 240mm will produce bigger temperature variances, ΔT. By this i mean, the 240mm will produce colder water and the sucessive waterblocks will produce the hottest water.

2- Cooling the water intermittenly (like in example loop) will bring freshly cooled water to CPU and GPU, lowering overall temperature variance and possibly the entire loop?

(Again this juist MY logic, not actual results. Please help me prove/disprove this!)

I do understand that since my logic is based on ΔT, it is greatly dependant on the quality of WC parts used. Especially radiator, and liquid composition/mixture.

This sounds like an awesome science experiment IMO!

 

[Xilikon]

Both would be equivalent in terms of heat dissipation but not in terms of restriction. 2 radiators is more restrictive than one

 

[ashmedai]

Both would be equivalent in terms of heat dissipation

This depends a bit on internal radiator design, but normally it goes something like up-down-up-down-up-down etc so that the temperature is fairly consistent across the radiator. For a series arrangement, it would already be cooled somewhat by the time it reaches each successive radiator, and cooling efficiency depends on the del T between the radiator and the external working fluid (the ambient temperature). So a series arrangement could potentially be MUCH less effective - but in the scenario of cooling your computer, it's unlikely to be nearly as pronounced as it could be.

Moral - don't do series radiators unless you have a damn good reason (like you already own them, or they're much cheaper ).

 

[ashmedai]

***

 

[LyL18]

I suppose water flow would be the biggest restricting factor of running two radiators (pun intended)

Especially in my example loop a few posts up. I dont even want to know what that would look like in a case. However a larger pump could be the solution! ... at pointless solution i suppose

This depends a bit on internal radiator design, but normally it goes something like up-down-up-down-up-down etc so that the temperature is fairly consistent across the radiator. For a series arrangement, it would already be cooled somewhat by the time it reaches each successive radiator, and cooling efficiency depends on the del T between the radiator and the external working fluid (the ambient temperature). So a series arrangement could potentially be MUCH less effective - but in the scenario of cooling your computer, it's unlikely to be nearly as pronounced as it could be.

Moral - don't do series radiators unless you have a damn good reason (like you already own them, or they're much cheaper ).

this helps me. i guess it all depends on radiator design. with the popular double pass style, a series of radiators doesnt seem logical (anymore )

 

[ashmedai]

For the goals you appear to have, you'd do a lot better to make two discrete loops. Or one loop with a single large radiator.

 

[ikellensbro]

The added restriction from 2-120's comes from the extra tubing run(s) you need to get both in the loop and the extra two barbs on the second radiator. AKA, nothing worth losing sleep over.

You'll also see pretty much no difference performance wise from 2-120s and a 2x120 of the same model, no matter if you have radiator(s) in between blocks. It doesn't really matter, just do what's easiest for you with the budget and parts you have available. If you're buying new, it is going to be cheaper and easier to route tubing if you can fit a single 2x120 in case, otherwise 2x120s will likely be easier to mod in and perform pretty much exactly the same.

 

[BillParrish]

What you will find in practice is that water is such an efficient conductor of heat there will be almost no difference in temps around the loop no matter if you use 1 big or 2 small rads. Of course there has to be a temp difference between the input and output of a rad or their would be no cooling because that is how rads work. But endless debate and attempts all over the web have shown that with normal water cooling of PCs, it really does not matter, one rad, two rads, rad placement in the loop etc. and the delta of loop temps measured at different points (a deg C or two) are usually within the error range of the probes used to measure them making them meaningless both from a measurement and a performance point of view. 2C, if it is even that much, is not going to gain you much anyway.

Go with what is easiest to install and maintain.

However the above sucks all the fun out of debating so please feel free.

 

[Xilikon]

What you will find in practice is that water is such an efficient conductor of heat there will be almost no difference in temps around the loop no matter if you use 1 big or 2 small rads. Of course there has to be a temp difference between the input and output of a rad or their would be no cooling because that is how rads work. But endless debate and attempts all over the web have shown that with normal water cooling of PCs, it really does not matter, one rad, two rads, rad placement in the loop etc. and the delta of loop temps measured at different points (a deg C or two) are usually within the error range of the probes used to measure them making them meaningless both from a measurement and a performance point of view. 2C, if it is even that much, is not going to gain you much anyway.

Go with what is easiest to install and maintain.

However the above sucks all the fun out of debating so please feel free.

Exactly... Don't bother splitting hairs which is more efficient and just concentrate on the total radiator surface area vs ease of installation.

 

[LyL18]

Great, thanks for the info.

As i have not started watercooling yet, it all seems like a fun physics experiment. Hence my curiosity.

Thanks guys