Heater Cores...Series? Parallel?

Just thinking about adding a second heater core to my loop in preparation for adding a GPU block (unsure which one) later on. My question is this- I see most people who have used two radiator/heater cores place them in series which I would *think* would add a bit to the amount of restriction in the loop. I was considering placing them in parallel using a Y splitter for the flow from the pump and for now rejoining them via another Y splitter before it hits the CPU block. Once I add the GPU block I am considering running a separate branch for my two different water blocks.

.........HC1 >> CPU WB
......./ ---------------\
Pump.........................Res >> Pump
.......\______________/
........HC2 >> GPU WB
(the dashes, periods, and underlines are only there for formatting purposes, they do not represent anything....yes, I am a forum formatting noob)

The idea being to avoid reduction in flow and to allow the pump to push more water through the entire loop by splitting the restrictions into two branches. I will have to be careful of course to get a GPU block that doesnt have a significantly higher or lower flow restriction than my GPU block. The heater cores will both be the '77 b-ville type and externally mounted. I will also probably have the pump directly connected to the reservoir.

Currently I am using a Hydor L35 but will most probably be changing to either an Iwaki or Panworld with a 13.5 to 17 foot head in order to sustain good flow through both loops.

Thoughts?

If you are already going to that much trouble, I would just go ahead and make separate loops. There isnt anything to be gained from combining them. Get that Iwaki and use your old loop for the cpu, then use the L30 on the GPU. It will most assuredly yield better results than a split loop and not be any more cumbersome. Alternatively, if you didnt want to go that far off the deep end, one double HC would handle your cpu and gpu quite nicely if for instance you aquired the Iwaki and just ran a single HC loop. That is what I would ultimately do, but if youre set on dueling HCs then go for it.

Thoughts?

Click to expand...

WB parallel = bad.

for example, in series th flow through your loop may be 5 LPM. putting the WB parallel may increase the flow to say 8 LPM. Assuming both blocks would recieve an equal amount of water they wouyld get 8/2 = 4 LPM each. Thats less than they would get when they're in series and so they wont be as effective.

My reasoning for considering this parallel setup are as follows-

1. Space internally. It would be difficult to fit 2 reservoirs (I prefer these) and 2 pumps into the case. I was looking to 'share' these two components since they are fairly bulky.

2. Performance- In a single loop with a single HC the option is to pass the already heated water from one wb over the other wb. By adding another HC I can basically send 'pre-cooled' water over both blocks.

3. Flow- This is my main question. Assuming I have a pump with enough head to handle 2 HCs, 2 blocks, tubing, etc then instead of running the those same high restriction points (blocks & HCs) in series I would spread them out to a parallel setup. This would allow a greater flow by reducing the total restriction across the loop by limiting the amount of restrictions on each branch.


The reason I was wondering if this would be more effective would be the ability send 'pre-cooled' coolant across both blocks. While flow itself may not be equal to the flow through a single loop the net cooling effect should be greater...at least I hope. My second reason for thinking this *may* be a better setup is that in most systems the restrictions deny a quality high end pump to work at or near its peak performance. Large amounts of pumping ability are not used simply because of the restriction values on the waterblocks. In creating a second loop branch I am hoping to harness more of the higher end pumps potential.

I guess that is my main question regarding the 2 branch option. Just how much flow do blocks like the Cascade, White Water, and RBX actually allow by comparision to how much the higher end pumps can manage?

put your rads in parralles that will give you less restriction.

The reason I was wondering if this would be more effective would be the ability send 'pre-cooled' coolant across both blocks.

Click to expand...

The water temp is only like 0.2C more after a block.

Senater_Cache said:

put your rads in parralles that will give you less restriction.

Click to expand...

Thats basically what I was thinking. My main question actually was....which is more restrictive- the waterblocks or the rads? I was thinking it was the waterblocks but I could be completely wrong on this. If its the waterblocks by putting them in parallel along with the rads wouldnt that improve flow?

Understand guys, I dont know this. Its just guesswork at this point. Main reason I posted wasnt to argue the idea but get some input. I am 99% sure that with some of the higher end pumps like the Iwaki's etc that a fair amount of pump capability is being wasted due to the restrictions in a system 'holding them back'. I am exploring ways to save space and fully harness those pumps...if they are indeed only using a percentage of their standard capability.

OK, in parallel OVERALL SYSTEM flowrate will increase, but each component will see LESS FLOW. Running the blocks in parallel is never (well almost) a good choice. You can run the rads in parallel ( I have done this) and leave the blocks in series. Each block will have more flow in series, but system flowrate will go down, but we don't care about overall system flowrate as much as what the WB's are getting. Since87 made some nice graphs of common WB's a while back with comparisons between two identical blocks in series and in parallel. The more restrictive the block, the more it makes sense to run in series. Somehow I don't think I'm explaining this in an understandable way...

Nah, ya are actually.

I didnt see those graphs unfortunately. I will do some searching. It may just be my belief that by separating the restrictions would actually allow the pump to move more overall fluid over a given time is wrong. I was thinking that instead of the fittings or tubing the actual water blocks were the determiner for how much fluid was going through the system. I was also thinking that if the pumps output could be split across two branches that overall flow rates would increase without negatively effecting the flowrate across each block. This of course would depend on having a fairly high head rate for the pump. By 'high' I meant around 14+ foot head for the pump. Even better would be the 20+ head of the Iwaki Z series.

I have seen people posting that these pumps (the Iwaki Z series) have almost ridiculously high head ratings and still maintain decent flow rates. The type of flow rates when divided in two for each branch would still max out the possible flow rates across the blocks.

Now if the pump cannot maintain a near maximum flow rate across each block in a branch then my idea is kaput.

With a pump like I currently have - the L35 - I have no doubt it would be asking too much of the pump with its average head rating to keep each block at near maximum flow levels. I definitely need to find his post about blocks in series vs parallel. Hopefully he used some different pumps including at least one of the 'high head, high flowrate' pumps

Thanks for the responses btw guys. I still have little experience in WCing and any input is certainly appreciated...even when its to tell me the idea is full of fertilizer

In short, if you run 2 indentical blocks in parallel, you HALVE the pressure to each block which is not a good thing.

Thats the part I am probably mistaken on.

I was thinking that the maximum pressure was limited by the restrictions in a series loop and that by dividing the primary restrictions into two loops you could actually gain more flow while maintaining pressure.

I thought that possibly one of the higher head pumps had more 'umph' available but the restrictions in series loops might be causing a loss of the possible output by the pump. By lowering restrictions the pump would actually be able to put out more flow/pressure. In short it might be able to maintain the same psi for each branch while not losing flow rate (assuming its maximum flowrate at this pressure was greater than the flowrate being achieved in series).

If the pump is already pushing as much coolant with as much pressure as it can then at that point dividing the flow would have worse net results.

Short form: By reducing the restriction in a loop would a quality high head pump be able to maintain pressure and increase flowrate?

if you really want to run the blocks in parrallel you might as well make a complete second loop though i think it would be overkill.


personaly i would not worry about it unless you are trying to squeez every clocck cycle