1/2 vs 5/8 Tubing

lemmyslender · Feb 5, 2003

Just wondering if anyone out there is using 5/8" tubing on their system?

I know that normally for most systems, 1/2" performs better than 3/8" due to the higher flowrate.

I finally have my new block setup, which I designed for based on the area of 1/2" tubing. Link to previous thread with design Before I set it up, I realized that 1/2" barbs have less area and would become the bottleneck, instead of the block. So I used 5/8" barbs, and currently am using 5/8" tubing (since my heatercore in/out is also 5/8").

I'm wondering if using 1/2" tubing would be better?

I'm still tweaking a little, since I'm trying to get the whole setup in my case.

Thanks in advance.

bigben2k · Feb 5, 2003

It isn't that simple.

The short answer is; it depends on the flow rate you achieve.

3/8 tubing is more restrictive than 1/2, but with a low flow rate, using 1/2 tubing won't make much of a difference.

I recently switched to 3/4 tubing, based on a (incorrect) assumption that I would reach 4 gpm. I also used the wrong formulae to calculate pressure drop, but essentially, 3 gpm is the highest flow rate for which you should use 1/2 tubing. 1.7 gpm would be the upper limit for 3/8 tubing.

For reference, keep the flow speed under 5 feet per second, and everything will be fine. If you can't do that with 3/8 tubing, then switch to 1/2.

BTW, your block should be significantly more restrictive than the tubing...(design hint).

roscal · Feb 5, 2003

The bigger the better and less pressure drop in tubing. You'll have a better flow using 5/8 than 1/2 but difference won't be extreme

, depend on lenght of tubing..
But if you can kill some pressure drop do it.. There no speed limit in tubing BB2K (or I misunderstand your sentence). Pressure drop increase at ² with flow rate, it's all you need (with pump curve).

lemmyslender · Feb 5, 2003

Having the block as the most restrictive part of the system is what I was trying to accomplish.

I based this on the area of the channels being slightly less than the area of a 1/2" circle.

The larger tubing won't give you greater flow unless it (smaller tubing) is the most restricitive part of your system.

I guess I am really asking if you can go too large when choosing the tubing size.

Thanks for the replies.

BBigJ · Feb 5, 2003

I'm using 5/8, but just because my rad and original pump had that size fittings (I made my own block.) I figured it would be easiest to keep everything the same size.

mr_kow · Feb 6, 2003

i think its best to use +1/8" size fittings for whatever size tubing you use. for example, use 5/8" barbs for 1/2" tubing. this is because the ID of the will equal the ID of the tubing, so the barbs wont be restrictive.

Wangster · Feb 6, 2003

mr_kow said:
i think its best to use +1/8" size fittings for whatever size tubing you use. for example, use 5/8" barbs for 1/2" tubing. this is because the ID of the will equal the ID of the tubing, so the barbs wont be restrictive.

Huh?

1/2" tubing generally has 3/8" ID

5/8" hose barbs would have between 3/8" - 1/2" ID

My experience researching hose barbs online has led me to believe that most hose barbs that state a size of 3/8", 1/2", 3/4", 1", etc., are in fact one size smaller in ID in comparison to OD.

Meaning 3/8" OD equals 1/4" ID, 1/2" OD equals 3/8" ID, etc.

Does that make sense?

Wangster

mr_kow · Feb 6, 2003

im talking about 1/2" tubing as 1/2" ID tubing.

one size smaller in ID in comparison to OD.

this is exactly what im saying... so you should buy 1/8" size larger barbs for that size ID tubing.
eg. 5/8 barbs for 1/2 (ID) tubing. the actual ID of 5/8 barbs is 1/2

SkiFletch · Feb 6, 2003

ah the wonders of confusing ID with OD

Hoot · Feb 6, 2003

This is an old diagram I posted many moons ago, but it gives you the perspective of brass barb size terminology. The Black numbers are how they are sold. The Red numbers are their typical ID. I am also a fan of using tubing who's ID is equal to my barbs ID. While barb sizes are typically specified by their OD, tubing is typically described by its ID. IE 1/2" barb = 3/8" tube.

In my system at home, I use an Eheim 1250 with 1/2" inlet and outlet barbs to a radiator with 1/2" inlet and outlet barbs to my CPU block with 1/2" inlet and outlet barbs to my NB block with 1/2" inlet and outlet barbs to my bleed/fill tee which, as you might guess is 1/2". All tubing is Silicone 3/8" tubing. Tight fit over barbs and no clamps necessary.

Hoot

lemmyslender · Feb 6, 2003

Yes, that is precisely why I was using 5/8" barbs (1/2" ID) thus keeping my block as the most restrictive part of the system.

However, I am debating switching to 1/2" ID tubing instead of the current 5/8" ID tubing.

My thoughts: The amount of head (pressure drop) difference between the 5/8" and 1/2" tubing over the length of tubing in my system (~18" ?) shouldn't be too significant. Therefore, I would expect roughly the same mass flowrate with either size tubing. This means that the velocity in the 1/2" tubing will be greater than in the 5/8" tubing (by around 50%, I think).

Therefore, a single "chunk" of water should be able to complete the circuit quicker, thus leading to better (slightly) heat transfer in both radiator and block (due to slightly better temperature differentials)

bb2k - just wondering how/where you derived your numbers?

I'm guessing that you would want laminar flow in the tubing (less head) and turbulent flow in the block (better heat transfer). So I'm wondering how your numbers relate to that.

Thanks

bigben2k · Feb 6, 2003

Again, it's not so simple.

First of all, you can't compare the size of the opening inside the block, and equate that to the tubing size: you first have to convert your block's openings into their hydraulic equivalent (i.e. what size tube behaves the same way).

Right now, if your channel cross sections have the same area as the tubing you're proposing to use (originally), then your block is already more restrictive than the tubing.

Here's a hint I picked up from myv65: when looking at very narrow channels, there is a point where the restriction due to the width is so small, that the height is almost irrelevant.

You're correct about laminar versus turbulent, except for one thing: you won't be able to achieve turbulent flow (Reynolds 4'000) with most pumps. Jet inpingement is a "work-around" solution, but it can't be applied in your design.

I originally used the Hazen Williams formulae, but using Darcy would be correct.

As for calculating the hydraulic equivalent, Google. If you can't find it, let me know.

Wangster · Feb 6, 2003

I think, depending on your pump inlet/outlet IDs, you might have greater velocity of flow through the tubing, but actually have less flow of water being circulated.

Think about drinking through a coffee straw compared to a frappuccino straw.

You can feel the high pressure of the liquid coming out when you drink from the coffee straw but your actual amount of liquid you are drinking is very small. On the other hand, when you drink from the frappuccino straw the pressure of the liquid coming out is less in comparison, but the actual amount of liquid is much greater!

Something to think about maybe.

lemmyslender · Feb 6, 2003

Crap, forgot (didn't think

)about hydraulic equivalency. I've got my copy of Perry's here and a Fluid Dynamics book at home. Should've thought about that. Thanks bb2k, I'll have to take that into account before I refine the design for the next block.

Wangster- I am assuming that the maximum flowrate (mass) that my pump can acheive is not limited by size of the tubing. As bb2k mentioned earlier, just because I'm using bigger tubing, doesn't necessarily mean that I can get more flow. I am assuming that my flow rate (mass) is limited by the pumps ability to overcome the resistance of the block/radiator, and that the rate is achievable in 1/2" tubing.

Wangster · Feb 6, 2003

I see.

What I was trying to point out to readers was that using tubing with smaller ID then your pump's ID would automatically cause the system to limit the flowrate to the maximum rate achievable through the hose (w/o bursting the hose).

lemmyslender · Feb 6, 2003

Fair enough

lemmyslender · Feb 6, 2003

Well, I determined the hydraulic of my block versus the hydraulic radius of a 1/2" ID tubing.

R(hydraulic) = cross-sectional area / length of the wetted perimeter

Or

For the tubing: R(h) = 0.25D = 0.25*0.50 = 0.125

For my block: R(h) = (a*b)/((2*(a+b)) where a and b are the width and depth of the channels. In my case the channels are 0.125 wide and 0.25 deep, so R(h) = 0.0417. Interestingly, if I use the first equation and include all 6 channels, R(h) stays the same at 0.0417

Wish I would have found out about this sooner. I wanted a block that was only slightly more restrictive than the tubing (low resistance = higher flow). Instead, it appears that I made a block that is significantly more restrictive than the tubing.

O well, off to the drawing board

Wangster · Feb 6, 2003

Hey you learned something at least, right!

bigben2k · Feb 6, 2003

lemmyslender said:
Well, I determined the hydraulic of my block versus the hydraulic radius of a 1/2" ID tubing.

R(hydraulic) = cross-sectional area / length of the wetted perimeter

Or

For the tubing: R(h) = 0.25D = 0.25*0.50 = 0.125

For my block: R(h) = (a*b)/((2*(a+b)) where a and b are the width and depth of the channels. In my case the channels are 0.125 wide and 0.25 deep, so R(h) = 0.0417. Interestingly, if I use the first equation and include all 6 channels, R(h) stays the same at 0.0417

Wish I would have found out about this sooner. I wanted a block that was only slightly more restrictive than the tubing (low resistance = higher flow). Instead, it appears that I made a block that is significantly more restrictive than the tubing.

O well, off to the drawing board

But that's correct: the tubing shouldn't be a restriction, within a loop! You DO want most of the restriction to be in the block, because that's where the pump's effort is being applied, making the water flow in turbulence and at high speed, which in turn improves the cooling ability!

ON the other hand, if you're trying toachieve ACTUAL turbulent flow (i.e. Reynolds 4'000+), then you're in for a surprise...

lemmyslender · Feb 6, 2003

Yes, I did learn something!

I realize that I would need a very different system for turbulent flow (though it would be nice from an ideal standpoint).

After some more thinking, using 0.0417 for each channel and converting that back to a "normal" diameter (*4 = 0.167) and figuring the area (0.022in^2) for each of the 6 channels, I figure I have an effective area of 0.132in^2 for my channels (versus the actual 0.1875in^2). When that is compared to the area of a 1/2" tube (0.196in^2), it's not as bad as I thought.

Still some room for improvement though. I think that I may try switching to 1/2" tubing after running the 5/8" for a while so I can make a realistic comparison.

Thanks for the input.

1/2 vs 5/8 Tubing

Member

Member

Registered

Member

Member

Registered

Official O/C Santa Ferret

Registered

Member

Inactive Moderator

Member

Member

Official O/C Santa Ferret

Member

Official O/C Santa Ferret

Member

Member

Official O/C Santa Ferret

Member

Member

Similar threads