Water coooling chamber, will turbulence on opposite side help?

[georgerosebush]

Will creating turbulence in a water cooling chamber on the opposite side surface that you're trying to cool help with cooling at all? I mean on the cover of the water cooling chamber, opposite of the heat source. We have an IGBT that we want to water cool, and the chief wants to add turbulators on the cover of the water cooling chamber. There was a study that I read that said that this kind of turbulence only helps if it's on the surface that you're trying to cool, and I tried explaining this but I'm not sure he understood. Will adding these things (bumps basically) on the cover help at with cooling?

 

[georgerosebush]

Am I only able to attach one thing at a time?

 

[georgerosebush]

2

 

[georgerosebush]

Here is a picture of the actual water cover and the turbulators.

 

[EdStainless]

If the current design is laminer flow and the 'bumps' increase mixing, then they will help.
If the flow is already turbulent then I don't see them helping at all.

= = = = = = = = = = = = = = = = = = = =
P.E. Metallurgy, Plymouth Tube

 

[georgerosebush]

Yeah thanks, I don't know if it's laminar or turbulent as I'm not sure of the speed, that's something I gotta find out, and I'm using FloXpress in Solidworks which is only laminar. But you say if it's laminar it would help mixing, and that's the purpose of those diamond shaped ones, those are actually on the side we want to cool, the "bumps" I'm talking about you can see in the third picture are on the cover. So they're on the opposite side, and I don't think they'll mix anything as they're about the width of the channel, so it's like a ramp that the water just goes over. Can't imagine that helping with mixing. The chiefs idea is that it'll create turbulence, but wouldn't that only work on the side we want to cool?

 

[georgerosebush]

Figured out how to embed pictures, couldn't see the button. The CAD pics are for a new revision.

This is the cooling side

Here's the ramps on the cover

The current water channel

 

[EdStainless]

So the pins are on the side that is heated? Good.
The 'bars' may just be getting in the way. You won't know until you run it. You have the old cover that is smooth? you could compare them.
Have you run it yet to see how much it increases the pressure drop?
They look big, smaller ones would work just as well and not restrict the flow.
The narrow inlet and outlet passage does not look good, you need the same cross section everywhere.
After all flow velocity is your first tool, don't reduce the flow.
If they obstruct too much just make them narrower, but the same length, () shaped with the long direction the flow direction.

= = = = = = = = = = = = = = = = = = = =
P.E. Metallurgy, Plymouth Tube

 

[georgerosebush]

When you say to make them narrower but the same length, do you mean the diamond shaped pins or the ramps on the cover? The diamond shape greatly increases the flow velocity and mixing versus the hexagons, I did not like the hexagon idea but they thought it would make the CNC faster by allowing it to go in straighter lines, turns out it just went in circles anyway (around the pins).

And no, there is no smooth cover, maybe we should make two versions for the next revision, we are having two made anyway.

I think increasing the inlet/oulet size would increase the size of the whole unit, and we're trying desperately to make it smaller (it's 33inches long).

 

[EdStainless]

I meant the pins on the hot side. Make them eye or teardrop shaped with the long dimension in the direction of flow.

But now back to where we should have started.
How much heat are you trying to remove?
How much flow do you have?
Let's see if it is even possible with this approach.

= = = = = = = = = = = = = = = = = = = =
P.E. Metallurgy, Plymouth Tube

 

[FredRosse]

"If the current design is laminer flow and the 'bumps' increase mixing, then they will help.
If the flow is already turbulent then I don't see them helping at all."

Don't agree with this, more turbulence always increases heat transfer coefficients (HTC). Transition from laminar to turbulent flow gives significant HTC increase, and the HTC will continue to have increasing HTC with more and more turbulence (Generally Reynolds Number).

"So the pins are on the side that is heated?" I think he is trying to say //The pins are on the side that is being cooled by the fluid flow?// Is this the proper clarification here?