Cascade improvements (?)

Dieter@be · 01-01-2004, 11:09 AM

I had a few ideas which might improve the performance of the Cascade block.

1)Enchanged inlet:

instead of injecting the water in the shape of a circle, I'd do it rectangular (the shape of the core in other words).
I'd use an "Adapter"(upper part of the block adapts the round fitting (from the water inlet) to a rectangular opening (to inject the water on the core).
You can compare it with a fan shroud.
(I think the less steep it is, the better the water gets divided over the jets)

2)Enchanged outlet:

First of all, dual outlets. (to get the water away better, and to do it more balanced)
Second, "preparing the shape of the waterflow to enter the outlet. This means I would change how the bottom area near the outlet.
It 'll be more clear when viewing these pictures: (it are drawings of a block that never made it, but the idea of enchanging the path to the outlet still makes sense to me)

3)Getting the water away better + higher "outlet flow"

I am aware that, when looking in the holes with the jets, the exit velocity should be smaller then the inlet velocity. (I believe this is to have a good pressure, to smash the boundary layer)
That is in fact determined mostly by the cup diameter I think. So I think this could be a nice improvement to make.

1) classic cascade: the water getting out of the cups meets the outer water getting out: chaos.
2) enchanged cascade: the water gets away nicely.

4)Much deeper cups (and longer cups). For the simple reason that when using a deeper cup the 2nd cooling effect is better (the water travelling upwards passes more copper)
Although it would make the block cost more, for just a very small performance increase.

5)Enchanged jets:

1)enchanged outer surface to create turbulence: some "obstacles" on the outer side of the jet. Lets say if you put tiny little rings over the jet.
This would create some turbulence, so the water would hit the copper better (2nd cooling effect)
2)echanged underside of jets to improve waterflow: very little detail.

Enchaning the jets like 1) and 2) would probably be very hard and with a very little performance gain (?), but it came in my mind, so why not put it with the others... ;-)

WAJ_UK · 01-02-2004, 05:14 AM

The cooling area in the cascade is a rectangular shape which I think was designed to be the best comprimise between many different core sizes.

The cascade originally had 2 outlets in development. I suggest you read the development thread for the reasons why it was changed to 1

The height of the jet exit from where it strikes the base is critical in an impingement setup and is different in different situations but usually falls between 4d and 5d and can only really be determined through experimentation so changing the depth of the cups would probably reduce performance.

Dieter@be · 01-02-2004, 08:14 AM

Quote:

Originally Posted by WAJ_UK

The cooling area in the cascade is a rectangular shape which I think was designed to be the best comprimise between many different core sizes.

I edited the first post with another drawing to show what I mean.

Quote:

Originally Posted by WAJ_UK

The height of the jet exit from where it strikes the base is critical in an impingement setup and is different in different situations but usually falls between 4d and 5d and can only really be determined through experimentation so changing the depth of the cups would probably reduce performance.

4d and 5d??
I dont get why it would reduce performance...please explain.

WAJ_UK · 01-02-2004, 11:17 AM

sorry I meant d as the diameter of the jet. 4d is 4*jet diameter and 5d is 5* jet diameter sorry it wasn't clear.

If you increase the distance from the nozzle exit to the impingement surface then there will be increased turbulance caused by the shear between the water jet and the surrounding water. However, the velocity of the jet is inversely proportional to z/d (Where z is the height of the jet exit and d is the jet diameter). This means that the further the nozzle exit is from the base plate the more the water jet is slowed by the interaction with the surrounding water.

The best height is where there is the perfect balance between the jet turbulance and the jet velocity. This is about z/d=5 in most of the experimental data I have come across.

I think the cascade is about z/d=4 probably due to the fact that the water surrounding the impingement jet is not stagnent due to the fact that it is coming back up out of the cup. This will slow the jet down more so it has to exit the nozzle closer to the impingement surface for the optimum heat transfer coefficient.

The graphs of experimental data I have come across show that the heat transfer coefficient drops off quite quickly either side of the optimum z/d value.

Hope this helps

Dieter@be · 01-02-2004, 04:33 PM

Oh, you misunderstood me I think.
I didnt mean to make the cups deeper, and leaving the jets. (which would offcourse lead to bad performance due to the water "interference")
What I meant was to make the cups deeper, but also making the jets larger equally

So the cooling system "down at the baseplate" would stay, I just think that if the cup is deeper, and the jet longer, the water needs to travel a longer dinstance to get back up, so it has more contact with the copper walls of the cup (the "2nd cooling effect of the jet impingement system)

I have thought about pressure loss, but I think that it wouldn't be the case, by just enlargening the distance the water needs to pass (in a straight line), because you can't compress the water. It does have more "friction" with the copper walls, but that's good for getting the heat away

WAJ_UK · 01-03-2004, 04:03 AM

If you increase the jet diameter you can't fit as many jets in the same area compared to when you use smaller diameter jets. I don't know how this will affect performance i guess you will either have to experiment or create a computer model to use in CFD.

Dieter@be · 01-03-2004, 04:42 AM

increasing the diameter? I never said anything about that

WAJ_UK · 01-03-2004, 09:51 AM

sorry I totally misunderstood I've got it now just being dumb. Yeah you could give it a try depends how far up the sides of the cup the heat will get.

Dieter@be · 01-03-2004, 09:57 AM

Quote:

Originally Posted by WAJ_UK

(...) depends how far up the sides of the cup the heat will get.

That's true, was thinking that too

01-01-2004, 11:09 AM	#1
Dieter@be Cooling Neophyte Join Date: Nov 2003 Location: Belgium\Wetteren Posts: 94	Cascade improvements (?) I had a few ideas which might improve the performance of the Cascade block. 1)Enchanged inlet: instead of injecting the water in the shape of a circle, I'd do it rectangular (the shape of the core in other words). I'd use an "Adapter"(upper part of the block adapts the round fitting (from the water inlet) to a rectangular opening (to inject the water on the core). You can compare it with a fan shroud. (I think the less steep it is, the better the water gets divided over the jets) 2)Enchanged outlet: First of all, dual outlets. (to get the water away better, and to do it more balanced) Second, "preparing the shape of the waterflow to enter the outlet. This means I would change how the bottom area near the outlet. It 'll be more clear when viewing these pictures: (it are drawings of a block that never made it, but the idea of enchanging the path to the outlet still makes sense to me) 3)Getting the water away better + higher "outlet flow" I am aware that, when looking in the holes with the jets, the exit velocity should be smaller then the inlet velocity. (I believe this is to have a good pressure, to smash the boundary layer) That is in fact determined mostly by the cup diameter I think. So I think this could be a nice improvement to make. 1) classic cascade: the water getting out of the cups meets the outer water getting out: chaos. 2) enchanged cascade: the water gets away nicely. 4)Much deeper cups (and longer cups). For the simple reason that when using a deeper cup the 2nd cooling effect is better (the water travelling upwards passes more copper) Although it would make the block cost more, for just a very small performance increase. 5)Enchanged jets: 1)enchanged outer surface to create turbulence: some "obstacles" on the outer side of the jet. Lets say if you put tiny little rings over the jet. This would create some turbulence, so the water would hit the copper better (2nd cooling effect) 2)echanged underside of jets to improve waterflow: very little detail. Enchaning the jets like 1) and 2) would probably be very hard and with a very little performance gain (?), but it came in my mind, so why not put it with the others... ;-) Last edited by Dieter@be; 01-02-2004 at 08:34 AM.

01-02-2004, 11:17 AM	#4
WAJ_UK Cooling Savant Join Date: Oct 2003 Location: Sussex Posts: 109	sorry I meant d as the diameter of the jet. 4d is 4jet diameter and 5d is 5 jet diameter sorry it wasn't clear. If you increase the distance from the nozzle exit to the impingement surface then there will be increased turbulance caused by the shear between the water jet and the surrounding water. However, the velocity of the jet is inversely proportional to z/d (Where z is the height of the jet exit and d is the jet diameter). This means that the further the nozzle exit is from the base plate the more the water jet is slowed by the interaction with the surrounding water. The best height is where there is the perfect balance between the jet turbulance and the jet velocity. This is about z/d=5 in most of the experimental data I have come across. I think the cascade is about z/d=4 probably due to the fact that the water surrounding the impingement jet is not stagnent due to the fact that it is coming back up out of the cup. This will slow the jet down more so it has to exit the nozzle closer to the impingement surface for the optimum heat transfer coefficient. The graphs of experimental data I have come across show that the heat transfer coefficient drops off quite quickly either side of the optimum z/d value. Hope this helps Last edited by WAJ_UK; 01-02-2004 at 11:23 AM.

01-02-2004, 05:14 AM	#2
WAJ_UK Cooling Savant Join Date: Oct 2003 Location: Sussex Posts: 109	The cooling area in the cascade is a rectangular shape which I think was designed to be the best comprimise between many different core sizes. The cascade originally had 2 outlets in development. I suggest you read the development thread for the reasons why it was changed to 1 The height of the jet exit from where it strikes the base is critical in an impingement setup and is different in different situations but usually falls between 4d and 5d and can only really be determined through experimentation so changing the depth of the cups would probably reduce performance.

01-02-2004, 04:33 PM	#5
Dieter@be Cooling Neophyte Join Date: Nov 2003 Location: Belgium\Wetteren Posts: 94	Oh, you misunderstood me I think. I didnt mean to make the cups deeper, and leaving the jets. (which would offcourse lead to bad performance due to the water "interference") What I meant was to make the cups deeper, but also making the jets larger equally So the cooling system "down at the baseplate" would stay, I just think that if the cup is deeper, and the jet longer, the water needs to travel a longer dinstance to get back up, so it has more contact with the copper walls of the cup (the "2nd cooling effect of the jet impingement system) I have thought about pressure loss, but I think that it wouldn't be the case, by just enlargening the distance the water needs to pass (in a straight line), because you can't compress the water. It does have more "friction" with the copper walls, but that's good for getting the heat away

01-03-2004, 04:03 AM	#6
WAJ_UK Cooling Savant Join Date: Oct 2003 Location: Sussex Posts: 109	If you increase the jet diameter you can't fit as many jets in the same area compared to when you use smaller diameter jets. I don't know how this will affect performance i guess you will either have to experiment or create a computer model to use in CFD.

01-03-2004, 04:42 AM	#7
Dieter@be Cooling Neophyte Join Date: Nov 2003 Location: Belgium\Wetteren Posts: 94	increasing the diameter? I never said anything about that

01-03-2004, 09:51 AM	#8
WAJ_UK Cooling Savant Join Date: Oct 2003 Location: Sussex Posts: 109	sorry I totally misunderstood I've got it now just being dumb. Yeah you could give it a try depends how far up the sides of the cup the heat will get.

Currently Active Users Viewing This Thread: 1 (0 members and 1 guests)