Pulsating or Intermittent Flow Waterblock

Proxes · 07-06-2004, 10:42 AM

I happened to run across a few articles on pulsating the jet flow in impingement cooling the other day. From what I gather, by making the flow pulsate or intermittent, it forces the convective cooling boundary layer to re-establish itself, thereby improving performance.

Has anybody ever heard of this or even know if it's been tried? :shrug:

I attached one of the papers for your amusement. Sorry the lady at the library was a retard and scanned every other page upside down.

krazy · 07-06-2004, 12:37 PM

Kind of like one of those ratchet-hammer lawn sprinklers or something? I think I see the theory you're presenting, but I'm not quite clear on the implementation.

Either way, it sounds to me like a lot of work for probably negligible noticable differences in performace. I'm interested to hear what the old timers have to say about this topic.

Proxes · 07-06-2004, 04:23 PM

Yeah, or like one of those pulsating shower heads. I wonder how those things work??? From the article, it sounds like the frequency of pulsation needs to be pretty quick and I'm not sure you could ever get a shower head type of system to pulse that fast. The only other option I could think of would involve a spinning disk with holes or slots in it....or maybe a spinning ball valve. I'm not fond of either of those solutions, since they probably won't be very reliable due to the fast moving parts. That also makes it pretty expensive.

Cathar · 07-06-2004, 05:38 PM

Hmmm - implementation - use a piston pump perhaps?

Groth · 07-07-2004, 12:19 AM

Quote:

Originally Posted by Proxes

I attached one of the papers

Where?

A lobe pump would give a fast pulsed flow.

Proxes · 07-07-2004, 09:55 AM

Quote:

Originally Posted by Groth

Where?

Wooops! LOL, I guess I'm the retard.

I tried to re-upload the paper and realized the file is too big to post, so I sent it to your email. Hope that helps!

Groth · 07-07-2004, 12:36 PM

Got the pdf. Haven't had much of a chance to look at it...in the meantime, if anyone else wants a peek at it, I've hosted it.

Pretty funny with half the pages upside-down.

Huckleberry · 07-07-2004, 02:34 PM

Interesting read. It looks like this approach was yielding good results, 10% improvement or so.

Most interesting is the non-steady-state approach to improving efficiency.

Maybe if I pick my 'puter up and shake it vigorously, my temps will improve.

Proxes · 07-07-2004, 06:54 PM

LOL, yeah, good idea. Just make sure you shake it at the optimal frequency.

Aviator747 · 07-07-2004, 07:16 PM

A cars fuel pump, for fuel injected vehicles, runs like this. But, I think the pressures would be too high & maybe not enough volume. They run on 12v too.

bigben2k · 07-07-2004, 09:38 PM

Also hosted here:
http://www.wbta.us/bigben2k/Effect%2...ulsations2.pdf

Nice work! will have to review it some more.

edit: fixed link typo

Proxes · 07-07-2004, 10:32 PM

Cathar, I just happened to think that piston pump idea you had wouldn't be to bad at all.

Piston pumps can operate very well at high pressure differentials, thus you could decrease the nozzle diameter on an impingement block design and decrease your nozzle spacing a bit to keep heat transfer uniform. That in turn would increase your exit velocity a ton! The boundary layer thickness would drop quite a bit and improve heat transfer immensely.

That doesn't even factor in the transient effects from the pulsating action of the pump, which won't hurt either.

OK, wow, I'm a nerd.... I'll stop now.

bigben2k · 07-08-2004, 08:47 AM

Quote:

Originally Posted by Proxes

...

OK, wow, I'm a nerd.... I'll stop now.

Welcome aboard!

Proxes · 07-10-2004, 07:37 PM

Quote:

Originally Posted by bigben2k

Welcome aboard!

I'm home!!!!

sandman · 07-10-2004, 09:24 PM

Quote:

Originally Posted by Proxes

I'm home!!!!

Could a propellor type thing on the inlet work?

Like, the water would kind of go in gushes alernating around the circle as the water spun the propellor.

Proxes · 07-11-2004, 12:38 AM

Quote:

Originally Posted by sandman

Could a propellor type thing on the inlet work?

Like, the water would kind of go in gushes alernating around the circle as the water spun the propellor.

I had thought of something similiar, but I don't know what the effect would be.

I can think of ways that it would be similar to pulsating the jet and I can think of ways that it is different. I honestly don't have a clue whether that would work or not.

Somebody with more money than me should research that possibility!

Bigmack · 07-11-2004, 07:56 AM

Just a thought,, Inject air bubbles into the water stream using a venturi type asperator

Bigmack · 07-11-2004, 08:00 AM

the bubbles would also provide a wipeing action on the boundry layer if the velocity was high enough.

bigben2k · 07-11-2004, 11:25 AM

All of these "add-ons" may or may not work, depending on the block design. A Cascade would probably not fare well.

The piston pump idea is the best yet.

I'm going to take a guess here, and state that adding any air in the loop would decrease performance, simply based on the resulting decrease in thermal capacity and increase in thermal resistance. I don't believe that the intended effect here is going to compensate for that.

This might be worth of experimentation. A small piston pump should do the trick. The final application would still have to be quiet though. A case for the ole' fuel pump?

Still going over the paper.

Proxes · 07-11-2004, 12:30 PM

I think I'd have to agree with bigben. Adding air to the system would complicate matters. You'd have to include a water trap to get the air back out of the system before it returns to the pump.

Well, on second thought, I guess the reservoir could act like a trap. Maybe there'd be no need for additional equipment other than a pressure relief valve on the reservoir to let the air out as fast as you are injecting it in. Another thing to consider is that by constantly injecting air, you'd open yourself up to contaminating the water.

I still like that piston pump idea way back there. I've been looking for something in the 2 gpm range that would be small, quiet, CHEAP, and run on 110VAC or 12VDC, but I haven't had any luck.

Speaking of luck, I had lightning hit my house a couple of nights ago and fry two of my computers' mobos (possibly CPU's too, but I don't have anymore mobos to troubleshoot with now).

Coincidently, the two computers that fried both had expensive (to me anyway) APC surge protectors on them. The third computer (a POS Dell) that had a cheapo surge protector is fine.

So, looks like this little waterblock project may get put on the back burner until I get those fixed.

PS - I don't think I'd recommend APC surge protectors to anyone.

Groth · 07-11-2004, 01:46 PM

APC does (or used to) have a protection guarantee. Check to see if you can make a claim.

Proxes · 07-11-2004, 03:27 PM

Quote:

Originally Posted by Groth

APC does (or used to) have a protection guarantee. Check to see if you can make a claim.

Yup, I checked my documentation and the model I have covers up to $50,000. I would assume that is for each one I use too. I wonder what two $50,000 computers would look like.

I've already called them and I'm sending the surge protectors in to them, so their engineers can look them over to determine whether they are at fault or not. Hopefully, they'll agree to replace everything I lost, but from my experiences with insurance, it usually doesn't happen like that.

SysCrusher · 07-11-2004, 04:50 PM

Quote:

Originally Posted by Cathar

Hmmm - implementation - use a piston pump perhaps?

Looks like people are catching on to what we discussed eons ago?

An idea. Find away to use a fuel injector so that you have a straight jet instead of a spray pattern. Not sure how it would act with water though. You could regulate the frequency by the voltage applied to it. Hopefully there is enough pressure to make it work.

Aviator747 · 07-11-2004, 05:13 PM

I am sure you could port them easily enough. Just like how grafitti artists port their spray bomb nozzles for different patterns.

Or you could just make your own also.

A few of the fuel pumps that I have seen around are like 2 GPM or so. I still don't know if they would be able to shove the volume though.

The piston pump idea is good. Especially if you have a 2 cylinder model. BUT, there is a very bad draw back too piston pumps. They heat up the water considerably. I used to work for a high pressure washing company. And we pushed 16GPM @ 10,000 PSI. The water was a little warmer than luke warm when it came out. We felt the water when it came out of the bypass tube on our gun of course hehe.

Bundles · 07-11-2004, 05:50 PM

Hmmm, i wonder if a rotary valve like the ones used on older 2 stroke motorbikes could be used for this. Making it/ buying it would be the hard bit though.

07-06-2004, 10:42 AM	#1
Proxes Cooling Neophyte Join Date: Jul 2004 Location: In the middle of nowhere Posts: 10	Pulsating or Intermittent Flow Waterblock I happened to run across a few articles on pulsating the jet flow in impingement cooling the other day. From what I gather, by making the flow pulsate or intermittent, it forces the convective cooling boundary layer to re-establish itself, thereby improving performance. Has anybody ever heard of this or even know if it's been tried? :shrug: I attached one of the papers for your amusement. Sorry the lady at the library was a retard and scanned every other page upside down. __________________ AMD Athlon64 2800+ @ 2.0GHz, 1 GB Corsair XMS PC3700, 2 X 36 GB Raptors 10k rpm RAID 0, ATI 9800 Pro Last edited by Proxes; 07-07-2004 at 09:34 AM.

07-06-2004, 12:37 PM	#2
krazy Cooling Savant Join Date: Mar 2003 Location: St. Paul, Minnesota, USA Posts: 123	Kind of like one of those ratchet-hammer lawn sprinklers or something? I think I see the theory you're presenting, but I'm not quite clear on the implementation. Either way, it sounds to me like a lot of work for probably negligible noticable differences in performace. I'm interested to hear what the old timers have to say about this topic. __________________ Abit NF7-S ... xp1800+ ... Abit GF4ti4200 ... Maxtor 120gb SATA150 Currently gathering parts to water cool everything! (read about the project)

07-06-2004, 04:23 PM	#3
Proxes Cooling Neophyte Join Date: Jul 2004 Location: In the middle of nowhere Posts: 10	Yeah, or like one of those pulsating shower heads. I wonder how those things work??? From the article, it sounds like the frequency of pulsation needs to be pretty quick and I'm not sure you could ever get a shower head type of system to pulse that fast. The only other option I could think of would involve a spinning disk with holes or slots in it....or maybe a spinning ball valve. I'm not fond of either of those solutions, since they probably won't be very reliable due to the fast moving parts. That also makes it pretty expensive. __________________ AMD Athlon64 2800+ @ 2.0GHz, 1 GB Corsair XMS PC3700, 2 X 36 GB Raptors 10k rpm RAID 0, ATI 9800 Pro

07-07-2004, 12:36 PM	#7
Groth Cooling Savant Join Date: Mar 2003 Location: MO Posts: 781	Got the pdf. Haven't had much of a chance to look at it...in the meantime, if anyone else wants a peek at it, I've hosted it. Pretty funny with half the pages upside-down. Last edited by Groth; 07-07-2004 at 11:07 PM. Reason: Lernin to spel

07-07-2004, 06:54 PM	#9
Proxes Cooling Neophyte Join Date: Jul 2004 Location: In the middle of nowhere Posts: 10	LOL, yeah, good idea. Just make sure you shake it at the optimal frequency. __________________ AMD Athlon64 2800+ @ 2.0GHz, 1 GB Corsair XMS PC3700, 2 X 36 GB Raptors 10k rpm RAID 0, ATI 9800 Pro

07-06-2004, 05:38 PM	#4
Cathar Thermophile Join Date: Sep 2002 Location: Melbourne, Australia Posts: 2,538	Hmmm - implementation - use a piston pump perhaps?

07-07-2004, 02:34 PM	#8
Huckleberry Cooling Neophyte Join Date: Dec 2003 Location: Utahr Posts: 22	Interesting read. It looks like this approach was yielding good results, 10% improvement or so. Most interesting is the non-steady-state approach to improving efficiency. Maybe if I pick my 'puter up and shake it vigorously, my temps will improve.

07-07-2004, 07:16 PM	#10
Aviator747 Cooling Neophyte Join Date: Jun 2004 Location: Cornwall,Ontario,Canada Posts: 13	A cars fuel pump, for fuel injected vehicles, runs like this. But, I think the pressures would be too high & maybe not enough volume. They run on 12v too.

07-07-2004, 09:38 PM	#11
bigben2k Responsible for 2% of all the posts here. Join Date: May 2002 Location: Texas, U.S.A. Posts: 8,302	Also hosted here: http://www.wbta.us/bigben2k/Effect%2...ulsations2.pdf Nice work! will have to review it some more. edit: fixed link typo __________________ WBTA (Water Block Testing Alliance) Nordic Hardware WC 101 Last edited by bigben2k; 07-11-2004 at 11:13 AM.

07-07-2004, 10:32 PM	#12
Proxes Cooling Neophyte Join Date: Jul 2004 Location: In the middle of nowhere Posts: 10	Cathar, I just happened to think that piston pump idea you had wouldn't be to bad at all. Piston pumps can operate very well at high pressure differentials, thus you could decrease the nozzle diameter on an impingement block design and decrease your nozzle spacing a bit to keep heat transfer uniform. That in turn would increase your exit velocity a ton! The boundary layer thickness would drop quite a bit and improve heat transfer immensely. That doesn't even factor in the transient effects from the pulsating action of the pump, which won't hurt either. OK, wow, I'm a nerd.... I'll stop now. __________________ AMD Athlon64 2800+ @ 2.0GHz, 1 GB Corsair XMS PC3700, 2 X 36 GB Raptors 10k rpm RAID 0, ATI 9800 Pro

07-11-2004, 07:56 AM	#17
Bigmack Cooling Neophyte Join Date: Jun 2004 Location: Miami Posts: 60	Just a thought,, Inject air bubbles into the water stream using a venturi type asperator __________________ A smart man won't brag about what he knows, but listens to others that may know something he dosen't.

07-11-2004, 08:00 AM	#18
Bigmack Cooling Neophyte Join Date: Jun 2004 Location: Miami Posts: 60	the bubbles would also provide a wipeing action on the boundry layer if the velocity was high enough. __________________ A smart man won't brag about what he knows, but listens to others that may know something he dosen't.

07-11-2004, 11:25 AM	#19
bigben2k Responsible for 2% of all the posts here. Join Date: May 2002 Location: Texas, U.S.A. Posts: 8,302	All of these "add-ons" may or may not work, depending on the block design. A Cascade would probably not fare well. The piston pump idea is the best yet. I'm going to take a guess here, and state that adding any air in the loop would decrease performance, simply based on the resulting decrease in thermal capacity and increase in thermal resistance. I don't believe that the intended effect here is going to compensate for that. This might be worth of experimentation. A small piston pump should do the trick. The final application would still have to be quiet though. A case for the ole' fuel pump? Still going over the paper. __________________ WBTA (Water Block Testing Alliance) Nordic Hardware WC 101

07-11-2004, 12:30 PM	#20
Proxes Cooling Neophyte Join Date: Jul 2004 Location: In the middle of nowhere Posts: 10	I think I'd have to agree with bigben. Adding air to the system would complicate matters. You'd have to include a water trap to get the air back out of the system before it returns to the pump. Well, on second thought, I guess the reservoir could act like a trap. Maybe there'd be no need for additional equipment other than a pressure relief valve on the reservoir to let the air out as fast as you are injecting it in. Another thing to consider is that by constantly injecting air, you'd open yourself up to contaminating the water. I still like that piston pump idea way back there. I've been looking for something in the 2 gpm range that would be small, quiet, CHEAP, and run on 110VAC or 12VDC, but I haven't had any luck. Speaking of luck, I had lightning hit my house a couple of nights ago and fry two of my computers' mobos (possibly CPU's too, but I don't have anymore mobos to troubleshoot with now). Coincidently, the two computers that fried both had expensive (to me anyway) APC surge protectors on them. The third computer (a POS Dell) that had a cheapo surge protector is fine. So, looks like this little waterblock project may get put on the back burner until I get those fixed. PS - I don't think I'd recommend APC surge protectors to anyone. __________________ AMD Athlon64 2800+ @ 2.0GHz, 1 GB Corsair XMS PC3700, 2 X 36 GB Raptors 10k rpm RAID 0, ATI 9800 Pro

07-11-2004, 01:46 PM	#21
Groth Cooling Savant Join Date: Mar 2003 Location: MO Posts: 781	APC does (or used to) have a protection guarantee. Check to see if you can make a claim.

07-11-2004, 05:13 PM	#24
Aviator747 Cooling Neophyte Join Date: Jun 2004 Location: Cornwall,Ontario,Canada Posts: 13	I am sure you could port them easily enough. Just like how grafitti artists port their spray bomb nozzles for different patterns. Or you could just make your own also. A few of the fuel pumps that I have seen around are like 2 GPM or so. I still don't know if they would be able to shove the volume though. The piston pump idea is good. Especially if you have a 2 cylinder model. BUT, there is a very bad draw back too piston pumps. They heat up the water considerably. I used to work for a high pressure washing company. And we pushed 16GPM @ 10,000 PSI. The water was a little warmer than luke warm when it came out. We felt the water when it came out of the bypass tube on our gun of course hehe.

07-11-2004, 05:50 PM	#25
Bundles Cooling Neophyte Join Date: Mar 2003 Location: Scotland Posts: 32	Hmmm, i wonder if a rotary valve like the ones used on older 2 stroke motorbikes could be used for this. Making it/ buying it would be the hard bit though. __________________ "Okay, okay! I take it back. Unf*ck you!!!"

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