Overclock.net banner

1 - 11 of 11 Posts

·
Registered
Joined
·
359 Posts
Discussion Starter #1
Regrettably, I have been on air for the past few months. The lexan top on my CPU block cracked and bricked my 88GT and so I've disassembled everything until I can get a new card and block. My idea is this: use the entire case as a shroud. Let me explain.
First, I set all my case fans to evac or inward. It doesn't matter which so long as they are all the same. Then I add a few more fans wherever I find space. Following that I seal up every air hole - ie. air leak - in the case, save, obviously, only the fan vents. Then I cut a hole in the top of the case (or bottom) that is exactly in the footprint of my MCR 320. The result (hopefully) will be that the only way for air to get into (or out of) the case will be through the rad.

Originally I thought that a low pressure area in the case (fans on evac) would be best because it would avoid the warm air from the MOBO heatsinks. But then realized that it would turn my rig into a dust magnet. Obviously if I do the experiment I will try both setups at some point.
One issue will be the problem that fans competing with one another may burn them out in short order.
Also, leak sealing will be difficult. Cases, on close inspection, are rather porous.
The only other issue is the PSU. For that I think I will isolate its air system with ducting.

I've always been intrigued by shrouding and the idea of running more fans than the rad has space for by some measure or another. Please share your comments.
 

·
Registered
Joined
·
359 Posts
Discussion Starter #3
Quote:

Originally Posted by Ovlovian View Post
I don't think the air pressure created will be high enough to push enough air through the rad to handle the kind of heat you're going to be putting out.
Hmmm, I see your point. The fin density is rather high on a MCR 320. However, with five 120mm fans and two 800mm fans, don't you think I would be okay? Thats what I have access to right now and they total about 500cfm.
I really do see your point about pressure though. It's a very important consideration. Thanks.
 

·
Registered
Joined
·
359 Posts
Discussion Starter #5
Ya, no question about that. Nevertheless, I suspect the static pressure probably won't be diminished by a huge margin.
 

·
Registered
Joined
·
373 Posts
Quote:


Originally Posted by Veltri
View Post

Hmmm, I see your point. The fin density is rather high on a MCR 320. However, with five 120mm fans and two 800mm fans, don't you think I would be okay? Thats what I have access to right now and they total about 500cfm.
I really do see your point about pressure though. It's a very important consideration. Thanks.

If you were to attempt this, you have to consider that the pressure gradient that you'd be maintaining between your case and ambient pressure with your fans will reduce your fan air flow rate.

Case 1: High pressure case
Your fans all blow inward to build up pressure, but because that pressure is trying to escape, the airflow will decrease as part of your fan power is being used just to keep the air inside.

Case 2: Low pressure case
Same problem, but in this case your fans are working to keep outside air from getting in.

Fans are rated by their "free air" cfm most of the time, which refers to their flow rate when pressure effects are assumed to be 0. For this plan to work, you will need fans with high static pressure ratings more than you will fans that are rated for high cfm. This of course also means that 500cfm worth of fans in free air will not blow nearly that much in the presence of a large enough pressure gradient.

Of the two plans, I prefer the high pressure case, since the only way that heat is absorbed to the air is through molecular collisions... More molecules, more collisions. Lower pressure would remove the medium for heat transmission.

However, I think both plans might result in very little actual air movement inside the case, and I think that could be trouble. Your motherboard and RAM and any other air cooled components rely on that airflow. Maxing the airflow through your rad might help your CPU, but if you're overclocking, your motherboard needs to stay cool just like your CPU does, and that lack of air movement might hurt its temps a lot.

So that's my look at this... Not a terrible idea, but I don't know if it would be beneficial. Interesting enough to test out for a day, though.
 

·
Registered
Joined
·
359 Posts
Discussion Starter #7
Quote:


Originally Posted by sivyr
View Post

If you were to attempt this, you have to consider that the pressure gradient that you'd be maintaining between your case and ambient pressure with your fans will reduce your fan air flow rate.

Case 1: High pressure case
Your fans all blow inward to build up pressure, but because that pressure is trying to escape, the airflow will decrease as part of your fan power is being used just to keep the air inside.

Case 2: Low pressure case
Same problem, but in this case your fans are working to keep outside air from getting in.

Fans are rated by their "free air" cfm most of the time, which refers to their flow rate when pressure effects are assumed to be 0. For this plan to work, you will need fans with high static pressure ratings more than you will fans that are rated for high cfm. This of course also means that 500cfm worth of fans in free air will not blow nearly that much in the presence of a large enough pressure gradient.

Of the two plans, I prefer the high pressure case, since the only way that heat is absorbed to the air is through molecular collisions... More molecules, more collisions. Lower pressure would remove the medium for heat transmission.

However, I think both plans might result in very little actual air movement inside the case, and I think that could be trouble. Your motherboard and RAM and any other air cooled components rely on that airflow. Maxing the airflow through your rad might help your CPU, but if you're overclocking, your motherboard needs to stay cool just like your CPU does, and that lack of air movement might hurt its temps a lot.

So that's my look at this... Not a terrible idea, but I don't know if it would be beneficial. Interesting enough to test out for a day, though.

Very helpful insights. Thank you.
The cooling of non-WD'd components is indeed a potential problem and I've recognized it early on. In a high pressure situation one could direct the air to these problems areas and mitigate the problem.

I like your considerations as they have scientific footing. However I fail to see how my case would not act analogously to a shroud. By your argument, if I used only three fans (as a shroud would) it should work fine, the pressure will be irrelevant. My idea for adding more than three fans may potentially create a log jam of air, so to speak - your insight here is very telling. However, I think that the pressure created by the fans will be enough to overcome this problem. Nevertheless, as you have noted, CFM is measured without a load on either side of the fan. It is very possibly that with a small load - precisely what I may find in my case - the CFM may drop at a rate disproportionate to the increase in pressure. It will be interesting to see.
 

·
Registered
Joined
·
373 Posts
Quote:

Originally Posted by Veltri View Post
Very helpful insights. Thank you.
The cooling of non-WD'd components is indeed a potential problem and I've recognized it early on. In a high pressure situation one could direct the air to these problems areas and mitigate the problem.

I like your considerations as they have scientific footing. However I fail to see how my case would not act analogously to a shroud. By your argument, if I used only three fans (as a shroud would) it should work fine, the pressure will be irrelevant. My idea for adding more than three fans may potentially create a log jam of air, so to speak - your insight here is very telling. However, I think that the pressure created by the fans will be enough to overcome this problem. Nevertheless, as you have noted, CFM is measured without a load on either side of the fan. It is very possibly that with a small load - precisely what I may find in my case - the CFM may drop at a rate disproportionate to the increase in pressure. It will be interesting to see.
I don't mean to indicate that your case wouldn't work as a shroud. There's every reason to believe it would. I'm just citing different concerns than those that a shroud has to satisfy. Obviously, a fan shroud has no other complications to its design other than to allow pressure and airflow to homogenize throughout such that the rad gets reasonably regulated airflow. Of course, the PC case has many more internal concerns to keep in mind.

If you're interested in testing it out, I'll find your results interesting and maybe helpful for my own projects. I guess I just feel skeptical about the effectiveness of the idea overall. But if you think about it and work out some design features to avoid problems like those I mentioned, you might just be on to something.

Good luck.
 

·
Registered
Joined
·
359 Posts
Discussion Starter #9
Quote:


Originally Posted by sivyr
View Post

I don't mean to indicate that your case wouldn't work as a shroud. There's every reason to believe it would. I'm just citing different concerns than those that a shroud has to satisfy. Obviously, a fan shroud has no other complications to its design other than to allow pressure and airflow to homogenize throughout such that the rad gets reasonably regulated airflow. Of course, the PC case has many more internal concerns to keep in mind.

If you're interested in testing it out, I'll find your results interesting and maybe helpful for my own projects. I guess I just feel skeptical about the effectiveness of the idea overall. But if you think about it and work out some design features to avoid problems like those I mentioned, you might just be on to something.

Good luck.



I understand entirely. In fact, I insist that you remain skeptical throughout the experiment. That is the absolutely most fundamental function of a scientific community. Without skepticism we'd still be burying chicken bones in our backyards, sticking needles in dolls, mixing superstition with politics (well we still do that to an unfortunate degree), reading tea leaves, praying to sky-gods (again, unfortunately people still do this), trusting omens, paying Shaman, and all other manner of stupid *****. Anyway, I look forward to your analysis of my results. I'll start a worklog when I get cracking.
 

·
Registered
Joined
·
373 Posts
Quote:

Originally Posted by Veltri View Post
I understand entirely. In fact, I insist that you remain skeptical throughout the experiment. That is the absolutely most fundamental function of a scientific community. Without skepticism we'd still be burying chicken bones in our backyards, sticking needles in dolls, mixing superstition with politics (well we still do that to an unfortunate degree), reading tea leaves, praying to sky-gods (again, unfortunately people still do this), trusting omens, paying Shaman, and all other manner of stupid *****. Anyway, I look forward to your analysis of my results. I'll start a worklog when I get cracking.
You don't know how happy it makes me to see people who care about scientific process. Until this recent water cooling project of mine and moving to OCN as my home on the net, I spent a great deal of time discussing the properties, merits, scope, etc. of science. One of my favourite topics.

Not that I'm a scientist by trade, even. I'm a test engineer. I keep automated trains from colliding with one another. But people don't appreciate that science can be used in almost every facet of one's life. Or well, they figure its no fun or too much work or something. But I try to think scientifically at all times. You get used to it.

Friend request sent
 

·
Registered
Joined
·
359 Posts
Discussion Starter #11
We can speak volumes on the scientific process. However, I'm sure we would just agree with one another on its nature and function. I'm also not a scientist... yet. I am training as a political scientist. I can't say that I think scientifically all the time - not day-to-day. I do, however, try to think analytically and rationally at all times. This is the way that proper science is conceived of and I'm sure that what you meant.
Sure, to some we're cold people because of it, you and I, but at least we're dedicated to veritas.

In my field - one that is currently under attack from subjectivist, reality- and truth-denying post-modernists - I apply traditional epistemology to my methods to produce the most undeniable conceptions of truth. However, I am not always right and I happily adjust my understandings to fit new data. This is something that cannot be said for idiot post-modernists. Holy I am ranting. Anyway, if someone tries to tell you that Foucault or Derrida are great, ask them why. Then turn around and tell them that for those reasons the theorist in question is not great. They can't reply because their earlier reasons will have been so disjointed and subjectivistic that nothing is clear. [rant over]
 
1 - 11 of 11 Posts
Top