if i wanted to fight physics i would be pointing the helper fan downwards towards the mac pro which would be ridiculous. i'm [attempting] to work WITH physics since (a) hot air rises and (b) the helper fan would only help the vacuum created by the tube to pull hot air out.
also - agreed - putting these things below a mac pro would be equally ridiculous.
i'm not buying the HYPOTHESIS that the heat from a mac pro would ruin two drives above it. hard drives have been proven to literally take a beating in high heat situations. and it's not like the air would be stagnant, it would still vent out the top. consider that GOOGLE has provided a very detailed report on failure trends of magnetic media from 2007 where they have concluded
"it is likely that there are other effects which affect failure rates much more strongly than temperatures do" - i'm betting they own a lot of drives and have a lot of experience in that field.
(
http://static.googleusercontent.com/media/research.google.com/en/us/archive/disk_failures.pdf)
just wait and see, when I marry a supermodel, I'll take all of her money and vest it into R&D for this. just you wait....
That paper reveals more than you imply. In their study, temperature alone, was a much poorer predictor of drive failure rates than shown in prior studies. An earlier study by a Seagate engineer in 2000 reported "that MTBF could degrade by as much as 50% when going from operating temperatures of 30C to 40C". The Google results were surprising because: "In our study, we did not find much correlation between failure rate and either elevated temperature or utilization."
Part of this discrepancy can be explained by advances in both drive design and manufacturing over the intervening 7 years. Another part of the difference appears to be due to the fact that google identified several sources of data which show much higher correlation to failure rates than temperature such as, SMART data, sector and seek errors, etc. Finally, the role that temperature plays is not uniformly distributed over drive lifespan.
Even though temperature alone was a poor predictor of failure, it was still statistically significant. At ages less than 2 years, temperature alone was insignificant, but at 3 years the failure rate attributable to temperature double when temperatures rise from 35C-40C, and triple when comparing 35C-45C. So, operating temperature still matters for older disk drives.
At this particular point in time, I think this is highly relevant because people migrating from older Mac Pro would use such an enclosure to migrate internal storage. If so, the drives are likely to already be 2 or more years old.
The optical drive is a greater concern than the spinning hard disks. These are much more sensitive to proper cooling. Most venders rate their drives for use only up to 35C (95F). Cooling and heating cycles for DVD and CD disks also impair the optical media.
Working against physics is far more than "heat rises". It has to do with the fact that heat travels from hotter to cooler temperature regions, and does so at a rate which depends on the difference in temperature not the absolute temperature.
When these external drive are under idle conditions or modest load, the exhaust from the Mac Pro is likely to be adding heat to your storage devices instead of letting them dissipate heat to the much lower ambient temperature of the room. This is working against physics.
When the Mac Pro is at high load, the exhaust may be at higher temperature than your optical drive is designed to function. No mater how much air flow you get, it will be heating the drive beyond its operating range rather than cooling it at all. This, too, is working against physics.
Beyond this, even though your design calls for an additional cooling fan, it runs in series with the Mac Pro's existing cooling system rather than in parallel. This means that when the Mac Pro is relatively cool, but the components above need greater air flow, the system below will be acting as drag: impeding air flow rather than accelerating it. This requires an even larger and beefier fan operating at a higher duty cycle.
In over 20 years as a systems engineer my most useful rule of thumb was, "If a problem seems really difficult, STOP and look again: you are probably trying to solve the wrong problem."
In this particular case, I don't think that your proposed design is one that is worth throwing a bunch of money or time at. The design pre-supposes a high thermal load which is independent of the problem being solved.
Adding a Mac pro inside a storage enclosure adds a large cooling problem to the design. From a physical standpoint your proposal says "To start with, let's quintuple the thermal load." As an engineer, this is tantamount to saying let's take a small problem and make it large. Simplicity in industrial design emerges from solving a few problems well. This is unlikely to emerge when the initial constraints add new problems which did not originally require a solution.
You obviously have the graphical design chops and creativity to produce some nice ideas, and you should be encouraged to continue. However, there is a huge gulf between working with pixels and working with objects. The real world has real constraints. If you want to transition from graphical design to industrial design, you should strive to gain some more experience with, and understanding of, physical limits.
I realize that this critique might be interpreted as pissing on your cheerios. That is not my intent. I hope that you can see it in the light in which it was intended: an attempt to get you to look at some issues which you may be unaware, and which pose larger problem than I believe you thought.
I wish you well.