Everything You Know Is Wrong - Mesh vs Non-Mesh Front Panel

What started as a mesh vs solid front panel test quickly turned into fan configuration testing when our RTX 3090 was getting hotter with more Noctuas in the case. Come find out why!

Purchases made through some store links may provide some compensation to Linus Media Group.

AFFILIATES & REFERRALS
---------------------------------------------------

FOLLOW US ELSEWHERE
---------------------------------------------------

FOLLOW OUR OTHER CHANNELS
---------------------------------------------------

MUSIC CREDIT
---------------------------------------------------
Title: Laszlo - Supernova

CHAPTERS
---------------------------------------------------
0:00 Intro
1:18 Mesh vs Solid
2:35 GPU
3:35 Gaming
4:28 Smoke
6:40 Conclusion
7:58 Outro

[johnsonwang]

Guys, if you do this fog test at home, pay real close attention to the type of fogger and fluid you use. Most combinations leave oil residue on whatever the smoke collects on.

[Maarrk]

I appreciate the testing guys, but you'll want to differentiate between turbulence and vortices. All of the flow at these speeds and scales will be turbulent, which is actually great for cooling, as it mixes air more. What you call turbulence are actually stationary vortices, and this is the thing that is giving you worse temps.

I hate to be the "um, acktchually" guy, but if you use the correct terms, maybe the viewers will find better resources when googling about this stuff. Hope to see more advanced flow testing as the labs get developed, keep up the good work!

[JokerTheDank]

As an Aerodynamicist I have one pet peeve: what you refer to as "turbulence" is absolutely not turbulence but just ricirculation bubbles. Thank you and sorry

[djrbaker]

I used to work with smoke machines for a living. If you intend to do the same test at home DO NOT DO IT. It coats everything with a thin glycerin layer, the more you do it, the thicker it gets. Then your components get all sticky and start to collect dust and turn into a goop.

[ImJBlu]

I was expecting some type of actual trouble shooting like seeing if lowering the fan rpm or adding a guide for the airflow would help at all

[martintrencovski]

Linus: “Yep we’ve got one”
I love how after all these years they are still throwing shade to the verge

[ThroughFallenEyes]

I have an H510 elite (I bought it before I knew anything about thermals) and found that by running the front fans at the lowest I can possibly get them spinning (cant remember the speed off hand), and increasing the speed on the top and back fans by a tiny amount, the thermals improved drastically from the stock configuration.

[TheForfeef]

I'd be very interested to see a deeper delve into this. There are so many unanswered questions. What about AIOs? does having the rad on the top or bottom make much of a difference? What about having your intake on top and exhaust on the sides? What about 1 exhaust vs 2/3/4/5 intakes? What about the same number of intakes and exhausts and just different RPMs? So much could be learnt from this and I look forward to what you will do next with these experiments.

[oglcn]

One interesting discovery I had was that, moving my single front intake fan from middle slot to lower slot decreased my GPU temps almost by 10C° It blows directly under the GPU so provides much better airflow for the GPU

[dividion]

I'm curious if flipping the PSU over to use as another exhaust would've broken the GPU air bubble.

[Amunre]

Would be interessting to see what would happen if you separate the airflow from the intakefans with a piece of cardboard (attached between the intakefans and the graphicscard). Basicly giving the cpu one intakefan and the graphicscard one intakefan, also limiting the posibility of turbulance... ?!

[nickryan]

Definitely would like to see more about this. I've been building PCs since before fans or even heat sinks were fitted to the CPUs, the graphics cards were more problematic due to their jumper configuration and PSUs that were built into the cases... the various attempts at improved or standardised improved air flow and other cooling have always been of interest. Not that I ever build a system that has the top end components in it therefore I don't care that much but cooling for even mid range PCs is becoming more and more important.

[ZeroUm_]

With such analysis I was expecting some exploration of fixes to the problem, like putting the GPU in a lower slot, getting a smaller CPU cooler, larger cases or trying a intake config with smaller but we'll placed fans.

Hopefully the new labs channel can get long form videos about that.

[sokka]

Would be interesting to see if some 3d-printed guide could improve the airflow to the GPU.

[Abcgum]

Amazing work guys! Was hoping to see some kind of solution in your build configuration (smaller GPU, being able to find a way to add that fan at the top, etc.) with results.

[Rebelguagez]

I'd like to see this with a shroud forcing the airflow to the deadzone...Be interesting to see how free cardboard to direct airflow might be better than strapping on more fans/vents.

[Mnc]

I'm interested in seeing this test performed on a case with bottom intake fans, like the Fractal Design Torrent.

[AatreshKarnam]

A few quick thoughts on the conclusions you guys drew about turbulence. Most real world air flow systems are inherently turbulent since laminar flow requires carefully controlled environments & pretty low air speed requirements(which are often possible only in laboratories). The fans that are sucking in the air are making what is already a turbulent stream even more turbulent by adding more velocity(energy) to the incoming air. This is a non issue for the CPU because the air reaching the CPU tower has unhindered access to the air brought in by the case fans.
But in the case of the GPU the shear width of the card blocks a bunch of incoming air deflecting it above & below the GPU. This sets up a local vortex system that is powered by the incoming air from the case fans. What does that do? It converts the linear velocity of the incoming air into rotational velocity making the air simply go in circles above & below the front of the GPU. This effectively cuts off any air to the rear of the GPU starving it of fresh air. The same thing happens in front of the CPU fans as well if you notice in the highlighted box at 5:58. What's even more interesting is there seems to be a vortex pair in the top section of the case that are feeding each other drawing in more air. You would see all of this in much more clarity if you use PIV as mentioned by @Spirit, but for starters using a much lower amount of smoke will greatly improve the clarity of all the phenomenon I discussed above.

[nomiSimple]

I test this sort of thing for work using 1-wire temp sensors for both ambient and attached directly to devices. Air cooling is a dark art and the results of small changes are often surprising when mapped over time.

[mikicerise]

Living in dry, dusty central Spain, negative vs. positive airflow has always been a big question mark. Around here, my own feeling is that you should prefer positive airflow, because you want to direct air intake through the dust filters. If pressure is negative, you get air coming in through other apertures in the case, along with all the dust it carries. On the other hand, positive air pressure tends to lead to dead zones and air recirculation. I try to get the balance to just barely positive, but it is tough to balance. If it weren't for the dust I would go for slightly negative.