Separate names with a comma.
Discussion in 'Custom Cases & Projects' started by AlexTzone, Oct 14, 2017.
Ok this has to be satire...
From the Kickstarter webiste: "Projects that involve the development of physical products must feature explicit demos of working prototypes."
We may need that prototype after all before this is listed on Kickstarter...
Hi @AlexTzone , so I'm just curious: What is the dimensions of that fan filter? approximately of course. I like making things and one thing I'm thinking of is a panel of 92mm fans. You don't have to oblige me by changing your design, just want to know is all
Any reason we're keeping the PCIe bar (between the slots)? is it for functional or aesthetic reasons we're keeping it?
Apparantly it is gone, Alex thought I was talking about something else initially.
Many questions and comments have been raised since the last update. Hopefully Alex will chime in soon with clarifications.
All the best.
The inside mount looks pretty clean, but I'm definitely concerned about temps. Also, would the glass be sturdy enough with just those four bars holding it in place? If I remember correctly, your first design had the same border around the glass, but you eliminated it due to concerns about structural integrity.
No one has ever claimed that Louqe invented the cheese grater look that dates as far back as the 2003 PowerMac. No need to be obtuse about it.
A window kit on stand offs without cutout will result in very very bad temps +15°C. The cpu heatsink will recycle a lot of air. The stand offs doesn't help that much.
Creators and community can only suggest so much based on their own experiences. I understand that you may have a schedule to follow, but it would be prudent to create prototype of the different panel designs and test out the thermal in real life (just acrylic will do?). At this moment it's all guess work about which panels give what kind of thermal performance, which is not good engineering. Many have suggested that completely solid panels will not work, thermals will be off the roof, but you should test it out also just to see how it performs. Who knows? Maybe it works on your case? If not at least you know how far off you are. You could make some simple prototype panels in a few days easily, definitely better than relying on our "comments" or pure guesstimation.
@macbosco I guess it's much better to run some simulations before and only then begin phisical tests on the small batch which gave the best results... Probably not the cheapest option (even if, one can opt for the trial of the choosen software for a couple of weeks) but for sure, the most scientifically driven approach, I guess...
there are already videos on youtube on how to set a flow simulation for a fan...
adding a solid panel ahead with some mesh in it shouldn't be that hard
#can't you see how original is my design? In mine the led is under the power button
I attempted to do exactly this for Macbosco's project with very little success and I have a PhD with a heavy focus in finite element modeling. There are some subtleties with CFD, and particularly with Solidworks Flow which make accuracy really challenging. Trying to model the flow profile from every individual perforation in a side panel creates huge solver times due to the meshing requirements. If you instead designate it as a porous meshed surface governed by area fraction of permeation, you lose out on all of the vorticity and boundary layer effects. Then you still have to worry about fan profiles (pressure vs flow) for approximate fan vendors and sizes unless you're going to model the fan blades and their housings in detail which once again provides a significant performance hit. Lastly, how finely are you going to model CPU coolers, thermal interface materials/thickness, and are you going to treat air as isothermal and incompressible or with full temperature/pressure dependent fluid dynamics?
The type of FEA that I typically work in is not fluids focused, so I'm by no means an expert at CFD and/or Solidworks flow, but it ends up being far more time consuming and challenging than that tutorial implies. If you have access to a laser cutter and some acrylic sheets, you'll probably be able to generate usable data far faster than with a CFD code (unless you're already an expert in it).
I believe you guys should definitely run some tests with different side panels to see how they alter the internal temps of the system on a prototype. Even if this was to delay the kick starter process, I think it would be well worth it to have a fully functional idea of the entire chassis to ensure you guys are delivering what you want. If these tests result in a need for cutouts in the glass panel, as long as the design is something clean, i think most would be ok with it. Im leaning towards the fact that some sort of cutout will need to be made, as thermals are more important than having a clean glass. Im sure some sort of design can be agreed on with most here.
+1 for cutouts or thermal testing to confirm temperatures
Glass is no good if the hardware you’re showing off is overheating.
Computer simulation can only do so much, that's why the aerospace companies still use wind tunnels for testing
If at all cutouts are an option, I do also think that @AlexTzone should test them with acrylic panels. I think I have come to accept that the tempered glass panels will be for pure looks while sacrificing thermals, and I will probably cut out some acrylic windows myself to get a true breathable panel that doesn't hinder acoustics either.
A video tells a thousand pictures worth of words, so here is a live demonstration of a honeycomb grille on a fan. Notice that the noise is not as bad if there is even a little space, but this may not be the case for all patterns.
The final configuration and how it comes out of the box should also to a degree should influence who you send out the product to review. If the performance is likely to be bad, then sending it to Gamers Nexus would be a really bad idea since they will rip the product to shreds. While I very much like Gamers Nexus content, if the objective was to put this product in a good light, I wouldn't send it there - Their focus is primarily on absolute airflow and temperatures "out-of-the-box". To take a recent example, they sort of have a stab at the Ophion Evo for not coming with a fan. Also their chart toppers are almost always the loudest cases and so to score "well" on their charts, its important to have a very breathable case with fans included, not what you are doing here.
Optimum tech is also a bit of a temperature nerd and so would be risky given his viewership. Bitwit is a bit more "gentle" and slightly less critical, and has a larger viewership. I think LTT is too big and they probably won't do another small case since "its not different from their Dan case video". I think since this is ultimately an aesthetic focus case, somewhere like Hardware Canucks would be the most suitable for review.
I like Gamers Nexus, but it's more technical for the lesser enthusiasts and I think Steve badly needs a haircut
Another off-the-cuff video to demonstrate the effects of a solid glass panel vs a fan. Sorry for the quality, unedited, and unscripted nature of the video (and the fact that I am speaking at all), but I just did it quickly to get my point across. The "flow reversal" seems to happen at about 10 mm for most fans, and at about 15 mm qualitatively performance seems to be restored.
Obviously I'm no professional tester, but you get the point.
Here's the thing, you can get acrylic panels but they scratch easily and will look like trash inside of 3 months unless you handle them with rubber gloves. We have seen this a couple of years ago with the bigger case manufacturers and their acrylic side panels
You can do only so much airflow and temp testing without a 1:1 mockup of the case. You need to decide what characteristics the case needs to have and stick to that the best you can.
The case is solid, all the pieces are there to launch the Kickstarter. Once that is over, final tweaks can be made and everything can be built.
I am a mod on the ghost s1 slack server (and the QC volunteer). Over the last year I have learned 1 thing, this case will not fit everyone's needs. There is a point where you need to say that's enough and take what you have and run with it.
Yes you can try and plan for future hardware but you need to keep in mind that at the end of the day this is a SFF case and you can only fit so much into a small space without making the case bigger and defeating the purpose of the case or worse, changing the essence of the design to the point where it's unrecognizable.
Do the best with what you have and go from there.
I am more than happy to lend a hand with the community and give you as much insight as I can to make sure you can learn from the ghost's production mistakes @AlexTzone