CNC aluminium wall/route thickness for static axial flange seal

Hi everyone, I've just joined this forum needing some design help, and from a brief flick through the various topics, it looks like I've got a lot of reading to do!

I'm designing an enclosure for a PCB, which requires sealing against external pressures (namely external dust and water, standard electronics enclosure requirements). Attached is a quick mock up of one side of the enclosure, showing the following:

1. A wall thickness of 5mm (machined from 6061 aluminium).
2. M2 screw threads drilled and tapped into the wall of the enclosure.
3. A channel in which the o-ring/cord will be placed.

The second attachment is an excerpt from the following site, which provides some basic sizing guidelines for static axial seals:

https://www.aceseal.com/gland-design...al-application

In my diagram, I've selected the 1.02mm o-ring (which corresponds with the -001 standard size), with a 'Groove' width (w1) of 1.4mm and a 'Gland' depth (D) of 0.65mm. While not exactly the same as in the above link, I've found these two dimensions elsewhere in similar guides.

The question I have, as someone who has never done any form of industrial design, nor any type of CNC or machining work, is the design as shown and explained above practical? I.e. if I take a CAD file to a CNC machinist, will they be able to route a 1.4mm wide, 0.65mm deep channel along a 5mm wall thickness piece of aluminium, with only 0.475mm between the wall and the channel?

Any input, guidance or alternative approaches to creating a seal against external liquids with a 5mm (or less) wall thickness with M2 screws would be greatly appreciated!

jars121

A 1mm O-ring would only need 20% crush for a static application. At 35% crush you risk damaging it if you calculate the volume of the oring and the volume of the groove make sure there is room in the groove for it. (I just calculated your oring at 0.79mm2 and the groove at 0.91mm2 so there's room)Wet sealing with a suitable sealer is the go these days. No need to machine a fiddly groove. The groove is possible, most CNC mills can do this with a 0.5mm mill say. But look up sealants for metal to metal and you will find many. Peter

https://www.permatex.com/products/ad...ubber-sealant/ something like this. I use it in mining equipment much easier than orings to fix in the field and in the workshop. One size fits all...

Thanks for that Peter, much appreciated. Would you still opt for a wet seal on a production enclosure where you might be manufacturing and assembling larger volumes? Once closed, the device won't (in theory) need to be opened again.

Most car parts are wet sealed using robotic placement these days. Wet seal is perfect for lifetime seal. Orings are for things being opened and closed often. Peter

Can't argue with that! If it's good enough for an automotive OEM, it's certainly good enough for me.

In that case, is the 5mm wall thickness reasonable for a Permatex wet seal, or could it be reduced considerably? 4mm seems like a good fit, as I can centre the M2 screw and allow some space for the ~3.2mm low profile M2 screw head diameter on the enclosure lid.

Hi Jars- For width advice talk to the Permatex techs. Permatex is a good brand used several of their products all good. Some products are designed for flat metal to metal seal others need a small groove or step so you don't get a zero thickness line (ie no seal!). Peter

Goo seals are quick and easy, but... If you need to remove and replace the thing regularly (or ever) I'd recommend for an o-ring over gasket goo because having to scrape and replace seals gets old real quick and, particularly on aluminium faces, often leads to the face getting wrecked on the first go.

My last client made their own enclosures for their mining and military mobile electronic equipment products. They tried a few different options but in the end they went with o-rings.

I've made my own electronics enclosures for dive related electronics where the pressure becomes a major factor and, if I was happy enough to use goo, I was happy enough to just fill the enclosure with potting mix. Conversely, if I'm not happy to pot the whole thing, I wouldn't use goo.

Bit of an update to this thread.

The enclosure has undergone a considerable design change, as I needed to reduce the overall footprint by a sizeable amount. This is the current approach:

- Wall thickness has been reduced to ~2mm. This should still be plenty strong enough, and reduces the weight of the enclosure somewhat as well.
- I've done away with screws on three sides of the enclosure, and just have two M3/4 screws on the top side of the enclosure where there is a small extruded section with more wall thickness. These screws also support the main moment of inertia generated from the various cables interfacing with the enclosure connectors.
- I'm going to use a sealing adhesive (possibly a Permatex as recommended above or a Sika compound) as the primary joining/sealing method, supported by the two aforementioned screws. I've done the calculations with some of the Sika products, and even with a ~2mm wall thickness, there's enough mating area available to generate over 1000N of clamping/bonding force, which should be plenty for my application.

The next question I have is around the mating interface itself. Peteeng mentioned above the potential need for a groove or step; this is something I've been trying to research but haven't yet come across anything of note. A flat metal-to-metal interface would certainly be ideal, mainly from a machining perspective, but in my mind it makes sense to have a small groove to situate the adhesive when the two surfaces are pressed together.

Does anyone have any input on this particular matter? Any links, forum posts, videos, etc. that explores mating interface design principles, etc?

Thanks!

Hi Jars - The sealant you choose will have a tech rep. Speak to them and they will be able to answer you Q's about minimum width of seal or if it needs a groove for minimum thickness to seal. Some sealants are for "thick" bond lines some are for thin. Peter

Thanks Peter, I'll be doing exactly that this week. I'll report back with any updates :)