Water proof sealing/gasket design

Hello to all

I am facing project related to water proof design(kind of measure tool for construction stuff), something close to IP68.

Have you any tips how to start designing jousings and “sealing between” them? How to determine compression force deflection?

How to select best metrial for sealing and appropriate section and, at the end, manufacture solution.

thanks in advance

There are a few ways, so it’s hard to know where to start. Here are some of my first thoughts:

  1. Find a product that is IP68 and tear it apart. Maybe a GoPro is, but you can definitely find an IP68 project box at your local electronics supply store. It will look like this:

Notice that the joint surface is planar and the fasteners are outside of the gasket. Non-planar joining surfaces are a pain to seal properly, especially at IP68.

  1. Consider welding and sealing with epoxy.

For welding, search “ultrasonic welding”. There are many guides available and your plastics supplier can help you out on the final design details. It’s widely available at both North Americana and Asian factories.

Sealing with epoxy is also widely available. You’ll find this done in a lot of electronics. Things like relays are often epoxy sealed. It makes the product nearly impossible to take apart (which may or may not be a benefit to your design), and resistant to vibration as well as water tight. It looks like this:

First off - you need to decide your spec.

“Close to IP68” isn’t a spec. IP67 is most common for tools that are designed to be rugged, but not used underwater. IP68 is typically reserved for things like underwater cameras, things that must not only be water proof, but waterproof over a longer duration of time and at greater depths where the external pressure is much higher. (IP67 is only waterproof to 1 meter for 30 min, 68 means you can specify waterproof up to 30 meters, etc)

Since you described this as a construction tool, unless it is being used by underwater divers than IP67 is more than adequate for rain/puddles/buckets of liquid.

Without this turning into a full engineering tutorials, most products do one of the following:

-Seal the enclosure permanently with welding/goop (like Ray mentioned above) - obviously this is more risky because it prevents any rework at the manufacturing level, and the housing would need to be destroyed to disassemble the device. This is usually only done on electronics that are meant to be tossed at the end of their useful life or small subcomponents.

-Leave the enclosure mostly unsealed, but protect your electronics - this can be done depending on the design of the actual product. Sometimes it’s OK for the product to get water inside of it, as long as your sensitive parts are properly protected with a smaller sub enclosure. You can also hydrophobically coat sensitive parts to protect them. In a construction environment however, you probably have a lot of dust and particles to protect against

-Build a compression seal between your two housings and any openings of the product - this takes your standard O ring design and modifies it to fit your product. A soft rubber/silicone/elastomer will be squished between your two housings and compressed by the force of the screws that hold your product together. This requires the most engineering because obviously with a complex ID you will need to make sure that your screw bosses are consistent enough to apply a proper amount of force (but not to look ugly as shit) to compress the seal. You will also need to make sure the seal is easy to assemble - some products have very complex seals that slight misalignments during assembly can cause the seal to flop over, twist, and then not properly seal. This means that even though some of your products will be IP67, others will fail sealing tests in the field.

Anything you seal with a compression seal will require you to also seal any openings. Take a look at some of the earlier Samsung phones and you can see the very tiny seals that were used to seal the battery door, USB ports, etc. Those seals are not terribly robust, especially in a construction space where dirt on the seal will create openings for water to enter the device.

I’ve done a number of products where the seal material itself is over molded on to one of the housings using elastomer and eliminates the need for the O ring (and additional space it drives for the channel and wall thickness on each side).

You can see some of the seal designs on this image - the outer material of the top cap was TPE (though due to quality issues we had to add a third softer durometer material to improve the sealing consistency) and had foam gaskets and poron seals around the openings for things like the speaker and receiver. This was not a flawless design by any means - it took a great deal of energy and rework to get the rubber durometers correct and the multi angled seal on the keypad was a nightmare since it often shifted during assembly.

If you can get away with a flat, and consistently compressed seal your life will certainly be the easiest. But cell phone companies have taught us that there is a lot of clever engineering that can be done to seal a device that doesn’t require a 5mm thick wall section to achieve.

Thank you guys

I admit: IP68 too much for my application., IP 67 would be appropriate.

Over molded solution seems to fit best to my product.

Question is:

  • how to select/develop right cross section of the seal?

  • how to calculate compression force?

The best way is to work it out with your supplier. They probably know how.

Second best way is to find something similar and copy it. That’s why I like the project boxes so much. They’ve been around for decades, so they are engineered to just work.

Third way is to wing and build in too much tolerance so that you can always reduce the tolerances later by machining in the mold. Basically, make everything fit a little loose and then adjust.

There are engineering calculations that can be done in deciding the correct profile of the seal, the size of the rib is going to have to do with multiple factors:

-Softness of the seal material (Shore #)
-Compression force (distance between your screw bosses)
-Flexibility/thickness of your substrate (at a certain point, your plastic will bend from the force of the seal and will reduce compression force)

You can look around at existing parts, but at the end of the day you will likely need more than just a copy and pasted design due to the varying number of factors. Since it’s difficult/impossible to test sealing on rapid prototyped parts, your best bet would be to try and find an outside engineering firm that could potentially help with some of those design decisions. It would save you money vs continued tooling modifications.

I’m going to have to disagree. I have used Accura sla materials in conjunction with Objet Tango materials that translated into direct dims for the two shot part. We utilized snap features instead of screws for assemblies that hold liquids.

Obviously there are limitations to 3D parts. Brittle. Can’t weld them. No guarantee that the tolerances will match IM. But I have found at the very least they will get you to a metal-safe start on your tooling. Once you have committed to 5-6 figures on your tool, an extra thousand or two to nibble a thousandths or two is not going to break my budget. Have you ever had a tool that didn’t need changes after first-shot parts?

I think 1 time in 100.

Fair enough - the last time I used an Objet 2 shot part was probably ~5 years ago, and at the time the materials were not very good, and had very bad flexibility (usually flexing would cause them to delaminate from the substrate) - but it’s good to hear you’ve had better luck with it. Of course that’s ensuring that the shore rating of your RP part is able to be similar to that of your real part (otherwise it would be tough to get match the testing).

I think the bigger concern in my mind is that without seeing the OP’s actual design, without any engineering support on the design of the ribs, spacing of the compression elements, etc it can be a bit of a crap shoot.

I was not implying that any tool would be done on the first try, but I was implying that if this is a new area for their design, it is worth spending the extra couple of thousand on a mechanical engineer with experience in sealing design. Especially for IP67 which is very challenging (even with a full team of ODM and internal ME’s we would spend weeks debugging and carefully modifying the tools and design to meet IP67 - and it was standard for most of our products).

I also live in the rare land of high margins. We can “over-engineer” parts. Extracting every penny out of cost takes a back seat to getting to market and we can move forward without spending the time to optimize the part.

And for the record, I have never used an Objet 2 shot part. I make separate parts, and assemble. I know if that works, the IM parts will work. The problem occurs when the printed parts don’t work and I need to convince the powers that be to sign a 6 figure PO for tooling.

Coming from the land of “Why is this so big? We need to shave off .5mm” every chunk of sealing material and wall thickness that went unused would have a big impact to the ID/ergonomics so I got used to fighting tooth and nail for wall thicknesses and seals.

But now I just move pixels around…ah the good old days. :wink:

An easy way to do the calculations is to use Parker’s O-ring handbook. There is a section that will tell you the range of forces/liner inch at a certain durometer and o-ring width. That will give you a starting point that can be translated into overmolded parts. In practice, 30-50 is very soft and probably right for IP67, 70-90 is better for greater protection. At high pressures (spray or depth), you will need other tricks.

Balance compression force with number of fasteners and the stiffness of the part.

In general, I would say the Parker book is a conservative approch, but right order of magnitude.

Figure 2.4 in the parker book here: https://www.parker.com/literature/ORD%205700%20Parker_O-Ring_Handbook.pdf

I’m attempting to create an IP67 seal at the moment. I tried to get a sealable 3D printed part before realising it was more or less impossible due to the bending of the part. I’m considering now using JLC’s SLA printing service to get a completely filled part hoping that there won’t be considerable deflection.

I’m wondering now if my problems prototyping sealing are only from the stiffness of the material or if there are problems with my design too. Would you be able to comment on this? Here are two designs I am considering, I’ve tried a 1.78mm FKM material but cost is too high although sealing seems easier.

https ://global.discourse-cdn com/business4/uploads/core77/original/3X/3/8/3870b517fae57cb1cd89d57abdd8c8b91aadb55d.png

https ://global.discourse-cdn com/business4/uploads/core77/original/3X/f/8/f890c236f3dfcfdb2f022b2f209c0717610d7e64.png

The first design is a 1mm nitrile o ring with a hardness of 70 Shore A.

The seal profile here is a flat construction and the groove is a racetrack. While the internal radii are poor the seal is temperamental but sometimes effective up to 500mm. The dimensions here are consistent with Parker although I used the Apple Rubber calculator for them.

https ://global.discourse-cdn com/business4/uploads/core77/original/3X/5/c/5c879afcae83ae18bd9265c88d9d539d4550d906.png

https ://global.discourse-cdn com/business4/uploads/core77/original/3X/1/1/113b5e5ccfd2c66a847b8282f4734756b089e0c2.png

In this first design I have a 1.5mm Silicone o-ring with a hardness of 60 Shore A. The seal profile is a tongue and groove construction.

The internal radius is good and the tongue should be able to compensate for some level of bending. The hardness of the seal has also been reduced as well. The dimensions used here are taken from an existing design that has been certified to work yet this only seems to seal at or above 150mm, and in that case chances are there is no sealing occurring at all.

Help! Will SLA fix all my problems and am I just trying to fix the impossible or are my designs faulty too?

note: links have space added after https and the fullstop removed before com

To add, I’ve been using clear coat to cover up irregularities of surfaces after sanding, and also to make water proof the general body of the FDM part. I doubt that the surface finish is the problem with my sealing.