Calculating Design Failure

[Vig16]

I'm designing a cylinder-shaped sealing component to a part that needs to have a face seal on the distal end of the part and a seal on the OD as well. I've attached a sketch of what I'm thinking for the component.

The part will need to be a single durometer material (to be chosen after the calculations of force required are made). So far, I'm thinking of a durometer material of 50-70 Shore D. The far right of the part will be the part used for the face seal and the curved geometry at the same side will be used as a radial seal.

I'm looking to determine how thick the material needs to be on the area on the far right the looks like an "S" but I'm not sure how I would go about doing this.

Thanks!

Nick

 

[Vig16]

I"m not sure if that file opens for everyone, but I attached a JPEG to this post...

 

[tr1ntx]

For similar type calculations on non-metallics in the past, getting the material properties has proven to be difficult, as they vary widely with different treatments and molding temperatures and such. Then it seems you almost need to be an elastomer or plastics specialist to know the rules of thumb on how to deal with it.

I think I understand that you're wondering about the thickness that is currently dimensioned as .009. The flat surface at the far right, which you said will be the face seal, will presumedly experience some sort of load to the left, i.e., compressing the material. It appears that such loading will cause bending in the .009 thick section, although it is pretty close to pure compression through the middle portion just below the R.005 inside radius.

I'm thinking it will require iterative calculations to home in on it, checking to make sure you don't exceed the limits in that top sloped face that will/might be in tension. Even then, you're probably going to have such wide-ranging values on the material strength that you'll need to CYA and make it plenty thick. If you want to utilize the material's reaction to generate a particular sealing force, that will likely require physical trials, unless you get lucky and talk to a material manufacturer who is confident enough in their data that they narrow it down for you. There is surely emperical data out there somewhere on a similar geometry seal, so if you're lucky and hit the manufacturer/supplier who has it, this might turn out to be a simple calculation based in dimensional compression of the face seal surface.

 

[Cockroach]

I have a problem with "face seal" in general, as used in this application. The integrity of such a seal relies on clamp forces induced from the face plate, presumably through bolting. Anyone experienced with face seals using gasket has a deep understanding of how tricky this can be, especially in achieving high pressure seal integrity.

The other observation is the departure from classic, commercial available elastomers. Are you suggesting yet another design for a specialty seal? Then you have opened up the design to added expense of seal development and testing using a quality manufacturer. This also opens up the technical uncertainty in maintaining performance over time of the application.

But in general, o-ring durometers I prefer are Duro 70/75 for static applications and Duro 85/90 in dynamic. The actual material preference is selected knowing the medium you wish to contain in the pressure vessel as well as a few minor concerns such as pressure class, installation and maintenance for future turn arounds.

Not understanding the overall design, it is quite difficult to add further comment. Good luck with it.

Kenneth J Hueston, PEng
Principal
Sturni-Hueston Engineering Inc
Edmonton, Alberta Canada

 

[Vig16]

The component is a threaded tip that allows it to stay in place but the reason for the face seal is to prevent electrical shock as the device is used in cautery applications.

 

[Vig16]

Would a deflection equation of F=Y*(3EI)/L^3 be the best way to calculate the defelection when I already know the force applied?

 

[Cockroach]

Treating it as a circular beam, I would imagine so. You are assuming a rigid support at one end and free movement at the other.

Kenneth J Hueston, PEng
Principal
Sturni-Hueston Engineering Inc
Edmonton, Alberta Canada