Alternative flange design

[LKH005]

Hello,

I'm working on reducing leakage from the interface between a steel manifold and a high temperature device running on natural gas.
The high temperature device has a flat bottom surface with XX inlet/outlet holes, diameter approximately 10 mm. This surface I cannot modify.
The steel manifold has inlet/outlet holes placed according to the high temperature device and this surface I'm allowed to modify.
We are going to apply a mica based gasket with thickness 1.0-1.3 mm between the two surfaces.
My plan is to consult the ASME B16.5 (2003) standard and machine a raised face around the holes according to paragraph 6.4.5.3. Height of the raised face is set to 0.4 mm.

I find the specifications in paragraph 6.4.5.3 not very well defined. Can anyone indicate if I'm on the right track or do you have a suggestion for a better solution?

Unfortunately, it's not possible to reveal all details, but please ask if something is missing and I'll try to answer.

2D of cross section of raised surface around hole (in millimeters):

 

[LittleInch]

Why aren't you using the 2009 edition? Not that it makes much difference for this clause, but you should use latest codes.

I think you've just about got it apart from the feature depth. 6.4.5.3 quotes 3.2 to 6.3 microns (0.0032 to 0.0063 mm) not the 0.02 you show above.

Not sure where your bolt holes are and you've clearly gicen us only a small fraction of the drawing so it all looks quite odd.

I was a bit alarmed at the "reduce" leakage bit so hope you haven't had a fire or explosion yet....

My motto: Learn something new every day

Also: There's usually a good reason why everyone does it that way

 

[LKH005]

2003 edition is what I currently have at hand, but I'll get the 2009 edition very soon. Thanks for pointing out.

I'm quite sure the feature depth is correct. As I see it, this is controlled by the other parameters (radius, number of grooves/mm). Ra specifies the average distance from the mean line to the theoretical surface. In this case the mean line must be at a depth of 0.01 mm and the average distance to the surface somewhere around 0.01/2 mm = 0.005 mm. Or am I getting something wrong?

Bolt holes are not shown because there aren't any, compression is maintained using a different principle. I can't show this because of IP rights, which is also the reason for the small section of the drawing.

Explosion is not possible, but a small flame can occur if gasket is severely damaged :-(

Thanks for input, it's appreciated!

 

[XL83NL]

As of 29 April 2013, there's also a 2013 edt. of B16.5.

 

[LittleInch]

Your feature depth works when you assume that the profile is a "wavy" one like you show. 16.5 is quite clear that the profile is actually "serrated" or small depressions with a surface roughness as detailed above (3 to 6 microns).

Buy an RF flange and have a look at the surface finish.

My motto: Learn something new every day

Also: There's usually a good reason why everyone does it that way

 

[LKH005]

I've now got hands on a commercial RF flange. It feels very similar to the sample I have had produced from the above shown drawing. The variation between the two seems to be mainly about the "pitch" or number of grooves/mm which is stated as 1.8-2.2 grooves/mm in the standard.

You're right that the standard says "serrated", but the commercial RF flange also feels more "wavy" than "serrated". You think that the 1.5 mm radius of the cutting tool stated in the standard refers to the radius of a ball nose cutting tool or to the radius of a V-shaped cutting tool with an angle defined by the remaining parameters (grooves/mm and roughness)?

Thanks for your feedback and have a nice weekend!

 

[LittleInch]

You would probably need to look at the difference under some serious magnification, but why don't you ask the supplier / manufacturer of your flange. I have no idea how they make the "serrations", only that the "wavy" approach just looked wrong.

Ultimately the waves or serrations are there to provide point impact on your gasket in many locations so that a single speck of dirt doesn't affect the sealing of the entire gasket, so in reality both will work providing your gasket is sufficiently flexible to be compressed in this manner.

My motto: Learn something new every day

Also: There's usually a good reason why everyone does it that way

 

[racookpe1978]

Kind of a basic question, but (obviously) you re concerned about propriqtary parts of this flange.

We are going to apply a mica based gasket with thickness 1.0-1.3 mm between the two surfaces.
My plan is to consult the ASME B16.5 (2003) standard and machine a raised face around the holes according to paragraph 6.4.5.3. Height of the raised face is set to 0.4 mm.

Are you worried about leakage "between the (several?) 10 mm holes" (between one 10 mm hole into an adjacent 10 mm hole) or are you worried about leakage between the "set" of 10 mm holes and the outside world?

If only the latter (leakage between the "set" of 10 mm holes (no matter how many) and the outside world, then run the serrated rings that enhance leakage protection only around the inside and outside 1/2 inch (10 mm or so) regions of the entire flange surface.

If the former (leakage between adjacent 10mm holes is a concern) then you could - as you stated - run circles around each hole. DON'T overlap circles or you will create leakage paths where the circles intersect.

Also! Don't fixate on "standard flanges" and "standard flange sizes" and "standard flange bolt holes and bolt hole circles" as a limit! You are far past "standard flanges" in this problem. Use the standards as a proven method that is only a foundation for a new solution.