ASME Flange Thickness Tolerance - Standards vs Reality

[TEIGRAY]

Hi Eng-tips community,

This is my first post here so hopefully it's nice and concise.

I'm currently doing some R&D for a couple of new products, one of them being a tool to measure the thickness' of ASME Flanges on site,
The thickness values (c) in the ASME 16.5 standards are taken as the minimum and usually to 0.1mm.
I know the ASME 16.5 standard also states that the thickness of flanges can have a tolerance of -0/+3.2mm, this seems very high, basically rendering such a tool non-practical, as there is sometimes only a 1.6mm difference between these flange thicknesses.

My question is, without going through and physically measuring batches, is this tolerance stated in the standard actually regularly translated into reality? Since I'm guessing these tolerances in 16.5 were taken when manufacturing techniques were much less advanced than today. Surely its in the interest of many manufacturers to get their tolerances lower than those in the standard (saving on material costs etc).

Usually I'd just drop the project, but having spent a lot of time on it and this issue of variation of flange thickness for raised face ASME flanges also affects other R&D projects where F-F space is an issue (so a potential F-F tolerance of 6.4mm is not ideal).

Any clarification on the feasibility and actual current manufacturing tolerances for ASME/PN flanges would be much appreciated.

 

[XL83NL]

I had the same question a few years ago, for a situation where I wanted to define standard bolting lengths for B16.5 flanges (since the B16.5 bolt length typically is too short for most cases).

The flange thickness impacts the bolt length, especially with the high tolerance from B16.5 you mentioned. What I did was measure a few dozen flanges for the typical range we use (1/2"-8") and check how much they were off. I dont recall the outcome, but the exercise was short, simple and useful, and gave me the info I needed. Maybe you can do this as well.

 

[SnTMan]

I'm missing something. I thought your tool was to measure the flange thickness.

The problem with sloppy work is that the supply FAR EXCEEDS the demand

 

[TEIGRAY]

The tool is essentially the different ASME flange thickness' as a cut out from a steel plate, you'd fit it over the flange to the correct section, combined with the Nominal bore you'd be able to quickly find the flange class from these two variables, which are etched into the tool itself.

Cheers Xl83NL! We came to the same conclusion, I'm going to take some calipers out on site for my next few surveys and note them down.
Still its strange that manufacturers and suppliers don't include these in their quality checks, maybe they would just rather not release such info.

SnTMan, I agree, seems to me any manufacturer would want to keep their material costs down for items produced in large volumes like this, so would most likely work to a lower tolerance, I'm guessing the standards tolerances were from when manufacturing techniques were not as good as they are now, and have been kept to cover manufacturers liability.

The -0/+3.2mm would make these sections overlap in size, as some are only around 2mm difference.. Here's hoping the actual on site checks will produce a tolerance lower than this and make the project actually viable.

 

[EdStainless]

The sizes are supposed to overlap, so that you can use the next thicker one as a substitute.
The actual tol largely depends on the size of the flange and the fabrication method.
When you have a large number of forged blanks which costs more, having them all a little too thick or having a few of them scraped for being too thin?

= = = = = = = = = = = = = = = = = = = =
P.E. Metallurgy, Plymouth Tube

 

[SnTMan]

I don't recall I've ever seen a B16.5 flange that wasn't identified w/ size, class and material. In fact is required by B16.5. In addition each size and class is available in a number of different bores.

I'm still missing something....

Regards,

Mike

The problem with sloppy work is that the supply FAR EXCEEDS the demand

 

[r6155]

See page vii of B16.5-2013
“Bolt length calculations based on worst case tolerances led to a revision of tabulated lengths”.

Regards
r6155

 

[FoxRox]

If you are indeed making flanges from plate, then it is to your benefit to leave them on the high end of the tolerance. Machining takes time and carbide.

Yes, machining techniques are more advanced than they used to be. I could (and would) leave the design thickness at +3mm over spec, and my machine shop would never be out of the +3.2mm tolerance. .0079" is child's play to those guys.

 

[bcd]

The tolerances are wide because valves use to be in the scope of B16.5 Now the valves are out in their own standard, but still refer to 16.5 for flange dimensions. Since many flange end valves are castings, they need a more generous tolerance. Many valve flanges are not machined on the back side (nut bearing surface). So a wide tolerance is necessary. Fully machined pipeline flanges will not require as generous a tolerance. For fully machined flanges, the resulting tolerance will be within the manufacturing process capability.