See pages 3-52 and 3-53. The standard is based upon SI units with a cross reference to approximate imperial units. Interpret the standard using 25 mm and accepting that it uses an approximate inch conversion. Perhaps this accommodates round numbers in imperial units, thus could be helpful to those who lack SI familiarity.
For example today we can buy soft drings in two-liter bottles in the USA. Oil is sold in quart containers. A quart is one fourth of a US gallon, not to be confused with a British gallon. Still I tell people to visualize a quart of oil to visualize a liter even though this lacks precision. The standard also lacks precision regarding temperature and pressure.
I'm not quite sure what you would expect. I understand that the correct answer is exactly 25.4mm, but what is 4 tenths of a mm? Not much. A mm is less than 1/16 inch, usual steel tolerance. It is close enough.
However, those tenths add up. Over the length of a foot, this is about a 5mm difference, about 3/16". It is up to us to convert precisely.
In API RP2a, have I not seen API use 50mm for 2 inches? In offshore construction, that is definitely close enough.
JLSeagull,
You ask "How many decimals would you prefer". That's an interesting question. With the Computers being used nowadays it's all to easy to get results with 3 to 4 decimal places (or more) and unfortunately there are younger engineers who think this accuracy is warranted. In the "olde" days calculations were not as accurate and several decimal places of accuracy were not provided (especially when slide rules were used) but plants were made and did not fall down (very often). It's now become the norm to provide displacement results to 3 decimals of a millimetre or three decimals of an inch which is ridiculous as confirmation of these displacements in the field is not possible without extensive surveying equipment. We all need to get back to reality as regards the required accuracy of computer results.
I am astounded at these responses. The USA is not the only place in the World where API has been applied in refineries. I would venture to say that there are more refineries in metric based countries than there are in inch based countries, so a big welcome was given when API indicated it had metric dimensions only to find that it gave big problems. Imagine a tank builder using metric dimensions from API 650 for a 30" Nozzle and using the bolt circle diameter of 900mm and I build a piece of equipment to fasten to this flange also using API but I use 36" the bolt circle will be different to the metric by 14.4mm (over 1/2")and the bolts will not fit! Both built to API650.... It is not necessary to have anymore decimal places, it is 25.4mm/inch and even a slide rule would have been reasonably accurate. When I was 7 years old we had 12" rules with metric on and at that time the UK was still using inch sizes as standard. I hope young engineers are being taught the importance of the relevance of decimal places, whether they are in 0.001" or mm or +/- 1" Perhaps it is good that API is not applied to the space program!
This problem was pointed out to API over 10 years ago (10th Edition) and so we are now on 11th Edition Add 2. So my question should more be why is API continuing to use 25mm to the inch.
While I agree totally with Bentlyman, API states that either SI or Imperial units should be used, not both.
This Standard has requirements given in two alternate systems of units. The Manufacturer shall comply with either:
1. all of the requirements given in this Standard in SI units, or
2. all of the requirements given in this Standard in US Customary units.
Thank you IFRs and you are correct, the problem comes when everybody thinks they are right and dont tell other parties which system they are using. We always base on inch size translations and we are happy to use 'soft' translations, but we do like to be told. The problem would go away if API removed the soft translation and used 25.4mm/". It is insulting to think that engineers involved in building these plants could not handle the correct translation.
Compare the content of most API standards with the typical ASME standards. API provides broad guidelines. API 650 includes metal thickness etc. but this is still not as precise as the machining requirements within ASME such as flange dimensions in B16.5 etc. If a standard applies to nominal dimensional concepts the degree of precision is less important than standards for manufacturing.
I would not have noticed if the dimension reflected 1-inch (25.4 mm). I would have noticed if the dimension reflected 25 mm (0.98425 inch).
