Refer back to API 650 ( Welded Steel Tanks for Oil Storages ) ; 10th edition with Addendum 3 Para 5.2.1.4 says about Acceptance Limit of Weld related Undercuts. It says .." For Vertical Joints, the maximum acceptable undercutting is 0.4 mm ( 1/64 inch.) of the base metal. For Horizontal butt joints, Undercutting not exceeding 0.8 mm ( 1/32 inch.) in depth is acceptable..." Seeking forum expertise to response that why this acceptance depth level differences between vertical and horizontal joints ? regards DM Mob- 94342 36119

Scott In reply to Debasis's question, my views are as follows. Hoop stress is more or less the same. Hoop stress does not vary much. It is related to the maximum length of weld and the forming of plate within the code permitted tolerance of mis-match for weld joints For Ex: The maximum width any steel plate manufacturer makes is 4 M wide. The same steel plate manufacturer can supply plate of length varying from 3 M to even as high as 10 M today. Steel plate manufacturers and major large diameter construction companies prefer to keep the number of joints to the minimum. At the same time no steel manufacturer produces a plate of width more than 4 M today due to limitations of space in steel roll mills. At site, while forming and matching weld joints - for large diameter storage tanks meeting to API 650 - both long seams and circ. seams - the long seam being of maximum width or height of 1.5 M to 2.5 M can be matched more easily with minimum mis-match within code permitted tolerance of 1/16". It is possible to make longitudinal weld joints with minimum mis-match at site by bridging support tack welds till the weld joint is completed - both from inside and outside even for large diamter storage tanks. For Circ. seams - with plate length varying from 3 M to 8 M for the next Long seam joint - the mis-match in weld joint cannot be so tightly controlled to keep within the code permitted 1/16" even good bracing supports and tight forming tolernace to suit to the over-all ovality and concentricity of the tank diameter and heights. Due to this, the tolerance in horizontal circumferential joint under cut depth has to be higher than given for longitudinal weld joints. To my mind, Hoop stress is the same whether it is Circumferential or Longitudinal joint . Small variations in hoop stress really do not matter. It is the mis-match tolerance when you form a long length circumference weld joint and the difficulties to keep the control within the mis-match in plate forming and site weld joints plays a part in giving this extra tolerance in undercut depth for horizontal weld joints in any storage tank construcion. Trust this is of help to you C.V.Srinivasan Nishi Engineers Pvt Ltd India Jan 04m, 2006 E-mail:nishi@vsnl.com >I would have to say because the long seams see a higher hoop >stress created by the static head in the tank. > >Scott123

Debasis A discussion forum -i guess - is meant for sharing of practical and problem solving experiences of common interest for users / engineers / designers / inspection / corrosion / material related topics. Idea of brain storing is OK. A dicussion forum may not be the right place. A separate training session or refresher course with active direct participation would be ideal for this than an international discussion forum - where basics of engineering fundamentals should not be discussed by practising engineers / managers. For this , a separate brain storming session at your country of origin would be ideal and can disuss based on common sharing of knowledge on individual experience and expertise with multi-various engineering hands-on expertise. Giving a little clarification on your point - to the extent i know : Weld under-cut or mis-match in fabrication in a large tank are both considered as point loading / considered as local stress raisers. Mostly the point load stress would be tensile in nature. Tensile stress is likely to affect more or likely to initiate problems in service with the service produce stored and the contaminants present in the stored product. Drawing an analogy between a forged component and a fabricated steel rolled component (rolled steel plate in this case) either for vertical or horizontal joints may not be appropriate. Code practices are evolved after review of practices, difficutlies experienced and with constant upgradation from industry engineering practioners and experts drawn all over the world. Tolerances are arrived based on the feed back and the best that would not affect the integrity of the structure or joint degradation over a period of time. For the same steel plate, construction code and fabrication practices and weld joint acceptable or rejection limit for defect toerances may differ. For Example: a "T" joint bewteen horizontal and vertical seams will behave adversely or peculiarly with tri-axial stresses compounding even with the code allowed under-cuts and allowed mis-match. The same code practising engineers would give the least tolerance between the vertical and horizontal weld joint - taking into account the tri-axial stress role at the "T' joint and the adverse impact on crack propagation for a given service duty or in code accepted practices to recognize this and the tolerance to which this has to be accepted or rejected. Trust this is of help to you C.V.Srinivasan Nishi Engineers Pvt Ltd India Jan 5, 2006 E-mail:nishi@vsnl.com >Thanks Mr. Srinivashan, > >You have considered mismatch as a function of weld undercut. >Now if I treat this depth of undercut values w.r.t. Tank >self body load factor - for the vertical welds ( long seam ) >the load vector is parallel to the weld axis and for the >horizontal welds ( circumferential welds) the load vector is >perpendicular to the weld axis. Now as in forgings we >consider load vector perpendicular to the fibre direction is >the best one. Now , if we treat here similarly , then the >horizontal welds are more comfortable compared to vertical >welds. Accordingly a higher defect ( undercut depth) >tolerance is permissible in horizontal welds. > >This is just my brainstorming and may be inappropriate. > >regards and for further discussion s please.