Conern with machined flanges, possible stringers and defects 1

[mshimko]

I’ve a situation where C-steel flanges to ASTM A105 are required to be manufactured from individual forgings; however, we’ve discovered (AFTER the fact, of course) that one vendor did not properly read our procurement documents and as a result we have flanges that were machined from either forged bar stock or hot-rolled bar stock.

These flanges are installed in various locations in steam and lube oil systems, and range from ½ inch to 6 inch in size.

One of my material folks has concerns with defects (stringers, end-grains) as well as grain orientation that may be in a direction that has significant impact on the performance of these flanges in resisting fatigue, as well as sudden extreme loading. While I think of myself as pretty experienced and knowledgeable of metallic materials, I do not share this concern, but obviously do not want to and cannot simply ignore my colleague’s advice.

Does anyone have experience, knowledge, or leads regarding the use of machined C-steel flanges and the possible/likely impact of stringers/defects/end grain orientation?

Thanks in advance//mjs

 

[metengr]

mshimko;
I am not a self proclaimed expert, but have been around the block several times in the Power Generation sector. I agree with your position, and do not share your colleague's concern, as well.

 

[metengr]

Fogot to mention that ASTM permits the use of bar in lieu of a forging to manufacture fittings, etc., up to a certain size. See below for ASTM A 182

1. Scope
1.1 This specification covers forged low alloy and stainless
steel piping components for use in pressure systems. Included are flanges, fittings, valves, and similar parts to specified dimensions or to dimensional standards such as the ASME specifications that are referenced in Section 2.

1.2 For bars and products machined directly from bar, refer
to Specifications A 479/A 479M and A 739, for the similar
grades available in those specifications. Products made to this specification are limited to a maximum weight of 10 000 lb [4540 kg]. For larger products and products for other applications,refer to Specification A 336 for the similar grades available in that specification.

 

[grampi1]

I think the issue is how was the starting bar made? May have been made using good practices and there will not be a problem. Could have been made with less good practices. Some examples would be:

Low reduction ratio - May still have coarse grains, hydrogen cracks from casting, non-homogeneous composition and directional properties.
As rolled condition - Again coarse grains.

Probability is on your side to not have a problem and would pick battles carefully to only those where the consequence of failure is high such that it is worth the investigation time to determine how the bar was made.

Its fairly easy to take one of these valves from stock and have a metallurgical characterization done.

 

[mshimko]

Providing a little additional info (for grampi1 and others):

We've already conducted a little testing on bar stock material; we had an independent inspection lab UT a sample of bar stock (don't recall the acceptance criteria off the top of my head, but "it passed"); and we also took a completed blind flange and from this machined tensile specimens; all met yield and elongation requirements; ~half met UTS, and the others "just missed" UTS requirements (less than 5% under the min).

From this, one could seemingly make a case for acceptance; HOWEVER, my collegue's concern is: what if the two pieces we tested simply had no stringers, but the next piece did?

It's not feasible (in my opinion, at least) to test numerous samples from numerous venders until we either develop the statistical probablily of sizable defects, or find a sample that contains a defect.

Hence, I'm primarily seeking leads to info that may document actual cases to validate my colleque's concern; or alternatively, make a case to discount his concern.

 

[salmon2]

Right now I think your colleague is leading as you said you found half UTSs failed, which is unacceptable. To me it is a strong enough case to return them to supplier.

I think the statistics should be if you tested a decent amount of material, you can roughly assumed you will have the same results. Say you have 1000 flanges, you sampled 100 pcs randomly, 20 pcs are bad. Then you can assume 200 pcs of total 1000 are bad.

Tensile test is too tedious, you can do hardness. One time, we checked every single pc's hardness, thousands pcs. Breaking bundle and re-bundle, a lot of work. 100% UT test at supplier should have been done.

 

[mshimko]

For salmon:

The flanges we're questioning have already been welded into their applications, some have been in service for over 7 years. Any flanges not welded in place when the issue was identified were returned to the vendor.

The issue isn't if the provided products meet our specs, (they don't), it's if we need to take systems out of service to cut out and replace existing flanges.

Based on the mechanical test results only, these flanges are certainly "good enough" sinceour design standards have ample margine.

The issue is only of performance under sudden extreme loading (and to a lesser concern - fatigue).

 

[salmon2]

This is a tough situation. Somebody has to put his name on it.

How about the original MTRs? Were they good? By metengr, it is OK to use machined parts (I didn't check carefully) in lieu of forging. So the current issue has become why UTS not meeting Spec? And is it a big deal?

Design standards have different safety margins for different conditions, such as normal, extreme, accidental, etc. Those margains should not be challenged, in my opinion.

People know forged parts are stronger than machined one. But it is hard to quantify. And actually machined parts will tend to have less inclusion stringers compared with forging, IMO.