Welded Strength of 5083-H321

I'm working through updating a foundation design, and I've noticed the previous design, that was accepted by NAVSEA, uses an as welded yield strength of 165 MPa.

Now from my experience, and from numerous sources I have read up on today I find that one should use the annealed strength of the material once it has been heat affected. In this case the yield strength of 5083-H0 is 145 MPa.

Could anyone refer me to a source for justification of using this higher value for yield strength of 5083-H321?

 

ClarkT

Firstly i would question why H321??...it is horrible. It must come with an ASSET cert to prove it is marine grade and secondly you wont be able to bend it, well, very easily, and in doing so you'll introduce small flaw/cracks which will cause problems in service.

You first need to address what the design is for and what factors of safety and the environmental conditions are, before you can really select a "design allowable" stress. Whatever is the published data,, if you do not make allowance for these factors, you'll have bits of metal in front of it in a short period of time.

Back to your question.

5083-O is nominally 120-125MPa.



Since this is a non-strain hardened allow and the temper is O, the un-welded and as-welded strength is the same. You also need to be careful which filler wires you use, some are a tad more brittle (in terms of ease of fabrication) than others yet give higher strengths.

 

We agree. The H321 is specified by the client. Not mine to argue. It's a very boxy shape with minimal shaping, so I'm not worried about bending.

However, I've found several sources, including the prior calculations reviewed and approved by NAVSEA that do not use the O temper yield strength as both you and I expect. Rather they use an as-welded yield strength of 165 MPa. Heck, this Reference Attached goes as high as 172 MPa.

I'm stumped by it and would feel a lot better about it if I had an authorative explanation for these values.

 

Never mind the welded strength. Have you read some of the reports about this stuff cracking, whole boat hulls being scraped? Perhaps you are right that their published values are not safe to use and that the material will fatigue at lower values prematurely.

 

I find this discussion interesting. A previous company I worked had a fair bit of experience in aluminum design. They always specified 5083-H321 for stiffeners such as built T-frames and yes they used a yield stress of 165 Mpa in the welded condition. I am no metallurgist so I trusted their numbers. Just for reference they used 5083-H111 for plating, and 6061-T6 for extrusion stiffeners like OA or bulb flats.

 

Not Marine Grade

Ok I couldn't find the original article but the link below is probably better to describe what I had read before. Basically it is not a suitable marine grade material and quite unlike 5083-0 from that side of things.
At the time I was looking into an alternative to the pricey Alustar, is this not an option for your client if he wants a higher performance alloy? I recall 20% stronger than 5083 was the marketing.

http://www.professionalmariner.com/...91&tier=4&id=A8E43FA7CD2746F8AACD1DCF82781F6A

 

Clark.

The values you posted are these for welded or un-welded?

The ref’s I attached are from DNV and LR; figures to trust, as these used worldwide in High Speed aluminium design/build. I also have many others from other Mills etc. But the values posted by DNV are consistent with others I have found. Once you weld a strain harden alloy, if you’re using the same filler wire you’ll go back to its annealed O temper condition. There is no rocket science there. This is also reflected in table J2 I posted, from DNV.

I have occasionally come across rogue values like 170MPa etc…I never trust these.

If you wish to have higher as-welded strength when using H321, then select 1) filler wire 5183 and 2) the 5383 alloy. But not going for a higher priced alloy.

The Alustar alloy that came out in the mid 90s, now called 5059 alloy does quote 160 MPa as an as-welded strength in the H321 alloy. But if I recall correctly this is much more expensive than 5083 and more than the alternative 5383.

David,

You wont be able to “extrude” H321 into a “T” shape. Unless you have a seriously major set of hydraulics in the press and a state of the art QA system, ie very very expensive and renders the product cost prohibitive to begin with. So I assume you mean a web and rider bar to form a fabricated Tee. The temper makes its next to impossible to extrude into a suitable and crack free end product.

Again, I would not trust any value above 125 MPa, unless the person/supplier can provide you with exact details of the alloy and filler wired used. Also make sure the H321 comes with a proper ASSET 66 & 67 cert.

 

In the part you quoted me on Ad Hoc I said "built T" which is the terminology I'm familiar with. I believe I meant the same thing you called "fabricated tee". 5083 was never specified for extrusions, as I said in my previous post they specified 6061 for those.

 

Quite right, I stand corrected. There are several definitions/ways of saying the same thing.
I should read more carefully