A-36 Cold Weather 2

[nich0142]

Don't know if this should be in the Structural Engineer or Materials section.

The project is a self supporting stack that would be behind a natural gas fired boiler.

We always use A-36 material but are aware of brittleness at low temperatures.

Is there a general "rule" when to switch to a cold weather steel? 516 Grade 70N / A572 / CSA 350WT

Our application is at -46C but the question always comes up at what temp should we consider using a different material rather than A-36.

Our customer specifies A-36 as the material but as the supplier of the stack we feel that we have an obligation to inform them of the issues of A-36.

If there is a brittle failure who is at fault us as the designer / supplier or the customer for failing to protect the stack from brittle impacts?


Appreciate your thoughts!

 

[JStephen]

There is an ASME steel stack standard. Does it address this?
The different tank and vessel codes work this in varying ways.
In some cases, cold temps are not considered where operating stresses are relatively low.
In some cases, base metals and weld procedures are qualified by impact testing at the design temperature.
In other cases, exempted based on material and thickness ranges in the standards and/or stress levels anticipated.
I would think generally, the customer is responsible for specifying the design temperature, while the designer is responsible for complying with that temperature. But that may be altered contractually or in the standards, and if the customer doesn't know the difference, that responsibility may land more on you.

 

[nich0142]

ASME STS - 1 Steel Stacks does not address this issue.

There is a requirement from the client to charpy test lifting lugs and weld procedures at -46C.

Correct the client specifies a minimum design temperature of -46C and A-36 materials.

It still comes down to 1) Is A-36 an acceptable material at this temperature? or at what cold temperature would you stray away from A-36.

I'm not an expert in this by any means but I would assume a brittle failure only occurs when there is an impact. If you use a hammer and hit an A-36 stack at -46C is there going to be a catastrophic failure and then who is at liability of a brittle failure. Does someone "bumping" into an A-36 stack at -46C result in a brittle failure?

 

[JStephen]

The brittle failure is not related to impact, but to how the material fails.

 

[Agent666]

It's about having toughness and ductility. At colder temperatures steel can fracture in a brittle way, rather than displaying ductility. The usual measure of this brittleness is the charpey v notch test. Basically how much energy it takes to fracture the steel at a given temperature. Material specifications should outline the minimum a given grade should comply with.

-46C is pretty cold, I don't know for certain but I'd imagine A36 specifications wouldn't go anywhere near that low.

 

[RPMG]

I believe -20F or -28C is the limit

 

[Geoff13]

A way of making an estimate would be ASME Sect VIII Div 1. In Figure UCS-66 there are exemption curves for different materials vs thickness and temperature (A36 is Curve A material). This implicitly assumes fully stressed material, but Figure UCS-66.1 provides a temperature reduction when the material is less than fully stressed. This might give you an estimate of what temperature A36 might be good for at whatever stress levels exist in your stack. Be careful as ASME's idea of "fully stressed" is much lower than most other Codes. While I suspect A36 is very unlikely to be suitable at -46°C, in the end it would depend on the thickness and stress.

The test is called an impact test but this does not mean only an impact load can cause brittle fracture. In general you need three things: cold, tensile stress and an initial flaw. The last one is always present, which leaves the first two as your variables, and you can only control one of them.

As for responsibilities, I would say the Owner's responsilbility finishes after specifying where they want the stack built. The designer should then be responsible to select design conditions appropriate for that location (wind, snow, seismic, temperature, etc).

 

[dhengr]

Nich0142:
Almost any steel can be ordered to a (some) Charpy V-Notch energy value, at a given temp., within reason, and your steel supplier can tell you if they can meet that or not. You will pay some premium for this mill cert. process. They can then suggest a steel that will work for your design and temp. needs, assuming they have one. But, every bit as important to the low temp. ductility of the steel, at some point, is the design, detailing and welding procedures used on the project. You can turn a good piece of steel into a piece of brittle junk if your design, details and welding introduce stress raisers (notches, reentrant corners, craters, arc strikes and the like) and high design stresses or residual stresses. The steel does not shatter like glass, of its own volition. But, of course, such things as highly restrained joints and details, which are highly stressed under normal conditions, and then maybe stressed further because of contraction forces, which can be very high in a highly restrained structure, can be the straw which breaks the camel’s back. If this is an exhaust stack, what is its normal operating temp. on the steel surface, isn’t it higher then -46°C?

 

[nich0142]

Thanks for the very detailed replies!

Dhengr- normal operating temperature of the stack is 400F. Of course the -46 C will only occur when the unit is not operating and during an extremely cold day.

The job site is near Edmonton Canada.

 

[JStephen]

One thing that might help is to consider when and how the stresses are applied. If they are due mainly to wind, it's reasonable to assume that your coldest temperature is also subject to high winds. But if the maximum stresses are due to some thermal effects so that maximum stress and minimum temperature just can't coincide, you can take that into account also.
For steel plate, there is a modified manganese version of A36 ("A-36 Mod")which is allow to be used at considerably lower temperatures than straight A36, and a good bit of the A36 sold meets that regardless. It's probably still not good for -46C, but would come closer than straight A36.
Also consider A516 Grade 60 as an option.

 

[r13]

nich0142,

This thread might interest you,

https://www.eng-tips.com/viewthread.cfm?qid=184793

 

[dik]

Your project is in Canada... why not use Canadian steels? Check into G40.21-50W, Type T.

From elsewhere:

TYPE T – Weldable, low temperature steel. A superior quality, weldable structural steel, with chemical composition that makes it suitable for low temperature applications. Possesses good notch toughness in the “as rolled” condition. Used for bridge construction for all types of loading. Certified minimum impact test requirement available. Easily weldable, using good shop or field practices, by all the usual methods.

A36 is available, but, not common.

Dik

 

[r13]

The responsibility is squarely falls on whoever specified the material for the project. As a supplier, it is appreciated if you can point out the potential short falls, but its not your duty, besides ethics and relationship with the client/customer.

 

[dik]

"The responsibility is squarely falls on whoever specified the material for the project."

I would suggest that you still may be liable if you use an inappropriate material when you should have known better.

Dik

 

[Ron]

I agree with dik...you might still be responsible, since you are likely more knowledgeable than your client with regard to material limitations.

 

[dik]

Thanks, guy...

Dik

 

[r13]

I don't think this case involves inappropriate material, but the question on the brittleness concerns of A36 steel in sub-cold applications. Since this likely a powerplant project, I think the steel was specified by an engineering professional, the customer then ordered the specified steel, and revealed the intend use, including temperature range, in the manner of a casual business talking. I doubt the customer had specifically requested the supplier to offer his opinion, as there was a designer/specifier by his back. For the foregoing premises, below is my defense on what I've said in comment above, regarding who is liable shall failure occurs.

Not if the supplier provide defective material, inferior or different material other than that asked by the customer, but think the replacement is specified equivalent without telling the customer. The supplier may have more knowledge and experience than the customer, and/or the specifier (an engineer), but its not his legal obligation to provide those information, except moral concern. Unless the customer came to store and tell the supplier the intent of use, and ASKED for suggestion and advice. In the latter case, yes, the supplier maybe liable if he's not offered what he SHALL knew, or the information that is considered common knowledge in his trade. But not all the suppliers have the complete knowledge on the goods he carries in the shop, nor they MUST to, just like the OP is uncertain/suspicious about the suitability of A36 for sub-cold application (he's lacking the specific knowledge on this), and seeking help from those who he thought are professionals, this is an extra, but not legally necessary, effort to help himself to gain knowledge, and maybe pass the knowledge onto his customer in a more convincing manner (against the designer/specifier). Do you really think he is liable? Or you think he is obligated/legally bounded to find out the information before he's allowed to sell the steel as requested by the customer, who in turn was backed by the designer/specifier? I don't think so. But I do agree that at the end of the day, the lawyers always win by simply put everybody on the stand. Good news for the engineer/specifier who practice recklessly, you won't be left alone when stand in trail, there will be the supplier stand next to you - a prevalent sentiment here.

 

[nich0142]

Thanks all for chiming in. I've never been to this site and great support.

Our customer, a boiler manufacture, sold an "A-36" exhaust stack as part of their package to their customer, a Canadian petrochemical company. This boiler supplier offered a higher grade material for the stack but that option was not selected from their customer therefore A-36 was put in the purchase RFQ.

As the designer / supplier of the exhaust stack we of course need to engineer and supply a quality solution that will match the environment (wind / seismic / ambient conditions, etc) specified regardless of the material stated by our customer.

Thanks again for the input.

 

[r13]

nich,

Thanks for telling us more on the story. I wish you have found useful information from guys in here. Lastly, your customer friend is liable, as he must have the knowledge as a boiler maker/supplier, or even been the designer, if anything went wrong.

 

[HS_PA_EIT]

From ASME STS-1 2016 Section 2.2.2 SHELL AND BASE PLATES (c) -
"Carbon steels such as ASTM A516, Grades 55 through 70 and low-alloy steels such as ASTM A517, Grades A through T and ASTM A537 are usually specified for service temperatures as low as −50°F (−46°C). Nickel-containing alloy steels such as ASTM A203, Grades A and B are usually used for service temperatures as low as −75°F (−59°C), and ASTM Grades D, E, and F are often used for service temperatures of −150°F(−101°C). Nickel-containing alloy steels and nickel stainless steels are used for even lower temperatures. Suppliers of structural quality steels will provide data on notch toughness when specified."