Railing System with Stainless steel

[McGill10]

Hello,
My question is on design of railing systems with stainless steel.
What would be the yield strength (fy) of stainless steel alloy 304, in practice. Railing Supllier says they have been using 50 -60 ksi and has material testing report to verify. It seems quite a high compared to 25-30 ksi per ASTM standard A554.
Shall we design per material report? Want an authentic input, this being a big difference on qunatity and hence cost.
Really appreciate your thoughts and experience. -McGil

 

[Enable]

You go based on the standard, ALWAYS. If they are using material made to a standard with a higher min Fy, you can use that obviously. However, you cannot take coupon testing or other tests taken only at select locations to be representative of what's going be in service. A number of rather famous failures have happened due to this.

 

[phamENG]

Depends, really. If this is a one off and you have verifiable material strengths from a testing lab that you trust taken in an acceptable manner (quantity, location, etc.) and the difference in cost is going to be substantial, then designing for the actual strength could be an acceptable option. But if this is a "typical detail" that is going to be used frequently for multiple projects or it's the difference between a 3/16" wall and a 1/4" wall on 8 posts, I would use the ASTM minimum for the steel provided. There's no guarantee they'll keep getting steel of that quality/strength from their supplier, so I wouldn't risk exposing myself on their behalf.

 

[phamENG]

Enable takes a harder line on this than I do. I agree random coupon testing with inappropriately small sample sizes would be a no go, but I'm curious about some of the famous failures you mention? Did they have rigorous testing to support it or was it just a gut feel/guess?

 

[phamENG]

It just occurred to me that your "50-60ksi" material is probably the tensile strength. In other words, your railing supplier probably doesn't know the difference between yield and tensile strength and the importance of that difference in design. I withdraw my comment about it maybe being okay. Don't. Follow Enable's advice and only use the standard minimum for your design.

Enable - I'm still interested in your viewpoint and the failures you mentioned.

 

[jjl317]

FWIW - I design in 304 stainless almost everyday, and use for 30 ksi for yield strength, but we don't always know who the manufacturer will be when we do our work (custom cladding systems). We have had a few suppliers claim that their material had a higher yield (though I don't remember ever seeing a claim beyond 45 ksi). And potentially the confusion could be yield vs tensile, as the tensile is so much higher.

For a manufactured product with the appropriate documentation, I would probably consider using a higher yield strength.

 

[McGill10]

I did doulbe check on that to confirm it is the Yield Strenght (Fy), and it is yield. The report also has Fu (about 100 ksi)
My concern is tracking the product, who and how accurately they do if we use higher yield based on test.And I belief also they won't test every prodcut. This report is based on last year products test.
They say put the fy value you used on the calculation and drawings and we will take care the rest.How much we as a professional need to involve to make sure the right grade is used? Who will confirm the products meets the listed value.

 

[JStephen]

You don't say if the material is actually A554. If so, then go per spec as noted above.
But I think some of the stainless materials can be furnished in cold-formed/cold drawn variations where the properties are higher than the basic hot-rolled properties.

 

[EdStainless]

A few points, Some (a lot) of the SS supplied for A554 types of use are not annealed.
The range of A554 is large;
7. Condition
7.1 The tubes shall be furnished in any of the following
conditions as specified:
7.1.1 As welded,
7.1.2 Welded and annealed,
7.1.3 Cold reduced,
7.1.4 Cold reduced and annealed.
7.2 The inside diameter bead shall be furnished in any of the
following conditions as specified:
7.2.1 Bead not removed,
7.2.2 Bead controlled to 0.005 in. (0.13 mm) or 15 % of the
specified wall thickness, whichever is greater, and
7.2.3 Bead removed.
7.3 Square and rectangular welded stainless tubing is supplied as cold worked unless otherwise specified.

The flat rolled sheet was annealed, the give a light sizing pass on a rolling mill.
After that it was slit, formed into a tube, and welded. A higher strength is common in this material.
In A554 tensile testing and NDT are supplementary requirements and not part of the base spec.
Even in fully annealed 304L we rarely saw yield strength below 45ksi.
UTS was commonly 65-75ksi.
In the old days when L grades were first being made and these strength levels were established the alloys were different.
The Ni content was higher (which makes them softer).
And they were not made via AOD with residual nitrogen left in the alloy (which makes it stronger).

= = = = = = = = = = = = = = = = = = = =
P.E. Metallurgy, consulting work welcomed

 

[dik]

Do they have a SS version of ASTM A500 for carbon steel?

When doing calcs with A500 involved, I reduce the Section properties to match the minimum dimensions of A500.

Rather than think climate change and the corona virus as science, think of it as the wrath of God. Feel any better?

-Dik

 

[Enable]

@phamENG

I say that for pretty much the same reasons that we are not allowed to rely upon Mill certifications when they indicate a greater yield than the design minimum: the factors involved in limit states design already take into account the fact that the actual Fy is likely to be greater than the minimum Fy. So even if you could be assured that testing revealed the actual Fy across the entire cross-section of each piece (and you likely will not get that - see attached document) you would still be reducing your safety factors by using the actual vs minimum.

I know you do a fair bit of work with older structures so you are accustomed to taking coupons to figure strengths for analysis. However, even there we don't actual use the obtained Fy for analysis. We use the Fy obtained from the coupons and back-calculate what grade of steel we are likely dealing with for that era and use the minimum specified for that grade! If we must, we use the obtained Fy from the coupons but discount it to account to A) ignorance and B) the fact the code is based on minimum with probability of exceedance baked in, not actual.

I've attached my old course notes that touch on this, which includes commentary in S16, and the Canadian bridge code.

One particularly bad failure was the Station Square building in Burnaby BC. An independent structural review firm found inconsistencies in the design, however they received mill certs for the actual beams installed (not of the group but of the actual ones that where the design was lacking) and found the Fy was 20% greater than the design Fy. They concluded that the structure was therefore safe using the new Fy.

In March 1988 the MSS Group was hired by the tenant, Save-On-Foods, to conduct a structural review of the entire project due to concerns about deflections of roof beams above the store. MSS found two beams to have "insufficient capacity" and made recommendations for correction. Both engineering firms, the architects, developer, and contractor took steps to initiate this action. The following day Tamm Tacy advised that this remedial work would not be necessary due to the steel mill who had fabricated the beam provided documentation indicating the actual capacity of the beam was significantly greater than documentation used during design. MSS Group advised that the beam was therefore "satisfactory to resist the design loadings imposed" and "this information is a blessing for all concerned".

Roof collapsed at 9:15am of opening day.

That same collapse led to significant changes in Gerber steel design.

 

[JLNJ]

From AISC Steel Design Guide 27, 304 material has an Fy of 30 ksi and an Fu of 75 ksi. That's a lot of spare capacity from yield to ultimate. (AISC has recently published a new stainless spec, but I have yet to look at it.)

It seems to me that we are extremely conservative in our guardrail post and rail designs. Truth be told, we probably spend too much time worrying about the posts and not enough time considering the post anchorages.

Once upon a time I felt perfectly fine using a 200# load and a one-third allowable stress increase for a temporary load. (Makes sense since the 200# came from ANSI/OSHA as a service load.) Now it's been defined as a Live Load with all the uncertainty and load factors which go with a Live Load. Methinks we've taken it too far.

 

[phamENG]

Enable - thanks for that. Makes sense. On the rare occasions that I have used coupon testing (in most cases it's an assumption of strength based on the time period, and testing is required only if the DCR reaches a certain threshold or welding is required) I do take care in where the coupon is taken from the beam. Residual stresses are fairly well understood and can be considered in the evaluation of the results. And I still use an additional safety factor based on the sample size. But it is a good reminder of the importance of making sure the entire system is understood. Engineering isn't just about plugging numbers into formulas - it's about understanding what and how we're designing from the time the material leaves the earth (or recycle plant) to the time the last person walks out and the building is taken down. There are a lot of factors at play, and we ignore/misunderstand them to our peril.

I wish I had time to find the report for that collapse and read it. Not to quibble, but it sounds more like a stability failure to me than a strength/yielding failure based on the rudimentary description in the Wikipedia article. I've seen tons of Gerber systems with similar flaws - continuous beam over column connections with no bracing at the column, not joist aligning with the column, no web stiffeners, etc. None of them have had problems, but they are also just roof structures, most in low to only moderate snow regions. I couldn't imagine using that for a parking deck. Yikes. I'll add it to my reading list...perhaps I'll get to it before I retire...

 

[McGill10]

Thanks all for your input and very helpful discussion. With so much input and other literature review ,I feel better on using material test data. The following link is very helpful:

http://www.steel-stainless.org/media/1036/english.pdf