Steel Box Section Made of Different Steel Grade (Eurocode design)

[neukcm]

Good day Everyone!

In Eurocode steel design code, is there any provision regarding the design of single member composed of different steel grades? I'm talking about a box section wherein top plate is grade S275 while that of side plate is grade S355 (see attached image).

Thanks in advance

https://res.cloudinary.com/engineering-com/image/upload/v1445561739/tips/DIIFERENT_GRADE_001_jlgw2f.jpg]

 

[dhengr]

Neukcm:
That is a common fabrication for a box section beam, etc., so there must be provisions for that arrangement. But, I think you will find that you want the stronger steel in the flanges, not in the webs. What are the dimensions and thicknesses, and span length? How is it loaded? These things might dictate how it is welded and its assembly sequence.

 

[neukcm]

Hi dhengr,

Thank you for that immediate respond.

Ok, the situation is this, we have installed a 40m (unsupported length) cross beam that will serve as a runway for gantry crane passing a tunnel. Since the beam (box type) is so critical for its purpose, they conducted a random test and found out that the steel grade used for side plate differs from that of top/bottom plate (as shown in the image).

Now, the entire set up is in question and the client want to see a structural design based on the actual material properties installed on site. Initial design uses S275 fyi. Is it using average yield stress, any idea?

This is actually the report overview;

Side Plate
Grade S460
Yield Stress (Lab results) 430Mpa

Top/Bot Plate
Grade S275
Yield Stress (Lab results) 235Mpa

*sorry for the image, S355 should be S460.

Thank you!
Neukcm

 

[neukcm]

Please find attached file for the Complete cross beam details including sections and elevations.

Thanks
Neukcm

 

[hokie66]

This should be referred back to the original designer, who I assume is up to date with Eurocode provisions. But if the whole thing was supposed to be of the lower grade, and for some reason, the higher grade was used for the webs, I doubt you have a problem. Both grades are common structural grades, and there should be nothing special about the welding requirements.

 

[dhengr]

Neukcm:
What is this thing, and how does it balance on 4 wheels on that 65mm sq. bar (rail), or are there two box beams and some interconnection btwn. the upper moving system halves? How much does it weigh, loaded and empty? You certainly have an original design for this box beam, don’t you? On this important a structural element, you must have more detailed fab drawings than the general arrangement that you have now shown. And, you should certainly have some traceability of the materials you used to fab this beam. Did you design, detail and fab this beam, or who did and what control did you have? If the original design was done with S275 stl., the fact that you ended up using some higher strength steel shouldn’t normally cause any problems, unless the higher strength steel is unweldable, poor impact values, or some such. You have to have all these things in proper order in your design file on this system, or your design is suspect. And, now that your customer has found some funny stl. where he would normal not expect it, you have all the more proving to do to regain his confidence. I certainly have plenty of other questions.

 

[neukcm]

hokie66,

Initial design uses S275 for web and flanges and design was acceptable but the random material test results shows much lower yield stress value of 235Mpa only. Even if the material used in the web is of Grade S460 with test result of 430Mpa, still there is a lower yield stress present in the member which of course need to consider. I know structurally the member is safe (i think so), but the calculation is inconsistent with what is actually on site and that alone leaves a big question to the client.

Now that the setup was completely installed, the only way to lead the client into a close perspective of assurance is a structural design calculation based on the actual yield stresses. That leads to question, what stress do we need to consider for both flange and web? If using average stress, any research that explains how to use average yield stress? Any idea?

Thank you,
Neukcm

 

[hokie66]

As I said, you need to go back to the design engineer. But you also need to go back to the steel supplier if the flange steel is weaker than specified. Where did the steel come from? There are pitfalls in using steel from the cheapest supplier, if that is what happened.

 

[neukcm]

dhengr,

I too have a lot of questions, this is actually designed by my colleague and he does not have full control of the fabrication. Yes, this is a two box beam that serve as a runway for the crane to pass over a tunnel.

The design system is in order, i can say that, the detailing shown in the drawing was accepted by the client with no comments but it is in the fabrication that we gain shortcomings. The client is not actually upset, they just want one thing from us - verify that the existing structure is safe.

neukcm

 

[kingnero]

I disagree that a higher grade in a construction is generally fine if it was designed using lower grade values.
S355 and S460 require preheat and low hydrogen filler material when welding due to it's (much) higher carbon content than S235/S275.
They also have other fatique properties, although the difference is not that much.
Most likely the welds will not be done correctly and will need attention.

If this is a recent (after 2013) constructed design, the requirements of the EN1090 should also have been taken into account, meaning (amongst others) that material certificates of all used steel should be available. This might not be applicable depending on the countr(ies) of fabrication and destiny.

 

[dhengr]

Neukcm:
If the original design is in proper order, you should know (be able to find, in the calcs.) what the critical stresses are in those box beams, for all of the critical locations of the crane wheels. Those stresses better be lower than the mechanical properties of the materials which the girder is made of, by some factor of safety, or you’ve got problems. Of course, welding design and welding procedures should be compatible with the materials actually used. I would do some more testing of the actual materials in place, to be sure that there is some consistency btwn. the various plates in each structural element btwn. the various splices. Do both mechanical and chemical testing to see if you can really pin down the material spec. Also, talk with your stl. suppliers and the fab’er. for the paperwork on the materials used. This whole thing is sounding more and more sketchy to me. Are you an engineer, so that you can actually interpret the original design calcs. or make adjustments and corrections to them? What is your company’s involvement in this project, did you furnish the crane and this bridge, or what? I do appreciate the fact that these types of problems can happen to almost anyone in today’s engineering world. But, I am often amazed at how indifferent companies are to good and qualified design and engineering or the management and control over their construction/fabrication, etc., even after something has gone wrong. Given the questions asked, the way they are asked, the lack of problem definition, and the predicaments that companies get into, it is a wonder that anything gets built right these days, or that more of it doesn’t fall down. It’s a scary world out there.

 

[hokie66]

I may have underestimated the welding requirements, but the attached reference indicates that generally preheating of S460 structural steel is not required.

http://www.constructalia.com/englis...gth_thermomechanical_steels_s460#.VitMpS_ouUk

 

[Agent666]

You have to remember that the specified strength (S275) is generally the lower 5% characteristic strength, therefore getting 235MPa on a test coupon is very suspect as will be several standard deviations lower than the 5% characteristic stress, which is already several standard deviations lower than the mean. This is especially concerning if its the result of multiple tests.

If you were to do a sufficient number of tests the yield stress distribution should in theory follow the typical bell curve we all know and love, with 95% of results expected to be above the 275MPa grade designation.

However if you are getting 235MPa (average?) from tests then something is definitely up with the steel that has been used. Seems like fabricator has substituted or inadvertently used a lower grade.

Remember as well depending on the thickness of the plate, thinner plates generally have a yield stress greater than the specified grade designation (i.e. S275) due to the working/rolling process to get the plates that thin, and thicker plates a little lower than the grade designation. Least thats how it works in my part of the world in the steel standards plate is produced to (New Zealand/Australia).

You didn't mention what thicknesses you were using, but in New Zealand some of the thicker plates say 40-50mm thick at say grade 350 specification get down to a yield strength of 300MPa, whereas 10mm thick G300 plate has a yield strength of 310MPa for example. The effect is more prominent in our higher grades of steel > 350MPa. So check the relavent plate standards and see what the yield strenght should be.

 

[kingnero]

Hokie,

just double checked because of the article you mentioned,
Depending on the type of S460 you're dealing with (S460M - S460ML - S460N - S460NL), you've got a carbon equivalent of around 0.50
Requiring preheat as per EN1011, and all codes of good practise.

I wouldn't try to do so without extensive testing, and either way you would not be working according to european standards.

 

[hokie66]

You would know far better than me about European practice, but the ArcelorMittal blurb does say that structural shapes of S460M and S460ML don't usually require preheat.

 

[kingnero]

That is also what I read. As I said, I wouldn't try to do so unless extensive testing. After all, that's what WPQR's are for.

Edit: thickness is also an important parameter in this discussion...