Splicing Large Plate-Girders by Complete Penetration Joint (CPJ) welding

[Catee]

Dear All,
In general, we plan to perform a bolted splice connection for the large built-up section steel structure. I have a question regarding the possibility of executing a splice connection using the complete penetration joint (CPJ) welding method. The attached image illustrates a bolted splice connection. If the contractor wishes to change the connection type to a complete penetration joint (CPJ), is it feasible according to engineering principles? Moreover, does it ensure structural safety? The mentioned structure is a portal frame with concrete columns and large box-section steel beams as the crossbeam. This structure supports the guideway beam of a train monorail system.

 

[TugboatEng]

Is there access to the inside of the beam for a welder to run a root pass or install ceramic tape?

 

[DrZoidberWoop]

Could you attach .pdf's, instead of screen captures? The devil could be in the details, and we can't read them too well.

 

[Catee]

Okay please find the attached file in pdf for more clear details.

 

[HTURKAK]

If you look from structural engineering point of view, the answer is yes; complete penetration joint (CPJ) is possible. I think the question is for site welded splice connection.
I do not have any idea for the proposed ASTM A572 steel plate ,regarding the material properties but a quick search shows equivalent of S275 Eurosteel.

Below my points and concerns,
- The steel plt thk. more than 20 mm so PWHT and preheating would be required,
- I have cold region experience and my concern would be reduced toughness at site welded splice,
- Suggest you to speak an experienced metallurgy eng. for preheating and PWHT requirements and consider posting this thread to the WELDING ,BONDING forum.





Use it up, wear it out;
Make it do, or do without.

NEW ENGLAND MAXIM

 

[Chessanyone]

Did they change the terminology? I’ve always heard and used CJP (Complete Joint Penetration).

To answer your question, a properly welded CJP splice is like the beam is a single piece, so yes, that would be structurally acceptable. There are some complications with large CJP field welds and fit up sometimes. And if you have thick material (thinking flanges of a major girder), you may need MANY passes to get the CJP. I had a project once with CJP of thick material that took a welder 8-12 hours to complete a single weld with hundreds of passes.

 

[EdStainless]

With a qualified weld procedure and individual welder qualifications, yes you can do this.
I would want to see this process impact qualified.
What would the NDT after the welding be?
Are you planning on stopping after every few passes and doing Mag Particle inspection?
How often you stop to check is a fine balance.
If you find an indication you would need to grind it out and reweld.
The more passes the more work the repair is.

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

 

[dik]

Can you use a stiffened end plate connection with bolts (even slip critical for shear) for the moment and shear? The end plates can be attached in the shop for better quality control and slightly lesser cost. Do you need a CJP weld? or can you design it for the moment and shear at that location?

-----*****-----
So strange to see the singularity approaching while the entire planet is rapidly turning into a hellscape. -John Coates

-Dik

 

[bridgebuster]

welding dyslexia I lot of us are guilty of the occasional CPJ and some of us still say full pen. Anyway, I've seen a bridge with field welded girder splices. It's feasible but it comes down to time and money. Bolting will take about 3 to 5 days. I would say the welding is at least 60 hours. EdStainless makes good points about testing; given the weld volume you don't want to wait until the end to find a flaw. Two other things to consider, if you're in the US, the box is a confined space you have to have watchman when someone is inside, whether you're welding or bolting. You need to have an exhaust fan going when the welder is in the box.

 

[dik]

Thanks bridge... forgot it was confined...

-----*****-----
So strange to see the singularity approaching while the entire planet is rapidly turning into a hellscape. -John Coates

-Dik

 

[bridgebuster]

you're welcome dik

 

[BridgeSmith]

It's near the limit for fabrication length in my region of the US, but doable to fabricate it as a single piece. Has that option been considered?

I'd be concerned about keeping the pieces from moving during the welding process. I'm surprised the contractor would want to field weld rather than bolting. In addition to all of the NDT that it would need, the shoring looks like it would be very extensive and cost way more than what would be saved. They're not planning to support these on cranes while they attempt to weld it, are they? That would be nuts, and would have little chance of success.

I don't have my AASHTO spec right now, so does a CJP weld have the same fatigue stress limit as a bolted connection?

 

[Catee]

Thanks all for your comments. Today I've got the welding detail form the contractor as shown below. Any idea for this? Is it a good practice for CJP?.

 

[dik]

The sketch should come with a written procedure... then OK.

-----*****-----
So strange to see the singularity approaching while the entire planet is rapidly turning into a hellscape. -John Coates

-Dik

 

[bridgebuster]

The contractor should also show the the root opening. If you're following AWS D1.1 or D1.5, the opening is 1/4" +1/16, -0.

 

[Catee]

Hello All,
Today I've got method statament for welding splice procedure. Please find the attached file for details. Any suggestions are welcome and appreciated.
Thanks,
Catee.

https://files.engineering.com/getfile.aspx?folder=f46017ab-9e0f-4c67-94cf-0d9cd928add6&file=MS.pdf

 

[Catee]

May I ask for more information? If the steel plate is not 2.50 m wide (crossbeam cross-section size), can we weld two plates of 1.0 + 1.50 m together?

 

[ifitsmoving]

Similar to previous answers, there's no reason you can't do a longitudinal weld - it's not much different from the weld to the flanges. Provided you've done the checks for fatigue / fracture tolerance of the final structure, have an appropriate weld procedure, and have an appropriate temporary works procedure you should be OK.

However at least in AUS, plates with widths >2.5m are easily available. Are you sure you need to do a longitudinal weld? It will likely add significant cost and significant risk of defects and rework, compared to the cost of sourcing a plate of the right size. Weld shrinkage could result deformations of the web plates which could be a buckling risk or at least very unsightly, or require additional stiffeners and rework to address. If you plan to do the splicing after the rest of the fabrication is complete, any internal stiffeners will also need to be welded up as well, or otherwise a new stiffener detail developed.

Also, I get the feeling that you are out of your depth - a 2.5m deep girder supporting a train / monorail viaduct is a pretty significant piece of structure. You shouldn't be relying on Eng-Tips to answer relatively basic questions like these. If you don't know / can't find the answers yourself, the first place to go should be your company's senior engineers. They should have been able to answer these questions. If you aren't working for experienced engineers, or they won't answer your questions, find somewhere else to work, where you get the support you need.