Welded Steel Plate Design

[EagerOne]

Hi All,

I am a student who is trying to do something new for a change. This is not yet in my course material but just interested in learning. I was reading about how vaulted ceilings can be framed and came across this article (

https://www.finehomebuilding.com/membership/pdf/17848/021178060.pdf)

which shows how a steel plate is used as a component in a flitch beam scenario. This allows the contractor/designer avoid collar or rafter ties and also avoids a ridge beam.

I have done very few welding design activities beyond the basics. I understand that I have to model the "a-frame" rafter assembly to find the moment at the peak (completed), and then design the welded connection of the two plates to sustain this moment.

Does anyone have any tips on how to complete this weld design?

Thank you.

 

[dik]

Do you need a CJP weld? What is the load? Can a 3/16" fillet weld do the trick... it's good for a factored load of several thousand pounds. Looks like BAR 3/8x4 minimum... Sorry it's a ridge... I thought it was a 90deg bend... CJP may be appropriate... To develop that moment, you may require a lot more fasteners. Can you use Glulam rivets? They are generally better for that type of connection.

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

-Dik

 

[EagerOne]

Hello dik, thank you for replying. Truthfully I have not thought of glulam rivets but I do have access to a copy of Design of Wood O86 so I shall give it a read through. In essence, I am looking for guidance on how to design the welded connection between the two plate sections. Perhaps I am overthinking it, sorry, but this is new to me, and my process so far is as follows:

1) Model the "A-frame" to find the required shear Vf and moment Mf resistances of each timber section
2) With the above complete, also resolve the frame to find the design moment Mf at the peak
3) Size the rafters based on internal Vf and Mf found in point 1
4) Determine weld connection details to sustain the Mf at the peak as per point 2
5) Depending on size of plate and whatnot, potentially revisit point 3 and check them as flitch beams rather than solely relying on wood sections

I have simply never done this before so I am not sure how to go about point 4. I've done moment and shear calcs for steel sections before but not this. Can you please give a few pointers? Also, when should a CJP be used rather than a double sided fillet weld? Thank you so much.

Edit: On a sample roof of 20ft span, A-frames at 4ft on centre, and with LL 1.0 kPa, DL 1.15 kPa, SL 2.44 kPa, my Mf at peak is 29.1 kn-m.

 

[dik]

I'm not into wood design very much, but have done a fair bit. Glulam rivets are very useful for that type of connection. Spacing of them is a lot closer. Are you looking at a cathedral type ceiling? Can you use a ridge beam? Spriinging apart at the bottom outside connections can be an issue.

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

-Dik

 

[BAretired]

A full strength butt weld, properly executed, will develop the full strength of the plate material.

 

[EagerOne]

@BAretired Thank you. That is helpful. So to make sure I understand correctly - with a full strength butt weld my main concern would then be to ensure that my flitch section itself can withstand the Mf and Vf and it turns into a simple beam analysis. I have the following questions in that case:

1) Do you have any recommendations on where to read about the welds. This butt weld - would it be recommended to be single sided or double sided? My understanding is that full penetration double sided V butt welds require less grinding and filler material than a single sided V.

2) Do butt welds have a standard for the size of root gaps (such as 1/8")?

Obviously any welding resources that you may share would be useful. Thanks again.

@dik Yes this endeavor of mine is to design a cathedral type ceiling. It's a challenge I set for myself as a learning exercise. I know of scissor trusses but I won't learn anything by "specifying" them A ridge beam is also something I am familiar with and can design without issue, but again, not the lesson I am trying to learn. This solution interests me because it eliminates the post at the end of a ridge beam (or other posts along the beam span) and allows for a much more open space, windows, etc. Thank you for your input.

 

[dik]

If the material is less than 1/16", I don't leave a gap, and then leave a gap equal to approx the material thickness... up to about an 1/8" for material up to 1/4" thick. I don't have a source for this, just something I've done.

A full length butt weld, because of the slope of the weld, has adequate strength to likely develop the strength of the BAR stock (at right angles). Check design calcs accordingly. You want to avoid CJP welds due to the cost. BART is correct about the butt weld... I should have suggested it, in lieu of something other than a CJP weld... missed that, other than a CJP weld was inappropriate.

I think glulam rivets are the way to go...

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

-Dik

 

[EagerOne]

@dik - Thank you but I am confused about one aspect of your comment just now - what's the difference between a butt weld and a CJP weld? Wouldn't a CJP weld be a form of a butt weld?

 

[BAretired]

I am not a welder, so would have to brush up on some of the techniques. Your drawing indicates 3/8" plate. For that thickness, I believe you could leave a gap between the plates, no bevel and weld with a backup plate. For thicker plates you may need one or two bevels. I consider a butt weld to be a CJP weld, but a qualified welder might not agree.

 

[dik]

There's a little more preparation for a CJP weld and there are well documented and prepared procedures, including welding with a backer plate or welding on the opposite side. For a CJP weld, the edges have to be prepared for a bevel weld. I'll see if I can dig up something.

Catch the link...

https://www.metalworkmasters.com/welding/cjp-welds/

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

-Dik

 

[dik]

From a welding engineer I know...

"Yes - up to 1/8" thick (this is as thick as we use AWS D1.3; any thicker we use CSA W59) can be CJP welded with no prep and no backing material (see attached), and if we have access to both sides we can CJP weld without backgouging using certain weld processes up to 1/4" thick (see attached)."

[URL unfurl="true"]https://res.cloudinary.com/engineering-com/image/upload/v1700595347/tips/CSA_W59-18_Figure_10.1_Square_CJP_Non-GTSM_z5kbaz.pdf

[/url][URL unfurl="true"]https://res.cloudinary.com/engineering-com/image/upload/v1700595347/tips/AWS_D1.3_Figure_5.2A_Sq_Groove_CJP_No_Prep_vt8r7e.pdf[/url]"

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

-Dik