BENT_BAR_TRUSS_Circa_1956 1

[BruceLow]

I am attaching a sketch of an unusual truss I have been asked to evaluate.
I can't find this design in the SJI 60-Year Steel Joist Manual, or in any current joist catalogs.
It consists of two .5" x .75" steel bars which are bent in a zig-zag pattern. Each bar runs the full length of the joist.
These are butt-welded together at their contact points, and to upper and lower T-sections. The joist is symmetrical about the center row of welds. The sketch is shown as a partially exploded view.
It is 12" deep with a 27 foot span and on 24" centers. Supporting a flat(built-up) roof on a single story building.

Since 12" deep joists have a max span of 24 ft, (even in 1956)...

Does this style joist have another name?
Does anyone have any info on it?

 

[hokie66]

This looks like a homemade truss, not any standard product. The main problem I see with its configuration is that the eccentricity of the web to chord connections, thus placing bending stresses on the chords. I wouldn't get too hung up over the span/depth ratio, but the analysis of the truss is not simple and will be best done by computer analysis with a plane frame program.

 

[BruceLow]

Thanks for your response, Hokie66.

I'm attaching a photo of the joist.
These joists must have been a nightmare to build.
The bars need to be formed very precisely to keep the joist straight & square. Also, the weld surfaces need to be flat & clamped in contact. It is hard to get good weld penetration into thick sections without warping.

One of the things I am most concerned about is the possibility of cold welds and cracks in the welds between the bars.

If one of the butt-welds between the bars failed, could the joist split apart along this center row of welds?

The building is also in an area of heavy lake-effect snow (Grand Rapids, MI). The owners are looking to add a ballasted solar panel array which will add 7 psf to the existing roof load.

To be safe, I'm leaning toward recommending adding a support beam underneath the joists in the center of the span, and to also request further analysis by a structural engineering firm with experience in reinforcing old joist systems.

How does this sound?

 

[hokie66]

You are on the right track. I would definitely advise the owners that their building is atypical and not something which you can evaluate without detailed structural analysis.

 

[BruceLow]

Thanks for your help, Hokie66.

I just submitted a letter to that effect.

 

[msquared48]

To me this does not look "homemade", Hokie as the welds look too uniform and, quite frankly, good...

I do agree with the computer analysis though. Personally,I would look harder - specifically as to who was manufacturing open web steel joists in 1956... Maybe even a call to SJI is warranted here...

Mike McCann
McCann Engineering

 

[youngstructural]

Just thought I'd add in my two cents:

- Eccentricity at the panel point in OWSJ (which is still what we're dealing with here) is the norm, not the exception. Obviously this is quite a bit of an eccentricity in this case, but it is probably still something that falls within the allowable stress range (otherwise this truss wouldn't be 50+ years old). I can't remember how they explain it away (and if I remember right they are more or less guessing and not knowing), but the fact is that eccentricity in this type of system is not normally of serious concern. I would DEFINATELY check them, I'm just saying that they are quite normal and I should imagine will not govern the design.

- Definately time for a by-time charge out, and a detailed computer analysis.

- Be certain of your material strength and applicability of your chosen code. I don't know about the US code, but both the AS/NZS and the CISC codes have Fy to be assumed when dealing with an unknown steel.

Keep us up to date, this is an interesting one!

Cheers,

YS

B.Eng (Carleton)
Working in New Zealand, thinking of my snow covered home...

 

[hokie66]

Mike,

I called it homemade based on the sketch, not the photo. After seeing the photo, I would still say it is homemade, not because of the fabrication quality, which appears to be very good, but because the design itself is illogical. It may be strong enough, but is not an efficient use of the material, and certainly does not represent an efficient manufacturing process.

 

[msquared48]

I can agree to some extent here Hokie, but most of trusses today utilize the "VVVV" pattern of the web, while this uses the "XXXX" pattern. For the same size bar, this should give more steel in the shear area of the web, with little effect on the bending strength. Might work better for shorter spans with heavier loads. Would be interesting to look at the output from the computer run.

Mike McCann
McCann Engineering

 

[hokie66]

That is why I said it was inefficient. The truss span/depth ratio is 27. Shear should not be an issue.

 

[msquared48]

True, very true.

Nevertheless, I feel that it most likely is a plant manufactured truss that either never caught on in the industry, or was pre-empted by more efficient designs.

There must be some info on it out there somewhere.

Mike McCann
McCann Engineering

 

[msquared48]

Thinking further...remember when W shapes used to be cut in a similar fashion in the web, the two halves pulled apart, then re-welded to achieve a deeper section with web penetrations?

This design pattern reminds me of that.

Mike McCann
McCann Engineering

 

[hokie66]

Castellated beams. Still used occasionally. I have designed those, but the trusses Bruce showed us are totally unfamiliar to me.

 

[msquared48]

Me too...

Mike McCann
McCann Engineering

 

[SlideRuleEra]

Looks like the fabricator made a copy of the 1930's Kalmantruss design. See the last page of "Carnegie Pocket Companion (1934) - Excerpt" on this page of my website

http://www.slideruleera.net/contributions.html

http://www.SlideRuleEra.net

 

[BruceLow]

Thanks all,

It certainly looks like a welded, formed-bar version of the Kalmantruss design.

This is getting interesting.

I will keep you posted as this saga unfolds.


Bruce