Attic Trusses 2

[medeek]

Has anyone else had much experience analyzing trusses (metal plate connected trusses) utilizing RISA3D?

I've pin jointed all of the web members and rigid jointed the heel and peak joints. For now I've ignored all of the load cases and just assumed one load case to get a better feel for the Axial, Shear and Moments in the various chords and webs.

Loads

Axial

Shear

Moments

A confused student is a good student.
Nathaniel P. Wilkerson, PE

http://www.medeek.com

 

[msquared48]

Well... instead of fix the apex and heel I would pin them. I would fix the sloped rafters as continuous embers and pin the rest.

Mike McCann, PE, SE (WA)

 

[XR250]

are you just bored…. :>

 

[medeek]

On the weekends I like to work on some of my various speadsheets/programs for analyzing trusses. Up until this point I haven't really considered attic trusses in any of my software however I've had enough instances where I did have to do the vertical load takedowns for them so I'm considering adding them in. However they are a bit more complicated than the run of the mill trusses I normally encounter and my lack of experience in dealing with them prompted me to put up this post.

I'm pretty sure I've got this thing figured out just fine but nine times out of ten if I post it here I will find something I have missed...

A confused student is a good student.
Nathaniel P. Wilkerson, PE

http://www.medeek.com

 

[Ron]

Agree with Mike...the only fixity you need is to make a member continuous as it exists in the construction. Pin all the others.

 

[medeek]

I've generated a model for a truss that I've recently worked on. My load case is #4 from the ASCE 7-10. I'm am trying to match the results as closely as possible to the Mitek report shown below. Changing the joints from pinned to rigid does affect the numbers slightly. The ANSI TPI1-2007 is somewhat confusing or at least a little unclear in my opinion on how exactly to treat rigid versus pinned joints at the peak and the heel. The H_r reduction factor given in section 8.3.2.2 implies that the heel joint does have some moment transferred across it and probably should be treated as a rigid joint. The peak joint is a little less clear. I typically treat all of the web members as pinned at their ends and hence only axial members, this simplifies things slightly.

All Joints Rigid:

All Joints Pinned except for Heel Joints:

A confused student is a good student.
Nathaniel P. Wilkerson, PE

http://www.medeek.com

 

[medeek]

My axial loads are for the most part quite close to the Mitek output. I am very curious though how they are treating the peak and heel joints as well as the web joints.

A confused student is a good student.
Nathaniel P. Wilkerson, PE

http://www.medeek.com

 

[medeek]

I am pretty stoked to find out that Risa3D will do a NDS code check on each member, have never really used this feature before. A word of caution though, you need to carefully setup the unbraced length for each member in one of the spreadsheets.

My combined stress indexes are coming in a little higher than the ones generated by Mitek so not entirely sure that I have all the adjustment factors correct. For example the highest stressed web member is the ceiling web that is in compression and bending. Its CSI is .58 which is slightly larger than the WB = .47

A confused student is a good student.
Nathaniel P. Wilkerson, PE

http://www.medeek.com

 

[XR250]

I believe they treat the heel and peak joints as pinned as MSquared mentioned. I don't think their software or testing is that sophisticated to do otherwise.
I have had instances where a contractor has hacked up a few of these for flues. They are really hard to fix as they are usually pretty highly stressed if the span is long.

 

[Ron]

I don't see anything in the heel connection that would transfer moment.
Remember...just because we model something in a particular way does not necessarily dictate its action or response in construction.
It appears that the original truss design (assuming that's your last sketch) considered the heels to be pinned.

 

[KootK]

I haven't scrutinized a version of TPI since the late 90's, when that was my thing. Back then there was a 'fictional heel modelling procedure" for heel joints to simulate the partial fixity generally expected there.

Truss heel heights are generally too high for a concentric joining of the chords and simultaneously not high enough to get a meaningful vertical web at the heel. You end up with wedge blocks, sliders, giant plates, and multie plates, all of which create significant potential for fixity,

I agree agree that the particular case illustrated in the sketches above would create something fairly close to pinned however.

I like to debate structural engineering theory -- a lot. If I challenge you on something, know that I'm doing so because I respect your opinion enough to either change it or adopt it.

 

[medeek]

Given the size of the plate at the heel I would agree that modeling should probably be pinned and not rigid. I have read a number of papers on modeling MPC trusses and I am still learning. Section 8.3.2.2 of the TPI1 doesn't make sense where the H_r decreases in value to .65 for higher pitch trusses 5:12 to 12:12 and then jumps up to 1.0 for truss with pitches greater than or equal to 12:12. I would really like to get into the head of whomever wrote the bulk of TPI1 to better understand some of the additional checks and exceptions. Most of it follows the NDS fairly closely but there are some things that make it special.

A confused student is a good student.
Nathaniel P. Wilkerson, PE

http://www.medeek.com

 

[medeek]

I would not discount the complexity of the software produced by the giants like Mitek. If you dig through their user manuals, which I have done on occasion, you will find various options for adjusting fixity at the heel and pitchbreaks (peak joint etc...) including a semi-rigid option somewhere in between the pinned and rigid. They seem to suggest modeling the truss with semi-rigid joints at the chords.

Read this Article:

http://www.mitek-us.com/WorkArea/DownloadAsset.aspx?id=1114

Being that the heel plates are usually rather large on most trusses I tend to agree with the article in paragraph two and I normally assume typical MPC trusses will have rigid heels and semi-rigid peak joints.

I would really like to get my hands on a copy of Mitek or Simpson's truss design software just to get a better idea of what goes on under the hood. It's really too bad they don't produce an engineer's or architect version of their software. I know the solar panel industry would definitely benefit from this.

A confused student is a good student.
Nathaniel P. Wilkerson, PE

http://www.medeek.com

 

[dhengr]

Medeek:
You make mighty nice sketches and diagrams, but I’ll be damned if I’ll follow this thread. I have to scroll all the way into the next room to the right to read the right most two thirds of each line of text. It just isn’t worth the effort. Why not show your sketches at a much smaller scale (pixel width?) so that they don’t expand the entire thread width so much? When you show them as an attachment, we can bring them up in a separate window and scroll around them without screwing up the reading of the entire thread. Many times I can adjust the scale on these attachments to see details, if I wish. I must be the only person on the E-Tips forums who has this problem, since no one else seems to be bothered by this thread width issue.

 

[medeek]

Sorry about that dhengr, your screen resolution is probably a lot smaller than my 24" monitors. Also this may have something to do with your browser, not sure. I'm using google chrome, when I shrink the browser window the width of the thread adjusts automatically so that all of the text and images show up without any additional scrolling. What browser are you using?

A confused student is a good student.
Nathaniel P. Wilkerson, PE

http://www.medeek.com

 

[XR250]

I would like to see their testing to back this up these assumptions of fixity. The amount of load these plates can take is suspect to begin with, in my opinion.
I have seen enough significant creep in floor trusses to know there is a problem somewhere. Never seen this in roof trusses, however. I guess the plates are not loaded as heavily due to the more favorable aspect ratio.

 

[PSEPK]

medeek is right, dhengr. You were not the only one, facing this problem.

I was using Mozilla Firefox on a 17" screen, and was facing the same problem, when I viewed the thread for the first time today. However, opening the page in Google Chrome has solved the problem.

You may try the same, too.

Thanks dhengr, for pointing out the problem, and medeek, for the tip.





A good structural engineer is often a blessing for others.

 

[PSEPK]

A little clarification.

First sentence in my previous post is not conveying the right sense. I actually meant that "medeek is right about the use of web browsers. Moreover, ..."



A good structural engineer is often a blessing for others.

 

[KootK]

@XR250: the situation with floor trusses is often exacerbated by the fact that they often get seriously overloaded during construction.

I like to debate structural engineering theory -- a lot. If I challenge you on something, know that I'm doing so because I respect your opinion enough to either change it or adopt it.

 

[XR250]

you mean they are not intended to have 4,000 lbs of sheetrock stored on them? :>