Ply gussets in wood truss 2

[Aton]

Hi. I am new to the group. I do solar, structural, and environmental engineering. I presently am analysing a series of wooden trusses with plywood gussets. Since all ply gussets are large (relative to the metal stampings commonly in use), the joints cannot be modeled as simple pin-connections, where only axial forces are transmitted. In many of these gussets, significant moments can be built up, which have a direct impact on the number of fasteners specified. I have gone through the labourious process of employing the slope-deflection equations with FEMs - ending up with a large matrix that I solved for the deflected slope angles for each member (and then plugging back into the SD Equation to solve for the moments. (This is a procedure developed by Prof. S.K. Suddarth from Purdue in the earky 60s) The trouble is that I do not have the time to painstakingly verify every plus or minus sign in this lengthy and iterative process - and do not entirely trust my results. Does anyone have any suggestions for alternative (and less time-consuming and error-prone) methods. I am interested in Finite Element Analysis Software (I have been exposed to NISA a good while ago), and perhaps this is the only route to take, but do not know which FEA software would be appropriate for the money. Any comments would be helpful. Thanks in advance.

 

[DaveVikingPE]

Is there a good reason to use plywood gusset plates in lieu of custom steel (not stamped) gussets? You probably have already found that you might was well sheath the entire truss in plywood, depending on your loads. Do you really think that the conections (nails - assumed) are that rigid to not allow some rotation? What is the size of the truss members themselves?

My advice, off the top of my head, would be to look at 1/8" or so thick steel side plates with bolts if the truss is of substantial size (the Timber Construction Manual,

http://www.aitc.org,

has example of timber truss connection design - though NOT stamped metal plates).

Are you doing research on plywood gusset plates? If so, I'd be very interested, nevertheless.

 

[Aton]

Thanks for your reply! Yes, ply gussets (especially if they are glued and nailed) are very close to rigid connections, and must be modeled so, or moments are neglected.

The reasons for ply are four: (1) In a floor system, standard sheet metal gussets on floor trusses are not terrific in a fire. The metal acts as a conductor of the heat, increasing the penetration of char. The metal ultimately warps in the heat, and can partly pull out of the wood (penetration is only about 3/8" anyway). Although upright chord floor trusses with metal stamped gussets are better in a fire than are I-type joists made with OSB webs, they DO tend to sag under load, and are unstable at best. (2) This particular analysis is on the repair of an old ('50s probably) site-built truss roof using ply (and, canyoubelieve, 1x8 spruce!) gussets on a large building. I was hired to see if repair was possible - rather than ripping off the entire structure and exposing the interior to weather (plus possible increased expense. (3) In many cases, where distance from a regular truss manufacturing plant is far, it is practical to produce site-built trusses. In the 50s and 60s these were advocated by CMHC, the housing department of the Canadian Govt. - but are now a forgotten relic of the past. However, ususlly site-built rafters (using traditional steel square layout) is more practical for roofs - which brings me to the main reason: (4) In sustainable construction, I advocate designing with several goals in mind - one of which is to design for adaptability of interior space (saving much embodied energy at every stage of a building's life). Full-span pright-chord floor trusses with no interior bearing walls are ideal for this purpose.

While properly engineered plywood box beams are structurally adequate, they are hardly cost-effective. Thus, my need to engineer for ply gussets. Now, split ring connectors are also an option (and have the added terrific advantage of being pin connectors), but these are hard to get (one source in the US and the other in Britain), and are expensive. I am also interested in exploring the metal plates of which you talk: that is my next option to analyse. However, from a carpenter's point of view (site-built, remember), neither split rings nor metal plates with bolts are as easy as air-nailing and clinching some plywood gussets.

I will continue the analysis - it may well end up as a FEA job - and will be happy to share any positive results. Talk soon! Sustainable, Solar, Environmental, and Structural Engineering - Solutions world-wide.

 

[DaveVikingPE]

What kind of glue would you add? Won't it crack under rotations unless it's particularaly springy? What about fatigue loading?

What kinds of nailing requirements have you come up with based on the NDS? Also, have you thought of using lag bolts to connect the gussets (kind of a waste of lag bolts, but depending on what size truss memeber you have). Additionally, have you asked anyone at the Wood Truss Council of America (

http://www.woodtruss.com)?

I'm assuming that the plywood gussets would be cut on-site from new plywood. I really thing that you're better off going with metal side plates plates. You can fabricate them on-site from steel plates, or even canabalize a spare W-shape if one's nearby.

 

[Aton]

Hello, again, DaveViking. You make some good points, and I'll do my best to respond.

1. Kind of glue: A high-strength glue like PL-2000, which develops 200#/in^2. If I were sure of (low) moisture conditions, I would use a good aliphatic resin. These glues are stronger than the lignins in the wood, so on the interface, the truss fibres would tear before the glue gave out. Due to cost and fabrication time, I would stay away from resorcinal or epoxy. Cracking of the glue bond under rotation is not something I have seen. Since both the wood truss member and the ply gusset are checked under Limit States Design codes (CSA 086.1-94 here in Canada) for axial tension and compression, as well as the various types of bending and shear if applicable, these aspects of the loading are considered. As far as fatique loading goes, no, I have not considered it - and perhaps I should. On the other hand, I would be surprised if the standard stamped metal gussets have ever been analysed for fatique either. If anyone has any information on this topic, please post.

2. Nailing requirements: This is, indeed my difficulty, and it is why I need to know what moments are developed in each gusset. Designing nailing for axial tension is not a problem. For example, one unclinched common wire nail (3" long, gauge # 9) will have a factored lateral resistance of 0.6 kN in Spruce-Pine-Fir if wood side plates are used (and a few other conditions are met). This is fine if the gussets are modeled as pure pin-connected, for then a simple Method of Joints or Sections analysis of the axial forces is all that is required under static loading. But if not pin-connected, I must have a way to compute the moments that is less onorous than the one developed by Suddarth that I mentioned in my first post. Knowing the M(max) in the gusset can be translated into shear on individual nails - which might change the number of nails needed....

3. No, I have not contacted the Wood Truss Council of America. I believe I had visited their website in my search, and, if memory serves, found no mention of plywood gussets - only the regular ones, all of which, I believe are proprietory anyway. I did contact the Canadian Plywood Association (which is how I received a copy of Suddarth's paper). Other than that, I have struck out, because, I believe, no one uses ply gussets any more.

4. Lags: A possibility, but, again, moments will develop, and I am back to square one - with the added reduction in cross-sectional area that the use of lags necessitates.

5. Ply cut on site: Well, perhaps. Most contractors have smaller portable table and radial arm saws for on-site work. Many have shops with heavier-duty equipment, where the mass cutting of the gussets would take place. This is the easy part. Air-nailing and glueing too (even clinching, which increases the lateral resistance by a factor of 1.6, is easy). Knowing construction as I do, using ply is somewhat easier than the use of metal plates - just a matter of air-nailing and spreading glue. Even hand-nailing is not too time-consuming, and a crew can have a good time doing it.

Moments.....

Sustainable, Solar, Environmental, and Structural Engineering - Solutions world-wide.

 

[DaveVikingPE]

My own experience with plywood gusset-connected wood trusses is sort of extensive. I was in a situation where, for six months on a 24-7 construction site, a lot of roof trusses were constructed by hand labor (some of them spanning 50 ft - though supported by columns at panel points). Gussets were cut on-site from new plywood. I considered the joints to be pinned in my designs, even with a lot of nails, as the nailing patterns and connection to bearing columns resulted in concentric forces, hence no moment. Now, you're still going to get small rotations, but the loading cycles, creep, etc. aren't going to be such that the truss plates/substantial nailing will cause splitting in the truss members.

The major Q&A job was to count nails in the gusset plates, as you can imagine.

Calculating moments at the connection points is an entirely different matter, though straight-forward. Moment Distribution ought to be pretty easy, though laborious (but not that laborious). On the other hand, if you're going to go to the trouble and expense of a full-blown finite element analysis (greatest cost is in buying the software and figuring out how to use it, though free programs are available as are low-cost truss design analysis programs, i.e., WinWood by Archon -

http://www.archoneng.com)

you might be spending more than it would cost to field-fabricate metal side plates from relatively inexpensive and easy-to-transport steel plates.

Questions:

1. Did you check with the Engineering Wood Association (

http://www.apawood.org)?

I think, since they are the definitive authority on plywood, that you should press them for answers.

2. What really is the issue? Is it a desire to re-use materials? Is it a desire to minimize any expenditures of "energy"? Is it a question of economics rather than engineering? You can quickly and easily fashion steel side plates - steel is a recycled material - that will give you pinned connections.

 

[Trussdoc]

Interesting discussion.....did you also get a copy of Stan Suddarth's Purdue Plane Structures Analyser (PPSA)program? this program allows you to lock joints to create a fully rigid truss, even modeling the webs as rigid-ended. Also, a demo version of "Visual Analysis" may be available at low cost from IES, and this does a handy job for heavy timber trusses.

 

[Aton]

Thanks to both of you for your replies.

To DaveViking: It is good to hear that you were able to deal with ply gussets as pin-connected joints, and did not worry about moments. For now, that is what I will do. As for the complexity of the analysis, I had not tried Hardy Cross. Instead, Slope-deflection equtions, ending up with a 6x6 matrix. I was not trusting my process: switching signs from "beam convention" to "joint convention", among other things (and, really would have loved to have an engineer more experienced with these methods looking over my shoulder!). The basic issue, really, is one of providing the contractor with some alternatives to standard manufactured trusses (besides having to solve this particular truss repair situation. I have no problem with the concept of steel side plates, and am now (thanks to your suggestion) thinking of easy ways to fabricate the trusses on site with the steel gussets and bolts. I will be following up on both the websites you mentioned in your last post. Thanks.

To Trussdoc: Thanks for the advice! I will do a web search for both PPSA and IES.

Sustainable, Solar, Environmental, and Structural Engineering - Solutions world-wide.

 

[dpa]

I agree with Dave Viking about treating the nailed connections as pinned ends. Nails are maleable. That's why with a few prequalified exceptions earthquake retrofits here in Seattle require nails to be used with Simpson Strong Tie fittings rather than screws. This even applies to Simpson fittings where screws are specified in the catalog. Screws are more brittle and the heads can pop off in earthquakes.

Good luck
Dave Adkins

 

[Aton]

Thanks, Dave Adkins! Interesting comment about SST and brittle screws. -Aton Sustainable, Solar, Environmental, and Structural Engineering: Appropriate technologies for a planet in stress.

 

[AJGIII]

dba
Except for bugle head drywall (and similarly made) screws, I'm wondering how properly sized metal can fail before wood. Doesn't shear value of the steel exceed forces resisted by the gussets?

In any case, Simpson produces a 1/4" lag screw that is self tapping and clearing that provides good values (SDS series), especially w/ 5/8" ply.

the problem I keep encountering is contractors not reading or following the plans, or thinking they know a better way. Oh, and people that can't follow directions by others (language in Calif?) or jsut don't care.

On the last truss repair job I did, the contractor was suprised I checked all of the trusses--figured I was going to check only a few near the entrance, and pretty much said so. Incredible.

 

[DTGT2002]

I am thinking the screws fail in tension with earthquake loads. I have no problem believing the nails elongate and continue to hold, even with a bit of local necking.

However, I do have a problem using the simpson catalog freely without meeting all asuumptions. I would think it to be proper to use an appropriately sized hanger that uses nails rather than use the wring fastener in a hanger. For nailed connections, do you use the simpson nails or allow common nails for use. The design capacited are based on all design assumptions being met.

 

[greggillis]

Before metal connectors, trusses were built with plywood gussets. I have a book with many, many examples of plywood gusseted trusses that was published by the Cornell University extension service. It has been a long time since I referenced it but I will dig it up for design info if you guys are interested.

I like the plywood gussets. I was embarassed once in a trial where I was an expert witness in a truss failure case. I was demonstrating how strong the metal connector was and the darn connection broke. I know that if it had been plywood it never would have broken

Are any of you in New York State?

 

[bylar]

you failed to mention what size trusses you are talking about and what kind of loads.

My first thought is that you are over engineering a rather simple problem but i don't know all the details.

You may look at the simpson site they have a new screw/nail made specificaly for truss repairs.

For me i try to avoid plywood gussets except for transfer of low forces.

 

[swBausch]

If you gents don't mind a question:

Last time I looked at "extension service" documents on nails, screws and trusses, seemed like screws were rated at twice the pullout of nails.

As far as screws being brittle, it depends on the screws. Brass isn't brittle, and neither is the last batch of "stainless steel" deck screws I used on to apply 5/8 sheathing on the side of my house. If I hit a knot, they would literally corkscrew on me, or I would see the head rotate while it failed to pull into the wood, just before the head would twist off. If I attempted a vigorous hammer tap to start the screw, it would bend just as a nail.

Drywall screws may be brittle, but I wouldn't say that about "deck screws".

As far as elongation of a nail, perhaps if it's clinched, but it's the rare situation that allows for clinching. And, clinching isn't a process lending itself to power tools.

When wood-to-wood is fastened, my rule of thumb is a nail will pull out, a screw will pull the head through the wood.

 

[cycleman99]

Does anyone know where I can find allowable shear values for nails in plywood gusset plates (values that are approved for building code use)? I have a stubborn client who wants to fabricate his own wood trusses with plywood gussets. The glue and nail solution sounds good, but is there test data to back up a particular assembly (ICBO Evaluation Reports?)?
Thanks.

 

[AJGIII]

It's been a long while since I checked this site, partly because someone stole my office computer. Anyways...

To DTGT2002, I don't recall recommending anyone use the Simpson catalog indiscriminately. If you were truly concerned, I'd expect you'd refer someone to the ICBO Evaluation Reports for the Simpson SDS screw---that is what I use.

Deck screws are not brittle like bugle head screws, and they have known design values. So, they would be acceptable.

Over the past year, with continued frustration dealing with contractors not reading, not caring, not having mental capacity to understand how to place fasteners in a pattern (the best pattern job I ever saw was by a musician--made a cardboard template and thought first, unlike any contractors I've met), I'm experimenting with specifying through bolts, much like Aton, except I'm getting resistance about using steel.

As Rosanne Rosanadanna's grandma said: "It's always something."
J.

 

[RARWOOD]

A lot of good responses to your question and a lot to think about. Hopefully you were able to get reference material on glued trusses from APA EWS or one of the other references listed.

I have a manual "Designs for Glued Trusses" published by the Midwest Plan service, which is tied in with 12 Land-Grant Universities in the Midwest. The universities include Iowa State, Michigan State, Ohio State, Purdue, ect. The manual covers trusses up to a 60' span. However my copy is from 1975 so I don't know if you can still get a copy of it.

I mostly wanted to reply for two reasons. First I would not specify the use of split rings, although there are people in the wood industry who would disagree with me. I would use shear plates instead. My company uses a lot of them in our designs both 2 5/8" and 4"..

We do not use split rings because of the difficulty of accurately fabricating the connection. One thing to keep in mind the installation of either split rings or shear plates requires a special tool. The only contractors I know of who have the ability to field install shear plates are the ones who did not rerurn the equipment we loaned to them.

Our standard practice is to shop install the shear plates, at one time we would shop fabricate for the shear plates and then send the shear plates loose. That practice was discontinued because of the difficulty of verifying in the field that the shear plates were actually installed.

The other reason I wanted to respond was because of the discussion about how to model and analze the trusses. I design a lot of glulam trusses using steel gusset plates with bolts and shear plates. I use the Purdue Plane Stress Analyzer treating the trusses as a series of linear members with either pinned or fixed joints.

The design of the actual connections requires a lot of judgement. On larger trusses I try to detail pinned connections. On smaller trusses I may use rigid truss plates.

I have talked with others about using a finte element program. The difficulty in taking that approach is in accurately modeling all the components considering all the variables found in the wood members. I would be interested in hearing from someone who has had success in doing that.