Truss design misconceptions 29

[RontheRedneck]

I mentioned in another thread that I've been designing trusses most of my adult life. I started working in a truss plant in 1984, and it went from there.

Over the years I've run into a lot of misconceptions about how trusses are designed, who designs them, etc. I thought I'd take a crack at explaining the typical process.

The vast majority of truss plants do not have engineers on staff at the plant. Trusses are designed by guys like me. I only have a HS diploma. But I do have a lot of training and experience.

We buy our truss plates from a company called Alpine. They provide the truss design software that we use. They also have engineers on staff. The cost of the software and engineering support is built into the price of the plates.

Since we're in a rural area we don't get an engineers seal on probably 95% of what we do. There's no reason to.

If we do need sealed drawings, it's usually because someone is building in an area where there's a building department that requires them. Or on commercial work we sometimes have to send them to the project architect/engineer for review.

Once we have the trusses designed in our system we can send a job down electronically to Alpine. One of their engineers is assigned to our account, and that person typically reviews our stuff.

The engineer does not alter the truss designs - They're either approved or not approved. If they want something changed we get an email or phone call explaining what they want to see. We revise them and send them back down.

The engineers at Alpine never see the plans. They typically do not know where the job is going or any details about it. They only review what we send down.

So that's the basic process. If you have any questions let me know.

 

[JoshPlumSE]

I've done delegated design for steel and wood joists. But, not for full truss systems. I'm a little shocked to hear KootK's description. In all the delegated design that I've done, I'm responsible for the determination of all loads applied to the trusses. It's part of the packet what we give them when we delegating that design.

The only times I've see it done the other way is for full metal building systems where the packet we give them is more generalized and they're giving us a complete system. We might give them the LL and DL, and then give them some parameters for them to determine wind and seismic. But, we'd probably specify what the lateral force resisting system is going to be (with the R value) for each direction. Along with a request that no connection exceed 80% of code limits under gravity loads, columns 90% and the rest of the members limited to a maximum of 95% of code limits.

For a truss system, I'm shocked that engineers are not specifying the dead and live loading, as well as identifying which elements are used as drags / collectors for the lateral system. Really?! The key being that I'd want to identify the trusses that we'd be doing our own analysis of when the design back from the truss guys.

 

[phamENG]

Josh,

I don't think it's entirely a question of the EOR here...it's a question of what the Specialty SE is doing as part of his "delegated design." When I do wood trusses, I define live load for the top and bottom chord, dead load for the top and bottom chord, any environmental loading (or at least the job specific parameters for them to arrive at the correct loading like wind speed, exposure, etc.), and I also identify drag trusses and other unique items along with loading.

The problem is, when I have something that's more than "I want 20/10/0/10 roof trusses at 24" o/c with 120mph wind, exposure B"...there's almost always a correction to be made on the first round of shop drawings. They used 115mph exposure B when it's really 135mph exposure D. They didn't put in the drag truss. They used the wrong duration factor. They used a repetitive member factor for a 2-ply girder. Etc. These are all things that 1) an experience tech probably wouldn't do, and 2) a SSE should pick up on a review of the entire truss system design. But according to RonTheRedNeck and KootK's description of the MPC truss industry MO...they aren't doing that. They're saying "yes, the truss design output is acceptable for the truss program input." No verification of the input, no garbage in/garbage out check. So...nothing of any worth.

 

[RontheRedneck]

I don't have time to deal with all the statements made here. Some of them insulting and false.

When the engineer puts a seal on a drawing, it's only for what's on that drawing. Nothing more, nothing less.

Whomever said that the truss designs coming from the computer are worthless is ignorant. The programs are very sophisticated.

 

[KootK]

Whomever said that the truss designs coming from the computer are worthless is ignorant. The programs are very sophisticated.

Nobody said that the designs coming from the software are worthless. What we're saying is that an engineer's review of a truss drawing is pretty worthless:

1) at the rate of 300/day and;

2) in the absence of any contextual knowledge of the project as a whole.

Yeah, the programs are very sophisticated. We should probably just let AI's "check" the truss designs.

I would urge you to not get overly defensive about the wood truss industry. In the opinion of most engineers, it is an industry that does have some problems from an engineering ethics perspective. As such, it is rather impractical to attempt to manage a "truss design misconceptions" thread without at least acknowledging those concerns.

As I said earlier, we value your input here as an industry insider. That said, you are not the only industry insider here. I've held every job that there is to have at a truss fabrication plant save general manager. In addition, I spent a couple of years at the Wood Truss Council of America back when it was called that. And I'm an SE in Illinois. I am many undesirable things but, I assure you, ignorant about the metal plate connected wood truss industry is not one of them.

 

[Brad805]

Ron, it would be helpful to understand the Alpine training for techs. Residential truss suppliers are my least favorite gang to deal with. The industry is the most cost conscious group I run into, and trades these days pay little attention. I had always wondered what info the truss eng is provided, and this has been helpful to clear that up. I suspected it was not much given what they are paid per truss design.

 

[KootK]

I don't have time to deal with all the statements made here.

Take your time. Threads don't get closed here until months of inactivity have passed. There's really no pressure to be responding in real time.

 

[KootK]

...it would be helpful to understand the Alpine training for techs.

Back in 2000-ish, the training function was sort of delegated to the Wood Truss Council of America (now SBCA) in the form of 4-day, in person courses held mostly at hotels and, sometimes, on site for the larger fabricators. Most of my job at WTCA was in developing, refining, and delivering TTT1, TTT2, and TTT3 from the list below. I've stayed at a ridiculous number of airport hotels in the continental US.

I would absolutely love to share the content of TTT3 with the gang here as that was my baby and there was some great stuff in there. As you can imagine, however, I don't need any additional intellectual property suits this year. Maybe in 2023.

 

[3DDave]

As an outsider - it seemed to me that the biggest problem facing anyone dealing with trusses is that the surety of the joints can only be determined by testing as there aren't great models of how steel teeth gripping into an inconsistent fibrous material functions, particularly if that fibrous material can vary so much with water content. In particular some plates seem to never have problems while others that at a glance seem similar enough do. On top of that they seem sensitive to handling - they can leave the factory fine and be mishandled and not have a visible disconnection when they are installed.

 

[phamENG]

When the engineer puts a seal on a drawing, it's only for what's on that drawing. Nothing more, nothing less.

Yes...and no. Context is important. It would be one thing if we were talking about a garage door - it's been analyzed/tested for up to 120mph wind loads, and the drawings say as much, and now it's a mass produced product that can be installed in any house with a design wind speed of 120mph. But these roof truss systems are one offs. They aren't the 'pick-the-truss-from-the-table' jobs of the 60s and 70s. They are sophisticated, unique designs tailored to a specific building in a specific place. So it's important that not only do the results match the input (which is how I understood your earlier statement about engineering reviews)...but also that the input is valid and consistent with the specific requirements of the job. Is it the SSE's job to determine that from local codes? No - the DOR should lay it out on a silver platter for him. But if the SSE doesn't have the structural plans (and hopefully the architectural plans) while reviewing the truss calcs, I don't think they're quite doing their job in full. They need to make sure the system works and meets the requirements from the EOR - and the EOR needs to make sure the system and the building will play nice together.

Whomever said that the truss designs coming from the computer are worthless is ignorant. The programs are very sophisticated.

I'm pretty sure you picked that up from my comment, though I assure you it's not the meaning I was trying to convey. As KootK clarified, I'm not saying the truss software is worthless. Far from it. I'd love to get my hands on some. TPI-1 checks are really tedious to do by hand when I'm checking an existing truss for somebody. But when an SSE is rubber stamping the output without any thought to how the truss fits in the system, they are worthless (at best).


If you read our responses more closely, you'll see that our issues with the industry aren't directed at you. In fact, you seem to be an incredibly knowledgeable truss technician that can probably design circles around most engineers when it comes to trusses. Look for posts about PEMBs. That's another industry niche we love to hate. But really, it's all about economics. Who can produce a code compliant structure the fastest with the least cost? That's PEMBs and pre-fab trusses. Hands down. Unfortunately, they are the least 'changeable' once they're in place. So when you really dig down, most of us just have issues with the construction industry in general and how many of us are constantly under pressure to do more faster and cheaper...like PEMB and Truss shops are able to do through automation...while still providing unique and complex structural designs in a variety of forms and materials.

 

[BridgeSmith]

I don't have time to deal with all the statements made here. Some of them insulting and false.

Ron, you began this thread, on a site primarily populated by engineers, by basically saying 'we don't have no stinkin' engineers, and we don't need no stinkin' engineers', and you didn't expect some pushback on that?

We appreciate your straightforward explanation of how the industry functions, but you should be open to some criticism of the apparent lack of 'responsible charge' in some situations. That is a big issue for us engineers, because if there's a failure, our profession as a whole gets blamed.

You may find it offensive, but regardless of how experienced you may be in 'designing' trusses, if you're not a PE, you're not qualified to engineer them, nor would you bear the legal responsibility for a failure. I've tried to follow the posts in this thread, but it's still somewhat murky to me who would bear the responsibility in some situations (such as residential) if a truss system were to fail and someone got killed. We (I think I can speak for pretty much all of us engineers here, at least) take our responsibility for the lives of people who rely on our work very seriously. After 16 years as a PE, I still get a lump in my throat every time I get ready to put my seal on a design, realizing people's lives are depending on me doing my job right. So, yes we get a little defensive about people who aren't engineers doing engineering, and engineers who 'rubber stamp' things without really knowing if they're going to perform like they need to.

Rod Smith, P.E., The artist formerly known as HotRod10

 

[KootK]

Speaking only for myself, I wish that I could go back in time and refrain from participating in this particular thread. The whole "if you don't have anything nice to say..." thing. I very much want for there to be MPCWT representation in this forum and that's just not going to happen if we're constantly beating these guys up over the responsible charge thing, valid as those concerns may be. Nobody wants to hang out in a venue where they're constantly made to feel like an unethical jerk. And it's all but impossible to not take this stuff a bit personally when it's your own industry.

@RontheRedneck, here's what I propose:

1) Stick around and help us out with our technical MPCWT issues. We'll shower you with adulation, interesting conversation, and little purple stars in return.

2) Let's do our best to simply not discuss the "responsible charge" issues. Those issues are:

a) Not of your making and;

b) Not your responsibility to solve.

 

[driftLimiter]

I have had pretty good experience with Truss designers, In my area the truss design is submitted with original construction documents, consequently we don't think of it like a deferred design.

We take shop drawing of trusses very seriously because we accept that the truss package is going to be included with our construction documents and ultimately we bear responsibility that it is designed correctly.

But to others' points in the thread, I really don't know what is going on in the truss 'calcs' other than verifying the load criteria and basic geometry matches what we have been asked. I look for the loading and drag truss loading, confirm the shape of the truss (span, heel, slope etc) and if it all looks good its approved.

If there were for example a setting within the truss company's software that changed some assumption drastically it is very unlikely that I would catch it because I have no knowledge in that software. I think this is the role of the truss mfg. engineer who does know the software and should understand what is going on from the loading through to the output.

The way we practice is that EOR is responsible for providing loads roof loading for dead, roof live, and snow loads, and location and axial load of drag struts additionally we provide the ultimate wind speed. We also from time to time supply the truss mfg with details showing special applications of trusses including drag connections, offsets, or large bearings.

I quite often catch the truss designers missing criteria we have specified, but they are always quick to pick it up and never once has anyone tried to argue with me. However I do get the sense that sometimes technicians just march forward with their own assumptions without much regard for what the EOR calls for. This is why the shop drawing review is so important.

 

[MotorCity]

Since no question was asked by the OP, this thread is soliciting opinions or trolling for a debate. Both good and fine. However, one cannot heave a touchy subject over the fence and not expect disagreements.

When I delegate any portion of a design, I expect sealed calculations (which I would have to provide if I were doing it myself), the design would be performed by a licensed structural engineer, and it would be a sound, economic, and constructible design. Basically, I hold the delegated design engineer to the same expectations and standard of care that I hold myself to since I am ultimately responsible for his/her design as well.

While I do not doubt that there are people out there perfectly capable of punching numbers into a truss design program and generating a truss that works, that is not engineering. A truss is part of a structure that does need to be engineered because if it fails, it could result in fatal consequences. It goes back to the age old "garbage in, garbage out" mantra. As with our junior engineers that love to sit down and start using software, they also need to be able to perform hand calcs to verify or approximate the results. If you cannot use first principles and basic mechanics, you should not be blindly using any software.

 

[RontheRedneck]

You know, I expected a forum of engineers to be more professional that some of you are. I'm particularly annoyed by the pompous ass who said I'm not qualified to design trusses. I know more abut truss design than most engineers do.


Someone mentioned training truss designers. Most places I know of like bringing guys in from the plant. That way they already know what a truss is. They know what works in the plant and what doesn't.

We start them out slowly - Designing trusses for something like a 23x36 garage. Then we teach them how to input plans in a layout. We have a series of "training plans" that progress from simple to more complex.

It takes time. There's a lot to learn and a lot to remember. I usually figure it will take a year before a designer can work on their own independently. Although that varies depending on how smart they are and how quickly they pick things up.


Jumping back to the issue of an engineer only sealing one truss at a time -

Trusses are components of a roof framing system. Just like a 2X4 stud is a component in a wall.

You wouldn't go to a lumberyard and tell them they have to provide engineering for an entire wall just because they're selling the components of a wall. And the same is true of trusses. We're selling components, not an engineered system. So when an engineer seals one truss design, he's only sealing one part of an overall system.


Someone mentioned truss fabricators being "cost conscious". Why do you suppose that is?

One of my favorite saying is "People are willing to pay for quality - As long as it doesn't cost extra".

If we're not competitive on pricing we don't get jobs. We lose jobs over ridiculously small amounts. It's a constant, daily battle.


Liability is another issue one of you brought up.

In my experience, if you have insurance you're liable for anything that goes wrong. Since I started in the truss business in the 1980s no company I know of has been involved in a lawsuit because of something they did wrong. Every lawsuit has been a money grab. i.e. someone gets hurt on a jobsite and sues everyone who has insurance.

I have seen mistakes made that ended up getting to the jobsite. Every company I have worked for has taken care of problems that were their fault. And often help solve problems that are not their fault.

If truss companies were making sub-standard designs and sending them to the field they wouldn't last long. The lawyers would have a field day with them, and it wouldn't be long before no insurance company would cover them. They'd be bankrupt in no time.


Someone mentioned plate testing. Truss plates have to undergo a tremendous amount of testing in various species and grades of lumber before they go to market. It literally takes years to do the testing, analyze the results, and assign values.

Full sized trusses are also tested to failure to verify design methodology.


KootK, I appreciate you asking me to stick around. For the moment I plan to. But there's a limit to what I'm willing to put up with.

 

[KootK]

KootK, I appreciate you asking me to stick around. For the moment I plan to. But there's a limit to what I'm willing to put up with.

I hear it. That said, I recorded your email address from the other thread. So, should I deem it necessary in the future, I'll simply summon you like a very particular superhero.

 

[sandman21]

Why are we acting like the code does not spell out the requirements for truss design, truss design drawings and sealing of the drawings, IBC 2303.4. Even if the IBC in the area does not have the section, ANSI/TPI-1 covers many of these items including the responsibility of the manufacturer. You as the manufacturer are providing an engineered system not a component like a 2x4. TPI-1 goes over all of the testing and design of metal plate trusses.

Additionally, some may want to look more into delegated design, as it varies by state. Link is a helpful start.

Design Delegation: Legal Definitions and Practical Effects

 

[DayRooster]

Random thought but there really should be like a small technician certificate for being like a certified truss designer, certified pipe support design, etc. Like an accreditation that won’t require complete structural engineer knowledge but just enough. And it should be recognized, governed, and monitored for corrective action like the P.E. license. It should also have strict limitations and outline the process. Heck we can barely hire enough engineers. Why not allow for something like this?

 

[Aesur]

@sandman21, I don't believe @RontheRedneck is saying his truss isn't a system, he is saying the truss itself (with all it's components) is the component to our building.

@RontheRedneck - thank you for sticking around, your input is very valuable here, and I think I can speak for most here in saying it is evident of your experience and I doubt anyone questions your knowledge in this area. The issue people have is who is responsible for making sure all the add loads, drags, etc are accurately in the truss designs, and being that the EOR is not familiar with the truss design software, nor has a copy of it, the EOR cannot with confidence say it is 100% correct and therefore relies on the truss engineer to make sure it matches the drawings. This is a common issue as many times in my area, these added loads are missed the first round and the EOR spends a good deal of time reviewing truss shop drawings over and over sometimes without just compensation for their time. Many engineers have notes they place on the drawings saying they didn't review everything and it is the contractor and truss designers responsibility to make sure the trusses meet the projects design criteria. Many truss shops are also reviewed by junior engineers which they themselves lack the knowledge to adequately understand the add loading and why it's required.

In an ideal world, the engineer sealing the truss calculations would seal the layout as well and they would have a set of the structural drawings as they are the delegated design professional. Based on my own experience I can say there are some amazing truss designers out there and some horrible ones, this can be said about ever profession. Lately I have rejected truss shop drawings over and over until the manufacturer finally brought in a professional engineer to review it and it got through without issue then. This was all due to proper application of terrain factors for wind, snow drift loading and drag loads. The first 2 reviews I gave them feedback explaining exactly what I saw was missing, after it still wasn't addressed after the first 2 reviews, it was a simple matter of look for those few calcs, and reject with a note to see previous comments, because like yours, my time is valuable as well and our small CA fees only cover so much.

One question I have that has bothered me for years is how truss manufacturers (at least in my area) don't use the specified species of lumber and many times use "weaker" lumber which affects the h clips, a clips, straps etc.. capacities. I have started just using the smaller capacities (which contractors are noticing since straps are longer and require more nails), but this is the only way to have a properly designed system. As an example, we use Douglas-Fir as the basis for design in my area and specify as such on the plans, however most trusses seem to go with Hem-fir or other weaker species. Many EOR's ignore or don't even realize this, the few I have asked in the past simply said they didn't care. What are your thoughts on this?

 

[phamENG]

our small CA fees only cover so much.

That's why all of my proposals now have hourly CA, non-negotiable.

RontheRedneck - thanks for sticking around. I hope you'll give most of us the benefit of the doubt - we're not trying to attack you and we mean no disrespect or offense to you. As others have said, we really appreciate the technical perspective and industry insider viewpoint you can provide. While I think it's true that many of us don't completely understand what goes on 'behind the curtain' in the MPC truss industry, it's also true that some of what you see as our misconceptions are actually a drive for us to push for changes in the way the industry operates. Or, at the very least, how it interfaces with the broader engineering community.

One thing that I hope you'll understand - for the average engineer working in a consulting office, the idea of giving somebody without an engineering degree access to structural analysis software is tantamount to blasphemy. It just isn't done. But it is done in the truss industry. All the time. That's something that many of us just need to make our peace with. But the flip side to that is that it's perfectly reasonable to expect a more detailed review from the engineer who is signing off on the truss system.

sandman21 - nice paper. I especially love this line regarding shop drawing review stamps:

Significant legal battles have been waged over the import of the fine print in these routinely employed stamps, which, truth be told, are often read by no one other than those who first created them and the lawyers that argue about them when something has gone wrong.

 

[stanleyshum1997]

It is clearly stated in the book, "Structural Analysis," by Hibbebler that a truss is statically determinate when b+r=2j and statically indeterminate when b+r>2j. And the truss is unstable when b+r<2j. Where b is number of member, r is number is number of external reaction and j is the number of joints. A truss member is usually moment jointed but the reactions at the end of member are considered as pinned. Similar to the concrete member sitting on concrete girder.

Yes, the regulator is right. Indeterminate truss is stable when b+r>2j and they can be solved by computer.

I have mistaken the truss to be something called external instability when the truss reactions are parrallel when there are interior joints on the bottom of the truss. There is no lockdown in the x direction. The truss can be overlooked if the ends of the truss are not sitting on rigid walls.