Residential Lateral Design Philosophy 3

SarBear · Dec 22, 2022

Hi, thanks for taking the time to read my post. I wanted to put some thoughts I have about residential lateral design out here and see if anyone has some advice or words of comfort lol.

I've been working at this engineering firm for a couple of years, and thus far my training has mostly been on small commercial buildings and small residential projects (tract housing). We have started to get an influx of large (like 10,000 square feet plus) custom residential homes with features including:
- 12:12 roof pitches
- Plate height changes galore
- Jogs, jogs, and more jogs
- Windows, windows, and more windows
- 3-sided glass dining nooks that stick out 15' from the rest of the house
- Covered decks that extend 15' away from the rear wall of the house
- Wings of the house at a 45 degree angle from the rest of the house
- Second floors with exterior bearing walls that are offset from main floor bearing walls below
- Rear walls of the home with a second floor, a main floor, and a walkout basement

We are in a heavy seismic area and have developed Excel spreadsheets for our lateral design. As I use our spreadsheet I can't shake the feeling that the formulas and calculations we have created are not even close to approximating what would really happen in an earthquake. Our spreadsheet or any other spreadsheet can't possibly capture the effect of 12:12 roof pitches, bearing walls that are offset from below, multiple jogs in each wall line, numerous plate height changes, etc. In my opinion if we were honest about really trying to analyze these structures and understand their true behavior in an earthquake then we'd need to do something way more robust than use our spreadsheet. Don't get me wrong, our spreadsheet is great for looking at rectangular-shaped buildings with flat roofs. But these custom homes we're working on are ridiculously complex. In my opinion, using our spreadsheet is just a way for us to say "See look, we ran some numbers". Well I think that those numbers are a sham and that it's dishonest to pretend that those numbers are anything besides eyewash. I've brought this up with my bosses. Their response usually boils down to "This is how all the other firms around here do it", with the reason for that being that's it's most cost-effective. Sure, we could build a scale model of the home and test it on a shake table, but who's going to pay for that?

So I guess some questions I have for you all are:

1) In my opinion, as engineers we should be able to "prove" mathematically that everything we call out on the plans and every construction detail we draw "works". Am I wrong? If I'm right, then I think that any engineer involved in residential engineering is not actually doing that.
2) What programs or solutions are you residential engineers using on these super complex custom homes? I looked at Woodworks, but from what I understand they don't have an option for changing plate heights on a level. In my mind that eliminates Woodworks from being a robust solution, considering that many of the homes we're seeing now have a 10' plate height, but then a 12' plate height in the master bedroom, and then a 19' plate height in the great room, etc.
3) Do you have any advice for someone just starting out in this field? I'm starting to feel a little disappointed in my career choice. In school we focus so much on calculating things and getting the right answer. In the field things are never as simple as they were in school, but even so it feels like there are a lot of things that we continue to do because they've always been done that way even if they don't "calc out", or we run these lateral design numbers just to say we did it rather than really trying to understand the true behavior of the structure.
4) Any other thoughts about lateral design on residential structures?

Thanks again for your time.

bones206 · Dec 22, 2022

There are others on here who can weigh in with much more experience in this arena, but my first suggestion is to try to discretize the complex building into smaller rectangular sections that can be analyzed with your existing spreadsheet. You can distribute the seismic forces to each of these segments in a rational manner, then do your typical lateral analysis for that individual segment.

When you have special conditions like an offset bearing wall, you can still utilize your spreadsheet for the bulk of the analysis, but then you can supplement with hand calcs for the remainder of the load path (transfer beam, overstrength requirements, columns, etc.).

driftLimiter · Dec 22, 2022

Hand calcs and engineering judgement are used heavily for this type of construction in my office. I agree that relying on a simple rectangular box spreadsheet for complex custom homes is not a good approach.

Generally for something like this I start off with a simple approximation to get some loading. Then as you detail the structure you address the offsets, slopes, and other items that impact the load path.

I would try to not get "too" hung up worrying about what is the actual behavior. Remember we are trying to anticipate a highly variable and transient event (earthquakes) even if you do a PHD thesis on a single building you still wont get the exact right behavior.

What you do want to worry about is load path, details, big discontinuities or changes in mass/stiffness. Whether you are 100% about the magnitude of the loading or not as important as providing a load path that is well thought out. For complex buildings different engineers might come up with different level of loading, but they better all have a diaphragm, shear walls, etc.

The last crazy custom home I did had a myriad of different roof planes and slopes. It required a lot of detailing and consideration for load path.

Residential wood frame construction has seen very little damage to earthquakes since we started considering the seismic load path in our designs. This is a reason that many designers are not so concerned with doing a time-consuming and sophisticated lateral analysis. I know that not everyone sees it that way.

phamENG · Dec 22, 2022

I totally agree with what bones said - you don't have the right tool for the job, but you're not trying to drive a nail with a wrench here...you're trying to pry something apart with a claw hammer. It can do the job, and was even designed to do the job, if you used properly...but a crowbar would be sooo much easier and more efficient.

I'm out on the East Coast and seismic isn't a concern (you can feel the ground shake here once every one or two generations). All my houses are SDC A. So I can't help you too much with seismic specific questions. We do, however, have hurricanes, so I pay close attention to lateral design all the same.

1) I generally agree with this. Either prove it works through engineering, or prove it fits the mold of the code in question (assuming you can apply some sort of prescriptive design to your project). A lot of the stuff in the code really doesn't "calc out" but nobody wants to be the guy to say that the millions of tract homes we've built that way are going to collapse any second...because they won't. I also agree that a lot of the engineers in this space aren't doing a great job of it. You can 'wing it' pretty easily on the tract homes...custom homes need some real engineering. Unfortunately, contractors think that the engineering for a 10,000sf custom home should be the same as a 1,000sf tract home.

2) I mostly use hand calcs and spreadsheets. This is a little easier with wind as the loading is a bit more simple than seismic. If you really want a tool, RISA does a decent job with wood framing. You'd need to combine 3D and Floor to get a complete package with shear walls, diaphragms, and framing.

3) You need to strike a balance. Efficiency in design isn't exclusively aimed at minimizing construction labor and material. For the sake of you firm, the design itself needs to be efficient. Now, doing less simply to do it in fewer hours is not efficiency. But you do need to figure out what really matters. There are lots of connections and framing members that you don't need to give special consideration to. Figure out what those are, and either remove them from your workflow (develop a standard that works 95% of the time and make sure you know when the 5% happens) or at least cut back on the time you give it.

4) Study failures of "conventionally" framed structures in your practice area. How did they fail? Are you paying attention to those details? Paying attention to all of it is important, but make sure you're getting this low hanging fruit. Soft stories are a big one in seismic zones, I believe.

For your projections with seemingly no LFRS - keep in mind that wood diaphragms can cantilever. I know there are more stringent requirements for seismic, but make use of subdiaphragms and drag the loads back in to the primary LFRS. Where I am, because diaphragms don't come up in the prescriptive residential stuff (the floor just...is), a lot of people seem to forget that wood buildings have diaphragms at all. If you are willing to put in the time to detail those diaphragms properly, they can take a good bit pressure off in some of the unique arrangements.

SarBear · Dec 22, 2022

Thank you bones206 and driftLimiter for your responses.

driftLimiter said:
Hand calcs and engineering judgement are used heavily for this type of construction in my office.

This is an interesting thing you bring up because I've thought a few times that I could spend an hour on the lateral analysis for a home or I could spend 16 hours on it, but I'd end up with the same design either way. Breaking up the house into small rectangular sections (as bones recommended) and doing hand calcs could take many hours. On the other hand, I could just look at the structure and use engineering judgement and what I call "thought experiments" to determine where the load path will go. Call out shear walls and moment frames where necessary and call it a day. I am torn between doing extremely detailed analysis and calculations vs. doing it by feel. If I call out shearwalls with 3" field nailing instead of 2" field nailing will that be the difference in whether the home stands or not? On the other hand, what was the point of my schooling if I'm not doing 12 hours' worth of hand calcs for these homes? One thing I do know is that none of the other firms around here are doing hand calcs in this situation.

phamENG said:
A lot of the stuff in the code really doesn't "calc out" but nobody wants to be the guy...

Thanks for your thoughts, phamENG. This statement you made right here is what I get hung up on. When I got into engineering I thought it was "science", where we're the nerds at the whiteboard writing calculations and proving that our design is foolproof. The more I get into the field the more I'm realizing that it's not like that at all. As an example, on a big custom home I did this summer I called out an 8' long shearwall in the basement of the home. I didn't think it was necessary, but the wall is there and it made me feel better to make it a shearwall, so why not use it? Without asking me the builder decided to put a 6' wide doorway in the middle of that wall instead. The inspector called him out on it. The builder wanted me to say that the opening is ok since it would take them two days and cost them $1k to remove what's there and redo it correctly. I can't say with any degree of confidence that the shearwall NEEDS to be there. On the other hand, I can't confidently say that it DOESN'T NEED to be there. I hate feeling that way, and yet I feel that way on basically every big residential project. I am lacking confidence. None of my designs has ever been tested by an earthquake so how can I possibly be confident that what I've engineered is good work?

driftLimiter · Dec 22, 2022

Lacking confidence is normal, it comes with experience and it comes slowly depending on who your working with. Really you need to look towards your mentors to be able to confident. Like phamENG said, understanding what failures or issues have been experienced in your area with your type of structures goes a long way.

In all I would say its normal to feel how you are, I definitely have and still do feel this way from time to time. Engineering is more art-form than may you realize. Coming up with a reasonable load for a given element takes practice. My boss can pick any element in a building and come up with a load in a few minutes. But he knows whats reasonable because he's being doing for years.

What I can say is that it is a very satisfying thing to have gone over a structural design and its details and feel confident in what you have proposed. If you feel you are pressed by others to be 'efficient' but you also want to do the best analysis/design as possible then it's on you to figure out how to strike the balance.

And finally its better that you are thinking about this and asking these questions than blindly accepting the standard approach, this is one of the things that makes a good structural engineer imho.

bones206 · Dec 22, 2022

phamENG said:
but make use of subdiaphragms and drag the loads back in to the primary LFRS

Just continuing this thought, you could implement this strategy in your spreadsheet by adding a manual input for "additional shear" in any given shear line. That way you can account for complex load paths with a relatively simple implementation. It takes some manual input and planning, but at least that way you can justify your analysis as rational.

SarBear said:
I am lacking confidence.

I think most conscientious structural engineers experience this. It's just part of the gig. Based on this thread, I think you are on the path to becoming a solid engineer. Just have to go through to process to build that experience base so you can use it to balance analytical certainty with pragmatic efficiency. Leaning on the experience of Eng-Tips folks has always helped me when I'm not feeling confident about something.

phamENG · Dec 22, 2022

SarBear - not sure if you're familiar with it, but a popular trope on this board is the Dunning-Kruger effect. All I can say is: welcome to the Valley of Despair. Don't fret too much as there is hope.

Though directed at driftLimiter, I'll comment on the first paragraph first. 3" vs. 2" nailing could be the difference. If it couldn't, we'd never use 2" nail spacing. But to get past the 1 hour vs. 16 hours....do both for the same project. Hopefully your employer will sponsor such an undertaking, but if not you may need to self fund your experiment. Do a house or two on the fly, spend an hour or maybe two on the lateral design and see what you come up with. Then do the same couple of houses in excruciating detail. Compare the results. Where is your intuition and judgement correct, and where is it wrong? Make adjustments and refine your methods incrementally. You may never get to 1 hour production design time, but hopefully you can avoid spending all week on it.

SarBear said:
When I got into engineering I thought it was "science", where we're the nerds at the whiteboard writing calculations and proving that our design is foolproof.

You're spot on and dead wrong at the same time. There are areas of practice where this is the case. They're generally in larger buildings with a higher consequence of failure. Those building are also more commonly built in steel and concrete. Frame structures tend to have fewer 'accidental' redundancies, so consequences of individual failures are much lower. We also have a much higher confidence in material strengths, so our factors of safety are a lot smaller. So in these cases, nerd out to your heart's desire (within budget, of course). But with a light frame wood structure...that level of detailed analysis is generally relegated to the halls of academia. The level of complexity is higher, but with that comes more redundancy in most cases. My beam header is undersized? That's okay, it's nailed off to sheathing that actually acts as a 3' deep plywood box beam. My shear wall hold down wasn't installed properly? I didn't account for 60% of the walls to resist shear even though they were built the same way but instead of a hold down they have ample dead load on them, so the shear capacity of the building is actually 130% of what I thought. Oh, and because wood is a naturally variable material, we apply our factor of safety to the 5th percentile strength. So before we even reduce the strength with a FoS, the wood is already stronger than we think 95% of the time. So the inherent level of conservatism lends itself to...liberties in design.

This is where a purely scientific engineer will run into problems. We're not in government funded labs designing buildings. We're in competition with others who are trying to design them faster and cheaper. Those liberties? Unscrupulous engineers will take them to cut corners and steal your clients. It sucks, it's wrong, and the bankruptcy judge won't care. So it's incumbent on us to find faster, more efficient ways of doing the same thing while maintaining a higher standard and protecting public safety.

Aesur · Dec 22, 2022

We do a good amount of custom homes similar to what you describe and most are located in SDC B, and in a few cases seismic does govern over wind. We tend to do all this with a combination of hand calcs and EnerCalc shear wall module for the shear wall and footing designs. For the shear walls, being that these are mostly flexible (wood) then I tend to distribute loading based on trib area to the shear wall. I then look at diaphragms, typically as smaller rectangles and make sure I have chords and drags as needed to get to said shear walls. I have done a few very detailed analysis to get a good feel and over time you start to see patterns and develop a sense of being able to design just one diaphragm on a project and it is representative for each diaphragm. From there you just look at chord forces for straps/nail connections for each diaphragm, eventually you will see a pattern here as well and can do a few representative worst case designs. Lastly you start looking at drags, making sure you can get your loading into the shear wall. A few ways to do this is to look at the actual loading being dragged from one diaphragm to another and into the shear wall, or you could conservatively take the shear wall loading and drag the whole amount if your straps/drags aren't too crazy. I have had a few where the available shear walls were limited and required a much greater analysis, accounting for interior diaphragms with seismic loading only surrounded by other diaphragms with wind and seismic loading and thereby reducing the drag forces or shear wall forces to be exact, these analysis can take a few days to complete if needed, but this is why high end residential gets the fees it does compared to small commercial/residential. If your company is pricing similar to small commercial/residential then you will never have the budget to design these structures to the extent needed.

Also, don't be afraid to use cantilever columns, moment frames or strongwalls if needed at cantilever diaphragms, they are common for high end residential.

SarBear · Dec 22, 2022

phamENG said:
SarBear - not sure if you're familiar with it, but a popular trope on this board is the Dunning-Kruger effect. All I can say is: welcome to the Valley of Despair.

Lol, yes I discovered this phenomenon a few months ago. I was both hopeful and disappointed when I found out that I am in the valley of despair.

phamENG said:
This is where a purely scientific engineer will run into problems.

I guess I need to get out of this mindset then, somehow.

Aesur said:
Also, don't be afraid to use cantilever columns, moment frames or strongwalls if needed

Yeah I would like to use more of these systems, but of course there's pushback. Especially when they just finished another big home engineered by someone else that didn't have any of these.

As an example of what I'm talking about is this sketch of a detached garage I worked on. When I look at this structure I see a lot of the lateral load going into that center beam where the plate height jumps from 10' to 14'. I proposed that the beam should be steel and form a steel moment frame with W-shape columns at each end of the beam. When my boss sees this structure he has the lateral load ending up in wall lines 1 and 2. I don't understand how that is realistic, and I don't understand how there could even be a calculation to show how the load "jumps" from the lower diaphragm to the higher one, and vice versa.

driftLimiter · Dec 22, 2022

The load is going where the stiff lines of resistance are. You may be correct that the diaphragm shear has to pass through the beam to make up the offset, but the load doesn't just flow into the beam because its there. Those two little wall segments attached to the center beam are very flexible compared to the walls on GL 1 and 2.

But your still correct the load doesn't just magically jump from the upper level to lower level, I would ensure you can get shear transfer boundary nailing from both the upper and lower diaphragm into the beam.

TBH it sounds like you could use some different mentorship. Hand-wavy attention to detail isn't a good look on an engineer imho.

How much detailing are you being asked to do? My firm would certainly require a detail for shear transfer at that offset.

SarBear · Dec 22, 2022

driftLimiter said:
How much detailing are you being asked to do? My firm would certainly require a detail for shear transfer at that offset.

We didn't have any shear transfer details for this. My thought is that the bottom of the beam is at the 10' plate height. The higher trusses bear on a 4' tall wall that sits on top of the beam. The 4' wall and the beam are sheathed on the left side. There should be a truss at the lower garage that rests right up against the length of the beam. The top chord of that truss should be nailed wherever it crosses the studs for the 4' wall. The bottom chord should be nailed periodically into the beam. The lower diaphragm should have edge nailing from the sheathing into the top chord of this truss. That's what I was thinking, is that what your detail would show or did you mean something different? This garage has been built already so I'm wondering what they actually did.

driftLimiter · Dec 22, 2022

I would need to see the geometry but it sounds like your thinking something similar to me.

Edge nailing needed from the upper diaphragm and the lower diaphragm on the same beam. Perhaps a 'ledger' type nailer on the side of the beam, something to finish the sheathing into that is also connected to the beam with the edge nails.

That detail is routine for this type of house.

skeletron · Dec 23, 2022

Yep. It's definitely a problem that there is a lot of inadequate engineering going into housing. On the other hand, there is absolutely no reason to make a house a PhD thesis project especially given the ceiling on fees. There's also false sense of security with using modelling software versus turbo-charged spreadsheets and hand calculations. So, in that respect, it's important to become comfortable with the simplifications of hand calculations before the intricacies of a paid-product. The best you can do is run "some numbers" and know what those numbers represent.

In closing, and in a very overly general sense that is hopefully not misconstrued, I found that there is so much initial focus placed on getting the magnitude of the force distribution when there should be more focus on the detailing for the forces. The computer programs will likely only distribute our forces and give us *some numbers* to play with but ultimately we need to trace those forces through the load path and draw details. I'd rather run "some numbers" and draw "all the details" then vice versa.

Tomfh · Dec 23, 2022

How is the garage stabilised in the perpendicular direction?

XR250 · Dec 23, 2022

SarBear said:
As an example of what I'm talking about is this sketch of a detached garage I worked on. When I look at this structure I see a lot of the lateral load going into that center beam where the plate height jumps from 10' to 14'. I proposed that the beam should be steel and form a steel moment frame with W-shape columns at each end of the beam. When my boss sees this structure he has the lateral load ending up in wall lines 1 and 2. I don't understand how that is realistic, and I don't understand how there could even be a calculation to show how the load "jumps" from the lower diaphragm to the higher one, and vice versa.

I'd be doing the same thing your boss does. Treat it as 2 three-sided buildings in the front to back direction and don't sharpen your pencil too much. That is how most houses end up working and it keeps you from having to worry about things like this.
Let's be honest, how many lateral failures do we see in houses?

gte447f · Dec 28, 2022

My advice is to do 100% of these types of design by hand as you are learning. Do this on your own time if your employer will not sponsor it. Never rely on a computer to do something for you that you do not understand how to do on your own. Custom homes are complicated, but only because they are a terribly inefficient conglomeration of disjointed parts and pieces and systems. The actual engineering of each part is usually not complicated. Once you understand how to accomplish these designs using traditional methods, then you will be able to judge whether or not the software applications are a reasonable tool and if so, by all means use them to increase your productivity if you can.

JoshPlumSE · Dec 28, 2022

1) In my opinion, as engineers we should be able to "prove" mathematically that everything we call out on the plans and every construction detail we draw "works". Am I wrong? If I'm right, then I think that any engineer involved in residential engineering is not actually doing that.

I don't know about proving it mathematically. I think there are times where engineering judgment comes into play. But, you should be able to convince yourself that it will work.

But, I know what you're saying. I did a curved glu-lam bridge (very small) a few years ago. I analyzed it as a straight glu-lam. The curve was gentle and the analysis was much easier. I think it was okay, especially since the resulting glu-lams were not all that highly stressed. Now, that was a single variable where I was able to convince myself the shortcut was okay.

In your case, there are so many items in one house that it compounds the concerns over and over.

3) Do you have any advice for someone just starting out in this field? I'm starting to feel a little disappointed in my career choice. In school we focus so much on calculating things and getting the right answer. In the field things are never as simple as they were in school, but even so it feels like there are a lot of things that we continue to do because they've always been done that way even if they don't "calc out", or we run these lateral design numbers just to say we did it rather than really trying to understand the true behavior of the structure.

No advice, just my sympathies. Residential engineering often sucks. Way more work with way more complications than you're really getting paid for. I really enjoyed doing work on commercial buildings. Tilt Up, concrete, concrete masonry in combination with wood roof or steel columns or such. I always felt like there was a reasonable solution and that the company was making money on the projects.

4) Any other thoughts about lateral design on residential structures?

For residential, I think the three most important things to keep in mind are load path, load path and load path. If you've really thought through your load paths in these complex regions and made sure there is good detailing for load transfer in those areas then you are probably doing a good job. The forces we design for in seismic are no where near the forces the structure will actually see. Right? We've got to make sure that the structure is allowed to fail with ductility. That means focusing on connectors, load paths and such. Over design those things and you will get ductility. If you don't properly think out the load path, the ductility goes away, no mater what forces you designed for.

In the end a structure that has code checks on the diaphragms that show a 40% overstress (due to a calculation error / over simplification) but is beautifully detailed will probably do pretty well in a seismic event.

Sometimes you might add some simpson connectors in locations where there the top plate jumps.

Or, maybe look at the roof as a sub-diaphram type of analysis with straps to ensure a continuous connection across the entire roof and all the way to the other side.

Maybe you design the cantilevers that pick up a bearing wall above with an Omega factor.

In your garage example, I think there are ways to do the detailing so that middle line doesn't get a ton of shear. Though I think it begs for a strong wall or something on either side of that middle line. Making sure you got something like drift limiter suggested to get force transfer from both diaphragm heights into that center beam. Personally, I think that's just easier than trying to pull all the load out to the far walls.

milkshakelake · Dec 30, 2022

3) Do you have any advice for someone just starting out in this field? I'm starting to feel a little disappointed in my career choice. In school we focus so much on calculating things and getting the right answer. In the field things are never as simple as they were in school, but even so it feels like there are a lot of things that we continue to do because they've always been done that way even if they don't "calc out", or we run these lateral design numbers just to say we did it rather than really trying to understand the true behavior of the structure.

You're in the right career. Besides maybe being an astronaut or CEO, this is a cool ass field.

If someone else is signing and sealing it, I don't think it's worth stressing out over it. Around when I started, I designed lateral systems for ~100 apartment wood buildings with extremely complex load paths, and I was 100% sure that everything was wrong because we analyzed it like a box. But the PE I was working under insisted I use that approach. I lost quite a bit of sleep imaging the building failing and crushing people in a lateral event, but got over it when I realized that it wasn't me who made the decisions. The boss needs production, and they want to use the super fast tools that they have.

And I think skeletron makes a great point. Correct detailing will iron out a lot of load path issues that are missed in the simplified calculations.

phamENG · Dec 30, 2022

milkshakelake said:
If someone else is signing and sealing it, I don't think it's worth stressing out over it.

MSL - I like you (I think - what does it mean to like an anonymous, nameless person on the internet?), but I disagree with this statement so strongly that I feel the need to break my vacation eng-tips fast to rebuke you for it.

From the beginning, every junior engineer should take as much ownership as they can in whatever they do. Of course that needs to be tempered with said junior's level of knowledge and experience. But if something feels wrong, the boss should be teaching the junior why it's right. Not justifying the business side, but showing why the engineering makes sense. If they don't, find a mentor who will.

That's part of the reason I am where I am. I got sick of insufficient answers to such challenges. I realized that if I have to teach myself anyway, I might as well get the profits from it! (That's only partly true - I learned a lot from the firm where I got my initial experience and have some good mentors now...but the 'just do it because that's how we do it' got old.)

EDIT: MSL - flip this on its head. You own a firm now. How would you feel if you found out that one of your junior engineers knew something wasn't quite right but decided "f&#@ it, I'm not sealing it..."?

Eng-Tips is the largest engineering community on the Internet

Intelligent Work Forums for Engineering Professionals

Residential Lateral Design Philosophy 3

Structural

Recommended for you

Structural

Structural

Structural

Structural

Structural

Structural

Structural

Structural

Structural

Structural

Structural

Structural

Structural

Structural

Structural

Structural

Structural

Structural

Structural

Similar threads

Log in

Part and Inventory Search

Sponsor