What are the limits of prescriptive design? 9

[DTS419]

The IRC, as we are all familiar, provides prescriptive design standards meant to cover common construction of “one and two family dwellings and townhouses up to three stories.” Examples of this prescriptive design include connections such as from wall to foundation, headers over openings, etc.

But what are the limits of these prescriptive provisions, and whose responsibility is it to identify them?

Let’s take a closer look at connections to foundations, for example. It’s not uncommon for large custom homes that fall within the IRC’s scope to have finished basements with tall ceilings resulting in deep foundation walls with significant unbalanced soil load. There can also be significant uplift loads that must also be transmitted to ground depending on the proportions of the superstructure. These forces can easily exceed the capacities of the prescriptive provisions, that, if I had to guess, were developed long ago with simpler construction in mind.

It’s also not uncommon for many home builders to skip architects and engineers and simply follow the IRC. I’ve seen too many projects where this happens, and the result is connections that are over capacity, lateral systems without adequate diaphragm and shear wall detailing, etc. This often doesn’t result in total failure, but rather a final product that doesn’t meet current standards of practice, making it hard to call the builder’s attention to flaws with “the way we’ve always done it” that might be code compliant but not necessarily sufficient. And of course, failures can and do happen in the worst cases.

So what mechanisms, if any, are in place to ensure that simply following prescriptive codes are adequate for every situation, and whose job is it to identify when an engineered design is required? And, who is responsible if a code compliant prescriptive design ultimately proves to be inadequate for the situation?

 

[phamENG]

Who's job? The AHJ and whomever is doing the design.

What are the limits? If it's not in the tables or spelled out, it's not covered.

 

[jerseyshore]

I can only speak for NJ, but around here generally the building officials are looking for the plans to match the IRC's prescriptive tables when submitted by a homeowner/ GC. When there is engineered lumber, steel, uncommon details, or something that doesn't quite fall into the IRC prescriptive sections, they require an architect or engineer's seal.

Obviously these days very little house construction doesn't use engineered lumber, so for smaller or simpler beams or additions, most GC's go thru the lumber yard and have Trus Joist send them a calc to submit along with the plans. The building officials approve that typically.

Besides shore towns (and only some of them), I've never heard of a building official request a braced wall plan. And regardless of location, I've never heard of a town requiring engineered shear walls.

The way I look at it is if they ask us to do the lateral, they get the engineered shear wall design. If they want to do it themselves they can do braced walls or nothing, don't matter to me. Same for the other engineered vs prescriptive elements in the IRC. We're engineers so we sign off on engineered designs. Prescriptive is for the archs/ GCs/ homeowners.

I think for anyone here that has worked in NJ it's a bit unique since there are a zillion building departments and officials and what one town will approve the next one will reject. I'm not sure what happens elsewhere in the US or what's supposed to happen, but that's how things have worked in NJ as long as I've practiced.

 

[lexpatrie]

So what mechanisms, if any, are in place to ensure that simply following prescriptive codes are adequate for every situation, and whose job is it to identify when an engineered design is required? And, who is responsible if a code compliant prescriptive design ultimately proves to be inadequate for the situation?

Rather a broad question.

The responsible parties are the G.C. etc.

This is why I have a lot of problems with AISC and ACI and the truss design people saying "design professional" when they mean engineer of record, and saying "building designer" when they mean architect, professional engineer, or random dude with a ruler (i.e. G.C.).

Traditionally when there's not really major malfeasance on the part of the building department (I think it was Nebraska), in the legal sense, they aren't actionable on liability.


Municipal Liability for Negligent Inspections in Sinning v. Clark - A 'Hollow' Victory for the Public Duty Doctrine, Campbell Law Review, Swindell, Jan 1996.

Who Is Liable when a Deck Collapses, Dermer Appel Ruder, LLC, Feb 22, 2021

What happens when a prescriptive design fails? Usually there's a deviation from the prescriptive design that's contributing to the problem, so, again, back to the G.C. (G.C. is supposed to supervise work and ensure work is done correctly before passing to the next sub) or "building designer" "specifier" or whatever the fuzzy terminology that is used says.

e.g. Shingle blow off? beyond high wind events, how were they nailed, install temp, specific instructions from manufacturer (hand sealing required), that kind of thing.

Ice dam? Was an actual Ice Barrier installed or just something that looks like an ice barrier, was it installed too cold, was it not rolled down per manufacturer requirements? Installed on a not-approved substrate? Was not enough installed, up the slope? Insufficient overlap of the layers? reverse shingled?

If you mean the light-frame provisions, I'm not sure I've seen an "in IRC" design fail like that. Window installs? Defective WRB, non durable construction of the sills, sure.

eta. Montana actually.
Montana City Loses Appeal in Lawsuit over Deck Collapse, Claims Journal, Feb 27, 2009.

 

[Eng16080]

I think for anyone here that has worked in NJ it's a bit unique since there are a zillion building departments and officials and what one town will approve the next one will reject. I'm not sure what happens elsewhere in the US or what's supposed to happen

Within a 2 hour radius of my location in NH we get the full spectrum, from building departments requiring an engineer's stamp on every house to towns that don't have a building department or any form of oversight.

 

[Mike Mike]

ICC needs to state the limits of applicability. If I were in charge of their committee, the screenshot below is an example of some mark ups I would have for my committee members to pick up. I would have similar comments for all prescriptive design tables including foundation connections.

If code compliant prescriptive design ultimately proves to be inadequate in a jurisdiction which has adopted IRC, contractor would not be liable if they followed the structural drawings, structural engineer of record would not be liable if they followed local laws, building official would not be liable if they acted "in good faith", and ICC would not be liable since it's a voluntary standard. Please correct me if I'm wrong here tho people, most of my experience is in larger commercial projects and I'm not as familiar with IRC.

 

[lexpatrie]

I'm not aware of any failures specifically attributed to the IRC itself. There are things that do not calculate per an engineering standard, this is an accepted fact, however.

Could you build something per code that failed "on paper" in engineering? Perhaps, maybe even probably, as you've noted on the stud chart above. I think a lot of the limits you're wanting on the table above are elsewhere in the code, as most engineers don't interact with the IRC all that often, there's a whole sequence of provisions and limitations you have to pass through to reach those tables. (Wind speed is 115 mph Ultimate, 90 mph allowable stress, for example. The IRC gives up at a certain level of seismic activity as well). The spacing and depth limitations on prescriptive floor systems restrict the loads that can reach a wall, etc. It is all interrelated and the pieces all interact with each other. Roof trusses, Floor trusses, I-joists and other similar systems may exceed these span limitations and void the stud table's use, for example. You might consider using the stud tables with floor trusses or I-joists if they fit the span limitations for joists, but it's an individual choice (by the engineer).

If code compliant prescriptive design ultimately proves to be inadequate in a jurisdiction which has adopted IRC, contractor would not be liable if they followed the structural drawings, structural engineer of record would not be liable if they followed local laws, building official would not be liable if they acted "in good faith", and ICC would not be liable since it's a voluntary standard. Please correct me if I'm wrong here tho people, most of my experience is in larger commercial projects and I'm not as familiar with IRC.

First off, this presumes a structural engineer, this is rare. Very rare. Structural engineer of record is also rare as hen's teeth. Maybe on larger palatial estates and what not, sure, but 99.9% of all residential construction is a) older housing stock and b) not engineered, even the McMansions, in my experience tend not to have an engineering involved, even for larger more convoluted structures. I feel like there are people on this forum who specialize in the really grandiose "residential" structures. I don't. I could, but I don't advertise and these people don't seek me out.

Most typically there's either a general contractor, or potentially a "building designer" who is as formally trained as your average person. No engineering formal education or background. Think of the Brian Fellow's Safari Planet sketch on SNL. Fellow's isn't a zoologist, he just loves animals.

There are also circumstances where there actually isn't a residential code (Kansas, I believe, doesn't have one).

Not that I'm an attorney, but..... the burden of proof is on the plaintiff (complaintant), so they have to prove that the house was constructed negligently, out of compliance with the IRC, etc, (And that is caused the alleged damages). Whether that involves the expertise and expert report of a general contractor, civil or structural engineer, architect, associate AIA (which is somebody who likes architecture enough to send a fee to AIA, there's no tests involved, again, Brian Fellows), interior designer, painter, plumber, framer, or (independent) code official is a question of legal strategy for the plaintiff's (complaintant's) attorney.

The "if code-complaint prescriptive design ultimately proves to be inadequate" is the key item here, you're presuming the building is constructed completely to code, this is rarely the case, particularly when the building has problems. The last time I looked at the requirements for studs, they wanted a modulus of elasticity that was only available in one species because it was so high. So if you have deflection problems, you will rarely be able to credibly argue the code was followed as the E wasn't met to use the tables. One example.

"We did nothing wrong", "We followed the code" and other similar concepts "...it worked fine until it broke" are what I'll (as a layperson) call affirmative defenses. It's admitting something happened and that you did it, but it's not your fault. You *(the defendant, contractor) are putting a burden of proof on yourself to prove it was adequate. Installed correctly per code and manufacturer's installation instructions, etc. As most houses are not constructed and documented during construction to the level of detail one might hope for should litigation arise, there's some measure of difficulty with this approach. (i.e. little to no evidentiary foundation for these expert opinions). The same applies for the plaintiff, as some things need to be taken apart ("at risk") to see if there's actually an installation problem, versus speculating.

As you outline it, the structural engineer would be expected to conform to the standard of care NOT the IRC, (this is a nebulous term, but that's what it is), but blindly following the code and refusing to do calculations would not be wise practice. Where you draw the line in a residential review (which I've participated in previously) would, I presume, vary, probably wildly, between various professional engineers. Some elements one would check compliance (i.e. braced walls), and disclaim designing them, other elements you'd at least have to admit are outside the code and either provide the engineering, if qualified, (steel beams, steel columns, LVL header, spread footings or walls that are taller than 10', bay windows, oriel windows, exterior decks that "hang" on the structure, roof helipads, etc.) or indicate who is needed.

As to standard of care itself, I think Josh Kardon is about the only person who publishes articles on this subject. Some of the earlier literature (Oulette), Morcarz appeared after the Hyatt Walkway collapse and are more-or-less a direct reaction (or defense, I suppose, depending on your viewpoint or how the article reads).

I think Bruce Barnes had a decent article that was particular to residential construction a decade or so ago.

Building officials receive some measure of immunity to lawsuits but this has been eroding over the years and it depends on the state, (read Sinning, it's really old but still informative) my impression is since the building official is on the site before the homeowner, it would be the general contractor who has to bring that aspect of the claim for damages. (No city or contractor seems to have invoked that defense however). Again, not an attorney. The more egregious the errors or negligence the more likely there is legal exposure (i.e. The Hard Rock Hotel the building official actually faked inspections, as I recall, but I'm not clear the city or department knew they were faking the inspections, it just came out later, I don't recall the circumstances of the Montana deck collapse).

Just because there's a building official doesn't absolve a general contractor of conforming to the code. Even if it passed inspection.

Good faith, to me, (or more accurately "bad faith") is more of a term I see in insurance and property liability. I don't think the accuracy of the term here affects the discussion much, but it would more likely be breach of contract or breach of warranty, negligence per se, some other negligence, or fraud, or whatever the correct legal term is depending on jurisdiction.

This is a pretty deep subject, frankly, or maybe it's more abstruse than deep. The States all have different roadmaps, legal reasoning, precedents that are unpublished or not binding, etc. I'm not convinced any state has something that fully matches another state, it's not like there's a Seven States Standards for it. (Actually it's TEN STATES, anyway). And then you have Louisiana which is a totally different legal landscape in the foundation (French).

Basically any conversation on this subject has about fourteen hundred exceptions, caveats, and footnotes.

 

[Mike Mike]

Lex, this is some great reading, but I couldn't get thru the paywalls on a lot of this, and I definitely struggle with some of the legalese in Sinning. Could you recommend any intro to structural engineering law books? How did you learn? I realize you're not a lawyer (Brian Fellows, that goat is weird) but you seem to know some stuff.

Where do you see the 115mph wind speed limitation? I spent about 10 hours digging thru the 2018 IRC at this point and the best I can find is figure R301.2.5(B), which permits use of the IRC over 95% of the country, but requires "wind design" along the coasts. Language also appears to have been added in 2024 to preclude IRC use in special wind regions where wind speeds exceed 140mph.

Where do you see the required stud modulus of elasticity? R602.2 prescribes no. 3, standard or stud grade lumber.

Similarly chapter 3 appears to preclude use of the IRC where ground snow load exceeds 70psf or floor dead load exceeds 10psf (with some exceptions).

Tabulated rafter spans per R802.4 go up to 26', meaning 52' between exterior bearing walls. Truss spans are limited to 36' by R802.10. Tabulated joist spans per R502.3 go up to 26'. Floor trusses are permitted and I could not find any limitation on span.

Given these limits, the 2x4x10' exterior stud wall I mentioned above doesn't calc out even if I cut all the loads in half. Doesn't seem to be in anyone's interests (contractor, engineer, public safety, building officials) to have wildly varying reliability factors from one code to the next. When you say "blindly following the code and refusing to do calculations would not be wise practice", are you saying calculations are required to determine if you agree with the prescriptive designs of the IRC tables? If you were to specify a 2x4x10' exterior wall in a 140mph wind zone from the IRC tables, are you saying that you would not be acting according to the standard of care?

DTS, back to your original example, R404.1.3 limits foundation walls to 10' and limits backfill to 60psf/ft clay. R403.1.6 requires anchorage of sill plate to top of wall with 1/2" bolts at 6' OC. Is this what you're referring to? I agree this is yet another example where the IRC is wildly outside the requirements of other standards. The ICC committees appear to be struggling and are in need of our help. My recommendation to them would be to consolidate and clarify all IRC use prerequisites, and to rewrite their prescriptions to be within a factor of 1.5 of compliant with other standards.

 

[gte447f]

I'm with phamENG on this. It's relatively straightforward. If it's not covered by the prescriptive code, then it's outside the limits of the prescriptive code.

It's the responsibility of the AHJ to make sure that construction complies with local laws, rules, regulations; in this case, with the prescriptive building code. As far as liability, however, the AHJ probably won't be held liable for nonconforming construction. If someone has been paid to design it, I would say that person or entity would be responsible and liable for either complying with prescriptive requirements or designing to other accepted standards. If there is no "designer", then I would say that responsibility and liability falls on the builder. The responsibility and liability of the designer and/or builder to conform to the code may be explicitly spelled out contractually, or it may be understood to be implied by the nature of the contractual relationship.

Now, the fact that my desk copy of the 2018 IRC is 964 pages long admittedly adds some ambiguity to something that should be relatively straight forward.

 

[lexpatrie]

Those articles you can't get to aren't going to transform your life. There are ways around paywalls but you can't learn them from a Jedi. It looks like AISC Engineering Journal has released a ton of documents from their paywall, by the way.

I've done a fair bit of construction defect work (mostly building envelope, bad window installation), and the rest is less formal, as far as how I learned about things.

There used to be a Legal article in Structural Engineer Magazine, and Matthew Rechtien has done a few articles in Structure Magazine, and teaching some ethics classes helped too. I read a lot of disciplinary actions, too, I feel like you can learn a lot there, like the unusually specific nature of some of the Florida requirements.

Tabulated rafter spans per R802.4 go up to 26', meaning 52' between exterior bearing walls. Truss spans are limited to 36' by R802.10. Tabulated joist spans per R502.3 go up to 26'. Floor trusses are permitted and I could not find any limitation on span.

I'd say floor trusses don't have span limitations in the IRC because they're an engineered system, so the P.E. on those is in charge of the "span" limitation (via depth, plate forces, connections, bearing, and deflection), so there might not be any hard span limits in the IRC for them.

I don't think that means 52', because the hand-framed rafters require a ridge (beam) board at the peak and then there's a requirement for a purlin brace that has to go down to an interior wall. Take a look at Figure R802.4.5.

The more I look at that figure the funkier it gets. If that's really a ridge board up there, it's not designed for downward load, and that means the rafter they show cantilevers past the purlin brace. I would say these figures are exceedingly old and haven't been revised or amended in ages. For some reason that guy at Reid Middleton who submits about five million code changes per cycle hasn't gotten near this part of the code. (I guess I've been looking at this wrong, Reid Middleton is representing the Washington State Building Officials so the code amendments come from them.)

Anyway.

For the E requirement, it was in 2000 IRC, Table R602.3.1, footnote b. It's probably moved around, or maybe it got deleted at some point. If you really want to track something, you have to go code cycle by code cycle and look for it because there will be a deletion mark if it gets deleted, a "bar" if it moved, etc. So it can be really time consuming.

As to the prescriptive design limits, you're allowed to use them, but it isn't exactly doing engineering, if you ask me, and as you've pointed out, sometimes the engineered calculation will fail where the prescriptive code will permit it (the IRC has a relatively large bending stress increase, that might be part of your issue with the 10' stud wall). It's mentally weird to me that the prescriptive code has .... you know what I mean, as you say, less reliable/robust elements in it, then the engineered portions are designed more, to a more specific set of parameters. I think there's "wiggle room" to design it for "less" than the full code, but it's an area of practice I suspect few engineers would delve into, trying to determine a rational basis for the loads out of the IRC and then conforming with that back-calculation, for example, the engineered design would be no less dangerous than the prescriptive design. It's perhaps just not worth anyone's bother to try to do it that way, so nobody does it. I've not seen any disciplinary actions where an engineer got that adventurous, but the Florida code does have language where you deviate from the code and specifically explain why, so it's an avenue. I think it's fairly risky (if not more) to try it that way, however. Not for me. Back to your point, I've not seen any engineers get a disciplinary ruling where they followed the IRC.

What I would posit is "normally done" (here is where we land on standard of care, that mercurial little pixie) is engineers who perform reviews of houses check the things in their scope of work that are obviously outside the IRC, (so LVL, large point loads, tall studs, steel beams, pad footings, that are outside the code), based on the typical residential loads (40 psf for a non-sleeping room, 30 psf for a bedroom [if they get that detailed ], and so on.

If there are areas that don't conform (a 10'-3" 2x4 stud wall, for example) they may not notice them, but as to exposure, legally, it also depends on what they have in their contract/review letter language, "This house conforms with the IRC provisions for structure" versus "we did not review the house for overall conformance with the IRC which is the responsibility of the building designer" or something similar. Overall conformance with the IRC most engineers don't do (so, braced walls you may note some deficiencies and see if they want you to engage, otherwise say "you need to show the braced walls" and it's done), they might notice floor joists are overspan, but I think that would be incidental, and a lot of production homes are floor trusses or I-joists and outside the tables, but with their own engineer and or product standard that's "pre-engineered" (I-joists). (This starts to sound a lot like the Fabricator's responsibility in the AISC code of standard practice, if they notice something they're supposed to report it, but they aren't on the hook explicitly for issues). Ultimately if you produce drawings and seal them, what is on your drawings is fodder for a "you knew about it", at least potentially.

In some circumstances, for a legal complaint to go forward against a professional (attorney, doctor, professional engineer, architect, etc), they need an affidavit from an expert that there's a problem (bad design, major error, negligence, lack of competence, standard of care), others (construction defect) need proof of damage/deterioration due to the defect, so it varies a lot based on the damages and the specific situation, and if the actual dispute or whatever you call it isn't specifically about the design work (i.e. a delay claim), that doesn't usually apply.

For engineers "standard of care" is pretty poorly defined, so an opposing expert to say you deviated from the "standard of care" seems to be the general trend. As you are likely aware, that sort of thing isn't all that rigid, it's not like there's a requirement for a calculation showing that the design is 1.15% overstressed or 133% overstressed or something, there's no "hard line" in the sand, one would perhaps hope that an expert would not say 1.01% overstress is cause for alarm but who knows. The state boards don't have this kind of legal obstacle, however. TBPE has advisory panels that will make recommendations, versus something perhaps more arbitrary.

Now, if something is blatantly wrong, a 16' 2x4 wall, that would be another story, potentially, or say a 4" thick masonry foundation wall supporting three stories of house above it, but even so that plan goes to the city, AHJ, and they are supposed to be experts on code conformance. I would expect better work from a building official on the residential/wood construction side of things, versus, say, a building official who's running steel composite beam deflection checks on a 10 story hotel. Or checking the allowable span of metal deck.... (I'm talking about the Hard Rock Hotel, here). It would depend, of course, on the work they routinely do, the Miami-Dade plan check folks seem to be crazy skilled, same with Las Vegas and Los Angeles, and Davis. Even there, though, I'm not convinced I'd expect a building official to notice something like that). Then again, I feel like a lot of building officials have backgrounds as electricians and don't know structure all that well.

Windspeed? IRC 2000, R301.2.2.1 had that 110 mph limit and forced the use of alternative documents above that (the Wood Frame Construction Manual goes up to pretty high wind speeds for the basic structure). Mentally I thought it was 90 mph, then I converted it to the "modern" 115 ultimate, so that's probably on the low end of what the IRC is intended for.

If there is no "designer", then I would say that responsibility and liability falls on the builder. The responsibility and liability of the designer and/or builder to conform to the code may be explicitly spelled out contractually, or it may be understood to be implied by the nature of the contractual relationship.

There's kind of always a designer, but it's not all that restrictive, if you've read the TPI standard anybody who sends them a layout is the "designer". Usually the General Contractor has a license (excepting, say, Indiana, I think), so there are obligations there. "Workmanlike fashion" is one of them, otherwise there's a lot of implied warranty in there, as to the building purchaser versus the contractor. I'd be surprised if a contract contained language like that, but most houses are built on spec, and there's no construction contract anyway. There's a sale contract, probably, but not some kind of performance contract unless it's a much higher end kind of house. The licensing requirements and whatever statutory warranties apply, regardless.

[ again, not a lawyer, not to be construed as legal advice.] Think of it as general information and if there's a specific case, be aware that almost any general information is not going to apply to a specific case. This is the sort of stuff you can ask the attorney about to confirm or amend, that's all.

 

[phamENG]

here is where we land on standard of care, that mercurial little pixie

Haven't read your whole post yet, but I want to say...this is certainly the best line I've read on this forum.

 

[Mike Mike]

lex, my interpretation of R802.4.5 remains as I previously stated, 52' max building width. Purlins are not required so long as the tabulated rafter spans are not exceeded. Difficulty in ordering a 26' long 2x12 not withstanding, just following the letter of the code. Let me know if you disagree.

For the modulus of elasticity requirement, do you mean table R602.3(1)? footnote b just says "Staples are 16 gage wire..." In light of code updates over the last couple decades, are you still of the opinion "building is constructed completely to code...is rarely the case"?

I have had similar thoughts to you about "adventurous" "back-calculation", but the way I would describe the foray is a bit different. Instead I would say as the engineer of record on a custom home, we are permitted to specify prescriptive design where prescriptive is cheaper, and we are permitted to specify engineered design where engineered is cheaper, mixing and matching just so long as our engineered design of one portion does not void any IRC prescriptive design limitations on other portions. Structural engineers well-versed in both IRC and IBC should have a big advantage over the competition. Having multiple options for meeting code should keep their clients happier than if restricted to either IRC or IBC. But are we acting ethically specifying prescriptive designs such as OP's example sketched below, knowing it's way outside the limits of other standards? The soil load on each bolt is 6 kips. The capacity in HF per NDS is 340 lbs. Demand/capacity = 18 even neglecting the effect of in-plane shear load from wind. If I were the ICC I would try to make my standards come to within a factor smaller than 18 of giving the same answer. But maybe I'm missing something here, I'm not an IRC expert.

gte, agreed. What do you think about DTS's other points? "who is responsible if a code compliant prescriptive design ultimately proves to be inadequate?" "These forces can easily exceed the capacities of the prescriptive provisions"

 

[phamENG]

I'd say basement walls are one of the biggest problems in the IRC. We don't have basements in my area (for the most part...there are some intrepid souls here that want to try their hand at building a boat under their house...), but from what I've seen of other posts here and general research, it seems that 'basement repair companies' are as ubiquitous in areas with basements and cellars as 'foundation repair specialists' are here with our soft clays, high water table, and frequent drywall cracks.

The IRC is fairly robust, if you take the time to understand it (with a few gray areas, as you've pointed out), until you get to the foundation. Then, it's just sort of a shrug and "I guess we need to do something here" approach. Some states have refined it - I know North Carolina has a "High Wind Zones" for coastal and beach areas that improve on foundation anchorage a lot - but the 'vanilla' IRC is pretty quiet.

 

[StrEng007]

I have no comments to the OP's question, but this thread has generated a question of my own that is of value to the topic of prescriptive design.

As some have mentioned, residential design often follows prescriptive design which doesn't not require detailed engineering work. I happen to practice in an area that goes against the norm. Having my of my work situated around Miami, FL, I'm typically outside the limitation of prescriptive design. Simply put, almost EVERYTHING gets engineered, reviewed, questioned, and disputed.

I'm interested in learning more about work that lends itself to light gauge framing in areas with decreased wind loads (115 ultimate wind speeds). Being that I haven't spent a whole lot of time doing light gauge framing, the subject feels very academic to me (ie lacking a feel for what is commonly done). Adding to that, prescriptive design feels cheap and impersonal (that's a really silly way to explain it but I'm struggling to find another expression). By my own method, I can get by and issue drawings no problem. My problem is that sometimes I feel like I'm using a sledge hammer when a framing hammer works just fine.

What are your opinions on the prescriptive designs presented by the IRC, Wood Framed Construction Manual, and ICC 600? Is there a whole lot of value in spending the time to learn how to prescribe those designs and try to recreate the calculations by which those tables work? Is it a wasted effort? Will I end up just frustrating myself because those table don't actually calc out to our typical engineered designs?

 

[lexpatrie]

Preamble: This is going to meander a bit, sorry, it's about 2 am here. Also, I'm really mixing the years here, 2018 NDS, 2000 International Residential Code... if anybody wants to cross reference and see where any of these tables may have moved to if they've been renumbered in the last 20+ years, do please feel free. I'm not looking at the 2000 IRC to be stubborn or anything, it's just what I have open at the moment.

@StrEng007 - I'm not convinced light gauge steel framing is popular outside Florida, so there's maybe not much need for the lower wind loads, but it's in the prescriptive code, never seen it used personally. You could use the WFCM for petty high wind loads, but again, wood isn't popular in that part of Florida for a house, even. I keep hearing masonry is the thing. I thought ICC 600 was for storm shelters, I guess not, haven't much seen it or used it. The residential code, speaking loosely, is the minimum, it's probably in most places less than what an engineered design via the IBC would produce. I don't have big bowls of calculations to show that, mind you, but given the origin of the IRC it's "what's worked fairly well" versus anything that was explicitly calculated out in say, 1926. I'd expect quite a bit of it won't calculate out strictly, but does that make is a "waste?" that you learn and can speak to the client about their expectations for IRC versus an engineered building/wall system?

@phamENG - anchorage for loads along the wall, sure, if you skip over the potential for needing a holdown at the end of the wall via the braced wall provisions, sure, but for a soil lateral load, well, Mike Mike has more on that point.

I feel like you're missing a ribbon board on that detail at the joist, Mike Mike, but it's not critical to the discussion at hand.

On the 52' span, I suppose, if the rafter table gives you a 52' span, then... maybe? R802.5/1(1) (2000 IRC, sorry, that's the one I have open at the moment, but the other versions aren't insanely different...), is for dead load of 10 psf, roof live load of 20 psf (no snow), and no ceiling (L/180 deflection), Even so, that's a 12" joist spacing, DFL select structural, not something you'll usually get in quantity, given the 26' length you need.

I mean, yeah, for a ceiling joist versus a vaulted ceiling, I suppose, you need a ceiling joist, too, though, which is another DF ss, 26' long 12" on center. I hear rafter span (and ceiling joist span) and I think vertically supported at both ends. Maybe the residential code doesn't come out and say that, the "ridge board or ridge beam" in the drawing sure doesn't mean there's a vertical support (ridge board isn't a vertical support), but the ceiling joist, that's got to be looking for an interior wall at the 26' mark, it's not a clear span. I suppose, yeah. 52' width. I would imagine while there may be tables for the rafter to ceiling joist connection, there is probably not enough space for that many nails. Then we have the splice in the ceiling joist. That is probably going to produce a large force demand that could exceed the IRC fastening schedule, as well.

That kind of bears on the rest of the conversation, if you can put it in, but there's no space for the nailed connection? Table R802.5.1(9) has one entry where 39 16d nails are required for the connection.... that may be enough to make the connection work, but that sounds like a lot of nails in a relatively small space. You can clinch the nails to reduce the number required, but a 16d nail 0.162 x 3 1/2" per 2018 NDS, I suppose for 3" of wood (2 x 1/5") you get the 3/8" length beyond to get a legitimate clinch. (3 diameters past the side member? 12.1.6.4?) 802.3.2 requires a 3" overlap between the two ceiling joists, that's not a lot of space for nailing, at least not for a 26' long rafter each side, R602.3(1) Fastening schedule seems to call for 3-10d nails there. That value never changes regardless of the rafter to ceiling joist connection.....

On the bolt at the top of the masonry wall - wouldn't you need preservative treated wood there for the sill, so DF or SP, not HF?

Look at the 2000 International Residential Code, R603.2.1, point one, not parentheses 1. You're looking at a fastening table, the table I'm talking about (back when) was called Maximum Allowable Length (THEY should say height? Okay maybe not) of Wood Wall Studs,.... 100 mph, SDC A, B, C, or D[sub]1[/sub] (snow load 25 psf or less.... provided the F[sub]b[/sub] multiplied by the repetitive stress is 1,310 or more. That's where the MOE requirement was, like I said it might have gotten deleted.

Per 2018 National Design Specification, Supplement:
Sitka Spruce, Select Structural, 1,600,000 psi MOE
Beech Birch Hickory, Select Structural and Grade #1,
Coast Sitka Spruce, Select Structural,
DFL #2 or better,
Hem Fir Select Structural,
Hem Fir (north) #2 and better,
Red Maple #1 or better
Yellow Cedar select structural

Nothing one usually constructs with, unless you're going to count DF #1 which I'd rather doubt is a typical stud in say, California.

I kind of wonder if you can get a 2x4 stud to work for that 26' span of roof above it. Is that an 8' height, then?

There's just so many provisions, I'd be really surprised if EVERYTHING was per code in any given building. Let alone the footnotes, and the longer the code gets, the less likely it's "perfect".

I want to say that the sill bolting at 6' is for the load in the braced wall along the length of the wall (into the page, from wind/seismic). When it comes to the bracing effect of the top of wall versus the retained soil (perpendicular to the wall), the code .. might not know about it. There's maybe other provisions elsewhere for anchor bolt spacing due to the lateral (soil) load with C[sub]D[/sub] = 0.9 in there somewhere, or the code gives a reaction at the top of the wall that's supposed to be "designed for" (by whom?).

I'm not convinced there's "cheaper" available via an engineered design, and I'm not overly excited about putting an engineered system in the middle of two prescriptive designs and calling that good. We have load path requirements, after all (R301.1), (The IRC – Does It Really Matter?, Young, Structure Magazine, November 2019). So the connections and the elements to which your engineered design attach to and interact with are supposed to be proven valid to some degree. You wouldn't attach an 8,000 pound holdown into floor blocking in a commercial building, after all, it's supposed to be connected to something that's good for resisting the load.

I think you get into more ethical exposure when you know something is 18x over the engineered limit and don't point it out. Phil, Reid Middleton, and the Washington building officials are VERY active on foundation walls every IRC code cycle I've seen. I don't think you are fully on the hook for not providing them a fix "for free" but failing to point it out and either say --- hey, somebody should look at that, it's way overstressed, or a) I know a few guys that could design this or b) I could.

I think the critical element here is you know (in this example) that it looks highly doubtful to work correctly as a prescriptive design, and from past experience you know it won't calculate out per an engineered standard).

The provisions themselves, don't quite seem to capture a consistent connection strategy, so I'd really hesitate to "mix and match" and put an engineered design onto a prescriptive load path. I may be the fifth dentist, or I'm one of the 4 out of five dentists. Can't be sure.

 

[phamENG]

@phamENG - anchorage for loads along the wall, sure, if you skip over the potential for needing a holdown at the end of the wall via the braced wall provisions, sure, but for a soil lateral load, well, Mike Mike has more on that point.

Sorry - my post was a bit vague. The soil retaining was what I was intending to allude to, followed by a mention that some states take some parts of the IRC into their own hands for particular areas that are known to be deficient.

One thing to note - the roof tie down forces are limited to a roof span of 48', so that would be a limit on width. Four feet isn't much, but it helps a bit, I'm sure.

For the basement walls, I wonder how much of this depends on 2-way action. Sure, if you have a 60' straight wall, you'll have some issues. But that's pretty rare. As houses get bigger, the layout often gets more complex. So you may have a 48'x60' house, but there will be several jogs and changes in the basement wall that break it up. The result is a bunch of wall segments. An 8" slab can span up to 25 or 30ft. I bet there's a way to justify at least 80% (highly accurate statistic) of the concrete walls out their and their connections. Masonry, probably less so.

There are no requirements for an engineer to design the connection of the floor to the top of the basement wall, by the way. Engineering design is only required for retaining walls that are not laterally restrained at the top and retain more than 4' of unbalanced soil. I don't believe they define what 'laterally restrained' is, though.

I agree that the limitations should be a bit more clear, but at what point is it just the wood frame construction guide? You end up with so many 'if this then this' scenarios that it would be almost impossible to keep track without doing a full engineering analysis, which is contrary to the purpose of the IRC.

 

[Greenalleycat]

This is an interesting thread for me to read - in particular, Lexpatrie's responses which are quite firm with how they see this working
We have a generalised building code but not the IBC and there are some clear differences between the US system and ours
Shows the importance of some of you US folk remembering that you aren't the entire planet....... and for us non-US folks to always take a grain of salt with the advice on this platform

Our building system is structure roughly pyramidal:
The Building Act sets out the legal framework
Then comes the Building Code which sets out performance requirements for building work but gives little specifics on design
Then there are generalised Building Standards e.g. for light timber housing - this is similar to the IBC, it collates stud and joists sizes, all that stuff
Then there are the specific engineering standards that we work with

1) Our system requires that either an architect or structural engineer sign off the plans
A builder cannot submit plans even if they are done off the generalised Building Standard
If they did, or the homeowner decided to get handy and wield some tools themselves, they can and do get held to account by the local or national Government departments that oversee building
Thus there is always someone who can be held to account

2) There have been cases where engineers have been sued for failures that have occurred from specifications within the generalised Building Standard
To my knowledge these have only been serviceability issues (mainly bouncy floors) but our building code does not indemnify someone just because "you followed code"
The ultimate thing is performance for the person who paid for the product - if it doesn't perform then you can be sued regardless of following the generalised Building Standard
The intention of this structure is for the person who paid for the service to never be stuck in a situation where everyone points fingers at everyone else, no one gets the blame, and the client gets stuck with a shit result (it's a nice theory...)
I'm not aware of strength failures from this generalised Standard though

3) Our Councils have the responsibility under the Building Act to only consent something that "meets" the performance expectations of the Building Code
Going back to the pyramid above, this obligation sits right at the top, above the generalised Standards etc
If the floor joist (or whatever) didn't perform satisfactorily then Council would be dragged in 100% of the time
This happens a huge amount

 

[Mike Mike]

lex, were you sipping the sauce too that night? what a delightful ramble. I agree your comments would be great for submission to Stephanie J. Young, P.E. Chair of the NCSEA Code Advisory IRC Working Group. She said she needs our help in the article you linked. Here's a summary of our comments so far that in terms of bringing codes closer to alignment. Let me know if you're in agreement:
1. Clarify prerequisites for use of IRC tables: 10psf floor dead, 15psf roof dead, 70psf snow, 140mph wind, any exposure category, max building width it doesn't appear we reached a conclusion on but there seems to be a practical limit of around 40' or so.
2. Decrease IRC permissible basement wall height or decrease IRC anchor spacing or increase tabulated capacities in NDS or a combination of the three (yes, maybe part of the solution to our conversation is to reduce safety factors on engineered design).
3. Decrease permissible stud wall height or increase allowable stresses per NDS or both.
4. Do not specify 39 16d nails at heel joints. Decrease permissible rafter spans. Add requirement for ceiling joist splice connections to match heel joints.

Yes, there are a few refinements that could be made to my basement wall calc, but it doesn't change the conversation:
1. Change sill to treated DFL
2. Account for load duration factor
3. Add rim joist to detail
4. Account for wind load
5. Account for concrete on wood friction
6. Account for partial rotational fixity at concrete footing
Are you saying within the IRC "there's maybe other provisions elsewhere for anchor bolt spacing due to the lateral (soil) load"? I spent many hours digging thru the IRC at this point, pretty sure there's nothing in here, but let me know if you see something.

I see your reference to the high E value in 2000 IRC table 602.3.1 now, sorry I was in the wrong table.

Here are a few examples where design per IBC is cheaper than design per IRC. I could give more but let me know if this is enough to "convince" you:
1. Tension ties per 2024 IRC R507.9.2 not typically required.
2. Single top plate might be justified.
3. Top plate nailing requirements of table R602.3(1) lines 13 and 14 are extreme overkill and can be reduced or deleted.
4. Wall headers may be reduced or deleted if load path can be justified thru rim board.

Floor trusses are specifically permitted by 2024 IRC R502.12 and require engineered design regardless of span. You can't get more "in the middle of two prescriptive designs" than that. If you were EOR would you not allow your clients to use floor trusses?

Are you saying you would be okay with specifying 1/2" bolts at 6' to brace the top of your basement wall just so long as you noted on the detail that the connection was selected from the prescriptive requirements of IRC only. Would you include a note requiring the owner or contractor to hire an engineer to verify the connection?

I think we're all the 5th dentist at this point. The National Association of Home Builders might not appreciate us asking questions. I don't know anything about them, but I see they are active in the ICC

https://www.sbcacomponents.com/medi...de-victories-including-truss-bracing-proposal

and they "lobby" politicians (not bribe, lobby)

https://www.nahb.org/advocacy/advocacy-overview

is it possible the reason the IRC exists in its current state is because money? No, that's not possible, sounds like a conspiracy, don't follow the money, nothing to see here people.

phameng, great find on the 48' limit! funny how 5 highly paid professionals are combing thru a thousand pages looking for clues as to what the hell the committee might be trying to say.

Agreed, typically there are wall jogs. Perhaps the route forward is to prescriptively require jogs.

Correct "floor to the top of the basement wall" fastening is also woefully out of line with other standards. Prescriptive design requires just a few toe nails per foot of sill plate per R602.3(1).

I agree we should refrain from creating "so many 'if this then this' scenarios". Let's focus on revising tabulated values instead of creating alternative prescriptions. I would not expect my comments above to add significant length to the code.

Green cat, sounds like your committees should either revise or delete their floor joist prescriptions. Does not generally seem helpful to provide guidance that engineers can't follow without designing it themselves anyway. What country?

 

[Mike Mike]

One more inconsistency I noticed when I was doing my nightly IRC reading: Am I looking at something wrong, or does the IRC table below prescriptively specify 7/16" floor sheathing for supports at 24" oc? Why doesn't it match APA E30, also below? I mean the IRC is probably not wrong, 7/16" probably holds up fine to 40psf uniform live, but panel design is point load controlled, so it seems misleading to allow IRC users to select sheathing based on uniform loading, particularly since IRC tells its users what the required uniform live load is in table R301.5.

 

[Greenalleycat]

@Mike this is New Zealand
Yes I think they do need to revise the joists (and other things)...same as you guys have highlighted, there are many items in the generalised standard that are inaccessible to standard engineering design