Contractor Talk - Professional Construction and Remodeling Forum banner
1 - 15 of 15 Posts

·
Registered
Joined
·
5 Posts
Discussion Starter · #1 ·
I'll start off by agreeing with the hundreds of previous threads that say an engineer is indispensable on any project requiring significant structural work.

From time to time however, most of us rely on experience and the prescriptive rules in the IRC to make calls on the fly. How big does the header on this two foot wide window need to be? Probably the same as the last few hundred two foot window headers you've installed.

So this brings me to my question. What is the margin of safety that is built into the IRC and engineering specs in general?

The reason I ask is that in the coming year I'll be building my own cabin in the Washington Cascades. While an engineer will review the final design I've been using this project to learn more about engineering. There is a really cool, free program from Boise-Cascade called BC Calc that will recommend lumber sizing based on the architectural design. It tells you exactly what percentage of the allowable load a given design uses. In theory it can be used to design a very efficient (in terms of material use) building.

It does make me wonder if there is any margin of safety built into the 88 psf snow load required by the county, the specs from the lumber manufacturer and the IRC?

Ultimately an engineer will figure all of this out, but they don't end up paying for the lumber. ;)
 

·
"Pro"
Joined
·
628 Posts
In a lot of our classes we use a Factor of Safety of around 2.5 (A lot of that was for steel "I-Beams")....I'm trying to think...for things like joist sizing there's different allowable stresses for different grades/species of lumber...I'll have to check my notes on what they are...some where around 1000-1500psi maybe.
 

·
Registered
Joined
·
10,471 Posts
The recommended value for anything carrying human loads is 3.

A company that I did some work for insisted on 5. They've been in business for over 50 yrs. now and have never been sued over a quality issue.

The building code is a minimum for your area. My Dad always over-built and I do too.
 

·
"Pro"
Joined
·
628 Posts
The recommended value for anything carrying human loads is 3.

A company that I did some work for insisted on 5. They've been in business for over 50 yrs. now and have never been sued over a quality issue.

The building code is a minimum for your area. My Dad always over-built and I do too.
It's never a bad idea to go a little overboard...especially when designing decks. However I think with all of this "green" movement crap and builder's trying to save materials, a lot less overbuilding will be done in the future.
 

·
Registered
Joined
·
5 Posts
Discussion Starter · #6 ·
what part of the foothills are you in? I'm in Enumclaw
I live and work in the Seattle area but I'll be constructing my cabin up in the mountains near Glacier (Mt. Baker area). It's going to be fun building for a required snow load of 74 psf along with the various seismic regulations. Looks like there are going to be some big glulams in my future.

Where I used to live in Minnesota structural requirements were mostly the same throughout the state. It's interesting working in Washington where a few miles can dramatically affect what you have to build for. For instance, I need to engineer for 74 psf at my cabin. 25 miles up the road at Mt. Baker the roof snow load is 588 psf (that's not a typo). If my math is correct that would be like having a 5 foot deep swimming pool on the entire roof -- awesome.
 

·
Registered
Joined
·
26 Posts
This is a good question. Unfortunately the answer is not straightforward. The required safety factor may vary significantly, depending on the type of material, failure mode anticipated, load applied, and type of design specification referenced.

The ‘old’ way of design, ‘allowable stress design’, or ASD, in general applies a single factor of safety to material strength, such as steel yield or wood modulus of rupture. In this method, no adjustment is usually made with regard to type of load (however wood has a ‘time duration factor’ that decreases the safety margin on short-term loads).

For example, in many common cases steel yield in bending has a safety factor of about 1.67 in the AISC Steel Construction Spec; that is the nominal yield stress is reduced by 1/1.67 for design purposes, but the loads are not altered from the nominal values.

The ‘new’ way of design, ‘load and resistance factor design’, or LRFD (came out around early 90’s for some structural standards; later for others), applies a reduction factor to the material strength as well as separate factors to the different types of loads. For example, a common combination is a reduction of 0.9 to material resistance (a ‘phi’ factor) and factors of 1.2 to dead loads and 1.6 to occupancy live loads.

In reality, the expected factor of safety is obscured from the designer, because the mean value of the material resistance is usually different from the nominal value specified in the code. For example, back to the steel example, say a 50 ksi yield stress beam is specified, with a code safety factor of 1.67. In reality, the mean value of your 50 ksi beam may be closer to 53 ksi. Thus, an additional hidden ‘safety factor’ is present in the code-specified nominal values of material strength, added by the material manufacturer to insure material strength is at least the minimum nominal value. Wood has very high ‘hidden’ safety factors; I believe nominal values are usually based on the lowest 5% or so of specimens tested, rather than the 50% (mean).

It becomes more complicated when loads are considered; nominal loads often refer to a single maximum load anticipated over a long time period, say 50 or 75 years. This return period varies depending on the type of load applied. Thus practically, safety factor cannot be separated from a time period of consideration, and effectively decreases as the age of the structure increases.

Unless one has done design code calibration work, which is the task of coming up with the appropriate load and resistance factors (which is generally done by a small number of engineering research professors, with industry input), an engineer doesn’t really know what the actual safety factors are.
 

·
Registered
Joined
·
3,977 Posts
Be careful not to go redundancy over redundancy, this happens a lot when designing concrete. Engineers will design a structure with the safety factor included, then the concrete plant will add their own safety factor to the design in the concrete mix. So they are adding a safety factor on top of a safety factor ending up with way more than what is actually needed. I don't know if your engineering software already includes a safety factor or not.
 

·
Registered
Joined
·
5 Posts
Discussion Starter · #9 ·
...
In reality, the expected factor of safety is obscured from the designer, because the mean value of the material resistance is usually different from the nominal value specified in the code. For example, back to the steel example, say a 50 ksi yield stress beam is specified, with a code safety factor of 1.67. In reality, the mean value of your 50 ksi beam may be closer to 53 ksi. Thus, an additional hidden ‘safety factor’ is present in the code-specified nominal values of material strength, added by the material manufacturer to insure material strength is at least the minimum nominal value. Wood has very high ‘hidden’ safety factors; I believe nominal values are usually based on the lowest 5% or so of specimens tested, rather than the 50% (mean).
..
Thanks for your very insightful reply. If I can grossly oversimplify what you wrote, the takeaway is that safety factors are already built into the calculations used to produce things like span tables and IRC regulations. If I'm building a structure that conforms with the code and I'm using the code specified span tables then, in general, no additional safety factor is necessary (assuming no other special conditions are in play).
 

·
Registered
Joined
·
141 Posts
The IRC or any Code book in general, isn't a bible or book on how to build a house. Its a book to refer to, for the MINIMAL things you can do, yet still be safe.

I was told in school, that the code book can only build you a house to the absolute minimal standards. For many builders they won't rest there reputation on Minimal, at least not the ones I work for.

When reading span tables, in the code book to see what size lumber you can use to span your opening. Just be sure to use some common sense. If you have a 10 foot opening and a 2x8 is rated to span 10 feet in the code book, go with a 2x10. Why? Because your at the absolute limit to something that is already minimal, confusing but true.

But that's easy, there are plenty of confusing things you encounter on site, that aren't as easy as i just posted. Expecially, since a lot of things in the code book can be very confusing.

Code books (Ours at least) give more than one option and a lot of leadway for the builder, witch can definitly cause problems for anybody, who isn't very comfortable with the situation at hand.

-Bill
 

·
Registered
Joined
·
5 Posts
Discussion Starter · #11 ·
Be careful not to go redundancy over redundancy...
I agree with you Kenn. It's so easy to start beefing up everything well in excess of what it probably needs to be. My whole reason for starting this thread was that if everyone (code developers, materials manufacturers, engineers, etc.) are building in safety factors along the way it is possible that there is a lot of redundant safety built into the specs and if I upsize lumber when I get within 25% of the max I may just be adding to the problem. Thanks to everyone who's provided clarity so far.

For me it's not so much an issue of "green" building, but more of cost. I know, if the building falls down it's way more expensive than the cost of a little extra lumber along the way. That said, if the specs have significant safety factors built into them then I shouldn't necessarily be afraid when I'm pushing something to 95% of spec.

If anyone is looking for software to help size structural members I would recommend you check out BC Calc (I'd post a link, but the forum software won't let me).

I wrote to Boise-Cascade and one of their engineers provided the following info about how the software incorporates safety factors.

"The calculations that BC Calc performs are based upon current building codes. Safety factors are applied only to the products' design values. Those factors vary depending upon the property but range from to 2 to 3. The exception is deflection; as with all structural materials, stiffness values do not have safety factors."

That seems pretty consistent with what others have said so far in the thread.
 

·
Registered
Joined
·
26 Posts
Thanks for your very insightful reply. If I can grossly oversimplify what you wrote, the takeaway is that safety factors are already built into the calculations used to produce things like span tables and IRC regulations. If I'm building a structure that conforms with the code and I'm using the code specified span tables then, in general, no additional safety factor is necessary (assuming no other special conditions are in play).
I don’t use span tables for design so I couldn’t say for sure, however I would think that any table like this would be made to a certain code standard; for wood I would assume it to be NDS. There should be some documentation or footnote to the table that says what criteria/assumptions were used to do the calculations for the table. However, I could not imagine that a span table would be made without the required safety factors included; without this the tables would be worthless.

Also note that for many (or even most) practical designs, serviceability criteria (deflection, vibration, crack control, etc.; non-structural issues) tend to control over strength issues. Therefore practically, the actual safety factor is usually even higher than theoretically intended if serviceability criteria are satisfied.

I am not suggesting that structures are ‘overly safe’ however, and I would never suggest to design below minimum accepted standards.
 

·
Registered
Joined
·
6,102 Posts
I have an amazing book, given to me by my father titled "Machinery's Handbook" printed 1942.

It contains specs on everything from mathematical tables to strengths of lumber by species to rope transmission loads, etc. etc.

I have used this to calculate strength of lumber for loads.(as an information tool only) I defer to the engineer always. But it's an invaluable sourse of info for me.

FYI - Printed by 'The Industrial Press' New York, NY. last printing was 1942
 

·
Registered
Joined
·
10,471 Posts
My copy of that book (AKA the machinists bible) is the 'twentieth edition, third printing, 1976. The first edition was printed in 1914.
 

·
Registered
Joined
·
5 Posts
If anyone is looking for software to help size structural members I would recommend you check out BC Calc (I'd post a link, but the forum software won't let me).

I couldn't help but reply - thanks first of all, jon, for praising the BCCalc software... not to be partial, but it IS very user friendly and intuitive and you can do a heck of a lot more than just size a floor joist with it.

Secondly, I work for Boise Cascade and this is exactly what I do - design using EWP. If you click through my company web link, you can download the software from the website. It's FREE, so I am not selling anything or soliciting.. just providing a very useful tool. ;)

If anyone does get/use the software and runs has questions, you can contact me via email.

Not sure if there is a contact via the forum... just in case: [email protected]
 
1 - 15 of 15 Posts
Top