CountryPlans Design/Build Forum
General => General Forum => The IRC: International Residential Code and Good Practices => Topic started by: Squirl on April 05, 2011, 08:43:43 AM
I wrote this this morning as a seperate post. I will place it here, I hope others find it helpful.
So I started a new thread based on what I saw as one of the least published parts of framing. This in turn, makes it one of the more common questions on the forum. One of the popular foundations used on the site is post and pier. The problem is that it is not widely published in most framing and foundation books. So there are many clear examples of framing charts of joist, stud, and rafter sizes, but there arenít many for outside bearing girders. After many discussions on this forum I have found this information in the building code itself. I will try and interject as many pictures as possible in this. Keep in mind, I am not an architect or engineer. If you notice any mistakes, please correct me and I will edit the post. The code section I am referring to can be found here:
This is the area I am referring to.
Since a 1.5 story 20í wide universal seems to be a popular option I will try and run a visual guide through the code with this as an example. Depending on design people can scale it back from there. Because the 1.5 story house normally has two clear span floors with no center bearing beam, I will use the 2 story chart as the reference because it should have close to the same weight load on the outside walls. This chart is for #2 lumber which seems to be the default for big blue and big orange.
I will focus on different sections.
This is the first category. The snow load can usually be found at your local building department. Also most state codes have a map with snow loads by area. The minimum is 30 lbs, but with upper elevations and northern properties it goes up to 70. In my area I have a 50 lb requirement. Sizing up never hurts and allows for additions later such as roof mounted solar panels and heavier floors. For this example I will use the 30 lb requirement.
The next section of the chart is the width. This is referring to the total width of the joists. We are using a 20 ft wide for the building. If you expanded the building to 24 ft, you would use the 28 ft column.
This is an example of the clear span floor. There is no center beam for the width.
The last section is to simply figure out the number of posts you want to span the length of your building. So if you want to go with 7 posts over a 30 foot long building you would have 6 equal sections of 5 feet. Due to the size of the post it would be just under a 5 foot unsupported area. So with a 30 lb snow load and a 20 foot wide building you would have your choice of two 2x12ís, three 2x10ís, or four 2x8ís to meet the code minimum.
If I missed anything, please let me know. Maybe someone with more experience and more articulate than I can explain the jackstud requirement. Also a key note make sure all parts were two piece of dimensional lumber meet are over top of a post, not in the middle of the span. It does not matter if these are nailed to the board next to it or not.
Very nicely presented, with enough back-up info such as the girder table R502.5(1) so that we are all discussing and looking at the same thing and can point to various items with a min. amount of fuss. In the future, I would suggest including the footnotes for that table in what you copy and post, since understanding them is essential to proper use of the table; maybe also the specific code sections or paragraphs which apply, again so we are all looking at the same thing with min. amount of hunting for it, since you already have it at hand.
Allow some time to look it over and comment, if we find things we want to explain or expand upon. This is exactly the kind of discussions I hoped to generate, and this is certainly a very good first topic. There are probably four or five additional sub-topics shown in your photos if you donít mind them being used as the examples for discussion.
The number of jack studs reference is to when the charts are used in sizing a header as in over a door, window or open arch or walk through in the wall. When a header reaches a certain size the header will require 2 or more jack studs per side to support the header properly.
Those images have a look of familiarity about them. ;) BTW, that particular set of piers and girders have had an undocumented change/addition made to them. The inside of each pier - girder connection has a 3/4" PT plywood gusset plate, about 16" square with lower corners lopped off, securely nailed over the pier-girder interface.
I tried to find clear accessible images of some proper examples. They should look familiar. d*
I realized I should have put the notes after I created the post. One of the ones that I left off or made a short cut was that in between sizes can be interpolated, such as the 24' example, and I just stated to go to the next size up. "c. Building width is measured perpendicular to the ridge. For widths between those shown, spans are permitted to be interpolated." At least they are accessible from the link.
This brought to mind something that is a little off topic so feel free to get us back on track.
The girder tables are severely lacking in the codebook. Notice the IRC does allow design "in accordance with accepted engineering practice" "or in accordance with AF&PA/NDS". The tables Squirl posted are "prescriptive". The methods and tables in the code prescribe one method you can use to easily satisfy the inspector. If you need something different that portion is supposed to be engineered.
This is a link to an AF&PA document that contains safe load tables for many stock timber dimensions. It does require more understanding to use than the codebook tables.
I'm also pointing out a grey area in the code, subject to interpretation. You can use these tables to size a girder . These tables were made using the methods referenced by the code, the NDS (National Design Specification for wood construction), these are the methods of "accepted engineering practice". A strict interpretation is that using these tables is the realm of an engineer. If you can demonstrate an understanding of how to use these tables, and if the inspector is comfortable in his knowledge, he may allow you do do this level of "engineering". These tables are set up the way the codebook tables were about 20 years ago.
Finally, crossing all the way over the line into engineered. The calculators I have posted perform the calculations from the NDS. At this level you are engineering, if done correctly it is according to accepted practice. At just about any level of enforcement, if you are performing these calculations, this area is required to be stamped by an architect or engineer.
In the case of a girder, I often use LVL's. I can have the supplier check the sizes needed and if necessary he can provide documentation sealed by a design professional. Engineered products often come with sealed documents as part of their service.
Squirl, you have presented an opportune time to expand on the girder theme you introduced. We were going to get here sooner or later, so now is as good as later, maybe better. Thank you.
Below I have inserted what a professional engineer (PE) sees and thinks about when presented with the issue of piers and beams. Please read it through with an open mind, there is a lot of good information there. I believe it develops the Ďbig pictureí in about as easy to understand manner as is possible.
The information presented is for educational purposes. It is not meant to point fingers at any individual or group of people. Anyone may choose to consider the information or choose to disregard it. My PE friend does not intend to design anyoneís foundation or structure for them here on the forum. He simply wants to assist owner Ė builders in their own design and evaluation. As a professional engineer he certainly wants to avoid any issue of liability having done someoneís design for them here and then having something go wrong because of information that was not fully presented, or someoneís shortcut, etc.
"One of the popular foundations used on the site is post and pier. The problem is that it is not widely published in most framing and foundation books."
And, at least one reason for this is that these kinds of pier foundations are not a real good solution for the foundation system for a habitable building. They work fine for out buildings and farm buildings where life safety is somewhat less an issue. The primary problem is that they really only consider gravity loading (vertical loading) and then donít do a real good job of that in some cases. No doubt, they are an economical solution for light gravity loads, but you must weigh their shortcomings in this analysis. All the piers will not be loaded the same, and they are pretty susceptible to differential settlement, unless they are on a proper strip footing or grade beam. They take essentially no account of lateral loading or uplift from either wind or earthquake and are very difficult to detail to make serve these purposes. You must get all the potential loads on the structure into the foundation and its bearing material (the soil), at an allowable bearing stress. And, these lateral loads can be applied in any direction, not just the most advantageous direction. Loads do not go away because you ignore them; they just come back around to haunt you.
"After many discussions on this forum I have found this information in the building code itself."
This type of find and discussion was exactly why we thought understanding and studying the IRC a bit more might be beneficial to many people here. Your example seems particularly germane for this forum and your use of Table R502.5(1) seems pretty much on the money too, with a few exceptions, caveats, or things that they just donít tell you because they donít expect you to stretch the use of this table as you are with your example. And, you must know when you are going beyond the limits of the code or a table like this, or when you are exceeding your own abilities and knowledge and should hire some experienced design or construction help.
A.) This girder table basically assumes simple span girders or headers in an exterior bearing wall, uniformly loaded (say w, lbs./ft.), but without concentrated loads, so watch out for point loads from above these beams; thus there must be jack studs (also called cripples, trimmers, shoulder studs, etc. depending on your neck of the woods) to support each end of the built-up girder. And, you can see in the table that as the girder (header) member gets deeper (that is x8, x10, & x12) more jack studs are required for its support, because the deeper member will carry more load on a given span. Generally the plywood is added to a built-up header to fill out its thickness to match the wall or pier thickness, it doesnít add much to the header strength.
B.) In your example, you have 7 piers and two short end cantilevers and 6 identical interior span lengths (call them L, about 5'), and Iíd call this a 6 span continuous girder. The big caveat or Ďdidnít tell yaí, is that continuous girders donít work quite the same way simple spans do. The two end piers are lightly loaded because they only carry about (0.5Lw, lbs.) of span length, the first interior piers can be loaded at about (1.2Lw), and the three middle piers can be loaded at (0.9Lw) or so. So, there is some potential for differential settlement if everything else is perfect, which is rarely the case. And, if one of the piers is particularly susceptible to settlement because of bad soil at that location, or some such, then you are starting to cause the built-up girder to have to span about 10', not just 5'. And, you can see from the table that that just wonít work. Not to worry though, with enough settlement of the weaker pier, things will likely stabilize, but maybe over stress the girder, or crack the window above, etc. This is probably not a catastrophic failure mechanism, but it will make for some interesting dinners when Suzieís spilled milk on the table ends up in your lap. I tell you this not to scare the hell out of you, but so you at least give these issues some thought and consideration as you decide how to build your cabin support system.
C.) The other big caveat or Ďdidnít tell yaí, is that deflections on continuous girders are generally less than on simple beams for the same load, span length and member size, and thatís a good thing. However, loading one span and not the adjacent span can actually start to cause an upward deflection in the unloaded span. Try this experiment: using short blocks of 2x4 on edge, support a yard stick laying flat at 0, 1, 2, & 3' points; load the middle span &/or one of the end spans, and watch the movement of the other spans. To a greater or less extend this will happen with a continuous girder, and the loadings on the piers will change too. Now, lay one of the interior blocks on the flat (a settled pier) and load various spans. Try to get some sense of the magnitude of force required to cause various deflections, up or down. Simple beams just load their own reaction points and donít change deflections or reactions in adjacent spans.
D.) This could start to happen at an interior span with a 6' window opening over it, where the adjacent spans are loaded and jambs on the window opening also fall in these adjacent spans; thus only 1st floor framing is loading the girder span in question, but 2nd fl. and roof framing loads are going to the adjacent spans through the opening header and jambs. Another place this type condition could happen is at an end span where you have a 5 or 6' entry opening over that span, but the adjacent interior span is loaded. Then you have a similar load path as mentioned above.
E.) Finally, the built-up girder table only considers a uniform load condition. We do consider studs and floor joists at 16" o/c as a uniform load for this condition, but a large opening with jamb loads at mid span on two adjacent girder spans could cause some girder over stress or deflection issues. Again, I donít bring these issues up to scare you, but to make you think a bit, and consider them in your design. These cabins, these types of buildings, are fairly loose and forgiving, so again, we are not likely talking about catastrophic failure mechanism here, and they are generally fairly lightly loaded except for snow and lateral loads, but this is some insight into how an engineer looks at these problems, and what you get when you retain one for a little design planning help. Iím not saying all or any of these conditions will cause you problems, the conditions which cause them are design specific, but I do want to make you aware of them and am suggesting that they deserve consideration in your planning process. Ultimately, you have to make these decisions, spend the money now, to avoid problems later; or how much is enough. Except for experience and professional judgment, it is a bit of a crap shoot, and it is even for experienced designers when they are not given enough info on which to base their design decisions.
3.) The part of the table you showed assumes that the roof, ceiling and the floor framing does clear span to the exterior walls, and that there are no interior bearing walls or supports. Floor framing systems can become fairly flexible (unacceptable deflections) at these 18-20' clear spans. Any splices in the built-up girder certainly should be made over the piers, and not all at one pier, and certainly not at mid-span. Contrary to your assumption however, the side-by-side pieces of 2x material making up the built-up girder or header most certainly should be face nailed together. See Table R602.3(1), and I would use the Ďcontinued headerí prescription with 16d, at 16" o/c each edge, staggered. You do want all the members to act in unison, and you do not want them to be able to roll with respect to each other as they make up this built-up girder.
"One of the popular foundations used on the site is post and pier. The problem is that it is not widely published in most framing and foundation books."
Post and pier often seems to be the most practical solution for the DIY builder, especially in remote areas where access for heavy equipment is poor or unavailable. If that system is going to be used it would seem to me that it is advisable to avoid the temptation to use code minimums or the minimums in the engineering tables cited above. Spending the $$$ for some soil analysis is also a good means to at least even the odds a little more for the builder-owner. "Overbuilding" at the foundation may be the saving grace for post & pier systems. It also seems to me that local knowledge and empirical experience does provide some level of confidence for the cabin builder. In the area of southcentral Alaska where I built my cabin it is rare to find continuous perimeter foundations. Literally thousands are built with P & P foundations and, while some do develop differential settlement problems, the great majority appear to give good service. I was advised that using the code and span table minimums was likely to be a false bargain so I built with extra and larger piers and footers under them as well as upgrading the girders size. I did use the minimums for the height of the piers above grade, hoping to help minimize lateral or racking forces. If my site had not been well drained gravely soils, I probably would not have used P & P since their use in soft boggy areas here appears to have a high rate of settlement problems.
Essentially, the tables and codes are uncertain guides for P & P builders especially in poorer soils. I did consult with a structural engineer friend who was kind enough to provide a little informal guidance on my foundation and roof framing hence my "overbuilt" foundation and use of a rather hefty glulam as a ridge beam in lieu of my intent to use a simple ridge board.
The information/cautions provided by the previous poster are certainly valid, highlighting the usual problems caused by the "we don't know what we don't know" trap.
Sorry I was not clear on the last point about face nailing. Some people are under the incorrect assumption that it is alright to put a joint mid span as long as it is face nailed to the board next to it. I have a picture to describe the two situations. Please excuse the crude drawing.
As you can see my point was that the built up girder on the left is not as strong as the one on the right, even if it is face nailed. Not that either shouldn't be face nailed. I was looking for a photo of this yesterday and didn't want to single out any individual construction project, so I inarticulately tried to describe it.
This is a good and helpful thread.
There is no doubt that in problem soils (poor draining, clay, differential bearing) or where strong lateral loads (high wind or earthquake) need to be considered that the standard reinforced concrete perimeter foundation is superior over the post and pier design. That is why it is the most common prescriptive choice for residential buildings.
I try to point this out in my plans that have the post and pier foundation (and most house with the P&P foundation also have a perimeter concrete foundation and slab plan as well).
That said, there are reasons and places where the post and pier foundation will still be built even if not in ideal soil and loading conditions. This thread can provide some guidance and design help for those builders and those situations.
To carry on....
The problem with being specific about where to put the splices in a built-up girder is that the splice itself and its exact location must be designed structurally and is highly dependent upon the span lengths and exact loads on the spans. And, you must know the bending moment and shear at the location you wish to make the splice to do this design; alternatively you move the splice location to a point where you can design around the moment and shear which exist there. So it is really dependent upon each individual girder design. One engineer might say to another, put the splice at a point of contraflexure, that is, where the moment is zero and only shear must be resisted. Thatís the problem with explaining these kinds of structural designs and problems to non-engineers, many times some of the folklore has a shred of truth, however ill purposed and advised. The most expeditious place to make a properly designed splice might be out in the span someplace, but you should probably not do this.
The easiest way to answer about splice location, to be fairly safe from the engineering standpoint, but also to be sure to offer advice which would not be dangerous to a non-engineer is to agree that splices should always fall over the piers. No splices would be the best solution, and can be done on a 30' girder, with LVLs or GlueLams, assuming you can transport them, but canít be done with normal dimensional lumber. The engineering logic behind my suggestion of splicing over the piers goes something like this: The discussion started with a built-up girder table based on allowable bending and shear stresses and some deflection limit, for a uniformly loaded simple beam, girder or header. Now to refine my comments, I would prefer to see no splices over the first two piers from either end of this 6 span cont. girder, and no splices out in the span, but as I mentioned above load conditions could change this thinking.
Normally, the largest bending moments and shears are at the first interior pier on this 6 span cont. girder so we certainly donít want any splices there. The cantilever might be fairly highly loaded by the gable end walls, to the point that there should probably be at least a double joist immediately under each gable end wall. The bending moments and shears at the other interior piers (middle 3 piers) are generally no worse than those found in the simple beam which was the basis of the original IRC table. However, when you make a splice at one of these piers you will have a weaker girder than you originally hoped for.
Assuming a built-up girder consisting of 3-2x10's, you effectively reduce the strength of that member by 1/3 when you butt splice one of the three members at a splice, and you reduce its strength with respect to bending by 2/3's when you butt splice two of the members at the same location. And, the design question becomes can you tolerate this reduction given the bending moment at that location. When you do this splice over the pier you do at least have bearing for the spiced members so you should be able to resolve the shear and bearing design issues. But, the reduced bending capacity at the splice will likely mean that the cont. members at that point are over stressed as compared to the normal allowable stresses, it will mean more member rotation at the support and will lead to greater deflection near mid span in the two adjacent spans.
Again, this is all dependant upon a better understanding of the beam loads on those spans. Under normal conditions, this approach will lead to a reasonably safe beam design, if you have compiled your loads correctly, this will not normally lead to a catastrophic beam failure. Iím not trying to make a big deal out of this subject, but you are asking for a simple answer to a complicated problem, which requires considerable experience and judgment during the design process. And, Iím not suggesting that each girder is a three week design job, but it took a lot longer than that to gain the education, experience and judgment to do that design efficiently and safely for all concerned. Consequently, the IRC type (prescriptive) tables for beam spans and the like, will of necessity, be fairly conservative when presented for use by the general public, and you exceed their intended limits at your own risk.
I think Rwanders is right on the money in his comments, not penny wise and pound foolish. "I was advised that using the code and span table minimums was likely to be a false bargain so I built with extra and larger piers and footers under them as well as upgrading the girders size. I did use the minimums for the height of the piers above grade, hoping to help minimize lateral or racking forces. If my site had not been well drained gravely soils, I probably would not have used P & P since their use in soft boggy areas here appears to have a high rate of settlement problems." He provided a little more Factor of Safety in his found. design, at a fairly small percentage increase in materials costs, and almost no change in labor. His ridge beam if supported properly, and rafters if connected properly to the ridge beam; as opposed to a ridge board, rafters and raised rafter ties; solved another issue which seems to be a regular topic of discussion on this forum, namely rafter thrust at the top wall plate.
He didnít actually Ďminimize lateral or racking forcesí on the structure or the found., those didnít change, but he reduced the stilt height, or lever arm, btwn. the foundation and the floor diaphragm, that is the distance over which those forces act. The projected area of the bldg. to the wind doesnít change, and must be considered in any and all directions. The width dimensions of the bldg. influence its overturning potential; and the height from some point below grade, the soil reaction elev. on the piers to the elev. of the fl. diaphragm determines in bending moments on the piers, acting like cantiís. out of the ground. Actually, beach houses on 8 or 9' high stilts are fairly common, but these founds. are long, heavy structural piles, properly located and braced and driven to an appropriate depth of penetration. This should be a highly engineered foundation and should not give DIYíers. license to put 6' long, 6x6's, 4' down, in a hole, in just any soil and expect good foundation performance with respect to all loads.
In many locations up in Alaska, one of their problems which leads them to use posts and ftgs. for a foundation is permafrost which dictates that the structure be elevated w.r.t. grade so as to keep its heat from melting the found. soils. Many of those soils turn into something btwn. liquid mud and just short of quicksand when they melt. For example, the oil pipe line from the North Slope is on stilt support structures, in many locations, which are manufactured to generate a freezing action to keep these soils frozen, unsettling, from the heat of the oil.
" the oil pipe line from the North Slope is on stilt support structures, in many locations, which are manufactured to generate a freezing action to keep these soils frozen, unsettling, from the heat of the oil."
They are called VSMs (vertical support members) and represent some nice, economical and virtually maintenance free design work. They are 36" diameter steel pipes driven to depths indicated by soils at their locations and a sealed loop is mounted in each one and filled with an ammonia based refrigerant liquid. on top is a finned radiator device. The liquid circulates without pumps of any kind up and through the radiator and keeps the soil around and below the VSM frozen year round. It essentially extracts heat energy from the earth and transfers that heat to the atmosphere. The same principles are used in geothermal systems with heat pumps and also in your refrigerators and air conditioners. All in all a very neat solution using some basic principles from physics and chemistry
nicely done!! thanks for taking the time! [cool]
bookmarked for future reference!
I wanted to add a couple things; some are repeats of what has already been stated. Just doing it for emphasis.
The IRC does state that all splices will occur over supports. Even if building with no inspections it makes sense to me to follow along. It makes sense to avoid having joints in different layers occur over the same support.
Any material used as spacing material does not add any strength as far as doing engineering calculations or reading any tables. As such it should not matter is the spacing material is used between two 2x's or between the 2x's and the hardware like Simpson brackets.
The codes make no mention of using adhesives or glues between the layers. So why bother? Construction lumber is not smooth like fine hardwoods used for furniture. Construction lumber frequently has a certain amount of cup to it. For a good solid glue joint the wood must have intimate contact. Gap filling glue is more a gap filler than a strong glue. So why bother? That's my opinion.
A lot of nails are not required. The real strength of a built beam comes from the thickness and height of the 2x's. The nails mostly hold the layers together in a neat package. The Southern Pine Council and the IRC nailing schedule calls for nails every 32 inches of length, with the nails along the top and bottom staggered. Two nail at splices. That's their minimum, so twice that would be excellent. But any more is likely only going to make the nail vendors happy.
Construction adhesive is great for use under subflooring; pretty much eliminates squeak risk. Also good for wall paneling. As an adhesive to hold structural members together it is less satisfactory. It does tend to creep. Personally I've never been happy with the way it does or does not compress between layers. Other glues like polyurethane will squeeze down to a thin film, but are lousy when the glued surfaces do not make intimate contact.
Taking the same example from above and using a free beam software
the diferential loading of each pier is shown.
I had this uneven loading on my BBQ barn plans, post to post - basicly I was going to install as per the worst post under each one as the diferences are quite considerable especially on marginal soils like I have
First post is roughly loaded by 2500lbs
the second ones in are 6064lbs
3rd in 5492lbs
living floor- trib 10 ft
sleeping loft- trib 10ft
roof with 2ft eave- trib 12 ft
snow load 30- evenly distributed
All loads even
looking at big foot for example and 2000#sqft soil
BF20 would do for the first posts
BF24 would just work for the second posts in - but better using the BF28's
If you used BF-28's all arround you'd have support of 119,840lbs (8560#/post)
6x6 posts alone on that same soil would be only about 500lbs support for each one- certainly would not work in my soil area !
even in a 5000#sqft area I'd be sinking
Feel free to check and correct me as this is a reference page and I'm an instrumentation engineer - not a structural one !
Report from software
First post is roughly loaded by 2500lbs
the second ones in are 6064lbs
3rd in 5492lbs
The load on the first pier in a row is always the least loaded pier. The second pier in from the end is the most highly loaded pier. This disparity is one of the potential issues with a pier and beam foundation and I think PEte touched on that somewhere above. Or maybe the loading was something we talked about between ourselves. ??? Anyhow that's usual, unless the building is very unusual with its load distribution.
The recent conversation started by vrf has been moved to it's own topic
.... to General Forum (http://countryplans.com/smf/index.php?board=5.0).
Thanks for starting this thread. In truth, the initial table you posted seems written to be incomprehensible. It's impossible to tell without extensive research that you're meant to understand the idea is to take a building width supported only along the longest sides, with piers every so often, as indicated.
I've learned in the last decade that codes are too important to be left to mortal man or woman. There is no thing so simple that a building code writing cannot make it impossible to understand. Cheers, all, and thanks.
I beg you, though, please, please proofread your posts. It's just not possible to make sense of the following:
"Also a key note make sure all parts were two piece of dimensional lumber meet are over top of a post, not in the middle of the span."
Thanks very much.
Code books are like law books - difficult for the layman to understand and constantly changing.
In the end you need to meet the needs of the local inspector. So ask HIM (it is usually a him) -- Or see an engineer. An engineer will offer you more freedom as alternative methods can be modeled in the calculations.
"Also a key note, make sure all parts where two pieces of dimensional lumber meet is over the top of a post, not in the middle of the span."All splices and joints are required to be over a load bearing element (post, pier, crossing wall, etc.).
However, a pier type foundation already requires an engineer, they will also detail the girder.
The codebook, although highly informational, is a lousy "how to" book. It is not only like a law book, it is the book of laws for construction.