Originally written: Jun 2004
SIPs "behave" differently than sticks when subject to loads; they are more akin to shell structures than assemblies of columns, beams and girders. Point loads are dispersed throughout the entire surface of a particular plane and may continue to be resolved across panel joints. This fundamental difference in how SIPs respond to imposed loads must be clearly understood by those who design with SIPs and those who are responsible for deciding both what size and how panels shall be oriented and where and how the joints between them shall go and what type they should be.
For simple structures such as a small gabled barn or flat or shed roofed box, the process is seldom complicated. When the forms start to get large, high and complex with several girders or purlins incorporated, we may then be talking about something else entirely.
The conflict occurs when one attempts to maintain the advantages of SIPs in the face of conventional wisdom and common practice. The SIP advantages accrue when we use the fewest, largest pieces we can manage, thus reducing the number of joints and connections required in the field. Additionally, utilizing the fewest number of supplementary structural members - headers, solid splines, and posts - so that the amount of traditional stick material and field man-hours are reduced to a minimum is "best practice" as far as I am concerned. We maintain these goals for several reasons:
1. Reduce stick material - not just spline material, but all unnecessary posts, solid splines, headers, and miscellaneous blocking.
2. Reduce joints that are both the structural and thermal weak points in the system. Although properly designed and executed spline joints may transfer over 90% of the stresses across them; it is still not as strong as having no joint. Where there are cantilevers or headers, this may be critical.
3. Reduce field man-hours to both save time and maintain the quality of the work achieved in the factory environment. No doubt about it; man-hours in the field are never as productive as factory man-hours, and quality control is always a field issue.
4. Reduce total erection time as to push to project schedule ahead. Time = money, but this is especially a concern in the winter or in climates with extreme weather.
5. All of the above in order to both hold down costs, and also maximize the "greenness" of the project by minimizing the amount of additional lumber required.
The above summarizes the advantages of using the fewest panels possible to get the job done. As admirable as it may be to be passionate and creative in pursuit of that goal, ignoring the structural requirements of the design in order to maximize the above is insuring disaster. Having said all this, there are solutions that handle some of these standard loading conditions in ways other than what we normally see around within the panel industry. One of my favorites is the ridge beam landing at the peak of the gable end.
Most SIP manufacturers and distributors recommend posting this load down within the wall thickness. This usually means a vertical panel joint directly under the end of the ridge beam with two 2 x 4s or 2 x 6s depending upon the SIP thickness as a solid spline joint. But if we can live without this solid spline joint, wouldn't we be better off?
If we look at the load/span tables we find that for a 6½" SIP gable end wall 14 feet high at the bottom of the ridge, we are good for about 1700 pound loading per linear foot of wall at 100 mph wind load condition. In our sample case the building is 28 feet wide and the ridge is spanning 20 feet from some interior column to our gable end. The collected load area is 14 feet wide by half the span of 20 feet. Thus 14 x 10 = 140 square feet at 50 pounds per square foot loading (total live plus dead) equals a reaction (load at the point of support) of 140 x 50 or 7000 pounds. Certainly this is too much for just a beam pocket 3½" wide - or even 6" wide, if we figure the blocking on each side. This would only be good for half a linear foot of allowable load or 850 pounds.
What oh what can we do?
If we could just get more of the wall to play a part here we might get some place. If one foot of wall is good for 1700 pounds, then 4 feet is good for 1700 x 4 = 6800 pounds; 5 feet is good for 8500 pounds - that would do just fine! But how do we get that beam pocket point load to distribute itself over 5 linear feet of wall? There are two economical ways I know of:
1. STEEL BOLT TO TOP PLATES
The gable wall has 2 x 6 top plates which are used for capping off the wall and accepting the screws through the top of the roof panels, but they can also be used to distribute the ridge beam reaction down the slope of the gable panel. If we use a ¾" diameter steel bolt through the plate on one side, through the ridge beam, and out through the plate on the other side, we are connecting the beam to theses sloping plates and distributing the load at least 30 inches out each side. This would give us our 8500 pound capacity. One may use lag screws from each side if you can trust that they are applied properly and the threads are not stripped out and therefore the screws are carrying their rated load. I prefer the through bolt.
2. DOUBLE BOTTOM POCKET PLATE EXTENDED OUT OVER SLOPING PLATES
Instead of constructing the normal beam pocket, stop the panel height full across at the bottom of the ridge beam, route the foam down 3 inches for a double plate and insert these two 2 x 6s so that they extend out over the sloping plates. After the ridge is up, foam and fill in the small triangular holes on each side with small SIPs.
In both cases you can see that the trick is to get those sloping top plates to distribute the load laterally just enough so that the SIP wall can do its stuff. One must take care to see that the through bolt or bolts are properly sized so that it can handle the load without crushing either the ridge beam at the top of the hole where the bolt goes through, or the 2 x 6 plates where the bolt lands. Check the allowable compression perpendicular to the grain. Also, one must check the fasteners that transfer the load from the top plate through to the SIP skins. In our example, it would work out as follows:
Each side of the ridge is taking 3500 pounds. Number 8 screws are good for 150 pounds shear in a 7/16" OSB skin. So 3500 / 150 = 23.3 screws or round up to 24. For our 30 inch length this equals 12 screws on each side at 2½" on center.
If all these numbers are too much for you and you just skipped down to this part to see how it all came out, I hope the point will still be taken that we should let what lumber we must use work a little harder for us. Those sloping top plates can really distribute considerable loads down their lengths and accept ridge reactions that are double those used in our example. For running exact numbers it is important to factor in the pitch, the type of lumber used for these plates, screw shear resistance and spacing. On one job we had to double up these plates but it still worked out just fine. Don't forget this trick when there is a window centered underneath the ridge and you can't post the load straight down any way. And of course, sometimes the most economical panel layout doesn't give us a joint under the ridge. Rather than add another joint, remember the little trick in this article.
Sometimes the most economical layout is one where the panels run horizontally. Here we have to be careful about having a "hinge" joint on the middle of a wall and discounting the allowable loading due to the fact that OSB, like plywood, has about 15% greater axial loading capacity parallel to the face grain than in the perpendicular direction.
There are many other ways to layout SIPs other than the common soldier-next-to-soldier fashion. Like everything else, reflexive response without careful analysis closes the door on creative solutions that may pay off in different ways. In our little panelization exercises the payoff can be in $20's and $50's that very quickly add up to significant numbers. Construction of all types these days is very expensive and you know that we are always arguing that SIPs deliver high value. Here is another place we can close the cost gap - where it still may remain - between the stick guys and ourselves. We know that the correct answer to "Which is cheaper, SIPs or sticks?" is really, "It depends." The above is an example of one way "it depends." When SIP jobs are estimated the one that is aggressive about reducing sticks and joints may price out well under the reflexively designed "standard" solution. This is true both in the factory where the amount of pre-cutting and foam routing is cut down, and in the field where assembly time is significantly reduced.
Working with SIPs is like working with any other material, the deeper you dig, the more there is to know. The more you know, the deeper you dig. A little square piece of paper appears to reveal its entire realm of possibilities in an instant...until you plunge into the world of origami.
1. Reduce stick material - not just spline material, but all unnecessary posts, solid splines, headers, and miscellaneous blocking.
2. Reduce joints that are both the structural and thermal weak points in the system. Although properly designed and executed spline joints may transfer over 90% of the stresses across them; it is still not as strong as having no joint. Where there are cantilevers or headers, this may be critical.
3. Reduce field man-hours to both save time and maintain the quality of the work achieved in the factory environment. No doubt about it; man-hours in the field are never as productive as factory man-hours, and quality control is always a field issue.
4. Reduce total erection time as to push to project schedule ahead. Time = money, but this is especially a concern in the winter or in climates with extreme weather.
5. All of the above in order to both hold down costs, and also maximize the "greenness" of the project by minimizing the amount of additional lumber required.
The above summarizes the advantages of using the fewest panels possible to get the job done. As admirable as it may be to be passionate and creative in pursuit of that goal, ignoring the structural requirements of the design in order to maximize the above is insuring disaster. Having said all this, there are solutions that handle some of these standard loading conditions in ways other than what we normally see around within the panel industry. One of my favorites is the ridge beam landing at the peak of the gable end.
Most SIP manufacturers and distributors recommend posting this load down within the wall thickness. This usually means a vertical panel joint directly under the end of the ridge beam with two 2 x 4s or 2 x 6s depending upon the SIP thickness as a solid spline joint. But if we can live without this solid spline joint, wouldn't we be better off?
If we look at the load/span tables we find that for a 6½" SIP gable end wall 14 feet high at the bottom of the ridge, we are good for about 1700 pound loading per linear foot of wall at 100 mph wind load condition. In our sample case the building is 28 feet wide and the ridge is spanning 20 feet from some interior column to our gable end. The collected load area is 14 feet wide by half the span of 20 feet. Thus 14 x 10 = 140 square feet at 50 pounds per square foot loading (total live plus dead) equals a reaction (load at the point of support) of 140 x 50 or 7000 pounds. Certainly this is too much for just a beam pocket 3½" wide - or even 6" wide, if we figure the blocking on each side. This would only be good for half a linear foot of allowable load or 850 pounds.
What oh what can we do?
If we could just get more of the wall to play a part here we might get some place. If one foot of wall is good for 1700 pounds, then 4 feet is good for 1700 x 4 = 6800 pounds; 5 feet is good for 8500 pounds - that would do just fine! But how do we get that beam pocket point load to distribute itself over 5 linear feet of wall? There are two economical ways I know of:
1. STEEL BOLT TO TOP PLATES
The gable wall has 2 x 6 top plates which are used for capping off the wall and accepting the screws through the top of the roof panels, but they can also be used to distribute the ridge beam reaction down the slope of the gable panel. If we use a ¾" diameter steel bolt through the plate on one side, through the ridge beam, and out through the plate on the other side, we are connecting the beam to theses sloping plates and distributing the load at least 30 inches out each side. This would give us our 8500 pound capacity. One may use lag screws from each side if you can trust that they are applied properly and the threads are not stripped out and therefore the screws are carrying their rated load. I prefer the through bolt.
2. DOUBLE BOTTOM POCKET PLATE EXTENDED OUT OVER SLOPING PLATES
Instead of constructing the normal beam pocket, stop the panel height full across at the bottom of the ridge beam, route the foam down 3 inches for a double plate and insert these two 2 x 6s so that they extend out over the sloping plates. After the ridge is up, foam and fill in the small triangular holes on each side with small SIPs.
In both cases you can see that the trick is to get those sloping top plates to distribute the load laterally just enough so that the SIP wall can do its stuff. One must take care to see that the through bolt or bolts are properly sized so that it can handle the load without crushing either the ridge beam at the top of the hole where the bolt goes through, or the 2 x 6 plates where the bolt lands. Check the allowable compression perpendicular to the grain. Also, one must check the fasteners that transfer the load from the top plate through to the SIP skins. In our example, it would work out as follows:
Each side of the ridge is taking 3500 pounds. Number 8 screws are good for 150 pounds shear in a 7/16" OSB skin. So 3500 / 150 = 23.3 screws or round up to 24. For our 30 inch length this equals 12 screws on each side at 2½" on center.
If all these numbers are too much for you and you just skipped down to this part to see how it all came out, I hope the point will still be taken that we should let what lumber we must use work a little harder for us. Those sloping top plates can really distribute considerable loads down their lengths and accept ridge reactions that are double those used in our example. For running exact numbers it is important to factor in the pitch, the type of lumber used for these plates, screw shear resistance and spacing. On one job we had to double up these plates but it still worked out just fine. Don't forget this trick when there is a window centered underneath the ridge and you can't post the load straight down any way. And of course, sometimes the most economical panel layout doesn't give us a joint under the ridge. Rather than add another joint, remember the little trick in this article.
Sometimes the most economical layout is one where the panels run horizontally. Here we have to be careful about having a "hinge" joint on the middle of a wall and discounting the allowable loading due to the fact that OSB, like plywood, has about 15% greater axial loading capacity parallel to the face grain than in the perpendicular direction.
There are many other ways to layout SIPs other than the common soldier-next-to-soldier fashion. Like everything else, reflexive response without careful analysis closes the door on creative solutions that may pay off in different ways. In our little panelization exercises the payoff can be in $20's and $50's that very quickly add up to significant numbers. Construction of all types these days is very expensive and you know that we are always arguing that SIPs deliver high value. Here is another place we can close the cost gap - where it still may remain - between the stick guys and ourselves. We know that the correct answer to "Which is cheaper, SIPs or sticks?" is really, "It depends." The above is an example of one way "it depends." When SIP jobs are estimated the one that is aggressive about reducing sticks and joints may price out well under the reflexively designed "standard" solution. This is true both in the factory where the amount of pre-cutting and foam routing is cut down, and in the field where assembly time is significantly reduced.
Working with SIPs is like working with any other material, the deeper you dig, the more there is to know. The more you know, the deeper you dig. A little square piece of paper appears to reveal its entire realm of possibilities in an instant...until you plunge into the world of origami.
Originally written: Jun 2004
RSS Feed