Wood is still the king here in North America for residential construction. Who knows maybe in a few years these 3D printed homes (of concrete) might take off and leave me in the dust but for now its all about studs, rafters, trusses and joists, and a few beams and columns for good measure.
On that train of thought I ran some numbers on what size of glulam beam it might take to span that garage at full span.
The post might be a little overkill, probably and 8x8 would work fine but I tend to error on the conservative side.
Even with a 27" beam you are still getting a lot of sag, about an inch. All of the furniture will slowly walk its way to the center of the room, Iāve seen this before. As much as I hate having a post in the middle of my garage it seems like the best solution.
Possibly another option is to run two glulam beams, divide and conquer, I might have to check those numbers.
P.S. Even with two glulam beams (tributary length is 10ā-8") in the garage (24" deep) you still get almost an inch of total deflection. That 32 foot span is brutal. Now Iām wondering what steel would look like.
P.S.
I also realized that once I upgraded the post to a glulam (8.75 x 9) I donāt have an appropriate post base in the plugin for this larger size. I just looked through my Simpson catalog and what Iāve included in the plugin built-in library. Iāve only included the CB44 and CB66 post bases as well as the full line of the CBSQ line. This is severely limiting. I need to add in 27 more post bases of the CB lineup. Working on itā¦
Out of curiosity, the glulam beam would obviously need posts at each end ā is this function part of the beam logic or would that be added as a function of the wall extension?
Sorry, I havenāt added the in-wall columns or posts at the beam ends yet. That is a function of the in-wall column tool. What would that deflection look like on that W flange beam? Is there any free tools or even paid tools that you would recommend for analysis of steel beams and posts?
I just did the quickie calc as a mental challenge using basic assumptions ā deflection would take more time and analysis but I bet it would be less than the glulam beam. Truth is I donāt do this everyday (anymore) - sadly Iām mostly in management nowadays trying to āherd the catsā.
I donāt know of any free tools but our engineers use STAAD Pro Structural Worksuite ā unfortunately not free and not inexpensive.
Iāve used STAAD in the past about 10 years ago and more recently Iāve used RISA 2D, which is also a really nice piece of engineering software (especially for matrix analysis and indeterminate structures). The last time I did anything significant with structural steel was in 2013 working for a company in Seattle called Pacifica Engineering, they have since been acquired by MTorres. I could probably hand calc the deflection but nowadays it is so much easier to find the right piece of software and let it run the numbers, we are too spoiled in the modern age of computers, programs and apps, most of us have become lazy (myself included).
When I first started doing a lot residential engineering I used to hand calc my beams, posts and headers. Breyerās book was the bible:
I quickly realized that hand calculating each time was tedious and so I wrote my own beam calculator (2014):
I then realized that free products like Forte (Weyerhaeuser) and BC Calc were also out there and were pretty decent so I typically use those products now for running the numbers on wood beams:
Forte now has a SDK or API and Iāve been thinking about how to integrate the extensions with their engineering software, that would be a huge timesaver for myself (and others) if I can pull it off.
Additional thoughts:
Forgive me if I missed it, but is there a reason to not run the glulam in the shorter direction? Either 1 or 2 floor beams to reduce the joist spans?
Or am I reading too much into this and this is merely an exercise of the features of your extensionsā¦
The other thought, as an academic exercise, is as suggested by @ivanjones of using the wall above as a ābeamā to stiffen the floor joist to reduce the ābouncinessā. The only thing I would add is to treat the wall like a shear wall and provide additional Simpson A35s at the corner framing of each door opening and Simpson LTP4s at the post/header or Simpson ST or CS or CMST straps across the headers and down each king/jack stud to the floor beam (purple below).
The garage is essentially a box so it is the same dimension either way (32 feet).
I ran the numbers for two beams and it still takes two 24" glulams to achieve a total deflection of about an inch, too much sag for me.
I really like the idea of the shear wall acting as a large beam and I think with enough strapping one could probably achieve an FTAO type shearwall that works reasonably well but I still think the overall sag is not to my liking. The shearwall would certainly stiffen things up considerably, I do fully agree with that, the rigidity created by a 10 ft deep/high wall would be significant.
Splitting the span in two (16 feet) dramatically changes everything. The only thing is the post will be supporting about 20,000 to 25,000 lbs, but that is easy to achieve with any decent sized post and spread footing, just donāt want to bump it with your car or truck.
My question is how much would the shearwall help with the overall deflection?
If I can sheath both sides of the wall with 7/16 OSB and 10D nails and strap the heck out of it around the openings and upsize the glulam to an 8.75x27, can I cut the total deflection to half an inch, or less? If I can, I would probably go that route instead of the post in the garage. The problem is there is no way to actually check this or actually calculate it, it would simply be trial by fire and hope it works as planned.
Am Iām being too picky, is an inch of sag okay? Maybe they can put some crown on the beam so that it negates some of that deflection and make it more reasonable? I know they typically do this anyways with these large beams. Anyhow a lot of fun engineering to contemplate and in doing so Iāve managed to derail my own thread.
Found a critical bug in the in-wall column module, so I spent the better part of today digging through the code to fix this issue. Now that it is fixed I finally inserted the columns under the beam:
I still need to bit more work on the floor plan (two more bathrooms would be nice), but here is what I have so far. Iām excited to make some video explaining some modeling hints that I have for users of the mdkBIM tools. Iāve learned a lot the last few days.
Structurally everything seems to be taken care of except Iām still a bit concerned about the octagon room and its full load path to the foundation.
The other issue is the roofing where the turret meets the main roof on the backside of the turret, not sure how one would frame that out correctly so that it sheds water while maintaining the correct look and feel, bit of a question mark.
I ran a few more numbers, specifically on the I-joists on the third floor in the turret. The central (widest span) part of the turret is 24ā out-to-out which is pushing the limits of any I-joist to span without noticeable deflection and ābounceā.
I didnāt realize it until now but Forte now includes some steel beam analysis (is this new or did I just not notice it before?), so I ran the numbers for a dropped steel beam, and it looks like a W16X40 (or larger) can work nicely here:
Note, that these are 10 foot ceilings so dropping down 16" or so is not going to be the end of the world, still plenty of headroom, and hopefully not too unattractive once it is wrapped in gypsum.
Also note that where the beam bearing point meets the Tee intersection of the three walls things get a bit interesting. In the field one would simply frame this as one wall (the octagon wall), however, as you can see here, to the one side of the beam is an external wall and the other side is an internal wall, so the model is split into two walls. Each wall at the tee intersection is set to āTerminalā for its corner configuration, and the one wall to the left of the beam extends beyond the beam by 1.5" and it is the wall that contains the in-wall column. Again there is more than one way to solve this problem with the Medeek Wall plugin but this seems to be a logical way to achieve the desired result.
I have a few W16 sizes built into the beam module but I should really add some of the larger sizes which I will do very shortly and then release an update to the Wall plugin at about 4:00AM MST.
P.S.
A number of updates and a critical bug fix released for the Wall plugin as a result of my findings. Probably some more to come after I fully digest my experience modeling with my own plugins. As I mentioned previously I need to do this more often.
Very nice! A couple questions: Iām assuming that the 2nd floor doesnāt have vaulted ceilings; if so, would the ceiling joists be created with the roof extension or the floor extension? With regard to the post load path to the foundation, if a spread footing was required how does that get implemented with the foundation extension?
I havenāt really spent much attention on the main roof only because I was going to truss it out and then I switched to a complex rafter roof. The complex rafter roof currently does not have a built-in option for ceiling joists but as you suggest these can easily be created with the floor plugin.
Iāve only ever worked on projects in Washington State and Utah so Iām not familiar with rafter built roofs. I understand rafters, hips, valleys etcā¦ however the ceiling joists and the framing that goes under the hood (between the rafters and ceiling joists) is a bit of a mystery to me.
Spread footings are already part of the foundation plugin, and in this model you will notice that they have already been placed in strategic places to pick up the point loads dropping down from the various beams/columns.
I could see why designing the complex rafter roof plug in to accommodate all of these members would be quite a task if not downright ridiculous for one person. Things really get hairy when you involve vaulted or cathedral ceilings, even just plain old raised ceilings for that matter. I usually will start with one of your roof assemblies as a base and then everything else is custom from there inside of it, most roof plans are simple enough for the builder I work with to figure out himself without all the details, but we have drawn every board before.
Thank you for the clarification with regard to the extensions.
The only thing I can add to @LinearGraphs reply is that the intermediate purlins can reduce the rafter sizes (with reduced span) and that the ceiling joists act as the bottom chord of the rafter triangulation (i.e. like a truss).
When I think of a rafter roof I generally think of the rafters, ridge and ceiling joists. All of the other framing members I typically lump together as bracing similar to bracing found in truss assemblies. Itās necessary but not something I would spend a lot of time planning out, in the field construction.
However I am curious about how these purlins would be constructed on a large roof like this. Also what is the typical depth used for ceiling joist?
