Designing a House 4

[abusementpark]

For those of you who have designed or regularly design stick-framed wood houses, what is your general approach?

I find stick-framed wood structures to be difficult to simplify. There many different angles and lengths of rafter braces and the exact locations of the braces aren't always known. Also, for complex roofs (which are usually the case with houses), the distribution of the loads down to the foundation can be very non-uniform and a pain to accurately trace, particularly with openings that don't line up from second floor to first floor.

Do you just develop a bunch of rules of thumbs through experience?

 

[sdz]

What country are you in? That affects the answer a lot.

 

[abusementpark]

United States

 

[BAretired]

I agree that running down loads in some designs can be a bit of a pain but nowadays, roofs are usually framed with trusses spanning wall to wall.

I try to avoid bearing walls which do not line up from main to second floor, but when analyzing an existing house, you must take what is there.

I have never developed any rules of thumb on such matters. I treat each job as a new challenge.

BA

 

[cancmm]

Haven't dealt with one of these in a few years but here's my shot at it based on memory.

Check your local building code - the contractor is likely familiar with its requirements that prevent you from needed to detail every single connection and member. They generally have "rules of thumb" that can be followed in most cases such as required sizes of exterior stud walls, typical joist connections, etc. Be sure to reference the code in your specs.

Follow the concentrated point loads originating from the roof down to the foundation (girder truss reaction loads, etc.). This will likely require built up columns in most cases, which are very simple to call out. Sometimes engineered lumber sections will be required as lintels, etc. Check the worst case first and work your way down, you'll soon get a feel for the building's reaction loads and won't have to put much thought into it as you get deeper.

Don't over-engineer the thing. The cost savings involved with reducing sections to their absolute minimum could be voided by the time it takes you to check. Keep it simple and conservative. Use consistent details and as few of them as possible. This will also help the contractor avoid mistakes and timely repairs if every component was a different size or material.

 

[dcarr82775]

My rule of thumb is to avoid designing houses

 

[MiketheEngineer]

In many parts of the country - an architect is required and he/she must be proficient in this area.

That said, designing houses really is not too hard. Just do it a few times, go on a few field trips, help build one, etc., etc. and it will seem quite reasonable.

The hardest part is getting the most bang for your buck and that comes from reasonably sizing all rooms, the right amenities and pleasing the buyer.

Construction details are also important to insure a safe, dry and safe structure. A good builder can be a big help here.

AND make sure you meet all the requirements of the local codes and officials. They can make life miserable if not satisfied.

 

[zstructural]

Follow your loads.

Rules of thumb-wise, my advice is any point load over 2 kips should have multiple studs all the way to the foundation. The number of studs will be governed by your plate crushing. The IRC has most everything covered in terms of required framing methods.

 

[msquared48]

My general rules:

Gravity sucks...

Lateral forces tend to be pushy.

Mike McCann
MMC Engineering
Motto: KISS
Motivation: Don't ask

 

[ToadJones]

I built houses for about ten years in a county, where, believe or not there were no building codes (except zoning) or inspections for structural work.
As the "framer" it was nice not to deal with them, but of course, the engineer in me forced me to do the right thing.
For the most part, unless the house is out-of-the-ordinary, the design is mostly empirical. And, for the most part, I believe it should be.
We would simply look over the architects pretty, usually dimension free drawings and decide what needed to be done for the home to be built properly. Most beam sizing was done by in-house engineers at the lumber suppliers. The roof trusses were of course done the same way.

 

[msquared48]

Vertical loads first.
Pre-manufactured trussed roofs are the cheapest.
Stack bearing walls and columns if possible.
The most direct route of the forces to the foundation is the cheapest.

Lateral next.
Stack shear walls where possible too.
Try to avoid using transfer beams and steel moment frames.
At the top story, use only external shear walls to avoid the interior shearwall to truss detail which is messy.

Mike McCann
MMC Engineering
Motto: KISS
Motivation: Don't ask

 

[vandede427]

don't over do it with sections and details. you'll draft your way out of a profit (if you had one to begin with).

have the basic sections:

typical footing section
bearing wall section
wall parallel to framing section
shearwall elevation

and my new favorite, the wood nailing schedule.

i basically reproduced the minimum nailing requirements schedule for wood to wood connections from the IRC table. it's kind of a CYA for any connection you don't specifically call out.

 

[csd72]

You need to be sequential and well documented otherwise any revisions are a real pain.

Start at the roof. Do a sketch with you truss directions, location of girder trusses e.t.c. you can number your truss and girder truss types on this for reference. Now calculate your loads on your trusses and then the reactions on the walls.

It helps at this stage if you also number your different wall sections for reference e.g a,b e.t.c.

Work yor way down the building starting with the worst case for each member and if this looks reasonable make them all this if not then check the next one down the ladder and use 2 sizes.

 

[abusementpark]

Many have brought up the use of pre-engineered wood trusses. Having this option, of course, simplifies everything with the roof. However, most new home construction involves very complex roofs that are not conducive to pre-engineered trusses, and you are forced to use stick framing from a cost standpoint.

On a another note... if I understand correctly, in most states, a structural engineer's stamp is not required for design of a residential structure. I assume that builders normally size members and build the structure based on the empirical/prescriptive guidelines that are given in the IRC (of which, I am not familiar). So, if a structural engineer designs a house, does he have the liberty to use these prescriptive requirements as well?

 

[SlideRuleEra]

If a structural engineer designs a house, does he have the liberty to use these prescriptive requirements as well?

Yes, anyone can use the prescriptive requirements.

If you want to get a general overview of how houses may be designed (all in one book), see "Residential Design Guide - 2000 Edition", free .pdf at this link:

http://www.pathnet.org/sp.asp?id=1442

http://www.SlideRuleEra.net

http://www.VacuumTubeEra.net

 

[zstructural]

To your point about using the IRC prescriptive requirements - I have actually not been able to get the numbers to check out on some of those prescriptive requirements, so sometimes using those prescriptive requirements can help you use a more cost-effective design. For example, in some cases the rebar requirements for a basement wall will not check out if you put the numbers to them. I assume that they've been building foundation walls like that for so long, they've accepted certain rebar configurations based on being empirical.

I've dealt with some hand-framed roofs that were literally impossible to followt the loads. If you design the hand-framing somewhat logical, then you won't have this problem. Or, worst case, you do end up with a ultra-complicated hand-framing system, then it's usually not too difficult to conservatively estimate where the loads are going, without being OVER-conservative.

 

[vandede427]

i've been able to verify some of the prescriptive stuff like floor and roof joists spans. there's nothing wrong with using them; save you time.

p.s. I just looked at a house this morning that had severe settlement issues. normally a geotech eng doesn't get involved with a house, but you might want to encourage the owner to hire one to come out and do at least once to do the "push the rod into the ground" test. please forgive me; i'm blanking on the proper term. what good is it to engineer the structure if it has a soft soil below it.

 

[BAretired]

In my opinion, an engineer should not use prescriptive requirements of a code which he is unable to justify by calculation. There are a number of examples in the Canadian Code which fall into that category.

I agree with vandede that a geotech should be involved in house construction more frequently than is the usual custom. Foundation movements can result in substantial claims, particularly if they were easily predictable by a cursory geotechnical examination.

BA

 

[ToadJones]

Some of the most complicated roofs I have ever framed were with trusses. Truss designers/manufactures can pretty much make any out of a trusses that a good carpenter can stick frame...sometimes they make life very simple as in the case of a hip roof with different pitches meeting at a hip such that the hip rafter doesn't sit at the corner of load bearing walls and is not a 45º angle (in plan).
Having said that, there were many occasions where I chose to stick frame a roof over trusses after seeing how messy the truss plans were going to be.

 

[abusementpark]

Thanks all for the responses.

As a sorta follow-up question: Which states require a structural engineer's involvement in house design? And how is the process usually handled, does a contractor send a set of plans which you review, make corrections, and stamp?