View Full Version : sizing a header

11-28-2006, 07:06 PM
I would like to create an opening in a wall, (approx. 9.5 ft.), in a downstairs wall on a two story residential house. The upstairs floor-joists lap above the area of the prospective opening. There is a 2x4 interior wall above this same area and ceiling joists for the upper rooms lap also above this area. But there is no roof supports of any sort above this area. The walls are sheet-rocked, and on either side of the new opening are two 32" return-air chases. How can I go about sizing the header for this new opening. I have been a carpenter for some 30 years, but have always relied on lumber tech to tell me what I should use. Is there a sizing calculater, (so to speak), that will help me size the header myself? Or can someone advise me what header will SAFELY carry this span? I am more that comfortable in supporting the downstairs ceilings temporarily, while cutting the opening. I just need to know what size of header or lamenant beam will carry this opening.
I appreciate your help....Thanks Ron

Bill Lacey
11-28-2006, 07:50 PM
If you go to WWPA.org, there is an Excel file that has all sorts of lumber calculating functions in it (Design Suite, or something like that). Other than that, you have to figure out how many pounds per foot you're trying to carry, and look up a span table to see how many plies of what height lumber you need to carry that load.

If all that fails, go to either a lumber yard or call an engineer.

Tom Thomas
11-28-2006, 08:08 PM

The rule of thumb that I learned so long ago I can't remember the source; prehaps Popular Mechanics:6ft span=4x6 or double 2x6,8ft=4x8 etc.
So 4x10 or double 2x10 would be sufficient for 9.5ft.

There is a way to retrofit a header without shoring inside. Once you have the
existing studs exposed, cut out a path for one side of the header only half way through the studs. Then install one 2x10 complete with jack stud. Now
you can cut the rest of the way through the other side and add the other 2x10. It is much easier to work on the inside without the shoring in the way.

Have fun,

11-28-2006, 08:11 PM
Thanks so much, Bill. I will check that site out right away. I get so frustrated trying to get info out of lumber yards or finding a cooperative engineer.... Again, I appreciate your quick response and help. Sincerely, Ron

11-28-2006, 08:18 PM
That is great information, Tom. That also gives me a good start on my peticular situation. Yes, I have done installations like that before, and it truely works well. I really appreciate your response and help. Thanks, Ron

11-28-2006, 08:37 PM
Deuce, if it were me, I would use 2-1 3/4" X 12" microlam beams for this size header. The cost is small compared to what might happen if it settles. Also if there are point loads above there is no way a dbl 2 x 10 header would work. Been there before!

11-28-2006, 08:44 PM
A good lumber yard should have an engineer(s) on staff. They do the sizing for all of the engineered lumber--I joists, LVL, truss joists, etc. I don't know what I'd do without the 2 engineers employed by my local lumber yard (The Contractor Yard, Wilmington, NC). I give those guys numbers and situations all of the time for the additions, repairs, and remodels that we do. They enter all of the conditions in their computer and they spit out exactly what's needed. And that printed piece of paper is what's really important. That's the engineer's seal and he's not going to make his recommendation without being 100% confident in what he's given.

Further, if you want to keep this header from sticking down from the ceiling too far, use LVL for the header. 9 1/2' is a pretty significant span with joists for rooms above bearing on that. Also, you probably don't want this header to be more than typical wall thickness w/sheetrock: 4 1/2".

11-29-2006, 06:58 AM
I'm with oldschool, but it might just be a NC thing. My lumber yard will also size beams for me for free. Whoever published that rule of thumb that Tom mentioned needs to be pushed back under the rock he crawled out from. That is the worse rule of thumb or advice I have read in a long time. Tom I hope your not using that for your projects. I really hope your inspectors aren't allowing you to either. A double 2x10 could never carry the load across a 10' span for a 2nd floor, its furniture and people, the ceiling above and any attic storage that is being used on those ceiling joists.

Ron, take the advice that Bill Lacey gave you and also go to your lumber yard and see what they can do for you.

11-29-2006, 05:45 PM
Maybe 15 years ago, engineer Robert Randall wrote an article for JLC on sizing headers that included tables. Even tells you how many jack studs you need at each end.

James Eggert
11-29-2006, 06:33 PM
Charles brings up a very interesting point. The IRC codes require multiple jacks as the opening increases.
Check your adopted code!

Dancing Dan
11-29-2006, 07:22 PM
Ron - are you really going to size a beam carrying a significant load based on advice you receive on an on-line forum? You are receiving some good advice, but span charts aren't that hard to find, and your lumber yard should have a pamphlet from the LVL manufacturer that can answer your question in about 5 minutes.

I don't mean to seem nasty, but it's an awful lot of liability and it all lands on you.

11-30-2006, 07:27 PM
This long adress is to the header tables on the CWC's Spancalc. Click on the house section that describes the situation and its chart will pop up.


12-01-2006, 06:26 AM
Good link Don. I won't use that as a guide to go by but as a reference for ideas.

Bill Lacey
12-01-2006, 03:30 PM
Why wouldn't you use that as a guide? I have all of those tables in a bound book form and I use it as a guide for spans all the time.

12-01-2006, 05:00 PM
Maybe I'm over my head on this one, when I say guide, I mean that I won't go by it alone. I'll still check my codebook and verify with my engineer in cases I'm not sure. Is that better? I didn't read the site enough to know whether the information contained was excepted in my area. Therefore its just a reference to me. One that I appreciated and would have left at that.

12-02-2006, 04:10 PM
I know what you mean Rob, nothing like reading from the same book. These are good tables though, they come from the people grading the lumber and are based on known design values, that they helped develop.

Here's a couple more worth bookmarking;


12-02-2006, 05:04 PM
thanks Don,
I love the southern pine one, everything we get here bigger than a 2X6x10 in width and length is syp. Sometimes even our studs are but not to often. I love it when it is fresh, but once that stuff gets a little age on it, its like a rock.

12-02-2006, 08:15 PM
It's a simple calculation but it can't be done accurately without knowing the span of the floor and ceiling joists, in other words, how big is the house?

12-03-2006, 10:30 AM
I've written this description of the math involved in sizing a uniformly loaded simple beam. I'd appreciate feedback on making it easy to understand and correct.

12-03-2006, 11:50 AM
In your example the 8x16 would work but the 10x12 is pushing a deflection limit of L/360 if manufactured to less than nominal size.

For something this big I would use LVL's from the manufacturer's tables.

Here's a cheap, easy to use manual calculator for beams and joists

Here is a calculator that might be useful to you if you like to size beams by calculation.

12-03-2006, 02:31 PM
I wrote that using an example a sawyer proposed for a shed he needed to build, as you could see his guesstimate was a little light. When I suggested those dimensions a few people thought they were pretty large. On a longer span or wider building, under different loads, rules of thumb or comparisons to dissimilar work can get dangerous fast. A squarer section is easier to find in a tree than the preferred deeper section was why I proposed both shapes. That is a good observation worth noting in there somewhere, given 2 beams of equal roughly section, the deeper will deflect less. I should rewrite it with a typical built up header or girder as the example. Having a sawmill myself, being able to safely size beams of different species and grades has been one of those side journeys I never anticipated when I first started out.

LVL's do have a much higher E value, or stiffness, than solid sawn wood. If floor vibration is a concern beams and headers probably should have an L/600 deflection or less. SYP is another wood with a high E value... good stiff stuff, about like iron when it dries :)

Thanks for the section properties link Sweep. I'm learning to understand timberframing connections and how that relates to the members. If you run across a section properties calc that lets me whittle notches in the section at will, please holler!

12-03-2006, 10:18 PM
I just need to know what size of header or lamenant beam will carry this opening.

You also need to know how many studs you need to support the beam and what you need UNDER those studs to carry the newly created point loads safely down to the earth.

12-04-2006, 08:40 AM
Compared to visually graded No. 2 SYP lumber, the strongest grade of LVL's are 25% stronger in deflection and 200% stronger in bending. Making big beams out of sawn lumber doesn't make a lot of sense to me.

Joe Bartok
12-04-2006, 09:26 AM
A squarer section is easier to find in a tree than the preferred deeper section was why I proposed both shapes.

Section Moduli (http://ca.geocities.com/xpf51/LOG_BEAM_SECTIONS/SECTION_MODULI.doc): For the folks who like to play around with the math the solution of the preferred or optimum rectangular section which can be extracted from the round is on Page #3.

Joe Bartok
12-04-2006, 11:56 AM
If you run across a section properties calc that lets me whittle notches in the section at will, please holler!
Yeah, there's a thought! Don, I usually end up doing this the rough and dirty way. Parallel Centroidal Axis Theorem for Second Moments of Area. Shear stress for irregular sections (http://ca.geocities.com/web_sketches/misc_html/shear_by_integration.doc) is adapted from the book referenced in the linked document.
I've been thinking of writing some code where one could trace the outline of a scale drawing of the section in question. The program would track the position of the cursor and estimate the section properties. The calculation wouldn't be a true integration but rather based on the sum of a series of rectangles, where each row of pixels represents a rectangle - this should be close enough for most real world applications.

12-04-2006, 09:12 PM
Good points Sweep.
If only my life were always so easy.
The design often dictates the choice. For one example, I've been looking at a 1790's resto. The exposed beam second floor is in some distress, a large solid sawn summer beam is probably the best solution there. Headroom is at a premium as is often the case in old work. The beam will need to be relatively wide for its depth. Not very material efficient, but that is not the primary concern in that case. Under the main floor and up in the attic I'll probably slide in LVL's.

My prints often spec #2 or better spf headers and girders, switching to a higher grade syp makes them both stiffer and stronger than design at little or no increased cost. An LVL would be stiffer and stronger yet.

I have rewritten the bending example on that page to something a little more typical, the pictures need some work yet.

Thanks for the links Joe,
That sounds cool, I'm rootin for you, thats way outta my programming knowledge. A professor at one of the engineering schools has been breaking notched heavy timbers and is working on the math. Holler if you want some notes :)

12-09-2006, 08:32 PM
I truley appreciate all of the help. I do not take this project lightly, and I have always tried to go overboard, rather than just trying the minimum. Unfortunately, since I moved to Memphis, I have yet to find a REAL lumber yard that can help in a situation like mine. No one wants to recommend of advise me of specs. for a 9 1/2 ft. span. But I will put to good use all of the help....Thanks Ron

12-09-2006, 10:47 PM
The rule of thumb I read about header sizing ie. 4x4 for 4' opening, 4x6 for 6' opening, etc.... only applies for single story structures and simply supported wood framing structures.

12-10-2006, 06:33 AM
What is a simply supported wood structure? Another crappy rule of thumb. Thank god someone invented span charts. Or who knows where the rule of thumb would take us. I was reading "Residential Structure and Framing" and one of the lines in the very beginning, said a good carpenter can not only frame the structure he is tasked with, but he can also advise on area's where the structure isn't sound. I get disgusted when I meet someone who says, yeah I'm a carpenter or a I'm a contractor and they couldn't even build a shed better than a Lowe's prebuilt one.

12-10-2006, 08:56 AM
A simple beam rests on a support at each end, the ends of the beam being free to rotate. Thats a typical header, joist, etc.

A cantilever beam is supported at one end only. A beam embedded in a wall and projecting beyond the face of the wall is an example.

An overhanging beam is a beam whose end or ends project beyond its supports. The old raising plate cj's projecting to form soffit and support rafters is an example.

A continuous beam rests on more than 2 supports.

A restrained beam has one or both ends fixed against rotation.

Loads can be concentrated or distributed. We most often suppose a uniformly distributed load on a simple beam.

A table is simply a series of points worked from formulas. A case could be made that a carpenter should know his formulas for when the tables run out on him. Rules of thumb can be remembered, often used points, from the tables. Making sure either applies to the situation at hand is the key.

The header tables in the codebook, the ones in the CWC Spancalc, WWPA's, or from various manufacturers should do the job here. If not, have an engineer check it out.

I hijacked above, sorry bout that. Sweep had asked for the building dimensions.

12-11-2006, 07:05 AM
I think a "Carpenter" should know his formula's or atleast where to find them and have a sense of how to use them. I expect a framer to know how to read a span chart. Swinging a hammer when told for a couple years doesn't necessarily make someone a carpenter or even a framer for that part. Rules of thumb can be dangerous if they have specific guidelines that come with them. When talking about headers and beams, a rule of thumb can get real dangerous when in the hands of an unexperienced individual. I would hate someone to read here that a rule of thumb for headers was to use dimensional lumber like 2x6 for 6' span. Its to vague.

I liked this guys signature from the bouncy floor thread.

"Disclaimer: I am a structural engineer, but I'm not your structural engineer. Check with a local PE before you try anything I suggest."

12-11-2006, 07:16 PM
Hmmmm...that would be me...the new guy here.

Can I ask why the correct answer isn't R502.5(2) in the IRC? Maybe the answer is that the OP isn't in IBC territory (almost everybody is now).

Look up the number of stories (1 or 2), the width of the the building (20,28, or 36 feet - take the higher number if your're not comfortable interpolating), and read down the "span" column to find the span you need (9'-6"), then across to the left to get the header size, and to the right to check the number of jack studs on each side. If your span is not in the table, then you need an engineered solution (LVL, PSL, steel)

The reason those tables are there are because residential builders encounter these types of conditions so frequently that it would be a waste of money to pay an engineer $100-$150 to size a beam and draw a detail or write a letter (and the associated paperwork) for the building department. Unless you're really comfortable with calculating tributary loads, I would go easy on using a "sizing calculator."

12-11-2006, 08:25 PM
Show of hands? How many framers own a codebook?

12-12-2006, 07:33 AM
I do, but not just because I'm a framer, I'm also a licensed contractor.

Jordan how about getting licensed to sign items in NC. My engineer charges $300, but he gives me peace of mind. I got one right now I'm taking over to him today or tomorrow. Steel I-beam spanning 27 ft. W12-22 I want to replace it with LVL's. Want to take a guess at what he is going to make me use?

Bill Lacey
12-12-2006, 08:37 AM
Doulgas Fir 3-1/8"x22-1/2" or 5-1/8"x19-1/2" would be my guess.

12-12-2006, 08:39 AM
What about multy ply 1 3/4" LVL's?
Thank you

Bill Lacey
12-12-2006, 08:55 AM
What is the tributary width you are trying to carry? Are there any bearing points along the beam? There isn't much in LVL that will span 27' and carry a lot of load.

12-12-2006, 09:12 AM
Your talking like an engineer, I'm a framer by trade. The beam is in a 27' wide X 30' deep garage, running parallel to the garage doors 16' in to break the span of the floor joists from the room above and keep the garage floor wide open. The floor above is a bonus room under the roof line, rafters are designed to span to exterior bearing walls, so there aren't any bearing points on this beam that I know of.
When the lumber yard did the material takeoff for the house, which I review for accuracy, they substituted a quadruple 1 3/4"X20" LVL for this. Of course I have to have an engineer stamp any changes so I will be taking it to him today probably. Just wondering what other engineers would spec, sort of a way to check on mine. Not that I don't trust him, just curious if he's giving me bare minimum or overdoing it. I appreciate the advice.

Bill Lacey
12-12-2006, 09:34 AM
4-ply 20" makes it, but it's going to flex (over 78% of allowed deflection at L/360). A 4-ply 22" or a 3-ply 24" would be a better choice (only 60% of allowed deflection). Are you loading it as a living space with 40/10 loading or a storage space with 20/10 loading. 20/10 will give better/smaller results.

Disclaimer: I'm not an engineer or an architect. I'm a CAD Technician who has 9 years experience selling engineered lumber and I've got sizing software to use, plus multiple span charts.

12-12-2006, 09:40 AM
Well I'm leaving this room as "unfinished attic storage", for the homeowners property tax purposes. But I suspect that he will finish it after the permit is closed out and we get a CO. So I want to make sure it can be used as living space. I don't want a callback. Thickness of the beam makes no difference to me, so if 3ply 24 was better I'd use it, so I'll see what my engineer requires me to do in order to stamp the prints. Thanks Bill.

12-12-2006, 10:30 AM
I'm licenced in NC, WV, TN, MD, and VA. And I probably just violated the no-AD TOS of the forum ;-)

Personally, I like steel. W16x26, W14x30, W12x35 all work as long as there's no roof load. The shallower you get, the bouncier - I'd go with the 16 if you have the headroom. 4 - 18.75" LVLs is what I get in lumber, or 5.25x23.25 PSL, even bigger in standard glulams 25-30+" deep (didn't try an Anthony Powerbeam).

12-12-2006, 11:07 AM
Thanks Jordan,
I guess the guy that works at my lumberyard was pretty close then with the 4-20" LVL's. Honestly the reason I dumped the steel was the extra logistics of it. The lumberyard stocks the LVL's and its a lot faster than the steel. Included with the steel beam was 2-2X12 on both sides between the flanges bolted 2' o.c., plus a 2X8 bolted to the bottom as a nailer. Just seamed like a heavy pain in the but, so I opted to pay an engineering fee and switch to wood.

12-12-2006, 12:39 PM
I'm a bit suprised more people don't use self drilling screws for attaching wood to lumber (maybe their engineers don't let them?). A #14 with a #4 point will go through just about anything a framer is likely to see. It's a little messy for the sides of the beam, (need to clamp all of it on or you'll never get the second side attached) and you probably have to order the screws too. But for a nailer on the bottom flange (or a top flange with a top bearing hanger) there's no reason not to use them.

Bill Lacey
12-12-2006, 12:44 PM
Use TrussLok-Z screws from FastenMaster for multi-ply LVL connections. They are designed specifically for LVL use (1-3/4" multiples) and have more shear strength than a 1/2" bolt (per thier ads). Rep did a demo on site for me and a few contractors using cupped LVL, and the screws brought them back in line.

12-12-2006, 01:53 PM
My rule of thumb regarding structural calculations is "The only person qualified is the engineer of record."

I'd fire the carpenter who took it upon himself to calculate the required beam for a span. Our buildings are too complex, the codes to strict and the liability too high.

How big does that 2'0" window header need to be that has a structural ridge point load hitting it 6" from one side?

If the engineer of record didn't provide it on the stamped plans, they will do it free of charge in every condition I've ever run into. If it's the result of a change, then the owner needs to pay for it anyways.

12-12-2006, 08:39 PM
I'd check his work. If it was within the prescriptive allowances of the code, done with understanding and done correctly we'd have a little talk and I'd consider him for a raise. I know you are mostly fully engineered out west, we are not, yet.

This exception on a readily available beam sizing program pretty much sums up my feelings;
Exception: Use of the Software by structural design novices is acceptable if the results are checked by a structural design professional who verifies and certifies the data entry and Software results to be accurate. Failure to do so may jeopardize the safety of the public at large and breaches a prime condition of this license, thus immediately terminating this license. You the licensee, assume full responsibility for the selection, installation, use, results, and application of the results of the Software.

Robs example above is a uniformly loaded simple beam, floor load only, its not rocket science. At 2.0E I'm not comfortable with the 18.75" calcs, I'm happier with 4- 20's and happier still if 3-24's will fit
If you understand you can verify for yourself whether the engineer is going light or heavy. Checking the math myself, if the engineer specced the 18.75's, I'd bump up to 20's to make myself happy if possible. Being able to double check is reason enough to know how to do this.

Minutes of looking turned up this set of LVL tables, http://www.gp.com/build/DocumentViewer.aspx?repository=BP&elementid=4372

Page 43 of the above pdf shows 4ply 24" 2.0LVL's will work...prescriptively, since this product has an ICC ES#. Bill's software is also giving prescriptive solutions and is leaner than the general table above. I don't have the coastal restrictions you likely have, for that job my inspector would not require an engineer, the entire garage could be done within the prescriptive allowances of the code.

Bill Lacey
12-13-2006, 04:05 AM
Actually, the PDF you linked to does allow for a 3-ply 24" 2.0E LVL, since we have a LL of 600 PLF and a TL of 750 PLF.

Using 40/10 Loading:

15 (Trib Width in feet) x 40 = 600 PLF LL (746 PLF per table)
15 (Trib Width in feet) x 50 = 750 PLF TL (984 PLF per table)

Using Rosburg's software, if the bearing size was modified to 5-1/2"x(beam width) the values allow for a 2-ply 24", but it's starting to max out.