Full Hard Covering Of An Attic Floor

This job experience was imagined as content in Fine Homebuilding Magazine. Subject: the protection of insulation integrity and value, by complete air-barrier hard covering. The inspiration? Suggestion by writer Martin Holladay, that complete hard covering should apply to skylights. Great idea, I say, but it is simple logical extension then, all of an attic floor should be covered too. The example which follows is quite fully doocumented in a job photo album at Google My Drive: Woodstock Attic, Challenged Roof Joists.pdf  . 

 And, it turns out that another writer is not much wanted, at Fine Homebuilding Magazine.

I have many similar jobs with substantial floor decking, upon strong supports, found by picking blog Label: Strong Attic Floors , here, or at the right. A unique feature of this job is use of attic floor supports to keep attic weight from loading very-long 2x4 roof joists. Second floor walls do not directly support the roof, here, and fairly often. Include label: Roof Strength for this post. Try that to find more-complete discussion.

This graphic shows structures in overlay upon a satellite photo, drawn in Adobe FrameMaker.

At 6/26/2026, observe this sign of public interest:

A genuine reader, in Argentina! This inspires me to revisit and update this post. 

Full Hard Covering Of An Attic Floor

An attic floor over composite beams does much more than the protection and enhancement of insulation value. Flooring is imperative as flanging of the composite beams. Anywhere, well-attached facing of a beam adds stability and strength far better than lumber blocking between beams. It is weird that we commonly block floor joists over a basement or crawl space, rather than applying an underside covering of modestly-thick plywood that encapsulates a full filling of insulation. The excuse is in wanting readily-accessible space for wiring and plumbing, but that is not compelling reason. Any attached flanging or webbing of a beam is interruptible with affordable penalty in strength. Absent flanging at both edges of a beam, compressive loads in a beam tend readily to cause rotation and buckling, dumping transverse loads.

Focus now on the job at-hand where roof 2x4@16 framing is extremely long. Expand an offered insight upon what is achieved in roof joist 2x4s where the attic floor is strengthened by composite beams:

This is the composite-beam floor thickening and strengthening that evolved, over North and South bedrooms. 2x4 floor joists, some twisted, some bowing downward, are locked by the beam plywood. 2x4 nailers, ten ft length, are bound too, to the roof joists. A short couple in each beam line acts against bowing of roof joists, none yet detectable. Surely, added deck structures are helpful, not detrimental, and all work in the widespread construction of composite beams, stabilized by attic flooring, is justified. 

New insulation is in three layers of unfaced batts: upper and lower R19, and wider R11 batts  filling 2x spaces, in between. I don't apply flooring out to the difficult edges. Consider covering the periphery instead, with heavy cotton throws. Those throws would collect dust carried up into the attic through the heat engine of roof/ ceiling slopes. The throws could be laundered in periodic attic maintenance that includes vacuuming of roof vent screens, reducing need of a respirator in-attic.

Always add as much insulation as space allows.












The total of insulation everywhere is R49, in a depth 9.25" where compressed under decking. I think the air barrier and the added boundaries of conductive layers is good compensation for the small batt compression. This attic is improved from R4, to better than R38. 






None of this is for storage, though the space is now usable.

A possibility not taken, was in the early discussion. Slope ceilings are very badly insulated, and room usefulness would be little-diminished, if slope ceilings were rebuilt much-strengthened, with the composite beam method. The entire envelope over first-floor rooms could be raised to R49 super-insulation levels.

In most bungalow situations, knee walls perhaps-needed, interfere with possibilities of demolition to enable insulation.

Strong Attic Floors

Find this and much more in writings at this blog, pick Label: Strong Attic Floors, then finding posts in reverse chronological order.

At 3/14/2024, add a link to a thorough conversation at Fine Homebuilding, in Fall 2023:

https://www.finehomebuilding.com/forum/old-house-2x6-joists . In a back-and-forth with several competent builders there is sobering demand of scarfing, as essential to beam repair with sistering. Refer to  Forestry Forum. Scarfing is first a means of lengthening a beam, while (wishing to) preserve strength as a single-piece beam. In fact, a scarfing joint (near an end), is not an acceptable means of placing a sistering joist that bears upon supporting walls. The scarf-joint reduces beam strength by nearly half. A precise penalty is not, and can not, be offered. Don't expect a professional engineer to weigh in on this. Builders generally agree that scarfing is not appropriate for a load-carrying beam. This leaves sistering a foolish solution, never ever doubling support strength.

Here is a scarfing illustration offered to the Fine Homebuilding discussion:

Here is a Forestry Forum illustration of a usual crudely cut/ traced pair of scarf-cut beam ends, necessarily preloaded in some tension with hammered-in wedges. Bind further with machine bolts and washers. The wedge space is made rectangular with difficult cuts at odd angle to the beam axis.

Here is my design of scarfing with steel templates that might permit accurate scarf cuts, loaded to tension with only a pair of lag screws.

Bond from the accessible top surface with a dozen smartly placed screws.

0.220" x 5" Strong-Tie SDWS22500DB-R50 

Timber Screw - Double Barrier Coating, 

Don't be diverted into trying scarfing. The solution in gaining load capacity is in making beams taller.

At 11/19/2012, this post is edited to include the lead-in phrase that follows, a sub-title:

Composite Beam Method In Residential Construction
I would like to claim those simple words for my practice of making wood beams taller as reinforcement, rather than wider (called sistering or doubling). I stitch rips of plywood (called web), to found 2x framing, as bridges to added 2x4 (called nailer). The added 2x4 are longest-possible match of the length of found framing.

 

















In fact, "composite beam" is more commonly descriptive of material mixes including wood/ steel (Flitch Beam), and wood-concrete.

In my work in attics, it is sometimes structurally necessary that I provide substantial decking of the attic floor. 

That attic flooring then provides novel, efficient roof supports.













Here are three examples in a Picasa Web Album. In the first two, I have repaired broken floor joists where they must support needed attic ladders. In the third, I have provided load bearing capability over 2x4 truss bottom elements otherwise flimsy. In all three, the strength of a composite beam is assured only where 2x elements at bottom (ceiling of rooms below) and at top (attic deck surface), are not free to rotate.

Here is a graphic illustration of the composite beam application with 2x4 trusses:

I apply ample decking, with needful disregard of restraints. Search at Google "Energy Trust of Oregon". "Fraction of attic floor decked." Find 2012 Weatherization Specifications. Find AT 2.4:

Sadly, there is no forum for public input to rules. I can only try to influence, by this kind of public expression. Here then, let me ask: What is a "cavity to be filled"? I will link this by email to Energy Trust, and see what they have to say, on the broad topic. I invite them to respond by comment here.

This post was imagined, in dealing with structural analysis of composite beams, as presented in another public forum I have joined, DIY Chatroom:
http://www.diychatroom.com/f19/strengthen-ceiling-joist-151512/#post1046734

Perhaps readers here will see value in composite beam construction, and might comment. I think floor reinforcement by "sistering" has no validity, where added lumber couples that found, with rotational freedom and unlikely bearing on supports below.

I wish for academic and professional comment. Lacking that, I can attempt my own defense. The ultimate would be deflection testing of sample beams, which is within my capability, but would best happen with support of an interested college or university. For now, note that beam stiffness and strength are in linear proportion to thickness, and cubic to height. Sistering at best doubles strength. A composite 2x10 is likely at least four times stronger than a 2x6, and more than ten times stronger than a 2x4. A composite 2x12 of the same 2x lumber is strengthened by a further factor of two. What wonderful and easy control of strength! Testing and questioning are hardly justified. Does it matter that you might increase 2x4 strength by times thirty? Maybe it is just the added insulation depth that matters. Often it is sufficient to just do all that the situation allows for strength and insulation, as in availability of headroom. Sistering is best effort, only if space for making the floor thicker is absolutely unavailable.

Interest in this has appeared in a re-post to Fine HomeBuilding. I have graphics to share there, which I will copy from expansion of this post. I explain in a further example, that of the repair and strengthening of a 2x8 floor joist in my crawl space. Here is a photo.

And, here is a cross-section detail. This beam bears more load than any other in my crawl space, yet suffers a carve-out passing the bathtub drain. I go way beyond compensating for the divot. Surely the reinforced beam is as strong as a joist of 1 1/2 x 12" cross section. Better strength is by the ratio (12/7.25), cubed. That is, by times 4.5.

I worried some about that beam before my remodel. I won't anymore.

Real Air Sealing In An Oregon Attic

I offer a contrast between charged silliness of Home Performance with Energy Star (HPwES) air sealing, and real sealing, of an Oregon attic. Work complained against, is by the most-respected HPwES contractor in metro Portland, Oregon. The HPwES work was  completed Fall 2008, and left the home occupants feeling no improvement of comfort, helpless and had. I claim a friendship with the owner of the HPwES business, and invited him to confront my complaints, as they hatched. Instead, shielded by the power of a large national commitment to HPwES, he simply challenged me: "How long have you been in this business?" Pathetic me? Six years of energetic learning. He? 25 years. Of such work? I think not. Businesses become corrupted by stupid practices, and the decline likely is only in recent years. There is a basis for defense and pride, but it is not likely that work tied to HPwES, has improved since 2008.


Here is an album showing what I found, and how I fixed it.


There is a lot in the link, for an average reader. Let me summarize. The HPwES contractor accepted a job of improving the "performance" of an attic with dangerous and limited access, and badly-placed found insulation, intended R19. Work done was messy slathering of goop on the home's air handler return ducts and header boxes (air source heat pump electric heat already installed), a wrap of unfaced R11 batts on those ducts, and an aimless hour or two of whimsically dispensing a couple of cans of orange spray foam. That orange foam was abhorrent to the home owner, oozing in kitchen cabinets, and in other areas of the living space. I take it, this is the taught practice of HPwES. Nothing more brainy happens in the "holistic" approach to weatherization. The foam touched gaps of about 5% of the home's drywall wall headers on the attic floor, and was all for a very offending "show" elsewhere in unattached piles and gobs. I infer that a door on the air handler closet whistled under blower door influence, and was quieted with spray foam and new gaskets. The "after" blower door test result can only have been with not-allowed closure of that closet door, for result of more than 1100 CFM50, vs required 300 CFM50, minimum for rebate. Quite an achievement! The HPwES contractor might claim even now, that there was no error or wrong-doing. The rebate sponsor, Energy Trust of Oregon, will not respond to other complaint in the job, against ineffective crawl space insulation. I think then that the only way to challenge arrogant HPwES, Energy Trust, and Bonneville Power Administration, is in this kind of public airing.


I have been waiting for HPwES contractors to defend themselves where I have accused them in several places on a public forum: Here, and  here. Responses are aw c'mon, or don't spoil the "game" of auditing. This isn't about taking a home owner's money to do an audit. It's about saving energy, and fixing things. We don't need auditors. We need people who fix things, and document their work for homeowner records. Anyone who does things or writes things that do not have honest value, is part of a problem.

Roof Cross Braces

Getting around in an attic is sometimes made difficult by head bangers, there for uncertain reasons. The function of ties under a roof peak, between roof joists, is to resist uplift and ruin of a roof, in storm conditions. Some will think the purpose of ties is as found in code requirement of collar ties not more than one third distant from tops of walls, to the roof peak, to hold walls in absence of ceiling joists. Collar ties will have little relevance to residential construction.


Here is a slide show review of the problem, and some solutions.


Here is another view of the solutions, in a web page.