Sweeton Window Wall and Roof Resto/Rehabilitation

To control SPAM, you must now be a registered user to post to this Message Board.

EFFECTIVE 14 Nov. 2012 PRIVATE MESSAGING HAS BEEN RE-ENABLED. IF YOU RECEIVE A SUSPICIOUS DO NOT CLICK ON ANY LINKS AND PLEASE REPORT TO THE ADMINISTRATOR FOR FURTHER INVESTIGATION.

This is the Frank Lloyd Wright Building Conservancy's Message Board. Wright enthusiasts can post questions and comments, and other people visiting the site can respond.

You agree not to post any abusive, obscene, vulgar, slanderous, hateful, threatening, *-oriented or any other material that may violate any applicable laws. Doing so may lead to you being immediately and permanently banned (and your service provider being informed). The IP address of all posts is recorded to aid in enforcing these conditions. You agree that the webmaster, administrator and moderators of this forum have the right to remove, edit, move or close any topic at any time they see fit.
Post Reply
DRN
Posts: 3944
Joined: Mon Jul 10, 2006 10:02 am
Location: Cherry Hill, NJ

Sweeton Window Wall and Roof Resto/Rehabilitation

Post by DRN »

Much as we wanted build a detached garage before we undertake the window wall and roof, recent deterioration from an unusually wet summer and an unusually harsh winter have caused us to reconsider our triage plan. I'm currently working on drawings and details which will direct the repairs and introduce reinforcements which should prevent reoccurrence of the structural deficiencies and address the issues of rot, hopefully extending the life of the repairs.

As part of the work, six structural mullions at the French doors are to be replaced. The originals were anchored to the floor slab by two 3/8" steel pins set into the concrete and let into holes drilled into the mullion endgrain. The mullion was not continuous to the roof framing, rather it was interrupted by the transom bar which was semi-continuous over length of the window wall. A short mullion segment was then toe-nailed to the top of the transom bar and extended to the underside of a rafter to which it was toe-nailed. When coupled with the deflection of the ridge ( two 2x6's spanning 28 feet!) the mullions, under lateral loading, hinged at the roof plane, the transom, and the slab, such that the wall bows outward in both plan and section. The roof overhang at this wall is only 12", and due to the slope of the roof and the absence of a gutter, the mullion bases and bottom rails of the windows and doors are soaked at every rainfall or snow melt. The original cypress lower mullions rotted through and were replaced with Douglas Fir in the 1980's which has rotted even faster. About half of the white pine French door bottom rails have been replaced, and all replacements and the originals that remain are rotted through currently.

Roof framing reinforcement and roof framing connection details aside, (they are being addressed) my current puzzle is the mullions. The replacements will extend from slab to roof, that is certain....the existing transom bar will be cut into segments and let into the new framing. Given the shallow eave and water shedding, the mullion and French door bases at Sweeton seem to be more vulnerable to weather and water than those at other Usonian houses which may have: deeper roof overhangs and/or no sheet flow drainage at the window wall. (For these reasons, I'm considering a shallow gutter at this fascia only to direct water to either end of the roof.) Mullion wood endgrain in contact with (often damp) concrete has led to a drinking straw effect. I'm heading toward a galvanized (Likely powder coated) steel mullion "shoe" which will separate the wood from the concrete, and provide a means to anchor the mullion directly to the foundation, rather than the unreinforced floor slab which the rotating point loads have snapped like a cracker. The "shoe" will be an "open back hightop"...the high sides (12" to match the top of the door bottom rails) will shield the wood from rain spatter and the open back will allow the mullion to be slid into place from inside the house once the shoe is set on the foundation. I'll send some pics and sketches to SDR for posting.

The big question is species of wood: Wright chose cypress...it rotted anyway. I'm providing protection, but I don't want to do this again when I'm 70.

Do I consider Ipe? It is hard, it resists rot. Thoughts?

pharding
Posts: 2253
Joined: Sat Jun 25, 2005 5:19 pm
Location: River Forest, Illinois
Contact:

Post by pharding »

I would not do the metal shoe. I would do old growth or recovered Cypress and treat it with a wood preservative. I would do the continuous vertical plate that usonians typically had. I would use polyurethane on the end grain. That will never leak in your lifetime. Plus I would slope the concrete pavement away from doors and windows.
Paul Harding FAIA Restoration Architect for FLW's 1901 E. Arthur Davenport House, 1941 Lloyd Lewis House, 1952 Glore House | www.harding.com | LinkedIn

SDR
Posts: 19311
Joined: Sat Jun 17, 2006 11:33 pm
Location: San Francisco

Post by SDR »

DRN's photos:


Image
1 Ceiling Cracks from Deflection


Image
2 Window Wall Bowing in Plan and Section -- interior


Image
3 Widening Slab Crack at Length of Window Wall


Image
4 Discontinuous Structural Mullions.


Image
5 Outward Bow of Window Wall


Image
6 Rot Through at Mullion Base


Image
7 East Facade


Image
8 Horizontal Deflection of Mullion Segments

pharding
Posts: 2253
Joined: Sat Jun 25, 2005 5:19 pm
Location: River Forest, Illinois
Contact:

Post by pharding »

You have numerous structural issues that need to be addressed. There are number of things going on that need attention in addition to the columns that were referred to as structural mullions in the original post. Just to name one, the columns were built incorrectly. They need to be one piece from slab to structure above. We worked on similar issues at the FLW 1951 Glore House.

The wood rotting issue at the base of the wood column is attributable to poor maintenance by prior owners. On a lot of these FLW Houses maintenance gets deferred and small problems became major expensive problems. The original design of the wood columns at the base is fine, if built and maintained properly.

Good luck.
Last edited by pharding on Tue Mar 18, 2014 7:45 am, edited 1 time in total.
Paul Harding FAIA Restoration Architect for FLW's 1901 E. Arthur Davenport House, 1941 Lloyd Lewis House, 1952 Glore House | www.harding.com | LinkedIn

peterm
Posts: 6196
Joined: Thu Mar 13, 2008 10:27 am
Location: Chicago, Il.---Oskaloosa, Ia.

Post by peterm »

Were the wood structural members originally specified to be painted? If so, what speaks against using a more rot resistant species like Ipe?

Is there space between the concrete and the bottom of the post/mullion, or does the mullion rest directly on the concrete?

I have seen examples of post and beam construction where the wooden post does not actually touch the concrete base, lessening the chance of rot.

DRN
Posts: 3944
Joined: Mon Jul 10, 2006 10:02 am
Location: Cherry Hill, NJ

Post by DRN »

pharding wrote:
I would do the continuous vertical plate that usonians typically had.
Do you have an example of this detail that could be posted? The Sweeton roof framing is bit unique. I'll have SDR post my preliminary framing section which reinforces the mullion to rafter connection...rather than toe nailing the mullion to the underside of the rafter, I'm extending the top of the new mullions to the underside of roof sheathing to allow the existing cut rafter to be sistered and gusseted to the sides of the mullion making a more rigid frame.

peterm:
The mullions are specified to be shellac'd or coated with spar varnish depending on which sheet of the drawings is consulted. The 1980's replacements were painted when epoxy resin patching was done intermittently during the ten years before I bought the house. The resin patches are falling out as the surrounding wood rots from within. The mullions are detailed on the original drawings to be in direct contact with the concrete.

By contrast, the vertical mullions at the front porch of Sweeton are original and in good condition. The portion of roof they support has not sagged causing rotation, and the 3'-6" to 8' deep eaves have kept the mullion bases relatively dry. That said, I believe the east window wall at Sweeton is an entirely different condition that calls into question for me the soundness and durability of the original detail. What may have worked for a flat-roofed or shed roofed house with 3' or 4' eaves and no sheet drainage, is not applicable to an unrestrained gable roof (no collar ties, no ceiling joists, inadequate ridge beam) with 1/3 or 1/4 the eave protection and sheet drainage. I believe endgrain of (topically treated or not) wood, particularly with 3/8" dia. x 1 5/8" deep holes drilled in them should not be in direct contact with concrete on an east facing facade that is subject to weekly, or more, soaking from roof eave spillage that is only 12" away. I'm also suspicious of pointloads at 4' o.c. bearing on an unreinforced 3 1/2" thick slab, that per the drawings, is not turned down at the window wall, thickened at the load points, and has no foundation beneath it....the ever widening slab cracks 6" inboard from the window wall bear that out. Two 3/8" diameter pins set in just a thin, unreinforced slab, let into the end grain of 2x6's at 4' on center is not a secure anchorage for the support of a roof, particularly one subject to lateral loading from wind, snow loads and heavy swinging French doors.

pharding
Posts: 2253
Joined: Sat Jun 25, 2005 5:19 pm
Location: River Forest, Illinois
Contact:

Post by pharding »

I'll put up a sketch when I have time which will be in a few days. Part of what is happening is that the design of the roof framing places outward thrust on the columns between the doors or windows. That is why you get a bulge in the center. Plus the wood columns are not continuous which is awful. Ultimately you will need a fine structural engineer to work with you. It is important to rely on engineering as opposed to just nailing more wood up. What is called for is surgical precision.

You may also wish to fix the deflecting cantilevered eave extensions at the same time. They can be easily jacked into place and reinforced. I know how to do that also.

If it were me, I would do all of above plus open up the roof from below to fix the eaves structurally, add insulation, conduit, ventilation space above insulation, ridge vent, and soffit vent, etc. Then use a skim coat of plaster on blueboard when the ceiling goes back in place.
Paul Harding FAIA Restoration Architect for FLW's 1901 E. Arthur Davenport House, 1941 Lloyd Lewis House, 1952 Glore House | www.harding.com | LinkedIn

DRN
Posts: 3944
Joined: Mon Jul 10, 2006 10:02 am
Location: Cherry Hill, NJ

Post by DRN »

Paul, thanks in advance for posting the detail.
Your suggestions are understood and already in motion....I've been planning, analyzing, sketching, dliberating, and detailing all aspects of this for the better part of 3 years. The work will include:

-reinforcement of the roof framing...I enlisted a structural engineer to review my framing schemes and assist in member and connection sizing. Steel flitch plate beams and steel channels will be interwoven to maintain the original roof depth and appearance but to increase load capacity.

-re-roofing all sloped roofs with the originally designed roll roofing, this time with synthetic wood battens and a waterproof membrane underlying all. Working roof venting in was a challenge...ridge was easy as it would appear as another batten, soffits don't want to be violated as they are so visible from the interior and Wright's concept of a continuous uninterrupted plane from interior to fascia needed to be maintained....I'm proposing a top mounted vent to blend with the battens and keep the soffit clear http://www.airvent.com/professional/pro ... ries.shtml

-re-insulation and replacement of the living room and dining area ceiling GWB...always was sand-finished paint on GWB, we'll restore as such.

-removal of existing exterior concrete slab that is dead flat and does not shed roof water adequately. To be replaced with sloped, red tinted concrete of same module with a turn down edge for stability.

-replacement and reframing of the window wall; mullions to be continuous from slab to roof. water and potential for rot issues to be addressed. Original wood to be re-used where possible.

-restoration of window wall French doors and stationary glass. Cracked sheets to be replaced all others reinstalled; rotted doors to be dutchman'd or replaced on case by case basis.

dtc
Posts: 739
Joined: Mon Mar 05, 2007 9:04 am

Post by dtc »

Dan,
I can relate to some of your problems. I believe picture 6 shows us the various patches on your bottom rails. Do we see 1/4 inch plywood, or worse yet, masonite attached to the bottom rails?
Are the bottoms of the stiles cut off as well?

As you know, this whole window wall will need to opened up which will necessitate erecting a plastic or temporary wall a few feet in to protect the interior from the elements.

The discontinuous mullions do not work...I believe you mentioned you will be replacing them with new wood, which the species has not yet been determined and running them from the concrete pad to the roof. I agree with Paul, Cypress should be installed. As for a finish on all the end grain the West System works like a champ. I will post some picks for you to see. For boat builders it is the glue/epoxy of choice, and for those of us who live with these window walls we can attest that they must function like a wooden boat.

This is a lengthy project, but I know you and Christine will solve it to your satisfaction.

DRN
Posts: 3944
Joined: Mon Jul 10, 2006 10:02 am
Location: Cherry Hill, NJ

Post by DRN »

Dan:
The bottom rails of the French doors and stationary windows were patched, cut, and cobbled in various ways prior to our purchase of the house. Of the three sets of operable doors, one was patched with epoxy resin which is falling out, one was not patched at all (it was operable until two summers ago when the bottom rail sagged while open and was nearly impossible to close..it is too soft to move now), the last pair was patched shut with a sheet of luan plywood screwed to the exterior and sealed with silicone. The stationary panel pairs have both been dutchman'd with stacked 2x6's faced with luan and painted.

None of these repairs was done in a manner approaching adequate craftsmanship or historical authenticity. They will all be undone and repaired using details closely matching the original condition per the drawings. Screens and associated frames per the original drawings will be made for the operable pairs.

A question: The drawings are thin on detail for the underside of the bottom rail of the French doors...from what I could tell, there was no metal strip on the under side to protect the endgrain of the stiles or to provide a positive weather or bug barrier. What was in place or not at Dobkins?

SDR
Posts: 19311
Joined: Sat Jun 17, 2006 11:33 pm
Location: San Francisco

Post by SDR »

Paul's mention of polyurethane adhesive as an end-grain protection is sound, in my opinion. Keeping the bottoms of the doors (indeed, all wood) off the concrete, and a sweep-type weatherstrip, might be acceptable -- especially if the sweep was buried in a slot in the door bottom ?

Why cypress, if a better specie is available ? Are the mullions to be painted ? Ipe is great, though very dark. "What Wright did is good enough" doesn't necessarily fly, in my opinion. (I'm still wondering how the roofing/fascia connection was made, without flashing, at Jacobs I -- the only such restoration I'm aware of !)

dtc sends photos of Dobkins; I'll post those next.


Here are DRN's drawings:


Image
Sweeton progress sheet


Image
Proposed mullion base detail (if CMU exists under slab)


Image
Proposed mullion base shoe


Image
Sweeton roof existing framing sketch 2.20.2014



Image
Sweeton roof proposed framing sketch 2.20.2014

KevinW
Posts: 1282
Joined: Sun Feb 06, 2005 6:41 pm

Post by KevinW »

Are the columns actually two pieces or one, notched to accommodate the horizontal? The Beharka house has a similar detail, but the vertical is solid with a notch at the horizontal clerestory.
KevinW

SDR
Posts: 19311
Joined: Sat Jun 17, 2006 11:33 pm
Location: San Francisco

Post by SDR »

Dobkins photos:


Image dtc 49


Image dtc 50


Image dtc 52


Image dct 53

dtc
Posts: 739
Joined: Mon Mar 05, 2007 9:04 am

Post by dtc »

Thanks SDR for posting these pics.

49:
Mullion repair.
I cut out the rot and a couple of inches of sound wood above it.
Glued in new piece/screwed & plugged.

50:
Detail of Dutchman joint.

52:
Bottom rail of mitered window. The joint glued/screwed & plugged.

53:
New stile & bottom rail.

I replaced stops where needed with fresh mahogany & stainless steel brads. All glue "West System".

dtc
Posts: 739
Joined: Mon Mar 05, 2007 9:04 am

Post by dtc »

I tackled this project in baby steps, one set of doors at a time.
The house reached its 50th birthday while I was in the middle of it.
They were in very rough shape when we purchased the house back in 1997.
Only one set of doors were operable.
I suspect the mortice and tenons of the doors failed when the house was 30 to 35 years old. Assembled with inferior glues and nails that started to rust a year or so in. Deferred maintenance always takes its toll.

Post Reply