Ahhhh, the sweet smell of fresh cedar!: The Roof Restoration Begins

Omar and Luis installing the first sections of the new roof. The sheeting above them is used to protect the attic below in the event of rain.

Craftsmen from The Historic Roofing Company have been working diligently since the beginning of the month to replace the well worn wood roofing system on The Hilleary-Magruder House. Installed during the 1980 restoration, and now over 35 years old, the system has held up amazingly well, a testament to quality specifications, materials and installation. With few exceptions the roof has kept bad weather out as intended for even a bit longer than expected.

The roof is composed of what is called a 4-layer wood shingle system. Using 24″ vertical grain western red cedar “Royals,” the roof essentially has a extra layer of protection as opposed to a more typical 3 layer roof of the same material using 18″ perfections. That, in combination with the traditional open roof sheathing has allowed the system to breath and maintain equilibrium far better than a hybrid approach common today in wood roof shingle systems. As this is a restoration project (replace-in-kind) the system will be replaced using the specifications from the 1980 restoration by James T. Wollon, AIA, since it performed so well. During the demolition reconstruction we documented some areas of particular interest and/or concern for evaluation and treatment.

The original spaced sheathing and rafters are white oak. The 1980 restoration  replacements are white pine. Our replacements are poplar. This attic hasn’t seen daylight for 35 years!

The process began with a visual inspection of the roof structure to determine any soft spots in the system which might present a safety concern for the workmen. A few rafters and sheathing areas were flagged as potential areas of concern, mainly where cracks or excessive sagging was observed in the structure and where spotting from rainwater leakage was detected. Close inspection of flashings at walls, dormers and roof penetrations was also performed.

Since the roof is gambreled (it has two pitches on each side with a break in between), the main roof replacement process was broken into four phases: 1&2 tear off the old shingles and replace the top sections of roofing on each side (including flashings around the chimneys), and 3&4 tear off and replace the lower half of each side (including, the 6 dormer roofs, valley, sidewall and sill flashing). A 5th phase of the roofing replacement includes repairing all of the associated trim, and finally the lower porch roofs and trim must be similarly treated.

This work all has to be carefully coordinated with the masons working on the chimneys and the carpenters restoring the dormers so a water tight envelope is maintained during the process and so that the flashing details are correctly completed.

Chimneys are Okay!

After much scrutiny it has been determined by AEON Preservation Consultants (AEON) and Federal Masonry Restoration (FMR) that the chimneys are essentially sound, and only minor repairs are necessary. Our concerns here were that they had been significantly “rocked” during the earthquake in 2011. Several of us were in the basement when we heard, felt and then saw the shock wave travel through the building on that day in August!

Our chimney story begins with photos from the early 20th c the first showing the building in its earliest photographed configuration. In the 1922 image, note that both chimneys are largely intact with corbelled and vaulted tops and appear to be unpainted. The brick on both chimneys is showing some signs of deterioration, as unevenness in the surface is apparent.

 

The next image taken in 1933 shows that the East chimney top has been rebuilt, but without the corbel and the vault is as a result a little smaller as well. By measurement the East chimney is shorter by the height of the corbel, and the quality of the vault is inferior to the original.They both remain unpainted and the deterioration does not appear to have gotten much worse.

The 1970’s images of the building show it in perhaps its worst state. Aside from cheap roll roofing and a lot of roofing tar applied to seal the chimney and dormer flashing areas, the chimneys have now both been stuccoed over. This is most evident as the corbel details on the West chimney have been smoothed out by the thick coating.

 

The “as-built”1980 restoration which we are working from in this project diverged somewhat from that proposed in the architects construction documents. Those drawings show restoration and repair of the existing brick work, but a decision was made to retain and repair the stucco over the brick work. Further, in an unusual detail, the stucco was extended over the chimney flashing as a counter flashing, making it impossible for us to replace the roof without disturbing the stucco at the intersection of the roof and both chimneys. This had already failed as cracks had telegraphed through the stucco at the point where it was no longer attached to the brick and simply floated on top of the flashing attached with a strip of galvanized metal lath.

FMR masons chase cracks in stucco surface on West chimney to reveal structural conditions.

So FMR’s first task was to remove loose stucco and chase cracks that had telegraphed through the stucco, and remove enough material at the base to free up the flashing, and look for signs of catastrophic structural weakness (system discontinuity from the quake and other lateral force events like hurricanes, micro bursts and Durrechos), and systemic weakness (general brick and mortar condition failure) if any.

Much to our surprise (and relief), we found very little of either! Almost all spot checks revealed that other than the changes described above, the masonry was in good, largely original (to 1742 or certainly that era) condition.

 

 

 

 

As with any building this old few areas of concern do have our attention though.

• 

  Aluminum flashing and stucco counter flashing condition at West chimney all appear to be from the 1980 restoration. flashing tuck location appears to be in original location, very low to roofing.

  The chimney step flashing once revealed was found to be consistently very close to the roof level, less that 3″ in some cases. This was contributing to some mortar deterioration below the flashing and perhaps some of the moisture related blistering of plaster in the 2nd floor chambers as water seeped through the system and found it way out.

• The roof along the front and back of the West chimney was flashed to a dramatically sloping roof condition which appears to have been sagging since early in its life. Further investigation of the structural reasons will be included in another post.

• The middle of the brick vault atop the West chimney was sagging significantly and loose brick work was evident, though a very heavy portland cement parge coat remained intact above the brick. From initial visual inspect there was concern it was a sign that the entire vault was in a near state of collapse.
  

  1980 stucco patch on top of West chimney vault (top), and new flue liner and falling brickwork (bottom).

  But looking at the 1980 restoration drawing and chimney brickwork below the roof line in the attic AEON and FMR determined that there was a clay flue liner repair apparently conducted at that time and the area of weakness was relegated to that repair only as they had to remove a portion of the vault to prepare chimney and slip the replacement clay liner into the the chimney mass.

 

Natural hydralic lime product to be used in filling voids and all masonry surface repairs.

So the good news is that the repairs will require less intensive measures that originally though might be required. FMR will finish chasing the cracks resetting the few bricks in the chimneys where loose, and pointing parging with with a natural hydralic lime in several lifts to ensure the integrity of the repairs and seal the surfaces. The surfaces will then be coated with a breathable masonry coating which will be determined once all repairs are complete.

Pulleys but no counter weights?!

Our first task in the window restoration and repair project involves investigating the condition of the windows to determine how they had fared since the 1980 restoration. We knew from visual observation that a lot of paint was peeling and there was evidence of moisture related deterioration on the inside of the sashes.Further, the 1980’s restoration documents did not provide any observational notes which might give us a clue as to their age or construction. While it is presumed that the sashes are not original, their multi-light configuration suggest either a pre or post Victorian era age (earlier as Federal/Italianate rarely used large panels of glass or late as Colonial Revivals often imitated the smaller muntin configurations of earlier architectural styles), and remains unresolved to this posting.

Phase 1 Window investigation.

So sashes from three representative openings were removed (one window on each level on the South or back side of the building) to determine conditions and get a general feel for how repairs would proceed on all sashes.

 

One of the first things noted was that sash pulleys had been installed for single hung windows (only one set per window opening), but there were no counter-weight ropes. Further inspection revealed that there was no dado slot for the rope or a knot hole to tie it off as one might expect. The blocks supporting the pulleys were removed to reveal that the pulleys were constructed of finely machined hardwood with a flared brass bearing shaft through which a brass axle pin was inserted, clearly shop made but field installed.

We drilled small holes in the counterweight boxes to insert a scope and discovered perfectly made rectangular boxes that were…completely…empty. Also, there are no obvious counter weight box doors through which the weights could be installed in the boxes for use.

Bob Schnabel carefully removes caulk and nails to open shutter box

We also carefully pried open one of the interior shutter boxes which had been nailed and caulked into place in the 1980 restoration. Those restoration documents only specified that inner shutter panels not be replaced. Once open we discovered that the shutter had been stripped of all paint, the attachment to the window frame modified, rendering the hinges inoperable, no secondary shutter leaves and no obvious access point for counterweight installation as that counterweight box panel (essentially the window frame), is one continuous piece of wood.

 

solid pulley box at back of  shutter box (no access door for counter-weights).

So what we have discovered so far provides no real “smoking guns.” Our working theory and summary conclusion as to the age of window systems thus far, based on observational and background information are as follows:

 

Documentary Evidence:

• Earliest photographs reveal the sashes (which appear to be the current units) are single hung, have relatively small glass panels and are all very delicate in woodworking proportion suggesting a Federal/Italianate period or Colonial Revival period – so 1785-1850 or post 1890. We are leaning more toward Federal/Italianate as the thinness of the materials and proportion is uncharacteristic of the Colonial Revival period.
• There were significant ownership changes during both of these periods, with more affluent owners apparently (but not conclusively) in the Federal/pre-Civil War periods (bracketed by Henderson and Magruder ownership periods respectively). At 50 to 100 years old, life cycle replacement of building systems and changes in contemporary style might prompt more affluent owners to make modifications which we see here today.

No dado slots for sash cords.

• Xerographic copies of photographic documents from the pre-1980 restoration suggest that the exterior was finished in ruled plaster creating an ashlar stone appearance on some or all of the exterior stone masonry walls. Unfortunately, and to our knowledge, that work was not dated prior to its removal. It seems unlikely that this was an original finish to the 1742 Middle Georgian era Colonial structure. If however it was an early modification – this technique was widely used during the Federal period- it suggests that an intensive improvement campaign may have been engaged then including window systems. This could be during the early construction of the Federal City (Henderson Period) or its rebuilding after the War of 1812, maybe as late as pre-Civil War (Magruder Period) or even later, as building in a “Federal Mode” continues to this day in Washington DC.
• Newspaper reports of the time document several instances of significant window damage in Bladensburg after powder mills exploded mainly in the Federal period, prior to consolidation of munitions manufacturing away from this area.

Physical Evidence

No evidence of hold open hardware (pins or bolts) on upper or lower sash faces.

• The window sash currently installed are not related to the sash pulley/window frame system on the first floor (typical condition for all 7 of the 9/9 windows to be verified).
• The lack of counterweights in the boxes or access panels to service them in window opening examined suggest that they may never have been used.
• Pulley system may be original to the building (counter-weighted systems were introduced to the colonies in the late 17^th), but hardware and the quality of the machining of the wood pulleys may suggest a later installation. The pulley blocks are installed with a hooked notch and a single cut nail. Cut nails are in regular production and use by the1780’s.
• The sash maker/installers do not appear to have employed any counterbalance or hold open system other than props. So far there are no clear indications of hardware shadows or hole patterns which suggest a bolt or pin type locking system (typically involving locking the lower sash to the upper).

Hole for pulley block.

• It appears that most of the window woodwork was stripped of paint in 1980 so only 2-3 coats of modern paint are evident on sashes and no records of paint analysis are in evidence in archival materials.
• The relationship of the shutter boxes to the sash counterweight boxes is unclear due to 1980 restoration work. While it seems clear that the shutter panels are very early, the counterweight boxes appear to be fabricated and finished differently. Additional lumber species and fabrication technique identification is needed here to be more conclusive.

The investigation continues…

Hilleary Magruder House – The Restoration Begins

 

 

Hilleary-Magruder House looking Southwest

Prince George’s Heritage will begin a major maintenance and restoration campaign for the Hilleary-Magruder House beginning on or about March 1, 2016. Acquired and restored by Heritage in 1980, the 1742 gambrel roofed stone structure where George Washington dined with merchant/owner David Henderson, is in need of typical exterior life cycle maintenance – a new wood shingle roof; window, woodwork and masonry repairs, and paint.

The work will be performed by notable area restoration contractors: Historic Roofing Company, Inc., Mozer Works, Inc., and Federal Masonry Restoration, Inc.; and overseen and managed by arnold & arnold architects, and AEON Preservation Services. It has been financed in part by a Historic Property Grant from Prince George’s County, and a State of Maryland bond bill.

Beginning in December of 2015, the Phase 1 investigation involved removing 3 sets of sashes to determine extent and method of restoration of those elements throughout the building.

As the work progresses over the next three months, we will post updates of the process as it unfolds. At the Hilleary-Magruder House, as with any historic  site, each restoration campaign gives rise to the opportunity to review the success of the previous, and compare and apply evolved methodology with the intention of improving upon discovery and restoration technique which will ensure better understanding and the enduring preservation of this cherished National Register historic site.

 

Mike Arnold, RA

Name(required)

Email(required)

Comment(required)

Submit

Δ