Knickerbocker Theatre, Washington D.C. (January 28, 1922)

Stephen Kijak, B.A.E./M.A.E., Penn State, 2011

The roof of the Knickerbocker Theatre collapsed on the evening of January 28th, 1922 after overstressing of the main roof truss caused it to become unseated. The theatre was built five years prior to collapse, often hosting in excess of 1700 guests. The evening that saw the collapse of the Knickerbocker roof was at the tail end of a violent blizzard that paralyzed Washington, D.C. for two days, blanketing the city with 28 inches of snow. Around 400 patrons were attending a silent film that night when, at approximately 9:00 PM, the roof gave way killing 95 people (NYT, Jan. 1922) and injuring another 133 more. The collapse garnered nationwide attention, bringing about criticisms to an insufficient building code and to the irresponsible building practices of the day.

Introduction


Figure 1: 18th Street Entrance, courtesy of the Library of Congress
Figure 1: 18th Street Entrance, courtesy of the Library of Congress
30825v.jpg
Figure 2: Theatre Auditorium, courtesy of the Library of Congress














Figure 3: Theatre Plan, via ENR archives
Figure 3: Theatre Plan, via ENR archives

The Knickerbocker Theatre was the largest and latest in a series of silent film venues in Washington. Commissioned by owner Harry Crandall, the theatre was located in a wealthier area of D.C. at the southwest corner of Columbia Road and 18th Street NW. Due to the oblique angle at which the two roads intersected, the theatre was designed to have a triangular shape in plan, with the West wall following the curvature of Columbia Road to maximize interior space (Washington Kaleidoscope 2009).

Aesthetics were inspired by Neoclassicism perhaps influenced by the surrounding buildings, which architect R.W. Geare chose to emulate in some aspects of the exterior design. Guests would enter along 18th street, seen in Figure 1, and enter into the lobby where a candy store was also available. One could choose to venture upstairs to the balcony or remain on the ground floor, entering the theatre through the doors in front of them, see Figure 3. The interior was designed like many other theatres at the time; the seating area was wedge shaped, tapering as it approached the orchestra, see Figure 2. Unlike modern movie theatres, silent film theatres had to cater to an orchestra, because artificial sound had not been adopted yet. Musicians sat at the base of the proscenium, immediately preceding the stage and screen. Up to 1700 guests could be fit into the theatre; it was by far and away one of the larger cinemas in the area (Mehren: B, Feb. 1922).

Keywords


Truss Failure, Overstressing, Eccentricity, Skewed Bearing, Inadequate Stiffness, Blizzard, Washington D.C.

Design and Construction


Construction began in the latter part of 1916 and finished in the fall of 1917. In total, the theatre cost owner, Harry Crandall, $77,000 to build (E&C 1922); a fairly minimal amount considering the size and prominence of the theatre. Architect R.W. Geare was commissioned to design the building, hiring engineer M.S. Rich to design the roof framing (Mehren: B, Feb. 1922).

As-Built Design

The building stood 40 feet above the sidewalk and was constructed mostly of masonry. Hollow tile was chosen for the walls along Columbia Ave. and 18th street with structural brick only being used in the south wall. The wall of most concern, the one along Columbia Ave., or the NW wall as it will be called from hence forth, was 22” thick and composed 18” of hollow tile and 4 inches of face brick. In order to assimilate with the surrounding architecture the NW wall was riddled with windows, which only served to weaken the wall (E&C 1922). Only four columns existed in the entire building as a result of the exterior load bearing walls. Two columns supported the balcony while the remaining two extended up to the roof structure providing intermediate support (Mehren: B, Feb. 1922).
Figure 4: Critical Areas, via ENR archive
Figure 4: Critical Areas, via ENR archive

The roof consisted of a three-inch concrete slab that rested on a series of trusses and girders. The main truss extended from the NW wall to one of the interior columns. This truss, in turn, supported 4 auxiliary trusses and one beam; it was crucial to the integrity of the entire roof. The remaining layout of the roof structure can be seen in Figure 4. Beams are spaced 9.5 feet apart giving a fairly large span to the concrete roof slab. In total, the entire roof, including concrete, trusses, beams, and roof finishing weighed approximately 75 lbs. per square foot (Contract Record 1922).

Changes to the Original Design

One disconcerting change occurred when the successful bidder of the steel framework suggested that some of the trusses be substituted for plate girders and heavy beams in order to expedite construction as materials were in short supply due to the war. The revisions were quickly approved by the city with neither the steel fabricator nor the architect running any computations to check for strength of the new design. The engineer had not even been notified of the changes; the resulting alterations were substantially independent of his original layout (Mehren: B, Feb. 1922).

During construction, the contractor had noticed that the NW wall seemed rather unstable due to insufficient stiffness, so a reinforced concrete belt course was added to strengthen the wall. At the same time, the contractor had also noticed a lack of lateral bracing for the two main supporting columns. Upon approval of the architect, a steel strut was added between columns C2 and C3 at the level of the bottom chord of the truss (Mehren: B, Feb. 1922).

Causes of Failure


On the evening of January 28th, 1922 the roof of the Knickerbocker Theatre collapsed after the main truss T11, seen in Figure 4, became unseated from the NW wall, overstressing the remaining roof members and causing them to be ripped from th
Figure 5: Aftermath, courtesy of the Library of Congress
Figure 5: Aftermath, courtesy of the Library of Congress
eir seats (Contract Record 1922). Instigated by one of the worst blizzards in the history of Washington, D.C. the roof fell suddenly and without warning (Wash. Post 1922), consuming the audience below in a torrent of steel and snow. No one, up until that point, had any suspicion that the building may have been dangerous or deficient. The theatre had been inspected regularly and had met code requirements in the opinion of the city (Mehren: A, Feb. 1922). So when the roof did collapse an investigation readily commenced revealing a slew of defects affecting nearly every aspect of the roof.

Defects Causing Collapse

Two committees investigated the collapse of the theatre; their findings were nearly identical. The unseating of Truss T11 was attributed primarily to the buckling of the northernmost gusset plate of truss T11 and the failure of the bottom chord where it met the NW wall (Mehren, Mar. 1922). It is important to note that the failures occurred from multiple deficiencies in the truss, the wall, and the connections. Seven design weaknesses were found to be contributors of the collapse; they are as follows:
  • Truss T11 experienced a large deflection caused by overstressing due to: design loads being greatly exceeded by the actual loads by around 30% totaling 75psf dead load, the truss being fabricated nine inches shallower than originally intended, and from moment plates not being spliced (E&C 1922).
  • Bearing stresses at the seat of Truss T11 were concentrated at the edge of the NW wall, causing an eccentricity (E&C 1922).
  • Anchorage of truss T11 to the bearing wall was not present. Only six inches of the bearing channels were extended onto the seat, perpendicular to the wall (E&C 1922).
  • The bottom chord of truss T11 was insufficiently reinforced at the truss seat, so that its channels, which extended independently of one another onto the bearing plate, were subjected to: the full weight of the truss, a twisting force caused by the angled bearing on the wall, and tension from the eccentric bearing on the seat (E&C 1922).
  • A cover plate was omitted in truss T11 at the connection of the top chord and batter post limiting rigidity to a ½ inch gusset plate (E&C 1922).
  • An Inadequate out-of-plane stiffness in the NW wall was caused by slenderness and openings (E&C 1922).
  • Lateral stiffness throughout the entire structure was lacking, especially concerning the area around columns C2 and C3 (E&C 1922).
Figure 6: Truss Bearing, via ENR archives
Figure 6: Truss Bearing, via ENR archives

Each one of these errors contributed to weakening a very small area of the building. The truss deflection created an outward thrust on the NW wall, which was already slender and riddled with openings, causing it to be five inches out of plumb at the time of collapse (Contract Record 1922). This thrust was heightened by the misaligned bearing center, seen in Figure 6 at the bottom left, which fell on the edge of the wall. And since the truss came to the wall at an angle, as seen in the plan view of Figure 7, the bottom chord experienced twisting as the truss deflected, pivoting at the wall edge (Contract Record 1922).
Figure 7: Channel Crushing, via ENR archives
Figure 7: Channel Crushing, via ENR archives

Because the truss was shallower than designed, the top chord was overloaded. Stresses culminated at the junction of the top chord and batter post, where a lone gusset plate was the only source of lateral stability, as seen in Figure 8. Buckling occurred as the stresses overwhelmed the joint. At the same time, two unreinforced channels, which extended from the gusset plate over the seat of the truss, acted as the bearing elements for truss T11 at the NW wall. These channels formed an eccentric bearing on the seat and were entirely independent of one another, as seen on the bottom left of Figure 6. Ultimately the combination of bearing, twisting, and tension at the seat crushed the channels (Contract Record 1922).
Figure 8: Elevation of Truss T11, via ENR archives
Figure 8: Elevation of Truss T11, via ENR archives

Other Potentially Critical Defects

Although the ones listed above are the only alleged contributing factors, other deficiencies existed throughout the building that could have just as easily compromised the stability of the building. Some of the deficiencies listed below, while not provoking the collapse, increased its severity:
  • The NW bearing wall lacked strength and rigidity (Mehren, Mar. 1922).
  • Trusses were shallow and did not meet the required cross sectional area, nor were they detailed correctly (Mehren, Mar. 1922).
  • Columns were improperly braced (Mehren, Mar. 1922).
  • Truss T11 did not have lateral support at intermediate panel points (Mehren, Mar. 1922).
  • Auxiliary trusses lacked diagonal bracing between them (Mehren, Mar. 1922).
  • The concrete roof slab was too thin to meet strength requirements (Mehren, Mar. 1922).

The building from day one was hopelessly inadequate in many areas. The blizzard of 1922 only exacerbated the extent of structural insufficiencies by depositing over two feet of snow, estimated at 12 psf. or 16% of the dead load, onto the roof (Contract Record 1922). Collapse was inevitable at some point, the official report claims, and it is a wonder how so many of its other structural defects did not contribute directly to the collapse.

Other Theories


Several other theories were posited to explain the collapse of the Knickerbocker Theatre. After an extensive investigation by two separate teams, the alternative theories were altogether debunked as the committees' accounts revolved around the same conclusion, namely the one that has already been stated.

Army and Navy Engineers

Upon request by the District Attorney, an investigation was undertaken by engineers of the Army and Navy. Their report attributed failure to the unseating of Beam 21 due to crushing of the bearing tile. Beam 21 was alleged to have fallen, bringing with it Beam 22, and releasing truss T12 from a point of lateral bracing. Truss T12 then buckled and fell from its seat, carrying with it the main truss and roof ensemble. This theory was readily discredited with a lack of evidence from the site, where the unseating would have allegedly happened. Eye witness accounts also corroborate that the collapse initiated from the area around the seat of truss T11 (Mehren, Mar. 1922).

Slab Failure Theory

As mentioned before, the slab was inadequate for the load applied to it. This theory suggests that the slab failed, causing Beams 21 and 22 to fall toward each other and off their seats. The committees' response to this allegation was that a slab failure could not have caused the two beams to fall inwards, toward one another. Furthermore, the slab at the suggested point of failure was, in fact, stronger than the surrounding areas and was therefore unlikely to fail (Mehren, Mar. 1922).

Column Failure Theory

Given that the column supporting the main truss had been short of lateral bracing, failure of column C2 seems plausible. As the theory claims, Purlin 11 yielded under the heavy roof load, causing column C2 to fail and subsequently unseat truss T11 at the south end. However, evidence recovered from the site eliminated the possibility of a failure at column C2, as the column appeared to be undamaged and void of failure (Condron 1922).

Legal Repercussions


After the collapse, nine individuals were held responsible by the coroner's jury. Only five of the nine were actually indicted for manslaughter. Reginald W. Geare, the architect, John Howard Ford, the man responsible for designing and furnishing the steel roof framework, Richard G. Fletcher, the hollow tile wall sub-contractor, Julian R. Downman, the reinforced concrete inspector, and Donald Wallace, the builder's foreman, were ultimately indicted by the grand jury. The four exonerated included the building inspector, the city's computer who had also approved the Knickerbocker drawings, and the steel and general contractors. According to the prosecutor, each party could be held accountable for the others' shoddy work (NYT, Apr. 1922). The project had been constructed out of a combined effort from all parties involved, and each individual was responsible for ensuring the proper completion of the whole. Negligence could be the only explanation for a project so flooded with errors. Despite evidence to suggest otherwise, all defendants were found to be innocent and thus were vindicated completely (Mehren, Apr. 1922).

The official report explores the topic of responsibility. In terms of the city inspectors being held accountable, the report suggests that it was not unto them to check every aspect of the drawings and plans. Inspectors serve as the police for the building industry, but they can only ensure that certain elements of a job are in compliance with code; they do not have the resources nor the time to investigate every piece of a building as it is erected. The investigating committee also found the general contractor to have little responsibility over the collapse. Despite a few minor construction errors, the building was largely built as intended (E&C 1922).

Regarding the architect and steel framework designer, the committee found them to be guilty of negligence. Both parties had ample opportunity to review the designs before being sent to the field. The steel fabricator chose not to design a competent truss, and the architect approved the drawings without considering their potential for catastrophic failure. On part of the owner, the committee found the him to have paid the architect a smaller fee than what was suggested by normal practice. Because of this minimal fee, the architect may not have been compelled to complete all work with as much assiduity as it may have warranted (E&C 1922).

Lessons Learned


Several recommendations were made by the investigating committee. If a fee was charged for building inspection and approval, it would allow the city to expand their resources for checking more critically the work sent to them. Although in no way is it their responsibility to check every detail, the city could possibly examine certain representative portions of a project to validate its integrity. It was also suggested that architect fees be representative of the work expected. As stated before, work may be substandard for a fee that is equally substandard. The committee also found that by standardizing the building code, certain codes fraught with errors could be eliminated and replaced with a more able universal code. Finally it was implored that licensure should be mandated for all architects and engineers who want to practice professionally; this is an easy method to winnow out the incompetent (E&C 1922).

Conclusion


This collapse exemplifies the glaring negligence of the architect and steel designer who did not bother to check their alterations to the engineered design. A hierarchical chain of professionals is established in order to maintain a level of workmanship and safety for the eventual owner and building occupants. When that system of checks and balances is deliberately avoided in order to expedite construction, designs are not checked and errors are given a chance to occur. The errors of the architect and steel fabricator cost 98 (NYT, Apr. 1922) people their lives. And even though neither party was charged for their carelessness, they will forever live with the ignominy of their irresponsible practices.

Bibliography


  • Mehren, E. J. (February 9, 1922). “The Knickerbocker Theater Disaster.” Engineering News-Record Vol. 88 No. 6: 221-222.
    Journal Article: This article summarizes the initial investigation into what caused the collapse of the Knickerbocker Theatre.
  • Mehren, E. J. (February 9, 1922). “Facts of the Knickerbocker Theater Collapse.” Engineering News-Record Vol. 88 No. 6 (1922): 224-229.
    Journal Article: Immediately following the surmised findings of the initial investigation, this article explores the structure of the Knickerbocker Theatre, specifically the roof trusses and their connections, and provides pictures of the aftermath of the collapse.
  • Mehren, E. J. (March 30, 1922). “Two Committees Charge Knickerbocker Collapse to Twisting Failure of Truss End.” Engineering News-Record Vol. 88 No. 18: 532-538.
    Journal Article: After nearly two months of investigation, a formal report was released for the theatre’s collapse and this article presents that report’s findings.
  • Mehren, E. J. (April 6, 1922). “Five Indicted in Knickerbocker Theatre Collapse.” Engineering News-Record Vol. 88 No. 6: 582.
    Journal Article: An article that explains who was indicted for the Knickerbocker Theatre collapse.
  • Condron, Theodore L. (April 1922). “Observations on the Collapse of the Knickerbocker Theatre at Washington, D.C.” Western Society of Engineers Vol. 27 No. 4: 97-112.
    Journal Article: Mr. TL Condron investigated the collapse of the Knickerbocker Theatre himself and presented his findings to a committee in March of 1922; this article is a record of those findings and the contents of that meeting.
  • “Hundreds, Dead or Injured, Buried Under Ruins As Roof of Knickerbocker Theatre Collapses.” The Washington Post, January 29, 1922.
    Newspaper: The morning after the collapse, the Sunday paper of the Washington Post publishes their report on the events of the night before.
  • “Theater Wreck Inquiries Started; Senate Will Act.” The New York Times, January 31, 1922.
    Newspaper: A lengthy article giving quotes from various city officials on their opinions of the collapse, expressing anger at the substandard building codes.
  • “Five Men Are Indicted for the Theatre Collapse.” The New York Times, April 4, 1922.
    Newspaper: Lawsuits were being hurled at alleged culprits, who were said to be colluding with building officials up until the time of the collapse; this article lists the extent of those indictments.
  • Engineering and Contracting (April 26, 1922). "Report of Associated General Contractors on the Knickerbocker Theater Collapse." Vol. 57; 395-398.
    Journal Article: Summary of a committee meeting for the collapse of the Knickerbocker Theatre.
  • Contract Record and Engineering Review (1922). "What Caused the Collapse of the Knickerbocker Theatre?." Vol. 63; 326-335.
    Journal Article: An analysis of the collapse
  • Gershon Fishbein. "A Winter's Tale of Tragedy." The Washington Post, January 22, 2009. <http://www.washingtonpost.com/wp-dyn/content/article/2009/01/21/AR2009012101298.html>
    Newspaper: A remembrance of the 1922 theatre collapse.
  • Library of Congress. "Knickerbocker Theatre" Library of Congress Prints and Photographs Division, October, 1917. <http://www.loc.gov/pictures/item/npc2008011327/> (Nov. 2011)
    Photograph: Pictures of the theatre after construction was complete.
  • Washington Kaleidoscope. "Lost Washington: the Knickerbocker Theatre." Accessed Nov. 2011 <http://dckaleidoscope.wordpress.com/2009/07/07/lost-washington-the-knickerbocker-theater/>
    News Article: A brief summary of the Knickerbocker Theatre history and collapse.