Ehitusinsener The engineer behind the architect

August Komendant

Born 2 October 1906 at Nurmsi in Järvamaa (English sources say Mäo parish); died 14 September 1992 — in Upper Montclair, New Jersey, as the sources repeat, though no document read here records the place

The Estonian engineer who made Louis Kahn’s buildings stand up. His Kimbell vaults span a hundred feet in a shell four inches thick.

Drawn to one scale, one unit an inch: the hundred feet are 1 200 units across the width of the drawing, and the shell is 4 of them. Edge: blue compression, red tension, full tone at the allowable; a tick — tension past its limit.
verdict
It holds. Nearest its limit is the tension in service: 0.26 MPa where 3.64 MPa is allowed.
span ÷ thickness, in inches
1 200 ÷ 4 = 300
span ÷ depth of the section
30.48 m ÷ 2.279 m = 13.4
force
4.64 MN — 35 strands, which carry 4.75 MN
window
3.19–8.83 MN, at 7 days
deflection
10.0 mm down at midspan, in service
cracked stations
0 of 81 in service, 0 of 81 at transfer
crowd
0 people of 75 kg: +0.00 kN/m on the page’s 2 kN/m

The tendon’s drape, 371 mm below the centroid at midspan, is fitted to his +37 psi at the edge — the one number of his the page fits anything to. The 2 kN/m of load and the 75 kg a person are this page’s own round numbers, and the strength at 7 days follows the ACI 209 curve, which is this page’s assumption.

Both readings, from the sources’ own figures: 4 in — 1 200 ÷ 4 = 300, span ÷ depth 13.4; 5 in — 1 200 ÷ 5 = 240, span ÷ depth 15.2.

He left a village near Paide for Dresden, came back with the newest thing in concrete, and spent ten years building the young republic’s stadium, hangars and silos. In 1944 he left again, rebuilt German bridges for the US Army for five years, and started over at forty-three or forty-four — no source gives the month he landed — with a consulting office in New Jersey. Then for eighteen years he made Louis Kahn’s buildings stand up, and quarrelled with him about who should be credited for them.

Wikidata’s item for him (Q762422): place of birth Mäo, place of death Montclair, no burial — one more claim beside the sources’, not a verdict.

01NurmsiNurmsi

01 · 1906 — 1929 · Nurmsi · 58°52′N 25°43′E

A village near Paide, and somebody who said: be an engineer.

Born on 2 October 1906 in the village of Nurmsi, Mäo parish, Järva County. The date is the same in every source; the place wobbles between village and parish, because English summaries name the parish and Estonian ones the village.

Nurmsi is a small farming village a few kilometres from Paide, today part of Paide municipality. Out of it came a boy who went into the Republic of Estonia’s army before any engineering training at all. No source found here names his unit, his rank or the exact years; that he served is established, everything about its shape is not.

There is one name behind the decision to become an engineer. Carl-Dag Lige writes in Sirp that, by Komendant’s own recollection, it was Konstantin Zeren who steered him towards civil engineering; the recollection itself is Komendant’s piece “Konstantin Zeren ehitajana” in the volume “Liivika võõrsil” (Toronto, 1984), which Lige cites. Zeren, in Lige’s account, carried weight in the Estonian building trade of the 1920s and 1930s — a contractor and engineer, and for years the head of the public works department at the Ministry of Roads. After his military service the young man from Nurmsi worked under him as a foreman and later as a site manager. It is the earliest documented employment of his Estonian career.

The order matters and is easily forgotten: he was a workman and a site manager before he was a student. He first saw concrete from the outside rather than from a drawing. By the time he reached Dresden he already had a picture of what actually happens on a site — and that is what later descriptions of him always single out: an engineer who knew what comes out of the formwork.

The name itself has three forms. The Estonian biographical lexicon of art and architecture carries him hyphenated — August-Eduard Komendant; English-language sources write it without the hyphen; in American practice he signed himself August E. Komendant. Anyone searching an archive for him needs all three, and the Estonian genitive Komendandi besides.

  1. 1906

    2 October

    Born in the village of Nurmsi, Mäo parish, Järva County.

    source
  2. 1920s

    The army

    Military service in the Republic of Estonia, before any engineering training. The sources read here give no unit and no years.

    source
  3. 1920s

    Under Zeren

    Foreman and later site manager under Konstantin Zeren.

    source
2 X 1906 · Nurmsi · 58°52′N 25°43′E 14 IX 1992 · 85 31 394 days 1910 1920 1930 1940 1950 1960 1970 1980 1990 1929 · 22|23 1934 · 27|28 1944 · 37|38 1950 · 43|44 1956 · 49|50
  1. 1929 · The technical college at Dresden 1 I 1929 – 1 X 1929: 274 days aged 22 · 2 X 1929 – 31 XII 1929: 91 days aged 23
  2. 1934 · The degree, and the Ministry of Roads 1 I 1934 – 1 X 1934: 274 days aged 27 · 2 X 1934 – 31 XII 1934: 91 days aged 28
  3. 1944 · Out 1 I 1944 – 1 X 1944: 275 days aged 37 · 2 X 1944 – 31 XII 1944: 91 days aged 38
  4. 1950 · Arrival 1 I 1950 – 1 X 1950: 274 days aged 43 · 2 X 1950 – 31 XII 1950: 91 days aged 44
  5. 1956 · The collaboration begins 1 I 1956 – 1 X 1956: 275 days aged 49 · 2 X 1956 – 31 XII 1956: 91 days aged 50
22|23 · 27|28 · 37|38 · 43|44 · 49|50 · 85 computed on this pageEvery move the sources date is dated to the year only, and he was born on 2 X 1906, so every move is two ages: the year is split at his birthday, and a year-only row is counted from 1 January. He died on 14 IX 1992 aged 85, 31 394 days after he was born; that row is dated to the day and gives one age. The village has Wikidata’s point, 58°52′N 25°43′E, and the compass measures from it.

· · The technical college at Dresden · · The degree, and the Ministry of Roads · · Out · · Arrival · · The collaboration begins ·

02DresdenDresden

02 · 1929 — 1934 · The Dresden technical college · 51°02′N 13°44′E

Five years where prestressing was still being tried out.

The University of Pennsylvania’s Architectural Archives, which hold his papers, date the Dresden years to 1929–1934: a five-year course, the normal length of a German Diplom-Ingenieur programme. A news page on the same university’s site has him return to Tallinn in 1936; EhitusEST reads 1936 as the year his teaching at the Tallinn technical institute began, not the year he came home. Two pages of one institution still disagree, and this page prints both.

The subject is on record: Vaba Eesti Sõna reports that at Dresden he studied the statics of reinforced-concrete structures and specialised in the possibilities and theory of reinforced concrete, then the most modern building material. That is structural analysis, not architecture and not materials testing. And there were people there to learn it from.

Kurt Beyer (1881–1952) held the chair of structural analysis and technical mechanics at TH Dresden and was professor there until 1944 — that is, throughout Komendant’s student years. His “Die Statik im Eisenbetonbau” (1927) was so authoritative that it was nicknamed the Beyer Bible; TU Dresden still calls its civil engineering building the Beyer-Bau. However much personal contact there was, a Dresden civil engineering student of 1929–1934 worked out of Beyer’s textbook.

And a second name, less often mentioned now: Willy Gehler (1876–1953), professor at the same school from 1913 to 1945, who directed the building-technology section of the Saxon state materials testing office. TU Dresden’s own research page says his work on prestressed concrete was significant and that the competing systems of Hoyer and Freyssinet were evaluated in Dresden for the first time. The same page adds why Gehler is little known today: his membership of the Nazi party.

Prestressing was then barely two decades old as a practical technique and was taught nowhere as routine. Estonian accounts say Komendant met the technology already during his Dresden studies in the 1930s. From there a straight line runs to the 1952 book. Two things in this chapter stay open: what the school of 1929–1934 should properly be called, which the sources do three different ways, and whether he was a doctor — in American architectural circles he was routinely addressed as “Dr. Komendant”, yet no source read here names a degree or a dissertation.

  1. 1929

    Dresden

    Studies begin at the Saxon technical college; a five-year course, per the archive.

    source
  2. 1930s

    Prestressing

    The Hoyer and Freyssinet systems were evaluated in Dresden for the first time. He met the technique as a student.

    source
  3. 1934

    The degree

    Graduation, and a post at the Estonian Ministry of Roads.

    source
the course, the Penn archive 5 years 1934 · Sirp 1936 · the Penn news page EhitusEST: his teaching begins Kurt Beyer · to 1944; no source gives the start Willy Gehler · 1913–1945 the course lies inside both: 5 + 5 1910 1915 1920 1925 1930 1935 1940 1945 1950
1934 − 1929 = 5 · 1934 / 1936 computed on this pageThe University of Pennsylvania's archive gives the course as 1929–1934: 5 years, both rows dated to the year only. The return is drawn twice because two sources date it twice: 1934 (Carl-Dag Lige in Sirp, the diplom row) and 1936 (the same university's news page; EhitusEST reads 1936 as the start of his teaching). On the same axis, the two chairs as the chapter gives them: Kurt Beyer to 1944 — no source gives the start, so the line starts dashed — and Willy Gehler 1913–1945. The course lies wholly inside both: 5 and 5 years of five.

· design.upenn.edu · 1934 · sirp.ee · 1936 · design.upenn.edu · EhitusEST · Kurt Beyer · Willy Gehler

03The republicVabariik

03 · 1934 — 1944 · The Kadriorg Stadium · 59°26′N 24°47′E

Ten years: a stadium, hangars, a grain elevator and a water tower.

Postimees reports that before he fled Estonia in the autumn of 1944 he took part in designing about forty reinforced-concrete buildings. Against the roughly two hundred projects of his whole life that is a fifth — done in one decade, in one small country.

The best known is the concrete grandstand of the Kadriorg Stadium in Tallinn, architect Elmar Lohk: designed 1936, opened 1938, replacing a wooden stand. The Museum of Estonian Architecture calls Komendant the principal engineer-constructor — he made the technical drawings and the structural calculations and supervised the work on site; Maris Suits and Carl-Dag Lige, writing in Sirp on the engineer Heinrich Laul, report that current research names Komendant and the engineer Tarmo Randvee together as the main designers, with Laul making only the canopy’s deformation calculations. The roof cantilevers 12.8 metres out over a stand 51 metres wide, with no columns in front. Estonian sources put it among the largest cantilevered concrete roofs in the world at the time, and Kenneth Frampton has singled it out.

In Tartu there are two things that matter more to this page than the stadium. The northern aircraft hangar at Raadi was built in 1935–1936 with Komendant as its engineer, and Tartu City Museum says it carries a so-called prestressed-concrete solution that he developed specifically for it. Carl-Dag Lige, in Sirp, goes further: the earliest known structure in Estonia using steel prestressing in its concrete is the Tartu aircraft hangar, probably designed in its entirety by Komendant. If that holds, the technique he made his American name with was already being tried on an Estonian military airfield.

The rest is infrastructure and industry rather than prestige architecture — exactly what a young Dresden-trained engineer does in a small state. The Tartu grain elevator was designed in 1939–1940 with the architect Ernst Kesa and built in 1941. For the water tower on Õpetaja Street the working drawings and structural calculations were made by the engineer Hans Tari, but the calculations for the tank itself were Komendant’s: a two-part reinforced-concrete reservoir 7 metres deep, holding 500 cubic metres, its floor 23.5 metres above the ground. Construction began on 3 April 1939 and the tower opened on 26 November 1941.

And he taught. In 1937–1938 he read reinforced-concrete courses at the technical university in Tallinn and passed new calculation methods on to future engineers; the University of Pennsylvania writes that with his lectures of 1937–1939 he laid the foundations of a school of concrete shells attached to what is now Tallinn University of Technology. In 1939, on the basis of those lectures, he began compiling an Estonian-language handbook of reinforced concrete; war and emigration left it unfinished. He was also a candidate for the chair of timber, masonry and reinforced-concrete construction — the reviewers Oskar Martin and Vladimir Paavel assessed him positively, but the professorship was never filled.

  1. 1935 — 1936

    The Raadi hangar

    A prestressed-concrete solution developed for this building — the earliest known in Estonia.

    source
  2. 1938

    12.8 metres

    The Kadriorg grandstand’s cantilever, with no front columns, over a stand 51 metres wide.

    source
  3. 1939

    The handbook

    An Estonian handbook of reinforced concrete from his own lectures. The war left it unfinished.

    source
The Kadriorg Stadium after its new reinforced-concrete grandstand was completed, 1937–1940; architect of the grandstand Elmar Lohk
The Kadriorg Stadium and its new grandstand, Tallinn, 1937–1940 photo: Hans Voolmann, Estonian History Museum (Eesti Ajaloomuuseum SA), AM F 19157:230, CC0 (as marked on Wikimedia Commons)
in metres, at one scale Kadriorg · 12.8 m cantilever · root moment w·a²/2 2a = 25.6 m · the same moment at midspan: w·(2a)²/8 Kimbell · 30.48 m 25.6 ÷ 30.48 = 0.840 the stand beneath the roof, 51 m wide, at the same scale in meganewtons, at one scale 4.903 MN · the tank’s water 4.644 MN · his six Kimbell cables 500 m³ × 1 000 kg/m³ × 9.80665 m/s² = 4.903 MN · ÷ 4.644 = 1.056 projects, one cell a project about 40 ÷ over 200 ≤ 0.20 · “a fifth”
0.840 · 1.056 · ≤ 0.20 computed on this pageA cantilever's root moment, w·a²/2, is the midspan moment of a simple span, w·L²/8, when L = 2a: the Kadriorg roof's 12.8 m cantilever carries a uniform load like a 25.6 m span, which is 0.840 of the Kimbell's hundred feet. No source gives the roof's section or thickness, so no stress is claimed here — only the lengths, and that the same load makes the same moment. The Õpetaja Street tank's 500 m³ of water, at 1 000 kg/m³ (the page's own round figure) and g = 9.80665 m/s², weigh 4.903 MN, 1.056 times the force of his Kimbell cables; its floor stands 23.5 m up and it is 7 m deep. About 40 buildings of 200 is a fifth only if the whole work is exactly 200; the source says “over 200”, so a fifth is an upper bound.

Wikipedia: Kadriorg Stadium · Museum of Estonian Architecture: the Kadriorg Stadium (archived) · Tartu Postimees: the water tower · Mikhail Osanov’s thesis · Tartu Postimees: Komendant in Tartu · Museum of Estonian Architecture: “Betoonist võlutud”

04BridgesSillad

04 · 1944 — 1950 · Dresden · 51°03′N 13°44′E

Five years of broken bridges, and then starting over at forty-three or forty-four.

How he got to Germany is not settled. English Wikipedia says that after Germany overran Estonia he was shipped back to Germany to work on war projects; the Museum of Estonian Architecture’s exhibition text says he emigrated to Germany at the beginning of the war; Carl-Dag Lige in Sirp and Estonian World say he fled in 1944 with his wife and two small children ahead of the returning Soviets. This page prints all three.

What followed is a little clearer. Karin Paulus writes in Sirp that after the flight of 1944 he made his way back to Dresden and began teaching there — in the city where he had taken his degree. Dresden was firebombed in February 1945, so the teaching period must have been short. Paulus names neither institution nor years.

Then five years for the US Army: rebuilding Germany’s broken bridges and the infrastructure destroyed in the fighting. The English Wikipedia version adds internment and General Patton to this; for that addition this research found no archival or institutional source, and the page leaves it out as something it cannot confirm. The documented core is reconstruction engineering rather than damage assessment.

Lige argues in Sirp that it was precisely the study of broken structures that gave him an unusually broad practical knowledge of types of concrete construction and their modes of failure — and pushed him to publish and to look for new openings. An engineer who had seen hundreds of buildings broken knew something nobody at a drawing board knows.

In 1950 the family reached New York and he opened a consulting practice in Upper Montclair, New Jersey. His first American position was therefore his own office rather than a salaried post in an established firm — unusual for a displaced person. Two years later, in 1952, came “Prestressed Concrete Structures”, which the Penn archives say grew directly out of the German bridge work. It was later translated into Chinese and Russian. The sources repeat that he died there in 1992; no obituary or record read here names the place of death.

  1. 1944

    Out

    Three incompatible versions of how he got to Germany. The page prints all of them.

    source
  2. 1945 — 1950

    Bridges

    Five years for the US Army: rebuilding broken bridges and infrastructure.

    source
  3. 1952

    The book

    “Prestressed Concrete Structures” — two years after arrival, straight out of the bridge work.

    source
Wikipedia: shipped to Germany for war projects · no year given The museum: emigrated at the beginning of the war · no year given Carl-Dag Lige in Sirp, Estonian World: fled in 1944, in the autumn by Lige 13 II 1945 · Dresden is firebombed 1944: the year he left Estonia; darker: the autumn, from 1 IX 1944 at most 409 days from 1 I 1944 at most 165 days from the autumn, 1 IX 1944 the US Army, 1945–1950 · 1 row 1950 · New York 1944 1945 1946 1947 1948 1949 1950 1951 the three versions stand apart: the page does not merge them
≤ 409 · ≤ 165 days computed on this pageHow he reached Germany the sources tell three ways, and the drawing keeps them apart: Wikipedia and the architecture museum give no year, Carl-Dag Lige in Sirp and Estonian World give 1944. Karin Paulus, in Sirp, has him go back to Dresden to teach after that; the city was bombed on 13 II 1945. From the start of the year-only row, 1 I 1944, to the bombing is 409 days — an upper bound. Lige says he fled in the autumn; this page reads the autumn from 1 IX 1944, and from there to the bombing is at most 165 days. Of the army years, 1945–1950, the timeline holds 1 row.

Wikipedia · the museum · Sirp · Estonian World · · 1945–1950 ·

05The benchPingutuspink

05 · The calculation · The Kimbell Art Museum, Fort Worth · 32°45′N 97°22′W

Too little and the bottom opens. Too much and the top does.

Concrete is strong in compression and weak in tension, and a loaded beam puts its bottom in tension. Prestressing is the trick of squeezing the member hard enough, and low enough down, that the load never gets it into tension at all. Komendant’s working life was spent finding the force and the tendon profile that do that — and doing it for shapes an architect had drawn for other reasons. Here is that calculation. Pull on the force and watch which end of it fails.

Four inches and 23 ft: Wendy Lesser, the Museum of Estonian Architecture, the Kimbell’s own page. Five inches and 20 ft: Wikipedia. The slab is the four-inch reading’s concrete laid flat.

−25 −20 −15 −10 −5 0 +5 0 5 10 15 20 25 30 m MPa tension compression +1.23+3.64−14.56−15.53 Komendant −6.23 +0.26 fitted top bottom

It holds. Nearest its limit is the tension in service: 0.26 MPa where 3.64 MPa is allowed.

Solid: in service, with the load and the long-term loss. Faint and dashed: at transfer, the day the jacks are released, with nothing on the member but itself. Red dashed: the tension allowed, +1.23 MPa at transfer (0.25√f′ci) and +3.64 MPa in service (0.62√f′c). Blue dashed: the compression allowed, −14.56 MPa at transfer (0.60 f′ci) and −15.53 MPa in service (0.45 f′c). Dotted at midspan: Komendant’s published figures.
1 m span 30.48 m (100 ft)
Deflection at midspan: 10.0 mm down in service, 6.4 mm down at transfer, 7.5 mm down before the long-term loss. Dashed: the deflected centroid, drawn 100× deeper than it is. The member itself is drawn to one scale along the span and through the depth; the bar is one metre.
Area 0.869 m², second moment 0.380 m⁴ — both integrated from the cycloid’s own equation rather than looked up. Dashed: the centroid, 0.824 m below the top fibre. Shaded: the kern, 0.300 m above the centroid and 0.531 m below. The dots are the tendon 371 mm below the centroid, where it meets the wall: inside the concrete.
At midspan
At midspantop fibre, MPabottom fibre, MPadeflection, mm (down +)
At transfer, day 7 (f′ci 24.3 MPa) −6.69 −2.35 +6.4
In service −6.96 +0.26 +10.0

Losses to midspan: friction and wobble 4.20 % (4.449 of 4.644 MN left); elastic shortening off; long-term 15.0 %, a rule of thumb. In service 81.4 % of the jacking force reaches midspan.

Creep, shrinkage and the steel’s relaxation as one number; a real calculation needs the humidity and the age, which no source gives.

The Kimbell’s page: one week; the curve is ACI 209’s, the cement unknown.

The drape was fitted with it off.

Tendons stressed one after another

The sources disagree: the museum says three, Osanov six. The count matters only to elastic shortening.

window 3.19–8.83 MN 0 2 4 6 8 10 12 MN
Window at 7 days: 3.19–8.83 MN, 24 to 65 strands.

This page’s arithmetic at midspan: top −6.96 MPa, bottom 0.26 MPa, deflection 10.0 mm downward. Komendant published −903 psi (−6.23 MPa), +37 psi (+0.26 MPa) and −0.3814 in (9.69 mm), which this page reads as downward: his figure grows from −0.152 in to −0.3814 in when the prestress is lost, and losing prestress can only let the shell sag.

Six cables at 174 kips: Osanov’s reconstruction. The strand count is in this page’s half-inch strand.

When the member changes, the slider reaches only as deep as there is concrete. +37 psi

0.35 is the page’s own figure; no source gives the tendon’s profile.

Everything the shell carries besides itself: the light fitting, the travertine, the finishes. No source gives a figure.

The section’s area and second moment are integrated here from the cycloid’s own equation, x = r(θ − sin θ), y = r(1 − cos θ) — the vault’s real shape, not a table value. The material constants are ACI 318’s in their SI forms: Ec = 4 700√f′c, modulus of rupture 0.62√f′c, allowable tension at transfer 0.25√f′ci, allowable compression 0.60 f′ci at transfer and 0.45 f′c in service. The strength on the day of the jacks, f′ci, follows ACI 209’s curve for moist-cured Type I cement — this page’s assumption, since no source names the Kimbell’s cement. The long-term loss is a single-number estimate rather than a calculation, and the page says so where it uses it.

Without JavaScript the bench shows its starting state: Komendant’s force, the fitted drape, one week of curing.

05 · His numbers

Seven numbers Komendant computed for the vault, read back

Mikhail Osanov’s 2014 thesis prints the figures Komendant computed for the Kimbell vault. Here they stand beside what this page’s arithmetic gives for the same member — live from the bench above, at whatever force, load or loss it is set to. Each row carries a ratio and a word for it: fitted where the page tuned itself to his number; within 15 % where the arithmetic checks; misses where it does not. Nothing is averaged.

His seven numbers

Komendant’s figures as Osanov gives them, beside this page’s arithmetic
Figure His In SI This page Page ÷ his Status
Jacking force 6 × 174 kip 4.644 MN 4.644 MN the force the bench is set to 1.00 within 15 %
Stress at the crown (top fibre), midspan −903 psi −6.23 MPa −6.96 MPa service, top fibre 1.12 within 15 %
Stress at the edge (bottom fibre), midspan +37 psi +0.26 MPa +0.26 MPa service, bottom fibre 1.02 fitted to it
Deflection before the losses −0.152 in 3.86 mm down 7.54 mm down service, without the long-term loss 1.95 misses
Deflection after the losses −0.3814 in 9.69 mm down 10.02 mm down service, with every loss 1.03 within 15 %
Design moment 4 700 kip·ft 6 372 kN·m 2 607 kN·m self-weight + load, at midspan 0.41 misses
Support reaction 90.5 kip 402.6 kN 342.1 kN per end · 171.1 kN per corner per end · per corner 0.85 · 0.42 misses · misses

Source: Mikhail Osanov, Bucknell University honors thesis (2014); the SI, the ratios and the statuses are this page’s arithmetic.

His moment implies a load of 8M/L² = 8 × 6 372 kN·m ÷ (30.48 m)² = 54.87 kN/m; less the section’s own weight, 20.45 kN/m, that leaves 34.42 kN/m on top of it. His reaction implies another: 2R/L = 26.41 kN/m if 90.5 kip is the reaction at each end of the span, or 4R/L = 52.83 kN/m if it is at each of the shell’s four corners (25.40 kN/m per end over Osanov’s 104 ft, ring beams included). The source says neither; the research reads the moment as “about the longitudinal axis”. Moment over reaction: 2.08 per end, 1.04 per corner — printed, not averaged.

The losses, walked along the tendon

Half the span, anchorage to midspan, at one scale both ways. The band’s width is proportional to the force in the tendon: dashed — at the jack; light — after friction; dark — in service, after every loss.

15.24 m from the anchorage · α = 0.0316 rad · P = 4.449 MN · 4.20 % lost · 3.782 MN in service

From the jack to service, at midspan

  1. At the jack 4.644 MN P₀ — his six cables at 174 kip (Osanov)
  2. After friction and wobble, at midspan 4.449 MN −4.20 % P₀·exp(−(μα + kx)), with α = 0.0316 rad the angle the tendon turns through by midspan: the codes’ relation (ACI 318). μ = 0.25 and k = 0.0007 per foot are this page’s generic values — no source gives the Kimbell’s.
  3. After elastic shortening 4.449 MN off: the drape was fitted without it. With six tendons (Osanov) it would be 1.02 %, with three (the museum) 0.81 %. (n − 1)/2n · Es/Ec · f_cgp, with the self-weight moment (ACI 318)
  4. After the long-term loss 3.782 MN −15.0 % Creep, shrinkage and relaxation as one lump sum of 15.0 % — this page’s rule of thumb. No source gives humidity, age at stressing or volume-to-surface; the figure is the page’s.

Left at midspan in service: 3.782 MN, 81.43 % of the jack — the same force the service run carries at its middle station.

The strands, counted

4.644 MN ÷ (1 860 MPa × 0.74 × 98.7 mm²) = 34.18 → 35 strands, rounded up because 34 give only 4.62 MN

This page

Grade 270 seven-wire strand, 98.7 mm², jacked to 0.74 × 1 860 MPa = 1 376 MPa, so 135.9 kN a strand; 35 strands carry 4.755 MN. The unit is the page’s: no source gives the Kimbell’s strand.

the page’s own unit

His six cables

0.98 in² = 632.3 mm² a cable, 3 793.5 mm² in all; 174 kip = 774.0 kN a cable, so 1 224 MPa in the steel — 0.658 of a 1 860 MPa strand’s strength, or 0.710 of a 250 ksi wire’s. No source gives his grade.

Mikhail Osanov, Bucknell University honors thesis (2014)

The museum’s three cables

For the same 1 044 kip, 348 kip = 1 548 kN a cable. The museum gives the count, not the force; the division is this page’s.

Museum of Estonian Architecture

06KahnKahn

06 · 1956 — 1974 · The Richards laboratories, Philadelphia · 39°57′N 75°12′W

Eighteen years, and the same quarrel every time.

The collaboration with Louis Kahn ran from 1956 until Kahn’s death in March 1974 — the eighteen years that give Komendant’s 1975 book its title. Counted from the Richards laboratories, for which Kahn got the commission in 1957, it comes to seventeen. English Wikipedia says they met in 1956 and names no occasion; the University of Pennsylvania’s School of Design says Kahn asked him that year to consult on a project for a competition in Chicago.

The first job was the Alfred Newton Richards Medical Research Laboratories at the University of Pennsylvania. Komendant designed the complex inner structure and introduced the precast, prestressed concrete elements that made the building’s spatial organisation possible: 1 019 precast members, post-tensioned in three directions, and in the finished frame the largest offset between any two elements was one sixteenth of an inch. It was the first multi-storey precast, prestressed, post-tensioned rigid frame in the United States.

Whose engineer Richards was is nevertheless disputed. Popular accounts say flatly that the building was “engineered by August Komendant”. The Philadelphia firm Keast & Hood states on its own project page that it was the structural engineer of record for the original construction, engineering the building “in association with Dr. August Komendant”, with Komendant in the consultant role. Three different completion years — 1960, 1961, 1962 — and two different definitions of “the engineer” sit side by side in the literature.

At the Salk Institute in La Jolla the wording is different. Both SAH Archipedia and Docomomo US say Komendant suggested the nine-foot-deep Vierendeel truss to span the 65-foot-wide laboratories — about 2.7 m deep over about 19.8 m. The truss is deep enough to hold a whole interstitial service floor, and that is the defining feature of the building’s section. It is the most generous attribution the sources give him anywhere.

And then the Salk redesign, which says everything about the division of labour. After approvals had been given and contractor meetings had begun, Kahn and the institute decided to cut the laboratory buildings from four to two and add floors to each. Komendant re-engineered the structure and produced new drawings at a speed that Thomas Leslie, in his book “Louis I. Kahn: Building Art, Building Science” (2005), called “legendary”, as English Wikipedia quotes him — and trained the construction workers in the techniques the refined exposed concrete finish demanded. Kahn and the client changed the programme; Komendant absorbed the consequences.

Some sentences of this chapter follow the wording of the English Wikipedia articles linked above, condensed and reworded. Wikipedia’s text is used under CC BY-SA 4.0.

  1. 1957 — 1961

    Richards

    1 019 precast members, post-tensioned in three directions; the largest offset 1/16 inch.

    source
  2. 1959 — 1965

    Salk

    Nine-foot Vierendeel trusses that make a whole service floor.

    source
  3. 1966 — 1970

    Olivetti

    The Harrisburg typewriter factory roof: 72 prestressed units in an eight-by-nine grid.

    source
The Richards and Goddard laboratories on the University of Pennsylvania campus, 2010
The Richards and Goddard laboratories, Philadelphia, 2010 photo: Smallbones, public domain
The Salk Institute from its courtyard, La Jolla, 2011
The Salk Institute’s courtyard, La Jolla, 2011 photo: Carol M. Highsmith, Library of Congress, Prints and Photographs Division, highsm.13139, public domain
1956 6 634 days · 18.16 years · 21.1 % of his life · 7 rows between 18 · the book’s title 1957 6 268 days · 17.16 years · 20.0 % of his life · 6 rows between 18 · not reached March 1974 · Kahn dies · counted from 1 III 1974 1955 1960 1965 1970 1975
18.16 / 17.16 computed on this pageTwo bands to Kahn's death, which is dated to the month and counted from 1 III 1974. From 1956, as the book's title counts, is 6 634 days, 18.16 years (365.2425 days a year), 18 whole years; from the Richards commission of 1957 it is 6 268 days, 17.16 years, 17 whole. The book's 18 is a tick only the first band reaches. Of his life (31 394 days) they are 21.1 and 20.0 per cent; between their ends the timeline holds 7 and 6 rows. A year-only row is counted from 1 January.

· The collaboration begins · · Kahn gets the Richards commission · · Kahn dies · “18 Years with Architect Louis I. Kahn”

07The KimbellKimbell

07 · 1966 — 1972 · The Kimbell Art Museum, Fort Worth · 32°45′N 97°22′W

A hundred feet, four inches, and a slot cut down the middle.

Kahn habitually called the Kimbell’s gallery roofs vaults, but by Komendant’s explanation each of them is a shell doing the work of a beam: a post-tensioned curved concrete beam spanning 100 feet between its four corner columns. That hundred feet is the most consistently repeated structural number for the whole building.

Three people stand behind the shape, and that is exactly why authorship at the Kimbell is contested. Kahn’s project architect Marshall Meyers pointed out that a cycloid curve would lower the ceiling and bring other advantages; the exact shape of the vault, the museum’s own page says, Kahn settled in collaboration with Komendant. A cycloid is the curve that a point on the rim of a wheel draws as the wheel rolls along a straight line. Its sides rise gently, and the museum itself gives that as the reason the vault feels monumental and yet does not weigh on the visitor.

Komendant came onto the project because nobody else could do it. The Fort Worth firm of Preston Geren, brought in to keep costs down, argued that the cycloid vaults would be too expensive and pressed for a flat roof. The dispute was settled by a division of responsibility: Geren took the foundations and basement, Komendant the upper floors and the cycloid shells. It is an unusually explicit documented example of how structural authority was carved up on a Kahn job.

The structural problem is that the shell is slit open along its crown. The light slot means the vault cannot work the way an ordinary barrel vault works; the continuity across the crown is restored by concrete struts at ten-foot intervals. Each shell ends in a thickened arch at both ends, each vault stands on four corner columns, and the cables were tightened with hydraulic jacks a week after the pour. The bench on this page is that calculation, in a cruder coat.

And after all that he did not come to the opening. Komendant deliberately stayed away from the Kimbell’s opening ceremonies — at a building he had contributed very substantially to. The reason given is the combination of Kahn’s reluctance to share credit and Komendant’s own sense of his due. It is the single clearest documented rupture in the eighteen years. Fifty years on, the Kimbell runs a public programme called “Komendant and Kahn: Engineer and Architect”.

Some sentences of this chapter follow the wording of the English Wikipedia articles linked above, condensed and reworded. Wikipedia’s text is used under CC BY-SA 4.0.

  1. 1966 — 1972

    The cycloid

    Meyers proposed the curve, Kahn fixed the shape with Komendant. Three people, one curve.

    source
  2. 1972

    One week

    The cables were tightened with hydraulic jacks a week after the pour.

    source
  3. 1972

    Did not come

    Komendant deliberately stayed away from the opening ceremonies.

    source
A Kimbell Art Museum vault and its light reflector, 2009
A Kimbell vault and its light reflector, Fort Worth, 2009 photo: grabadonut, “kahn kimbell” (Flickr), CC BY-SA 2.0, thumbnail toned

The one sentence about this form that Wikipedia reports as his explanation is quoted below ↓ Wikipedia, reporting his explanation · the source holds it word for word

Salk truss, 62 ft ÷ 9 ft · 6.89 Salk truss, 65 ft ÷ 9 ft · 7.22 the Kimbell vault by its depth (4 in) · 13.37 the Kimbell vault by its depth (5 in) · 15.23 Olivetti shell, 60 ft ÷ 6 in · 120 the Kimbell by its thickness (5 in) · 240 the Kimbell by its thickness (4 in), and the same concrete laid flat · 300 5 10 20 50 100 200 500 span ÷ depth, on a logarithmic axis
the same concrete laid flat
at his force, the tendon on its centroid (e = 0): 171.7 MPa past its limit (tension, the service case) and a computed sag of 12.47 m — a number that says only that the plate does not carry
the vault against the plate
I × 620 · A × 1.22 · the same concrete, the same width; the shape makes the difference
Richards
Richards: no truss depth is sourced, so it is listed, not drawn
6.89 → 300 computed on this pageOne axis: span over depth. The Salk truss is 9 ft deep and the sources give its span as 62 or 65 ft — both are drawn. The Kimbell's depth is that of the cycloid section this page computes, 2.279 m (the 4 in reading), and by its thickness the same span is 300 or 240. The Olivetti shell is 6 in thick and 60 ft across. The same concrete laid flat — a plate of 0.712 m², the tendon on its centroid — is computed on this page's bench at the same force.

SAH Archipedia: Salk Institute · Wendy Lesser, “Walking through the Kimbell” · Wikipedia: Kimbell Art Museum · Wikipedia: Olivetti-Underwood Factory · Wikipedia: Richards Medical Research Laboratories · Mikhail Osanov’s thesis

What is held of his words

Komendant explained that it was actually a shell playing the role of a beam.

Wikipedia, reporting his explanation Reported speech, not his own sentence · the source holds it word for word

On Wikipedia’s Kimbell Art Museum page: Kahn generally called the museum’s roof form a vault, but Komendant explained that it was actually a shell playing the role of a beam. This is Wikipedia reporting his explanation, not his own sentence. Wikipedia’s page prints the sentence word for word. en.wikipedia.org

08CreditAu

08 · 1975 — today · Tallinn · 59°26′N 24°45′E

Whose building is it when one drew it and the other worked it out?

A year after Kahn’s death Komendant published “18 Years with Architect Louis I. Kahn”. Maris Mändel, reviewing the Estonian edition in Sirp, writes that Komendant neither looks up at Kahn adoringly nor paints a glossy picture of him: the portrait is friendly and honest, of the kind only somebody who feels himself the sitter’s equal can give. That claim of equality is the heart of the matter.

The daily dynamic is on record in English Wikipedia’s blunt words: Kahn “would pointedly remind Komendant that he was an engineer, not an architect”, while “Komendant in turn delighted in needling Kahn about his lack of engineering knowledge”. The dispute is about a boundary: who is entitled to decide form, and who merely makes it stand up.

On credit the sources point two ways at once, and the page does not choose. Komendant, in the words of Kahn’s biographer Carter Wiseman, which Wikipedia quotes, “was not above expanding on his share of the credit for shared projects”, and Kahn, in Wikipedia’s own words, “found it difficult to share credit with anyone”. These are presented as two halves of one problem rather than as one man’s grievance. The Spanish critic Luis Fernández-Galiano places Komendant in a line with Anne Tyng and Robert Venturi as people whose debts Kahn failed to acknowledge properly — that is a critic’s judgement, not an archival finding.

One superlative this page does not repeat: that he brought prestressed concrete to America. The Minnesota Department of Transportation’s historic bridges survey states that the first major prestressed-concrete bridge in the United States is considered to be Philadelphia’s Walnut Lane Bridge, completed in late 1950 and designed by the Belgian engineer Gustave Magnel with his student Charles Zollman. Komendant only reached America in 1950. Every source read calls him a pioneer; none calls him the first.

The return to Estonia came after his death. In late 2019 his central book appeared in Estonian for the first time, in Epp Aareleid’s translation; in January 2020 the Museum of Estonian Architecture opened the exhibition “Miracles in Concrete”, to which the University of Pennsylvania lent over eighty objects and to which his estate came back from America; in 2022 Birkhäuser published the monograph edited by Carl-Dag Lige. In Sirp, Maris Mändel’s review of the Estonian edition ran under the heading “18 aastat koos Louis Kahniga – paarsada lehekülge ehituspõnevikku” — eighteen years with Louis Kahn, a couple of hundred pages of building thriller; of the exhibition “Ehituskultuuri õppetund: August Komendandi valitud tööd 1934–1988” at Tallinn University of Technology, Karin Paulus wrote in the same weekly in April 2026 under the heading “Komendandist. Mitte küll köitvalt, ent tummiselt” — on Komendant, not captivatingly but substantially.

Some sentences of this chapter follow the wording of the English Wikipedia articles linked above, condensed and reworded. Wikipedia’s text is used under CC BY-SA 4.0.

  1. 1975

    The book

    “18 Years with Architect Louis I. Kahn”, a year after Kahn’s death.

    source
  2. 2019

    In Estonian

    Translated by Epp Aareleid. It had been translated into Japanese, Spanish and French long before Estonian.

    source
  3. 2020

    The exhibition

    “Miracles in Concrete” at the Museum of Estonian Architecture; the Cultural Endowment’s annual prize.

    source
the book to the Estonian edition: 16 422 days March 1974 · Kahn dies 1975 · the book 14 September 1992 · his death 2016 · the research begins 18 December 2019 · in Estonian 10 January 2020 · the exhibition 2022 · the monograph 9 September 2023 · the Kimbell symposium 1980 1990 2000 2010 2020 his death to the research: 8 509 days Kahn’s death to the book: 276–670 days, between a row dated to the month and one dated to the year
16 422 · 8 509 · 276–670 computed on this pageThe book appeared in 1975 — the row is dated to the year and counted from 1 January — and in Estonian on 18 XII 2019: 16 422 days. From his death on 14 IX 1992 to the research of 2016 (counted from 1 January) is 8 509 days. Kahn died in March 1974 and the book appeared in 1975; neither row gives a day, so the gap is a window: at least 276, at most 670 days — “a year after”, as the chapter says, in calendar years.

· · · · · · ·

09People and placesInimesed ja kohad

09 · The map

Who proposed the shape, and who made it stand.

The line follows the years: from Nurmsi to Dresden in 1929, Tallinn in 1934, Dresden again in 1944 and Upper Montclair in 1950. From there thin spokes run to the buildings in Philadelphia, La Jolla and Fort Worth — the office’s reach, not moves. Every point is Wikidata’s, with its item and its revision, and Nurmsi, the village he was born in, is a dot of its own now.

Nurmsi 1906 Dresden 1929 · 1944 Tallinn 1934 Upper Montclair 1950 Philadelphia 1957 Fort Worth 1966 La Jolla 1959
Land: Natural Earth, public domain. The line is drawn as great circles rather than straight segments on the sheet — the crossing he made is a great circle, and a straight line on a flat map would be a different journey. Dresden is on the line twice: to study in 1929 and back again in 1944. The thin ones are the office’s reach. Every point is Wikidata’s, read on 24 September 2026; each place’s card links its item.
  • Louis I. Kahn Architect; eighteen years · 1901–1974 Born in 1901, when Estonia lay in the Russian Empire, on Saaremaa or in Pärnu — Karin Paulus and Olavi Pesti’s article in Eesti Ekspress (2006) leaves both open. By English Wikipedia’s account Kahn would pointedly remind Komendant that he was an engineer, not an architect, and Komendant needled him back about his lack of engineering knowledge. In the same Eesti Ekspress article the architect Toivo Tammik says the two of them made up the most powerful architect-engineer combination in twentieth-century world architecture. source · arhliit.ee
  • Konstantin Zeren Contractor and engineer By Carl-Dag Lige’s account, for years the head of the public works department at the Ministry of Roads. Komendant himself recalled that it was Zeren who steered him towards civil engineering — and it was under Zeren that he had his first jobs, as a foreman and later as a site manager. source
  • Kurt Beyer Professor of structural analysis, TH Dresden · 1881–1952 Professor until 1944, which covers the whole of Komendant’s student years. His “Die Statik im Eisenbetonbau” (1927) was so authoritative it was nicknamed the Beyer Bible; TU Dresden still calls its civil engineering building the Beyer-Bau. source
  • Willy Gehler Professor and head of materials testing, Dresden · 1876–1953 Directed the building-technology section of the Saxon state materials testing office. TU Dresden’s own page says the competing prestressing systems of Hoyer and Freyssinet were first evaluated in Dresden — and that Gehler is little known today partly because of his Nazi party membership. source
  • Elmar Lohk Architect, the Kadriorg Stadium · 1901–1963 Architect of the grandstand. Komendant made the technical drawings and the structural calculations and supervised the work; more recent research names Komendant and the engineer Tarmo Randvee as its main designers. source
  • Heinrich Laul Engineer; the deformation calculations for the Kadriorg roof · 1914–2007 Was in close contact with Komendant at the technical university in 1937–1938, when Komendant was reading the reinforced-concrete courses. He went on to become a central figure in Estonian twentieth-century structural engineering himself. source
  • Ernst Kesa Architect, the Tartu grain elevator They designed the elevator together in 1939–1940; it was built in 1941. One of the very few of his Estonian works for which both the design years and a named architectural collaborator are on record. source
  • Marshall Meyers Kahn’s project architect, the Kimbell It was he who pointed out that a cycloid curve would lower the ceiling and bring other advantages. Kahn then fixed the exact shape in collaboration with Komendant. Three people behind one curve — which is exactly why authorship at the Kimbell is contested. source
  • Preston Geren The associated Fort Worth architect, the Kimbell Brought in to keep costs down, he argued the cycloid vaults would be too expensive and pressed for a flat roof. The dispute was settled by a division of responsibility: Geren the foundations and basement, Komendant the upper floors and the shells. source
  • Moshe Safdie Architect, Habitat 67 354 modules making 158 dwellings, a suspension system of post-tensioned boxes. Safdie gives the Kahn question a control case: the same engineer, a very different architect, in the same years. source
  • Nicholas Gianopulos Co-founder of Keast & Hood; Kahn’s other engineer · 1925–2018 Komendant was not Kahn’s only engineer. Keast & Hood was the structural consultant for Kahn, Venturi and Giurgola, and the firm calls itself engineer of record at Richards with Komendant “in association”. Whether the two ever overlapped, no source found says. source
  • Anne Tyng Kahn’s collaborator from 1945 · 1920–2011 The critic Luis Fernández-Galiano writes that between 1945 and 1955 she brought “the abstract order of structural geometry” into Kahn’s work, and calls Kahn’s resistance to acknowledging that debt deplorable. The same shape of grievance as Komendant’s, from the other side of the office — though the sources do not tie her to Richards. source · en.wikipedia.org
  • Carl-Dag Lige Architectural historian; curator and editor Began in 2016 the research that asked exactly what nobody knew: what Komendant did in Estonia up to 1944 and what he did abroad. Out of it came the 2020 exhibition and the 2022 monograph. His research field is precisely Estonian architecture and structural engineering, 1918–1944. source
  • Kenneth Frampton Architectural historian Singled out the Kadriorg grandstand. That an internationally prominent historian of modern architecture noticed a 1930s Estonian stadium stand is the clearest sign that his Estonian work is not merely local heritage. He also wrote for the 2022 monograph. source
  • Merike Komendant Phillips Daughter · b. 1939 Born in Tallinn in 1939, she left with her parents for Germany in 1944 and for America in 1950. She and her brother G. Jüri Komendant gave their father’s drawings, photographs, books and personal effects to Estonia — the foundation of the exhibition and the collection. source
  • Mikhail Osanov Bucknell honors thesis, 2014 Reconstructed Komendant’s Kimbell calculations and re-analysed the shells by finite elements. His thesis is where the figures this page’s bench measures itself against come from: six cables at 174 kips, −903 psi at the crown, 37 psi at the edge. source
  • Gustave Magnel Belgian engineer; the Walnut Lane Bridge · 1889–1955 He and his student Charles Zollman designed the first major prestressed-concrete bridge in the United States, completed in late 1950. Komendant reached America that same year — and every source calls him a pioneer, none the first. source

Places

  • 58°52′N 25°43′E

    Nurmsi

    A small farming village, today part of Paide municipality; in his childhood, Mäo parish. Its point is Wikidata’s — a village in Paide, Järva County — and the compass, the footer and the map measure from it.

    His own Wikidata item gives the place of birth as Mäo. Wikidata Q762422

  • 58°53′N 25°34′E

    Paide

    The county town of Järvamaa, in whose municipality the village of Nurmsi lies today.

  • 51°03′N 13°44′E

    Dresden

    Five years of study, in Beyer’s and Gehler’s school, where the prestressing systems of Hoyer and Freyssinet were first compared. And then back in 1944, to teach — the city burned in February 1945, so that period must have been short.

  • 51°02′N 13°44′E

    The Dresden technical college

    The Saxon technical college, today the Technical University of Dresden, where the Penn archive has him study from 1929 to 1934. Its point is Wikidata’s for today’s university.

  • 59°26′N 24°45′E

    Tallinn

    Where his 1975 book first appeared in Estonian in 2019, and where the Museum of Estonian Architecture opened its exhibition of his work in 2020. Chapter 08 points here.

  • 59°26′N 24°47′E

    The Kadriorg Stadium

    A 12.8-metre cantilevered roof over a stand 51 metres wide, with no front columns, opened in 1938. The best-known thing he built in Estonia, and the one Kenneth Frampton has singled out.

  • 58°24′N 26°47′E

    The Raadi hangar, Tartu

    Built 1935–1936. Tartu City Museum says it carries a prestressed-concrete solution Komendant developed for it; Carl-Dag Lige, in Sirp, calls the Tartu aircraft hangar the earliest known structure in Estonia using steel prestressing in its concrete. Its point is Wikidata’s for the Raadi airfield, whose complex the museum says the hangars belonged to.

  • 58°23′N 26°43′E

    Tartu

    The town of the Raadi hangar, the grain elevator and the water tower on Õpetaja Street, whose tank Komendant calculated.

  • 40°51′N 74°12′W

    Upper Montclair

    The New Jersey town where he opened his consulting practice in 1950, and where the sources say he died in 1992 — a place repeated rather than documented. Forty-two years of American practice run from one suburb.

    His own Wikidata item gives the place of death as Montclair. Wikidata Q762422

  • 39°57′N 75°10′W

    Philadelphia

    Kahn’s city: the Richards laboratories, and the University of Pennsylvania’s Architectural Archives, which hold Komendant’s papers.

  • 39°57′N 75°12′W

    The Richards laboratories, Philadelphia

    The first multi-storey precast, prestressed, post-tensioned rigid frame in the United States: 1 019 members, the largest offset one sixteenth of an inch. Flawed in use — glare, heat, partitions before move-in day — and decisive in influence.

  • 39°57′N 75°12′W

    The Architectural Archives, University of Pennsylvania

    Holds his papers from 1945 to 1981: drawings and correspondence with Kahn, Safdie and Mitchell/Giurgola, construction and completed-work photographs, the production materials for both books. In 2020 it lent over eighty objects to the Tallinn exhibition. Its point is Wikidata’s for the University of Pennsylvania.

  • 32°50′N 117°17′W

    La Jolla

    Where the Salk Institute stands, its laboratory floors carried on his Vierendeel trusses.

  • 32°53′N 117°15′W

    The Salk Institute, La Jolla

    Column-free laboratory floors 65 by 245 feet, carried on nine-foot Vierendeel trusses. To get permission in a seismic zone he put lead-zinc slip plates between each truss and its columns.

  • 32°45′N 97°20′W

    Fort Worth

    The Texas city where the Kimbell Art Museum stands.

  • 32°45′N 97°22′W

    The Kimbell Art Museum, Fort Worth

    Sixteen cycloid vaults in units of six, four and six, each a hundred feet long and four inches thick, each on four corner columns. He did not come to the opening. Fifty years on the museum runs a programme called “Komendant and Kahn”.

10TimelineAjajoon

10 · 1906 — 1992

Eighty-five years, and thirty-one after.

34 rows: 7 dated to the day, 3 to the month and 24 only to the year; 6 of them after his death. Choose what to look at and scroll; the line in the margin grows with you and turns grey over the German years.

34 rows 7 to the day 3 to the month 24 to the year 6 after his death

9 958 km 4 counted legs, summed before rounding; rounded one by one, 9 959 km · the longest Dresden → Upper Montclair 6 481 km, 65.1 %

The office’s reach from Upper Montclair: spokes, not legs

  1. Philadelphia · 128 km
  2. Fort Worth · 2 238 km
  3. La Jolla · 3 894 km
The legs as the crow flies, from the same coordinates the compass and the map use, each as long as its distance. The coordinates are Wikidata’s, each place with its item and revision. The counted legs are summed before rounding, 9 958 km; rounded one by one they add to 9 959 km. The longest, Dresden → Upper Montclair, is 65.1 per cent of the whole. The line starts at Nurmsi, the village he was born in, and the first leg is counted too. The office's work from Upper Montclair is spokes from one point, not a journey from one site to the next.

34 of 34 events

The grey band on the thread: 1944 → 1950, six years and 2 192 days, both ends dated only to the year and counted from 1 January; under it, 2 rows (1945 · Five years of broken bridges; 13 February 1945 · Dresden is firebombed).

  1. 1900s
    1. Nurmsi Life

      Born

      In the village of Nurmsi, Mäo parish, Järva County. English summaries name the parish, Estonian ones the village.

      source
  2. 1920s
    1. Estonia Life

      Foreman under Zeren

      After his military service, by Carl-Dag Lige’s account as a foreman and later a site manager. Konstantin Zeren steered him towards civil engineering; Lige gives no year, so this row sits at the end of the decade.

      source
    2. Dresden Life

      The technical college at Dresden

      A five-year course, per the Penn archive. The statics of reinforced-concrete structures, out of Kurt Beyer’s textbook.

      source
  3. 1930s
    1. Tallinn Life

      The degree, and the Ministry of Roads

      Graduation and a post at the Estonian Ministry of Roads. The University of Pennsylvania dates the return to 1936; Estonian sources to 1934.

      source
    2. Tartu Work

      The Raadi hangar

      1935–1936. A prestressed-concrete solution developed for this building — the earliest known in Estonia.

      source
    3. Tallinn Work

      Lectures at the technical university

      1937–1939. They laid the foundations of a school of concrete shells attached to Tallinn, and taught future engineers new calculation methods.

      source
    4. Tallinn Work

      The Kadriorg grandstand opens

      Designed 1936, architect Elmar Lohk. A 12.8-metre cantilevered roof over a stand 51 metres wide.

      source
    5. Tallinn Work

      An Estonian handbook, unfinished

      From his own lectures. War and emigration left it unfinished.

      source
  4. 1940s
    1. Tartu Work

      The Tartu grain elevator is built

      Designed 1939–1940 with the architect Ernst Kesa. The same year saw the opening of the Õpetaja Street water tower, whose tank calculations were his.

      source
    2. Estonia — Germany Life

      Out

      Three incompatible versions: shipped out for war projects, emigrated at the war’s start, or fled in the autumn of 1944 with his wife and two children.

      source
    3. Germany Work

      Five years of broken bridges

      For the US Army, 1945–1950. Studying broken structures became his professional capital.

      source
    4. Dresden Life

      Dresden is firebombed

      The Allied bombing of Dresden begins on the night of 13 February 1945 and destroys the city centre.

      source
  5. 1950s
    1. New York Life

      Arrival

      To New York with his family, and at once his own consulting practice in Upper Montclair, New Jersey — an unusual start for a displaced person.

      source
    2. Philadelphia Legacy

      The Walnut Lane Bridge — not his

      The first major prestressed-concrete bridge in the United States was completed in late 1950, designed by Gustave Magnel and Charles Zollman. Komendant reached America that same year.

      source
    3. New York Work

      “Prestressed Concrete Structures”

      Two years after arrival, grown straight out of the German bridge work. Later translated into Chinese and Russian.

      source
    4. Philadelphia Kahn

      The collaboration begins

      From 1956, if the book title is to be trusted; from the Richards laboratories, 1957. English Wikipedia says they met in 1956 and names no occasion; the University of Pennsylvania’s School of Design says Kahn asked him that year to consult on a project for a competition in Chicago.

      source
    5. Philadelphia Kahn

      Kahn gets the Richards commission

      The Alfred Newton Richards Medical Research Laboratories at the University of Pennsylvania, on which Komendant served as structural consultant — the project the partnership is usually said to begin with. From here to Kahn’s death is seventeen years.

      source
  6. 1960s
    1. Philadelphia Kahn

      The Richards laboratories

      1 019 precast members, post-tensioned in three directions. The completion year is given as 1960, 1961 or 1962.

      source
    2. New York Legacy

      MoMA’s Richards exhibition

      Flawed in use, decisive in influence: the building got its own exhibition at the Museum of Modern Art.

      source
    3. La Jolla Kahn

      The Salk Institute is finished

      Nine-foot Vierendeel trusses that make a whole service floor. Four buildings became two, and he re-engineered the lot.

      source
    4. Montreal Kahn

      Habitat 67, with Moshe Safdie

      354 modules making 158 dwellings, boxes of 70 to 90 tons apiece. The same engineer, a very different architect, in the same years.

      source
  7. 1970s
    1. Harrisburg Kahn

      The Olivetti-Underwood factory

      72 prestressed units in an eight-by-nine grid. The structural module is the architecture.

      source
    2. Fort Worth Kahn

      The Kimbell opens — without him

      Sixteen cycloid vaults, 120 000 square feet. Komendant deliberately stayed away from the opening ceremonies.

      source
    3. New York Kahn

      Kahn dies

      Eighteen years end. A year later Komendant’s book appears.

      source
    4. Englewood Work

      “18 Years with Architect Louis I. Kahn”

      Maris Mändel in Sirp: a portrait only somebody who feels himself the sitter’s equal can give.

      source
  8. 1980s
    1. USA Work

      Patent 4 195 453

      A modular multi-floor building with vertical post-tensioning — his own system, under his own name.

      source
    2. USA Work

      “Practical Structural Analysis for Architectural Engineering”

      Prentice Hall. A fourth book, fifty-three years after the Dresden degree.

      source
  9. 1990s
    1. Upper Montclair Life

      Death

      The place is given as Upper Montclair, New Jersey, where he had kept his office since 1950 — repeated by the sources, documented by none read here.

      source
  10. 2010s
    1. Tallinn Legacy

      The research begins

      The Museum of Estonian Architecture, curator Carl-Dag Lige: what he did in Estonia up to 1944 and what he did abroad.

      source
    2. Tallinn Legacy

      The book in Estonian

      “18 aastat arhitekt Louis I. Kahniga”, translated by Epp Aareleid (In Nomine, 2019) — translated into Japanese, Spanish and French long before Estonian.

      source
  11. 2020s
    1. Tallinn Legacy

      “Miracles in Concrete”

      At the Rotermann Salt Storage from 10 January. The University of Pennsylvania lent over eighty objects; his hat, his site helmet and his pipe were shown too.

      source
    2. Tallinn Legacy

      The Cultural Endowment’s annual prize

      The exhibition won the annual prize for architecture.

      source
    3. Basel Legacy

      “Miracles in Concrete” from Birkhäuser

      Edited by Carl-Dag Lige; inside it, Kenneth Frampton’s text and Oscar Tenreiro’s 1985 interview, the only known in-depth interview with him.

      source
    4. Fort Worth Legacy

      The Kimbell symposium

      “Komendant and Kahn: Engineer and Architect” — at the museum whose opening he did not attend.

      source

34 rows; a month-only row is counted from the first of its month and a year-only row from 1 January. The last row is from 2023, 31 years after his death.

11LegacyPärand

11 · What still stands

An archive in Philadelphia, an exhibition in Tallinn, a symposium in Fort Worth.

What has been said of him, what the sources count, what carries his name, and where all of it is written down. And at the end, a list of what this page does not know — longer than usual for this man, because nobody asked these questions before 2016. Every line carries its source.

  • “Kahn determined the exact shape of the vault through his collaboration with a structural engineer, Dr. August E. Komendant.”

    Kimbell Art Museum the museum’s own page The museum draws the line precisely: the idea Kahn’s, the shape a collaboration. That is a division of labour, not an attribution of the cycloid to either man alone — and no page reached carries Komendant’s own counter-claim from his book. the source’s own words · the source could not be opened to check it source · kimbellart.org
  • “What Komendant describes is neither an adoring glance from below nor a glossy journalistic picture, but a friendly and honest portrait, of the kind only a person who feels himself the equal of the one portrayed can give.”

    Maris Mändel Sirp, 20 December 2019 — reviewing the Estonian edition of the 1975 book (translated by Epp Aareleid; In Nomine, 2019) That claim of equality is how the book’s bluntness is read in Estonia. In the same weekly, on 10 April 2026, Karin Paulus wrote of the Komendant exhibition at Tallinn University of Technology under the heading “Komendandist. Mitte küll köitvalt, ent tummiselt”. this page’s translation from the Estonian · the source holds it word for word source · sirp.ee
  • “To August whose genius has made my buildings his buildings and are the very ones that have wonder about them.”

    Louis I. Kahn a handwritten dedication, as Wikipedia reports it By Wikipedia’s account this is Kahn’s handwritten note, reproduced in the introduction to Komendant’s memoir; the grammar is the note’s own. Wikipedia dates their first meeting to 1956 in another paragraph and names no occasion; the University of Pennsylvania’s School of Design says Kahn asked him that year to consult on a project for a competition in Chicago. the source’s own words · the source holds it word for word source · en.wikipedia.org

One more sentence said of him stands in chapter 08, Credit, which quotes it with its source — Wikipedia once; it is not printed a second time here.

The record in numbers

26 figures, each with its source. 9 of them are counted on this page, and those show their arithmetic.

Honours and memory

In his name

  • The August Komendant collection, Museum of Estonian Architecture His children, Merike Komendant Phillips and G. Jüri Komendant, sent their father’s belongings from their American homes, and these became the foundation of the collection; at the family’s wish part of the books, drawings and personal effects went to the library and museum of Tallinn University of Technology, to which Martin Komendant, a grandson of his brother Johannes, gave the lecture notes, the manuscript of the reinforced-concrete handbook and the papers of the 1930s. The war-years documents, if they survive, are in those holdings rather than online.
  • The August E. Komendant collection, University of Pennsylvania His professional practice from 1945 to 1981: drawings and correspondence with Kahn, Safdie and Mitchell/Giurgola, construction photographs, the production materials for both books. The obvious primary source for settling questions of division of labour and credit.
  • Philadelphia’s Roundhouse The former Philadelphia Police Administration Building (1959–1962). The register of Docomomo US, describing its precast, prestressed concrete system: “Engineered by August Komendant as one of the first of its kind in the country”. An example of his name surviving in American documents outside the Kahn projects.

Written about him

  • 2022 Miracles in Concrete: Structural Engineer August Komendant ed. Carl-Dag Lige. Birkhäuser with the Museum of Estonian Architecture. The first comprehensive account, covering the Estonian and German years that English sources otherwise skip. Inside it, Kenneth Frampton’s text and Oscar Tenreiro’s 1985 interview — the only known in-depth interview with him.
  • 2022 Betoonist võlutud. Ehitusinsener August Komendant Carl-Dag Lige. The Estonian edition and exhibition catalogue. A career from the 1930s to the 1980s and more than two hundred projects; most of its roughly five hundred illustrations appeared there for the first time. Anyone wanting the full list of his Estonian structures has to go to this book.
  • 1975 18 Years with Architect Louis I. Kahn August E. Komendant. Aloray, Englewood, New Jersey. Published a year after Kahn’s death. Not digitised or searchable; into Estonian only in 2019 — “18 aastat arhitekt Louis I. Kahniga”, translated by Epp Aareleid, In Nomine — long after Japanese, Spanish and French.
  • 2014 Structural Analysis of the Kimbell Art Museum Shells Mikhail Osanov. A Bucknell University honors thesis that reconstructs Komendant’s own design calculations and re-analyses the shells in ADINA. Every figure this page’s bench measures itself against comes from it.
  • 1952 Prestressed Concrete Structures August E. Komendant. McGraw-Hill, two years after he reached America. The Penn archives say it grew directly out of his experience rebuilding war-damaged German bridges. Translated into Chinese and Russian.
  • 1972 Contemporary Concrete Structures August E. Komendant. McGraw-Hill. A second major technical book twenty years on, in the same year the Kimbell opened.
  • 2019 Julgus liigutada piire Carl-Dag Lige. Sirp, 10 May 2019. The fullest Estonian account read here of the Dresden years, the Estonian decade and the German period.
  • 2023 Review: Miracles in Concrete Journal of the Society of Architectural Historians. Volume 82, issue 3, p. 348. A peer-reviewed review is a useful check on how historians received the book’s account; the review itself could not be opened.

What this page does not know

Two countries, two wars, two languages and a life’s work whose most important questions nobody had even asked until 2016. What follows is not modesty but a list: each line is a place where the sources parted or fell silent, and which this page therefore does not assert.

  • Whether he was a doctor. In American architectural circles he was routinely addressed as “Dr. Komendant”, yet no source reached here names a degree or a dissertation. One search result claimed that in the autumn of 1944 he still hoped to finish a doctorate under Kurt Beyer; the page behind it could not be opened. en.wikipedia.org, kimbellart.org
  • How he got to Germany in 1944. English Wikipedia says he was shipped out for war projects; the Museum of Architecture’s exhibition text says he emigrated at the start of the war; Carl-Dag Lige in Sirp and Estonian World say he fled with his wife and two children. No source gives a month or a route. en.wikipedia.org, arhitektuurimuuseum.ee, sirp.ee, estonianworld.com
  • Whether he came to the US Army through General Patton. The story circulates widely. English Wikipedia tells it with a footnote to Carter Wiseman’s biography of Kahn (2007, p. 96), and gives the bridge anecdote in it as Komendant’s own account; no archival or institutional source confirms it. The documented core is five years of bridge reconstruction, not an internment and a recruitment. en.wikipedia.org
  • What exactly he did in Germany between 1945 and 1950. Carl-Dag Lige writes in Sirp that he worked for the US Army for five years, helping to rebuild bridges and other infrastructure; no source read here names the unit or the place. sirp.ee
  • Who the engineer of Richards was. Popular accounts say Komendant; the firm Keast & Hood calls itself engineer of record with Komendant “in association”. And the completion year is given as 1960, 1961 or 1962. keasthood.com (archived), en.wikipedia.org
  • How thick the Kimbell shell is. Four inches (Wendy Lesser), five inches (Wikipedia), or ten centimetres, which is four inches (the Museum of Estonian Architecture). Lesser adds that the arch thickens towards the crown, so a single figure may describe only one point. This page’s bench carries both readings as separate members. fsgworkinprogress.com, en.wikipedia.org, arhitektuurimuuseum.ee, architecturelab.net
  • How wide each vault is, and how wide the crown slot. Wikipedia, SAH Archipedia and the contractor give 20 feet; 23 feet is reported for the Kimbell’s own page, yet the same research elsewhere reads that page as printing 20 and could not pin 23 feet to any page — this page’s four-inch reading carries 23 feet all the same, and the five-inch reading 20. For the slot the only citable figure is 15 inches; the three-foot figure that circulates elsewhere could not be attached to any source. kimbellart.org, en.wikipedia.org, sah-archipedia.org, architecturelab.net
  • How many cables are in a shell. Osanov’s thesis reconstructs six; the Museum of Estonian Architecture’s exhibition text says three. The tendon profile — how deep it runs at midspan — no source gives, and this page’s bench chooses it itself, by fitting Komendant’s published edge tension. digitalcommons.bucknell.edu, arhitektuurimuuseum.ee
  • How many projects he had altogether. Forty (Estonia before 1944), at least fifteen (with Kahn), more than two hundred (the lifetime). Three different scopes, routinely quoted side by side without their scopes attached — which is how one career gets both inflated and shrunk in the same paragraph. sirp.ee, tartu.postimees.ee, blogs.lt.vt.edu (archived)
  • Whether he was professor at the University of Pennsylvania from 1959 to 1974, or simply “a professor”. One Wikipedia page gives years that bracket the Kahn period exactly; another gives none, and no university source that would adjudicate was reached. en.wikipedia.org, en.wikipedia.org
  • What his own book actually says. “18 Years with Architect Louis I. Kahn” is not digitised or searchable, which is why the unsparing passages attributed to it are so hard to verify at second hand. There is not one sentence from that book on this page. openlibrary.org
  • Whether the cycloid’s shape is Meyers’s, Kahn’s or Komendant’s. The Kimbell’s own page gives the idea to Kahn and the shape to the collaboration; another source says the project architect Marshall Meyers pointed to the curve. Three people for one curve, and no page reached carries Komendant’s own counter-claim. kimbellart.org, en.wikipedia.org

The limits in numbers

  • 4 photographs of his buildings are on this page, each under the licence Wikimedia Commons gives for it; none of him — Commons holds no picture of him under his name. Photographs of him and of his work are held by the University of Pennsylvania’s Architectural Archives (photographs of construction and of completed works) and the Museum of Estonian Architecture (over two hundred photographs in the 2020 exhibition).
  • 0 sentences in his own words are on this page, because no source it could read prints one. 1 sentence is a second-hand report of what he said and 4 are other people’s words about him. Of the sentences quoted, 4 were found word for word in their sources, and for 1 the source could not be opened to check.
  • 3 sentences have been taken off this page: read against their sources, they are not there in the wording the page had — “Komendant explained that the form was really a shell acting …” (en.wikipedia.org), “Komendant was not above expanding on his share of the …” (en.wikipedia.org) and “Kahn failed properly to acknowledge his debt to Anne Tyng, …” (arquitecturaviva.com).
  • 7 of the 34 events on the timeline are dated to the day, 3 to the month and 24 to the year.
  • 0 places on this page lack coordinates: all 16 take theirs from Wikidata. His own Wikidata item (Q762422) gives place of birth Mäo, place of death Montclair and no burial, where this page says Nurmsi and Upper Montclair. And no source names his grave.
  • 0 videos: there is no film of him on this page. No documentary about him was found, only the museum’s own video tour of the 2020 exhibition.
  • 9 of the 26 figures in the record are counted on this page, and each of those shows its arithmetic; the other 17 are printed as their sources give them.
  • 1 parameter is fitted rather than sourced: the tendon runs 371 mm below the centroid at midspan, chosen so that the bottom fibre lands on his +37 psi, which is +0.255 MPa (the page’s arithmetic gives +0.259). His crown stress and his deflection are then checks, not fits.
  • 9 numbers are this page’s own, not a source’s: 2 kN/m of load on top of the self-weight (the bench, the opening figure, the ledger); 15 % of long-term loss, a rule of thumb (the bench, the ledger); the ACI 209 strength curve, 0.70 f′c at 7 days, since no source gives the cement (the bench, the opening figure, the window figure); Grade 270 strand jacked to 0.74 of 1 860 MPa (the strand count); 98.7 mm² a strand (the strand count); 25 mm of concrete under the tendon (the drape’s limit on the bench); 75 kg a person (the opening figure’s crowd); the tendon’s depth at the anchorages at 0.35 of the midspan one (the bench, the opening figure); 2 400 kg/m³ of concrete (the self-weight everywhere).
  • 3 of his 7 figures the page’s arithmetic does not reproduce in its opening state: the deflection before losses, 7.54 mm against his 3.86 mm; the design moment, 2 607 kN·m against his 6 372 kN·m; the support reaction, 342.1 kN an end or 171.1 kN a corner against his 402.6 kN. The ledger under the bench prints all seven.
  • 2 places where the notes behind this page contradict themselves, so the page prints both readings and leaves the matter open where the source could not be opened: the first meeting — one note says 1956, when Kahn asked him to consult on a Chicago competition entry, and another that nothing documents a date or an occasion of first contact; Wikipedia says 1956 and names no occasion, and the University of Pennsylvania’s School of Design says 1956 and a competition in Chicago; the vault’s width on the Kimbell’s own page — one note reads 23 feet there, two read the same page as 20 feet, and 23 feet could not be pinned to any page.
  • 13 sources could not be read for this page: 10 web pages that would not open or gave no text (kimbellart.org, sirp.ee, digitalcommons.bucknell.edu, sah-archipedia.org, sah-archipedia.org, keasthood.com, arhitektuurimuuseum.ee, ekabl.ee, openlibrary.org and degruyterbrill.com) and 3 works that are in print only or catalogued without their text (openlibrary.org, degruyterbrill.com and worldcat.org).
  • 8 of the 12 lines above are places where the sources differ, and the page shows both sides; the other 4 are places where they are silent.

What is not claimed here: that he brought prestressed concrete to America — the Walnut Lane Bridge was finished in late 1950 to Gustave Magnel and Charles Zollman’s design, the same year Komendant arrived; that Kahn was his only architect — Habitat 67 is with Moshe Safdie and gives the whole Kahn question a control case; that every Kahn building has him behind it — Witold Rybczynski, in his review “Genius in Concrete” (The New York Review of Books, 10 May 2007), names Bryn Mawr, Exeter and the Yale Center as buildings where Kahn did not work with him; or that the credit was his. English Wikipedia says that Kahn found it hard to share credit, and quotes Kahn’s biographer Carter Wiseman, by whose account Komendant was not above expanding his own share of it.

Trails

↑ ↓ to move · Enter to open · Esc to close ·