Eiffel, Angot, Brazier, 1889-1914

Rémi

Introduction

Here is what Charles-Emile Brazier wrote in 1914, about his PhD thesis on cup anemometer calibration:

When I began my research, I did not think it would extend as far as it did in the course of my thesis.

This statement won’t surprise those who have attempted to better understand the craftmanship and science of measuring the wind. The topis is still very much at the heart of wind & site studies, due to the large sensivity of the results to the wind speeds (more or less of all scales and durations). I first encountered the work of Mr. Brazier while browsing the digital library of the National Library of France (BNF), http://gallica.bnf.fr/, but later on realised that he did author one the earliest scientific studies about cup anemometer calibration and design. He is cited in this very nice paper by Leif Kristensen from Risø for instance. The thesis is about 140 pages long, here is the full reference (the full thesis was only published afte WWI):

What I really like with this work is how much it connects with the ongoing experimental research in boundary layer meteorology, by the author and his peers. Many of the questions originally stem from the experimental setup deployed at the Eiffel tower by the Bureau Central Meteorologique de France whose data were scrutinised by Mr. Alfred Angot. The 1900 report can be found in the Aeolians.net library (also from Gallica). Beside the great amount of details about the technical setup (documented shortly in 1891 already), the document provides a great amount of details about the wind conditions measured at the Tower and at the Central Weather Bureau office located about 1 km away. Here is the full reference:

  • Travaux scientifiques exécutés à la tour de 300 mètres, de 1889 à 1900, by G. Eiffel (1900) (Aeolians.net library link)
Illustration of the measurement sensors and data loggers (located 1km away, at the Central Bureau offices) from the Eiffel tower atmospheric measurement setup. It was first documented by Alfred Angot in 1981, see text.
Location of the data loggers at the Central Weather Bureau, and the measurements at the top of Eiffel tower, on a 1900 map of Paris.

What is especially interesting is the discussion on the combined variations of temperature and wind speed with height, and the occurence of thermal inversion (where the temperature at the top of the tower is larger than close to the ground). There are also a number of statistics reported about the wind speed and direction, as well as considerations about storms events logged at very high frequency. All of this deserves a specific post, and I will certaintly come back to this report shortly. A few highlights below though, before continuing with the work of Brazier on the cup anemometer:

Mean diurnal wind speed profiles at the top and the bottom of the Eiffel Tower, from (Eiffel, 1900).
A temperature inversion event captured at the Eiffel Tower in November 1889. The text underlined in red reads: “conditions were completely calm at the bottom of the Tower, whereas at the top a warm SSW wind was blowing at 6 to 8 m/s”.
Example of slow and high-frequency wind speed measurements during the 1896-03-01 storm over Paris, at the top and at the bottom (Central Weather Bureau office) of the Eiffel Tower. From (Eiffel, 1900).

If you are curious about it, and read French, you can find details about the measurement technique in this report from Angot (1890). The anemometers and the data loggers all came from Richard Frères (the company still  exists today: http://www.jri.fr/en/about-jri/history). The basic idea is simple and elegant, see below:

  • the stylet is attached to a threaded rod (called “T” below)
  • the gearwheel “P” is spinning proportionnally to the rotation speed of the cup anemometer, it spins clockwise to it moves the rod to the right
  • the disk “D” spins at constant speed, and drives the wheel “R” which is connected to the rod, the disk spins clockswise, and R makes the rod moves to the right (it unscrews it from P)
  • it turns out that the distance x is proportional to the rotational speed of P, and therefore to the wind speed
Schematic illustration of the cinemographe connected to the cup anemometer. For a more complete view of the system, see (Angot, 1880) and http://patrimoine.bourgognefranchecomte.fr/connaitre-le-patrimoine/les-ressources-documentaires/acces-aux-dossiers-dinventaire/etude/69d5f931-c3a6-49ad-8614-9c8ab46a5c5c.html

Angot was (of course) quite enthusiatic about those observations, and suggested to Brazier that it would be a good idea to assess the accuracy of the wind measurements. He did so in this PhD thesis, and also contributed to linking temperature and wind speed variations, see three mentions of his work in the Monthly Weather Review of October 1919:

Series of mentions of the work from Brazier regarding wind speed and temperature profiles. All refer to the article “Sur les rapports du vent avec le gradient dans les couches basses de l’atmosphère” by Brazier in 1919 (link: http://visualiseur.bnf.fr/CadresFenetre?O=NUMM-3122&I=871&M=tdm)

Brazier’s research

So… are those wind measurements correct ? Well that is what Brazier investigated in this thesis. The thesis is (of course) long, and I have not read all the sections in details (so you know). Basically he tested several types of anemometers (not just cups) in the Eiffel wind tunnel (which still exists today). When I spoke to people around me, it came as a surprise to them that Eiffel was an big supporter of the early experimental research in aerodynamics. It is more widely known in France, as far as I am aware, but if you are curious about it please have a look to this nice article (it includes cool pictures).

Cup anemometer in the Eiffel wind tunnel (Brazier, 1914)

Brazier was not the first one to attempt validate cup anemometers, but from what I could find he was the first one to do it in a wind tunnel, using nowadays procedures. For previous works, see for instance:

  • Quaterly Journal of the Royal Meteorology Society, Vol. 18, 1892 (online link). See pp 165 and 257 for articles on comparisons between anemometers.
  • Symons’ Monhtly Meteorological Magazine, May 1879. (online link) See pp 56 for the derivation in open air of the cup anemometer constant by Dohrandt (1874), which involves testing the cup on a locomotive.

Brazier measured, for a range of reference wind speeds, the ratio between the rotational speed of the anemometers, and the reference wind speed for a number of configurations: different geometries (size of cups, lentghs of booms, etc), and different flow inclinations (by tilting the cup anemometer). As you can see below, the anemometer “I” turns out to have a pretty constant ratio.

Ratio between the rotational speed of the anemometer and the reference wind speed as a function of the reference wind speed (left), and the tilt angle (right).

Brazier, for every curves, fits the data to a quadratic expression (y=a+bx+cx^2), see below, he shows that the quadratic term is very much dependent of the distance D, which is twice the distance between the center of the cup and the axis of roation of the cup (please not: all cups have the same diameter, R).

Numeric values of the terms of the quadratic fit form the previous plot.

He then concludes that by shortening the arms of anemometer H to 122m, the quadratic term should be negligible, and the relationship wind speed / rotational speed become linear. An indeed, he derives that using ratio R/H=0.5 (he uses cups of 62mm diameter) the relationship is linear:

Results obtained by Brazier using a ration R/D of 0.5.

Something which is quite modern too in his work, is that he makes another set of tests, for two types of anemometers, in another wind tunnel (located in Saint-Cyr, build in 1916 for the French Air Force), see below how the results compare pretty well:

Comparison of tests in the Eiffel wind tunnel, and in the Saint-Cyr wind tunnel. Pretty neat (I think).

Lastly

The work of Brazier seems atrouciously modern to me, because it stems from boundary layer meteorology, and focuses on the same basic aspects of cup anemometer calibrations carried out today. That is: concerns about cup geometry, and reproduceability between different wind tunnels for instance. Brazier also sets the focus on measurement of the vertical component of  wind speed because he observed such upward movements at the top of the Tower (see below). He seemed to have intuited the link between temperature and wind speed gradient, and overall seemed like a very nice guy. He moved on later on to become the director the Saint-Maur observatory, and helped edit the 4th edition of Angot’s treaty on Meteorology after Angot passed away (his former supervisor was quite dear to him).

From (Brazier, 1914).

Angot himself seem to have been a weather and measurement enthusiast too, if you read French see below…

Nice words about Alfred Angot by Marcel Brillouin, in the preface to the 4th, revised edition of Angot’s treaty on Meteorology (published in 1928 after Angot passed away). Angot appears to have been someone quite passionate about weather and weather monitoring, securing good quality data, from France and overseas. No surprise he suggested to Brazier that calibration cups in a wind tunnel would be a good idea.

Comments and suggestions are welcome 🙂
Rémi

Published by Rémi