Fluctuation of Wind Pressure on the Roof of a House

Publisher’s version / Version de l'éditeur:

Technical Translation (National Research Council of Canada), 1967

READ THESE TERMS AND CONDITIONS CAREFULLY BEFORE USING THIS WEBSITE.

https://nrc-publications.canada.ca/eng/copyright

Vous avez des questions? Nous pouvons vous aider. Pour communiquer directement avec un auteur, consultez la première page de la revue dans laquelle son article a été publié afin de trouver ses coordonnées. Si vous n’arrivez pas à les repérer, communiquez avec nous à PublicationsArchive-ArchivesPublications@nrc-cnrc.gc.ca.

Questions? Contact the NRC Publications Archive team at

PublicationsArchive-ArchivesPublications@nrc-cnrc.gc.ca. If you wish to email the authors directly, please see the first page of the publication for their contact information.

NRC Publications Archive

Archives des publications du CNRC

For the publisher’s version, please access the DOI link below./ Pour consulter la version de l’éditeur, utilisez le lien DOI ci-dessous.

https://doi.org/10.4224/20331689

Access and use of this website and the material on it are subject to the Terms and Conditions set forth at

Fluctuation of Wind Pressure on the Roof of a House

Ishizaki, H.; Huh, C.

https://publications-cnrc.canada.ca/fra/droits

L’accès à ce site Web et l’utilisation de son contenu sont assujettis aux conditions présentées dans le site LISEZ CES CONDITIONS ATTENTIVEMENT AVANT D’UTILISER CE SITE WEB.

NRC Publications Record / Notice d'Archives des publications de CNRC:

https://nrc-publications.canada.ca/eng/view/object/?id=27a6c6bb-957a-4904-9c3f-7ae3643fcbeb https://publications-cnrc.canada.ca/fra/voir/objet/?id=27a6c6bb-957a-4904-9c3f-7ae3643fcbeb

PREFACE

Wind effects on buildings and structures have been extensively investigated through model studies in wind

tun-nels. Much of the work presently available is limited to

a consideration of static, or steady wind loads arising from

low-turbulence airflow. Natural wind, however, is turbulent

and wind loads on structures are dynamic, or at least have a fluctuating component superimposed on the average load.

This paper by Dr. H. Ishizaki and C. Huh presents

results of wind tunnel tests on models as well as

measure-ments on an actual house roof in natural wind. The

objec-tive was to investigate the dynamic aspects of wind

load-ing on low slope house roofs. The Building Structures Section

has an interest in measurements of wind pressure on full-scale structures as well as in work dealing with the dynamic effects of wind.

The Division of Building Research is pleased to make this research information available to other Canadian research workers through the Technical Translation Series of the

National Research Council. The contribution of the translator,

Dr. Makoto Inaba, is gratefully acknowledged, as is that of

W.A. Dalgleish of this Division who checked the translation.

Ottawa

November 1967

R.F. Legget Director

NATIONAL RESEARCH COUNCIL OF CANADA

Technical Translation 1306

Title: The fluctuation of wind pressure on the roof of a house

Authors: H. Ishizaki and C. Huh

Reference: Annual Report of the Disaster Prevention Research Institute,

Kyoto University, (6): 95-99, 1963

THE FLUCTUATION OF WIND PRESSURE ON THE ROOF OF A HOUSE Abstract

This paper presents the results of wind tunnel tests on the three different sized models of a house and the measurement of wind pressure on the roof of the actual house under natural wind

to investigate the dynamic behavior by the suction. In

conclu-sion, we found that the fluctuations of wind pressure behind the house were periodic and the behavior depended upon the wind veloc-ity.

1. Introduction

Houses damaged by strong winds often show that the damage was caused not only by static wind pressure but also by some dynamic, secondary force due to the change of the wind pressure.

Further, it is often found that some parts of the houses, i.e. windward roof surfaces of houses with roofs of small slope, the leeward roof surfaces of ordinary houses and their wall surfaces, are subject to considerable suction

force. The fluctuation of the wind pressure is especially great on windward

side of houses with roofs of small slope.

Repeated fluctuations of wind pressure appear to accelerate the

destruc-tion of such houses.

Evidently estimates of wind pressure for design purposes do not generally include the effect of the dynamic force due to fluctuating wind pressure, per-haps because the nature of the fluctuation of the wind pressure is not very

well known, especially from a quantitative standpoint. As a result, wind

pres-sure estimates for the design of buildings has been based solely on static wind pressure.

The authors planned to investigate the nature of the wind pressure fluc-tuation, and they began with a study of the behaviour of the wind pressure

fluctuation on a windward roof surface of small slope. The fluctuation was

de-termined with the aid of models of such houses in a wind tunnel and also from a full-scale house under natural conditions.

In order to check any possible variable factors that might depend on the size of the models, three models of different sizes but completely homologous

shape were used in the wind tunnel experiments. The experiments on the

full-*

scale house were conducted with an experimental house of the Shionomisaki wind pressure research laboratory of this institute.

*Translator's note: At N

37°45',

E

135°40'.

Known as the point where typhoons

o

8

=

15 , and their sizes are shown in Fig. 1.

and were made of lauan wood. Their surface

-4-2. Experimental and Determination Methods

(a) Wind tunnel experiments on models

The experiments were conducted in the Gottingen wind tunnel of this

in-stitute. It has an exhaust diameter of 1.0 m and a maximum flow velocity of

60 m/sec.

The models had a roof slope, They were all homologous in shape was polished carefully.

Each model was placed on a holder as shown in Fig. I, and was sUbjected

to four different air flow velocities, namely 15 m/sec, 20 m/sec, 25 m/sec, and

30 m/sec. The changes of wind pressure on the roof surface were fed to a

pres-sure gauge through a vinyl tube. The wind pressure gauge was of the wire

re-sistance strain gauge type, the strain being recorded by electro-magnetic

os-cillograph. The diameter of the hole for determining the wind pressure values

was 4 mm, and the wind pressure gauge was made to record the fluctuation of the

wind continuously on a wind pressure plate. It was sufficiently sensitive to

record a fluctuation of wind pressure of 0.1 g. The natural frequency of this

gauge is

73

cps. Therefore the wind pressure gauge could be used to measure

wind pressure fluctuations of approximately 40 cps. The measurement was

car-ried out simultaneously at the two points shown in Fig. I, and both were re-corded.

(b) Determination on the full-scale house

The size and the determination points are shown in Fig. 2. The building was

erected in the research laboratory at Shionomisaki. It had a lightweight iron

frame with panel walls reinforcing the construction. The roof was covered with

aluminium plates. Photograph 1 shows the bUilding. At the two determination

points shown in the Fig. 2, there were two wind pressure plates, 30 cm x 30 cm,

and changes of wind pressure on them were fed to a gauge through a vinyl tube.

The determination was carried out on January 23, 1963. The wind speed that

day was 10 m/sec - 18 m/sec, 10 m above the ground surface.

The main wind direction was northerly and the wind blew against the longer

side of the house. The change of the wind pressure was fed to a differential

type wind pressure gauge through a vinyl tube, and was recorded on a recording

oscillograph. The differential pressure gauge was designed so as to produce a

-5-a v-5-ari-5-able p-5-art of the pressure g-5-auge, The charge of electric potential change

was recorded by a potentiometer, The characteristic frequency of the variable

part containing the iron piece was 18 cps, and therefore the differential wind pressure gauge was estimated to have response of 10 cps frequency change,

3, The Results of Determination (a) Models

Photograph 2 shows part of the recorded wind pressure fluctuation data, the

amplitude is equivalent to the wind pressure, Table 1 shows the periods of

fluctuating wind pressure of the three models for various wind velocities, and

Fig, 3 shows the ratio of the fluctuating wind pressure to static wind

pres-sure,

As shown in Photograph 2, the period of wind pressure fluctuation is of two

types, one small and the other large, Also there is a phase difference

be-tween the two sets of wind pressure fluctuations 'at determination points No, 1

and No, 2, Apparently the phase difference was caused by transfer of an

air-eddy, If so, the velocity of transfer of the air-eddy can be estimated at

about three-tenths of the general air flow,

The phases of the wind pressure fluctuations at the centre and at the ends of the models are identical, and the fluctuations occur with the same amplitude,

(b) Full-scale house

Photographs 3a and 3b give part of the recorded data of the wind pressure

fluctuations, Photograph 3a gives data for the roof and the eave of the house,

while Photograph 3b gives them for the wall and the eave, The wind velocity

was 10 - 18 m/sec, 10 m above the ground surface,

Fig, 4 and 5 are the power spectra obtained from the wind pressure

fluc-tuation data at determination points No, 1 and No, 2, respectively, The

spec-tra were obtained by statistical processing of the energy distribution data of the wind pressure fluctuations with periods between 0 and 10 cps, using a

KDC-l computer, The wind pressure fluctuations, on which the power spectrum

is based, were taken from a continuous record for 30 seconds, the readings being taken every 0,05 seconds,

As seen in Fig, 4 and Fig, 5, both determination points, No, 1 and No, 2,

have a peak at 5 cps, The amplitude, i,e, the value of wind pressure

fluc-tuation, is also of interest, but since the wind itself in the natural field undergoes certain fluctuations, and hence the relationship of the combined fluc-tuating wind pressures to the static wind pressure is extremely complex, it was decided not to consider these data any further,

-6-4.

Discussion of the Results

The periods of the wind pressure fluctuation show a certain pattern,

al-though they are not completely regular. The period has a tendency to become

smaller with increasing wind velocity.

Changes of periods as a function of the size of the model are negligible. This fact and the data obtained on the full-scale house suggest that the period of wind pressure fluctuation does not change with the size of the constructions.

Assuming that the fluctuation periods are inversely proportional to size, in the case of full-scale houses, a wind velocity of 10 m/sec should result in

a period of a few seconds. However, the actual determined value is about 0.2

sec, and it can therefore be safely concluded that the size of the construction

does not affect the period. Of course, for full-scale constructions, because

of the perturbed flow of the air blowing against the construction, the effect of the velocity fluctuations does not lend itself so well to discussion as

does the data from a wind tunnel. Nevertheless, the data obtained from the

simultaneous determinations at the eave and on the wall of the full-scale house permitted us to conclude that the wind pressure fluctuations under nat-ural conditions are almost the same as those in a nearly homogeneous flow of air, i.e. as though the full-scale house had been placed in a wind tunnel.

The fluctuating wind pressure value may reach 0.7 - 0.8 times the static

wind pressure in the wind tunnel experiment. This shows that parts of the

house, especially the roof, which is likely to be subjected to a suction force,

are therefore exposed to continual fluctuations of wind pressure. In the case

of full-scale houses it is difficult to determine the ratio of the fluctuating wind pressure to the static wind pressure owing to the fluctuations of wind

speed under the natural conditions. Nevertheless, the same tendencies and

results as those obtained in the wind tunnel experiments were discerned.

5.

Conclusion

The above results indicated that both the models in the wind tunnel and the full-scale house under natural conditions were subjected to wind pressure fluc-tuations with fair regularity at those parts of the models or house which were exposed to a negative pressure, even when the wind blew at a constant speed.

It was found that the periods of fluctuation changed considerably as a function of the change of wind speed, but remained practically constant,

re-gardless of the size of the model. The fluctuations of wind pressure were very

great at parts exposed to a negative pressure - as much as 0.8 times the static wind pressure.

-7-With these results, the generally accepted method of wind pressure deter-mination for design purposes was not found to be satisfactory for the safety

of bUildings, since the designing of a building only for the static wind

pres-sure fails to take into account a fluctuating wind prespres-sure acting continuously in cycles.

Henceforth, there should be a more careful estimation of dynamic force which may affect the safety of the building constructions and this value should be

employed for wind-proofing purposes. The conventional method of designing

wind-proof constructions on the basis of the wind-pressure coefficient must be re-considered.

In closing, the authors wish to thank Mr. Mitsuda of this laboratory ror his co-operation in the preparation of the power spectra.

Table 1 _{The period of fluctuating wind pressure} by various wind velocities

ｾ

･ ｬ ｯ ｣ｩ ｴ ｹ

I115 m/sec

I

20 m/.sec

I

25 m/sec

I

30m/sec

Model

I

Fig. 1 The model size and it's position of the measurement points in wind tunnel test (0.73)

I

sec 0.100 sec 0.076 sec 0.060 sec Large (_) (0.58) (0.32) (0.28) 0.120 0.078 0.060 (0.58) (0.32) (0.28) 0.120 0.078 0.062 (0.76) (0.60) (0.34) (0.28) Small Middle HODEL B ｖ］ｊｏｾｬＳＭ･｣ ｎｾＧ I NoZ ,",sr"·.:'".::::: \ ｾ｜｟ｾＭＭＭ i i i ･ＮｾＺＺｾﾷＵｾｾＺＧＭｘＮｬＭＭＮｾ｟

___

J " i' I I I I p c .._1_ I

!'

I IT]

rn

_IT]

I

I

It

' -ｾ ---ｾＭＭ _.--- 200 0.50

Fig. 2 The actual house and stations of the measurement points

Fig. 3 The ratio of fluctuaing wind pressure value to the statical wind pressure value

-8-00 pi 'CpS 05 0.4 QJ • 0

.

ｾＭｾｾＭＭＢ

\

OJ Ｎｾ .9 B 7 3 o 4 5 6 ｾＭｴｐｓ

Fig. 5 Power spectrum of the wind pressure fluctuations of measurement point No.2

I . , I ! 'I I ! 1 !

J J 4 s 6 7 a .9

Cps

Power spectrum of the wind pressure fluctuations of measurement point NO.1 Fig. 4

Photo. 1 View of lllIlctual house

Photo. 2 The record of measurement of wind pressure fluctuations in wind tunnel

-9-Photo. 3a