SAND92-7009

WIND LOAD DESIGN METHODS FOR GROUND BASED

HELIOSTATS and PARABOLIC DISH COLLECTORS
By J.A. Peterko a n d R.G. Derickson

iv it N I U N Lifvi'i 1ED
DISTRSiSU , iw.-'*
 DISCLAIMER

This report was prepared as an account of work sponsored by an

agency of the United States Government. Neither the United States
Government nor any agency thereof, nor any of their employees,
makes any warranty, express or implied, or assumes any legal liability
or responsibility for the accuracy, completeness, or usefulness of any
information, apparatus, product, or process disclosed, or represents
that its use would not infringe privately owned rights. Reference
herein to any specific commercial product, process, or service by
trade name, trademark, manufacturer, or otherwise does not
necessarily constitute or imply its endorsement, recommendation, or
favoring by the United States Government or any agency thereof. The
views and opinions of authors expressed herein do not necessarily
state or reflect those of the United States Government or any agency
thereof.

D IS C L A IM E R

Portions of this document may be illegible in electronic image

products. Images are produced from the best available
original document.
Issued by S a n d ia N a tio n a l L a b o ra to ries , o p e ra t e d for th e U n ite d S ta te s
D e p a r t m e n t of E n e rg y by S a n d ia C o rp o ratio n .
N O T I C E : T h is re p o r t was p r e p a r e d as a n a c c o u n t of work sp o n s o red by an
agency of th e U n ite d S ta t e s G o v e rn m e n t. N e ith e r th e U n ite d S ta t e s G o v e r n
m e n t n or a n y agency thereo f, n or an y of th e i r em ployees, n o r an y of th e ir
co n tr a c to r s, s u b c o n tra c to rs , or th e ir em ployees, m a k e s any w a rra n ty , express
or im p lied, or a ssu m es an y legal liability or re spo nsibility for th e accuracy,
co m p leten ess, or usefuln ess of any in fo rm a tio n , a p p a r a t u s , p ro d u c t, or
process disclosed, or r e p r e s e n ts t h a t its use w ould n ot infringe p riv ately
ow ned rights. R efere n c e h erein to a n y specific c om m ercial p ro d u c t, process, or
service by t r a d e n a m e , tr a d e m a r k , m a n u f a c tu r e r , or o therw ise, does no t
necessarily c o n s t it u te or im p ly its e n d o rs e m e n t, re c o m m e n d a tio n , or favoring
by th e U n ite d S ta t e s G o v e rn m e n t, an y agency th e r e o f or an y of th e i r
c o n tr a c to r s or su b c o n tra c to rs. T h e views an d op in io n s exp ressed herein do
n o t necessarily s ta te or reflect th ose of th e U n ite d S ta t e s G o v e rn m e n t, any
agency th e r e o f or an y o f th e ir co ntracto rs.

P r i n t e d in th e U n ite d S ta t e s of A m erica. T h is re p o r t has been re p r o d u c e d

d irectly from th e best available copy.

A vailable to D O E a n d D O E c o n tr a c to r s from
Office of Scientific a n d T e c h n ic a l I n f o rm a tio n
P O Box 62
O ak R idge, T N 37831
P rices available from (615) 576-8401, F T S 626-8401

Available to th e p ub lic from

N a tio n a l T e c h n ic a l I n f o rm a tio n Service
U S D e p a r t m e n t of C o m m erc e
5285 P o r t R oyal Rd
Sprin gfield, VA 22161
N T I S price codes
P r i n t e d copy: A03
M icrofiche copy; AOl
 SAND92-7009

DE93 002737

SAND92-7009
Unlimited Release
Printed September 1992

WIND LOAD DESIGN METHODS FOR GROUND-BASED HELIOSTATS

AND PARABOLIC DISH COLLECTORS

J.A. Peterka
R.G. Derickson
Fluid Mechanics and Wind Engineering Program
Fluid Dynamics and Diffusion Laboratory
Department of Civil Engineering
Colorado State University
Fort Collins, CO 80523

Sandia Contract 69-1865

ABSTRACT

The purpose of this design method is to define wind loads on flat heliostat and parabolic
dish collectors in a simplified form. Wind loads are defined for both mean and peak loads
accounting for the protective influence of upwind collectors, wind protective fences, or
other wind-blockage elements. The method used to define wind loads was to generalize
wind load data obtained during tests on model collectors, heliostats or parabolic dishes,
placed in a modeled atmospheric wind in a boundary-layer wind tunnel at Colorado State
University. For both heliostats and parabolic dishes, loads are reported for solitary
collectors and for collectors as elements of a field. All collectors were solid with negligible
porosity; thus the effects of porosity in the collectors is not addressed.

&
 IV

ACKNOWLEDGEMENTS

Development of th is document was supported by Sandia National Laboratories

under contract 69-1865. The active support and guidance of Mr. Jim Grossman of

Sandia Laboratories is acknowledged. The data contained herein were developed

under other contracts with Sandia, the Solar Energy Research In s t it u t e (now the

National Renewable Energy Laboratory) and other sources as referenced within the

document.
 TABLE OF CONTENTS

Section Page

ACKNOWLEDGEMENTS ................................................................................................................ i

LIST OF T A B L E S ......................................................................................................................... i i i

LIST OF FIGURES.................................................................................................................... iv

INTRODUCTION ........................................................................................................................ 1

ATMOSPHERIC WINDS ................................................................................................................ 3

MEASURED PEAK AND QUASI-STEADY PEAK WIND L O A D S .................................................. 5

SPECIFICATION OF DESIGN WIND SPEEDS ........................................................................... 7

PRINCIPLES OF WIND LOADING ........................................................................................... 9

COORDINATE SYSTEM OF COLLECTORS - Forces and Moments ..................................... 10

EFFECTS OF ATMOSPHERIC TURBULENCE ........................................................................... 13

LARGEST WIND LOADS ON ISOLATED COLLECTORS - Heliostats and Parabolic

D i s h e s ..................................................................................................................................... 17

LOAD COMBINATIONS FOR ISOLATED HELIOSTATS ANDPARABOLIC DISHES ................... 17

LOCAL PRESSURE DISTRIBUTION ........................................................................................... 32

MAXIMUM WIND LOADS ON COLLECTORS IN A FIELD -General Blockage Area . . . 41

MAXIMUM COEFFICIENTS FOR IN-FIELD HELIOSTATS ANDPARABOLIC DISHES . . . . 45

LOAD COMBINATION COEFFICIENTS WITHIN A FIELD........ .................................................. 45

EXAMPLE PROBLEM 1 ................................................................................................................ 57

EXAMPLE PROBLEM 2 ................................................................................................................ 61

REFERENCES............................................................................................................................. 65
 VI

LIST OF TABLES

Table Page

1 Estimated values of the surface roughness ...................................... 6

2 HELIOSTATS - Wind load c o e ffic ie n ts on isolated

h e lio s ta ts when one component is a m axim um .................................. 18

3 PARABOLIC DISHES - Wind load c o e ffic ie n ts on isolated

parabolic dishes when one component is amaximum ..................... IB

4 M ultip ly in g factors fo r isolated load combinations fo r

i n - f i e l d h e lio s ta t performance .......................................................... 55

5 M u ltiplying factors fo r isolated load combinations fo r

i n - f i e l d parabolic dish performance .................................................. 56
 VII

LIST OF FIGURES

Figure Page

1 Coordinate system for c o lle c to r (h e lio s ta t or parabolic

d i s h ) ................................................................................................................ 11

2 V ariation of Cpx> Cp^ and with turbulence in te n s ity for

an isolated he lio s ta t ............................................................................... 14

3 V ariation of mean Cy with elevation angle and turbulence

i n t e n s i t y ....................................................................................................... 15

4 Drag force c o e ffic ie n ts fo r the h e l i o s t a t .............................. 19

5 L i f t force c o e ffic ie n ts Cp^ fo r the h e l i o s t a t .............................. 20

6 Azimuth moment coefficients fo r the h e lio s ta t ...................... 21

7 Hinge moment coefficients fo r the h e lio s ta t ...................... 22

8 Mean drag force coefficients MEAN Cp^^ fo r the parabolic

d i s h ................................................................................................................ 23

9 Peak drag force coefficients PEAK Cp fo r the parabolic

d i s h ................................................................................................................ 24

10 Mean l i f t force coefficients MEAN Cp^ fo r the parabolic

d i s h ................................................................................................................ 25

11 Peak l i f t force coeffic ie n ts PEAK Cp fo r the parabolic

d i s h ................................................................................................................ 26

12 Mean azimuth moment coe ffic ie n ts MEAN Cp, fo r the parabolic

d i s h ................................................................................................................ 27

13 Peak azimuth moment c o e ffic ie n ts PEAK C fo r the parabolic

d i s h ................................................................................................................ 28

14 Mean hinge moment coe ffic ie n ts MEAN C for the parabolic

d i s h .................................................................. 29

15 Peak hinge moment coefficients PEAK Cp, for the parabolic

d i s h ................................................................................... 30

16 Mean pressure coefficients (a = 90 and ^ = 0) . 33

17 Mean pressure coe ffic ie n ts (a = 90 and ^ = 60) 34

18 Mean pressure coe ffic ie n ts (a = 90 and ^ = 90) 35

19 Mean pressure coefficients {a = 90 and ^ = 180) 36

 V III

LIST OF FIGURES (c o n t.)

Figure Page

20 Mean pressure coe ffic ie n ts (a = 60 and ^ = 0 ) .............................. 37

21 Mean pressure c o e ffic ie n ts (or = 50 and /9 = 6 0 ) .......................... 38

22 Mean pressure coefficients (or = 60 and 0 = 9 0 ) .......................... 39

23 Mean pressure coefficients (a = 60 and = 1 8 0 ) .......................... 40

24 Layout fo r sim plified GBA calculations for c o lle c to r units

in rows 1 through 5 .................................................................................. 43

25 Mean and peak drag force c o e ffic ie n ts , Fx, of a h e lio s ta t

unit w ithin a f i e ld of h e l i o s t a t s ...................................................... 46

26 Mean and peak l i f t force c o e ffic ie n ts , Fz, of a h e lio s ta t

un it within a f i e ld of h e l i o s t a t s ...................................................... 47

27 Mean and peak hinge moment coefficpents, MHy, of a h e lio s ta t

unit within a f i e ld of h e l i o s t a t s ...................................................... 48

28 Mean and peak azimuthal moment c o e ffic ie n ts , Mz, of a

h e lio s ta t unit within a f ie ld of heliostats ................................. 49

29 Mean and peak drag force c o e ffic ie n ts , Fx, of a parabolic

dish unit within a f ie ld of parabolic d i s h e s .............................. 50

30 Mean and peak l i f t force c o e ffic ie n ts , Fz, of a parabolic

dish unit within a f i e l d of parabolic d i s h e s .............................. 51

31 Mean and peak hinge moment c o e ffic ie n ts , MHy, of a parabolic

dish unit within a f i e l d of parabolic d i s h e s .............................. 52

32 Mean and peak azimuthal moment c o e ffic ie n ts , Mz, of a

parabolic dish unit within a f i e ld of parabolic dishes . . . 53
 WIND LOAD DESIGN METHODS FOR GROUND BASED HELIOSTATS
AND PARABOLIC DISH COLLECTORS

J. A. Peterka^ and R. G. Derickson^

INTRODUCTION

Optimum design of isolated units or fie ld s of heliostats and parabolic dish

c ollecto rs is dependent upon obtaining r e a l i s t i c design wind loads. In the

future these structures may become even more sensitive to wind loads as gravity

loads decrease through innovative technology.

The purpose of this design method is to define wind loads on f l a t

h e lio s ta ts and parabolic dish collectors in a sim plified form. Wind loads are

defined fo r both mean and peak loads accounting for the protective influence of

upwind c o lle c to rs , wind protective fences, or other wind blockage elements. The

method used to define wind loads was to generalize wind load data obtained during

tests on model c o lle c to rs , heliostats or parabolic dishes, placed in a modeled

atmospheric wind in a boundary-layer wind tunnel at Colorado State University.

For both h e lio s ta ts and parabolic dishes, loads are reported fo r s o lita r y

co lle cto rs and fo r collectors as elements of a f i e l d . All collectors were solid

with n e g lig ib le porosity. Thus the effec ts of porosity in the collectors is not

addressed.

This methodology is based on experimental research performed for Sandia

National Laboratories and the Solar Energy Research In s tit u te (now the National

Renewable Energy Laboratory) over a period of several years, Peterka et a l .

^ Professor, Fluid Mechanics and Wind Engineering Program, Colorado State

U n iv e rsity, Fort C ollins, CO 80523.
Graduate Student, Fluid Mechanics and Wind Engineering Program, Colorado State
U niv e rsity.
(1990, 1988, 1987 and 1986), and e a r l ie r references cited there in . A summary of

these e ffo r ts is provided in Peterka et a l . (1990 and 1989). The measured mean

and peak wind loads have been compiled and presented here in a s im plified design

procedure th a t should provide r e a l i s t i c design guidance fo r the range of module

shapes and f i e l d layouts considered. This methodology represents current

research knowledge which may be modified through future research.

The user of th is design method is encouraged to become fa m ilia r with the

various p r in c ip le s , assumptions, and lim ita tio n s of application outlined herein.

Design parameters fo r heliostats and parabolic dish collectors are

contained in separate tables or sections to minimize confusion. It is not

exhaustive in it s treatment of case scenarios. I t is possible that higher loads

than those outlined herein are possible, although wind-tunnel tests were designed

to id e n t if y the higher load cases. The geometries studied were idealized and did

not have attached to them a ll of the truss structures which might be on a

production u n it. I t is not clear that truss structure loads are simply additive

to loads shown herein. In many cases that assumption w ill be v a lid - - in others,

the truss structure might s ig n ific a n tly modify the wind flow about the c o lle c to r

and thus the loads as calculated in this document (including possible s ig n ific a n t

load decreases).

The design methods outlined in this document are based only on wind-tunnel

te s ts . While wind-tunnel tests have proven highly useful in defining wind loads,

i t must be emphasized that f u l l- s c a l e v a lid atin g experiments would be highly

desirable, in view of the high s e n s it iv it y of loads to turbulence (gustiness in

the wind) observed in the wind-tunnel tests. The procedures proposed herein are

not intended to replace but to supplement existing codes, standards, and other

design methods and procedures.

ATMOSPHERIC WINDS

Discussions of atmospheric wind flow and it s physical model simulation can

be found in Simiu and Scanlan (1986), Counihan (1975), Cermak (1971, 1975), and

Reinhold (e d ., 1982). A b r ie f summary of the important c h a rac te ris tic s fo r use

in th is document are summarized below. Some material is included fo r educational

purposes and to c l a r i f y reasons for the design approach suggested in this

document.

Several types of atmospheric winds can lead to design level winds on a

solar c o lle c to r. Strong low pressure areas can generate design level winds.

These are called e xtratropical lows i f t h e i r orig in is in temperate regions;

t h e ir winds are the normal s traight lin e winds with which we are most fa m ilia r .

Low pressures of tropical origin give r is e to tropical cyclones which are

indistinguishable from e x tra -tro p ic a l lows unless they become strong enough to

be c la s s ifie d as tro p ica l storms or hurricanes. Tornados are concentrated

ro ta tio n al winds with a r e la t iv e ly small spatial extent of high speed; the weaker

ones could be resisted by a solar c o lle c to r, but more intense versions are not

w ithin the normal design range of solar c o lle c to rs . Dust devils are also

ro ta tio n al winds s im ila r to a tornado, but with lower speeds, which ty p ic a lly

o rig in a te with a c le a r sky as opposed to a storm orig in for tornadoes. Downslope

wind storms occur in local areas in the lee of some mountain ranges and can reach

speeds above hurricane magnitude.

Winds of in te re s t for solar c o lle c to r design have some common

c h a ra c te ris tic s near the ground. Mean (time averaged) wind speed increases with

height above ground, and winds have gusts of short duration which are

s ig n if ic a n t ly higher than the mean and which define peak wind loads on solar

c o lle c to rs . Both the variation of speed with height and the gustiness,

frequently called turbulence, must be accounted fo r in the design of solar

c o lle c to rs . For example, wind load coe ffic ie n ts determined by wind tunnel in a

flow without v e rtic a l v a ria tio n of speed and without s ig n ific a n t turbulence, such

as ASCE (1961), are not appropriate for design. The q u a n tita tiv e impact of

proper modeling w ill be shown in the section on atmospheric turbulence.

Design of solar collectors fo r wind outlined in th is document assumes a

boundary layer structure to the wind; the classical example is the s tra ig h t lin e

winds of the e x tra -tro p ic a l low or tropical cyclone. The structure of the wind

fo r other types of high wind event is not as well known, but is assumed to be

s im ila r enough to the boundary layer structure fo r design purposes. Boundary

layer wind tunnels capable of simulating the structure of boundary layer winds

are av aila b le fo r modeling wind loads on structures, and have been used to define

wind loads on various solar c o lle c to r shapes. The results of several research

te s t series were used in development of this design methodology.

In an atmospheric boundary layer wind, the v a riatio n of wind speed with

height is frequently represented by a power law

U (Z )/U (Z re f) = (Z /Z re f)"

where

U(Z) = mean v e lo c ity at height Z;

u(Zre f) = mean ve lo c ity at reference height Zref; and

n = power law exponent, a measure of ground roughness; 0.13 to 0.15

in open country.

An a lte rn a te expression which works well is the logarithmic law

U(Z)/U* = (1 /k ) ln ( Z /Z J .

A form of th is equation fo r tran sferring mean v e lo c itie s between heights is

U (Z )/U (Z re f) = l n ( Z / Z J / l n ( Z r e f / Z J

where

U* = the shear v e lo c ity ;

 5

k = von Karman constant, 0.4;

In = natural log function; and

Zq = e ffe c tiv e roughness length, another measure of ground roughness; 0.01

to 0.05 meters in open country.

An expression fo r turbulence is (Simiu and Scanlan, 1986)

Tu = (B)(U*)/U(Z)

where

B = a constant, approximately 2.5 for open country.

Examples of ground surface roughness c h aracteristics are included as Table 1.

This ta b le is fo r informational purposes; i t is not e x p l i c i t l y used in the design

methodology of th is document.

Gust v e lo c itie s of 2 to 3 second duration can be related to mean v e lo c itie s

at 10 meters height in open country by

Upk(lOm) = (R) U(lOm)

where

Upk(lOm) = 2-3 second gust magnitude at 10 m; and

R = constant, 1.53 to 1.6 in open country.

MEASURED PEAK AND QUASI-STEADY PEAK WIND LOADS

Wind load codes and standards have t r a d it io n a lly assumed a "quasi-steady"

wind load in which a time averaged, or mean, load c o e ffic ie n t is used with a peak

gust wind speed to determine peak wind loads. The "quasi-steady" approach

assumes th a t the short duration peak wind load is e s se n tia lly a "steady" wind

load. I f that assumption were s t r i c t l y true , then we would obtain the same load

by using a mean wind with a measured peak c o e f f ic ie n t . Experiments have shown

th a t the "quasi-steady" assumption works well fo r some cases, fo r example the

 Table 1. Estimated values of the surface roughness.

Repre Turbulence
sentative Terrain n In te n s ity , %
Value of at 10 m*
(m)

0 .5 -1 .5 0.7 Center of large towns, 0.35 34

c i t ie s , forests

Dense forests of 0 .2 7 -0 .3 0 ** 34
r e la t iv e ly non-uniform
height

Dense forests of 0 .2 3 -0 .2 5 ** 34
r e la t iv e ly uniform height

0 .1 5 -0 .5 0.3 Small towns, suburban 0.24 26

area

0 .0 5 -0 .1 5 0.1 Wooded country v illa g e s , 0.20 21

outskirts of small towns,
farmland

0 .015-0.05 0.03 Open country with 0.17 17

isolated trees and
buildings

0.007-0.015 0.01 Grass, very few trees 0.15 14

0.0015-0.007 0.003 RUNWAY AREAS (Average) 0.13 13

Surface covered with
snow, rough sea in storm

<0.0015 0.001 Calm open sea, lakes, 0.11 11

snow covered f l a t
t e r r a in . Flat desert

*Turbulence in te n s itie s calculated from information in Simiu et a l .

(1986)
Zg = e ffe c t iv e surface roughness
n = power law exponent fo r mean v elocity v ariation with elevation

**A11 roughness entries in table except these are from ESDU (1982)
peak force where mean force is large. The peak force is e s s e n tia lly the gust

fa c to r in wind, squared, times the mean force.

For other cases the "quasi-steady" assumption does not work well at a l l .

For example, the mean moment about the elevation axis fo r a f l a t h e lio s ta t in

stow position (horizontal orientation with minimum area exposed to wind) is quite

low. The flu c tu a tin g part of the moment leading to the peak value o s c illa te s

about the near-zero mean and is due in large part to v e rtic a l components of the

turbulent approach flow. Thus the peak moment is determined mainly by the

flu c tu a tin g part of the moment and is not predicted well by a gust factor

(representing the square of the r a tio of actual wind gust to mean wind speed)

m u ltip lie d by the mean.

For reasons given above, i t is best to determine peak loads by using a

measured peak c o e ffic ie n t in combination with the mean wind. Load coeffic ie n ts

are thus defined using the dynamic pressure, Q, of the mean wind speed. Since

both mean and peak wind loads may be of in te r e s t, this document includes both

mean and peak c o e ffic ie n ts .

SPECIFICATION OF DESIGN WIND SPEEDS

Since wind speed varies with height above ground, with gust duration and

with upwind surface ground roughness, i t is important that the wind speeds used

fo r c o lle c to r design be specified for height, duration and exposure. A clear

s p e c ific a tio n of required wind speed might be: X miles per hour mean hourly wind

at 10 meters above ground in open country, or a peak gust of Y miles per hour at

10 meters height in open country. A peak gust t y p ic a lly means the highest point

traced on an anemometer recording chart which is about a 2 to 3 second duration

gust. For a s tra ig h t lin e boundary layer wind, the peak gust is about 1.5 times

the mean hourly wind based on H o llis te r (1970) and about 1.53 times the hourly
 8

speed based on ANSI/ASCE 7-88 (1988). Conversion of mean wind from one height

to another is performed using the wind p r o f i le equations li s t e d above. Peak

winds follow a d if fe r e n t p r o f ile .

Converting winds from one exposure to another is not completely

straightforw ard. If the upwind ground roughness is constant fo r 5 to 10

kilometers, then the conversion can be accomplished by

Ur = Uo (Zgo/Z10)" (ZlO/Zgr)"'

where

Ur = mean speed at 10 meters in roughness r exposure;

Uo = mean speed at 10 meters in open country exposure;

Zgo = e ffe c tiv e gradient height in open country exposure;

Zgr = e ffe c t iv e gradient height in roughness r exposure;

ZIO = 10 meters, standard specification height for wind;

no = power law exponent for open country, 0.143; and

nr = power law exponent for roughness r exposure.

Consistent pairs for Zgr and nr (Zgo and no fo r open country) taken from

ANSI/ASCE 7-88 are:

Exposure Category Z meters nr

A - Large City Centers 450 0,.333

B - Urban and Suburban Areas 350 0..222

C - Open Country 275 0..143

D - F la t Unobstructed Coastal Areas 210 0,.100

It is recommended herein that collectors be designed fo r an open country

environment and that design wind requirements be specified also at 10 meters in

open country environment. The primary reason to change the exposure would be for

co lle cto rs located on the edge of a large body of water or on exposed te rr a in

( h i llt o p s , ridges) where wind speeds might be higher than inopen country.
H illt o p exposures have special problems (not addressed herein) in defining wind

speeds due to accelerated flows.

PRINCIPLES OF WIND LOADING

Structural f a ilu r e from wind loads can be due to d if fe r e n t mechanisms. One

type is overstressing in which the peak stresses induced by the near s ta tic wind

loads exceed the material capacity. Measurement of peak loads in the studies

leading to th is method provide a method fo r design against over-stressing

fa ilu re . Fatigue f a ilu r e can be caused by dynamic, repeated loading at stresses

less than the s ta tic , allowable material capacity. Fatigue design is not

addressed in th is methodology. Fatigue is not e x p l i c i t l y addressed in Peterka

et a l . (1990 and 1989), but the range of flu c tu a tin g load magnitudes fo r many

c ritic a l situations is shown in figures in these two references. The load

spectra, or frequency d is trib u tio n of loads, was not measured.

Mean loads are defined as an average over a period ranging from 10 minutes

to an hour. Values of these loads are reported herein. However, i t is the peak

flu c tu a tin g loads that provide the stresses fo r design. Peak forces may be a

re s u lt of peak, n e a r-s ta tic applied wind loads, or may be due to s t a tic or

dynamic wind loads augmented by resonant vibrations in the structure. Resonant

additions to the applied wind loads of th is design method may be s ig n ific a n t i f

structural damping is low. The research studies leading to this method included

only determination of peak applied wind loads and did not include the prediction

of resonant e ffe c t s . The importance of resonant e ffe c ts is not c le a r, since

aerodynamic damping due to c ollecto r motion (resistance to motion caused by

v e lo c ity of the c o lle c to r through the a ir ) may l i m i t resonance. As the natural

frequency of c o lle c to r increases, any resonant response w i ll decrease.

 10

With respect to resonant loads, one reviewer of this document made the

following comment which was s u ffic ie n tly s ig n ific a n t that i t is included here

verbatim: "I must highly emphasize to designers of large pedestal supported

arrays that resonant vibrations must be considered. In our e a r l i e r designs, we

experienced several drive fa ilu re s in the f i e l d when only s t a tic wind loads were

considered in the design. Naturally we were forced to consider dynamic e ffects

to correct the problems."

COORDINATE SYSTEM OF COLLECTORS - Forces and Moments

The coordinate systems fo r a h e lio s ta t or parabolic dish c o lle c to r are the

same. Forces or moments fo r e ith e r an isolated or i n - f i e l d c o lle c to r are based

on the same set of equations. In the equations, the values of the various

c o e ffic ie n ts d i f f e r , depending on a p a rtic u la r isolated or i n - f i e l d c o lle c to r

configuration. P a rtic u la r c o e ffic ie n t values fo r isolated or i n - f i e l d h e liostats

and parabolic dishes are presented la te r .

Based on the coordinate system shown in Figure 1, the defining set of

equations are as follows (forces in lb, moments in I b - f t ) :

Drag Force: ^ Q ^

L i f t Force: F^ = * Q* A (2)

Base Overturning Moment: * Q* A * H (3)

Hinge Moment: * Q* A * h (4)

Azimuthal Moment: MZ * Q* A * h (5)

C is obtained from C = Cp + * (h/H) (6)

y y ^ y
 11

A xis a t Hinge
X Axis at Hinge

(Azimuth Angle)

Cp z:::^

Y Axis at Base, X Axis at Base,, x

Z A xis, z

Figure 1. Coordinate system fo r colle cto r (h e lio s ta t or parabolic dish ).

where

= dynamic pressure of the mean approach wind (psf)

= 0.00255 * with U in mph (0.00256 includes a i r

density at sea level at standard conditions);

H
= mean approach wind at elevation H = Umean
^wind

= the power law exponent fo r the approaching wind;

mean = mean approach wind at elevation = U^astest^niie/*^

mean gust'

gust = 2-3 second gust magnitude at height Zy^^ above ground;

ufastest mile = fastest mile wind speed at height above ground;

 12

R = 1.30 i f fa s te s t mile wind is 90 mph obtai ned from

Hon is t e r
= 1.24 i f fastest mile wind is 60 mph (1970)

R has no units;

A = solid surface area of h e lio s ta t (include openings in

solid area i f they constitute less than 15 percent of

the to ta l area) ( f t ^ ) ;

h = chord length of the h e lio s ta t ( f t ) ;

H = height of the center of the h e lio s ta t area from ground

( f t ) ; and

Cpj^,CFj,Cy,etc. = the respective force and moment c o e ffic ie n ts (no u n its ).

* indicates m u ltip lic a tio n

Note: The approach flow is identical fo r the isolated or

i n - f i e l d cases, i . e . outside the f i e l d .

The equation fo r U assumes a wind with a r e l a t i v e l y stable mean value and

d ire c tio n over at least a 10-minute period. Thunderstorm outflows of short

duration, dust d e v ils , or tornados may not have the same gust-to-mean-wind r a t i o .

However, i f the mean wind is based on Ug^jg^l.S using the above equations fo r

these wind events, then the provisions of th is document might reasonably apply.

No c o e ffic ie n ts are included fo r y-axis forces or fo r x-axis moments. In

some wind-tunnel tests an attempt was made to measure these components fo r f l a t

p lates, but the forces and moments were s u f f ic ie n t l y low that r e lia b le

measurements were not obtained. Since those components were never a requirement

fo r any of the wind-tunnel te s ts , extra e ffo r ts were not made to obtain these

values. For parabolic c o lle c to rs , these forces and moments can be obtained from

x-axis forces and y-a xis moments by a suitable coordinate transformation.

 13

The wind-tunnel models were mostly ideal shapes which did not include

supporting trusses. The forces on the supporting trusses may, in many cases, be

la rg e r than the y -a xis forces on the collecto rs.

A ll tested h e lio s ta ts were nearly square in shape or were round f l a t discs.

Both f l a t shapes had e s s e n tia lly the same load c o e ffic ie n ts . The influence of

other shapes or aspect ra tio s is not known from tests leading to th is document.

The s p e c ific location of the hinge point about which hinge moments are

defined is centered on the c o lle c to r geometry and is 0.062 h from the rear

surface (downwind side when beta = 0 and alpha = 90), where h is the h e lio s ta t

chord or parabolic c o lle c to r diameter.

EFFECTS OF ATMOSPHERIC TURBULENCE

The wind load data in this document were obtained in a boundary-layer wind

tunnel in which the mean v e lo c ity and turbulence in te n s ity v a ria tio n with height

were modeled. As part of the research studies, the e ffe c t of varying turbulence

in te n s ity was studied. A s ig n ific a n t e ffe c t of turbulence in te n s ity was found.

Figures 2 and 3 show th is v a r i a b i l i t y . Figure 2, taken from Peterka et a l .

(1989), shows that load c o e ffic ie n ts increase dram atically fo r turbulence

in te n s itie s above 10 percent. Turbulence in te n s ity fo r an open country

environment on these graphs is about 15 to 18 percent. Collectors designed fo r

a rougher environment than open country should include the increase in load

c o e ffic ie n t from turbulence. Table 1 provides some guidance in the level of

turbulence in te n s ity expected in various exposures. Note that some of the

increase in load from turbulence w ill be removed by lower mean wind speeds in the

rougher exposures. Higher turbulence does not always mean higher loads. For

example, in a dense f i e l d of collectors where mean v e lo c itie s are very low and

turbulence in te n s itie s are high, there is a net decrease in peak wind loads.
 14

2.2 " I I I r I 1 1 1 1 ------------ 1 I i i...r 1 1 1 1 T I ITTTTT

cO B earm an ( 1 9 7 1 ) o ''
CD C erm ak, et. ol. ( 1978)

s 2.0 P eterka, et. al. (1 9 8 6 )0

P eterka. et. al. ( 1987) a

N C urrent Study
c
1.8 C urrent Study t Cp^ CO
o Current S t u d y Cj^^
(D

a 1.6
cO
CD
1.4
E CJ

X 1.2
(4
o
1.0
10 10 ' 10
T urbulence In te n s ity , %

6. 1 1 I I 111 1 1 1 I 1 1II .8
_ C erm ak, ei. al. ( 1978)
cO - Peterka, et. al. (1 9 8 6 ) O -

CD "C p,
n p eterka , et. al. (1 9 8 7 ) a
a 5. - Current Study v^
J,

I
N C urrent Study t
4^
j L
- C m z
cO
a
4. Current Study C W
CD
- - a
- N
"S 3. C r x - ^ / F

CD - / / T .4 o
a. -

^Fz \ 7
X 2.

o
1. 1 1 ................ 1 1 I I 1 1 1111 1 1 1 1 1 111 p

10 10 10 10
07
T u r b u le n c e In t e n s i t y , /O
Figure 2. V a ria tio n o f Cp^ and w ith turbulence in te n s ity f o r an
is o la te d h e lio s ta t.
 2 .4 T T
CMy = My/ qAH
- - - ( 18%)

2.0 -

A
-(I4 % )0
1.6 -

o (12%)
c A o
o o V
4>
A
E
O
2 O
I- '-2 O
o
A KEY
A
^ J ^ O ^ V A T u = 18% (P e terka , I 9 8 7 ^ \ ,
0.8 . o
A Tu=l4 % (Peterka, 1986)
o Tu=l2 % (C e rm a k, 1978)
0 .4 m Tu=l.2 % ( C e rm a k , 1 9 7 8 )
^ -------------- -------------- T u 0 (Peglow, 1979)
y -------------------- -------------- A S C E ( 1 9 6 1 )_______________
0.0

0 20 40 60 80 100 120 140 160 180

a
Figure 3. V a ria tio n of mean Cy with elevation angle and turbulence in ten sity.
 16

Figure 3 is also taken from Peterka et a l . (1989). I t shows the variation

of base moment Cy with elevation angle alpha. The solid lin e represents ASCE

(1961) which was based on wind-tunnel data obtained with near-zero turbulence

in te n s ity and no mean v e lo c ity variatio n with height. The dashed lin e in the

fig u re is data obtained by Peglow (1979) on a f u l l- s c a l e 20 x 20 foot h e lio s ta t

in s ta lle d in the 40 x 80 foot wind tunnel at NASA Ames Research Center. This

wind-tunnel te s t also had a near-zero turbulence in te n s ity . Comparison to the

ASCE (1961) data is e x cellent. The solid squares in Figure 3 show base moment

c o e ffic ie n ts on a Colorado State University h e lio s ta t model placed in a 1.2

percent turbulence in te n s ity in a boundary-layer tunnel in which the atmospheric

simulation was removed from the tunnel. The comparison with both 1961 ASCE and

Peglow data is good - the CSU measurements are s lig h tly higher than the two near

zero turbulence cases due to the low, but non-zero, level of turbulence in the

CSU tunnel. Three other data sets in Figure 3 show the influence of increasing

turbulence in a CSU boundary-layer simulation (which also included the variation

of mean v e lo c ity with height) on base moment. The v a riatio n of load c o e ffic ie n t

with turbulence was not an e ffe c t of the mean v e lo c ity v a riatio n with

height since th is v a riatio n was the same in the 12, 14 and 18 percent turbulence

cases.

The conclusion based on these data is that atmospheric turbulence

s ig n if ic a n t ly increases wind loads on solar collectors and cannot be ignored.

Data based on low-turbulence te s ts , such as those ty p ic a lly obtained in

aeronautical-type wind tunnels and in ASCE (1961), is not appropriate fo r solar

c o lle c to r design.
 17

LARGEST WIND LOADS ON ISOLATED COLLECTORS - H e lio s ta ts and P a ra |)c lic Dishes

Table 2 presents the largest mean and peak wind force and moment

c o e ffic ie n ts as a function of wind d ire c tio n , and c o lle c to r t i l t , a, fo r a

i
f l a t (or nearly f l a t ) rectangular or c ir c u la r h e lio s ta t alone in an open-country

atmospheric wind. Table 3 presents the same information f()r an isolated

parabolic dish c o lle c to r. Note that the maximum drag force and maximum base

overturning moment occur at the same a and ^ as might be expectfed for both the

h e lio s ta t and the parabolic dish. Less i n t u i t i v e l y , the maximutij l i f t force and

maximum hinge moment occur at the same a and fo r the h e lio s ta t. Stow loads

have been included in Tables 2 and 3. Stow load c o e ffic ie n ts are small but are

usually used with higher survival wind speeds. The to ta l vector force acting on

a h e lio s ta t acts almost p a ra lle l to the normal to the c o lle c to r surface. This

indicates a small component of force in the plane of the c o lle c tp r (wind-tunnel

I

colle cto rs had minimal truss supporting s tru c tu re ). Where MH^ is largest, the

h e lio s ta t normal force is non-uniform with largest (but unknown) value near the

upwind edge.

LOAD COMBINATIONS FOR ISOLATED HELIOSTATS AND PARABOLIC DISHES

Load combinations at conditions other than those where one component is a

maximum (which are tabulated in the previous section) are usei^ul for design.

Mean and peak load c o e ffic ie n ts for heliostats for a range of elevation angle a

and wind d ire c tio n ^ are shown in Figures 4 to 7. S im ilar c(|efficien ts for
i
parabolic dishes with depth-to-diameter r a tio of 0.1 ( f/D = 0.62^) are shown in

Figures 8 to 15.

Some guidance is helpful in the use of Figures 4 to 15:

1. Uncertainty in portions of graphs is s ig n if ic a n t ly hijgher than for

the data in Tables 2 and 3. The number of experiments in most

 18

Table 2. HELIOSTATS - - Wind load c o e ffic ie n ts on isolated h e lio s ta ts when one

component is a maximum.

Fx, My Fz, MHy Mz Stow

Max Max Max Loads

PEAK LOADS
a, degrees 90 30 90 0
degrees 0 0 65 0
Fx 4.0 2.1 3.7 0.6
Fz 1.0 2.8 0.5 0.9
MHy 0.25 0.5 0.15 0.2
Mz 0.29 0.06 0.7 0.02

MEAN LOADS
a, degrees 90 30 90 0
degrees 0 0 65 0
Fx 2.0 1.0 1.6 0.1
Fz 0.3 1.35 0.3 0.1
MHy 0.02 0.25 0.02 0.02
Mz 0 0 0.25 0

n = 0.15 Tu = 18% Zo = 0.03 meters

e 3. PARABOLIC DISHES - - Wind load c o e ffic ie n ts on isolated parabol

dishes when one component is a maximum.

Fx Fz MHy Mz Stow
Max Max Max Max Loads

PEAK LOADS
a, degrees 90 30 60 90 0
degrees 0 0 180 60 0
Fx 3.5 1.9 -1 .8 3.03 0.33
Fz 0.31 3.1 0.8 0.33 0.98
MHy 0.31 0.25 0.35 0.35 0.22
Mz 0.15 0.06 .08 0.35 0.02

MEAN LOADS
a, degrees 90 30 60 90 0
/3, degrees 0 0 180 60 0
Fx 1.75 1.1 -1.0 1.32 0.13
Fz 0.17 1.7 0.5 0.09 0.1
MHy 0.11 0.13 -0.17 0.14 0.09
Mz 0 0 0 0.13 0

n = 0.15 Tu = 18% Zo = 0 . 0 3 meters

Note: A ll data in Table 3 is fo r a parabolic dish shape with depth

diameter r a tio of 0.1 (f/D = 0 .6 2 5 ).
 19

'Fx
90
2.0
80

70

60

50
a MEAN
40

30
0.5
20

0.8

20

30
2.5
40
a PEAK
3.0
50

60
3.5
70
.3.8
80 3.9
.4.0
90,

/3

Figure 4. Drag force c o e ffic ie n ts f o r the h e lio s ta t

 20

90

80

70 0.5

60

50
a MEAN
40
11.35
30

20

2.5
a PEAK
2.0

60

70

80

90

Figure 5. L i f t force c o e ffic ie n ts f o r the h e lio s ta t

 21

Mz
90
0.25
80

70

60
0.2
50
a MEAN
40 0.15

30 0.10

20

20
0.2
30

40
a 0.4. PEAK
50
0.6
60

70

80
0.7
90

Figure 6. Azimuth moment c o e ffic ie n ts fo r the h e lio s ta t

 22

MHy
90

80 0.05
70 0.10

60 0.15

50
a MEAN
40
0.20
30
0.25
20

0.60
30

40
a PEAK
50 0 .5 0

60
0.40
70

80 0.30

90

Figure 7. Hinge moment c o e ffic ie n ts fo r the h e lio s ta t

 23

MEAN C
90
1.75
80

70

60

50 FRONT
a
40

30
0.5
20
0.25

-0 .2 5
20

30 rO.5
40 BACK
a - 0.8
50
- 1.0
60

70

80

90
180 170 160 150 140 130 120 110 100 90

Figure 8. Mean drag force c o e ffic ie n ts MEAN Cp^ f o r the parabolic dish,
 24

PEAK Cp^

FRONT
a

10 20 30 40 50 60 70 80

BACK
a

^?B0 170 160 150 140 130 120 110 100 90

/3

Figure 9. Peak drag force c o e ffic ie n ts PEAK C . f o r the parabolic dish

 25

MEAN Cp2
90

80
^ 0.2
70

60
d .o
50 FRONT
a
40

30 >

20

20 0 .2
vO.4
30 0.5
0.6
40 0.65 BACK
a
50

60

70

80

90
180 170 160 150 140 130 120 110 100 90

Figure 10. Mean l i f t force c o e ffic ie n ts MEAN C. f o r the parabolic dish

 26

PEAK CFz
90

80
^0.75
70 d .o

60

50 .- 2.0
FRONT
a
c2.5
40
; i 2.8

20

20

30
) 1.2
40 BACK
a
50

60

70 0.75

80

90
180 170 160 150 140 130 120 110 100 90
/3

Figure 11. Peak l i f t force c o e ffic ie n ts PEAK f o r the parabolic dish.

 27

mean
90
0.13
80
0.07 -0.05
70
0.03
60

50 FRONT
a
40

30

20

- 0.01
- 0.02
20
-0 .0 4
30

40 BACK
a -0.05,
50
-0.0 7
60

70 -0.08.

80
,- 0.12
90
180 170 160 150 140 130 120 110 100 90

Figure 12. Mean azimuth moment c o e ffic ie n ts MEAN fo r the parabolic dish
 28

PEAK C Mz
90
0.35
80

70 0.25
0.20
i0.30
60
0.15
50 FRONT
a
40

30
0.05
20

20
-0.05,
30
-O .IQ
40 BACK
a
50

60 - 0.20

70 -0 .2 5
-0.30;
80
-0.35
90 _Z=tx_
180 170 160 150 140 130 120 110 100 90

Figure 13. Peak azimuth moment c o e ffic ie n ts PEAK C f o r the parabolic dish
 29

MEAN C M H y
90

0.12 0.13 \
80

70
0.10
60

50 0.05 FRONT
a
40
\
0.02
30 r o F ) - 0 - 0 ^ ^ )-0.02
20

02
20

30 0.14 .05
40 .08
BACK
a
50
0.17
60

70

80

90
180 170 160 150 140 130 120 110 100 90

Figure 14. Mean hinge moment c o e ffic ie n ts MEAN f o r the parabolic dish.
 30

PEAK C MHy
90
0.35
80
0.30
70 -
0.25
60
<0.25 > -0 .2 5
50 FRONT
a
40 -0 .2 5
-0 .3 0
30
-0 .3 5
20

10 -

20

30

40 0 .3 5 0.30 BACK
a 0.20 0.10
50

60

70

80

90
180 170 160 150 140 130 120 110 100 90

Figure 15. Peak hinge moment c o e ffic ie n ts PEAK f o r the parabolic dish
 31

low-coefficient-magnitude areas of the graphs is not as large as those

near maxima where the wind-tunnel data was concentrated. The range of

uncertainty is d i f f i c u l t to quantify, since no f i e l d measurements are

a v a ila b le fo r comparison and since the s e n s it iv ity to wind turbulence

is high as shown in Figure 2. Parabolic dish loads have a smaller

uncertainty than h e lio s ta ts . As a rough guide h e lio s ta t (parabolic

dish) c o e ffic ie n ts probably have an uncertainty of 15-20 (10-15)

percent of the largest values in each graph.

2. The h e lio s ta t loads of Figures 4 to 7 for 90 < < 180 were folded onto

the 0 - 90 degree range since the loads are nearly symmetric about

90 degrees when the h e lio s ta t has no supporting truss work. For this

reason, the load c o e ffic ie n ts do not have signs. To determine signs,

r e fe r r in g to Figure 1, the sign of Fx is po sitive fo r ^ < 90 and

negative fo r $ > 90. Fz is negative when the wind impinges on the

upward surface, positive when the wind hits the lower side. The sign

of a ll h e lio s ta t moments is such that the wind action tr ie s to force

the c o lle c to r toward a maximum drag orientation with the wind impinging

d i r e c t ly on it s face.

3. The parabolic dish loads of Figures 8 to 15 are presented separately

fo r ^ < 90 (ca lle d FRONT in the figures) and fo r ^ > 90 (called BACK).

C o e ffic ie n t signs are placed d ir e c t ly on the graphs. For some graphs

( f o r example. Figure 15), the sign of the peak changes suddenly from

p o s itiv e to negative. This occurs when the absolute value of the

negative peak moment becomes larger than the magnitude of the positive

peak load.
 32

LOCAL PRESSURE DISTRIBUTION

The d is tr ib u tio n of local mean or peak pressures over the surface of

colle cto rs cannot be determined unambiguously from the to ta l forces and moments

presented in previous sections. There is an in fin ite set of pressure

d is trib u tio n s which can be integrated over the c o lle c to r surface to obtain the

recorded forces and moments. The research studies leading to th is design method

were directed toward integrated loads, and so pressure d is trib u tio n s were

measured in only one study involving parabolic dishes. In s u ffic ie n t local

pressure d is trib u tio n s on heliostats have been measured fo r general use.

Figures 16 to 23, taken from Peterka et a l . (1990), show mean pressure

d is trib u tio n s on an isolated solid parabolic c o lle c to r at a va riety of

o rie n ta tio n s . The defining equation fo r mean pressure in terms of the pressure

c o e ffic ie n ts in the figures are:

Mean Pressure: P = Cp * Q. (7)

In each fig u re , the concave face is displayed above the convex face.

Contours on the concave face correspond to the viewer looking d ir e c t ly at that

face, while the contours on the convex face are viewed from the concave side

through the c o lle c to r as i f the concave face were in v is ib le . Thus the viewer can

re a d ily make a correspondence between values on the opposing surfaces of the

c o lle c to r. Since the pressure taps were not placed on the c o lle c to r edge, the

inner ring in the figures corresponds to the maximum radius of tap placement.

The space between the inner and outer rings in the figures represents the area

on the c o lle c to r surface fo r which pressures were not measured.

Peak pressures were measured on the parabolic dish co lle cto rs but have not

been presented in plots in Peterka et a l. (1990). The largest values of

peak pressure c o e ffic ie n t were +2.6 and - 5 .9 . These occurred fo r the same
 33

Q/

(a) concave face (windward)

- 0 .8 5
m

(b) convex face (leeward)

Figure 16. Mean pressure c o e ffic ie n ts { a = 90 and ^ = 0)

 34

WIND

(a) concave face (windward)

WIND
c=C>

(b) convex face (leeward)

Figure 17. Mean pressure c o e ffic ie n ts (a = 90 and ^ = 60)

 35

WIND ro

c=C>

(a) concave face

WIND

(b) convex face

Figure 18. Mean pressure c o e ffic ie n ts { a = 90 and $ = 90)

 36

- 0.6

-0 .6 5

(a) concave face (leeward)

(b) convex face (windward)

Figure 19. Mean pressure c o e ffic ie n ts (a = 90 and = 180)

 37

(a) concave face (windward)

O'

(b) convex face (leeward)

Figure 20. Mean pressure c o e ffic ie n ts (a = 60 and = 0)

 38

WIND
'= = [>

(a) concave face (windward)

WIND
<=c>

(b) convex face (leeward)

Figure 21. Mean pressure c o e ffic ie n ts (a = 60 and ^3 = 60)

 39

0/0

WIND
f= 0

(a) concave face

WIND rpna

in

(b) convex face

Figure 22. Mean pressure c o e ffic ie n ts (a = 60 and = 90)

 40

(a) concave face (leeward)

(b) convex face (windward)

Figure 23. Mean pressure c o e ffic ie n ts (a = 60 and ^ = 180)

 41

orientatio ns which gave the largest mean c o e ffic ie n ts (shown in Figures 16 to 23)

of +1.1 and - 2 .1 . Peak pressure differences across the c o lle c to r must be

measured d i r e c t ly to be v a lid ; these were not part of that research program.

MAXIMUM WIND LOADS ON COLLECTORS IN A FIELD - General Blockage Area

This section may be used when calculating maximum loading on h e liostats and

parabolic co lle cto rs as part of a f i e l d fo r a given load component. For loads

not at t h e i r maximum, in s u f f ic ie n t data is available to r e lia b ly use th is section

and the following section should be used.

A convenient parameter c orrelating to the wind loading of an i n - f i e l d

c o lle c to r is it s Generalized Blockage Area or GBA that is defined as

Solid area of upwind blockage projected to wind direction

boA == ' .
Area of ground occupied by the blockage objects ^3 ^

The GBA includes the e ffec ts of wind fences and upwind c o lle cto rs .

The c alculatio n of the GBA can be a cumbersome task using the above general

d e fin it io n since the wind d ire c tio n , /3, r e la tiv e to the c o lle c to r, the c o lle cto r

tilt, Qt, and the angle of the wind r e la t iv e to a wind fence must be taken into

consideration in determining the area projections of the various blockage

elements upwind of a p a r tic u la r i n - f i e l d c o lle c to r u n it. Therefore a sim plified

method fo r GBA calculation has been developed that uses a fa c to r, K, to

correspond to the maximum loads fo r each component of loading. We redefine the

GBA as

(K)(AH) + AS
GBA =
AF (9)
 42

where

AH is the actual surface area of the collectors (chord times width fo r a

rectangular u n it; Trh^ fo r a parabolic u n it) in the AF;

AF is the representative ground area occupied by the c o lle cto rs ;

AS is the solid area of fences (external and in te rn a l) w ithin AF; and

K is a fa c to r to account fo r elevation and azimuth angle of c o lle c to r fo r

maximum loading fo r a given force or moment component;

in which the fa c to r K has the following values:

Comoonent K

Fx 1.0

Fz 0.5

Mhv 0.5

Mz 0.5

Figure 24 il lu s t r a t e s the layout fo r a s im plified GBA calculation fo r fiv e

unit lo catio ns, rows 1 through 5. I t is not necessary fo r the user to know the

wind d ire c tio n or c o lle c to r t i l t in performing the sim plified GBA c a lc u la tio n ,

because th a t knowledge is im p lic it in the fa c to r K.

The treatment of c o lle c to r units in rows 1 through 4 should be s e lf-e v id e n t

from the fig u r e . Collectors denoted by a square, solid symbol within each

representative f i e l d ground area are included in the calculation of AH in the

s im p lifie d GBA equation, Eq. (9 ). Note th a t a unit in the f i f t h row is treated

d i f f e r e n t l y from the other cases. By the f i f t h row, an external wind fence is

no longer an e f fe c t iv e blockage element (by the fourth row fo r a denser f i e l d of

c o lle c t o r s ) . Therefore the representative f i e l d ground area, AF, fo r a c o lle c to r

in the f i f t h row is reduced as shown in Figure 24.

 43

Row , External
Fence

-Internal
Fence

!
iF",

==d-

^

H e lio s ta t or P a ra b o lic D is h C o lle c to r

U n d e r C o n s id e ra tio n

J AF R e p re s e n ta tiv e F ie ld G r o u n d A rea

AH S o lid H e l i o s t a t A r e a in A F

AS S o lid A rea of F e n c e in A F

Figure 24. Layout fo r s im plified GBA calculations fo r c o lle c to r units in rows

1 through 5.

Field density is an issue of s ig n ific a n t importance in designing h e liostats

and parabolic dishes. Research has shown two separate regimes of i n - f i e l d unit

performance fo r f i e l d densities delimited by a GBA value of 0.15 to 0.2.

(A) I f the GBA < 0.15 to 0 .2 , wind loads on i n - f i e l d units ac tu a lly

increase in the f i r s t two or three rows from the edge of the fie ld .

Percentage increases are la rg e r fo r peak loads than fo r mean loads.

External wind fences w ill abate th is e ffe c t s ig n if ic a n t ly . I f the

f i e l d is quite sparse ( i . e . GBA < 0.1 - 0 .1 5 ), internal wind fences

may be used to increase GBA and lower wind loads on units in t e r io r

to the f i e l d .
 44

(B) I f the GBA > 0.15 to 0 .2 , wind loads on i n - f i e l d units w i l l decrease

with row placement from the perimeter. However, an external fence

w i ll s till have a s ig n ific a n t beneficial e ffe c t on reducing wind

loads on the perimeter collectors and on those up to three rows into

the f i e l d . By the fourth row, unit performance is the same with or

without an external fence. Internal fences have n e g lig ib le e ffe c t

and therefore are not recommended fo r in -fie ld cases with the

GBA > 0 .2 .

Note: I f a fence is to be used in the calculation of GBA, c r e d it fo r

blockage area can only be obtained fo r wind directions w ithin 60

degrees of perpendicular to the fence. Wind fences should be no

more than 30-50 percent porosity and at least 1.15 to 1.3 H t a l l fo r

maximum beneficial shielding. These conditions must be met fo r

proper use of th is data.

The following special conditions hold fo r calculating the GBA with Eq. (9 ):

(a) GBA= 0.01 for row 1 with no external fence;

(b) GBA= 0 . 0 2 fo r row 2 with no external fence;

(c) calculatio n fo r rows 6 , 7, etc. are the same as fo r row 5 when the

basic f i e l d density corresponds to a GBA < 0 . 1 5 - 0.2; and

(d) c alculatio n fo r rows 5, 6 , etc. are the same as fo r row 4 when the

basic f i e l d density corresponds to a GBA > 0 . 2 .

The wind load within a f ie ld is calculated using the curves in Figures 25

through 28 fo r helio s ta ts and Figures 29 through 32 fo r parabolic dishes. Curves

are given fo r both mean and peak loads. The calculated GBA is used fo r the

abscissa.
 45

MAXIMUM COEFFICIENTS FOR IN-FIELD HELIOSTATS AND PARABOLIC DISHES

The force and moment c o e ffic ie n ts for a h e lio s ta t placed in an array of

collecto rs are given in Figures 25 through 28. C oefficients are a function of

the Generalized Blockage Area (GBA). The i n - f i e l d parabolic dish c o e ffic ie n ts

are presented in Figures 29 through 32. The thin solid lin e s shown in the

figures represent upper bound values determined by wind-tunnel te s ts . All in

f i e l d cases f e l l on or below these lin e s . The thick v e rtic a l s olid lines insome

graphs show the highest GBA values fo r which data e x is t.

The dotted lines in the figures denote GBA values separating the thin lin e

segments and serve as an aid to the user in calculating the appropriate force or

moment c o e ff ic ie n t. The sloping lines have an associated equation th a t may be

used d i r e c t ly , with the appropriate GBA value. We have included the fig ures,

instead of equations only, to illu s tra te the dramatic reduction in the

c o e ffic ie n ts with increasing GBA values.

In each fig u re , only GBA values in the range of 0 to 0.3 are presented

corresponding to available data. The following special conditions hold:

(a) GBA = 0.01 fo r row 1 with no external fence, and

(b) GBA = 0.02 for row 2 with no external fence.

Note that load c o e ffic ie n ts fo r these f i r s t two rows without external fence may

be la rg e r in magnitude than the c o e ffic ie n ts for an isolated u n it.

It should be noted th a t data for this guideline were obtained in array

fie ld s with a regular pattern as shown in Figure 24. Data were not obtained fo r

other array geometries.

LOAD COMBINATION COEFFICIENTS WITHIN A FIELD

Load combinations within a f i e l d of heliostats or parabolic dishes at con

ditions other than those where one component is a maximum are presented in th is
 Heliostats
Mean Drag Force Coefficient Peak Drag Force Coefficient

5.0 5.0

-13.6'GBA + 6.52
4.0 4.0

cd = 2.98 cd
CD
B CX
3.0 3.0
>< cn

-8.0+GBA + 3.06
2.0 2.0

1.0 1.0

0.01 007

0 0

GBA GBA

Figure 25, Mean and peak drag force c o e f f ic ie n t s , Fx, o f a h e lio s ta t u n it w ith in a f i e l d o f h e lio s ta ts .
 Heliostats
Mean L ift Force Coefficient Peak L ift Force Coefficient

5.0 5.0

3.78
4.0 4.0

CO cO
CD QJ
B CL.
3.0 3.0
tS3
-10.6*GBA + L31
II tl
1.81
2.0 2.0

-6.5*GBA + 1.94
1.0 1.0

0.25
002 026
0.05
0.1 0.2 0.3
GBA GBA

Figure 26. Mean and peak l i f t force c o e ffic ie n t s , Fz, o f a h e lio s ta t u n it w ith in a f i e l d o f h e lio s ta ts .
 Heliostats
Mean Hinge Moment Coefficient Peak Hinge Moment Coefficient

0.8 0,8

0.7 0.7

0.6 0.6
cd
O) 03
cu
B 0,5 Cl, 0,5

re
0.4 0.4 -3.2*GBA + 0.90
II II CO
0.27
0.3 0.3

0.2 y = -1.0'GBA +0.31 0.2

0.15
0.1 0.1
0.05

004 026 009 0 23

0.1 0.2 0.3 0.1 0.2 0.3

GBA GBA

Figure 27. Mean and peak hinge moment c o e f f ic ie n t s , MHy, o f a h e lio s ta t u n it w ith in a f i e l d o f h e lio s ta ts .
 Heliostats
Mean Azimuthal Moment Coefficient Peak Azimuthal Moment Coefficient

0,8 0.8

0.7 0.7

0.6 0.6
CO
CD

s 0.5 fX 0.5 -4.6*GBA +0.92

of
0.4 0.4
VO

0.3 0.25 0.3

0.2 -2.4^GBA + 0.40 0.2

y = 0.15
0.1 01
0.05

0 06 016 0 05 017

0.1 0.2 0.3 0.1 0.2 0.3

GBA GBA

Figure 28. Mean and peak azimuthal moment c o e ffic ie n t s , Mz, o f a h e lio s ta t u n it w ith in a f i e l d o f h e lio s ta ts .
 Parabolic Dishes
Mean Drag Force Coefficient Peak Drag Force Coefficient

5.0 5.0

= 3.74
4,0 4.0
C
cd
CD Cd
CD
E CL.
3.0 3.0
x" >< -10.65*GBA + 4.54
ti-.
1.91 Cn
O
2.0 2.0
-6.23*GBA + 2.10

1.0 1.0

y - 0.30
0.03 0.29 0.75

0.1 0.2 0.3 0.1 0.2 0.3

GBA GBA

Figure 29. Mean and peak drag force c o e f f ic ie n t s , Fx, o f a parabolic dish u n it w ith in a f i e l d o f parabolic
dishes.
 Parabolic Dishes
Mean L ift Force Coefficient Peak L ift Force Coefficient

5.0 5.0

4.0 4.0 3.55

cO
O)
0-.
3.0 3.0 -10.52*GBA + 3.87
CSI t-3

^ 2.0 2.0

1.0 1.0

y 0.03

0 03 003

0.1 0.2 0.3

GBA GBA

Figure 30. Mean and peak l i f t force c o e ffic ie n t s , Fz, of a parabolic dish u n it w ith in a f i e l d o f parabolic
dishes.
 Parabolic Dishes
Mean Hinge Moment Coefficient Peak Hinge Moment Coefficient

0.8 0.8

0.7 0.7

0.6 0.6
a 0.49
cd
CL> cd
B 0.5 cu
CX 0.5
ac :r^
0.4 0.4
cn
ro
0.3 0.3

0.2 0.17 0.2

-0.63^GBA+0.21
0.1 0.1
-y - 0.03
007 0 29

0.1 0.2 0.3 0.1 0.2 0.3

GBA GBA

Figure 31. Mean and peak hinge moment c o e ffic ie n t s , MHy, o f a parabolic dish u n it w ith in a f i e l d o f parabolic
dishes.
 Parabolic Dishes
Mean Azimuthal Moment Coefficient Peak Azimuthal Moment Coefficient

0,8 0.8

0.7 0,7

0.6 0.6
a
CO
<D CU

B 0.5 fX 0.5
0.41

0.4 It
0.4
It CJl
0.88'GBA i 0.42 OJ
0.3 0.3

0.14
0.2 0.2
y = -0.6i*GBA + 0.16
0.1 0.1

0 03 0 21 001

0.1 0.2 0.3 0.1 0.3

GBA GBA

Figure 32. Mean and peak azimuthal moment c o e ffic ie n ts , Mz, o f a parabolic dish u n it w ith in a f i e l d of
parabolic dishes.
 54

section. Because the research leading to th is method was directed at obtaining

maximum loads, there is very l i t t l e data to support the calculation of load

combinations in a f i e l d . What data exists tends to show smaller reductions

within a f i e l d than occurs for the largest values of component loads. For that

reason, what data was available is presented. While the data presented here is

thought to be reasonable and conservative fo r design use, there is no guarantee

th a t the data used to prepare this section found the la rgest load combination

values. The data are presented with the caveat that use of th is data might lead

to under-prediction of some loads. The data leading to the results of th is

section were obtained fo r GBA values near 0 .1 .

Tables 4 and 5 show factors to be used as m u ltip lie rs of c o e ffic ie n ts shown

in Figures 4 to 15. The data are divided into four c la s s ific a tio n s : heliostats

or parabolic dish collectors with and without an external wind fence with

porosity no la rg e r than 30 to 50 percent.

 55

Table 4. M u ltiplying factors fo r isolated load combinations fo r in -fie ld

h e lio s ta t performance.

(a) without external fence

Row 1 2 3 4

p X.peak
0.69 0.69 0.69 0.69
x,mean
1. 0 1 0.76 0.73 0.69

p Z .p eak
0.93 0.95 0.91 0.72
z,mean
1.26 1.15 1.03 0.86

M *
H y.peak^
Hy.mean
- - - -

M
u Z ,p e a k
0.51 0.66 0.70 0.72
M
z,mean
0.68 0.78 0.96 0.95

(b) with external fence

Row 1 2 3 4

pX.peak
0.25 0.56 0.65 0.69
x,mean
0.26 0.48 0.56 0.69

p Z .p e a k
0.32 0.48 0.57 0.72
z,mean
0.32 0.53 0.72 0.86

MH y.peak^*
- - - -
Hy.ntean

Mz,p e a k 0.25 0.54 0.72 0.72

Mz,mean 0.43 0.62 0.98 0.95

* in s u f f ic ie n t data

Note: For cases where a component load is near it s maximum, use

Figures 25 to 28 instead of this ta ble.
 56

Table 5. M ultip ly in g factors for isolated load combinations fo r in -fie ld

parabolic dish performance.

(a) without external fence

Row 1 2 3 4

p 1.07 1.00 0.97 0.93

x .p e a k
1.03 0.81 0.58 0.55
x,mean

c 1.10 1.04 0.88 0.80

z,p e a k
1.06 0.76 0.63 0.61
z,tnean

MHy.peak 1.07 1.07 1.06 1.16

1.07 0.78 0.75 0.72
Hy,mean

Mz.p e a k 1.00 1.38 1.26 1.27

0.94 0.94 0.94 0.94
z,n>ean

(b) with external fence

Row 1 2 3 4

Fx.p e a k 0.45 0.71 0.86 0.93

0.36 0.43 0.54 0.55
x,mean

F 0.48 0.70 0.75 0.80

' z.p e a k
0.44 0.52 0.65 0.61
z,mean

Muy .p e ak 0.60 0.92 1.04 1.16

"H
0.37 0.55 0.65 0.72
Hy.mean

Mz.p e a k 0.46 0.90 1.12 1.27

0.43 0.61 0.96 0.94
z,mean

Note: For cases where a component load is near i t s maximum, use

Figures 29 to 32 instead of this ta b le .
 57

EXAMPLE PROBLEM 1

An example problem is presented fo r an i n - f i e l d h e lio s ta t to demonstrate

the c alculatio n of peak wind forces and moments when a c o lle c to r is embedded

within an array. The same procedure is followed fo r an i n - f i e l d parabolic dish.

For an isolated h e lio s ta t or parabolic dish, Step 1 and Step 3 are omitted and

Table 2 or Table 3 are used d ir e c t ly in Step 4.

PROBLEM: Calculate peak operational drag force, lift force, hinge

moment, and azimuthal moment fo r a 40 x 40 foot square

h e lio s ta t with a center, H, 22 f t above ground. Perform

c alculatio n fo r (a) h e lio s ta t in th ird row and (b) h e lio s ta t

in f i f t h row.

Given: H = 22 f t ; a 25 ft-h ig h external fence with a 50% porosity is

located 80 f t from edge of f i e l d of c o lle c to rs ; no internal

fences; h e lio s ta t spacing is 80 f t along rows (p a r a lle l to

external fence) and 80 f t between rows (perpendicular to

fence). Wind has peak gust speed of 55 mph at 10 m (32.8 f t )

eleva tio n .

Note: Refer to Figures 1 and 24 in doing th is problem. C oefficient

values fo r an i n - f i e l d h e lio s ta t are presented in Figures 25

through 28. If the calculations were to be performed for

an isolated h e lio s ta t, use Table 2 in place of Figures 25

to 28.

Steo 1. GBA Calculations (see Figure 4 and Eqs. (1) through ( 6 ) ) .

(a) H e lio s ta t at the th ird row

AF = (160 + 80) X 80 = 19200 ft^ (includes 2 upwind collectors and

external fence)
 58

AS = 80 X 25 X 50% = 1000 ft^ (external wind fence only)

AH = 2 X 40 X 40 = 3200 ft^ (2 upwind c o lle cto rs included as

shown in Figure 24)

K AH AS
GBA
c
For
c
Fx:
roi, 1-0 ^ 3200 + 1000
GBA = ----------------------------- = 0.219
19200

AF

c 0-5 X 3200 + 1000 -

Fz: GBA = ----------------------------- = 0.135
19200

MHy: GBA = ^ ^ = 0 .1 3 5
19200

0.5 X 3200 + 1000 -

Mz: GBA = ------------------------------ = 0.135
19200

(b) H e lio s ta t at the f i f t h row

AF = 160 X 80 = 12800 ft^ (2 upwind collectors included as shown in

Figure 24)

AS = 0 .0 ft^ (external fence is not e ffe c t iv e at 5th row)

AH = 2 X 40 X 40 = 3200 ft^

K AH AS
GBA
C C 1-0 ^ 3200 + 0.0
For Fx: GBA = = 0.25
12800

AF
 59

Fz: GBA = 1 i 1 2 L ! 2 1 J L ^ = 0.125

12800

MHy: GBA = = 0.125

12800

Mz: GBA = M J L ! 2 2 J lM = 0.125

12800

Step 2 . Dynamic pressure calculation

Umean at 10 IH = U g^ ^l.S = 55/1.6 = 34.4 mph

H = 22 f t , Z,.^ = 10 m = 32.8 f t

Open country te r r a in , n = 0.14

UnanatH = 34.4 ^ = 3 2 . 5 mph

Q = 0.00256 X 32.5^ = 2.70 Ib /f t ^

Step 3 . Maximum c o e ffic ie n ts on the bounding curves (use Figures 25 through 28

fo r h e lio s ta ts )

(a) The t h ir d row

Slope GBA Const.

Peak: CFx = ( -13.6 x 0.219 + 6.52 ) = 3.54

CFz = (-1 0 .6 X 0.135 + 4.31) = 2.88

CMHy = ( - 3 .2 x 0.135 + 0.90) = 0.468

CMz = ( - 4 .6 X 0.135 + 0.92) = 0.299

 60

(b) The f i f t h row

Peak: CFx = (-1 3 .6 x 0.25 + 6.52) = 3.12

CFz = (-1 0 .6 X 0.125 + 4.31) = 2.98

CMHy = ( - 3 .2 x 0.125 + 0.90) = 0.50

CMz = ( - 4 .6 X 0.125 + 0.92) = 0.345

The ca lc u la tio n of the base overturning moment, CMy fo r the th ird and

f i f t h rows is determined from Eq. ( 6 ):

CMjjgse = CMy = CFx + CMHy x (h/H) (h = 40 f t in th is problem)

Step 4 . Forces and moments (r e f e r to Eqs. (1) through ( 6 ))

(a) The th ir d row

Peak: Fx = 3.54 x 2.70 x 1600 = 15300 lb

Fz = 2.88 x 2.70 x 1600 = 12400 lb

Mbase = My = (3.54 + 0.468 x 40/22) x 2.70 x 1600 x 22

= 4.18 x10^ I b - f t

Mhinge = MHy = 0.468 x 2.70 X 1600 x 40

= 8.14 X 10"^ I b - f t

Mazimuth = Mz = 0.329 X 2.70 x 1600 x 40

= 5.54 x 10^ I b - f t
 61

(b) The f i f t h row

Peak: Fx = 3.12 x 2.70 x 1600 = 13500 lb

Fz = 2.98 X 2.70 x 1600 = 12900 1b

Mbase = My = (3.12 + 0.50 x 4 0 / 2 2 ) x 2 . 7 0 x 1600 x 22

= 3.83 X 10^ Ib -ft

Mhinge = MHy = 0 . 5 0 X 2.70 X 1600 x 40

= 8.72 X 10^ I b - f t

Mazimuth = Mz = 0.375 X 2.70 X 1600 x 40

= 6.36 X 10^ I b - f t

EXAMPLE PROBLEM 2

An example problem is presented fo r an isolated or i n - f i e l d parabolic disi

to demonstrate the calculatio n of peak load combinations fo r the major components

of the c o lle c to r where loads are not at t h e ir largest values. The same procedure

is followed fo r a h e lio s ta t.

Problem: Calculate the peak load combination of the major structural

components fo r a parabolic dish with a diameter, h, of 40

f e e t. The dish is in the second row of an array of

c o lle c to rs . Perform the calculations with and without an

external fence.

Givens: 1) H = 22 f t (height of center of dish from ground), 2) wind

has a peak gust speed of 55 mph at 10 m (32.8 f t ) elevation,

3) wind d ire c tio n , 0, is 50 , 4) c o lle c to r t i l t , a, is 70 ,

and 5) open country t e r r a in .

 62

Note: Refer to Figure 1, to Figures 8 to 15, and to Table 5 in

solving th is problem.

Steo 1. Dynamic pressure c a lc u la tio n . See Example Problem 1.

Q = 0.00256 (32.5)^ = 2.70 I b / f t ^

Steo 2 . Calculate isolated c o lle c to r peak loads fo r = 50" and a = 70".

(a) Drag force, F^^ (Figure 9 and Eq. (1 ))

^FX Q A

f"x,peak = (3.0) (2.70) (1257) = 10,200 lb

where A = TrhV^ is c o lle c to r area.

(b) L i f t force, F^ (Figure 11 and Eq. (2 ))

Cfz Q a

^ , peak = (-0 .9 5 ) (2.70) (1257) = -3200 lb

(c) Azimuthal moment, M,, (Figure 13 and Eq. ( 5) )

Cz Q A h

Mz.peak = (0.30) (2.70) (1257) (40) = 40,700 I b - f t

 63

(d) Hinge moment, MH (Figure 15 and Eq. (4 ))

CMHy Q A h

MH, .peak = ( 0 . 2 8 ) ( 2 . 7 0 ) ( 1 2 5 7 ) ( 4 0 ) = 3 8 , 0 0 0 I b - f t

(e) Base overturning moment. My (Eqs. (3) and ( 6 ))

CF^ CMHy h,H

r40
'M y,peak = 3.0 + 0.28 = 3.51
22

Therefore, the peak base overturning moment is

CMy Q A H

M y.peak = (3-51) (2.70) (1257) (22) = 2.62 x 10 ^ I b - f t

Step 3 . Calculate i n - f i e l d c o lle c to r performance at second row fo r ^ = 50* and

a = 70. See te x t fo r discussion of Table 5.

(a) Drag force, F^ (use values from Step 2 and Table 5)

Without external fence:

isolated Tbl 5

^x.peak = ( 10, 200) ( 1 . 0 0 ) = 1 0 , 2 0 0 lb

With external fence:

f'x.peak = (10.200) (0.71) = 7000 lb

 64

(b) L i f t force,

Without external fence:

f'z.peak = (-3200) (1.04) = -3300 lb

With external fence:

f'z.peaic = (-3200) (0.70) = -2200 lb

(c) Azimuthal moment,

Without external fence:

Mz.peak = ( 4 0 , 7 0 0 ) (1.38) = 56,200 I b - f t

With external fence:

Mz.peak = (40,700) (0.90) = 36,600 I b - f t

(d) Hinge moment, MHy

Without external fence:

MHy,peak = ( 3 8 , 0 0 0 ) (1.07) = 40,700 Ib - f t

With external fence:

MHy,peak = (38,000) (0.90) = 34,200 I b - f t

From th is example problem one can see the obvious advantages of an external

wind fence in reducing loads fo r an i n - f i e l d c o lle cto r in the second row. The

above calculatio ns y ie ld a component load combination fo r one s p e c ific c o lle c to r

o rie n ta tio n and i n - f i e l d placement wind d ire c tio n , and wind speed. The method
 65

allows f o r a wide range of c o lle c to r orientations and placements in studying load

combinations.

REFERENCES

ANSI/ASCE 7-88, (1988), "Minimum Design Loads For Buildings and Other
Structures," American Society of C iv il Engineers and American National
Standards Association, 1988 version of the national wind load standard.

ASCE, (1961), "Wind Forces On Structures," Transactions of the American Societv

of C iv il Engineers. Paper 3269, Vol. 126, p. 1124.

Cermak, J. E., (1975), "Applications of Fluid Mechanics to Wind Engineering," A

Freeman Scholar Lecture, ASME, J1. of Fluid Engineering. Vol. 97, No. 1.

Cermak, J. E., (1971), "Laboratory Simulation of the Atmospheric Boundary

Layer," AIAA J1. . Vol. 9.

Counihan, J . , (1975), "Adiabatic Atmospheric Boundary Layers: A Review and

Analysis of Data from the Period 1880-1972," Atmospheric Environment. Vol.
9, pp. 871-905.

ESDI), (1982), "Strong Winds in the Atmospheric Boundary Layer: Part 1: Mean-
hourly Wind Speeds," Item No. 82026, ESDU International L td ., London.

H o l l i s t e r , S. C., (1970), "The Engineering In te rp re ta tio n of Weather Bureau

Records fo r Wind Loading on Structures," Building Science Series 30 - -
Wind Loads on Buildings and Structures, National Bureau of Standards, pp.
151-164.

Peglow, S. G., (1979), "Wind-Tunnel Study of a Full-Scale H e lio s ta t," Sandia

Laboratories Report, DOE Contract DE-AC04-760P00789.

Peterka, J. A ., R. G. Derickson and J. E. Cermak, (1990), "Wind Loads and Local

Pressure D istributions on Parabolic Dish Collectors," Solar Energy
Research In s t i t u t e Report SERI/TP-253-3668, Golden, CO.

Peterka, J. A ., Z. Tan, J. E. Cermak and B. Bienkiewicz, (1989) "Mean and Peak

Wind Loads on H e lio s ta ts ," Transactions of the ASME, J1. of Solar Enerov
Engineering. Vol. I l l , May, pp. 158-164.

Peterka, J. A ., Z. Tan, B. Bienkiewicz and J. E. Cermak, (1988), "Wind Loads

on H eliostats and Parabolic Dish C ollectors," SERI/STR-253-3431, Technical
Report fo r Solar Energy Research In s t i t u t e , Golden, CO.

Peterka, J. A ., Z. Tan, B. Bienkiewicz and J. E. Cermak, (1987), "Mean and Peak

Wind Load Reduction on H e lio s ta ts ," SERI/STR-253-3212, DE87012281, Solar
Energy Research In s t it u t e , Golden, CO.
 66

Peterka, J. A ., N. Hosoya, B. Bienkiewicz and J. E. Cermak, (1986), "Wind Load

Reduction For H e lio s ta ts ," Solar Energy Research In s t i t u t e Report
SERI/STR-253-2859, Golden, CO.

Reinhold, T. A. (Ed.) (1982), Wind Tunnel Modeling fo r C iv il Engineering

Applications, Proceedings of an International Workshop, Cambridge
U niversity Press, NY.

Simiu, E. and R. H. Scanlan, (1986), Wind Effects on Structures, 2nd Edition,

John Wiley & Sons.
 67

UNLIMITED DISTRIBUTION Acurex Corporation (2)

INITIAL DISTRIBUTION 555 Clyde Avenue
REVISION 12/2/91/jwg M ountain View, C A 94039
A ttn: H. Dehne
U.S. Departm ent o f Energy (4)
Forrestal B uilding Code Advanced Therm al Systems
CE-132 1000 7600 East Arapahoe
Independence Avenue, SW Suite 319 Englewood, CO 80112
Washington, D C 20585 A ttn: D. Gorman
A ttn : G. Burch
S. G ronich Allegheny Ludlum Steel
80 Valley Street
U.S. Departm ent o f Energy (2) W allingford, CT 06492
Forrestal B uilding A ttn : J. H alpin
Code CE-13
1000 Independence Avenue, SW Analysis Review & C ritique
Washington, DC 20585 6503 8F Street
A ttn : B. Annan Cabin John, M D 20818
A ttn: C. LaPorta
U.S. Departm ent o f Energy
Forrestal B uilding Code Arizona Public Service Company
CE-10 P.O. Box 53999
1000 Independence Avenue, SW M /S 9110
Washington, DC. 20585 Phoenix,AZ 85072-3999
A ttn : R. San M a rtin A ttn: W. J. M cG uirk

U.S. Departm ent o f Energy (3) Arizona Solar Energy O ffice

Albuquerque Operations O ffice 3800 N orth Central
P.O. Box 5400 Phoenix, A Z 85012
Albuquerque,N M 87115 A ttn: R. W illiam son
A ttn : C. Garcia
G. Tennyson Australian N ational U niversity
N. Lackey Departm ent o f Engineering Physics
P. O. Box 4
U.S. Departm ent o f Energy Canberra A C T 2600 A U S T R A L IA
San Francisco Operations O ffice A ttn: S. Kaneff
1333 Broadway
Oakland, C A 94612
A ttn : R. Hughey Barber-Nichols Engineering
6325 West 55th Avenue
Arvada, CO 80002
A ttn: R. Barber
 68

B attelle Pacific Northwest Laboratory (2)

P.O. Box 999 Cummins Engine Co.
R ichland,W A 99352 MC 60125
A ttn : D. Brown P. O. Box 3005
Columbus, IN 47202-3005
B D M Coporation A ttn: R. Kubo
1801 Randolph Street
Albuquerque,N M 87106 Cummins Power Generation,Inc.,South
A ttn : W. Schwinkendorf 150 Tanne H ill D rive
Abilene, Texas 79602
Bechtel N ational, Inc. A ttn: M . M cGlaun
50 Beale Street
50/15 D8 Dan Ka
P. O. Box 3965 3905 South Mariposa
San Francisco, C A 94106 Englewood, CO 80110
A ttn : P. D eLaquil A ttn: D. Sallis

Black & Veatch ConsultingEngineers D LR

P.O. Box 8405 Pfaffenwaldring 38-40
Kansas City, M O 64114 7000 Stuttgart 80
A ttn : J. C. Grosskreutz FE D E R A L R E P U B LIC OF G E R M A N Y
A ttn: R. Buck
Tom Brumleve
1512 Northgate Road D LR (2)
W alnut Creek, C A 94598 Linder Hohe
5000 K oln 90
C alifornia Energy Commission F E D E R A L R E PU B LIC OF G E R M A N Y
1516 N inth Street, M-S 43 A ttn: M. Becker
Sacramento, C A 95814 M. Bohmer
A ttn : A . Jenkins
DSET
C alifornia Polytechnic U niversity P. O. Box 1850
Dept, o f Mechanical Engineering Black Canyon Stage I
Pomona, C A 91768 Phoenix, A Z 85029
A ttn : W. Stine A ttn: G. Zerlaut

C alifornia Public U tilitie s Com. E lectric Power Research Institute

Resource Branch, Room 5198 P.O .Box 10412
455 G olden Gate Avenue Palo A lto , C A 94303
San Francisco, C A 94102 A ttn: J. Schaeffer
Attn: T. Thompson
 69

Engineering Perspectives
20 19th Avenue
San Francisco, C A 94121 G arrett Turbine Engine Co.
A ttn : John Doyle I l l South 34th Street
P. O. Box 5217
Energy Technology Engr. Center Phoenix, A Z 85010
Rockwell International Corp. A ttn: E. Strain
P. O. Box 1449
Canoga Park, C A 91304 Georgia Power (2)
A ttn : W. Bigelow 7 Solar Circle
Shenandoah, G A 30265
E N IE C H , Inc. A ttn: W. King
P. O. Box 612246
DEW A irp o rt,T X 75261 H arris Corporation (2)
A ttn : R. W alters Govermnent and Aerospace Systems
D ivision
Flachglas Solartechnik GmbH P. O. Box 9400
Muhlengasse? Melbourne, F L 32902
D-5000 K oln 1 A ttn: K. Schumacher
F E D E R A L R E P U B LIC OF G E R M A N Y
A ttn : J. Benemann Industrial Solaj Technologies
5775 West 52nd Avenue
Flachglas Solartechnik Gm bH Denver, CO 80212
Sonnefistr. 25 A ttn: R. Gee
D-8000 Munchen 1
F E D E R A L R E P U B LIC OF G E R M A N Y Institute o f Gas Technology
A ttn : M . Geyer 34245 State Street
Chicago, IL 60616
Florida Solar Energy Center A ttn: Library
300 State Road 401
Cape Canaveral, F L 32920 ISEIR
A ttn : Library 951 Pershing D rive
Silver Spring, M D 20910
Ford Aerospace A ttn: A. Frank
Ford Road
Newport Beach, C A 92663
A ttn : R. Babbe
Foster W heeler Solar Development Lawrence Berkeley Laboratory
Corporation (2) MS 90-2024
12 Peach Tree H ill Road One Cyclotron Road
Livingston, NJ 07039 Berkeley, CA 94720
Attn: M. Garber Attn: A. Hunt
R. Zoschak
 70

Luz International (2)

924 Westwood Blvd.
Los Angeles, C A 90024
A ttn : D. Kearney Pacific Gas and E lectric Company (2)
3400 Crow Canyon Road
3M-Energy C ontrol Products (2) San Ramon, C A 94526
207-lW 3M Center A ttn: G. Braun
St. Paul, M N 55144 J. lannucci
A ttn : R. Dahlen
Peerless W insmith, Inc.
Mechanical Technology, Inc. (2) 172 Eaton Street
968 Albany Shaker Road P. O. Box 530 Springville, N Y 14141
Latham ,NY 12110 A ttn: W. H eller
A ttn: G. Dochat
J. Wagner Polydyne, Inc.
1900 S. N orfolk Street, Suite 209
M eridian C orporation San Mateo, C A 94403
4300 King Street A ttn : P. Bos
A lexandria,V A 22302
A ttn: D. Kum ar Power Kinetics, Inc.
415 R iver Street
N ASA Lewis Research Center (4) Troy,N Y 12180-2822
21000 B rook Park Road A ttn: W. Rogers
Cleveland, O H 44135
A ttn : R. Corrigan,500-221 Rocketdyne D ivision
L. Greenlee, 500-221 6633 Canoga Park Ave.
T. M roz, 301-5 Canoga Park,CA 91304
J. Calogeras, 301-5 A ttn: W. M arlatt

Nevada Power Co. San Diego Gas and E lectric Company

P. O. Box 230 P.O. Box 1831
Las Vegas, N V 89151 San Diego, CA 92112
A ttn : M ark Shank A ttn: R. Figueroa

SCE
P. O. Box 800
N R E L (5) Rosemead, C A 91770
1617 Cole Boulevard A ttn: C. Lopez
Golden, CO 80401
A ttn : T. W illiam s
L.M . Murphy
G.Jorgensen
T. Wendelin
A . Lewandowski
 71
Schlaich, Bergennann & Partner Steams Catalytic Corporation
Hohenzollernstr. 1 P.O. Box 5888-
D - 7000 Stuttgart 1 Denver, CO 80217
West Germany A ttn: T. B. Olson
A ttn : W. Schiel
Stirling Therm al Motors
Science A pplications International 275 M etty D rive
Corporation (2) AnnArbor, M I 48103
10343 Roselle Street, Suite G A ttn: T. Godett
San Diego, C A 92121
A ttn : K. Beninga Sun Power, Inc.
J. Sandubrae 6 Byard Street
Athens, O H 45701
Science Applications International A ttn: W. Beale
C orporation
M a il Stop 32 Tom Tracey
10206 Campus Point Court 6922 South Adams Way
San Diego, C A 92121 Littleton, CO 80122
A ttn : B. B utler
U nited Solar Tech, Inc.
Solar Kinetics, Inc. (2) 3434 M artin Way
P.O. Box 540636 Olympia, W A 98506
Dallas, T X 75354-0636 A ttn: R. Kelley
A ttn : J. A . Hutchison
P. Schertz U niversity o f Chicago
D. Konnerth Enrico Ferm i Institute
5640 E llis Avenue
Solar Power Engineering Company Chicago, IL 60637
P.O. Box 91 A ttn: J. O G allagher
M orrison, CO 80465
A ttn : H. W roton

Solar Steam U niversity o f Houston

P. O. Box 32 Solar Energy Laboratory
Fox Island, W A 98333 4800 Calhoun
A ttn : D. W ood Houston, T X 77704
A ttn: L. V ant-H ull
SRS Technologies
990 E xplorer Blvd., NW University o f U tah
H untsville, A L 35806 Mechanical and Industrial Engineering
A ttn : R. Bradford Salt Lake City, U T 84112
A ttn: B. Boehm
 72

E ric Weber
302 Caribbean Lane
Phoenix, A Z 85022

WGAssociates
6607 Stonebrook Circle
Dallas, T X 75240
A ttn : V. Goldberg

140 R. B. Loehman
1846 D. H . Doughty
1846 C. S. Ashley
7141 Technical Library (5)
7613-2 Document Processing (10)
forDOE/OSTT
7151 Technical Publications
4051 Disclosure D ivision (3)
6000 D. L. H artley
6200 B. W. M arshall
6215 C. P. Cameron
6215 R. M . Houser
6216 C. E. Tyner
6216 L. Yellowhorse
6216 D. J. A lp e rt
6216 J. W. Grossman (20)
6216 T. R. M ancini (3)
6216 J. E. Pacheco
6217 P. C. Klim as
6217 R. B. D iver
6220 D. G. Schueler
6221 T. C. Bickel
6221 A. R. Mahoney
6223 G. J. Jones
6224 D. E. Hasti
7470 J. L. Ledman
7476 F. P. Gerstle
7476 S. T. Reed
8523-2 Central Technical Files