RFP-3304

UC-60

MASTER
DO WOT MICROFILM m c DONNELL 40-kW
COVER as®»i GIROMILL WIND SYSTEM
Phase II
Fabrication and Test

JUNE 1980

ROBERT BRULLE

M cD O N N E L L A IR C R A FT C O M P A N Y
P.O. Box 516
St. Louis, Missouri 63166

Prepared for
Rockwell International Corporation
y rw g n m r>;T Energy Systems Group
Rocky Flats Plant
Wind Systems Program
P.O. Box 464
Golden, CO 80402

Subcontract No. PF-64100

As a Part of the
U N IT E D S TA TE S D E P A R T M E N T CP E NERG Y
W IN D ENERG Y T E C H N O L O G Y D IV IS IO N
FEDERAL W IN D E NERG Y P R O G RAM

Contract No. D E -A C 04-76D P 03533

DISTRIBUTION OF THIS OOCUMtHT IS UNLIMITED

 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.
 DISCLAIMER

Portions of this document may be illegible in

electronic image products. Images are produced
from the best available original document.
 J i-

D IS C L A IM E R

T h is re p o rt w as p re p a re d as an a c c o u n t of w o rk s p o n s o re d by th e U n ite d S ta te s
Government. Neither the United States nor the United States D epartm ent of Energy, nor
an y of th e ir e m p lo y e e s , m a k e s a n y w a rra n ty , e x p ress o r im p lie d , o r a ssu m es a n y legal
lia b ility o r re s p o n s ib ility fo r th e a c c u ra c y , c o m p le te n e s s , o r u sefu ln ess of an y
in fo rm a tio n , a p p a ra tu s , p ro d u c t, o r pro cess d is c lo s e d , o r re p re s e n ts th a t its use w o u ld
n ot in frin g e p riv a te ly o w n e d rights. R e fe re n c e h e re in to a n y s p e c ific c o m m e rc ia l
p ro d u c t, process, o r service by tra d e n a m e , m ark, m a n u fa c tu re r, o r o th e rw is e , d oes not
n e c essa rily c o n s titu te o r 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 , o r fa v o rin g by th e
U n ite d S tates G o v e rn m e n t o r a n y a g e n c y th e re o f. T h e view s an d o p in io n s of a u th o rs
exp res s e d h erein do not n e c e s s a rily state o r re fle c t th o s e of th e U n ite d S tates
G o v e rn m e n t o r a n y a g e n c y th e re o f.

P rin te d in th e U n ite d S ta te s o f A m e ric a

A v a ila b le fro m

N a tio n a l T e c h n ic a l In fo rm a tio n S e rv ic e
U .S . D e p a r tm e n t o f C o m m e r c e
5 2 8 5 P o rt R o y a l R o a d
S p rin g fie ld . V A 221 61

P rin te d C o p y: $7 . 00 M ic ro fic h e : $ 3 .5 0
 RFP.=
R F P — 3 304 UC-6 0

DE82 015145

M C D O N N E L L 40-kW G I R O M I L L
W I N D SY S T E M

Phase II
Fabrication and Test

June 1980

Ro b e r t Brulle

McDonnell Aircraft Company

P.O. Box 516
St. Louis, M i s s o u r i 63166

P r e p a r e d for
Rockwell International Corporation
E n e r g y Systems G roup
R o c k y F lats Plant
W i n d Systems P r o g r a m
P.O. Box 464
Golden, C o l o r a d o 80402

Su b c o n t r a c t No. PF-64100

As a Part of the

U N I T E D STATES D E P A R T M E N T OF ENERGY
WIN D EN E R G Y T E C H N O L O G Y D I V I S I O N
F E D E R A L W I N D E N E R G Y PROGRA*M

C o n t r a c t No. DE-AC04-76DPO3533

- DISCLAIMER ■
This book wdb Di’epjred dS an dccount of work sponsored by <jn agency of the United States Government.
Neither the United States Governrnent nor any agency thereof, nor dny of their employees, makes any
warranty, express or implied, or assumes any Imjal liab ility nr responsibility fo r the accuracy,
completeness, or usefulness o f any information, aopdratus, prtxJuct, or process disclosed, or
represents that its use would not infringe privately owned rights. Reference herein to any specdic
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.

BISTBIBUTION OF THIS OOCtlMENT IS (W ill

 ABSTRACT

This report summarizes the results of Phase II of a program

to design and test a 40 k W vertical axis windmill called a
"Giromill". Phase I of this program covered trade off studies,
selection of configuration, and detail design of that configura­
tion. In Phase II the unit was fabricated, erected, and tested.
This p r o g r a m was conducted under contract PF64100F, awarded by
the Rockwell International Energy Systems G roup at Rocky Flats,
Colorado, as part of the Department of Energy (DOE) small win d ­
mill development program.

McDonnell Aircraft Company (MCAIR) was prime contractor,

with major assistance from Valley Industries through a subcon­
tract and license agreement, and from McDonnell Douglas Elec­
tronics Division (MDEC) through an intercompany work order.

Mr. J. W. Anderson was Program Manager for MCAIR, Mr.

W i l l i a m Duwe was Engineering Manager for Va l l e y Industries, and
Mr. T o m Schmidt was Engineering Manager for MDEC. The principal
engineers for M CAIR were Messrs. Burt Birchfield, Bob Brulle,
Howard Clark, and Willis PCunz; for V a l l e y Industries, Mr. J i m
Herr; and for MDEC, Messrs. Bob Udell and Jerry Swaitlowski.
 TABLE OF CONTENTS

Section Page

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

2. DESIGN AND ASSEMBLY OVERVIEW .............................. 2

2.1 Design Trade Studies ............................... 2

2.2 Design Criteria Development ........................ 2
2.3 Detail Rotor Design .................................. 2

2.3.1 B l a d e .......................................... 5
2.3.2 Support A r m s ................................... 6
2.3.3 Rotating Tower ................................. 6

2.4 Fixed Tower Design .................................. 6

2.5 Control System D e s i g n ............... 11
2.6 Mechanical and Electrical Output Power System . . . 14
2.7 P e r f o r m a n c e ............................................ 14
2.8 Assembly and C h e c k o u t ................................ 14
2.9 Organization and S c h e d u l e ........................... 24

3. FABRICATION DESCRIPTION .................................. 28

3.1 B l a d e s ................................................ 28
3.2 Support A r m s ......................................... 29
3.3 Fixed and Rotating T o w e r s ........................... 29

4. PHASE II - DESIGN C H A N G E S ................................. 32

4.1 Design Change Descriptions ......................... 32

4.2 Drawing Summary ................ 33

5. UPDATE OF MANUFACTURING COST ESTIMATES ................... 42

5.1 Prototype Cost E s t i m a t e s ........................... 42

5.2 Production Design Considerations .................. 42

6 . T E S T I N G ...................................................... 45

6.1 Control System Acceptance Tests ...................... 45

6.2 Giromill System Acceptance Tests .................. 46
6.3 Operational T e s t ..................................... 46

6.3.1 S t r u c t u r a l ..................................... 47
6.3.2 M e c h a n i c a l ..................................... 53
6.3.3 Control S y s t e m ................................. 55
6.3.4 Electric Generation .......................... 58

6.4 P e r f o r m a n c e ............................................ 58
 TABLE OF CONTENTS (Continued)

Section Page

7. CONCLUSIONS & RECOMMENDATIONS ............................ 60

7.1 C o n c l u s i o n s ............................................ 60
7.2 R e c o m m e n d a t i o n s ....................................... 60

8 . R E F E R E N C E S .................................................... 62

11
 LIST OF ILLUSTRATIONS

Figure Page

1 Selected Giromill Configuration ................. 3

2 Giromill at Rocky Flats Test Center ............ 4

3 40 k W W i n d Conversion Sy s t e m Specifications . . 5

4 Blade Cross Section ............................... 6

5 General Arrangement Support A r m ................. 7

6 A r m /Tower Attachment ............................... 7

7 Rotating Tower ...................................... 8

8 Fixed T ower ........................................ 9

9 Upper Bearing Mounting ............................ 9

10 Lower B e a r i n g ...................................... 10

11 Lower Bearing Mounting ............................ 11

12 Foundation P i e r ................................... 12

13 Programmed Rock Angle Profiles ................... 13

14 Blade Actua t o r A s s e m b l y .......................... 13

15 Electrical Drive System .......................... 15

16 Giromill Control Performance ..................... 16

17 Fixed Tower A f t e r Installing the Upper

S e c t i o n ............................................ 17

18 F ixed Tower w i t h Lower Section of Rotating

T o w e r ............................................... 18

19 Attachment of Rotor to Assem b l y S t a n d .......... 19

20 Rotor A s s e m b l y 20

21 Aligning Blade w i t h Support A r m s ................. 21

22 Rotor on Assembly S t a n d .......................... 22

23 Lifting Rotor Assembly to Fixed Tower .......... 23

24 Giromill Operating Checkout . . . . 25

iii
 LIST OF ILLUSTRATIONS (Continued)

Figure Page
25 MCAIR 40 k W Giromill O r g a n i z a t i o n .............. 26

26 40 k W Giromill Development Schedule ............ 27

27 A s s embly Tooling for Giromill Blades ............ 28

28 Support A r m Assem b l y Fixture ..................... 30

29 Strain Relief Cycle for Support A r m s ............ 31

30 Blade Shims for C l e a r a n c e ........................ 34

31 Giromill Control System Constants .............. 35

32 Giromill Control System CSMP Simulation

C o n t r o l l e r .......................................... 36

33 MCAIR Blade Drawings ............................... 37

34 V a l l e y Top Assembly Drawings ..................... 38

35 Giromill Control System Drawings ................. 39

36 Giromill Actuator Drawings ........................ 40

37 Ground Rules for Estimating Cost of 1000th

U n i t ................................................. 43

38 Giromill Budgetary Cost Estimate - 1977

D o l l a r s ............................................ 43

39 Lightning Pole-to-Rotating Tower Interface

J o i n t .............................................. 49

40 Washer Partially Extruded into H o l e ............ 50

41 Lightning Pole-to-Rotating Tower Interface

G e o m e t r y ............................................ 51

42 Limit Loads on Lightning Pole vs Wind

V e l o c i t y ............................................ 52

43 Was h e r Bending and Shear Stress ................. 52

44 Blade Actuator Belt Tension Measuring Access

H o l e ................................................. 54

45 Giromill Control System - Processor (Phase

Sequence) .......................................... 56

46 Giromill P e r f o r m a n c e ............................... 59

IV
 1. INTRODUCTION

The design, fabrication and test of a 40 k W Giromill was

accomplished in two phases. A preferred configuration was
selected and designed during Phase I. This was done from 15
September 1978 to 15 June 1979, as detailed in Reference 1.

This report details the building, erection, and test ac t i v i ­

ties of Phase II from the start on 15 June 1979 through April
1981.

Fabrication was completed by January 1980. The fixed tower

was erected in December 1979, and the complete assembly including
the rotating tower, was erected in February 1980. First turn was
accomplished on 5 March 1980, and checkout testing, including
running at operational RPM, was completed on 14 March 1980.
These first checkout tests were done at Valley Industries plant
in Tallulah, Louisiana.

A f t e r the checkout tests were completed, the unit was d isas­

sembled and shipped to Rocky Flats. First turn at Rocky Flats
was on 3 July 1980.

The Giromill is presently connected into the local utility

grid and has produced 40 k W o r more of electric p ower when o p e r ­
ating and with adequate winds. Operational testing in the
electrical mode is continuing and may be followed by testing in
the mechanical mode.
 2. DESIGN AND A S S EMBLY OVERVIEW

The contractual design requirements specified the d e v e l o p ­

m ent of a prototype Giromill ha v i n g an electrical p ower output
plus an adapter ki t to convert the system to a mechanical p ow er
output. A stand alone capability requiring no external power was
also required.

Design tasks included: (a) Trade-off Studies, (b) Design

Criteria Development, (c) Detail Rotor Design, and (d) Detail
Control System Design. Supporting studies were conducted in
aerodynamics, performance, loads, structural dynamics, weights,
failure modes and effects, and test instrumentation.

A fter design the unit was fabricated and asseiribled.

McDonnell Aircraft Company (MCAIR) made the blades w hile Vall ey
Industries made the remainder of the structure. McDonnell
Douglas Electronics Company fabricated the controller and blade
actuators.

2.1 DESIGN TRADE STUDIES - To arrive at the preferred Giromill

configuration for detail design, a series of trade off studies
was conducted covering geometry, control system, blade, support
arms, rotating tower, and fixed tower. Geometric trade-offs
considered number of blades, number of support arms, blade solid­
ity, blade thickness, and rotor aspect ratio. M i n i m u m system
cost was an important selection criterion.

Figure 1 shows the selected configuration and Figure 2 shows

the Giromill erected and running at Rocky Flats test center. The
rotor has three blades and is 58 ft in diameter b y 42 ft long.
Each blade is supported b y two support arms to give a mini m u m
blade bending moment. The support arms are h e l d up w i t h stream­
lined cables. The rotating tower extends to ground level to
reduce bearing side loads and to locate the transmission and
generator near the ground for easy maintenance access. The fixed
tower is a truss tower made of structural angles. A microproces­
sor controller drives electrical actuators that modulate the
blades to maintain constant rotor RPM.

2.2 DESIGN CRITERIA DEVELOPMENT - The system specifications are

presented in Figure 3. These specifications were related to
design criteria for the various components. Included were operat­
ing loads, maxi m u m storm loads, and snow and ice loads.

2.3 DETAIL ROTOR DESIGN - The rotor consists of a central steel

tube rotating tower w ith three aluminum blades, each supported b y
two steel support arms. The support arms are pinned at the
rotating tower and supported at the outboard end b y streamlined
steel cables. The rotating tower is supported b y two bearings:
one at the top of the fixed tower and another near the ground.
 0

GP13-0189-5

FIGURE 1
SELECTED GIROMILL CONFIGURATION
 OP13-0189-30
FI GU RE 2
GIROMILL AT ROCKY FLATS T E S T C E N T E R
 ELECTRICAL MECHANICAL
OUTPUT 40kW M IN IM U M IN 9 m/s
40 kWMIN. IN 9 m/s (20 MPH) AT
POWER WIND
SEA LEVEL DENSITY. 60 Hz, POWER
MODE HORIZO NTAL SHAFT AT CON­
FACTOR OF 0.8 OR HIGHER
OPTIONS STANT SPEED OF EITHER 440,
1. MATCH WITH EXISTING UTILITY
GRID: 3-PHASE, 480 VOLT ±5% 880, OR 1760 RPM. SHAFT SPEED
2. INDEPENDENT OF U T IL IT Y GRID: NOT TO V A R Y MORE THAN ±1%
SINGLE MACHINE, 3-PHASE, 480 FOR WIND SPEED GREATER
VOLT ±5% THAN 9 m/s
3. INDEPENDENT OF U T IL IT Y GRID:
SINGLE MACHINE, 240 V D L T ±5%
4. 3-PHASE 480 VOLT ±5%, FOR TIE-
IN OF TWO OR MORE MACHINES
HEIGHT CENTERLINE OF ROTOR SWEPT AREA SAME
TO BE AT A HEIGHT OF 75 FT.

WIND RANGE
CUT-IN M IN IM IZ E WITH REGARDS TO ECONOMICS SAME
OF POWER PRODUCTION AND SYSTEM
COST.
CUT-OUT* 27 m/s (60 MPH) M IN IM U M . SELECTION SAME
OF A LOWER SPEED TO BETTER MEET
PROGRAM OBJECTIVE OF LOW-COST
POWER PRODUCTION MUST BE ADE Q UATE ­
LY JUSTIFIED.
PEAK GUST 56 m/s (125 MPH) M IN IM U M WITH A 1.5
PROTECTION SAFETY FACTOR

CONTROLS
START/STOP AUTOMATIC SAME
SHUTDOWN/CONTROL A UTOM ATIC FOR ROTOR OVERSPEEO
BACK-UP SHUTDOWN MECHANISM.
OPERATION AUTOM ATIC CUT-IN, AND CUT-OUT
AUTOM ATIC RE-ENGAGE AS WINDS
DROP BELOW CUT-OUT SPEED
OUTPUT AS REQUIRED TO PROVIDE PROPER
OUTPUT POWER MODE

*A cut-out wind speed of 40 mph was selected for the prototype. QP13-0189-6

FIGURE 3
40 kW WIND CONVERSION SYSTEM SPECIFICATIONS

2.3.1 Blade - Each Giromill blade has an N A C A 0018 shape

and is a two-cell sheet metal airfoil consisting of a 0.16 inch
leading edge skin, a 0.125 inch channel spar, and a 0.020 inch
be a d e d trailing edge skin. A cross-section of the b lade is shown
in Figure 4. Blade bending, shear, and torsion are carried by
the leading edge and spar. The beaded trailing edge structure
acts as a truss member to transfer local air loads to the leading
edge torque box.
 0.160 6061-T6 ALUMINUM
0.020 6061-T6 A L U M IN U M

0.125 6061-T6 A L U M IN U M

FIGURE 4
BLADE CROSS SECTION

The blade is attached to the rotor ar m through a 4140 steel

support tube fitting inserted into the end of each blade section.
Blade bending is transferred to the tube b y a coupling between
two machined aluminum ribs. Torsion in the b lade is transferred
through bolts attaching the root rib to a flange on the support
tube fitting.

2.3.2 Support Arms - The support arms are wel d e d sheet

steel construction, tapered and streamlined to minimize a e r o ­
dynamic drag. The outboard h a l f of the arms has a smaller cross
section and is more streamlined. The outline of the support arms
is shown in Figure 5. A ttachment of the support arms to the
rotating tower is shown in Figure 6 . The streamlined w ire braces
are formed from 0.625 inch diameter stainless steel bar.
Turnbuckles are used for rigging adjustments.

2.3.3 Rotating Tower - The rotating tower is made of

flanged tubular sections which are bolted together. The middle
sections are A36 steel tubing, 24 inch diameter b y 0.2 50 inch
wall thickness. Figure 7 illustrates the rotating tower.

2.4 FIXED TOWER DESIGN - The fixed tower is a truss type made of
A S T M A36 structural steel angles. The joints are bolted. Figure
8 shows the fixed tower.

T h e upper bearing for the rotor is a sealed ball bearing.

The inner bearing race is bolted to a flange on the rotating
tower, as shown in Figure 9. The outer race is bolted to a steel
ring for reinforcement. To provide a flexible mounting so that
this bearing carries only radial loads, the bearing assembly is
attached to the fixed tower wit h four thin sheets of steel.
 VIEW A \ D^
n

... ____1
/ D—
c J

WELDED (TYP)
A36 STEEL

B-B C-C D-D

GP13-0189-8

FIGURE 5
GENERAL ARRANGEMENT SUPPORT ARM

STEEL

WELDED
(TYP)

GP79-063e-28

VIEW E-E
A36STEEL

FIGURE 6
ARM/TOWER ATTACHMENT
 U T IL IT Y POLE
TENSION MEMBER
ATTACH LUGS

UPPER SUPPORT ARM

LOWER SUPPORT ARM

SUPPORT ARM
UPPER BEARING AREA

24 IN. D IA TUBE BRAKE DISC

24 IN. D IA TUBE

LOWER BEARING AREA

GENERATOR SPEED
INCREASER
SLIP RINGS

QP1M189-10
FIGURE 7
ROTATING TOWER

8
 GP13-0189-11

FIGURE 8
FIXED TOWER

R O T A T IN G TOW ER

B EA R IN G

STEEL SHEET

TOW ER LEG

^ S T E E L R IN G

L IG H T N IN G BRUSH
GP13-01B9-12

FIGURES
UPPER BEARING MOUNTING
 A tapered roller bearing is used for the lower bearing
(Figure 10). It is bolted to a plate that is suspended from the
four tower legs b y four tension members. Turnbuckles are used to
adjust the length of the four members for proper rigging of the
lower bearing, as shown in Figure 11. Four horizontal members
attached to the tower legs take the side load on the bearing.
This bearing carries the entire thrust load (weight) of the rotor
and radial load.

[— 9 -----

22 1/2

25.50

18.50
9.50

1.25 BOLT DIA

Dimensions in inches.

G P13-0189-13

FIGURE 10
LOWER BEARING

10
 TOWER LEG

TURNBUCKLE
ROD

BEARING

HORIZONTAL Q. ROTOR
BEARING SUPPORT

GP13-0189'14

FIGURE 11
LOWER BEARING MOUNTING

A disc brake system is designed into the prototype for

emergency situations. The brake disc is bolted between two
flanges of the rotating tower. The caliper is actuated b y
internal springs and released by hydraulic pressure. In all
normal operating and standby modes the caliper is in the released
position making for a fail safe configuration.
The fixed tower has a spread foundation made up of four
concrete piers. Figure 12 illustrates a typical pier. Two 1.5
inch diameter anchor rods extend out of each pier.
2.5 CONTROL SYSTEM DESIGN - The control system uses a m i c r o ­
processor controller that sends commands to electric blade
actuators. Constant RPM control is achieved b y a proportional
plus integral feedback on rotor RPM, summed with a measured blade
speed command. The blade rock angle (relative pitch angle) com­
mands generated by the control unit are transmitted to individual
electrical blade actuators. Each actuator consists of an elec­
tric motor, power amplifier, and gear box. The actuator
positions the blades according to the commands resulting from
position feedback which is monitored b y a potentiometer. This
potentiometer is mounted in the actuator gear box and related to
blade position through a separate set of gears.

11
 2 SQUARE

GROUND LEVEL

6.5

2.5 TYP

4 SQUARE
9 SQUARE

Dimensions in feet. GP13-0189-15

FIGURE 12
FOUNDATION PIER

The control system is housed in two enclosures: (1) control

unit which contains the microprocessor and (2 ) control system
power switching unit (CSPSU). These enclosures are mounted on
the control panel at the base of the fixed tower.

The control unit utilizes a microprocessor in conjunction

with three programmable read-only memories (PROMS) and associated
interface circuits. The rock angle profiles. Figure 13, stored
in the PROMS are used for commanding the blade actuator as a
function of the blade phase angle, , and rotor RPM. Self tests
in the controller assure a fail-safe system.

The Giromill rock angle actuator is a self contained closed

loop servomechanism which controls the angular position of the
output shaft in response to a position signal. This servomecha­
nism consists of an electronic control amplifier and a direct
drive actuator powered b y a dc motor operating from the 48 volt
supply. An isometric view of the actuator is shown in Figure 14.

Power for the control system is generated b y a 48 volt

alternator, driven by a toothed belt from the main gear box at
the bottom of the rotating tower. Four 12 volt batteries,
mounted on the control panel, provide power for starting and
standby.
 Vw
Cam
(MPH)
0 1.45 46
1 1.45 46
2 1.73 40
3 1.93 36
4 2.17 32
5 2.48 28
6 2.89 24
7 3.47 20
8 3.85 18
9 4.33 16
10 5.78 12

ROCK
ANGLE

CAM 10

270 360
BLADE PHASE ANGLE DEG

FIGURE 13
PROGRAMMED ROCK ANGLE PROFILES

TO BLADE
PULLEY

GEAR
BOX

O
POWER
AMPLIFIER

MOTOR

FIGURE 14
BLADE ACTUATOR ASSEMBLY

13
2.6 MECHANICAL AND ELECTRICAL OUTPUT POWER SYSTEM - A shaft-
mounted helical g e a r box speed increaser, having a g ear ratio of
24.3 to 1, is mounted on the lower end of the rotating tower.
The output shaft of the g e a r box drives the generator through a
toothed belt final stage for an overall increase of 54.675 to 1.
For the electrical output design the generator speed is 1830 RPM.
Figure 15 shows the electrical drive system assembly.

An induction g e nerator is used to feed 480 volt, 3 phase, 60

Hz power into a large utility grid. A magnetic contactor, c o n ­
trolled from the control system through a relay, is used to
connect the Giromill to the utility grid.

The mechanical output kit converts the Giromill from an

electrical output to a 1760 RPM mechanical output through a
horizontal shaft. A right angle gear box and mounting bracket
replaces the electrical generator.

2.7 PERFORMANCE - Performance calculations were made employing

the Larsen Cyclogiro Performance Computer Program. Figure 16
shows the overall performance computed for blades being modulated
to give constant effective angles of attack e* The maximum p r e ­
dicted Cp is approximately 0.5 and occurs at an e ~ 9° a
blade speed ratio \ = 3.4. The cyclogiro performance program
does not account for aerodynamic loss due to blade damping. This
loss in Cp was computed separately and is shown in Figure 16.
Note that its effect is predominate at the higher blade speed
ratios and is negligible at blade speed ratios less than 3.5.

Also shown in Figure 16 are lines of constant rotor p ower

(10 k W to 50 kW) and the performance point for the rock angle cam
in discrete winds from 12 to 40 MPH (circled points) . These
points show what cam o r rock angle profile would be required to
achieve that part i c u l a r performance at the specified wind speed.
The double dashed lines emanating downward from these discrete
wind points show how the power coefficient varies when that con­
stant blade rock angle profile is maintained. For example,
modulating the blades along a profile defined by cam 6.5 ( \ =
3.15 see Figure 13) o r 22 MPH wind speed cam, would result in a
Cp of .385 at X = 3.15. Maintaining this same blade modulation
as the blade speed ratio was increased would result in the Cp
decrease shown b y the double dashed line leading downward and to
the right. The fixed cam Cp variations from 22 MPH to 40 MPH are
shown. The Cp variations for fixed cam modulations from 20 MPH
down to 12 MPH are not shown since they lie very close to the
dashed line obtained by correcting for aerodynamic damping.

To achieve an output of 40 k W from an electrical generator

requires the rotor to have an output of about 50 kW. The 50 k W
line therefore shows the rock angle variation needed for constant
p ower above a wind speed of 20 MPH.

2.8 ASSEMBLY A N D CHECKOUT - The unit was assembled and checked

out at the Valley Industries Plant at Tallulah, La. A fter check­
out it was disassembled and shipped to Rocky Flats.

14
 GEARBOX
- :o SPEED
ROTATING TOWER SENSOR
SHAFT

TOOTHED BELT

BELT GUARD

TOOTHED BELT ROTATING TOWER SPEED SENSOR

SHAFT
BELT GUARD SHEAR PIN

itm_

GEARBOX

GENERATOR
a p i3 -o is » it
FIGURE 15
ELECTRICAL DRIVE SYSTEM

15
 0 .6

50 kW 40 CONSTANT ROTOR
POWER
20 MPH
0.5

0.4

POWER
COEFFICIENT, 0.3
12
AERODYNAMIC
DAMPING LOSS

0.2

0.1

FIXED CAM
k (22 MPH)
(24)
(28)

BLADE SPEED RATIO, X GP13-0189-19

FIGURE 16
GIROMILL CONTROL PERFORMANCE

Pictures of the unit assembly at Tallulah are shown in

Figures 17 through 23.These pictures illustrate the key
assembly steps as follows;

1. Lower portion of the fixed tower was first erected.

2. Upper portion of the fixed tower was assembled and then

raised and attached to the lower p a r t .

3. A subassembly of the lower mai n bearing with the lower

p ortion of the rotating tower and speed increaser/
generator was completed and then this assembly was
attached to the fixed tower.

4. Installation of the remaining portion of the lower

rotating tower and upper main bearing was then
completed. At Tallulah the rotating tower was eased
through the side of the fixed tower where several
members had been removed. At Rocky Flats the tower was
lowered through the top.

16
 GP13-01S9-24

FI GURE 17
FIXED T O W E R A F TE R INSTALLING T H E U P P E R S E C T I O N

17
 'A

GP13-0189-2S

FI GURE 18
FIXED T O W E R WITH L O WE R S EC T I O N O F ROTATING T O WE R

10
 GP13-0189-28

FI GURE 19
A TT A CH M EN T O F R O T O R TO A S S E M B L Y S TA ND

19
 QP134)1S9-27

F I G U RE 20
ROTOR ASSEMBLY

20
 QP13-0189-26

FI G UR E 21
ALI GNI NG BLADE WITH S U P P O R T A R M S

21
f ;

FI GURE 22 QP13-01S9-25

R O T O R ON A S SE M BL Y S TA ND

22
 G P13-0189-23
FIGURE 23
LIFTING ROTOR ASSEMBLY TO FIXED TOWER

23
 5. The rotating tower was then rigged to be vertical and
turn smoothly b y adjustment of the lower bearing tension
members and alignment of the lower bearing.

6 . Rotor assembly stand was then erected.

7. Upper rotating tower was assembled and placed on

assembly stand.

8 . The lower and then the upper support arms were attached
to the rotating tower.

9. The blades were installed and the entire rotor rigged

for level support arms orientation and free turning of
the b l a d e s .

10. Blade actuators were installed and temporary wiring

connected running from the controller to the actuators.

11. Actuator and control unit checkout was then completed.

12. Rotor was lifted and attached on top of the fixed tower.

13. All other parts were installed and all wiring connected.

14. The unit was then certified for final checkout and
running.

System checkout was accomplished without connecting the gen­

erator to the grid. The unit was therefore forced to run at zero
power output. It did it very well, stabilizing at synchronous
RPM even with moderate wind gusts. A picture of the unit running
at Tallulah is shown in Figure 24. About one wee k of testing was
accomplished prior to shipment to Rocky Flats.

2.9 ORGANIZATION AND SCHEDULE - Figure 25 presents the ke y

membei: organization and Figure 26 shows the schedule for accom­
plishing the effort.

24
 aP13-01S»-20

FIGURE 24
GIROMILL OPERATING CHECKOUT

25
 VICE PRESIDENT
ETD

W .J. BLATZ

DIRECTOR
ETD - DESIGN

H.H. COLE

PROGRAM
MANAGER

J.W . ANDERSON

PLANNING AND
MARKETING A D M IN IS TR A TIO N

G.L. HIBBARD J.M . PURSES

TECHNOLOGY DESIGN VALLEY PUM P PROCUREMENT CONTRACTS

INTEGRATION INTEGRATION DIVISIO N

R.V. BRULLE E.B. BIRCHFIELD W .D . DUWE P. WALOMAN O.F. PETRU

MDEC
CONTROL SYSTEM DESIGN VALLEY DESIGN

T.G. SCHMIDT W .H. KUNZ J. HERR

MDEC STRENGTH
ACTUATORS

R.F. UDELL H.T. CLARK

GP13-0S61-2

FIGURE 25
MCAIR 40 kW GIROMILL ORGANIZATION

26
 1978 1979 1980 1981

S 0 N D J F M A M J J A S 0 N D J F M A M J J A S 0 N D J F M A M J J A S 0 N 0

G O -AH EAO ......................... ............................................................

1.0 P H A S E l-D E S IG N

1.2 DEVELOP DESIGN CRITERIA (TASK 2)

1.3 PREFERRED DESIGN SELECTED_______________ .4k
▲ ▲
1.4 ROTOR DESIGN (TASK 3)
1.5 FIXED TOWER DESIGN (TASK 4)
▲
1.6 CONTROL SYSTEM DESIGN (TASK 61
1.7 MECHANICAL AND ELECTRICAL OUTPUT
▲
POWER SYSTEM DESIGN (TASK 8)
1.8 FAILU R E MODE AND EFFECT ANALYSIS A,
(TASK 7)
1.9 DRAFT TEST PLAN PREPARATION (TASK 8)
1.10 TEST INSTRUMENTATION DEFINITION ANO
DESIGN (TASK 9)
N-)
-J 1.11 PR ELIM IN ARY PRODUCTION SYSTEM COST
ANALYSIS (TASK 10)...................................................
112 DECISION TO PROCEED TO PHASETT 4.
2.0 P H A S E X - COMPONENT DEVELOPMENT/PROTOTYPE
FABRICATION AND TEST
71 RnrnR FARRinATinN (ta < :k I I 1
2.2 FIXED TOWER FABRICATION (TASK 12)
▲
2.3 CONTROL SYSTEM FABRICATION (TASK 13)
2.4 MECHANICAL AND ELECTRICAL OUTPUT
POWER SYSTEM FABRICATION (TASK 14)
2.8 SYSTEM ASSEMBLY (TASK 15)
• T A L L U L A H L O U IS IA N A ________ ___________
• ROCKY FLATS, C O L O R A D O __________ _____
2.7 SYSTEM OPERATIONAL TEST (TASK 18)
2.8 G IRO M ILL R E P A IR ......... .................................. ..........
QP13^1-10

FIGURE 26
40 kW GIROMILL DEVELOPMENT SCHEDULE
 FABRICATION DESCRIPTION

3.1 BLADES - The b l a d e s w e r e f a b r i c a t e d i n t h r e e s e c t i o n s ;

upper, m id d le and lo w e r. The u p p e r an d l o w e r s e c t i o n s a r e
id en tical. The same m e t h o d o f f a b r i c a t i o n was u s e d f o r e a c h
sectio n . The a s s e m b l y t o o l i n g i s shown i n F i g u r e 2 7 . The p r i ­
m ar y a s s e m b l y t o o l i n g ( r i g h t s i d e o f f i g u r e ) c o n s i s t s p r i m a r i l y
o f seven yoke f i t t i n g s which s u p p o r t t h e t o t a l b l a d e assem bly and
in s u re p ro p er alignm ent o f th e th r e e b lad e s e c t i o n s .

GP13-0S21-4

FIGURE 27
ASSEMBLY TOOLING FOR GIROMILL BLADES
 The first part placed in the yoke fixture was the leading
edge skin. Machined ribs were then riveted to the leading edge
skin with the connecting torque tubes inserted through the ribs
to maintain alignment. Spars and trailing edge skins were then
concurrently riveted to the leading edge skin. A second set of
yoke fixtures, shown in the middle of the figure, were used for
aligning the trailing edge tab prior to riveting the two trailing
edge skins together, through the tab, and thus closing up the
blade structure. Finally, the formed sheet metal ribs at the end
of each blade section were riveted to the trailing edge skin and
spar.

3.2 SUPPORT ARMS - The support arms were assembled in the fix­
ture shown in Figure 28. Initially the two spars were placed in
the fixture and bolted to the root fittings which attach the sup­
port arm to the rotating tower. The root fittings were pinned to
the assembly fixture. The trailing and leading edge sections
were then welded to the spar caps. Internal ribs and o t h e r m i s ­
cellaneous fittings were welded next. Miscellaneous fittings
included the in-plane attach lugs located at mid-span, the box
structure used for attaching the streamlined struts to the sup­
port arm, and the blade bearing support plate. The c over skins
were then tack welded to the spar caps, one side being completely
tack welded before tack welding of the remaining skin commenced.
Completion of the skin-to-spar welds was done very carefully by
intermittently filling in the weld beads between the original
tack welds to minimize warpage due to heating. A p a i r of support
arms were then placed in a separate fixture and strain relieved.
The strain relief fixture is an open box structure consisting
primarily of two large I-beams, connected b y six spacer bars on
top and bottom. Strain relieving was done according to PW-39,
Section I of the ASME Boiler and Pressure Vessel Code. The heat-
up and cool-down rates used are tabulated in Figure 29.

3.3 FIXED AN D ROTATING TOWERS - The fixed tower was fabricated

b y the Delta Steel Company of Jackson, Mississippi. This tower
is a standard size used for many o t h e r purposes throughout the
country. The pieces are standard steel angles cut to length on
an automated sawing machine. Holes for connecting fasteners are
automatically punched b y numerically controlled machines.

The rotating tower was fabricated in five sections by the

Custom Machine Company of Monroe, Louisiana. All sections were
standard size commercially available pipes. Flanges were welded
to the pipes to permit bolting sections together. The three m i d ­
sections were 1/4-inch thick, 24-inch diameter pipes with stan­
dard weld-neck flanges welded to each end. These butt-weld
joints were ground flush for maximum fatigue life. After welding
and strain relieving, the mating flange faces were machined to
insure flatness and alignment when assembled. Rings were fusion
welded to the pipes to provide attachment structure for the
support arms and streamlined struts.

29
 $

GP13-0821-3

FIGURE 28
SUPPORT ARM ASSEMBLY FIXTURE

30
 STEP INSTRUCTION TEMPERATURE (°F)

1 PREHEAT SLOWLY 0 -6 0 0

2 HEAT-UP RATE NO FASTER THAN 50°F/H O U R 600 - 925

3 HOLD FOR TEN HOURS 925 ± 25

4 COOL-DOWN RATE NO FASTER THAN 50°F/HOUR 925 - 600

5 COOLING IN STILL AIR 600 - AMBIENT

GP13-0821-2

FIGURE 29
STRAIN RELIEF CYCLE FOR SUPPORT ARMS

Upper and lower sections were much smaller pipes than used
in the mid-sections and therefore required large adapter plates
as well as gusset plates we l d e d to the pipe ends. Simple shop
tooling was used to insure perpendicularity of the plates with
the pip e centerline, Post-weld hea t treatment was in accordance
w i t h PW-39, Section I of the ASME Boiler and Pressure Vessel
Code.

31
 4. PHASE II - DESIGN CHANGES

During the fabrication of the various conponents and the

testing of the Giromill, some desirable design changes in the
configuration described in the Phase I report (Reference 1)
became apparent and the drawings, delivered at final design
review, wer e changed accordingly. In addition, continued control
system analysis finalized the controller gains and other con­
stants .

4.1 DESIGN CHANGE DESCRIPTIONS

Support Arms - Fabrication of the support arms revealed that

the sheet metal skins tended to "oil can" while b eing welded. To
alleviate this, additional support arm ribs were we l d e d to the
skins. Six additional ribs p e r support a r m were used.

The rib locations are shown on Va l l e y Industries Drawing

40213. The various rib shapes are shown on Valley Drawings 40520
through 40531 for the upper and lower support arms. The only
difference between the upper and lower support a r m ribs is a slot
in the lower ar m that fits around the pipe carrying the actuator
wiring.

In addition to installing additional support arm ribs, the

forming of the upper and lower support a r m skins was simplified.
Initially, the skins wer e to be rolled into their streamlined
shape. However, this was determined to be expensive and time
consuming. Instead, the skins were braked at seven places to
form the moldline. Each discrete ben d subtended an angle of
approximately 7 degrees. The effect on aerodynamic drag was
considered negligible. Valley Drawing 40213 was changed to
reflect this revision.

Rotating Tower - Strength and dynamics calculations were

b ased on using a 24 inch diameter b y 3/16 inch wall thickness
p ipe for the rotating tower. This was thought to be a standard
off-the-shelf pipe. However, we found it required a special
order. Rather than incur an extra cost and time delay an a vai l­
able 24 inch diameter by 1/4 inch wal l thickness pipe was
procured. Strength and structural dynamics checks of the greater
thickness tower wer e completed. They indicated a greater s truc­
tural safety margin and a more favorable structural frequency.
This change is reflected in Valley Drawing 40119.

Control Panel - A five foot square panel made from 1/4 inch
steel plate is used to mount the controller unit, CSPSU, and
generator contactor boxes. This panel is attached to one of the
horizontal torque members used between the lower main rotor
bearing and the fixed tower corner members. It was noted that
this panel swayed quite a bit w h e n it w a s windy. Therefore, an
additional brace was installed. This change is reflected in
Valley Drawing 40340.

32
 A voltage regulator assembly box was added to the panel,
when it was determined that an integral regulator was not
a vailable for the battery charging alternator. This change,
along wit h the voltage regulator wiring diagram, are shown in
V a l l e y Drawing 40340.

B lade Support Tube - The blade support tube connects the

blade center section w i t h an end section, either upper or lower,
and mounts the blade bearing that is connected to the support
arm. This tube must react to large loads and h e n c e is made from
h i g h strength steel. A 4130 steel was specified for this part.
However, it wa s unavailable in the form required. To prevent a
p r o g r a m delay, 4140 material billets wer e procured. This m a t e ­
rial has slightly better strength properties. The change is
shown in M C A I R Drawing 71A090151.

Blade Shims - W h e n assembling the Giromill it was determined

that tolerances could add up and cause the blades to rub against
the support a r m fairings. To increase the clearance, blade shims
w e r e made to fit b e t w e e n the b lade support tube mounting flange
and the blade rib, as shown in Figure 30. Shims ha v i n g a thick­
ness of 3/15 or 1/4 inch w ere used as required to provide
clearance. M C A I R Drawing 71A090101 shows that change.

Control S y s t e m - A t the completion of Phase I, the final

control system constants h a d not bee n determined. The final
values are tabulated in Figure 31. Figure 32 presents these
gains and other constants as used in the control system simula­
tion program. This is an updated version of Figure 108 of
Reference 1, the Phase I Final Report.

Several control system modifications were also done as a

result of the operating tests completed. These are discussed in
S ection 6.3.3.

Blade Bearing Lightning Protection - A lightning strike

could pi t the surface of the blade bearing. Therefore, p r o t e c ­
tion was p r o vided b y mounting a grounded pole on top of the
rotating tower. The pol e height was selected to provide a cone-
shaped protection zone over the blades.

The lightning pol e attachment failed during Giromill o p e r a ­

tion damaging a blade and two support arms (See Section 6.3.1).
Because of this carbon brushes w ith h i g h copper content wer e
installed to shunt the current around the bearing. The brush
installation is shown on M C A I R Drawing 71A090101. Removal of the
lightning p r otection pole is reflected in Val l e y Drawing 40102.

4.2 DRAWING SUMMARY - Figures 33 through 36 summarize the

pertinent drawings completed b y the four plants associated w ith
bu i lding the Giromill. These drawings contain the various
modifications described in Section 4.1, as w ell as corrective
actions, dictated b y the testing, w hich are covered in Section
6 .0.

33
 BLADE
SHIMS
BLADE
LEADING
EDGE BLADE RIB

BLADE WIPERS
(NO T USED)

SUPPORT ARM
F A IR IN G
SUPPORT
TUBE

BLADE SUPPORT
TUBE FLANGE

blade SHIMS BLADE SHIM S

FIGURE 30
BLADE SHIMS FOR CLEARANCE

34
 PARAMETER VALUE

CONTROL SYSTEM GAINS (CAM UNITS PER U N IT PERIOD)

START UP GAIN (K i x K4 ) ..................................................... -3 .6 6 2
OPERATING INTEGRAL GAIN (K 2 x K4 ) ........................... -1 7 .3 9 5
OPERATING PROPORTIONAL GAIN (K 3 x K4 ) ................ -2 3 4 .3 7 5
(29.98 < RPM < 3 2 .6 6 )
-58 5.9 3 75
(RPM > 32.66)

BLADE SPEED RATIO GAIN (K 5 ) (CAM UNITS PER

U N IT BLADE SP E E D ...................................................................................... 0.5

LIM IT CAM VALUES (Xp) AS A FUNCTION OF WIND VELOCITY

(h = 75 FT FOR STANDARD SEA LEVEL CONDITIONS)
Vw GREATER THAN 19 MPH ................................................................ Xp = 3.47
Vw GREATER THAN 17 MPH .................................................................. Xp = 3.85
Vw GREATER THAN 14 MPH ................................................................ Xp = 4.33
Vw GREATER THAN 10 MPH .................................................................. Xp = 5.78

CAM BIAS VALUE (XCMq ) ............................................................................. 0.8

L IM IT CAM VALUES (BLADE SPEED RATIO)

CAM 0 (AND CAM 1 ) .................................................................................. 1.45
CAM 10 ........................................................................................................ 5.78

OPERATING WIND SPEED, Vw (ONE MINUTE AVERAGE VALUE

RELATED TO h = 75 FT FOR STANDARD SEA LEVEL CONDITIONS) 10 MPH < V w < 4 0 MPH

START UP W IND SPEED (h = 75 FT FOR STANDARD SEA LEVEL

CONDITIONS)
LOW W IND ................................................................................................. ONE MINUTE ABOVE
13 MPH LESS ANY TIME
BELOW 10 MPH
HIGH W IN D ............................................................................. ................ ONE MINUTE BELOW
37 MPH LESS ANY TIME
ABOVE 40 MPH

ROTOR RPM DISCRETES

CONSTANT CAM START U P ....................................................... TO 7.17 RPM
OPEN LOOP START UP USING Ki K4 GAINS ...................... TO 29.98 RPM
GENERATOR CUT IN /C U T O U T ................................................. 32.92 RPM
N O M IN A L ........................................................................................ 33.5 RPM
OVERSPEED .................................................................................... 33.83 RPM

M AXIM UM NUMBER OF OVERSPEED CUT-OFFS W ITHIN AN

OPERATING P E R IO D ......................................................................... 5

ROCK ANGLE COMMAND LEAD ANGLE A T OPERATING RPM 15 DEG

QP13.0ia9-3

FIGURE 31
GIROMILL CONTROL SYSTEM CONSTANTS

35
 BACKEM F

= 0.00 05
A C TU A TO R MOTOR K b « 0.12 7 V /R A D /S E C
POWER J M = 0.0 0 1 1 3 3 S L U G - F r
JL = 3 .0 S L U G F T Z
Kt = a 0 9 4 7 9
SP2 = + 3.79.

A C TU A TO R
MOTOR SP1 = - 3 .7 9 c o > 3 3 .9 RPM N A* N a = 67.875
TORQUE (R E L E A S E BLAD ES)
La = 0
SP31 SP41 Ro = 0 .3 2 n -4 0

NA
^RO 'ro
SP41 SP31 B L A D E W E IG H T BLAD E AERO
C A L C U L A T IO N
C E N T R IF U G A L M O M . D A M P IN G

‘A F TA

0— BLAD E B LA D E A ER O
‘AF F R IC T IO N H IN G E M O M E N T
T , = 0.00 5
C A L C U L A T IO N
q C L A N D C Q = f(ae)
= 0.1

B LA D E E F F E C T IV E
5.9247 at
ANGLEOFATTACK
G „ = 1095 T Vw

B L A D E SPEED
= f(V w ) R A T IO C O M P U T A T IO N K3 = 4 6 .8 7 5 ,2 9 .9 8 < RPM < 32.1
CM
L IM IT K3 = 1 1 7 .18 75 , R PM > 3 2 .6 6
5.78.
ROCK A N G L E PR O ­
u> F IL E S = f( X p , 1//) R O TO R T O R Q U E
cr» (Q U A N T IZ E A T
3.47
t K4 = 0 , 0 < w < 7 . 1 7 R P M K2 = O . O ^ w < 2 9 . 9 8 RPM
C A L C U L A T IO N

12 .8 ms) ^ K4 = - 5 .0 , o ; > 7 .1 7 RPM K2 = 3 .47 9, o ; > 2 9 . 9 8 RPM P £ P = 0.0 5 8 2 3 6 SEC

CM,

Note: A T and Tg have m axim um value of

0.00052917 sec.
K i = 0 .0 < c o < 7 .1 7 R P M
has a m a xim u m value of 2.0.
K i = 0 .73 24 , 7 . 1 7 < w < 2 9 . 9 8
K^ = 0 , a ; > 2 9 .9 8 RPM

RPM SENSQR G E AR RPM GEAR

^^ = M E A S U R E D R O T Q R FREQUENCY P E R IO D -S E C
0 Q = R O T O R R O T A T IO N
R O T A T IO N R A TE (Q U A N T IZ E A T
R A TE = w 1^1 2 4 .3 x 1 6 2
58 .2 ms)

I g i r o = 7 6 ,3 5 0 SLUG-Fr
'r = 'Ro^'GEN ^9
Ip = 4 6 0 S L U G -F t2
' g ir o
lfiEN=D-22MSLUGFr N . = 54 .6 75 a ; < 3 2 .9 2
Ka = 1 2 .7 k W /R A D /S E C

K p n » 86 8 6 F T L B /R A D /S E C GENERATOR
RD
P O W E R -k W

K p s = 2 . 5 x 1 0 ° F T L B /R A D

0 .7 4 6 c j.

FIGURE 32
GIROMILL CONTROL SYSTEM CSMP SIMULATION-CONTROLLER
 DRAWING
NUMBER REVISION TITLE

71A090000 Giromill Structural Arrangement

71A090101 Blade Top Assembly
71A090120 Blade Assembly
70A090121 Blade Leading Edge
70A90122 Blade Trailing Edge
71A090123 Blade Channel
71A090124 A Blade Rib Leading Edge Machining
71A090125 A Blade Rib Formed Trailing Edge
71A090150 A Support Tube Assembly
71A090151 A Support Tube Machining
71A090160 A Support Tube Rings
71A090170 Bearing Support Upper
71A090171 A Bearing Support Lower
71A090180 A Sprocket Blade Drive
71A090190 Ballast, Fixed
71A090191 Ballast, Adjustable
71B090200 Bearing, Roller, Self Alignment -
Spec. Control Drawing
71B090205 Belt Drive - Spec. Control Drawing
71A090211 Supports - Fairing
71A090212 Skins, Fairing - Leading Edge and
Actuator Cover
71A090213 Skins, Fairing - Hub and Trailing Edge
71A090214 Seals - Blade Wiper
71T090000 Test Instrumentation - Giromill Blade

FIGURE 33
MCAIR BLADE DRAWINGS

37
DRAWING
NUMBER REVISION TITLE

J40G0G B1 4GKW Giromill

D4GGG1 Foundation
J4GGG3 A1 Fixed Tower Assembly
D40G52 B1 Lower Bearing to Fixed Tower Intf/Assy
D4G1G2 B1 Rotating Tower Assembly
D4G186 B1 Rotating Tower Assembly
D4G186 B1 Mechanical Drive System Assembly
D4G2G5 B1 Electric Drive System Assembly
D4G213 B1 Upper and Lower Support Arms
D4G253 B1 Brake System and Assembly
D4G3G1 A1 Speed Sensor Assembly
D4G34G Cl Wire Assembly Interface
D4G387 A1 Control System Gen. Drive Assembly
D4G389 B1 Control System Slip Ring Assembly
D4G392 B1 Upper Bearing to Fixed Tower intf/Assy
B4G4G1 Electrical Power System Design 1
A4G423 Wind Speed and Direction Sensor Assembly
D4G5G2 Rotor Assembly Fixture Foundation
D4G5G3 Rotor Assembly Fixture

FIGURE 34
VALLEY TOP ASSEMBLY DRAWINGS

38
 DRAWING
NUMBER REVISION TITLE

H05G0075 A Control Unit - Giromill

PLH05G0075 A Parts List
H05G0084 A Power Switching Unit
PLH05G0084 A Parts List
H06G1863 A Circuit Card Assembly Logic - A1
PLH06G1863 A Parts List
H06G1864 A Electronic Component Assembly
PLH06G1864 A Parts List
H08G1735 Wiring Diagram - Giromill Control System
H09G2325 Schematic Diagram - Giromill Control Unit
H09G2326 Schematic Diagram - CSPSU
H11G0264 Envelope Drawing, Control Unit
H11G0266 A Envelope Drawing, CSPSU
H14G3393 A Cable Assy, Special Purpose, Electrical
PLH14G3393 A Parts List
S52H0059 Housing - Giromill Control Unit
S52H0060 A Housing Giromill CSPSU
S76M0109 A Plate, Mounting Component
S76M0110 Plate, Mounting Component
S76M0119 Plate Mounting, Transistor
S76B0065 A Plate, Identification
S76B0064 A Plate, Identification
H12G0867 Acceptance Test Procedure - Control Unit -
Giromill P/N H05G0075
H12G0868 Acceptance Test Procedure - Power Switching
Unit - Giromill P/N H05G0084
H77D0014 Microcircuit Specification Giromill Control
Unit

FIGURE 35
GIROMILL CONTROL SYSTEM DRAWINGS

39
DRAWING
NUIffiER REVISION TITLE

7010A 6 Servomechanism - Giromill Blade Control

7010A0001 6 Servomechanism - Giromill Blade Control
7010-0002 4 Housing - Cover
7010-0003 5 Housing - Main
7010-0004 2 Gear - Spur
7010-0005 2 Gearshaft. Spur
7010-0006 3 Gearshaft. Spur
7010-0007 1 Gear, Spur
7010-0008 2 Gear Cluster, Helical
7010-0009 2 Gear, Helical
7010-0010 2 Pin, Hollow
7010-0011 2 Cover, Access
7010-0012 1 Cover, Access
7010-0013 1 Cover, Access
7010-0014 1 Plate, Mounting Potentiometer
7010-0015 2 Base, Amplifier
7010-0016 2 Cover, Amplifier
7010-0017 1 Heat Sink, Electrical Electronic Equipment
7010-0018 1 Gasket
7010-0019 1 Spacer
7010-0020 2 Guide - Card
7010-0021 1 Bushing
7010-0022 1 Insulator
7010-0023 2 Motor Assembly
7010-0024 1 Potentiometer Assembly
7010-0025 2 Motor, Direct Current
7010-0026 1 Terminal
7010-0027 2 Diagram, Schematic Electrical
7010-0028 2 Terminal, Electrical
7010-0029 2 Printed Circuit Board Envelope

FIGURE 36
GIROMILL ACTUATOR DRAWINGS

40
 DRAWING
NUMBER REVISION TITLE

7010-0030 1 Clamp, Loop

7010-0031 2 Diagram, Connections, Interconnect Board
7010-0032 1 Screw, Machine
7010-0033 2 Printed Circuit Board Assembly Amplifier
7010-0034 2 Printed Circuit Board Assembly Driver
7010-0036 1 Bracket, Mounting Capacitor
7010-0037 1 Bracket, Mounting Capacitor
7010-0038 1 Screw, Machine
7010-0039 1 Shield-Electromagnetic Interference
7010-0040 1 Spacer - EMI Shield
7010-0041 1 Potentiometer
7010-0042 1 Shield - Electromagnetic Interference
7010-0043 1 Shield - Electromagnetic Interference
7010-0044 1 Armature Assembly
7010-0045 1 Art Work - Interconnect Board, Back Side
7010-0046 1 Art Work - Interconnect Board, Component Side
7010-0047 1 Interconnect Board Assembly
7010-0048 1 Interconnect Board
6917-0002 3 Schematic, Servo Driver Board
, 7010-0033-01 3 Parts List - Printed Circuit Board Assembly
Amplifier
PL 7010-0034-01 3 Parts List - Printed Circuit Board Assembly
Driver
PL 7010A001-01 7 Parts List - Servomechanism - Giromill Blade
Control
PL 7010-0047-01 1 Parts List - Interconnect Board Assembly
PL 7010-0023-01 1 Parts List - Motor Assembly
PL 7010-0024-01 1 Parts List Potentiometer Assembly
ATP7010A 1 Acceptance Test Procedure

FIGURE 36 (Continued)
GIROMILL ACTUATOR DRAWINGS

41
 5. UPDATE OF MANUFACTURING COST ESTIMATES

5.1 PROTOTYPE COST ESTIMATES - The preliminary estimates of

production costs, itemized in Section 15 of Reference 1, are
still valid, except for the support arms. The estimates are for
the use of the Giromill as an electric power generator tied into
an electric utility grid. The basic ground rules under w h i c h the
estimates were made and an itemized list of the estimates are
shown in Figures 37 and 38. Learning curves applied to direct
labor and quantity discounts for vendor items, specified in
Reference 1 for various components, remain unchanged.

The prototype cost estimates shown in Figure 38 are h i g h e r

than for units designed for production. For a prototype design
m any parts are designed for expediency, i.e., the quickest and
least expensive method for one unit rather than for several units
in a long p roduction run. For example, the blades w ere made from
formed sheet metal parts, joined b y rivets, because this is the
least expensive method for a single unit. For long p r oduction
runs, an extruded leading edge shape w o u l d be more economical.
The initial tooling cost and long lead time for such a part,
however, made it an impractical choice for the prototype design.

5.2 PRODUCTION DESIGN CONSIDERATIONS - The cost of the support

arms exceeded original estimates because of difficulties
encountered during manufacturing and because the large number of
small pieces fabricated and w e l d e d on assembly were more labor
intensive than anticipated. In a company sponsored effort,
several of the more expensive conponents, including the support
arms, are b eing reviewed for possible cost saving changes to be
incorporated into a production design. Several of the proposed
cost reducing designs are discussed in the following paragraphs.

Blades - The production version of the blades w o u l d h ave an

extruded leading edge section to replace the sheet metal leading
edge and spar. The trailing edge would be either injection
molded plastic sections or formed aluminum much like the present
configuration. Ribs are no longer needed since blade bending is
transferred to the support tubes through socket action in the
leading edge extrusion.

T h e blade support tubes were redesigned to utilize standard

size pipes and thereby eliminate costly machining operations.

Support Arms - The current design is made from many formed

and wel d e d parts. It would be replaced b y a standard steel pipe
w ith fittings welded to the root end for attaching to the rotat­
ing tower, and to the tip for mounting the blades and actuator.
A simple aerodynamic fairing w ould be we l d e d to the outer one-
h alf of the support ar m to reduce drag. The effect on cost e sti­
mates of this design on 1000th unit cost is shown in Figure 38.

42
• ALL COSTS IN 1977 DOLLARS
• COSTS INCLUDE G&A AND PROFIT
• MARKETING AND TRANSPORTATION COSTS NOT INCLUDED
• FOUNDATION AND ERECTION COSTS NOT INCLUDED
• COMMERCIAL STEEL FABRICATORS TO BUILD ENTIRE U N IT
• RDT&E AND TOOLING COSTS NOT INCLUDED
• ROTOR CENTERLINE PLACED TO PROVIDE A 30-FOOT GROUND CLEARANCE
• CUT-OUT WIND VELOCITY IS 40 MPH. ALL OTHER REQUIREMENTS ARE AS
SPECIFIED IN REFERENCE 2
• APPROPRIATE LEARNING CURVE APPLIED FOR EACH COMPONENT
PART OF GIROM ILL
QP13-0S61-4

FIGURE 37
GROUND RULES FOR ESTIMATING COST OF 1,000th UNIT

1000th UNIT
1st UNIT
A
FIXED TOWER $ 5,961 $ 3,636
ROTATING TOWER 11,686 4,006
SUPPORT ARMS 28,541 9,878 3,980
STREAMLINE RODS 1,556 467
BLADES 13,693 4,944
UPPER BEARING 878 263
LOWER BEARING 1,624 487
CONTROL SYSTEM 7,039 4,084
SPEED INCREASER 3,810 1,143
MAIN DRIVE PULLEY 559 167
MAIN GENERATOR PULLEY 194 58
MAIN DRIVE BELT 89 27
INDUCTION GENERATOR 1,060 318
ELECTRIC COMPONENTS 369 111
TOTAL MATERIAL, LABOR, OVERHEAD $77,059 $29,589 $23,691
G&A (7%) 5,394 2,071 1,658
PROFIT (10%) 8,245 3,166 2,535
TOTAL $90,698 $34,826 $27,884
DOLLARS/KILOWATT (41.7 kW) $ 2,175 $ 835 $ 669

QP13-0581-7
/ ? \ R edesigned s u p p o rt arm s

FIGURE 38
GIROMILL BUDGETARY COST ESTIMATE
1977 Dollars

43
 Control System Modifications - In the cost estimate for the
controller presented in the Phase I report, it was assumed that a
circuit board would replace the wired-up board of the prototype
unit. The same assumption was made in the cost estimate of the
production version. Additional control modifications would
consist of eliminating the batteries and charging alternator for
machines tied into the grid. The controller and actuators would
use grid power. This means that these would have to be rede­
signed for grid power. However, analysis shows the changes would
not appreciably affect the presently estimated controller and
actuator costs.

General Cost Reduction - In addition to the redesign of

various components for production, o t h e r changes are also being
considered. Analysis to date has shown that a two bladed rotor,
with the same solidity and RPM as the present rotor, would be
more cost effective than the current 3-bladed rotor. Cost is
reduced by eliminating two support arms, one blade, and one
actuator.

Many of the structural components were designed by stiffness

requirements to provide satisfactory vibration, flutter, and
structural response characteristics. Because this provided
additional margins of safety for structural considerations, the
machine could be uprated to a higher capacity output with little
o r no structural changes. Initial indications are that the
rating could be as high as 60 k W in a 23 mph wind. Both cost of
energy and cost pe r kilowatt installed would therefore be
decreased.

44
 TESTING

The control system, the controller, and actuators, were

subjected to a series of b ench acceptance tests p r i o r to the
installation of the units in the Giromill. The Giromill system
was also p u t through an acceptance test p r i o r to deliveiry to
Rocky Flats.

Testing at Rocky Flats commenced w ith first turn without the

generator connected in July 1980. Development problems and lack
of w i n d delayed first turn w i t h the generator connected to the
gri d until September. The unit w e n t t o unattended operation in
November. O ver the Thanksgiving holiday, while running una t ­
tended, the lightning pole fell, damaging a b lade and two support
arms. The rotor was removed from the tower and placed on the
rotor assembly fixture. The damaged b lade and support arms wer e
t hen removed for repair. Ove r the Christmas holiday, the
assembly fixture stand pulled out of the concrete in a h i g h wind,
and the remaining blades and support arms wer e damaged. The unit
was repaired, reassembled, and put back in operation in April
1981.

Development and data collection problems and the inconsis­

tency of the w i n d h ave combined to limit the amount of data
collected. To date not enough consistent data has b e e n obtained
to define the performance of the unit. Also, the mechanical
drive system has not yet bee n tried.

6 .1 CONTROL SYSTEM ACCEPTANCE TESTS - Control system acceptance

tests w e r e done in two phases. The first p hase was acceptance of
the controller unit; the second was acceptance of the actuators
as they w e r e driven b y the controller. Both of these acceptance
tests were conducted at MDEC - Grand Rapids Division.

The controller acceptance test procedure was conducted on 10

October 1979. It was conducted pe r the procedures outlined in
MDEC drawing H12G9867. No major discrepancies w e r e found.

The actuator acceptance tests along with integrated

controller/actuator tests were conducted on 12 December 1979.
T h e actuator acceptance tests wer e conducted according to MDEC - G R
d rawing ATP7010A. No major discrepancies in actuator operation
w e r e found from these tests. However, some electronic adjust­
ments wer e made that smoothed out the actuator running and
d ecreased the p o w e r required. The biggest was an electronics
revision of the second order rock angle command filter time
constants of 0.035 and 0.040 to 0.055 and 0.0 20. This reduced
the actuator p o w e r required b y half, through elimination of h i g h
current spikes.

T h e controller unit and actuators were shipped to Va l l e y

Industries Plant in Tallulah, L a . , on 19 December 1979,

45
6.2 GIROMILL SYSTEM ACCEPTANCE TESTS - The Giromill was erected
and subjected Eo a series o? acceptance tests at the Valley
Industries Plant at Tallulah, La. It was structurally completed
and fully erected on 25 February 1980. Control system and instru­
mentation wiring, alignments, and non-rotor rotating tests of the
control system wer e completed b y 27 February. Emergency brake
h y draulic problems delayed first start-up until 5 March. The 40
k W generator was not connected to the local utility grid for
these tests. Since only a limited amount of testing was going to
b e done at Tallulah, it was not cost effective to spend the man-
hours and money to run a special grid connect line to the unit.

Instrumentation consisted of a 12 channel Honeywell

Visicorder, digital RPM indicator, and a w i n d speed and direction
indicator. Basically the three blade rock angle commands, the
three blade actuator feedback values, and the two tower a ccel­
erometer values wer e always recorded. Additional parameters wer e
recorded as necessary depending on the test. The vibration
sensor cut-off located on the top of the fixed tower was set at
1.0 g.

For first start the control system was configured to give a

blade modulation profile corresponding to c a m 0 (see Figure 13).
This modulation profile will normally limit rotor RPM to about 20
in a 20 MPH wind. The winds w ere from the N N W at 10-15 MPH. The
Giromill ran up to about 15 RPM in 30 seconds. A f t e r it ran for
about 60 seconds, power to the actuators was cut and the Giromill
coasted to a stop. Several similar runs w e r e completed to check
the emergency brake system and battery charging alternator. Run
6 was configured to give normal operation and develop full RPM.
The Giromill came up to speed and operated exactly as programmed.
Full operational RPM was maintained without even h a v i n g the
generator connected.

Approximately 30 additional runs were completed over the

next several days to check out the various s y s t e m s .

F r o m 11 M arch through 13 March 1980, the Giromill was demon­

strated to be in compliance with the checkout requirements.
Authority to ship the Giromill to Rocky Flats was received on 13
March 1980.

6.3 OPERATIONAL TEST - The Giromill was erected and functionally

checked at Rocky Flats during M a y and June 1980. First turn
without the generator being connected was accomplished on 3 J uly
1980. Various problems, mostly w ith the actuators and grid
connect electrical system, and lack of w i n d delayed first turn
wit h grid connect to 15 September 1980.

This section summarizes the problems and corrective actions

taken during the tests at Rocky Flats.

46
 6.3.1 Structural

Blade Seals - To prevent an aerodynamic performance loss due

to ai r flow between the blade sections separated b y the support
arms, blade seals were designed and fabricated. Wipers were
installed on the blades that rubbed ove r the fiberglass support
arm fairings. Due to construction tolerances and blade deflec­
tions u nder load, the wiping surfaces were not always perfectly
aligned. This caused several of the blade wipers to bind and
create high friction. For the prototype Giromill, all blade
wipers were removed before beginning operational testing. In
addition, blade shims were installed to increase the clearance
between blade and fairing (see Section 4.1).

It has bee n estimated that a performance degradation of 10%

could result from not having seals between blade sections. To
evaluate the seals' effect, simple temporary seals made from tape
should be installed and the performance compared.

Blade Alignment - Self-aligning bearings are used to allow

for blade deflections caused by manufacturing tolerances and
blade loads. The largest contributor to the possible m i s align­
ment of a blade is the deviation of the radial centerlines
between the upper and lower support arms. This deviation will
cause a blade to be off vertical.

Analyses were performed to establish the allowable tolerance

of this deviation b y determining the effect of this condition on
performance and system dynamics. For centerline deviations
causing a one o r two degree blade misalignment, no detrimental
effects could be calculated. It was therefore decided that the
total allowable angu l a r tolerance between support arm centerlines
would be arbitrarily specified to lie just within the bearing
misalignment capability. Since for all practical purposes, the
angular tolerance between the support ar m centerlines is domi­
nant, that tolerance should be less than the 1.0 degree bearing
misalignment capability.

Whe n the Giromill was assembled, it was visibly noted that

Blade 3 was not vertical. Measurement showed it was off by 0.75
degrees, which is within the tolerance specified. However,
because of the visible nature of this deviation, production Giro-
m ills should attempt to hold the upper and lower support arm
centerline tolerance closer.

Battery Charging A l t e r n a t o r Attachment Bracket - The battery

charging alternator is mounted on two brackets attached to the
speed increaser. Two slots in each bracket a llow for alternator
drive belt adjustment. These slots are 5/16 inch wide. The
bolts through the slots are 1/4 inch. The resultant freeplay
allows the a l ternator to tilt, loosening the belt and causing it
to ride on the pulley rim.

47
 A permanent fix is to make the brac k e t slots 1/4 inch wide.
A temporary fix w a s made by shimming the alternator to prevent it
from tilting.

L i ghtning Rod Structural Failure - A standard 30 foot

commercial utility light pole w a s mounted on top o f the rotating
tower to provide lightning protection for the blade b e a r i n g s b y
providing a 90® (included angle) cone o f protection. The pole
specification required it be fastened w i t h four 3/4 inch b o l t s and
double washers. W h e n the pole w a s delivered it w a s found to
contain four 1-1/4 inch holes in the mounting flange. Spacers
were made that had a 3/4 hole and fit snugly into the 1-1/4 inch
mounting holes. The pole w a s mounted on top o f the rotating tower
using 3/4 inch b o l t s through the spacers, and fastened w i t h one
w asher and nut per bolt.

A s mentioned, while the Giromill w a s running unattended over

the Thanksgiving holiday, the lightning pole came off, damaging
one blade and two support arms.

The rotor w a s removed from the tower and placed on the

assembly fixture. The damaged blade and support arms w ere then
removed.

A visual inspection revealed that the nuts had pulled through

the washers, allowing the bolts, nuts, and spacers to pass through
the holes in the base o f the pole and remain w i t h the rotating
tower as shown in Figure 39. The wash e r s were partially extruded
into the pole base holes, as shown in Figure 40, indicating that
the washers failed in shear. In addition, paint w a s chipped along
the weld line attaching the pole tube to the base plate indicating
possible high oscillatory loads.

The pole w a s analyzed as a cantilever beam, fixed at the root

b y four 3/4 inch bolts. Bending loads on the pole were assumed to
be caused b y air drag (W) on the pole. Air drag load w a s
calculated by the formula;

W = Cd q S

where W = total air drag load on the pole

= drag coefficient for a cylinder conservatively assumed
to be 1.2)
q = pressure loading in po u n d s per square foot
= V^2/295 = velocity o f w ind in k nots = )
S = plan form area o f pole

In order to analyze the failure o f the washers, bol t loads

were calculated for several win d velocities up to 125 MPH. The
bolt pattern at the pole-to-tower interface and a cross-section
through an attach bol t is shown in Figure 41. Bolts w ere assumed
to r e a c t to the p o l e root b e n d i n g m o m e n t as a c o u p l e b e t w e e n b olts
at diagonally opposite corners. Figure 42 lists, for each incre­
mental w i n d velocity, the total drag load, root bending moment on
the pole, and the m a x i m u m individual bolt load.

Bending and shear stresses on the w a s h e r are p l o t t e d versus

w i n d velocity in Figure 43. Stresses wer e calculated assuming
that the w a s h e r wa s divided into six equal segments, each segment
b e i n g fixed at its intersection w ith the edge of the nut face.
Thickness of the w a s h e r used in the analysis wa s scaled from an
actual part. Bending of each segment was caused b y an upward
force on the w a s h e r applied b y the pole bas e as shown in Figure
41.

G P13-0189-35

FIGURE 39
LIGHTNING POLE-TO-ROTATING TOWER INTERFACE JOINT

49
 QP13-0188-34

FIGURE 40
WASHER PARTIALLY EXTRUDED INTO HOLE

T h e location of the load is based on measurements of m a r k ­

ings on the failed washer. As the wa s h e r deflects under incre as­
ing load, this point of load application w o u l d move towards the
bolt centerline, reducing the eccentricity for bending and
approaching a pure shear loading. Bending stresses calculated in
the manner described w ill therefore be h i g h e r than the actual
stresses. (The actual failure mode of the wa s h e r was in shear,
as shown in Figure 40.)

At a win d velocity of 40 mph, the Giromill is prog r a m m e d to

shut down. At this w i n d velocity, there is ample margin of
safety on y ield strength, as shown in Figure 42, for bot h shear
and bending. A positive margin of safety exists on y ield
strengths for bending up to a w ind velocity of 60 MPH. A posi­
tive margin of safety exists on y ield strength for shear stresses
up to a w i n d velocity of 130 MPH.

50
 11.0

LIGHTNING POLE BOLT PATTERN

Pe l e m
(TYPICAL 6 PLACES)

0.63
2.0

1.065
NUT
(PART NO. 40361

WASHER
0.15 (PART NO. 40515)
(TYP)

LIGHTNING POLE
(PART NO. 40352)
0.125

ELEM ELEM
0.875

SPACER
(PART NO.
40508) 1.0
1.25

ROTATING TOWER TOP SECTION

(PART NO. 40133)
3/4 IN. DIA BOLT (PART NO. 40348)

SECTION A-A QP13.0S81-S

FIGURE 41
LIGHTNING POLE-TO ROTATING TOWER INTERFACE GEOMETRY

51
 ^W IND W ■v ir g g t Pb g l t
(MPH) (LB) (IN.LB) (LB)

20 18 2,767 252
40 71 11,072 1,007
60 159 24,913 2,265
80 283 44,290 4,026
100 442 69,205 6,291
125 691 108,133 9,830
GP13-0561-3

FIGURE 42
LIMIT LOADS ON LIGHTNING POLE vs WIND VELOCITY

160

120

BENDING STRESS

STRESS
IN
WASHER 80
KSI ULTIMATE TENSILE
STRENGTH (F^u)

YIELD TENSILE STRENGTH (F _ )

40

ULTIMATE SHEAR STRENGTH (Fju

YIELD SHEAR
STRENGTH (F „ ,)
SHEAR STRESS

40 80 120 160
W IN D -M P H
QP13-0S61-6

FIGURE 43
WASHER BENDING AND SHEAR STRESS

52
 The location of the pole on the ground after the accident
indicated that the attach bolts failed w hile the Giromill was
running. The w i n d speed at the time of failure was proba b l y
40-45 MP H (possibly 5 MPH over cut-off s p e e d ) . Since consider­
able margins of safety exist for the joint at this w i n d velocity,
it is speculated that dynamic forces added to the wind drag loads
on the pole to cause failure. These forces could have resulted
from a soft spring constant o f the joint, from a loose joint at
the pole base caused b y bending o f the wash e r s from p r e vious high
win d loads, from spacers too long to allow proper seating o f nuts,
or from insufficient torque on the nuts.

I f the joint at the root w a s indeed loose or essentially a

soft spring, additional forces on the joint could be caused by the
rotating motion o f the pole in a deflected position.

Based on the preceding analysis and the uncertainty of

dynamic loadings in addition to static w ind loadings, it w a s
decided to eliminate the need for a lightning arrestor pole. This
w a s accomplished b y installing carbon b r u s h e s on the upper support
ar m at the blade support tube. T h i s p r o v i d e s a grounding p a t h for
lightning current from the blade through the support arm.

6.3.2 Mechanical

Manual Brake Control - An emergency brake is provided con­

sisting o f a caliper acting on a brake disc attached to the
rotating tower. The caliper is k e p t open b y hydraulic pressure so
that any power loss actuates the brake.

During Giromill checkout it became apparent that m a n y times

we wanted to shut the Giromill o f f without also actuating the
brake. Therefore, a manual b rake switch w a s mounted in the
contactor enclosure. T his switch, when connected to the brake
control connector, bypassed normal controller control for the
b rake and allowed manual control. Thi s device worked very well
for the variety o f testing that had to be done.

A c t u a t o r Pu l l e y L o s s - The actuators are connected to the

b l a d e s by a toothed bel t and pulley system. Whe n the Giromill is
running, the blade is continually being cycled bac k and forth by
the actuator. Thi s cyclic motion has resulted in the actuator
p u l l e y s coming loose and falling off. These p u l l e y s are held on
by a taper lock and keyway. The taper lock specification requires
torquing the set screws to 175 in-lb while tapping the taper lock.
To obtain the torque value is difficult, since the actuator will
turn while the screws are being tightened. A strap wrench, that
fits around the pulley, w a s used to hold the actuator while
tightening the set screws.

53
 A c t u a t o r Belt Slipping - Once during checkout it w a s noticed
that two bl a d e s were not correctly aligned, in that a zero rock
angle command did not result in the blade being at zero. Check­
ing revealed that the toothed blade actuator b e l t s had slipped
several teeth. T his w a s attributed to b e l t stretch and the
inability to tension the belt correctly.

The b elt specification indicates that w hen the b elt is first

installed it will stretch somewhat, and should be retightened
after a few days. This should become a standard operating
procedure.

To enable measurement o f bel t tension a b e l t access hole w a s

drilled in the support arm, as shown in Figure 44. This hole is
located near the bel t mid span, and b y inserting a 1/4 inch rod
through the hole, belt tension and deflection can be measured
using a b e l t tension tester such as is shown on Page 30-55,
Reference 3. The force and deflection as measured b y the tester
should be 10 lbs for 0.44 inches deflection. T h i s should give 500
lb tension in the belt.
LEA D IN G EDGE
BELT TENSION
M EASUREM ENT
BLADE A C TU A TO R
COOLING DUCTS HOLES
FA IR IN G
FRONT SPAR

BLADE
PU LLEY

BLADE BELT
ACTUATOR BLADE

PULLEY

REAR SPAR

T R A IL IN G EDGE

FIGURE 44
BLADE ACTUATOR BELT TENSION MEASURING ACCESS HOLE

54
 6.3.3 Control System

A c t uator Gea r Box Oil Seepage - On inspection of the parts

after they h a d b e e n shipped to Rocky Flats, one blade actuator
showed evidence of leaking oil. Disassembly showed that the oil
h a d leaked pas t the motor oil seal. All actuators w ere modified
b y adding another m otor oil seal and providing an external drain
h o l e b e t w e e n the two seals.

A fter running about one month, oil leakage from the act u ­
ators was again noted. This was found to b e coming through the
gear case mating seal. The g loads at the actuator were suffi­
cient to exude oil through the porous seal. The fix was to
install non-porous seals in all actuators.

Inadvertent Controller Shutdowns - The first several weeks of

normal running w ere plag u e d by a controller command shutdown
caused b y h a v i n g the controller/actuator self-check tolerance too
tight. This self-check of the controller/actuator electronics is
p r o vided to shut down the Giromill immediately w h e n e v e r an anomaly
is detected, preve n t i n g a dangerous situation from developing.
This is part of the failure detection system p r ogrammed in the
controller.

The shutdown p r o b l e m was attributed to the self-check on h o w

w e l l the actual rock angle was following the commanded value. The
shutdown occurred w h e never a near-zero rock angle was commanded
for an appreciable p o r t i o n of the blade orbit. The checking logic
and the tolerances wer e modified to eliminate this problem. The
o ther failure detection functions wo r k e d as plan n e d and on several
occasions shut down the Giromill w h e n a failure occurred.

In addition, the controller logic was changed to k e e p the

controller on w h e never a no-go shutdown occurred. This allowed
diagnostic checks to be run to isolate the p a rticular self-check
that h a d caused the Giromill shutdown.

A l s o some m inor changes in the initial controller turn-on

diagnostic w e r e made. These changes resulted in a more orderly
c omputation scheme.

W i n d Direction Time Constant Reduction - Initially a filter

h a v i n g a time constant of 500 ms was used in the w i n d direction
controller circuit. This filter caused the rock angle commands to
be erratic w h e n the w i n d was from the direction w h e r e the 3° w i n d
direction sensor potentiometer dead ban d was located. T h e fix was
to essentially eliminate the time constant filter circuit,
changing it t o a 2.2 ms value. This solved the erratic rock angle
commands problem. However, the long ter m effect on actuator life
due to following all rapid w i n d direction fluctuations w ill have
to be determined.

An alternate way of correcting the erratic rock angle

commands w o u l d b e to install a w i n d direction sensor h a v i n g dual
potentiometers. N o dead zone w o u l d therefore exist.

55
 Controller Processor Phase Sequence - Start-up in a low wind
w a s found sometimes to take more than the 120 seconds allocated.
Also when coasting down from an overi^eed or aborted start,
weathervaneing friction w a s sometimes sufficient to k eep the
rotor turning about 15 to 16 RPM. Thi s w a s especially true if
the wind w a s about 9 MPH. To maintain better control o f the
operation, values o f the phase sequences wer e modified as shown
in Figure 45.

CONDITIONS FOR
ACTUATOR GRID NEXT TIME SHUTDOWN
PHASE RPM RANGE GOING TO
POWER CONNECT PHASE LIMIT CONDITIONS
NEXT PHASE

0 - OFF NO 1 PROCESSOR -
POWER TURN ON

1 - OFF NO 2 INITIALIZE COMPLETE SELF TEST FAILURE

AND SELF-TEST GO

2 CO < 7 RPM ON NO 2A ROTOR HAS 120 SEC TIME LIMIT

FORWARD ROTATION

2A CO < 7 RPM ON NO 3 FIRST RPM SENSOR TIME LIMIT

FLAG

< 32.95 RPM NO 6 TIME LIMIT 240 RPM FROM RPM SENSORS
3 CO ON
SEC
4 c j > 3 2 .9 2 RPM (1) RPM FROM ROTOR ANGLE POT

4 32.92 < CO < 33.83 ON YES 3 cu <32.92 RPM NO RPM SENSOR FLAG
LIMIT PERIOD > 6 4 MILLISECONDS
5 CO >33.83 RPM

5 33.83 > CO > 32.92 OFF YES 6 CO<32.92 RPM 60 SEC TIME LIMIT

6 32.92 > CO > 1 5 RPM OFF NO 3 co< 19.7RPM<3) 10 MIN TIME LIMIT OR PHASE 6
(2) ^PHASES LOOP COUNT > 5

(1) Was 120 sec

(2) Was 60 sec
(3) Was 15rpm
FIGURE 45
GIROMILL CONTROL SYSTEM - PROCESSOR
Phase Sequence

Control System Power Supply - Two failures o f the control

system power supply were experienced. The first w a s diagnosed as
a short in the power input circuit. The second w a s a loss o f the
5 vdc output circuit.

Since b oth o f the failures wer e different, and no evidence

o f external causes existed, they were attributed to quality
control in power supply manufacture. No Giromill changes were
made.

Battery Charging Alter nator Marginal - Cont i n u o u s running o f

the Giromill resulted in a slow discharge o f the batteries. It
appeared that the brake solenoids and actuators are pulling more
power than calculated, and that the alternator output is slightly
low. T h i s resulted in a discharge o f the ba t t e r i e s and eventual
shutdown o f the Giromill.

56
 The production fix will be to install a larger alternator
for stand-alone Giromill systems, and use a grid driven power
supply (eliminate b a t t e r i e s and alternator) system connected to
the power grid.

A temporory fix w a s made to the unit at Rocky Flats b y

installing battery chargers to supplement the present alternator.

A c t uator Power Transistor F a i lures - Several power transis­

tors, located on a heat sink attached to the actuator amplifier
case, failed during the test. Mos t o f these failures w ere in
conjunction w i t h other failures or anomalies. Power transistor
failures were evident after an actuator motor had jammed due to a
ma g n e t chip breaking off. Another case o f power transistor fail­
ure w a s when a m o t o r short occurred, p r o b a b l y caused b y pulling
on the w i r e s w hen replacing the actuator gear case gasket.
Anot h e r occurred on the actuator where the blade b e l t had slipped
three cogs, w h i c h could have created large actuator loads.

There were other power transistor failures where no cause

w a s evident. Analysis has indicated that the actuator electron­
ics m a y be experiencing a low voltage, causing loss o f electronic
control. Low voltage can occur when the actuator mo t o r s draw a
h i g h current, causing a large voltage drop through the power slip
rings.

The actuators and wiring wer e modified so as to have a

separate power line to the actuator electronics not affected b y
the m otor current variations. T h i s is expected to eliminate the
power transistor failures. Verification o f this fix has not been
completed at this writing.

P r emature Over speed Shutdown - Whe n first running the

Giromill w i t h the generator connected to the grid, the controller
w o u l d sometimes shut down the Giromill, signalling an over speed
condition w hen none had occurred. Analysis showed that drive
train flexibility could cause an occasional overspeed indication.
This is because the speed increaser/generator drive sprocket is
used to get an RPM indication b y relating to the time to count
125 teeth o f the sprocket.

The fix w a s to prog r a m that five consecutive indications had

to be present to trigger an overspeed shutdown. Since an RPM
me a s u r e m e n t is obtained every 0.058 second (at synchronous RPM),
this m e a n s a delay o f only 0.24 second in initiating a shutdown
in event o f an actual overspeed. During this time the rotor RPM
would increase 0.3 RPM, calculated assuming the existence o f
m a x i m u m rotor torque and no generator load. Thi s m u c h increase
does not create a critical situation.

57
 6.3.4 Electric Generation

Grid Connect Relay Chatter - During first attempts to con­

nect the generator to the grid, a grid connect relay chatter
occurred. This in turn caused the controller to sense an error
signal, shutting off the Giromill. The chatter was attributed to
the sensitivity of the RPM measurement and the fact that there
was no RPM tolerance between grid connect and disconnect.

A temporary fix was to install a time delay relay which

prevented grid disconnect for 1.5 seconds after disconnect
command. The final fix was to essentially incorporate that time
delay in the controller program. This solved the grid connect
relay chatter, but does allow the generator to drive the rotor
for that short time.

Generator Circuit Breaker Tripping - At times the grid

connect contactor heaters will trip. This causes p o w e r to drop
off line. It occurs w h e n the win d is marginal and the generator
is b eing cut in and out intermittently. The p o w e r surges due to
inrush currents eventually trip the heater.

This p r o b l e m was corrected b y installing slightly larger

heaters and setting the m to reset automatically after a three
minute cooling period.

6.4 PERFORMANCE - The Giromill has been operated successfully by

producing full p ower and feeding it into the utility grid. Blade
modulation, the k e y to the Giromill concept, has wo r k e d as
planned. A constant rotor RPM is maintainable even in a gusty
w i n d wit h or without the generator connected to the utility grid.
However, only a limited amount of performance data has bee n co l ­
lected due to a variety of e v e n t s . The damage due to the
lightning rod failure and ensuing repair time consumed the
greater port i o n of the w indy season at Rocky Flats. Other data
was lost w h e n the grid connect heaters tripped and the Giromill
ran unconnected to the grid. Still other times the data collec­
tion system was down w h e n the Giromill was running.

The data collected is plotted in Figure 46. This is binsed

data of several hours of continuous running. The data w e r e co r ­
rected to sea level conditions.

The data indicate that the prototype performance is lower

than predicted. This reduction of performance is related to the
increase in rotor drag due to: (1) The blade seals not installed
which effectively decrease blade aspect ratio increasing induced
drag (see Section 6.3.1), (2) One-half inch wide gaps existing
between the support arms and blade fairings due to manufacturing
deviations (see Section 4.1), and (3) The blade fairings are not
streamlined to account for the flow angularity changes around the
blade orbit. They should be shaped disk-like rather than rectan­
gular to prevent blade/support a r m juncture interference drag.
Correcting these design deficiencies should raise the performance
to nea r the predicted level.

An attractive alternate method of running the Giromill was

demonstrated. The control system was temporarily configured to
hold the blades at a rock angle of zero when the RPM reached
synchronous speed. This essentially converted the Giromill to
run in the Darrieus mode after it had started using the normal
Giromill blade modulation. The unit ran very smoothly in the
Darrieus mode. Unfortunately the data acquisition system was not
operating at the time (had been hit by lightning), so performance
in this mode is not available at this time. Further testing in
this mode is recommended to explore its possible advantages.

60

50

PREDICTED
40
ACTUAL

OUTPUT
POWER 30
kW

20

10

WIND SPEED-m/s
GP13-0821-1

FIGURE 46
GIROMILL PERFORMANCE
Rocky Flats Binsed Data Corrected to Sea Level

59
 7. CONCLUSIONS & RECOMMENDATIONS

7.1 CONCLUSIONS

o A prototype 40 k W Giromill w a s designed, constructed, and

operated successfully by generating 40 k W or more power
for a utility grid.

o The Giromill blade modulation scheme w o r k s as planned.

It is able to maintain a constant rotor RPM in a gusty
wind w i t h or without the generator connected to the
utility grid.

o The unit runs smoothly. V e r y little vibration is evident

during normal operation under all wind conditions.

o The Giromill has operated as expected, however,

performance is lower than predicted due to poor design
and fabrication o f the blade/support arm feirings.

o Satisfactory blade actuator reliability has not bee n

achieved.

o Controller-commanded shutdown due to failure detection

w a s successfully demonstrated.

o Aerodynamic blade noise is minimal. The loudest noise

while running comes from the generator drive belt.

o The fixed blade (Darrieus) mode has been successfully

demonstrated. This hybrid operation has advantages that
should be explored.

o Mechanical drive remains to be tested.

o No p r o blems have b een encountered w h i c h m ight preclude

the future development o f the Giromill concept.

7.2 RECOMMENDATIONS

On the b a s i s o f the tests conducted to date, the following

recommendations fbr further testing or follow-on effort are made.

o Extensive testing using the fixed blade mode should be

conducted. The advantages o f reduced blade actuator duty
cycle, and possible blade positioning to control power
output should be investigated.

o The noisy generator drive bel t should be replaced by a

direct gear drive making the machine very quiet.

60
 CONCLUSIONS & RECOMMENDATIONS (Continued)

o More and better instrumentation should b e installed to

verify theoretical analyses of the system structure and
dynamics.

o The blade actuators reliability should be increased.

o Uprating pres e n t 40 k W unit to approximately 60 k W should

be investigated.

o A two-bladed unit should be investigated.

61
 8. REFERENCES

1. Anderson, J. W . , et al, "Development o f a 40 k W Giromill,

Phase I," Volume II - Design and Analysis, DOE Report RFP-
3032/64100/3533/79-17-2, A u g u s t 1979.

62
U.S. G OVERNM ENT PRINTING OFFICE; 198 2-5 77-7 80/144 Region No. 8