AV HV aD

oO

 

THESIS
Deh Ue
SUPT na eee

F.O, GRAHAM F. W. OPENLANDER

19.17
Copy TRESIS

Cor. \
C
Ofy THESIS

Co } |

 
The Construction
of a

Doublie-current Generator

A Thesis
Submitted to

The Faculty
of
MICHIGAN AGRICULTURAL COLLEGE

By

¥F,0O.Graham F .W.Openlander
Candidates for the Degree of

Bachelor of Science

June, 1917.
TH esis

~
-&

4 \
Preface.

The object of this thesis was the reconstruction
of a dynamo-electric machine from a direct current motor
into a rotary converter, or double- current generator. Also
the construction of three s.ngle phase transformers, suite
able for changing the voltage of the alternating current
obtained from the above machine, to a number of convenient
voltages for laboratory use.

In working up the best methods to employ in the
different steps of the construction, we are greatly in-
debted to Professor A.R. Sawyer and Professor M. M. Cory
for their supervision and helpful suggestions. We are
also very grateful to the men in charge of the pattern
shop, the machine shop, and foundry for sce generously aid=
ing us in the matter of machinery tools and timely advice.

In tne text which follows, we have endeavored to
give a very careful discussion of the problems solved and
the methods used, so that they may be readily understood
by those having a general knowledge of mathematics and

machine shop methods.

Signed,

The Authors,

 
Chapter I.
ELIP-RINGS AND END FRAME CONSTRUCTION,

“The machine used in this work was a Westinghouse
two horse power, direct current motor driven by another
‘motor of the same size and type. Our object was to re-
puild the machine so that it would deliver three phase
alternating current in addition to direct current.

The first thinz done was to build a new end
frame for the pulley end of the motor, which would place
the bearing far enough away from the armature to allow
space for three slip-rings. Blue-nrint No.I sows the are
rangement of slip-rings and brushes. Instead of building
a pattern for the frame, we decided to remove the one
from the commutator end end duvlicate it for the slip-
ring end. To do tnis, it was necessary to build a core-
box, as shown in blue-print No.ITI. This core was turned
out on a lathe and then tne core-box was made by means
of a plaster-of=-Paris cast. In order to use the frame for
a pattern, it was necescury to plug all openings. The
plugs for the ends of the bearing were made in the form
of core-=prints. In places where finish was needed, we
fastened heavy strips of naste-board to the pattern. This
gave plenty of extra stock for machining.

The ulternating current brushes were ordered
from the Generali Electric Company and it was necessary

te design a suitable yoke to support them. Blue-vnrint No.
IIIT shows a full-size drawing of the yoke which. we built
for this purvose. It is secured to tne collar on the inner
end of the bearinz, by means of a set-screw.

A pattern was made for a brass cylinder 4.75"
outside, and 3.75" inside diameter. Three slip-rings were
turned from this cylinder, each of them being 0.625" wide
and 0.375" thick. These slip-rings were insulated from the
shaft by a wooden cylinder whose outside diameter was equal
to the inside diameter of the rings and whose inside dia-
meter was the same as the diameter of the shaft. This wood-
en cylinder was secured to tne shaft by a set-screw with a
counter-sunk head, the head being covered with insulatinz
wax. Short screwe were used to fasten the slip-rings to
the wooden cylinder. Holes were drilled laterally in the
insuletinz cylinder to provide conduits for the connections
to the armature windings.

It was found that the shaft of the machine vas too
short for the new arramgement, therefore a short niece of
steel was welded on to the end by the oxye- acetylene process.

The studs which carried tne brush-holders, were

carefully insulated from the frame to vrevent grounding.
ChapterII.

TRANSFOR RS,

the transformers constructed were of the low-
voltage shell type, end were designed for 78 volts primary
and 6, 18, 12, and 110 volts secondary. The most simvle
transformer to meet these requirements is the auto-trans-
former, three of which were constructed. We were somewhat
handicapped in our design by the fact that it was necessary
to use punchings which were originally designed for another
transformer. As a result tne core appears somewhat out of
proportion to the coil.

Ail computations for size of wire, number of turns,
volume of core, etc., were based on a design as given in
"The Flements of Mechanical and Flectrical Engineering", (Sup-
plementary Volume), published by the International Text-book
Company.

Design of a Single Phase, 0.5 K, W., Shell-type,
Auto-transformer:

As we have a 2H.P. machine, the total power

will be approximately 1500 watts. Hence the watts per phase will
be about 500. Therefore we shall design a 500 watt trans-

former for each phase.

D.C. voltage = 220 volts.

A.C. voltage, (Y-connection) = 78 volts.
Secondary voits = 78.
Desired primary volts = 110.

Assumed efficiency = 93.5 7,
Input = 500/.935 =535 watts
Therefore the total loss at full load should not

exceed 35 watts. This loss is divided approximately as

follows:
Copper loss = 20 watts
Hysteresis loss = 12 watts
Eddy-current loss = 3 watts
Total loss = 35 watts

The loss per cubic inch of core, for 60 cycles,
at a density of 30000, is about 0.15 watts. Therefore the
volume of the core is,

12/.15 = 80 cu.in.

The area of the punching is,

(5,53 x €.92) - 2 (2.2 x 3.23) = 35 sq.in.

Use 34 sq. in. to make allowance for champfered
corners. Then the tnickness of the core becomes:

80/34 = 2.35",

Make the core 24 inches thick.

The total amperes input at full load is,
Primary watts / Primary volts = 535/78 = 6.856 amps.
The maximum current which the wire must carry is,

6.85/K

78 x 6.86/110

110/72
X

if

4.76 amperes.
Allowing 1200 circular mils per ampere to get the
size of the conductor, we have,

4.76 x 1200 = 5712 or approximatel¥ 6000 cir. mils.

From this we see that #12 or #13 will be close

enough for our purpose.

Width of core = 2.187"
Depth of core = 2.500"
Length of core = 3.230"

#12 wire is about 0.1" thick including insulation.

The primary coil has to be provided with a suffi-
cient number of turns to generate a counter electromotive
force equal and opposite to the impressed electromotive
force. The number of turns required to set up this electro-
motive force will depend upon the magnetic flux ¢ which
threads through the turns. The maximum magnetic flux tnrough

the coil will be;

J = Bmax X A.
Where Bmax is the maximum

value which the magnetic density reaches during a cycle,
and A is the area of cross-section of the iron in the core.

In this case Bnax is 30000 lines per square inch,
and the area of cross-section of the iron is,

2.5 X 2.187 = 5.46 sq.in. hence,

Z = 30000 x 5.46 = 163800.
Taking tne electromotive force generated in the

primary as the equal and opposite of the line voltage, we

may write,
Eo = 4.44 9 Tp) n/10° where,
@ = tne maximum value of the magnetic flux thru
tne core;
Tp = the number of turns on the primary coil;
n = the frequency, ( cycles per second ).
Ep = the impressed primary voltage.
Apolying tne above formula to the case in hand,
we have,
4.44 x 163800 x Tp x 60
78 = SB 22 ener wn@Qeeen eee Banana a
108
78 x 10°
Tp —- Seow we oe an © oe ww ow me wo
4.44 x 1638 x 6

17@.6 turns.

Use 180 turns for the primary. the total number
of turns for the 110 volts will be,

Tt = 180 x 110/78 = 254 turns.

Knowing that the core is 3.23" long and tnat #12
wire is 0.1" in diameter, we find the number of turns per
layer of wire to be,

3.23/o.1 = 32.3

Use 30 turns to allow for tapeing.
Now there are 254 turns in all; hence the number
of layers required is,

254/30 = &.5 layers.

Allowing for wedging tyne coil and for clearance,
we find tne inside perimeter of the coil to be,
P) = 10.5620"
Adding tne thickness of 8.5 layers
of wire, we find tne outside perimeter of the coil to be,
Po = 17.42"

The average perimeter is then,

Pj + P./2

Pa

10.625 + 17.42

2
= 14.02"
The total length of wire, in feet, for all three
transformers is found to be,
Pa x N
Lt = cecn--ee X 35
12
14.02 X 254

12

890 feet.

No. 12 wire weizhs 19.8 pounds per 1000 ft. Therefore

the total weight of the wire needed becomes,

19.8 x 0.890 = 17.5 lbs.
The number of turns of wire reuuired to give one volt
1s

254 / 110 = 2.31 turns.

Knowing this relation, it is easy to get any desired
voltage from the transformer, by simply winding on the reqe-
uilred number of turns or by taking off a t@p at the proper
turn.

After constructing one transformer, it was tested

with tne following results:

No Load
E | 78 volts
-045 amps
W 16 watts

P.F. 64 7%
Chapter III.
SWITCH BOARD.

For convenience in laboratory use, we constructed
a wooden switch-board tnree feet wide and four feet long.

The idea, in the design of this board, was to have all wir-
ing amd apparatus in plain sight for the benefit of the
student, rather than concealed wiring as is done commercially.
For example, our transformers were not cased, thus rendering
them easy of access for inspection and study.

The direct-connected motor generator set was placed
on the floor, directly below the switch-board which was ele-
vated upon a standard two feet high. 220 volt leads to the
motor were brought to the top of the board. Space was provided
on the board for the starting box and also for a speed reg-
ulator, so that alternating current of a variable frequency
could be obtained.

The rest of the apparatus on the board included
the following generator auxiliaries; one alternating current
voltmeter, one direct current voltmeter, and three trans-
formers mounted along the lower portion of the board, one
side of each being connected to one ring of the armature
and the other to a common neutral.

The general arrangement of the switch-board and

connections are shown on blueeprint No. IV. The diagram
shows tne method of makinz connections in order to obtain
alternating voltages ranging from 6 to 155 volts, also

110 or 220 volts of direct current.

10.
 

 

 

aupeaesmee

 

 

 

_
 

 

yvIQ7I0H HSAYT

7
‘MIONUINTSO BP WiHuw> 1
eye) Me hg

ae

 

 

 

 

 

 

 

 

 

 
 

Be Sh ae ota

 

rh
NS Ae ke Lat
Per Te A cl a

 
ROOM |

c

2

eb
c CH! Y
hs, ps Oe he

 

eB e
4
. ‘4