- Email: [email protected]
Oxidation of carbon and metal coated carbon fibers
OXIDATION OF CARBON AND METAL COATED CARBON FIBERS F. S. GALASSO and J. PINTO
United Aircraft Research Laboratories, East Hartford, Connecticut (USA) (Received : 2 January, 1970)
SUMMARY
Courtaulds Type B, Type B (aluminum coated), Type C, Thornel T-25, T-50, T-50 (surface treated), Hitco HMG-50 and Morganite Type I (nickel coated) carbon fibers were heated at temperatures from 200° to 600°C in air. The fibers were found to oxidise (as evidenced by a reduction in fiber cross-sections and loss of weight) rapidly above 500°C. For fibers with similar moduli, those prepared from polyacrylonitrile precursors oxidised more rapidly than fibers formed from rayon precursors and, for fibers prepared from the same precursor, lower modulus fibers oxidised faster than higher modulus fibers. Carbon fibers coated with aluminum oxidised at a rate similar to that for the uncoated fiber below 500 °, but above 500°C oxidation was slower than for the uncoated fiber.
INTRODUCTION
In recent years, high modulus carbon fibers have been widely employed in epoxymatrix composites. Some consideration has also been given to the use of carbon fibers in metal matrices such as aluminum, nickel and cobalt. Three of the main concerns in using carbon fibers in metal matrices at elevated temperatures have been the reactivity of the carbon fibers with the metal, the effects of oxidation on the strength of the carbon fiber and the effects of thermal expansion differences in the carbon fiber and matrix on the properties of the composites. A thorough study was conducted by Jackson 1 in which carbon fibers and metal coated carbon fibers were heated in vacuum and the effect of heating and reaction with the metals on the strength of the fibers was evaluated. In the present study the second problem was investigated by determining the effects of heating carbon fibers and aluminum 303
Fibre Science and Technology (2) (1970)--O Elsevier Publishing Company Ltd, England--Printed in Great Britain
304
V.S. GALASSO,J. PINT()
and nickel coated carbon fibers in air on the strength, modulus, and cross-sectional area of the fibers.
EXPERIMENTAL
METHOD
The carbon fibers used in this study were Courtaulds Type B and Type C, Thornel T-25, T-50, and T-50 surface-treated at United Aircraft Research Laboratories (UARL), Hitco HMG-50, and Morganite Type 1. All of these carbon fibers, except for the U A R L treated fibers, are commercially available. Fiber tensile testing was conducted using a procedure in which individual fibers from the tows
Fig. 1.
Surface p h o t o g r a p h o f Hitco H M G - 5 0 c a r b o n yarn showing area of increased oxidation rate, heated at 500°C for 100 h o u r s ( ," 40).
or yarns were tested in a machine similar to the one described by Schile and Rosica. 2 The modulus values were obtained using the same apparatus. Since most of the high modulus carbon fiber is produced from rayon or polyacrylonitrile (PAN) fibers, a representative rayon based fiber, Hitco HMG-50 yarn, and PAN based fiber, Courtaulds Type B tow, were selected for use in the
OXIDATION OF CARBON AND METAL COATED CARBON FIBERS
305
initial oxidation experiments. Care was taken in handling the fiber, since it was found that handling the carbon yarns and tows to any great extent increased the oxidation rate (as evidenced by a decrease in fiber diameter) in the particular area that was touched (Fig. 1). The representative cross-sectional area for the individual fibers in each batch of fibers was obtained by imbedding the as-received and the heat-treated carbon yarn or tow in Bakelite, photographing the polished cross-sections and measuring 10 individual fiber cross-sections with a planimeter. The average of the readings for fibers subjected to each heat treatment was used as the cross-sectional area for that heat-treated batch of fibers. The cross-sectional area, tensile strength, and modulus of the as-received Hitco H M G - 5 0 yarn and Courtaulds Type B tow were measured by the procedure described above. Then samples of the fibers were placed in Combax boats, which are made primarily of ZrSiO4, heated from 24 to 500 h in air at temperatures from 200 ° to 600°C, and the cross-sectional area, strength and modulus of the fibers were measured. Once it was found that the carbon oxidised most rapidly at temperatures above 500°C, the rate of oxidation of these fibers as well as several other types of carbon fibers was investigated at 550 ° and 600°C. High modulus Hitco HMG-50, Thornel T-50 (treated and untreated), and Courtaulds Type B as well as lower modulus Courtaulds Type C and Thornel T-25 carbon fibers were used in these studies. Samples of these fibers were weighed, heated in air at 550 ° and 600°C for various times and reweighed to determine the loss in weight. Aluminum coated carbon fibers (aluminum coating thickness ~0'03-0"07 mils) were prepared by vacuum evaporating aluminum on Courtaulds Type B carbon fibers strung on a cardboard holder in bundles of 20-30 fibers. Nickel coated fibers were produced by electroplating the nickel on Morganite Type I carbon fibers. These coated fibers also were heated in air and the cross-sectional areas measured. The strength and modulus were measured on the heat-treated aluminum coated carbon fibers without removing the aluminum. The strength and modulus of the heat-treated nickel coated carbon fibers could not be measured, as they were not long enough due to a shredding of the fibers once NiO was formed during the heating process.
EXPERIMENTAL RESULTS AND DISCUSSION
Plots of the average room temperature ultimate tensile strength and modulus for Hitco H M G - 5 0 and Courtaulds Type B carbon fibers as a function of time at temperature in air are shown in Figs. 2 and 3. These plots indicate that there is no appreciable reduction in the strength or modulus of the carbon fibers after they
306
F . s . GALASSO~ J. PINTO
SYMBOL: II 0 /x 0 TIME, HRS: AS REC'D 2 4 150 5 0 0 SAMPLE SIZE=IO R A N G E = O N E STANDARD DEVIATION
u,
80
..=.
7o
~
6o
0 O~
50
"o ~ 40 30
t I
I
I
I
I
- 500
~
400
~ Z 300
•~
~ 2oo
it -
t
.z. loo
IJJ I,--
,~
Fig. 2.
0 o
100 200 300 400 500 EXPOSURE TEMPERATURE-°C
600
Average room temperature UTS and modulus as a function of time at temperature for Courtaulds Type B carbon fibers.
O X I D A T I O N OF C A R B O N A N D METAL C O A T E D C A R B O N FIBERS
SYMBOL:
•
~
z~
307
O
TIME, HRS: AS REC'D 24 150 SOD SAMPLE SIZE = 10 RANGE = ONE STANDARD DEVIATION
60 bill
o.~
50
O~ •v a
,.u 0
3(~
,< "~
400
L~I I - -
o~Z200 o~ "~ uJ 10G I.U ,..I
ON ,~ <' u ,zI u.i ~.-
F i g . 3.
,
0
T
T
T
T
100 200 300 400 500 EXPOSURE TEMPERATURE-°C
600
Average room temperature UTS and modulus as a function of time at temperature for H i t c o H M G - 5 0 c a r b o n fibers.
308
F . S . GALASSO, J. P I N T O
have been treated in air at temperatures up to 600°C and times up to 500 h. However, from Figs. 4 and 5 it can be seen that the cross-sectional area of the fibers decreases after they have been heated to temperatures of 5 0 0 C and higher. More
SYMBOL: • (> TIME, HRS: AS REC'D 24 SAMPLE SIZE=IO RANGE:
6
¢11 ~O
V A 100 150
C] O 300 500
DATA SPREAD
0.14
I o
<-
0.12
0.10 Z
o
0.08
u 144
J,
0.06
o ,v u
T
>
0.04
LM
~n 14.
0.02
144
(.9 O 144
>
O
i
-F
-F
100
200
300
-1400
500
600
TEMPERATURE-°C
Fig. 4. Averagefiber cross-sectional area as a function of time at temperature for Courtaulds Type B carbon fibers.
rapid oxidation is observed at 600°C, as seen in the photomicrographs of the cross-sections of carbon fiber samples shown in Figs. 6 and 7. Examination of the data in Figs. 4 and 5 also shows that the Courtaulds carbon fiber is reduced in size more than the Hitco HMG-50 fiber. This agrees with the data in Table 1, which indicates that the rayon based fibers, Hitco HMG-50 and
OXIDATION OF CARBON AND METAL COATED CARBON FIBERS
SYMBOL:
•
~
A
D
309
0
TIME, HRS: AS REC'D 24 150 3 0 0
500
SAMPLE SIZE = 10
o
Z
DATA SPREAD
RANGE:
6
'~
o
0.12
3n,,0 . 1 0
'~
Z
0.08
II
0
g o.o6 0 0.04 U
~ o.o2 14.
,
o
'~
0
w,i
> Fig. 5.
100
T
T
T
T ~.-u--J
200 300 400 TEMPERATURE-°C
500
600
Average fiber cross-sectional area as a function of time at temperature for Hitco H M G - 5 0 carbon fibers.
TABLE 1 W E I G H T L O S S O F C A R B O N F I B E R S H E A T E D 1N A I R
Weight loss (%) Type of fiber Hitco HMG-50 yarn Thornel T-50 yarn Courtaulds Type B tow* Courtaulds Type B tow** (aluminum coated) Thornel T-25 yarn Thornel T-50 yarn (surface treated) Courtaulds Type B tow** Courtaulds Type C tow
550°C 16 h 24 h
6h
600°C 18 h
24 h
2 3 13
2 5 14 20
2 3 14 16
15 14 17 20
20 18 28 23
8 3
16 6
!0 5
45 22
58 7l
26 32
62 63
19 90
74 87 completely oxidised
* This is a selected sample of very high modulus carbon fiber ( ~ 7 0 × 106 psi). ** These carbon fibers are from the same batch.
310
v . s . GALASSO, J. PINTO
(a)
(b) Fig. 6.
Cross-sections of Courtaulds Type B carbon fibers (a) as received and (b) after heating in air at 600°C for 50 hours ( x 500).
OXIDATION OF CARBON AND METAL COATED CARBON FIBERS
311
(a)
(b) Fig. 7.
Cross-sections of Hitco HMG-50 carbon fibers (a) as received and (b) after heating in air at 600°C for 50 hours ( x 500).
312
v . s . GALASSO, J. PINTO
Thornel T-50 oxidise less rapidly than the Courtaulds PAN based fibers. In addition, comparing the results for Thornel T-25 and T-50 with Courtaulds Type B and Type C, it appears that the lower modulus fiber of each series oxidises faster than the higher modulus fiber. This is illustrated quite dramatically by the oxidation data shown in Table 1 for a selected very high modulus ( ~ 7 0 x 106 psi) Courtaulds Type B sample which shows that this fiber oxidises almost as slowly as the typical high modulus rayon based carbon fibers. Since higher modulus fibers
TIME AT TEMPERATURE = 24 HOURS • UNCOATED O A L U M I N U M COATED ~, NICKEL COATED ~*ALUMIN UM COATED HEATED IN VACUUM SAMPLE SIZE = 10
0.12 7 0
0.10
i.j
O
0.08 0
0
4
u, o
0.06
,.--~
0.04
e
(.9 14J
~
0.02 0
~ 0
I 200
I 300
i 400
TEMPERATURE-
I 500
I 600
700
°C
Fig. 8. Average cross-sectional area as a function of temperature for Courtaulds Type B carbon fiber uncoated and aluminum coated and nickel coated Morganite Type I carbon fiber.
contain more aligned graphite platelets, it is logical that these less reactive platelet surfaces should oxidise more slowly than low modulus fibers which contain more active graphite platelet edges at the surface. The oxidation data for Thornel T-50 untreated and fibers treated with HNO3 to activate the surface appear to bear this out. The data from studies of aluminum coated Courtaulds carbon fiber show that the strength and modulus of the fiber are not appreciably lowered after heating to
OXIDATION OF CARBON AND METAL COATED CARBON FIBERS
313
temperatures up to 600°C for periods up to 24 h. However, the cross-sectional area o f the fibers is reduced as is the case for the uncoated fiber. This oxidation o f the fiber is slowed by the aluminum coating above 500°C as evidenced by the data in Fig. 8.
Fig. 9(a)
Fig. 9Cb)
The legend for Figs. 9a and 9b appears under Fig. 9c on page 314.
314
F . S . GALASSO~ J. PINTO
Fig. 9(c) Fig. 9.
Transverse cross-sections of nickel coated Morganite Type I carbon fibers heated in air for 24 hours (a) 300°C, (b) 400°C, and (c) 500°C ( ~ 500).
The nickel on the nickel coated Morganite fiber oxidised at 400°C and above to form NiO and disrupted the continuity of the fiber (Fig. 9). However, the oxide also slowed down the oxidation of the carbon fiber above 500°C (Fig. 8). It can be concluded that oxidation of carbon fiber becomes quite rapid above 500°C. Oxidation takes place more rapidly for the P A N based carbon fibers than the rayon based fibers studied and it is more rapid for lower modulus fiber than higher modulus fiber. The aluminum coating slows down the oxidation process, but does not eliminate it. Jackson's studies in vacuum showed that an aluminum-carbon fiber reaction takes place at 600°C which reduces the strength of the fiber. This reaction was not detected in this work in which heat treatments were conducted in air. Jackson's studies also showed that the strength of nickel coated carbon fibers was reduced when heated at temperatures above 1000°C. In these studies, the oxidation of nickel broke up the fibers at a temperature of 400°C even below the temperature at which rapid oxidation of carbon fibers takes place. If the nickel coating were alloyed to slow down its oxidation it is probable that the carbon fibers would still oxidise above 500°C, which is below the temperature at which the nickel degrades the carbon fiber. Thus these studies indicate that unless air can be prevented from reaching the carbon fibers, oxidation of these fibers will be a problem at a lower temperature in aluminum and nickel matrix composites than the temperature at which metal-carbon reactions decrease the strength of the fiber.
OXIDATION OF CARBON AND METAL COATED CARBON FIBERS
315
ACKNOWLEDGEMENTS We wo u l d like to t h a n k W. Tice for the a l u m i n u m coated c a r b o n fibers, N. Bornstein for the nickel coated fiber a n d D. Scola f o r the treated T-50 fiber.
REFERENCES 1. P. W. JACKSON,Some studies of the compatibility of graphite and other fibers with metal matrices, Western Metal and Tool Conf. and Exposition, Los Angeles, California, 1969. 2. R. SC.ILE and G. ROSICA,Simple tester for the rapid determination of the tensile strength of fine filament, Rev. Sci. Instr., 38 (1967) 1103.
United Aircraft Research Laboratories, East Hartford, Connecticut (USA) (Received : 2 January, 1970)
SUMMARY
Courtaulds Type B, Type B (aluminum coated), Type C, Thornel T-25, T-50, T-50 (surface treated), Hitco HMG-50 and Morganite Type I (nickel coated) carbon fibers were heated at temperatures from 200° to 600°C in air. The fibers were found to oxidise (as evidenced by a reduction in fiber cross-sections and loss of weight) rapidly above 500°C. For fibers with similar moduli, those prepared from polyacrylonitrile precursors oxidised more rapidly than fibers formed from rayon precursors and, for fibers prepared from the same precursor, lower modulus fibers oxidised faster than higher modulus fibers. Carbon fibers coated with aluminum oxidised at a rate similar to that for the uncoated fiber below 500 °, but above 500°C oxidation was slower than for the uncoated fiber.
INTRODUCTION
In recent years, high modulus carbon fibers have been widely employed in epoxymatrix composites. Some consideration has also been given to the use of carbon fibers in metal matrices such as aluminum, nickel and cobalt. Three of the main concerns in using carbon fibers in metal matrices at elevated temperatures have been the reactivity of the carbon fibers with the metal, the effects of oxidation on the strength of the carbon fiber and the effects of thermal expansion differences in the carbon fiber and matrix on the properties of the composites. A thorough study was conducted by Jackson 1 in which carbon fibers and metal coated carbon fibers were heated in vacuum and the effect of heating and reaction with the metals on the strength of the fibers was evaluated. In the present study the second problem was investigated by determining the effects of heating carbon fibers and aluminum 303
Fibre Science and Technology (2) (1970)--O Elsevier Publishing Company Ltd, England--Printed in Great Britain
304
V.S. GALASSO,J. PINT()
and nickel coated carbon fibers in air on the strength, modulus, and cross-sectional area of the fibers.
EXPERIMENTAL
METHOD
The carbon fibers used in this study were Courtaulds Type B and Type C, Thornel T-25, T-50, and T-50 surface-treated at United Aircraft Research Laboratories (UARL), Hitco HMG-50, and Morganite Type 1. All of these carbon fibers, except for the U A R L treated fibers, are commercially available. Fiber tensile testing was conducted using a procedure in which individual fibers from the tows
Fig. 1.
Surface p h o t o g r a p h o f Hitco H M G - 5 0 c a r b o n yarn showing area of increased oxidation rate, heated at 500°C for 100 h o u r s ( ," 40).
or yarns were tested in a machine similar to the one described by Schile and Rosica. 2 The modulus values were obtained using the same apparatus. Since most of the high modulus carbon fiber is produced from rayon or polyacrylonitrile (PAN) fibers, a representative rayon based fiber, Hitco HMG-50 yarn, and PAN based fiber, Courtaulds Type B tow, were selected for use in the
OXIDATION OF CARBON AND METAL COATED CARBON FIBERS
305
initial oxidation experiments. Care was taken in handling the fiber, since it was found that handling the carbon yarns and tows to any great extent increased the oxidation rate (as evidenced by a decrease in fiber diameter) in the particular area that was touched (Fig. 1). The representative cross-sectional area for the individual fibers in each batch of fibers was obtained by imbedding the as-received and the heat-treated carbon yarn or tow in Bakelite, photographing the polished cross-sections and measuring 10 individual fiber cross-sections with a planimeter. The average of the readings for fibers subjected to each heat treatment was used as the cross-sectional area for that heat-treated batch of fibers. The cross-sectional area, tensile strength, and modulus of the as-received Hitco H M G - 5 0 yarn and Courtaulds Type B tow were measured by the procedure described above. Then samples of the fibers were placed in Combax boats, which are made primarily of ZrSiO4, heated from 24 to 500 h in air at temperatures from 200 ° to 600°C, and the cross-sectional area, strength and modulus of the fibers were measured. Once it was found that the carbon oxidised most rapidly at temperatures above 500°C, the rate of oxidation of these fibers as well as several other types of carbon fibers was investigated at 550 ° and 600°C. High modulus Hitco HMG-50, Thornel T-50 (treated and untreated), and Courtaulds Type B as well as lower modulus Courtaulds Type C and Thornel T-25 carbon fibers were used in these studies. Samples of these fibers were weighed, heated in air at 550 ° and 600°C for various times and reweighed to determine the loss in weight. Aluminum coated carbon fibers (aluminum coating thickness ~0'03-0"07 mils) were prepared by vacuum evaporating aluminum on Courtaulds Type B carbon fibers strung on a cardboard holder in bundles of 20-30 fibers. Nickel coated fibers were produced by electroplating the nickel on Morganite Type I carbon fibers. These coated fibers also were heated in air and the cross-sectional areas measured. The strength and modulus were measured on the heat-treated aluminum coated carbon fibers without removing the aluminum. The strength and modulus of the heat-treated nickel coated carbon fibers could not be measured, as they were not long enough due to a shredding of the fibers once NiO was formed during the heating process.
EXPERIMENTAL RESULTS AND DISCUSSION
Plots of the average room temperature ultimate tensile strength and modulus for Hitco H M G - 5 0 and Courtaulds Type B carbon fibers as a function of time at temperature in air are shown in Figs. 2 and 3. These plots indicate that there is no appreciable reduction in the strength or modulus of the carbon fibers after they
306
F . s . GALASSO~ J. PINTO
SYMBOL: II 0 /x 0 TIME, HRS: AS REC'D 2 4 150 5 0 0 SAMPLE SIZE=IO R A N G E = O N E STANDARD DEVIATION
u,
80
..=.
7o
~
6o
0 O~
50
"o ~ 40 30
t I
I
I
I
I
- 500
~
400
~ Z 300
•~
~ 2oo
it -
t
.z. loo
IJJ I,--
,~
Fig. 2.
0 o
100 200 300 400 500 EXPOSURE TEMPERATURE-°C
600
Average room temperature UTS and modulus as a function of time at temperature for Courtaulds Type B carbon fibers.
O X I D A T I O N OF C A R B O N A N D METAL C O A T E D C A R B O N FIBERS
SYMBOL:
•
~
z~
307
O
TIME, HRS: AS REC'D 24 150 SOD SAMPLE SIZE = 10 RANGE = ONE STANDARD DEVIATION
60 bill
o.~
50
O~ •v a
,.u 0
3(~
,< "~
400
L~I I - -
o~Z200 o~ "~ uJ 10G I.U ,..I
ON ,~ <' u ,zI u.i ~.-
F i g . 3.
,
0
T
T
T
T
100 200 300 400 500 EXPOSURE TEMPERATURE-°C
600
Average room temperature UTS and modulus as a function of time at temperature for H i t c o H M G - 5 0 c a r b o n fibers.
308
F . S . GALASSO, J. P I N T O
have been treated in air at temperatures up to 600°C and times up to 500 h. However, from Figs. 4 and 5 it can be seen that the cross-sectional area of the fibers decreases after they have been heated to temperatures of 5 0 0 C and higher. More
SYMBOL: • (> TIME, HRS: AS REC'D 24 SAMPLE SIZE=IO RANGE:
6
¢11 ~O
V A 100 150
C] O 300 500
DATA SPREAD
0.14
I o
<-
0.12
0.10 Z
o
0.08
u 144
J,
0.06
o ,v u
T
>
0.04
LM
~n 14.
0.02
144
(.9 O 144
>
O
i
-F
-F
100
200
300
-1400
500
600
TEMPERATURE-°C
Fig. 4. Averagefiber cross-sectional area as a function of time at temperature for Courtaulds Type B carbon fibers.
rapid oxidation is observed at 600°C, as seen in the photomicrographs of the cross-sections of carbon fiber samples shown in Figs. 6 and 7. Examination of the data in Figs. 4 and 5 also shows that the Courtaulds carbon fiber is reduced in size more than the Hitco HMG-50 fiber. This agrees with the data in Table 1, which indicates that the rayon based fibers, Hitco HMG-50 and
OXIDATION OF CARBON AND METAL COATED CARBON FIBERS
SYMBOL:
•
~
A
D
309
0
TIME, HRS: AS REC'D 24 150 3 0 0
500
SAMPLE SIZE = 10
o
Z
DATA SPREAD
RANGE:
6
'~
o
0.12
3n,,0 . 1 0
'~
Z
0.08
II
0
g o.o6 0 0.04 U
~ o.o2 14.
,
o
'~
0
w,i
> Fig. 5.
100
T
T
T
T ~.-u--J
200 300 400 TEMPERATURE-°C
500
600
Average fiber cross-sectional area as a function of time at temperature for Hitco H M G - 5 0 carbon fibers.
TABLE 1 W E I G H T L O S S O F C A R B O N F I B E R S H E A T E D 1N A I R
Weight loss (%) Type of fiber Hitco HMG-50 yarn Thornel T-50 yarn Courtaulds Type B tow* Courtaulds Type B tow** (aluminum coated) Thornel T-25 yarn Thornel T-50 yarn (surface treated) Courtaulds Type B tow** Courtaulds Type C tow
550°C 16 h 24 h
6h
600°C 18 h
24 h
2 3 13
2 5 14 20
2 3 14 16
15 14 17 20
20 18 28 23
8 3
16 6
!0 5
45 22
58 7l
26 32
62 63
19 90
74 87 completely oxidised
* This is a selected sample of very high modulus carbon fiber ( ~ 7 0 × 106 psi). ** These carbon fibers are from the same batch.
310
v . s . GALASSO, J. PINTO
(a)
(b) Fig. 6.
Cross-sections of Courtaulds Type B carbon fibers (a) as received and (b) after heating in air at 600°C for 50 hours ( x 500).
OXIDATION OF CARBON AND METAL COATED CARBON FIBERS
311
(a)
(b) Fig. 7.
Cross-sections of Hitco HMG-50 carbon fibers (a) as received and (b) after heating in air at 600°C for 50 hours ( x 500).
312
v . s . GALASSO, J. PINTO
Thornel T-50 oxidise less rapidly than the Courtaulds PAN based fibers. In addition, comparing the results for Thornel T-25 and T-50 with Courtaulds Type B and Type C, it appears that the lower modulus fiber of each series oxidises faster than the higher modulus fiber. This is illustrated quite dramatically by the oxidation data shown in Table 1 for a selected very high modulus ( ~ 7 0 x 106 psi) Courtaulds Type B sample which shows that this fiber oxidises almost as slowly as the typical high modulus rayon based carbon fibers. Since higher modulus fibers
TIME AT TEMPERATURE = 24 HOURS • UNCOATED O A L U M I N U M COATED ~, NICKEL COATED ~*ALUMIN UM COATED HEATED IN VACUUM SAMPLE SIZE = 10
0.12 7 0
0.10
i.j
O
0.08 0
0
4
u, o
0.06
,.--~
0.04
e
(.9 14J
~
0.02 0
~ 0
I 200
I 300
i 400
TEMPERATURE-
I 500
I 600
700
°C
Fig. 8. Average cross-sectional area as a function of temperature for Courtaulds Type B carbon fiber uncoated and aluminum coated and nickel coated Morganite Type I carbon fiber.
contain more aligned graphite platelets, it is logical that these less reactive platelet surfaces should oxidise more slowly than low modulus fibers which contain more active graphite platelet edges at the surface. The oxidation data for Thornel T-50 untreated and fibers treated with HNO3 to activate the surface appear to bear this out. The data from studies of aluminum coated Courtaulds carbon fiber show that the strength and modulus of the fiber are not appreciably lowered after heating to
OXIDATION OF CARBON AND METAL COATED CARBON FIBERS
313
temperatures up to 600°C for periods up to 24 h. However, the cross-sectional area o f the fibers is reduced as is the case for the uncoated fiber. This oxidation o f the fiber is slowed by the aluminum coating above 500°C as evidenced by the data in Fig. 8.
Fig. 9(a)
Fig. 9Cb)
The legend for Figs. 9a and 9b appears under Fig. 9c on page 314.
314
F . S . GALASSO~ J. PINTO
Fig. 9(c) Fig. 9.
Transverse cross-sections of nickel coated Morganite Type I carbon fibers heated in air for 24 hours (a) 300°C, (b) 400°C, and (c) 500°C ( ~ 500).
The nickel on the nickel coated Morganite fiber oxidised at 400°C and above to form NiO and disrupted the continuity of the fiber (Fig. 9). However, the oxide also slowed down the oxidation of the carbon fiber above 500°C (Fig. 8). It can be concluded that oxidation of carbon fiber becomes quite rapid above 500°C. Oxidation takes place more rapidly for the P A N based carbon fibers than the rayon based fibers studied and it is more rapid for lower modulus fiber than higher modulus fiber. The aluminum coating slows down the oxidation process, but does not eliminate it. Jackson's studies in vacuum showed that an aluminum-carbon fiber reaction takes place at 600°C which reduces the strength of the fiber. This reaction was not detected in this work in which heat treatments were conducted in air. Jackson's studies also showed that the strength of nickel coated carbon fibers was reduced when heated at temperatures above 1000°C. In these studies, the oxidation of nickel broke up the fibers at a temperature of 400°C even below the temperature at which rapid oxidation of carbon fibers takes place. If the nickel coating were alloyed to slow down its oxidation it is probable that the carbon fibers would still oxidise above 500°C, which is below the temperature at which the nickel degrades the carbon fiber. Thus these studies indicate that unless air can be prevented from reaching the carbon fibers, oxidation of these fibers will be a problem at a lower temperature in aluminum and nickel matrix composites than the temperature at which metal-carbon reactions decrease the strength of the fiber.
OXIDATION OF CARBON AND METAL COATED CARBON FIBERS
315
ACKNOWLEDGEMENTS We wo u l d like to t h a n k W. Tice for the a l u m i n u m coated c a r b o n fibers, N. Bornstein for the nickel coated fiber a n d D. Scola f o r the treated T-50 fiber.
REFERENCES 1. P. W. JACKSON,Some studies of the compatibility of graphite and other fibers with metal matrices, Western Metal and Tool Conf. and Exposition, Los Angeles, California, 1969. 2. R. SC.ILE and G. ROSICA,Simple tester for the rapid determination of the tensile strength of fine filament, Rev. Sci. Instr., 38 (1967) 1103.