Development of Non-contact NDE Method for Autoclave … · 1 Development of Non-contact NDE Method for Autoclave Cure Monitoring of Carbon-Phenolic Re-entry Shells of Missiles by

1

Development of Non-contact NDE Method for Autoclave Cure

Monitoring of Carbon-Phenolic Re-entry Shells of Missiles by On-line

Gas Chromatography Technique

Kammari Veera Brahmam, Vemana Venkateswara Rao

Advanced Systems Laboratory, DRDO, Kanchanbagh, Hyderabad- 500 058, INDIA

[email protected], [email protected]

Abstract

During re-entry phase of the long range missiles, they experience severe

aerodynamic friction and high surface temperatures. In this context, carbon-epoxy shell is

used as internal structural layer and carbon-phenolic shell is used as external layer for

production of re-entry vehicle structure. Autoclave curing is used for production of

external thermal shells of Re-entry vehicle structures of the missiles. Autoclave curing of

phenolic composites involves a complex regime of time – temperature -pressure cycle

where selection of pressure application point is the most important parameter, decides the

quality of the component. Early pressure application with respect the gelation produces

resin starved component, which degrade the thermal performance of the component

whereas late pressure application produces more porosity in the component due to

trapping of volatiles generated due to curing chemical reaction. Production of high resin

content and minimum void distribution is most essential and which is achieved by

analyzing the volatiles evolved during the process.

During curing process, m-phenol and water are the one of the indicators and the

concentrations are determined by suitable detectors as a function of temperature of the

component. Based on the diminishing trend of the curves the gelation region for pressure

application was determined. A new contact NDE method for cure monitoring of carbon-

phenolic composites was developed right from development of concept, realization,

development of proto type system for assessing the advancement of curing reaction and

for determination of gelation region for pressure application. The system contains gas

sample injection port, gas-chromatograph, detector unit, software for signal analysis. The

equipment and experimental details were presented in the paper.

National Seminar & Exhibition on Non-Destructive Evaluation, NDE 2014, Pune, December 4-6, 2014 (NDE-India 2014)

Vol.20 No.6 (June 2015) - The e-Journal of Nondestructive Testing - ISSN 1435-4934www.ndt.net/?id=17838

2

Key Words: Gas-chromatography, Carbon-Phenolic Composites, Cure Monitoring

Technique,

1. Introduction

Aerospace structures are made by embedding high strength fibers in a light weight

thermoset plastic by adding hardener or supplying heat energy. In this process, addition

or condensation polymerization type of thermoset resins are used. Epoxy resins used in

structural applications undergo addition polymerization without volatiles evolution and

hence control of curing process is easy. In case of resins undergoing addition

polymerization, the degree of cure is evaluated by conventional methods like differential

scanning calorimetry (DSC) or dielectric methods [1-6]. Carbon-phenolic(C-P)

composites are used as thermal protection layer in aerospace structures as ablative liners

[7-9]. Phenolic resins undergo condensation polymerization and produce water and

methylol phenol (M-phenol) by-products during curing. These by-products come out as

volatiles/vapors during curing process and form porosity in the component. The curing

reaction is a function of temperature, time and pressure. Therefore to control the curing

process the component is cured in autoclave by keeping in vacuum bag. Volatile

management with selection of vacuum levels, rate of heating and gelation region for

pressure application are crucial parameters to minimize the porosity and better

consolidation among the fabric layers. Among the above parameters identification of

gelation region for pressure application is most sensitive parameter which depends on the

advancement of resin/prepreg properties. Due to above criticalities, an on-line cure

monitoring system for selection of on-line pressure application point is most essential.

Cure monitoring is required to track the real-time changes in a chemical reaction

that occurs during advancement of the resin [8]. In case of epoxy resins, during curing

process reactions have been studied by many researchers with the help of (i) modeling the

cure process[2,6] (ii) in-situ monitoring and control of curing during the actual

manufacturing of composite products. Many authors reported the cure monitoring

techniques by measuring the specific property of the material [2-6] by embedding the

3

sensors in the component. But in case of C-P composites, the embedded sensors act as

hot spots, generate high temperature under thermal environment and the structure fails

easily. Therefore a contact and embedding sensor type of cure monitoring method is not

suitable. In view of the above circumstances, the author developed a low cost, non-

contact and non embedding sensor type of cure monitoring technique using on-line gas

chromatograph.

Phenolic resin is a thermoset type of resin with aromatic network and is obtained by

condensation of phenol with formaldehyde as shown in Fig.1. In the first step of curing

reaction, the M-phenol interacts with phenol and forms a polymer chain with

methylene-bridge. In the second step, the M-phenol reacts with M-phenol and forms a

polymer chain with ketone-bridge. Further continuous supply of heat energy forms a

3D-network of cured solid. Therefore the liquid phenolic resin initially in the low

molecular weight monomeric stage undergoes long chain pre-polymer formation and

subsequent gelation (rubbery state) to the final stage of chemical cross linking (solid

glassy state).

Fig.1. Condensation reaction in phenolic resin

In the present technique the evolved gaseous by-products as a mixture (with M-

phenol and water) are injected into the gas chromatograph, separated as individual

components and their concentrations are determined by suitable detectors. The

concentrations of M-phenol and water are monitored as a function of component

temperature by connecting gas chromatograph (G.C) to autoclave facility. Finally based

on the falling trend of M-phenol concentration and post analysis of laminate properties,

the criterion for pressure application is established.

M-phenol

OH

CH2O

H

OH

OH

CH2

CH2OC

H2

OH

OH

+

+

H2O

H2O

OH

HCHO +

Phenol Formaldehyd

e (Methylol –phenol)

(Water

) Cross-linked polymer

Stage- 1

4

2. Experimental

2.1. Sample Preparation

Laminates were made by hand lay-up process with 45 layers and subsequently

cured in autoclave by keeping the component in a vacuum bag. Teflon treated releasing

fabric is used as a separator ply followed by nine layers of bleeder material. The bleeder

material is used to absorb excess resin and to provide path to volatiles at vacuum ports.

The total lay-up was kept in a kapton made vacuum bag and cured in autoclave. The

laminates were cured at different pressure application points based on the falling trend of

M-phenol and the samples are designated as CP-HP-B, CP-1/3-C, CP-1/2-D and CP-3/4-

E.

2.2. Coupling of gas chromatograph equipment with autoclave facility

Fig.2 On-line gas chromatograph coupled with autoclave facility

Suction pump

Air N2 H2

RS 232

Main Vacuum

Pump Gas Hut

Autoclave

5

Chromatography is an analytical technique, which separates the gas mixture in to

individual components to identify and quantify the concentrations [10]. Fig.2 shows the

coupling of on-line gas chromatograph with autoclave facility.

The total experimental setup was developed indigenously and the setup contains four sub

parts as

i). Gas distribution panel: The nitrogen, oxygen and hydrogen gases from different

cylinders were purified by sending through the molecular sieves. Hydrogen and air

mixture is used to produce flame in the flame ionization detector (FID) and where as

nitrogen is used as a carrier gas.

ii). Sample connection line: The gas chromatograph is coupled to autoclave through a

suction pump followed by a gas-sampling valve (GSV). Sample line was maintained at

1400C to avoid condensation of gaseous sample mixture. The gas mixture from vacuum

bag is injected into gas chromatograph for every 4 minutes by operating (GSV) through

electronic timer circuit module and all electronic modules are interfaced with P.C

iii). Gas chromatograph: The gaseous mixture is separated into individual components

using 2 meter length Porapak-Q column. The separated components of M-phenol and

water concentrations are determined at the other end of the column with thermal

conductivity detector (TCD) and flame ionization detectors (FID) respectively.

Chromatograms are obtained at every four minute interval of time and the concentrations

of the M-phenol and water is determined from the area under the peaks.

iv).Electronic data acquisition and Interfacing: The electronic data acquisition module

converts the analog response into digital file format. The digital data is transferred to

computer through RS-232 interfacing cables and stored with .TCD and .FID extension

files. The process parameters are controlled through user-friendly “AUTOCHROWIN”

software. On-line cure monitoring of the process is carried out and the

evolution/concentration curves of M-phenol and water are obtained as a function of

component temperature/time.

6

2.3. Optimization of testing parameters in gas chromatograph

Separation of individual components from the gas mixture depends on the carrier

gas flow rate through the column, column temperature and sensitivity of the detector.

AUTOCHROWIN’ software is used for analysis of the retention times and area under the

peaks in the chromatograms. M/s. National Physical Laboratory (NPL) standards are used

for calibration of water and M-phenol peaks interms of retention time and area of the

peak. The retention times for water and m-phenol are calibrated as 0.4 minutes and 0.8

minutes respectively. The calibration process is carried out in a lab G.C and the same

calibration parameters are implemented for on-line analysis of the gaseous samples. The

following testing parameters are optimized in the calibration process to obtain consistent

and error free results. The area under the peaks in the chromatogram is measured as the

concentration of the constituents.

Oven Temperature : 1500C

Injector Temperature : 1200C

TCD Temperature : 1500C

FID Temperature : 1500C

GSV Temperature : 1400C

Filament current for TCD : 85mA

FID range : 100

Attenuation : 1 dB

Gas flow rates :

Carrier-L-TCD : 1.1 Kg/cm2

Carrier-R-TCD : 1.5 Kg/cm2

Hydrogen : 0.9 kg/cm2

Air (Oxygen) : 1.00 kg/cm2

Retention time for M-phenol : 0.70 min

Retention time for M-phenol : 0.38 min

7

2.4. Chemical analysis of composite samples

The laminates were prepared by varying pressure application point and small samples

from the laminates were subjected to chemical analysis as per the ASTM-D3171

standard for determination of solid resin content, void content and fiber volume fraction.

Resin bleed out volume is measured by weight difference of the laminate at before and

after curing.

3. Results and Discussion

3.1. Role of pressure application point in autoclave curing process

Phenolic resin undergoes condensation reaction, and produces condensation by-

products like M-phenol and water. These by-products along with the residual solvents are

sucked out of the vacuum bagged product by a suitable vacuum pump, while pressure is

applied inside the autoclave for consolidating the layered composite structure. Therefore

a complex temperature-vacuum-pressure regime needs to be carefully selected for

producing composite products of acceptable quality. It is important to note that early

pressure application (before gelation) tends to bleed more resin and forms low resin

content in the component; whereas, late pressure application (after gelation) tends to

generate defects like voids and de-laminations, leading to rejection of expensive

products. Hence an on-line monitoring of the curing reaction and identification of the

correct pressure-application point are of crucial importance.

3.2. Determination of broad gelation region from evolution curves

Fig.3. shows typical volatile evolution curves of m-phenol and water during curing

process. In all the experiments the trend of m-phenol evolution is consistent and

reproducible. Water evolution curves are found inconsistent due to moisture absorption

from ambience and the water by-products produced due to advancement of resin during

storage of prepreg. Hence, based on the trend of m-phenol evolution, the pressure

application criterion is decided.

8

The m-phenol evolution behavior is divided into three parts). I).Increasing

evolution region (AB), ii). Decreasing evolution region (BD) and iii). Constant evolution

region (DE) corresponding to liquid, gelation and solidification states of the resin

respectively, as shown in Fig.4.

20 40 60 80 100 120 140

0

50

100

150

200

250

M-Phenol Peak

Water Peak

Average component temparature(0C)

M-P

he

no

l Pea

k A

rea

(mV

.s)

250

300

350

400

450

500

550

600

650

700

750

800

850

900

950

1000

Wa

ter p

ea

k are

a( m

V.s)

Fig.3. Typical volatile evolution curves - without pressure application

At the early stage of curing, the reaction rate of phenol with formaldehyde is slow

at low temperatures due to low M-phenol formation up to the point A. From A to B

region, free phenol reacts with formaldehyde and forms more M-phenol and the M-

phenol concentration increases up to B. In BE region, the M-phenol again reacts with M-

phenol and forms cross-linked polymer and shows a decreasing trend of M-phenol

concentration up to E. In EF region due to formation of three dimensional network of the

polymer, the reaction completes and evolution of M-phenol reaches to a saturation due to

solidification. Therefore a broad gelation region for pressure application is observed from

B to E. The broad gelation region (BE) is further divided in to four sub parts as BC, CD,

DE and EF, to identify the exact pressure application point. Based on the falling trend of

the m-phenol curve the BE region is divided as i). At highest peak (B point) ii). At 1/3

fall of highest peak area (C point) iii). At ½ fall of highest peak area (D point) and iv). At

9

¾ th

fall of highest peak area (E point). To identify the correct pressure application point,

four laminates were produced by applying the pressure at highest peak, 1/3 rd fall of

peak, at ½ fall of peak and at3/4 th fall of peak respectively and the cured laminates are

designated as CP-HP-B, CP-1/3-C, CP-1/2-D and CP-3/4-E. During the curing process

of the four laminates the M-phenol and water evolution curves are plotted as shown in

Figs. 5 and 6. The properties like resin content, void content, fiber volume fraction and

density of the different laminates are compared in the Table.1.

40 60 80 100 120

0

500

1000

1500

2000

2500

3000

3500

M-p

he

no

l P

ea

k A

rea

(mV

.s)

Average component tempearture (0C)

FID

Fig.4. Typical m-phenol evolution curve

A

B

D

E

F

C

20 30 40 50 60 70 80 90 100 110 120 130 140

0

50

100

150

200

250

300

350

400

450

500

550 CP-HP-B

CP-1/3-C

CP-1/2-D

CP-3/4-E

M-P

henol A

rea (

mV

sec)

Avg Component Temperature (0C)

10

Fig.5. M-phenol evolution versus pressure application point curves

3.3. Determination of on-line pressure application criterion - Post properties of

laminates

High resin content and low void content are preferable for better thermal

performance of the ablative components. High resin content in the carbon-phenolic

composites produces more pyrolysis gases and chars yield, which keeps the top surface of

the material cool and protects from the thermal environment [11-12]. Correlation of

laminate properties interms of resin content, void content, fiber volume fraction with

respect the pressure application point have been discussed in the following sections.

Fig.6. Water evolution versus pressure application point curves

3.3.1. Pressure application at highest peak on the curve

Pressure application at B-point (at highest concentration of M-phenol peak) on

the curve indicates starting of gelation region. At this point, the viscosity of the liquid

resin is low and it can flow easily through the reinforcement layers. Therefore due to

20 30 40 50 60 70 80 90 100 110 120 130 140

0

200

400

600

800

1000

1200

Wate

r A

rea (

mV

sec)

Avg Component Temperature (0C)

CP-HP-B

CP-1/3-C

CP-1/2-D

CP-3/4-E

11

more bleed of resin, resin content is low and fiber volume fraction is high. In this stage

the by-products in the vapor form can be removed effectively due to low shearing force

among the trapped gaseous molecules with respect the liquid molecules of the resin and

therefore the volatiles can be removed effectively without trapping in the component.

Hence CP-HP-B laminate possess minimum porosity. At this stage of pressure

application the bleeding of resin is more due to low viscous nature of the resin and

produces more bleed out of the resin compared to other cases of pressure application.

Therefore lowest resin content and minimum porosity is obtained in the samples, which is

also supporting the chemical analysis results of the samples as shown in the Table.1.

3.3.2. Pressure application at 1/3 fall of peak on the curve

Pressure application at C-point (at 1/3 fall of M-phenol concentration) on the

curve indicates increase in resin viscosity due to pre-polymers formation from its

monomeric state. Due to beginning of advancement of resin, the volume of resin bleed

from the component is very low compared to CP-HP-B case. Hence more resin content is

produced in the laminate. In this stage the by-products in the vapor form can be removed

effectively due to low shearing force among the trapped gaseous molecules with respect

the liquid molecules of the resin but slight increase in viscosity of the resin retards the

movement of the volatiles slightly. Therefore it produces minimum porosity. Therefore

high resin content with low porosity is observed in CP-1/3C laminate. Increase in resin

content will decrease the fiber volume fraction and the density of the laminate.

Property CP-HP-B CP-1/3-C CP- 1/2 -D CP- 3/4 -E

Solid resin content (wt%) 28.25 34.37 37.64 39.08

Fiber volume fraction (%) 65.38 55.95 56.02 56.66

Density (gm/cc) 1.50 1.45 1.40 1.41

Void content (%) 0.20 0.60 0.81 1.10

Bleed out ( kg) 1.30 0.72 0.20 0.10

Table.1. Laminate properties with respect to pressure application

12

3.3.3. Pressure application at 1/2 fall of peak on the curve

Pressure application at the point ‘D’ (at 1/2 fall of M-phenol concentration) on the

curve indicates further increase in resin viscosity due to long chain formation from pre-

polymer state of the resin. Due to further advancement of resin, the bleed of resin from

the laminate is relatively very low as compared to CP-1/3-C case. The increase in

viscosity of the resin further retards the movement of the volatiles through the gel stage

of the resin and produces considerable porosity. Therefore high resin content and

considerable porosity is produced in CP-1/2-D laminate.

3.3.4. Pressure application at 3/4th

fall of peak on the curve

Pressure application at the point ‘E’ (at 3/4 fall of M-phenol concentration) on the

curve indicates sudden increase in resin viscosity due to formation of solid state of the

resin from long chain of molecules. Due to formation of three dimensional network of

solid, the m-phenol evolution reaches to a minimum and saturates at above the point’ E’,

which indicates glassy state of the resin. At this point the advancement of resin has been

completed and the bleed of resin from the laminate is lowest compared to other cases. At

this stage the viscosity of the resin is very high and completely retards the movement of

the volatiles through the resin. Therefore the reaction by-products are completely trapped

in the component and produces accumulated porosity (voids) and delaminations.

Therefore high resin content with more porosity and delaminations have been produced

in CP-3/4-E.

Table.1 indicates that the pressure application at highest peak produces low

porosity and low resin content. The pressure application at 3/4th

fall of peak produces

intense defects like voids and delaminations but the resin content is high. Therefore to

obtain high resin content with low porosity, the pressure should be applied at 1/3 fall of

M-phenol concentration on the curve. To obtain high resin content with considerable

porosity, the pressure should be applied at ½ fall of m-phenol concentration curve.

13

4. Conclusions

Cure monitoring of carbon-phenolic laminates was carried out by on-line gas

chromatography technique. The volatile evolution curves of M-phenol and water were

recorded with respect to the component temperature. M-phenol evolution curves are most

consistent and based on the falling trend of the curve the criterion for pressure application

was decided. Low porosity can be obtained in the component by applying pressure at

1/3rd

fall of m-phenol concentration on the curve; whereas, more resin content can be

obtained by applying pressure at ½ fall of M-phenol concentration.

References

[1]. Duncan Chadwick, Cure monitoring, U.S.Patent 6675112B1, 2004

[2]. Delwin.O.Johnson, Ultrasonic cure monitoring of advanced composites, U.S. Patent

5009104, April,1991

[3]. Vincent.K.Choo and Manuel V. Gomez, Resin cure monitoring, U.S.Patent 5911159,

June,1999

[4]. J.Y.Chen, S.V.Hoa, C.K. Chen and H.Wang, "Fiber-optic and ultrasonic

measurements for in-situ cure monitoring of graphite/epoxy composites", Journal of

Composite Materials, vol.33 (20), 1860-1881, 1999

[5]. J.S.Leng and A.Asundi, "Real-time cure monitoring of smart composite materials

using extrinsic Fabry-Perot interferometer and fiber Brag grating sensors", Smart

Materials and Structures vol. 11, 249-255, 2002

[6] M.Stewart and M.Cain, Optical sensor characterization and calibration device for

integrated optical systems", NPL Report MATC (A) 57, 2001.

[7]. E.Fitzer and W. Schafer, Carbon Vol. 8, 353,1970

[8]. R.P.Nathan, S.Bindu, Low Temperature Ablative Heat Shield for Re-Entry Vehicles.

In: Proceedings of AIAA thermo physics conference, Toronto, Ontario, Canada, 2005

[9]. L.Torre, J.M.Kenny, G.Boghetich, A.M.Maffezzoli, Degradation behaviour of a

composite material for thermal protection systems. Part III – char characterization. J.

Mater Sci. 35, 4536–4566, 2000

14

[10]. J.C. Giddings, theory of Chromatography in “Chromatography”: A laboratory hand

book Heftmann Ed., Van Nostrand Reinhold, NY,1975

[11]. K.Kobayashi, S.Sugawara, S.Toyoda and H. Honda, carbon, 6, 359, 1968.

[12]. S.J.Mitchell, R.S.Pickering and C.R.Thomas, J.Appl.Polym.Sci.14, 175, 1970.