УДК 541.64:542.943


© 2013 г. Svetlana Terekhina", Marion Mille*, Bruno Fayolle*, and Xavier Colin*

a Arts de Metiers ParisTech, LAMPA, 2 Boulevard du Ronceray, BP 93525, 49035Angers, France b Arts de Metiers ParisTech, PIMM UMR 8006, 151 Boulevard de l'Hôpital, 75013 Paris, France e-mail : Svetlana.TEREKHINA@ensam.eu

Received October 26, 2012 Revised Manuscript Received February 7, 2013

Abstract — The thermal ageing of a neat epoxy matrix has been studied at 473 K in air by three complementary analytical techniques: optical microscopy, dynamic mechanical analysis and nano-indentation. Thermal oxidation is restricted in a superficial layer of about 195 ^m of maximal thickness. It consists in a predominant chain scission process involving, in particular, chemical groups whose P motions have the highest degree of cooperativity and thus, are responsible for the high temperature side of P dissipation band. As a result, chain scissions decrease catastrophically the glass transition temperature, but also increase significantly the storage modulus at glassy plateau between Tn and Ta. This phenomenon is called "internal antiplasticization". Starting from these observations, the Di Marzio and Gilbert's theories have been used in order to establish relationships between the glass transition temperature and number of chain scissions, and between the storage modulus and P transition activity respectively. The challenge is now to establish a relationship between the P transition activity and the concentration of the corresponding chemical groups.

DOI: 10.7868/S0507547513090080


There is a lack of organic matrix composite materials for civil aeronautical applications above 473 K in Europe. Highly aromatic epoxy matrices are candidate for such applications because of their high ther-momechanical performances. These thermostable matrices have the ability to maintain their elastic and fracture properties up to temperatures close to their glass transition temperature (typically up to Tg — 30 K). However, they will be used by airline companies only if their long-term durability in current use conditions is clearly demonstrated.

There is large amount of literature works devoted to the thermal degradation mechanisms and kinetics of epoxy matrices showing that oxidation is clearly the predominant ageing process [1—4]. A key feature of such process is that degradation is non-uniform in the sample thickness because oxidation kinetics is diffusion controlled. A non-empirical kinetic model has been derived from a realistic oxidation mechanistic scheme for epoxy matrices in order to access the spatial distribution (in the sample thickness) of structural changes at the molecular scale as a function of exposure time [1—4]. This mechanistic scheme can be summarized as follows:

(1) Initiation: SPOOH ^ aP° + p PO° + H2O (kx) (2) Propagation: P° + O2 ^ PO° (k2)

1 The article is published in the original.

(3) Propagation PO° + PH ^ POOH + P° (k3) (4) Termination: P° + P° ^ inactive products (k4)

(5) Termination: P° + PO° ^ inactive products (k5)

(6) Termination: PO° + PO° ^ ^ inactive products + O2 (k6)

where POOH, P°, PO°, PH and O2 represent hydroperoxides, alkyl and peroxy radicals, polymer substrate and oxygen respectively.

8 is the molecularity of initiation reaction such as: if 8 = 1, then a = 2 and в = 0 else if 8 = 2, then a = в = 1 and k are elementary rate constants.

It is a radical chain oxidation producing its own initiator: hydroperoxide POOH. This closed-loop character is responsible for the sharp auto-acceleration of thermal oxidation at the end of the induction period [5].

There is also an important amount of literature works devoted to the consequences of oxidation on viscoelastic and/or mechanical properties of epoxy matrices [6—8]. Most of the authors have tried to relate directly the oxidation conversion ratio to these macroscopic properties, whereas it was clearly established that these latter are essentially altered by structural changes taking place at upper scales, in particular at the macromolecular scale: chain scissions and crosslinking [9, 10].

The objective of the present article is to complete the previous non-empirical kinetic model in order to predict, in a near future, the consequences of oxidation on the viscoelastic properties of epoxy matrices. This operation will involve two successive stages:

Firstly, main reactions responsible for structural changes taking place at the macromolecular scale, i.e. chain scissions and crosslinking, will be briefly recalled and introduced in the previous mechanistic scheme. As a result, two new kinetic equations, giving access to the numbers of chain scissions S and crosslinking events X versus exposure time, will be derived from the mechanistic scheme.

Secondly, the Di Marzio [11] and Gilbert's theories [12] will be used to establish non-empirical relationships between S and X and some particularly important viscoelastic properties for aeronautical applications such as: glass transition temperature Tg and Young's modulus E respectively. Since the coefficients of these

equations have a real physical meaning, their values will be assessed from the theoretical structure of the virgin epoxy network under study.



The ideal epoxy network under study results from the reaction of a mixture of two epoxide monomers: tris(4-hydroxyphenyl)methane triglycidyl ether (Tac-tix 742) and diglycidyl ether of bisphenol A (Tactix 123), with an aromatic diamine hardener: 4,4-diami-nodiphenyl sulfone (DDS). The chemical structures of these products are given in Scheme 1. The mass fraction for synthetizing the ideal epoxy network of Tactix 742 was 54.3%, Tactix 123 - 18.1%, and DDS - 27.6%.

Neat epoxy plates (of 2 mm thick) were processed by autoclave moulding in accordance with the recommended cure cycle: 1 hour at 413 K and 3h at 453 K (with a temperature ramp of 2 K/min). Then, the samples were post-cured at 503 K for 10 hours under primary vacuum in order to reach the maximum crosslinking density while preventing any pre-oxida-tion before thermal ageing.

Test Conditions and Characterization

Ageing Conditions. All the plates were stored in the dry atmosphere of a desiccator prior thermal ageing experiments. They were subjected to isothermal exposure at 473 K in an air-circulating oven and were removed intermittently to be machined into rectangular samples and characterized by optical microscopy, dynamic mechanical analysis and nano-indentation.

Optical Microscopy Examinations. Optical microscopy examinations were performed using an interfer-ential contrast in order to better visualize the superficial oxidized layer and, then, try to determine its aver-

age thickness. Sample cross-sections were polished with a series of abrasive papers classified according to a decreasing order of the particle size of silicon carbide (from 600 to 4000). Then, a mirror finish was obtained with a diamond paste (of particle diameter ranged between 1 and 3 ^m). Finally, all surfaces to be examined were ultrasonically cleaned in ethanol for 5 min.

Dynamic Mechanical Analysis (DMA). DMA is a convenient and sensitive technique for a rapid determination of the viscoelastic properties of polymers and polymer-based materials as a function of frequency and temperature. It consists in the observation of the time-dependent behaviour of a material under a dynamic strain or stress.

In this study, DMA was performed on parallelepi-pedic barrels (25 x 2 x 1 mm3) machined from plates in a tension/compression mode with a Dynamic Mechanical Analyzer (DMA Q800, TA Instruments). The tests were carried out at a controlled sinusoidal strain in the linear domain of material viscoelasticity and the corresponding stress was measured. From

TOL, um 200 -







3000 Time, h

100 urn i_I_i

• "I

culated from the slope of the curve at F = Fm Eq. (2):

El =




where O is a parameter depending on the indenter type (O = 1.013 for a Vickers indenter) and Ap is the contact area between the indenter and the sample, projected on the plane perpendicular to the indenter axis: Ap = 1/2(h x b), h and b being the height and the base of the projected equilateral triangle.


Determination of the Thickness of Oxidized Layer

A superficial oxidized layer was clearly evidenced by optical microscopy according to the procedure described in the experimental section. This latter appears rougher and brighter than the specimen core. Example of micrograph for sample aged 3402 hours at 473 K in air is given in Fig. 1. By choosing the whole depth of visible changes as arbitrary criterion for defining the skin/core boundary, a thickness of oxidized layer (TOL) was determined. In Fig. 1, it can be seen that TOL increases with exposure time to finally tend towards an asymptotic value of about 195 ^m after 1000 hours of exposure.

Fig. 1. (a) Average thickness of oxidized layer (TOL) versus exposure time at 473 K in air. (b) Micrograph of sample aged for 3402 hours.

stress and strain values, the complex modulus was calculated:

E * = E' + jE",


where E' is the storage (elastic component) modulus, E'' is the loss (viscous component) modulus and tanS = E "/E' is the loss factor or damping, from which the phase shift S, between stress and strain, is extrapolated.

The frequency and heating rate were set at 1 Hz and 2 K/min respectively. The temperature was ranged from 163 to 603 K. The principal relaxation temperature Ta, associated to the glass transition temperature Tg, was taken at the maximum of the a dissipation band.

Nano-Indentation Measurements. Indentation measurements were performed with the nano-indent-er of an atomic force microscope (AFM, Veeco) [13]. The force-displacement curve (F = f(d)) collected during the indentation experiment provides

Для дальнейшего прочтения статьи необходимо приобрести полный текст. Статьи высылаются в формате PDF на указанную при оплате почту. Время доставки составляет менее 10 минут. Стоимость одной статьи — 150 рублей.

Показать целиком