Key Engineering Materials Vol 626 (2015) pp 182-186 Online: 2014-08-18

© (2015) Trans Tech Publications, Switzerland

Understanding the formation mechanism of residual stress during

Keywords: Precision glass moulding, Residual stress, Finite element analysis, Optical lens,
Constitutive model

Abstract. Precision glass moulding is a technique that enables the production of optical lenses of
complex geometries in a single step. However, it has been reported that the product quality highly
depends on the properties of a raw material, the design of a die, and the selection of a processing
program. This paper aims to reveal the formation mechanism of the residual stresses by optical lens
moulding. To this end, a modulus-based constitutive model was developed to integrate with the
deformation and stress analyses by the finite element method. The investigation showed that the
residual stresses are caused by the variability and heterogeneity of thermal expansion in a lens, but
that they can be reduced effectively by decreasing the rate of cooling within the stage from the
molding temperature to the glass transition temperature.

Introduction
Traditional processes such as grinding and polishing are time consuming, are associated with high
production cost, and are inefficient in the manufacturing of various advanced optical products with
small-scale and complicated features [1]. Alternatively, precision glass moulding (PGM) with
accurate force and temperature control, enables to produce aspherical lenses and other irregular
optical component in a single step. Hence, this process has received significant attention in past
decades [2-5].
However, to develop an effective PGM also faces significant scientific and technical challenges.
For example, PGM is carried out within the glass transition region of opitical glass above its glass
transition temperature Tg, in which the material has an unstable non-equilibrium micro-structure [6].
Within a narrow temperature variation window of 100 °C, the viscosity of the glass can change from
10-5 Pa·s to 1012 Pa·s [7], making its mechanical performance sensitive to the environment
temperature. This is because the structural relaxation of the atomic configurations in this temperature
window is strongly dependent on the time and thermal history [8]. The complexity of these issues
leads to shape distortion and residual stresses in a lens formed. Hence, deep scientific understanding
has become imperative to underpin the technology development.
To minimize the traditional trial-and-error process in the PGM development, numerical
simulations have been considered a useful tool [4, 9], of which the finite element (FE) method is
probably the most widely used. For example, Arai et al. [4] estimated the residual stresses in a glass
lens by a fundamental FE simulation of press moulding. Ananthasayanam [2, 3] proposed a coupled
thermo-mechanical FE model for the simulation of PGM. Jain et al. [9] discussed the viability of
using numerical simulation to predict the performance of PGM, but highlighted that a more
sophisticated model is needed to make a accurate prediction and to understand the key issues during
the process such as residual stresses.
This paper will comprehensively analyze a typical process of PGM using the FE method. A focus
will be to establish a modulus-based constitutive model of glass materials to reveal the formation
mechanism of residual stresses during PGM.

All rights reserved. No part of contents of this paper may be reproduced or transmitted in any form or by any means without the written permission of Trans
Tech Publications, www.ttp.net. (ID: 130.216.129.208, University of Auckland, Auckland, New Zealand-11/06/15,01:39:01)
 Key Engineering Materials Vol. 626 183

Methodology
Constitutive Modeling. Optical glass P-BK7 (composition: SiO2-BO-KO-NaO-BaO) has been
widely used for the thermoforming of precision lenses because of its low Tg [10]. Thus this section
will develop a modulus-based constitutive relationship to describe the material’s behavior.
Above the Tg, the optical glass becomes a viscoelastic material whose deformation has strongly
resistance to volumetric changes but can comply with shear easily. As such, the strain tensor εij and
stress tensor σij can be divided into volumetric and deviatoric parts, i.e.,

ε ij = (∂ui ∂x j + ∂u j ∂x i ) = eij + tr(ε )δ ij 3 , σ ij = Sij + tr(σ )δ ij 3 , (1)

where u is the displacement vector, eij and Sij are the deviatoric strains and stresses, respectively,
tr(ε) and tr(σ) are the traces of the strain tensor and stress tensor, respectively, and δij is the Kronecker
delta. The bulk viscosity of the material can be assumed to be infinite, i.e. time- and
temperature-independent. Hence, the constitutive relationship of the volumetric stress and strain can
be written as

1 1
tr (ε ) − α ∆T = tr (σ ) , (2)
3 9K

where K is the bulk modulus, and α is the coefficient of thermal expansion (CTE). The relationship of
deviatoric stress and strain can be described by a standard linear solid (SLS) model [11]

Gr G S
ij Sij
(1 + )e
ij + r eij = + , (3)
GM ηs 2GM 2ηs

where Gr is the modulus in the elastic branch of the SLS model, reflecting the residual modulus of the
glass material, and can be assumed as a constant, GM is the modulus in the Maxwell branch of the SLS
model, and ηs is the shear viscosity.
To accomplish the constitutive model, one has to determine the temperature-dependent properties,
i.e., GM, ηs, and α. In this study, we obtain the temperature dependence of the shear modulus by using
an impulse excitation technique (IET). The shear viscosity is derived from an elastic model (shoving
model) [12] which assumes that the main contribution to the activation energy of a flow event is the
energy needed for atoms to shove aside the surrounding, proportional to the instantaneous shear
modulus of the liquid. Hence, the viscosity becomes a function of shear modulus, i.e.,

ηs = η0 exp(VcGM / k BT ) , (4)

where η0 is a characteristic viscosity, kB is Boltzmann constant, and Vc is a characteristic microscopic

α l = α G + (α L − α G )δT f δT , (5)

where Tf is the fictive temperature of glass material, determined by fitting the relaxation data of
modulus [13], and αG and αL are the reference CTEs at the glassy state and liquid state, respectively.
The above constitutive model can be programmed into a commercial finite element software
(ABAQUS in this study). The model validity has been verified by a series of relaxation tests.
Simulation of PGM. Based on ABAQUS and the new model, a numerical platform of PGM has been
established, as shown in Fig. 1a. The raw material is a P-BK7 glass ball with a diameter of 5 mm. Both
the lower and upper dies are made of Tungsten Carbide, and have a sphercial concave surface with the
radius of 8.5 mm and depth of 1 mm. A typical process of PGM can be divided into four stages:
184 Advances in Engineering Plasticity XII

heating, moulding, cooling, and demoulding. The cooling stage can be further divided into two parts
with differenct cooling rates: the inital cooling stage above Tg and the second cooling stage below Tg.
Considering the symmetry of the set-up, the process can be analyzed by using only a quarter of the
model to improve the efficiency and accuracy of a simulation by applying the corresponding
symmetry boundary conditions. The time-dependent pressure is applied on the top surface of the
upper die, and the enviroment tempearture is set on the surfaces of the dies, as shown in Fig. 1b. In
additional, a fixed boundary condition is applied to the bottom of the lower die. The entire model is
divided by explicit temperature-coupled elements. Frictionless and thermal conductance can be
applied on the interfaces.
(a) (b)

Fig. 1. (a) The FE model of PGM, and (b) typical time history curves of temperature and pressure.

Results
Fig. 2 presents the residual stress distributions in a lens after glass moulding, including both the
hydrostatic stress and von Mises stress. It can be seen (Fig. 2a) that the interior of the lens sustains
tensile residual stress up to -8 Mpa. On the contrary, the exterior surface of the lens is under high
compressive stress. Between these two zones, there is an intermediate area with a low residual stress.
This can also be observed from the distribution of von Mises stress (Fig. 2b). The minimums of von
Mises stress are at the top and bottom subsurfaces symmetrically. The heterogeneity of the residual
stresses can vary the local density and the refractive index in the glass material [14].

(a) (b)

Fig. 2. The distributions of residual (a) hydrostatic stress, and (b) von Mises stress.
To reveal the formation mechanism of the residual stresses, the evolutions of the von Mises
stresses at three points, i.e. at the top, middle and bottom of the lens, are shown in Fig. 3a. It can be
seen that initially the internal stress is very low. When the temperature approaches Tg at around the
forming time of 270 s, however, the internal stresses jump to a high plateau. Such a high stress
remains until the finish of the PGM process, and becomes the final residual stresses. The distributions
of the stresses along the middle line are given in Fig. 3b. It can be seen that from 279.5 s to 299.5 s,
both the magnitude and gradient of the internal stresses increase significantly.
 Key Engineering Materials Vol. 626 185

(a) (b)

Fig. 3. (a) The evolutions of the von Mises stresses at three points, and (b) the stress distributions
along the middle line of the lens.
The stress jump is because of the variability and heterogeneity of the thermal expansion in lens. To
clarify this, the evolutions of the CTE of the top and middle points are examined, as shown in Fig. 4a.
When the internal stress jumps, the CTE deceases from a high plateau to a lower one. Moreover, the
decrease of the CTE at the middle point is slower than that of the top because of the slow cooling that
the former experiences. Fig. 4b presents the difference between the CTEs of the two points, showing
that it reaches its maximum at around 289.5 s corresponding to the sudden change of the stress
distribution revealed in Fig. 3b.
(a) (b)

Fig. 4. (a) The evolutions of the CTEs at the middle and top points, and (b) their difference.
As the jump of internal stress occurs during the cooling stage, it is expected that one can reduce the
residual stress by controlling the cooling rate. Fig. 5a shows the evolutions of the von Mises stress of
the middle point under three different cooling rates in the first cooling stage (above Tg). It indicates
that if one reduces the cooling rate of the first stage, the residual stress also decreases accordingly. In
contrast, the final residual stress is hard to be influenced by the cooling rate of the second stage (below
Tg), see Fig. 5b. Thus one can decrease cooling rate above Tg to achieve a low residual stress, and
increase the cooling rate below Tg to save the processing time.
(a)
(b)

Fig. 5. The effect of cooling rate on internal stress. (a) initial cooling stage, and (b) second cooling
stage.
186 Advances in Engineering Plasticity XII

Summary
This paper has developed a new model for describing the constitutive behavior of optical glass at a
temperature above the material’s glass transition. Base on this, the study has revealed the formation
mechanism of residual stresses during a PGM process by using the FE simulation. The investigation
concludes that during the cooling process the residual stress in a lens after thermoforming is due to the
stress jump induced by the variability and heterogeneity of the thermal expansion across the lens. The
residual stress can be reduced effectively by decreasing the cooling rate from the molding temperature
to the glass transition temperature.

Acknowledgement
The authors appreciate the Australian Research Council for its financial support to this work.

References
[1] Firestone, G.C., A. Jain, and A.Y. Yi, Precision laboratory apparatus for high temperature
compression molding of glass lenses. Review of Scientific Instruments, 2005. 76(6): p.
6115-6122.
[2] Ananthasayanam, B., et al., Final Shape of Precision Molded Optics: Part I-Computational
Approach, Material Definitions and the Effect of Lens Shape. Journal of Thermal Stresses, 2012.
35(6): p. 550-578.
[3] Ananthasayanam, B., et al., Final Shape of Precision Molded Optics: Part Ii-Validation and
Sensitivity to Material Properties and Process Parameters. Journal of Thermal Stresses, 2012.
35(7): p. 614-636.
[4] Arai, M., Y. Kato, and T. Kodera, Characterization of the Thermo-Viscoelastic Property of Glass
and Numerical Simulation of the Press Molding of Glass Lens. Journal of Thermal Stresses,
2009. 32(12): p. 1235-1255.
[5] Yi, A.Y. and A. Jain, Experimental and numerical analysis compression molding of aspherical
glass lenses - A combined. Journal of the American Ceramic Society, 2005. 88(3): p. 579-586.
[6] Debenedetti, P.G. and F.H. Stillinger, Supercooled liquids and the glass transition. Nature, 2001.
410(6825): p. 259-267.
[7] Angell, C.A., Structural Instability and Relaxation in Liquid and Glassy Phases near the Fragile
Liquid Limit. Journal of Non-Crystalline Solids, 1988. 102(1-3): p. 205-221.
[8] Scherer, G.W., Relaxation in Glass and Composites. 1986, New York: Wiley.
[9] Jain, A. and A.Y. Yi, Numerical modeling of viscoelastic stress relaxation during glass lens
forming process. Journal of the American Ceramic Society, 2005. 88(3): p. 530-535.
[10] Optics, J.D.R.P., Precision Glass Molding Technical Brief. 2007.
[11] Richter, F., Upsetting and Viscoelasticity of Vitreous SiO2: Experiments, Interpretation and
Simulation Viskoelastizität von Kieselglas: Experimente, Interpretation und Simulation, in
Technische Universität Berlin. 2006, Technische Universität Berlin: Berlin.
[12] Dyre, J.C., N.B. Olsen, and T. Christensen, Local elastic expansion model for viscous-flow
activation energies of glass-forming molecular liquids. Physical Review B, 1996. 53(5): p.
2171-2174.
[13] Liu, W.D., H.H. Ruan, and L.C. Zhang, Revealing structural relaxation of optical glass through
Young’s modulus. 2014. (To be submitted)
[14] Ritland, H.N., Relation between Refractive Index and Density of a Glass at Constant
Temperature. Journal of the American Ceramic Society, 1955. 38(2): p. 86-88.
Advances in Engineering Plasticity XII
10.4028/www.scientific.net/KEM.626

Understanding the Formation Mechanism of Residual Stress during Precision Glass Moulding
10.4028/www.scientific.net/KEM.626.182