Journal of Materials Processing Technology 210 (2010) 116–121
`
`Contents lists available at ScienceDirect
`
`Journal of Materials Processing Technology
`
`j o u r n a l h o m e p a g e : w w w . e l s e v i e r . c o m / l o c a t e / j m a t p r o t e c
`
`Study of thermal–mechanical properties of polyurethane foam
`and the three-dimensional shape of molded bra cups
`Kit-lun Yick ∗, Long Wu, Joanne Yip, Sun-pui Ng, Winnie Yu
`
`Institute of Textiles & Clothing, The Hong Kong Polytechnic University, Hong Kong
`
`a r t i c l e
`
`i n f o
`
`a b s t r a c t
`
`Article history:
`Received 4 December 2008
`Received in revised form 28 July 2009
`Accepted 9 August 2009
`
`Keywords:
`Thermo-mechanical analysis
`3D shape
`Polyurethane foam
`Bra cup molding
`
`Molded bras nowadays are dominating the overall bra market place, while the bra cups are mostly
`made of polyurethane foam sheets. The smooth and seamless inner surface of a molded bra cup gives
`three-dimensional nice fit to the wearer and provides unlimited designs for different softness and
`thickness. However, it has been a difficult question to determine the optimum molding conditions
`for various types of foam and there was no reliable method to measure the cup shape conformity. In
`this study, the properties of five polyurethane foams were investigated by thermal–mechanical analy-
`sis (TMA) and their 3D shapes formed in various molding conditions were measured by a Steinbichler
`Comet scanner and a new parameterization-based remesh algorithm method. The results revealed
`that the optimal temperature and dwell time for molding bra cups were greatly affected by the
`thermal–mechanical properties of polyurethane foams. The softening temperature and deformation
`properties of the foams tested by TMA can facilitate the determination of optimal molding temperature
`for the desirable cup shape and thickness. This study provides an effective and quantitative approach
`to eliminate the time-consuming “trial-and-error” in the molding tests traditionally being used in the
`industry.
`
`© 2009 Elsevier B.V. All rights reserved.
`
`1. Introduction
`
`The elastometric microcellular polyurethane (PU) foam is avail-
`able in a wide range of hardness, thickness, density, thermal,
`physical and mechanical properties. It is widely used in medical
`and apparel products as cushioning materials and wound dressing
`(Campbell et al., 2008; Park et al., 2007; Sakurai et al., 2007). For
`intimate apparel, PU foam is commonly used to mold 3D seamless
`bra cups to fit women’s breast shapes (Yu et al., 1998). The smooth
`inner surface of molded cup prevents irritation to the nipples and
`the outer surface provides a full and round breast shape as desired.
`Nevertheless, the hasty growth of market has brought significant
`challenges to the bra industry due to the severe shortages of appro-
`priate molding technologies. How to control the molding process
`and product quality accurately and effectively becomes a critical
`question.
`When producing bra cups, a contour molding machine with a
`pair of aluminum male and female molds in specific 3D shape is
`used. The heated male molds stretch and compress the originally
`flat foam sheets toward the female molds at a temperature over the
`foam’s material softening point. The high temperature was main-
`
`∗ Corresponding author. Tel.: +852 27666551; fax: +852 27731432.
`E-mail address: tcyick@inet.polyu.edu.hk (K.-l. Yick).
`
`0924-0136/$ – see front matter © 2009 Elsevier B.V. All rights reserved.
`doi:10.1016/j.jmatprotec.2009.08.004
`
`tained for about 1 min to allow the heat to transmit through the
`foam sheets which are held in a space between the male and female
`molds, so that the foam sheets are heat-set to the desired cup shape.
`Traditionally, the molded cups are assessed by visual examination
`against the “plastic shot” which is a transparent cup template used
`for checking the shape conformity of the outer cup surface (Yip and
`Ng, 2008). The quality of molded bra cups often varies with dif-
`ferent foam materials, molding temperatures and time (Yu et al.,
`1998).
`The processes of developing new master aluminum molds,
`molding experiments and quality inspection have been involv-
`ing lots of “trial-and-error” for at least 3 cycles (Myer and
`Montgomery, 1995). The quality of molded cups such as hand
`feel, appearance, shape conformity was assessed subjectively
`based on experience (Yu, 1996). There has been little literature
`found to provide accurate and reliable guidelines for bra cup
`molding.
`This study therefore aims at studying the relationships between
`the thermal–mechanical properties of PU foam materials and the
`optimal molding conditions for the required quality and shape
`of bra cups. It also presents an objective and quantitative way of
`assessing the 3D shape of PU foam cups by using 3D scanning and a
`new parameterization-based remesh algorithm method. This pro-
`vides useful knowledge for the determination of optimal molding
`conditions for bra cups.
`
`NIKE-1037
`
`

`

`Table 1
`Specifications and physical properties of the 5 PU foam materials.
`
`Style
`
`Density (kg/m3)
`
`Cell count (cells per 25 mm)
`
`Test standard
`
`ISO845-1988
`
`AS2282.5-1999
`
`A
`B
`C
`D
`E
`
`35.56
`28.23
`39.37
`45.07
`30.71
`
`47.2
`46.2
`49.5
`44.3
`43.5
`
`2. Experimental
`
`2.1. Physical properties of polyurethane foams
`
`Five PU foam sheets with thickness of 10 mm which are com-
`monly used for bra cup molding were sourced from a large bra
`cup molding company in the present study. Their physical proper-
`ties including density, cell count, extension, compression, hardness,
`and cell type were examined and summarized in Table 1.
`
`2.2. Thermal–mechanical analysis
`
`The thermal–mechanical properties of polyurethane foams
`were tested by the Thermomechanical Analyzer (Perkin Elmer TMA
`7, USA) which determines the deformation of a sample as a func-
`tion of temperature according to ISO11359-1:1999. Considering the
`high-strain deformation of foam at molding, a high tension probe
`load of 50 mN was used to compress the samples of same thickness
`(10 mm) that were heated from 25 ◦C to 250 ◦C at a heating rate of
`10 ◦C/min (Yu, 1996). The compression strains of foam at x ◦C (εx),
`in absolute values can be calculated by Eq. (1).
`εx(%) = |tx − to|
`× 100
`
`(1)
`
`to
`
`where to is the original thickness of foam, tx is the thickness of
`foam at x ◦C while x is set as room temperature 25, the molding
`temperatures 180, 200 or 220 respectively which are within the
`common temperature range of 180–220 ◦C generally used at foam
`cup molding.
`
`2.3. Foam cup molding process
`
`The PU foams were molded by using a contour molding machine
`of New Pads molding machine type DM-021HP4-2PR with a pair of
`size 34B mold heads (including male and female molds) as shown
`in Fig. 1a and b (Yu et al., 2006).
`PU foam sheet was placed on top of the fixed female mold head
`while the mold heads were heated to a pre-determined tempera-
`
`K.-l. Yick et al. / Journal of Materials Processing Technology 210 (2010) 116–121
`
`117
`
`Extension stress at 8%
`strain (kPa)
`ISO1798-1983
`
`Compression stress at 40%
`strain (106 kPa)
`ISO3386/1-1986
`
`Hardness (◦ShD)
`
`Cell Type
`
`ASTM D2240-05
`
`5.7
`2.8
`5.8
`4.8
`5.1
`
`5.27
`2.16
`4.72
`4.27
`1.15
`
`52.50
`21.96
`53.96
`43.98
`23.02
`
`Closed
`Open
`Open
`Open
`Open
`
`ture under control of thermostats, then the male mold heads were
`brought down vertically by compression air and released automat-
`ically after molding. The setting of molding temperature could be
`ranged from 180 ◦C to 220 ◦C and each adjustment was set at 10 ◦C
`interval. The dwell time of the molding process, ranging from 60 to
`180 s with 30 s interval for every adjustment, was determined by
`a time controller. The thickness of the foam cup is affected by the
`gap distance between the two mold heads, the material properties
`and molding condition. Specimens were allowed to cool down at
`room temperature atmosphere for no less than 24 h before the cup
`shape and cup thickness were measured (Yu et al., 1998).
`
`2.4. Measurement of 3D shape conformity
`
`An optical digitization system (Steinbichler Comet scanner, Ger-
`many) was used to measure the 3D shape conformity of outer
`surface of molded bra cups produced in various molding condi-
`tions. It applies projected grids and a high-resolution camera to
`capture 3D images, then accurate data sets can be generated by
`image analysis (Brosky et al., 2000).
`First, the highest bust point of the male mold head was defined.
`To measure the deviation of corresponding points on different
`scanned cup surfaces, a parameterization-based remeshing and
`registration algorithm method was used to characterize the 3D
`shapes of the convex surface of the scanned bra cups and a soft-
`ware interface was developed. The female mold head was set in a
`reference mesh, the outer surface of foam cup samples were then
`remesh based on the reference mesh. A rigid transformation was
`computed by matching the corresponding vertices around the bra
`cup peak. As shown in Fig. 2, the scanned surface of aluminum mold
`head and the fitting surface of a molded cup have been transformed
`respectively into 2D meshes for easy comparison.
`By aligning the bust point, the shape difference between the
`mold head and the outer surface of molded cup can be measured
`by a summation of deviations at all corresponding points in the two
`meshes as shown in Fig. 3.
`To quantify the degree of overall shape conformity between
`the molded cups and the mold head, the percentages of deviations
`
`Fig. 1. A contour molding machine. (a) The contour molding machine with male and female mold head and (b) schematic diagram of contour molding method.
`
`

`

`118
`
`K.-l. Yick et al. / Journal of Materials Processing Technology 210 (2010) 116–121
`
`Fig. 2. A software interface of calculating deviation.
`
`Fig. 3. Shape deviation between mold head and molded cup.
`
`(cid:2)
`
`T2 =
`
`(2)
`
`(3)
`
`between corresponding points are calculated as
`(cid:2)
`Num(|Xi − Mi| ≤ 1 mm)
`T1 =
`× 100%
`(cid:2)
`Num(|Xi − Mi| ≤ 10 mm)
`Num(|Xi − Mi| ≤ 1 mm)
`× 100%
`(cid:2)
`Num(Xi − Mi)
`where T1 is the calculated percentage of points which have devia-
`tion less than 1 mm in comparing with the number of points which
`have deviation less than 10 mm; while T2 is the calculated percent-
`age of points which have deviation less than 1 mm at 3 cm radius
`around the cup tip since the shape conformity at the cup tip region is
`critical and traditionally assessed in the industry. Mi is the arbitrary
`coordinate on the mold head and Xi is the corresponding coordi-
`nate on the molded cup. The function Num(i) is a number count of
`sampling points.
`
`2.5. Measurement of bra cup thickness
`
`In the industrial practice, the thickness of molded bra cups were
`measured by cutting cross-sectional horizontal and vertical lines
`passing through the cup peak. This has to destroy the cup sam-
`ple and involved handling errors. Yu et al. (1997) used non-contact
`moire topography and curve fitting to measure the thickness of
`molded foam cup along West–East (W–E) and North–South (N–S)
`directions. However, moire method gave limited resolution for
`images and the determination of cup peak relied on human eyes.
`
`In this study, a non-contact optical microscopy measuring
`instrument (LEICA QWin, Germany) was adopted to scan the foam
`cups as shown in Fig. 4. The cup peak was identified by com-
`puter software. Measurement points on the outer surface with
`equal distance of 10 mm between each point were used for thick-
`ness measurement along the horizontal and vertical cross-sectional
`lines passing through the cup peak. As the outer curve is longer than
`the inner curve, perpendicular lines intersecting each measure-
`ment point on the corresponding tangent touching the outer cup
`surface were drawn to measure the thickness at specific marked
`locations.
`
`3. Results and discussion
`
`3.1. Thermal–mechanical properties of polyurethane foams
`
`The TMA scans of the 5 PU foams under constant load of 50 mN
`are presented in Fig. 5. The extrapolated onset temperatures of the
`compression curves indicate the softening points of the 5 foams
`respectively. It is obvious from the TMA scans that foams B, C, D
`and E are very stable and do not show significant deformation until
`120–160 ◦C. On the other hand, foam A started deforming at a much
`lower temperature (around 60 ◦C) and the foam further deformed
`at temperature of 165 ◦C. Physical cross-linking between hard seg-
`ments was substantially damaged around 220 ◦C in foam A. For the
`other 4 foams, such failure of physical cross-linking did not start
`till a relatively higher temperature of 240 ◦C.
`
`

`

`K.-l. Yick et al. / Journal of Materials Processing Technology 210 (2010) 116–121
`
`119
`
`density foams showed major deformation behaviour at high tem-
`peratures, whilst the effects of high temperatures on low density
`foams were relatively less significant. The values of compressive
`strains only increased gently with the increase of heating temper-
`ature in foams B and E.
`Although foam A has a relatively high density of 35.56 kg/m3
`with a low compressive strain of 50.40% at 25 ◦C (ε25), it started
`deforming obviously at around 60 ◦C. The compressive strain
`increased extraordinarily to 96.92% at 220 ◦C (ε220). This may be
`explained by the fact that foam A has a closed-cell structure which
`has different plastic compositions from that of the open-cell struc-
`tures in other four foam samples (Klempner and Sendijarevic,
`2004). PU is made up of the PEG (polyethylene glycol) and the TDI
`(Toluene diisocyanate). The melting point of the PEG soft segment
`is about 50 ◦C (Su and Liu, 2007). Foam A has a higher content of
`soft segment as compared with the other four foams in this study.
`When the temperature exceeds the melting temperature of PEG,
`the plentiful soft segments start to fuse and the elastic proper-
`ties of the foam began to decline. Foams B, C, D and E have higher
`contents of hard segment and thus withstand deformation at high
`temperatures more effectively.
`As the soft PU foams under investigation have cross-linked
`thermosetting properties, the processing temperature of foam cup
`molding could range between the lower region of the softening
`point of hard segment and the upper region of pyrolysis tem-
`perature (Jiang et al., 2006). This phenomenon indicates that the
`softening temperature obtained from TMA scans can be taken as the
`lower bound of bra cup molding temperature, for example, 165 ◦C
`for foam A.
`
`3.2. Foam molding performance and 3D shape conformity of
`foam cups
`
`The molding results revealed that molding temperature and
`dwell time greatly affect the foam cup performance. It is evident
`that high molding temperatures and long time applied generally
`resulted in better shape conformity in all of the 5 foams studied.
`Nevertheless, excess molding temperature not only resulted in high
`incidence of yellowing, bubbling and hand feel problems, but also
`failed to conform to the desirable shape of the mold head. Fig. 6
`shows the spectrums of the mean deviations between the mold
`head and the cup samples of foam D obtained at various molding
`conditions.
`When the molding temperature was 180 ◦C and the dwell time
`was 60 s, large deviations were observed between the two surfaces
`as in Fig. 6(a). The minimum deviations were found at 190 ◦C and
`120 s as in Fig. 6(b). When molding temperature reaches 210 ◦C
`and/or the dwell time prolonged to 180 s, the level of shape confor-
`mity declines readily as shown in Fig. 6(c) and (d). It is noteworthy
`that relatively large deviations existed at two sides of the cup rim in
`all molding conditions. The results can be explained by the mechan-
`ical loadings applied to the foam sheet during the molding process
`as the cup rim had the largest geometric deformation of compres-
`sion and extension (Yu et al., 1998).
`Using Eqs. (2) and (3), the shape conformity is defined as the per-
`centage of point deviations between the mold head and the foam
`
`Fig. 4. Thickness measurement of molded cup sample along W–E and N–S direc-
`tions. (a) Horizontal cross-section line (W–E) and (b) vertical cross-section line
`(N–S).
`
`Fig. 5. TMA scans of different polyurethane foams.
`
`The compression strains of foam materials in a 50 mN load under
`different temperatures are reported in Table 2. The value at 25 ◦C
`(ε25) of foams B and E are much higher than those of foams C and
`D due to their lower density and hardness as presented in Table 1.
`Foams with lower densities (foams B and E) generally have lower
`compression resistance (i.e. indentation hardness) than high den-
`sity foams (foams C and D). However at the softening temperature
`range of 180–220 ◦C, the compressive strains (ε180, ε200 and ε220) of
`foams C and D increased substantially. The results indicate that high
`
`Table 2
`Thermo-mechanical properties of foam at 50 mN load.
`
`Sample
`
`to (mm)
`
`t25 (mm)
`
`A
`B
`C
`D
`E
`
`10.12
`9.97
`9.98
`10.00
`10.27
`
`5.02
`1.85
`5.66
`5.12
`2.08
`
`ε25 (%)
`
`50.40
`81.44
`43.30
`48.81
`79.75
`
`t180 (mm)
`
`ε180 (%)
`
`t200 (mm)
`
`ε200 (%)
`
`t220 (mm)
`
`1.72
`1.29
`3.24
`3.44
`1.32
`
`83.00
`87.06
`67.54
`65.61
`87.15
`
`0.8
`1.01
`2.36
`2.18
`0.97
`
`92.09
`89.87
`76.36
`78.20
`90.56
`
`0.31
`0.75
`1.68
`1.41
`0.70
`
`ε220
`
`96.92
`92.50
`83.22
`85.87
`93.21
`
`

`

`120
`
`K.-l. Yick et al. / Journal of Materials Processing Technology 210 (2010) 116–121
`
`Fig. 6. Spectrums of mean deviations between mold head and cup samples (Foam
`D) at various molding conditions (small deviations in blue and large deviations in
`red). (a) Foam D at 180 ◦C and 90 s, (b) foam D at 190 ◦C and 120 s, (c) foam D at
`190 ◦C and 180 s and (d) foam D at 210 ◦C and 90 s.
`
`Table 3
`Optimum molding conditions for the 5 foam materials studied.
`Temperature (◦C)
`180
`200
`190
`190
`200
`
`Dwell time (s)
`
`T1
`
`180
`90
`120
`120
`90
`
`38.62%
`40.10%
`38.62%
`41.85%
`35.54%
`
`Foam
`
`A
`B
`C
`D
`E
`
`T2
`
`94.34%
`89.47%
`96.97%
`98.68%
`80.67%
`
`curve that foam A is highly sensitive to thermal changes and starts
`deforming at 60 ◦C. To achieve the desirable cup shape, a lower tem-
`perature must be used in foam A, whilst a relatively longer dwell
`time is recommended. For low densities foams (foams B and E),
`a higher molding temperature is recommended, whilst the dwell
`time should not be too long to avoid potential problems of foam
`yellowing and bubbling. On the other hand, the molding tempera-
`ture of high densities foams (foams C and D) should be lower than
`those low densities foams.
`It is also noteworthy that even though the optimum mold-
`ing conditions were determined, the shape of foam cup may
`not necessarily fits the aluminum mold head perfectly. Amongst
`the 5 foam materials studied, the degree of shape conformity
`of the foam cups was around 40%. The degree of shape con-
`formity at the cup tip region ranged from 80.67% to 98.68% at
`deviations less than 1 mm level. Bra manufacturers therefore can
`quantify the desired degree of shape conformity at cup tip region
`at 80% target level so as to control the molding process more
`effectively
`
`3.3. Thickness of foam cups
`
`The thickness of foam cups at various locations along the N–S
`and W–E directions is measured. It is obvious that the foam
`cup thickness of all of the 5 foams decreases with increased
`dwell time until the optimum dwell time is reached. When
`the dwell time is too short, significant thickness variations are
`observed around the bust point position, whilst small variations
`are found at cup rims because it has the largest geometric defor-
`mation during the molding process. However, prolonged dwell
`time could also result in excessive thermal shrinkage, in which
`the foam cup thickness is thinner than the gap between mold
`heads.
`Fig. 7(a) shows the thickness (gap distance) between the male
`and female mold heads and the cup thickness of foam D at the
`molding temperature of 190 ◦C with dwell times of 60, 90 and 120 s
`respectively. After comparing against the mold head curve, the opti-
`mum dwell time of foam D at molding temperature of 190 ◦C is
`found to be 120 s, the thickness of the cup sample matches per-
`fectly with the gap distance between the male and female mold
`heads. Fig. 7(b) reveals that the optimum dwell time of foam B at
`molding temperature of 200 ◦C is 90 s. These results also coincide
`with the optimum molding conditions obtained by measuring the
`3D shape conformity of foam cups.
`
`cups obtained at various molding conditions. The minimum devi-
`ation determines the optimum molding conditions for the 5 foam
`materials as presented in Table 3.
`Foam A has a lower optimum molding temperature as com-
`pared with the other 4 foam materials. It is evident from the TMA
`
`Fig. 7. Thickness of cup samples measured at various locations. (a) Cup thickness along W–E direction of foam D molded under 190 ◦C and (b) cup thickness along N–S
`direction of foam B molded under 200 ◦C.
`
`

`

`4. Conclusion
`
`Acknowledgement
`
`K.-l. Yick et al. / Journal of Materials Processing Technology 210 (2010) 116–121
`
`121
`
`In this study, five different polyurethane foam materials were
`molded to a size 34B’s bra cup in various molding condition. It is
`evident that the optimal molding temperature and dwell time is
`greatly affected by the thermal–mechanical properties of the foam
`materials. The TMA scans revealed that the compressive strain of
`molded foams is closely related to the density of the foam materials.
`Of which, foams of high densities or closed-cell structures are more
`sensitive to thermal changes than those of low density foams. Thus,
`lower molding temperatures of 180–190 ◦C are recommended so as
`to achieve the most desirable 3D cup shape and thickness. To some
`extent, foams that may deform rapidly during the heating process
`and result in extraordinarily high values of compressive strain at
`high temperatures, so they are not recommended for the use of
`foam cup molding since a low molding temperature and long dwell
`time must be used.
`The 3D shape of the foam cups molded at various molding
`conditions were assessed and results indicated that foam samples
`molded at low molding temperatures generally failed to conform
`to the desirable shape. Prolonged dwell time may result in high
`incidence of foam yellowing and shrinkage in cup thickness, not to
`mention the low efficiency. It was also found that the softening tem-
`perature of hard segment obtained from TMA scans can be taken
`as the lower bound of bra cup molding temperature, where mold-
`ing temperature must be higher than the softening temperature of
`foam material.
`This experimental work also presents an objective and quanti-
`tative way of assessing 3D geometry of polyurethane foam cups.
`By using the parameterization-based remesh algorithm method,
`the shape conformity can be evaluated and quantified effectively.
`The optimal molding conditions for particular types of foam mate-
`rials can also be determined easily. This not only provides useful
`information for selecting a suitable foam material for the molding
`purpose, but also facilitates the control of bra cup molding pro-
`cess and communication amongst different sectors of the intimate
`apparel industry.
`
`The authors would like to thank the Research Grant Council for
`funding this research through the project account PolyU 5317/06E.
`
`References
`
`Brosky, M.E., Pesun, I.J., Lowder, P.D., DeLong, R., Hodges, J.S., 2000. Laser digitization
`of casts to determine the effect of tray selection and cast formation technique on
`accuracy. In: 78th General Session of the International-Association-for-Dental-
`Research. Mosby, Inc., Washington, DC, pp. 204–209.
`Campbell, P.E., Smith, G.S., Smith, J.M., 2008. Retrospective clinical evaluation of
`gauze-based negative pressure wound therapy. International Wound Journal 5
`(2), 280–286.
`Jiang, Z., Yuan, K., Li, S., Chow, W.K., 2006. Study of FTIR spectra and thermal
`analysis of polyurethane. Spectroscopy and Spectral Analysis 26 (4), 624–
`628.
`Klempner, D., Sendijarevic, V., 2004. Handbook of polymeric foams and foam tech-
`nology. Hanser Gardner, Munich, Germany, pp. 22–24.
`Myer, R.H., Montgomery, D.C., 1995. Response Surface Methodology—Process and
`Product Optimization Using Designed Experiments. John Wiley & Sons, NY, p.
`12.
`J.N., 2007. Effects of hardness and thickness of
`Park, C.C., Choi, W.S., Lee,
`polyurethane foam midsoles on bending properties of the footwear. Fibers and
`Polymers 8 (2), 192–197.
`Sakurai, A., Hashikawa, K., Yokoo, S., Terashi, H., Tahara, S., 2007. Simple dress-
`ing technique using polyurethane foam for fixation of skin grafts. Dermatologic
`Surgery 22 (8), 976–979.
`Su, J., Liu, P., 2007. Effect of hard and soft segments on the heat storage proper-
`ties of polyethylene glycol-based polyurethanes. Acta Polymerica Sinica (2), 97–
`102.
`Yip, J., Ng, S.P., 2008. Study of three-dimensional spacer fabrics: physical and
`mechanical properties. Journal of Materials Processing Technology 206 (1–3),
`359–364.
`Yu, W.M., 1996. The effects of polyurethane foam properties and moulding condi-
`tions on the shape characteristics of brassiere cups. PhD thesis, Department of
`Textile Industries, The University of Leeds.
`Yu, W.M., Fan, J., Harlock, S.C., Ng, S.P., 2006. Innovation and Technology of Women’s
`Intimate Product. Woodhead Publishing, Cambridge, p. 223.
`Yu, W.M., Yeung, K.W., Harlock, S.C., Leaf, G.A.V., 1998. Predicting energy require-
`ments for moulding flexible polyurethane foam sheets. Textile Research Journal
`68 (4), 289–295.
`Yu, W.M., Yeung, K.W., Harlock, S.C., Leaf, G.A.V., 1997. Mathematical measures of
`shape conformity. Journal of China Textile Institute 14 (3), 21–27.
`
`