Logo Euroceram Carte d'Europe
Welcome pagePresentation of the regions participating in the networkPresentation of the ceramic industry in the regions participating in the networkSeminars calendarNewsletter on lineAddressesAsk questions to expertsAdministering SME's data
FR  |  ENGL
Fabrication of ceramic parts by stereolithography
T. CHARTIER*, F. DOREAU**, C. CHAPUT**, M. LOISEAU**
* SPCTS, UMR CNRS 6638, ENSCI, 47 avenue A. Thomas, F-87065 Limoges
** Centre de Transfert de Technologies Céramiques, Ester Technopole, BP 6915, F-87069 Limoges

SUMMARY

Numeric fabrication technologies used in rapid prototyping have many advantages in the area of forming of ceramics. In particular, they allow prototype parts or small series of parts to be produced without the need for the design or fabrication of tooling, such as moulds or dies, normally required for production. Stereolithography, a well known technique in polymer science, may be used for the production of ceramic parts by including a UV reactive material in the ceramic suspension during processing. Sintered parts obtained by this process had properties similar to those of parts fabricated by conventional techniques.

1. INTRODUCTION

Most advanced ceramic processing techniques, such as pressing, extrusion or injection moulding, require the use of tooling (dies, moulds…) which are costly and difficult to justify for the production of a small number of parts. On the other hand, machining of sintered parts using diamond tooling, often necessary for the fabrication of complex shapes, is very expensive and is the origin of flaws in the material. In this light, numeric fabrication techniques through additive processes (layer upon layer), which have been developed in the last 15 years for the fabrication of polymer parts, offer an attractive alternative for the production of complex ceramic parts without using expensive tooling and eliminating or at least minimising the use of machining. Thus, research has focussed on adapting existing polymer processes and on developing new techniques that allow the direct production of near net shape parts with desired properties. Such techniques include selective laser sintering (SLS)1,2, 3 dimensional printing (3DP)3-5, fused deposition modelling (FDM)6-9 and laminated object manufacturing (LOM)10-15.

An attractive option is to adapt the stereolithography (SL) method, which is mainly used for the fabrication of three dimensional polymer parts, to the fabrication of 3D ceramic pieces with final properties (mechanical, thermal, electrical…) close to those obtained by conventional processing techniques16-19. In stereolithography, parts are constructed by the deposition of layer upon layer (typical thickness 25 to 100µm) of a ceramic paste containing a reactive system (monomer or photoinitiator). A CAD controlled UV laser beam then consolidates by polymerising or curing a programmed surface on each layer, a process that also ensures the integrity between successive layers. This process, which is completely innovative in the field of ceramics, allows the design and direct fabrication of final shape parts such as moulds, refractory moulds and cores, electronic components, parts for microsystems, and medical implants.

2. EXPERIMENTAL DETAILS

2.1. Raw Materials

Studies were carried out using a calcined alumina (CT1200SG, Alcoa, USA) of mean diameter 1.5µm and specific surface area 3.4 m2/g. An effective dispersant acting both by electrostatic and steric repulsion was used in order to increase the concentration of ceramic powder in the suspension. A high powder concentration (>60% by volume) is necessary in order to limit, and thus control, shrinkage during polymerisation and sintering. A thickener confers a high yield value to the paste to prevent settling of particles and to support the piece during fabrication.

The UV reactive system, consisting of a monomer (diacrylate – HDDA, UCB, Belgium) and a photoinitiator, absorbs in the range of the UV laser emission (Irgacure 651, Ciba, Switzerland).

2.2. Preparation of Suspensions

The photoinitiator (0.5 wt% of the monomer), the dispersant (2 wt% of the ceramic powder) and the thickener (0.5 wt% of the monomer) were first dissolved in the monomer before addition of the ceramic powder. Suspensions containing 62 vol% of alumina powder, were milled for 30 minutes to break down agglomerates and to achieve a good homogeneity.

In order to exploit stereolithography a specific rheological behaviour is required. Typical shear thinning behaviour is presented in figure 1. The high yield value (1200 Pa) inhibits the flow of non-cured zones during the construction of parts and allows them to be supported. The shear thinning behaviour allows deposition of homogeneous layers with thicknesses from 25 to 200µm. At a shear rate of 100 s-1, corresponding to the minimum value generated by the deposition device, the measured viscosity was 110 Pa.s, compared to 3500 Pa.s at rest.


Figure 1: Typical rheological behaviour of a stereolithography paste (62 vol% alumina)

2.3. Fabrication of Parts by Stereolithography

Unlike conventional stereolithography equipment used to deposit layers of liquid resin, the machine used in this study (Optoform, France) uses pastes or suspensions whose rheology (high viscosity and shear thinning behaviour) do not require the use of a container (figure 2).


Figure 2: Schematic diagram and a general view of the stereolithography machine

The ceramic paste is introduced by a piston, which delivers a controlled quantity onto the working area. Homogeneous layers, of different thicknesses (down to 25µm), and smooth surfaces were spread by means of a blade before polymerisation.

The laser (Ar+, Coherent, l = 351-364 nm), controlled by a CAD programme, polymerised zones of each layer of the part. When polymerisation of the first layer was performed, the process continued with deposition of another layer of paste on the polymerised layer. When fabrication and polymerisation of successive layers was complete, the part was cleaned and non-polymerised paste was removed using a solvent. At this stage a green part similar to that obtained after injection moulding was obtained.

2.4. Binder Removal and Sintering

A binder removal cycle, based on thermal analysis of a green sample, was used to eliminate all the organic phases (resin, photoinitiator, dispersant, thickener) before the densification step. After this step the parts were sintered at 1650°C for 2 hours.

2.5. Characterisation

The UV reactive system has to fulfil two main requirements, (i) the cured depth must be large enough to avoid an excessive fabrication time and (ii) the cured width must be small to ensure good dimensional resolution. Thus, cured depth and width were measured on small, polymerised lines cured in one thick (2 mm) layer by one scan of the laser beam. The brittle polymerised lines were included in an epoxy resin in order to cut sections for observation. The values of depth and width correspond to the average of four measurements.

The flexural strength of as-sintered and polished alumina bars (3.5 x 5.5 x 40 mm3) was measured by three-point bend tests (average of five values).

3 . RESULTS AND DISCUSSION

3.1. Depth and Width of Polymerisation

The main objective of reducing the fabrication time of parts was achieved by using a high scanning speed while maintaining a sufficient cured depth. The secondary objective, to obtain good dimensional resolution, was achievable by controlling the polymerization width.

The changes in cured depth and width, as a function of the energy density (DE) are given in figure 3. For a concentrated system (60 vol% alumina) a polymerised depth of 300µm can be achieved with a high scanning speed (1 ms-1) corresponding to an energy density of the order of 0.3 Jcm-2.


Figure 3: Variation in polymerised (a) depth and (b) width as a function of energy density (60 vol% Al2O3)

The polymerised depth (Ep) may be expressed as a function of the energy density (DE) transmitted to the paste, of the penetration depth of the beam (Dp) and of the critical energy density (DEc), which is the minimum DE value required to initiate polymerisation. It is exponentially related to DE as shown in equation 1:

(1)

DE is a function of the irradiating power (P), the scanning speed (v) and the diameter of the laser beam (w0 = 120µm) at the working surface:

(2)

The polymerised width is always larger than the diameter of the incident laser beam (120µm), demonstrating diffusion phenomena due to the presence of small ceramic particles. For example, for a depth of 300µm, DE=0.3 Jcm-2, the cured width (approx 600µm) is five times larger than the diameter of the laser beam. Conventional diffraction theories (Rayleigh, Gans, Mie) can be applied to dilute systems (<10 vol%) but are not applicable to more concentrated systems as occurs in this case.

A large polymerisation depth requires high energy densities while high resolution demands small energy densities. Thus, a compromise must be found, particularly for the scanning of zones on the surfaces of the parts.

3.2. Dimensional Resolution

DBidimensional surfaces, such as meshes, were obtained under different scanning conditions (power, speed) with a laser beam (diameter ~ 120µm) perfectly focussed on the surface. The resolution limit obtained was 230µm for an energy density of 0.05 Jcm-2 (figure 4).


Figure 4: Mesh type 2-dimensional structure, 120 layers of thickness 25 µm, DE=0.05 Jcm-2

3.3. Mechanical Strength (3-point flexure)

A mean flexural strength value of 394 MPa was measured on sintered and polished alumina bars prepared by stereolithography. Similar strength values were obtained for samples obtained by pressing the same alumina powder.

Thus, stereolithography allows the fabrication of parts with final properties similar to those obtained by conventional techniques while offering much greater possibilities in terms of achievable geometries and without the need for specific tooling.

4. CONCLUSIONS

The manufacture of ceramic parts by numerical fabrication technologies has been the subject of significant research, particularly in the USA, because of their advantages over conventional technologies, which require the design and production of expensive tooling. The feasibility of a number of these technologies has already been demonstrated: stereolithography, selective laser sintering, 3 dimensional printing, fused deposition modelling and laminated object manufacturing. Despite this, the technologies have found very few applications outside of the laboratory.

The stereolithography technique, used in the polymer field, has been adapted to the fabrication of complex shaped ceramic parts (figure 5) with final properties after sintering similar to those obtained by conventional processing. A formulation, concentrated with ceramic particles, reactive to UV radiation and having the correct rheological behaviour, has been developed. The correct choice of scanning parameters (power, speed…) controls dimensional tolerances to +/- 0.5% in parts with overall size of several centimetres.

Turbine – alumina Refractory core – zircon/silica
Aerator – alumina Gas injector – zirconia

Figure 5: Examples of parts produced by stereolithography

It is probable that the further development of numerical technologies and the reduction in price of the required equipment (lasers…) will progressively drive the integration of these technologies as fabrication processes for small to medium volume series of ceramic parts.

REFERENCES

  1. P.K. Subramanian, J.W. Barlow, H.L. Marcus, Proc. Of the SFF Symposium, 1995, p.346
  2. C. Griffin, S. McMillin, Proc. of the SFF Symposium, 1995, p.25
  3. M.J. Cima, J. Yoo, S. Khanuya, M. Rynerson, D. Nammour, B. Giriliglu, J. Grau, E.M. Sachs, Proc. Of the SFF Symposium, 1995, p.479
  4. J. Grau, J. Moon, S. Uhland, M. Cima, E. Sachs, Proc. Of the SFF Symposium (Ed.: D.L. Bourell, J.J. Beaman , R.H. Crawford, H.L. Marcus and J.W. Barlow), 1997, p.371
  5. M.Mott, J-H. Song, J.R.G. Evans, J. Am. Ceram. Soc., 1999, 82, 1653-1658
  6. A. Safari, S.C. Danforth, Proc of the 7th European Conference on Rapid Prototyping and Manufacturing, (Ed.: R. I. Campbell), 1998, p. 259
  7. R. Clancy, V. Jamalaba, P. Whalen, P. Bhargava, C. Dai, S. Rangarajan, S. Wu, S. Danforth, N. Langrana, A. Safari, Proc. Of the SFF Symposium (Ed.: D.L. Bourell, J.J. Beaman , R.H. Crawford, H.L. Marcus and J.W. Barlow), 1997, p.185
  8. C. Dai, G. Qi, S. Rangarajan, S. Wu, N. Langrana, A. Safari, S.C. Danforth Proc. Of the SFF Symposium (Ed.: D.L. Bourell, J.J. Beaman , R.H. Crawford, H.L. Marcus and J.W. Barlow), 1997, p.411
  9. E. Sachs, M. Cima, J. Cornie, Proc. Of the Second International Conference on Rapid Prototyping, 1991
  10. D. Klosterman, R. Chartoff, N. Osborne, G. Graves, A. Lightman, Proc. Of the SFF Symposium (Ed.: D.L. Bourell, J.J. Beaman , R.H. Crawford, H.L. Marcus and J.W. Barlow), 1997, p.247
  11. D. Klosterman, R. Chartoff, N. Osborne, G. Graves, Proc. Of the SFF Symposium (Ed.: D.L. Bourell, J.J. Beaman , R.H. Crawford, H.L. Marcus and J.W. Barlow), 1997, p.537
  12. M.J. Pope, M. Patterson, W. Zimbeck, M. Fehrenbacher, Proc. Of the SFF Symposium (Ed.: D.L. Bourell, J.J. Beaman, R.H. Crawford, H.L. Marcus and J.W. Barlow), 1997, p.529
  13. Z.E. Liu, T.C. Ku, J. Best, J.D. Cawley, A.H. Heuer, Proc. Of the SFF Symposium (Ed.: D.L. Bourell, J.J. Beaman , R.H. Crawford, H.L. Marcus and J.W. Barlow), 1997, p.379
  14. D. Klosterman, R. Chartoff, N. Osborne, G. Graves, A. Lightman, G.han, Proc. Of the 7th International Conference on Rapid Prototyping, 1997
  15. D.M. Kupp, H. Eifert, M. Greul, M. Künstner, Proc. Of the SFF Symposium (Ed.: D.L. Bourell, J.J. Beaman , R.H. Crawford, H.L. Marcus and J.W. Barlow), 1997, p.203
  16. C. Hinczewski, S. Corbel and T. Chartier, Journal of the European Ceramic Society, 18, 583-90 (1998).
  17. C. Hinczewski, S. Corbel and T. Chartier, Journal of Rapid Prototyping., 4 [3], 104-11 (1998).
  18. M.L. Griffith and J.W. Halloran, Journal of the American Ceramic Society, 79 [10], 2601-08 (1996).
  19. H.Liao and T.W. Coyle, Journal of the Canadian Ceramic Society, 65 [4], 254-262 (1996).
     
Issues  |  Table of Content