View/Open - Scholarly Commons Home

4.1 Experimental conditions
Chapter 4
Experimental evaluation of CLFDM
The main aim of this research is to find out the characteristics and behaviour of the
parts printed with the FDM test bed, mainly when they are printed in two styles of
deposition: flat and curved layer. Two parts (model 1 and model 2) are chosen and
support structures as required for each model is printed and kept ready for both the
models. The support structures can be printed with either silicon or fabepoxy. Silicon
was chosen as the material for support structures in all experiments as this is the
least expensive material available to be used with the current FDM test bed. Another
major advantage of using silicon is that it can be removed easily from the main part
without much difficulty and any adverse effects on the structure of the main part.
After the support structures are ready, the main parts are printed on top with varying
conditions of deposition style and speed. Once all the parts are printed, they are then
cured for nearly two days and then tested for their ultimate stress for varying
conditions. For the second part structure, model 2, a curing time of nearly two weeks
is given before they are tested in the Hounsfield Tensile testing equipment. This test
will give clear representation of strength qualities of parts when printed in flat and
curved layer style of deposition. Testing of model 2 in the same process will also
give results for strength comparison as well as tell if curing time has any effect on the
strength of the part.
The number of experiments is statistically planned by Taguchi L8 design. Three input
parameters namely curvature, speed and deposition style are selected for the
investigation. Curvature is represented by the two models 1 and 2 which are used for
this purpose. Speed of deposition can be changed from the tool information file, a
part of the Fab@home software. Two speeds of deposition are chosen and the
deposition style is the most important parameter in these experiments as it is either
Flat or curved style of deposition. A brief review of the need for statistical design of
85